Preface

Update on clinical veterinary behavior

Guest Editors

The goal of this issue of the Veterinary Clinics of North America: Small

Animal Practice is to make available the latest techniques and medications
in the diagnosis and treatment of small animal behavior problems. In the 6
years since the last behavior issue of this Clinic (May 1997), the American Col-
lege of Veterinary Behaviorists has grown to 32 members. All of the North
American authors who contributed to this issue are board certiﬁed; in addi-
tion, the issue includes articles by European behaviorists whose approach
to behavior problems and their solutions should provide fresh insight.

Pageat presents an in-depth review of feline and canine pheromones, the

vomeronasal organ that may detect them, and the uses of pheromones in
treatment of separation anxiety, as well as feline spraying and aggression.
Appleby and Pluijmakers consider predisposing factors, behavioral re-
sponses, and management considerations with separation anxiety in dogs.
Reisner classiﬁes canine aggression in a rational manner and oﬀers the latest
treatment options. Feline aggression is much more diﬃcult to classify, but
Frank and Dehasse oﬀer a new diagnostic approach and, therefore, a more
rational choice of treatments; the use of selegeline is especially intriguing.

Nielson addresses elimination behaviors in cats by oﬀering methods for

eliciting the proper information from the owner, for creating a conducive
and stress-free elimination area, and providing a review of recent research
on pharmacological treatments for spraying. In contrast, Houpt and Zicker
review older literature on the eﬀects of diet on behavior and point out when
diet is and is not important in treating behavior. One area in which diet is
important is in the older animal, in which the use of antioxidants for slowing
or reversing age-related cognitive decline is a surprisingly powerful tool.

Vet Clin Small Anim 33

(2003) 185–186

Katherine A. Houpt, VMD, PhD

Vint Virga, DVM

0195-5616/03/$ - see front matter

PII: S 0 1 9 5 - 5 6 1 6 ( 0 2 ) 0 0 1 3 4 - 1

An area of canine behavior that is growing in importance since Sep-

tember 11, 2001, relates to the use of substance-detecting dogs and other
classes of working dogs. Burghardt considers the assessment of aptitude and
performance, problems that may arise with performance of duties, and
behavioral therapy in working dogs. His evidence suggests that methyl-
phenidate may be useful in the treatment of some of the stereotypic behav-
iors that arise in these valuable animals.

Whether or not animals obsess remains controversial and diﬃcult to

prove, but Luescher contributes a diagnostic framework for and describes
many types of compulsive behaviors, including oral locomotor repetitive
nonfunctional activities. It is of paramount importance that medical causes
are ruled out before making a behavioral diagnosis. Overall presents
medical conditions that may cause behavioral manifestations by consider-
ing traditional diagnostic categories: developmental, acquired, metabolic,
neurological, infectious, and toxic. Virga establishes a framework for the
classiﬁcation of dermatologic conditions with behavioral causes, diagnostic
criteria for their clinical diﬀerentiation, and an integrative approach to their
assessment and management in canine and feline patients.

No issue devoted to behavior would be complete without a discussion of

the available pharmacological treatments. Simpson and Papich thoroughly
consider neurotransmitter functions in behavioral pharmacology and pro-
vide a rational and structured framework for the use of psychotropic drugs.
The area that oﬀers most promise for the future is behavioral genetics.
Takeuchi and Houpt review what is currently known about canine and feline
behavioral genetics and oﬀer the promise of blood tests for some behav-
ioral problems similar to those now used for progressive retinal atrophy.

This issue of the Veterinary Clinics of North America: Small Animal Prac-

tice represents the work not only of the authors but also John Vassallo, Editor,
and the editorial staﬀ at WB Saunders. By bringing together the thoughts and
eﬀorts of all these individuals, we hope to provide a worthwhile compilation
reviewing many of the most recent advances in small animal clinical behavior.

Katherine A. Houpt, VMD, PhD

Animal Behavior Clinic

College of Veterinary Medicine, Box 31

Cornell University

Ithaca, NY 14853-6401, USA

E-mail address: kah2@cornell.edu

Vint Virga, DVM

Behavioral Medicine for Animals

Veterinary Healing Arts, Inc.

PO Box 219

Newport, RI 02840-0219, USA

E-mail address: vvirga@worldnet.att.net

186

K.A. Houpt, V. Virga / Vet Clin Small Anim 33 (2003) 185–186

Current research in canine and

feline pheromones

Patrick Pageat, DVM, PhD*,

Emmanuel Gaultier, DVM, MA

PHEROSYNTHESE s.n.c., Le Rieu Neuf, 84490 Saint Saturnin les Apt, France

Everyone who has observed dogs or cats, if only once, has been impressed

by how important olfactory communication is in these species. Carnivora
are known as the order that has developed the greatest variety of glands
secreting chemical signals. Among all the molecules secreted by these glands,
some seem to transmit highly speciﬁc information between animals of the
same species—the pheromones. In 1959, Karlson, Luscher, and Butenand
created this word by combining the Greek verbs pherein (to carry) and
horman

(to stimulate). Initially, this kind of chemical communication was

supposed to exist only in invertebrates. At that time, more attention was
paid to the pheromones of insects, and a few products to ﬁght against some
pests have been developed. Many ethologists thought that this really strict
and biologic way to communicate should not be considered in mammals
because of the complexity and plasticity of their social behaviors [1].

The description of the ‘‘ram eﬀect,’’ which is the activation of ovulation

in the ewes by the ram’s skin secretions, has shown that pheromones could
exist in mammals [2]. The ﬁrst pheromone to be analyzed was the boar’s
pheromone, which is produced by the submaxillary glands. The main com-
ponent of this pheromone is 5 a-androsterone, a steroid with 19 car-
bons, which has a urine odor [3]. This pheromone induces an immobility
response in the sow when she is in estrus. Synthetic analogues of this
pheromone have been developed to detect estrus in sows before artiﬁcial
insemination.

For the last 10 years, the functions of some pheromones in dogs and

cats have been elucidated, and synthetic analogues of some of these are
beginning to be used as a therapeutic approach in behavioral medicine [4,5].

Vet Clin Small Anim

33 (2003) 187–211

* Corresponding author.
E-mail address:

pherosynthese@wanadoo.fr (P. Pageat).

0195-5616/03/$ - see front matter

PII: S 0 1 9 5 - 5 6 1 6 ( 0 2 ) 0 0 1 2 8 - 6

Marking behaviors seem to be more and more interesting parameters for the
evaluation of the behavior of our patients.

It seems to be important for veterinarians to be able to identify scent

marking during the behavioral examination and to take care of the activity
of the glands that produce these secretions.

Olfactory communication in carnivores in general, particularly in dogs,

has been the subject of a number of studies over the years. Among the
substances produced, pheromones have an important place. The complexity
and great number of substances produced led many authors to consider
the chemical analysis of these messages to be impossible or at least
unpredictable.

Many recent studies as well as the development of synthetic analogues

of pheromones have improved our knowledge. An important point has been
to determine the diﬀerence between pheromones and odors. Pheromones
are not simple smells. The perception of the odors is a spontaneous
phenomenon. During respiration, part of the inhaled air (up to 30% in some
species [4]) is deviated to the olfactory mucus. In that way, animals perceive
the odors when breathing. The perception of the pheromones does not work
in the same way. The vomeronasal organ (VNO) is not easily accessible
during normal respiration. To be stimulated, it has to be opened, and the
pheromones can then go to receptors on the membranes of the nervous cells
of this organ. The pheromones may have particular olfactory character-
istics, but they do not act only as an olfactory stimulus. The odor of the
pheromones can be a stimulus that induces the opening of the VNO.

Perception of pheromones

The perception of pheromones is not completely understood. The best-

known hypothetic is that there is a stimulation of the VNO. The VNO is a
part of the accessory olfactory tract (Fig. 1). There are two VNOs situated
on each side of the nasal septum in a small fossa. Each VNO is about 4 cm
long in the dog (Figs. 1 and 2) [7,8]. Three nerves innervate the VNO. The
nasopalatine nerve (part of the trigeminal nerve), which includes ﬁbers of
the parasympathetic and sympathic nervous system, may control both the
vascular activity and secretion of mucus by the glands. There is also a
functional relation between the nasopalatine nerve and some mechano-
receptors included in the nonnervous part of the VNO. The vomeronasal
nerve is dedicated to the transmission of stimulation initiated by the
pheromone. This nerve is connected to the accessory olfactory bulb and then
to the amygdala through the limbic system (in the medial nucleus and the
cortical posteromedial nucleus). In contrast to the main olfactory tract,
there is no connection between the VNO and the neocortex even through the
thalamus [9]. The connections between the sensory cells and the glomeruli
in the accessory olfactory bulb and in the main olfactory bulb are abso-
lutely diﬀerent. In the main olfactory system, there is a rich variety of odor

188

P. Pageat, E. Gaultier / Vet Clin Small Anim 33 (2003) 187–211

receptors. Each kind of sensory cell (meaning cells carrying the receptors for
one speciﬁc odor) is connected with only one glomerulus; it enhances the
precision and sensitivity of detection. In the accessory olfactory bulb, each
sensory cell is connected with several glomeruli, which, in that manner,
receive the information from cells carrying several types of receptors,
enabling a complex coding with a limited variety of receptors [10,11].

Fig. 1. Photographic montage of the dissection of the vomeronasal organ in a dog. Right
canine and a part of the incisive bone are removed. 1

¼ nasal septum; 2 ¼ body of the

vomeronasal organ (outlined); 3

¼ vomeronasal duct; 4 ¼ incisive duct (needle).

Fig. 2. Vomeronasal organ (with cartilage) of a dog. X

¼ cranial extremity.

189

P. Pageat, E. Gaultier / Vet Clin Small Anim 33 (2003) 187–211

The third nerve, the nervus terminalis, which has no precisely known

function, travels from the nose to the brain and includes gonadotropin-
releasing hormone LHRH cells.

The VNO is wrapped up in the vomeronasal cartilage (Figs. 2 and 3) to

form a tube closed at its caudal end. There are many elastic ﬁbers and
smooth muscular ﬁbers that show motor activity during the suction of the
pheromones [11–14]. The lumen of this organ is surrounded by two kinds of
epithelium: the medial one, which is thick and includes nervous cells, and the
lateral one, which is thin and consists of a respiratory mucous membrane
(Figs. 3 and 4). The axons of the nervous cells of the medial epithelium
merge together to create the vomeronasal nerve. On the lateral side are

Fig. 3. Cross section through the nose in a kitten (original magniﬁcation

25). 1 ¼ respiratory

mucosa; 2

¼ conjunctive chorion; 3 ¼ nasal septum bone; 4 ¼ vomer bone; 5 ¼ vomeronasal

cartilage (U-shaped conformation); 6

¼ medial (receptor) epithelium; 7 ¼ crescent-shaped

lumen of the vomeronasal organ; 8

¼ lateral (respiratory) epithelium; 9 ¼ vomeronasal glands.

190

P. Pageat, E. Gaultier / Vet Clin Small Anim 33 (2003) 187–211

many blood vessels and mucus-producing glands that are important during
the suction of the pheromones. The components of the mucus are those
usually observed in respiratory mucus and, in addition, some speciﬁc
lipocaline proteins that may be pheromone-binding proteins (PBPs). These
proteins, weighing around 20 kd, show a high aﬃnity for hydrophobic
molecules like fatty acids [15–17].

In many species of mammals, the suction of the pheromones by the VNO

follows a behavior described as ﬂehmen. This behavior is speciﬁc to some
mammals, including the cat. Flehmen consists of raising the upper lip with
the mouth half-open and, in cats and dogs, movements of the tongue. The
movement of the upper lip is achieved by contraction of the muscle levator

Fig. 4. Cross section of the vomeronasal organ in a kitten (original magniﬁcation

400).

Details of Fig. 3. 1

¼ receptor epithelium; 2 ¼ microvilli; 3 ¼ sensorial cell; 4 ¼ capillary

blood vessel; 5

¼ fascicle of the vomeronasal nerve; 6 ¼ vomeronasal cartilage.

191

P. Pageat, E. Gaultier / Vet Clin Small Anim 33 (2003) 187–211

labii maxillaris [12,13]. During this contraction, the accessory head of this
muscle, which attaches on the incisive papilla, opens the incisive ducts that
communicate with the vomeronasal duct and then the VNO. In the dog, the
occurrence of ﬂehmen remains controversial [3]. We do not observe any
real ﬂehmen in this species, but the analogous behavior may be tonguing, in
which the dog pants, raises the upper lip, creases the nose, and rapidly ﬂicks
the tongue against the incisive papilla during exploration of feces, urine, or
proestral blood. The role of ﬂehmen has sometimes been limited to detection
of sexual pheromones, but it is also known that castrated subjects carry it
out and that this behavior appears in circumstances having nothing to do
with sexual activity [3–5]. This is most noticeable in cats when a subject is
located close to a urine mark or even in close proximity to a facial mark
[6,18].

The mechanism of ﬂehmen induces aspiration of the pheromone into the

VNO, where it is mixed with the mucus. This suction occurs because of vaso-
constriction in the wall of the VNO. This increases the diameter of the lumen of
the organ and thus creates the fall in pressure necessary for the suction. The
components of the pheromone are mixed with the mucus, which consists
primarily of hydrophobic molecules. The components of the pheromone bind
with the pheromone binding proteins (PBPs) and so can stimulate the
receptors located on the membrane of the sensorial cells [5,14,16].

After the suction period, the washout of the VNO begins. The blood

pressure increases, and the diameter of the VNO’s lumen decreases,
expelling the mucus, including the PBP-pheromone complexes. Because of
the opening of the VNO into the incisive channel and then into the mouth,
the expelled mucus is not visible in carnivores. Conversely, this expulsion is
easily observed in the stallion just after the ﬂehmen to the urine of a mare
in estrus.

What the response of the brain is after the stimulation of the VNO’s

receptors by the pheromones is not really known. A number of hypothesis
related to sexual pheromones have been proposed. In the doe and the ewe,
the pheromones produced by the postcornual gland (sebaceous glands from
the skin between the horns in Artiodactyles) of males may induce a peak of
LH in females that is responsible for the ovulation [3]. The action of the
other types of pheromones has not been demonstrated, but the use of new
techniques like positronic cameras and radiolabeled pheromones should be
of great help to explore these phenomena.

Even if we do not precisely know the whole mechanism, it is possible to

summarize the action of the pheromones by dividing these biologic
mediators into two groups [19]:

Primers: pheromones that induce some major modiﬁcations of the

physiology in the receiving individual

Releasers: pheromones that induce an immediate modiﬁcation of the

behavior of the receiving individual

192

P. Pageat, E. Gaultier / Vet Clin Small Anim 33 (2003) 187–211

Structures secreting pheromones: the main known pheromones

Carnivora are identiﬁed as the mammalian species that have the most

developed and varied types of pheromone-secreting glands. Diﬀerent types
of glands present in the skin and in certain mucous membranes are involved
in producing pheromones. The histologic structure of these glands does not
reﬂect a functional speciﬁcity, except for the anal sacs, which are often
mentioned and seem to be the most typical pheromone-secreting structures
in carnivores [3,6,20–22].

If we try to explore the pheromone-secreting glands from the nose to the

tail of dogs and cats, we encounter six major sources of pheromones.

The facial area

The area of the cheek and perioral glands brings together a whole set of

secreting structures spread throughout the chin, lips, vibrissae, and cheeks.
These glands exist in both dogs and cats. The dog has one more gland,
however, the ear gland, which consists of some ceruminous glands of the ear
duct as well as some sebaceous glands from the external ear.

In the cat, ﬁve diﬀerent facial pheromones named F1 to F5 have been

isolated from the sebaceous secretions of the cheeks (Table 1). At the present
time, we know the function of only three of them (F2, F3, and F4). These
pheromones are involved together in territorial marking behavior (in its
widest meaning) in cats and in complex social exchanges that can easily be
observed.

The cat seems to mark some points around his preferred pathways in his

territory by rubbing his face on them (Fig. 5). In so doing, he deposits the
pheromone F3, which may appease him and helps in organizing the

Table 1
Chemical components of the facial secretions in the cat

Secretion

Components

Function

Oleic acid, caproic acid, trimethylamine

5-aminovaleric acid, n-butyric acid,
a-methylbutyric acid

Unknown

Oleic acid, palmitic acid, propionic acid,

p-hydroxyphenylacetic acid

Sexual facial marking in tomcats

Oleic acid, azelaic acid, pimelic acid,

palmitic acid

Facial marking on items, antagonist

of urine marking and scratching

5b-cholestan acid 3b-ol, oleic acid,

pimelic acid, n-butyric acid

Allomarking, antagonist of

territorial or irritative aggression

Palmitic acid, isobutyric acid,

5-aminovaleric acid, n-butyric acid,
a-methylbutyric acid, trimethylamine,
azelaic acid, p-hydroxyphenylacetic acid

Unknown

Data from

MacDonald DW. The carnivores: order Carnivora. In: Brown RE, MacDonald

DW, editors. Social odours in mammals. Oxford: Clarendon Press; 1985. p. 619–22.

193

P. Pageat, E. Gaultier / Vet Clin Small Anim 33 (2003) 187–211

environment by classifying it into ‘‘known objects’’ and ‘‘unknown objects’’
[3–5]. During sexual behavior, when detecting and attracting females in
estrus, the male cat rubs his face on some points around the place where he
is with the female cat and deposits the F2 pheromone. This pheromone
seems to improve the eﬃcacy of the sexual display [4,5,18]. In tomcats, the
development of the cheek glands is dramatic, and we observe an extreme
reduction of this gland in neutered male cats [23]. The social function of
another facial pheromone is also known in the cat—the F4 pheromone,
called the allomarking pheromone. The allomarking behavior is observed
between cats who live together or between a cat and a dog or a cat and a
human being when the cat is socialized with these other species. This
pheromone decreases the probability of aggressive behavior between the cat
and a marked individual [4,5,24–26]. In this species, it has been proven
clearly that the sebaceous glands synthesize a protein that is known to be the
main feline allergen for human beings—Fel-d-1 [23,27]. This protein is a
dimer whose structure is similar to PBP, and it has a high aﬃnity for
hydrophobic small molecules like the fatty acids that are components of the
pheromones [15,17,23,27]. This protein could be secreted to bind and

Fig. 5. Facial rubbing of a cat. This part of the ethogram of the cat is well known to all cat
owners. The cat rubs its head against an object from the side of the chin to the base of the ear.
Cats of both sexes show this behavior, and frequency of rubbing depends on the individual. The
facial secretions combined with urinary secretions inform male cats about the sexual receptivity
of female cats, but rubbing is not a behavior implicated only in reproduction. Rubbing seems to
have a visual communication function associated with the placing of facial pheromones. In fact,
cats seem to perform this marking behavior when a known individual approaches. Facial
secretions are divided into ﬁve fractions, with each of them having a speciﬁc action. The F4
fraction of facial pheromones has a relational function—allomarking. The F3 fraction of facial
pheromones represents spatial orientation and emotional stabilization functions. These marks
seem to be involved in the cat’s geographic orientation; they are placed on objects that form the
boundary between a passage and a territorial zone.

194

P. Pageat, E. Gaultier / Vet Clin Small Anim 33 (2003) 187–211

protect the pheromones and so enhance their probability of being detected.
According to some experimental data, Fel-d-1 seems to be a persistent sub-
stance, because it is possible to ﬁnd the protein even 3 weeks after a cat has
marked a place [21,27].

In the dog, the facial complex seems to be involved, especially in social

relationships (Fig. 6) [3,20,26,48]. Some information has recently been
acquired about the chemical structure of these secretions. The secretions of
the ear gland seem to be related to the social status of the dog. Some
secretions seem to be secreted only by dogs that show assertive behaviors (eg,
eating before the others, taking other dogs’ food because they show a
submissive attitude, exhibiting sexual behaviors in front of others). Just after
submission, a dog generally approaches the winner of the ﬁght and smells the
opening of his ear with small movements of the tongue and the lips [18,26].

The pedal complex

This area consists of the pedal glands of the four legs. These are diﬀuse

structures present both in the plantar pads and in the skin of the interdigital
region. The presence of glands in the plantar pad is not as clear in the dog as
in the cat. The cat shows many sweat glands in the plantar pads, which
secrete the sweat emitted, for example, during fear reactions [3,6,18]. The
histologic structure seems to be really diﬀerent in the dog, and the pedal area
is perhaps limited to the interdigital area.

In both cats and dogs, this complex is involved in marking territory and

in producing alarm pheromones [3,4,6,18,20,26,28]. During territory mark-
ing, the pedal secretions are associated with a scratching behavior, which
creates a persistent visual signal [3,4,6,18,26,28]. In the cat, the scratching

Fig. 6. Main pheromone-secreting glands in the dog. 1

¼ labial glands; 2 ¼ auricular glands;

¼ perianal glands; 4 ¼ vulvar or preputial glands; 5 ¼ interdigitous glands.

195

P. Pageat, E. Gaultier / Vet Clin Small Anim 33 (2003) 187–211

marks are generally made on vertical places, and this behavior is considered
as unacceptable by many owners (Fig. 7) [3,18]. In the dog, the scratches
are on the ground and are commonly remarked with urine by the male
dogs [18,28].

Some alarm marks seem to be emitted by the pedal center. Even if there

are no precise data about these secretions, it is easy to observe that the sweat
secreted by fearful cats enhances avoidance behaviors in the cats who
encounter these marks. It is possible that this kind of pheromone is involved
in the fear elicited by the veterinary clinic in dogs and cats. The alarm
pheromones secreted by the previous animals could create the fear reactions
that are commonly observed by the veterinarian [18].

Fig. 7. A cat is scratching a vertical object. This object is covered by a special resin containing a
synthetic pheromone that enables scratching. Scratching usually occurs near a place associated
with a precise activity (eg, litter, bowl) and in the pathway of a human being or other animal.

196

P. Pageat, E. Gaultier / Vet Clin Small Anim 33 (2003) 187–211

The perianal complex

This area consists of the supracaudal glands, the circumanal glands, and

the anal sacs. The supracaudal glands are well developed in the cat. They
consist of a high concentration of sebaceous glands located at the dorsal
aspect of the root of the tail [3], sometimes extending distally on both sides of
the tail. Each gland is visible, with the size varying between 0.85 mm

0.5 mm

in female cats and 1.84 mm

1 mm in tomcats. In each organ, we may iden-

tify tubular apocrine glands that lie between large sebaceous glands com-
prising cistern-like cavities running into ducts [3]. This structure is well
known to veterinarians because of the inﬂammations that can occur,
particularly in tomcats, but we do not precisely know their function, except
that it seems to be related to the identiﬁcation of the male by the female in
estrus. Neutered male cats show reduced supracaudal glands.

In the dog, the supracaudal glands are less developed. In the hunting

wild dog Lycaon pictus and in the domestic dog, these glands seem to be
undeveloped in some individuals [3]. These glands are associated with the
sebaceous disorders that can occur in some dogs. These glands generally
form an elliptic patch of large sebaceous glands on the dorsal surface of the
tail 5 to 40 mm below the base and are more developed in male dogs than in
female dogs. In male red foxes, it has been reported that there was an increase
in the sebaceous activity in this gland during spermatogenesis [29]. Extra-
polating from that, we can suppose that this secretion could be involved in
the stimulation of the bitch during sexual display. Conversely, the supra-
caudal glands seem to be active throughout the year and thus could be
involved not only in sexual behavior but in social communication (see Fig. 6).

The circumanal glands include the sebaceous and modiﬁed sweat glands

that are disseminated all around the anus. These glands seem to be more
developed in the dog than in the cat. The circumanal glands of male dogs are
the more developed, and their size increases with age. In old male dogs, the
diameter of the circumanal area is 3 to 4 cm and there are more glands on
the upper part of it. These secretions seem to be important for the social life
of the dogs, and it seems that the special color of the hairs around this area
plays an important role in improving the eﬃcacy of the chemical signal (see
Fig. 6). During estrus, these glands produce trimethylamine-rich secretions
in bitches [30].

In cats, these glands are well developed only on the skin below the tail.

As has been described in many Felidae (eg, lions), the secretions of the
circumanal glands of cats diﬀer from the canids’ secretions in that they lack
trimethylamine [1].

The anal sacs are found in both dogs and cats, and their physiology is

really important in understanding some critical points about chemical
communication.

Dogs show two symmetric sacs that open by a small ostium at the limit

between the anus and skin. The wall of this sac is stratiﬁed and keratinized

197

P. Pageat, E. Gaultier / Vet Clin Small Anim 33 (2003) 187–211

epithelium with several apocrine glands. These particular glands are
modiﬁed sweat glands and sebaceous glands. Some of them excrete their
secretions into the sac and some into the duct of the sac, which is 4 to 5 mm
wide [3,6,31]. The secretions change in color from cream to brown and often
contain large ﬂocculent globules. Anal sacs are fermentation chambers
where aerobic and anaerobic bacteria metabolize the original secretions of
the glands to produce aliphatic acids and amines (eg, putrescine, cadaverine,
methylamine trimethylamine). These sacs are surrounded by smooth
muscular ﬁbers that make possible the strong expulsion of these secretions
in marking behavior and especially in alarm marking during fear-induced
reactions; spraying of these secretions over 0.5 to 1.2 m has been observed by
some authors [31]. In bitches, the composition of the secretions varies during
ovarian activity. During estrus, these secretions seem to be highly attractive
to male dogs, and their concentration of C2 through C5 aliphatic acids,
acetone, and trimethylamine increases [3,30]. Variations of the bacterial ﬂora
are common in the anal sacs [30–32]. The consequence is a modiﬁcation of
the composition of the chemical signals, which may result in behavioral
problems in groups of dogs, such as aggression toward the dog with an anal
sac infection. This point is critical for the veterinary behaviorist, and it is
absolutely necessary to examine the anal sacs of dogs when there are some
social behavioral problems between dogs that live together [18].

In the cat, the same anal sacs exist but their opening is into the rectum;

thus, their secretions participate in the communication in the feces. The
walls of their anal sacs contain more sebaceous glands than in dogs, and,
correspondingly, they produce lipid-rich secretions [3]. These glands secrete
large amounts of Fel-d-1 protein, which could be important in binding the
pheromones so as to maintain them in the environment [33].

The genital complex

This area includes sebaceous glands of the prepuce or the vulva and

urethral or genital mucous glands together.

In the dog, this complex is intensely explored during each social contact

(see Fig. 6) [34]. These secretions participate in both social and sexual be-
haviors. In bitches, during estrus, a secretion of methyldihydroxybenzoate
seems to be highly attractive to male dogs and enhances sexual excitement
[3,18,20,35]. It is interesting to note that this component is commonly used
as a preservative in many human cosmetic products. It could be involved in
some sexual behavior shown by some dogs toward their owners.

In the cat, these glands are not so well studied and there is a lot to be

learned about their functions.

The mammary complex

This complex has been discovered in recent years. Some authors

described olfactory interactions between mothers and babies in primates

198

P. Pageat, E. Gaultier / Vet Clin Small Anim 33 (2003) 187–211

and especially in human beings, but there was no proof that there was any
pheromone secreted in this area. The ﬁrst pheromone isolated in this area
was in the sow [36]. Shortly thereafter, we isolated the same kind of
pheromones in bitches, mares, cows, ewes, queens, and does. Because these
pheromones seem to have an appeasing action on the babies and adults, we
proposed calling them appeasines. These pheromones are secreted by the
sebaceous glands of the sulcus between the two mammary chains [18,26].

The appeasines of the bitch or the queen have the same chemical

structure as those of the other species; three fatty acids could be considered
as the ‘‘mammal appeasing message’’: oleic acid, palmitic acid, and linoleic
acid. They are always associated in the same ratio. The other components
could be considered as the species-speciﬁc message, which always begins
with myristic acid (in a varying ratio). The speciﬁc message of the bitch is
in the following order: myristic acid, lauric acid, pentadecanoic acid,
and stearic acid [36]. In the queen, the speciﬁc message is similar but the
pentadecanoic acid is lacking and thus the ratios of each component are
diﬀerent. As in the anal sacs, the association between the secreting glands
and the saprophytic bacteria is critical. The glands produce fatty acids that
are not too volatile. It seems that the temperature of the skin rises as the
result of an increase in blood circulation but that it is not suﬃcient to
vaporize the acids. The chromatographic analysis of the sebum shows that
there is a balance between acid and methyl-esters. Both have the same
eﬃcacy according to the pheromonal message, but they do not have the
same volatility [36]. The methyl-esters are much more volatile; all of them
are liquid at 20

°C when many acids are solid.

These pheromones are not secreted immediately after the birth of the

litter. The secretion appears 3 to 4 days after parturition and persists 2 to 5
days after the weaning of the puppies or kittens (4 months of age for puppies
and 6–12 weeks for kittens) [18,36].

Urine and feces

The importance of urine and fecal marking is well known in both dogs

and cats. Both are a complex source of pheromones. In both urine and feces,
the chemical components of the pheromones are produced by both glands,
which emit their secretions in the lumen of the urinary tract or in the anal
duct, and also by saprophytic bacteria, which metabolize some components
of urine or feces [3,20,26,29].

Urine marking is certainly the most well-known marking behavior in the

cat. It is considered the main behavioral problem of cat owners. The posture
of the cat during urine marking is typical (Fig. 8). The cat stands up and
sprays small amounts of urine on vertical surfaces. The spots of urine on the
marked surface and the attitude of the cat are visual signals specifying that
there has been an emission of pheromones. It is easily visible, because the
cat that detects such signals exhibits a ﬂehmen [18]. Several situations can

199

P. Pageat, E. Gaultier / Vet Clin Small Anim 33 (2003) 187–211

enhance urine marking in cats. Spraying as sexual activity is well known in
tomcats but also exists in female cats during estrus. Some components of the
sexual pheromones in the tomcat are aromatic and are provided by the
seminal secretions [18,21,37]. After neutering, these aromatic components
disappear and the occurrence of this kind of urine marking decreases. Urine
marking is also used in situations of territorial modiﬁcations (arrival of an
unknown individual or modiﬁcation of the physical environment [eg,
introduction of a new piece of furniture]). Facial marking with F3 is an
antagonist of this urine marking [4,25]. Fecal marking is less common in
cats than in dogs. The signiﬁcance of this marking is not well known, and
it seems that this behavior could be related to alarm marking [18].

In dogs, urine marking is also recognized easily because it is a visible

marking. This behavior can be observed in both female and male dogs even
if it is more frequent in male dogs (Fig. 9). The movements of the rear limb
are pronounced in dominant male dogs when they mark in front of
conspeciﬁcs [3,18,26,38]. In the absence of conspeciﬁcs and especially when
there is no challenger, this marking is really rare. Urine marking can be
associated with scratching the ground, especially when there are female dogs
in estrus or if there is a situation of severe hierarchic challenge [18,26,28].
In the situation of hierarchic disorders between dogs and owners, urine

Fig. 8. Urinary marking in the cat is characterized by distinct behavioral patterns. After
selecting a vertical support, the cat sniﬀs it, kneads the ground and turns its back to the support.
With the tail held vertically, the cat sprays a horizontal jet of urine. The two most common
types of urine marking are sexual and reactional urine marking. Sexual urine marking occurs by
male cats when they are sexually excited (by the presence of a female in heat). Urine marks are
left on exits to the outside (doors and windows). Female marking is related to the onset of
estrus. Sexual urine marking often comes together with vocalizations. Reactional urine marking
is often caused by a change in the cat’s environment (eg, new piece of furniture, moving to a new
house, a visitor or new occupants).

200

P. Pageat, E. Gaultier / Vet Clin Small Anim 33 (2003) 187–211

marking increases and sometimes becomes the main complaint of the
owners [18]. Fecal marking is also well developed in the domestic dog as in
many Canids (eg, fox, coyote) and Mustelidae (eg, badger). The feces are
laid on high surfaces like stones, stumps, or furniture. Generally, only a
small amount of feces is associated with the secretions of the anal sacs and
glands of the anal duct. This marking is common in females too, especially
during metestrus [18,26].

Because we understand the mechanisms that control the emission of

pheromones better and better, it is possible to use them as therapeutic
measures.

Pheromonotherapy

We proposed this word to describe the use of pheromones to treat

behavioral disorders [5]. Because our pets are living in a world full of
smells, this particular clinical approach seems interesting and is considered
acceptable and pleasant by owners. The limits of this approach are related to
the precision of this communication. The right pheromone has to be chosen
and emitted at the right time and on the right place so as to obtain the
expected results. It means that the propedeutic approach to behavioral
disorders also has to be precise.

The precise mechanism of action of most pheromones is still unknown,

but they induce some modiﬁcations in both the limbic system and the

Fig. 9. Dog’s posture of urinary marking. This behavioral pattern involves a complex
association of messages. Visual cues inform others dogs in a wide range of the presence of
olfactory information in this precise area. The urine-marking behavior of the dog can be
completed by laying of podal pheromones on the ground by scratching the ground with its legs.
This behavior is signiﬁcantly more frequent in male dogs [20,28]. Performance of this marking
seems to bear a relation to certain situations of competition between male dogs as well as to the
intrusion of individuals not belonging to the pack [20,26,28].

201

P. Pageat, E. Gaultier / Vet Clin Small Anim 33 (2003) 187–211

Fig. 10. Usefulness of F3 in the treatment of urine marking in the cat [5]. The studied
population (1–6.5 years old) is composed of 30 male (29 neutered) and 31 female (22 spayed)
cats. These cats are treated for reactional and sexual type urine marking that had been evolving
for between 1.5 and 3 months. The treatment consists spraying F3 once a day on the urine
marks and on salient objects and locations in the house. This treatment is continued for 28 days,
and the cats are followed up to day 49 so as to monitor any relapse.

Fig. 11. Usefulness of F3 to decrease excessive scratching. Fifty-three cats are included in this
trial. F3 is sprayed daily for 28 days on every scratch mark. All cats are followed up for 7 weeks
so as to monitor any relapse.

202

P. Pageat, E. Gaultier / Vet Clin Small Anim 33 (2003) 187–211

hypothalamus. In that way, the emotional status and the way the animal
reacts are altered during the behavioral modiﬁcation program. According to
the type of pheromone prescribed, behavioral patterns can be induced (eg,
facial marking in the cat, scanning the environment) or can inhibit some
unacceptable or uncomfortable behavior (eg, urine marking, fear reactions).
With this approach, the synthetic analogue of pheromone is used as a kind
of psychotrope, the transmitted messages of which are speciﬁc.

Pheromonotherapy has some speciﬁc technical problems. In natural

conditions, the pheromones are not expelled alone. The emitting animal
shows a particular posture or mimicry (eg, the posture for urine marking in
cats or dogs), sometimes shows a part of its body that is usually hidden (eg,
the anal area), modiﬁes the marked substrate (eg, scratches on the ground in
dogs or on vertical surfaces in cats), and also expels some individual odors
and the odor of the pheromone itself. The function of all these messages is to
induce the opening of the VNO, which is usually closed as mentioned pre-
viously. We call these signals emphasizing signals because they emphasize
the emission of the pheromone and improve the probability of the receiving
animal detecting the expelled pheromone. Of course, it is diﬃcult to produce

Fig. 12. Usefulness of F3 in preventing behavioral problems in cats during holidays [39]. This is
a controlled, randomized, and double-blind clinical trial. The studied population includes 68
cats without any behavioral problem at home. When arriving at the holiday home, the owners
had to spray the main pieces of furniture and walls before introducing the cat. Three parameters
were studied: occurrence of urine spraying, time elapsing before eating for the ﬁrst time in the
new surroundings, and number of cats coming back to the new home after a walk outside
(return).

203

P. Pageat, E. Gaultier / Vet Clin Small Anim 33 (2003) 187–211

emphasizing signals except by using the odor of the pheromone. This means
that we diﬀuse more pheromone than the necessary natural amount to make
sure that there is a really high probability for the treated animal to smell the
odor of the synthetic analogue and thus to open his VNO.

Another particularity of the use of pheromones in behavioral therapy is

the need to inform the owner of the particularities of pheromones so as to
help him or her use the product under the right conditions.

Prescribed and used in the right way, pheromones show a great eﬀect and

can help the practitioner solve many problems in a totally safe way. The
pheromones are only messengers that do not penetrate into the organism.
Reception of pheromones creates an input and sets oﬀ internal and
physiologic reactions. Therefore, there is no toxicity or side eﬀects. This
property of pheromones is particularly helpful in old or ill animals and
allows associations with psychotropes.

The ﬁrst pheromone of Carnivores that we have been able to synthesize is

the F3 facial pheromone of the cat [4,5]. This pheromone shows interesting
eﬃcacy in inhibiting urine marking (Fig. 10) [39,41–51] and scratching

Fig. 13. Eﬀect of F3 on manifestations of stress in cats during transport [40]. The studied
population includes 32 male (30 neutered) and 26 female (21 spayed) cats from 1 to 7 years old.
This is a controlled, randomized, and double-blind study. Each cat stays in a crate during the
travel period (100–500 km). The crate is treated with F3 or controlled spray 10 minutes before
the cat enters the crate. The neurovegetative score is the number of defecation, vomiting, and
spoiling episodes occurring during travel. The assessment of the driver is a score based on a
scale of seven degrees ranging from 0 (comfortable travel) to 6 (continual annoying
manifestations that do not end when the driver stops the car).

204

P. Pageat, E. Gaultier / Vet Clin Small Anim 33 (2003) 187–211

(Fig. 11) [4,52]. The observation of the behavior of F3 (Feliway)-treated cats
shows not only a decrease in the unacceptable marking behaviors but an
improvement in feeding, scanning, and playing (Fig. 12) [4,5,39,44,46,48]. It
emphasizes that the positive eﬀects of F3 in treating spraying are not only
linked to the antagonism between facial and urine marking but are related to
a decrease in anxiety. This anxiolytic eﬀect becomes really evident when we
observe the results of the clinical trials evaluating the eﬀects of F3 in treating
transport-related disorders in cats (Fig. 13) [40], in calming cats before
intravenous catheterization [53], or in preventing stress-induced anorexia in
hospitalized cats [54]. In these cases, F3 shows a great emotional stabiliza-
tion function.

The other synthetic facial pheromone used in the cat is the F4 fraction

(Felifriend). The use of this facial fraction makes intra- or interspeciﬁc
interactions easy (Figs. 14 and 15). When an unknown animal is treated with

Fig. 14. Eﬀect of F4 for prevention of intraspeciﬁc aggression in poorly socialized cats [24]. This
trial is designed to assess the eﬃcacy of the F4 pheromonal fraction to get a cat that had never
accepted any speciﬁc in its home to welcome an unfamiliar cat. Seventy cats (35 pairs, 6 months
to 7 years old) are used in this monocentric pilot study. Treatment consists in daily application
of F4 on both sides of the head and ﬂanks of the 35 pairs of cats for 28 days. The cohabitation is
considered successful as soon as the owners see allomarking between the 2 cats.

205

P. Pageat, E. Gaultier / Vet Clin Small Anim 33 (2003) 187–211

the F4 fraction, the cat is misled about the correct status of the protagonist
and should consider it as a familiar. Therefore, the probability of peaceful
contact increases with the use of this allomarking pheromone; it helps the
resident cat to accept a newcomer [24], and it reduces the risk of aggression
caused by handling [25,55]. The success of the treatment is determined by

Fig. 16. Main comparative results of the use of clomipramine and dog-appeasing pheromone in
the treatment of separation anxiety [56]. Fifty-seven dogs are included in this multicentric double-
blind clinical trial. The inclusion is based on the presence of disturbances (excessive barking,
howling, destructive behavior, and

/or house soiling) and the presence of signs of hyperattach-

ment. The pheromone treatment group receives a plug-in diﬀuser delivering pheromone and
placebo capsules. The clomipramine group receives clomipramine capsules (1–2 mg

/kg twice a

day) and a placebo plug-in diﬀuser. The same behavioral modiﬁcation program is employed in
the two groups. The comparison of eﬃcacy is assessed by an evaluation by the owner at day 28
based on (A) destruction, (B) soiling, (C) vocalization, and (D) global assessment by the owner.

Fig. 15. Eﬀect of F4 to enable handling of cats with phobia of the veterinarian during the
consultation [25]. This is a monocentric, randomized, and single-blind clinical trial. The studied
population includes 26 cats (6 months to 7 years old) that systematically demonstrate aggressive
behavior directed against veterinarians. F4 solution is applied on the hands and arms of the
veterinarian 5 minutes before opening the cat’s traveling crate. For the ﬁrst 2 minutes after
the opening of the crate, the veterinarian has to stand still, leaning on the opposite side of the
examination table on which the crate has been set and presenting his hands to the cat. Once the
ﬁrst stage has elapsed, the investigator can slowly put forward his hands toward the cat and
gradually come into contact with it. Three parameters are recorded: the time before the ﬁrst
peaceful contact (TBPC), the number of aggressions, and the point at which the cat is the
peaceful contact initiator (PCI).

206

P. Pageat, E. Gaultier / Vet Clin Small Anim 33 (2003) 187–211

the willingness of the cat to mark unfamiliar people or animals
spontaneously (this spontaneous renewal of facial rubbing is also the sign
of a treatment success with F3).

In the dog, pheromonotherapy is based on the use of dog-appeasing

pheromone (DAP), which is a pheromonal analogue of the appeasing
pheromone secreted by nursing bitches. This appeasine shows great eﬃcacy
in a wide range of fear-inducing situations. Preliminary results of a clinical
trial have shown that DAP delivered via a plug-in diﬀuser can be used to
treat signs of separation anxiety (eg, destruction, vocalization, house soiling)
(Fig. 16). A preliminary study performed in the United Kingdom conﬁrmed
the use of DAP in the treatment of ﬁrework phobias [50]. Updated trials
attempt to assess the eﬃcacy of DAP on dogs suﬀering from deprivation
syndrome.

Summary

Pheromonotherapy seems to be a new therapeutic approach allowing

practitioners to tackle the treatment of behavioral disorders in a natural,
speciﬁc, and safe way. Although the eﬃcacy of pheromones has been
assessed in some speciﬁc behavioral problems, it seems that their range of
action could cover the wide ﬁeld of reduction of stress. Therefore, the use of
pheromones should not be reduced to treatment of behavioral disorders
(potentially associated with psychotropes or a behavioral modiﬁcation
program) but should be included in a strategy of improving the welfare of
pets in veterinary structures (during examination and hospitalization) and
in breeding networks (separation from the mother and transport).

Moreover, further studies may allow the veterinary practitioner to use

pheromone analogues in the ﬁeld of diagnostics to determine the behavioral
status of a pet (eg, anxious or not, dominant or not). Pheromonotherapy is
at its beginning, and the use of pheromones in various ﬁelds of medicine is
heartening.

References

[1] Bossert WH, Wilson EO. The analysis of olfactory communication among animals.

J Theor Biol 1963;5:443–69.

[2] Alberts AC. Constraints on the design of chemical communication systems in terrestrial

vertebrates. Am Nat 1992;139(Suppl):S62–89.

[3] MacDonald DW. The carnivores: order Carnivora. In: Brown RE, MacDonald DW,

editors. Social odours in mammals. Oxford: Clarendon Press; 1985. p. 619–722.

[4] Pageat P. Properties of cat’s facial pheromones. European Patent EP 0 724 832 B1.

February 3, 1995.

[5] Pageat P. Functions and uses of the facial pheromones in the treatment of urine marking in

the cat. In: Johnston D, Waner T, editors. Proceedings of the XX1st Congress of the World
Small Animal Veterinary Association, Jerusalem, 1996. WSAVA; 1996.

[6] Stoddart DM. The ecology of vertebrate olfaction. London: Chapman and Hall; 1980.

208

P. Pageat, E. Gaultier / Vet Clin Small Anim 33 (2003) 187–211

[7] Adams DR, Wiekamp MD. The canine vomeronasal organ. J Anat 1984;38:771–87.
[8] Wada A, Kogure N, Mizoguchi S, et al. Topographic anatomical study of the feline

vomeronasal organ. Bull Coll Agr Vet Med 1991;48:89–97.

[9] Meredith M. Sensory processing in the main and accessory olfactory systems: comparisons

and contrasts. J Steroid Biochem Mol Biol 1991;39(B):601–14.

[10] Halpern M. The organisation and function of the vomeronasal system. Annu Rev Neurosci

1987;10:325–62.

[11] Keverne EB. The vomeronasal organ. Science 1999;286:716–20.
[12] Salazar I, Sanchez Quinteiro P, Cifuentes JM. Comparative anatomy of the vomeronasal

cartilage in mammals: mink, cat, dog, cow and horse. Ann Anat 1995;177:475–81.

[13] Salazar I, Sanchez Quinteiro P, Cifuentes JM, Garcia Caballero T. The vomeronasal organ

of the cat. J Anat 1996;188:445–54.

[14] Eccles R. Autonomic innervation of the vomeronasal organ of the cat. Physiol Behav

1982;28:1011–5.

[15] Guiraudie G, Pageat P, Nagnan-Le Meillour P. Sus scrofa odorant protein m-RNA.

GeneBank NCBI April 17, 2002.

[16] Karn RC. The mouse salivary androgen binding protein (ABP) alpha subunit closely

resembles chain 1 of the allergen Fel d 1. Biochem Genet 1996;32:271–7.

[17] Tessier Y. Dosage de l’allerge`ne majeur Fel d 1 produit par le chat (Felis silvestris catus L.)

en fonction de son e´tat e´motionnel. Me´moire pour le D.E.A. de Biosciences de
l’environnement et sante´. Marseille; 1999.

[18] Pageat P. La communication dans l’univers des carnivores domestiques. Point Ve´t

1997;28:1055–63.

[19] Dicke M, Sabelis MW. Infochemical terminology: based on cost-beneﬁt analysis rather

than origin of compounds? Funct Ecol 1988;2:131–9.

[20] Bekoff M. Scent marking by free ranging domestic dogs: olfactory and visual components.

Biol Behav 1979;4:123–39.

[21] de Boer JN. The age of olfactory cues functioning in chemocommunication among male

domestic cats. Behav Processes 1977;2:209–25.

[22] Doty RL. Odor-guided behavior in mammals. Experiencia 1986;42:257–71.
[23] Zielonka TM, Charpin D, Berbis P, et al. Effects of castration and testosterone on Fel d 1

production by sebaceous glands of male cats: I—immunological assessment. Clin Exp
Allergy 1994;24:1169–73.

[24] Pageat P, Tessier Y. Usefulness of the F4 synthetic pheromone for preventing intraspeciﬁc

aggression in poorly socialized cats. In: Mills DS, Heath SE, Harrington LJ, editors.
Proceedings of the First International Conference on Veterinary Behavioural Medicine,
Birmingham, 1997. p. 64. Herts, UK: UFAW; 1997.

[25] Pageat P, Tessier Y. F4 synthetic pheromone: a means to enable handling of cats with a

phobia of veterinarian, during consultation. In: Mills DS, Heath SE, Harrington LJ,
editors. Proceedings of the First International Conference on Veterinary Behavioural
Medicine, Birmingham, 1997. p. 108. Herts, UK: UFAW; 1997.

[26] Pageat P. Pathologie du comportement du chien. 2nd edition. Paris: Ed. Le Point

Ve´te´rinaire; 1998. p. 28–32.

[27] Carayol N, Birnbaum J, Magnan A, et al. Fel d 1 production in the cat skin varies

according to anatomical sites. Allergy 2000;55:570–3.

[28] Bekoff M. Ground scratching by male domestic dogs: a composite signal. J Mammal

1979;60:847–8.

[29] Albone ES, Gosden PE, Ware GC, et al. Bacterial action and chemical signalling in the red

fox, Vulpes vulpes, and other mammals. Proc Am Chem Soc 1978;67:78–91.

[30] Donovan CA. Canine anal glands and chemical signals (pheromones). JAVMA 1969;

155:1995–6.

[31] Ashdown RR. Symposium on canine recto-anal disorders. I. Clinical anatomy. J Small

Anim Pract 1968;9:315–22.

209

P. Pageat, E. Gaultier / Vet Clin Small Anim 33 (2003) 187–211

[32] Preti G, Meutterties EL, Furman J, et al. Volatile constituents of dog (Canis familiaris)

and coyote (Canis latrans) anal sacs. J Chem Ecol 1976;2:177–87.

[33] Dornelas de Andrade A, Birnbaum J, Magalon C, et al. Fel d 1 levels in cat anal glands.

Clin Exp Allergy 1996;26:178–80.

[34] Gorman L, Trownbridge B. The role of odor in the social lives of carnivores. In: Gittleman

JL, editor. Carnivore behavior, ecology and evolution. Ithaca: Cornell University Press;
1989. p. 57–88.

[35] Goodwin M, Gooding KM, Reginer F. Sex pheromone in the dog. Science 1979;203:

559–61.

[36] Pageat P. Pig appeasing pheromone to decrease stress, anxiety and aggressiveness. United

States Patent 6,077,867. June 20, 2000.

[37] Johnson RP. Scent marking in mammals. Anim Behav 1973;21:521–35.
[38] Doty RL, Dunbar I. Attraction of beagles to conspeciﬁc urine, vaginal and anal sac

secretion odours. Physiol Behav 1974;12:825–33.

[39] Pageat P, Tessier Y. Usefulness of the F3 synthetic pheromone Feliway in preventing

behaviour problems in cats during holidays. In: Mills DS, Heath SE, Harrington LJ,
editors. Proceedings of the First International Conference on Veterinary Behavioural
Medicine, Birmingham, 1997. p. 231. Herts, UK: UFAW; 1997.

[40] Gaultier E, Pageat P, Tessier Y. Effect of a feline appeasing pheromone analogue on

manifestations of stress in cats during transport. In: Veissier I, Boissey A, editors.
Proceedings of the 32nd Congress of the International Society for Applied Ethology,
Clermont-Ferrand, 1998. INRA; 1998. p. 198.

[41] Frank D. Treatments of spraying cats. In: Proceedings of the Annual Convention of the

American Veterinary Medical Association, Baltimore, 1998. p. 225–6.

[42] Frank D, Erb HN, Houpt KA. Urine spraying in cats: presence of concurrent disease and

effects of pheromone treatment. Anim Behav Sci 1999;61:263–72.

[43] Hunthausen W. Evaluation of the use of pheromone analogue (Feliway) to control urine

marking in cats. In: Proceedings of the American Veterinary Society of Animal Behavior,
Baltimore, 1998. p. 6–7.

[44] Hunthausen W. Evaluating a feline facial pheromone analogue to control urine spraying.

Vet Med 2000;143:151–6.

[45] Landsberg G. Case challenge series—feline housesoiling. In: Proceedings of the North

American Veterinary Clinics, Orlando, 2001. p. 47–9.

[46] Mills DS, White JC. Long-term follow up of the effect of a pheromone therapy on feline

spraying behaviour. Vet Rec 2000;147:746–7.

[47] Mills DS, Mills CB. Evaluation of a novel method for delivering a synthetic analogue of

feline facial pheromone to control urine spraying by cats. Vet Rec 2001;149:197–9.

[48] Mills DS. Pheromonotherapy—an integral part of modern companion animal practice.

UK Vet 2002;7:1–3.

[49] Ogata N, Takeuchi Y. Clinical trial with a feline pheromone analogue for feline urine

marking. J Vet Med Sci 2001;63:157–61.

[50] White JC, Mills D: Efﬁcacy of synthetic feline facial pheromone (F3) analogue (Feliway) for

the treatment of chronic non-sexual urine spraying by the domestic cat. In: Mills DS, Heath
SE, Harrington LJ, editors. Proceedings of the First International Conference on
Veterinary Behavioural Medicine, Birmingham, 1997. Herts, UK: UFAW; 1997.

[51] Pageat P. Experimental evaluation of the efﬁcacy of a synthetic analogue of cat’s facial

pheromones (Feliway) in inhibiting urine marking of sexual origin in adult tom-cats. J Vet
Pharmacol Ther 1997;20(Suppl 1):169.

[52] Mertens PA. Snifﬁng around for information on feline olfaction. In: Proceedings of the

North American Veterinary Conference, Orlando, 2001. p. 52.

[53] Kronen PW, Ludders JW, Erb HN, et al. The F3-fraction of feline facial pheromones

calms cats prior to intravenous catheterization. In: Proceedings of the 7th World Congress
of Veterinary Anaesthesia, Berne, 2000. p. 71.

210

P. Pageat, E. Gaultier / Vet Clin Small Anim 33 (2003) 187–211

[54] Grifﬁth C, Steigerwals E, Bufﬁngton CAT. Effects of a synthetic facial pheromone on

behavior and food intake of caged cats. In: Proceedings of the 3rd ESVCN Congress,
Lyon, 1999.

[55] Kakuma Y, Bradshaw JWS. Effects of feline facial pheromone analogue on stress in

shelter cats. In: Overall KL, Mills DS, Heath SE, Horwitz D, editors. Proceedings of the
3rd International Congress on Veterinary Behavioural Medicine, Vancouver, 2001. Herts,
UK: UFAW; 2001.

[56] Pageat P, Gaultier E. Treatment of separation anxiety in dogs with a synthetic dog

appeasing pheromone—preliminary results. In: Dehasse J, Biosca Marce E, editors.
Proceedings of the 8th ESVCE Meeting on Veterinary Behavioural Medicine. Paris:
Publibook; 2002. p. 49–54.

211

P. Pageat, E. Gaultier / Vet Clin Small Anim 33 (2003) 187–211

Medical diﬀerentials with potential

behavioral manifestations

Karen L. Overall, MA, VMD, PhD

Psychiatry Department, University of Pennsylvania School of Medicine, 415 Curie Drive,

50 B-CRB, Philadelphia, PA 19104, USA

How to think logically about situations where so much is unknown

Behavioral medicine is complex. All signs are nonspeciﬁc, and we can

never hold the behavioral pathologic proﬁle in our hand. We can only assess
a verbal, spectral, or chemical manifestation of the behavioral pathologic
ﬁndings.

Behavior is the great and ﬁnal integrator between any animal’s internal

(physiologic

/neurochemical, neuroanatomic, and genetic components) and

external environments and the extent to which all these environments
interact to produce long-term potentiation (cellular

/molecular memory and

learning) [1–3]. Not only are behavioral problems the most common
concerns for pets and veterinary clients [4–7], but when an animal is truly
‘‘organically’’ ill, the mechanism by which the client recognizes the illness
involves a change in behavior.

Common documented ‘‘organic’’ or ‘‘medical’’ causes of behavioral

changes include congenital, inherited, and genetic infectious, inﬂammatory
or immune-mediated, metabolic and endocrine, nutritional, degenerative,
neoplastic, toxic, and traumatic conditions. Known medical conditions
with a neurologic or behavioral eﬀect include epilepsy, narcolepsy, hydro-
cephalus, polyphagia, and some forms of stereotypic behaviors. It is
important to remember, especially in light of the recent ﬁndings about
narcolepsy, that the extent to which these neurologic conditions seem
separate from ‘‘behavioral’’ ones may be a function of our ignorance: the
same set of regulatory genes that contributes to the narcoleptic pathologic
changes may, in fact, be at work in many anxiety-related conditions
involving dysregulation of neurochemical systems [8–10].

Vet Clin Small Anim

33 (2003) 213–229

E-mail address:

overallk@mail.med.upenn.edu

0195-5616/03/$ - see front matter

PII: S 0 1 9 5 - 5 6 1 6 ( 0 2 ) 0 0 1 0 3 - 1

It is also equally important to realize that when we evaluate the behaviors

involved in a behavioral complaint or pathologic ﬁnding, our observational
skills and behavioral history allow us to make assessments only at the
phenotypic level—the level at we can describe what something looks like.
We need to remember that there are many mechanistic ‘‘causes’’ for any
behavioral condition and that we are lacking suﬃcient data for most
conditions to evaluate the extent to which they are heterogeneous, poly-
morphic, multifactorial, or representative of spectrum disorders. In too
many cases, we are lacking the heuristic context in which to evaluate
various presentations of the same condition, which is why a structured
thought process is so critical [3]. For example, obsessive-compulsive
disorder (OCD) has many forms. It is likely that the locomotor forms
diﬀer from those involving hallucinations or ingestive behaviors in some
profound mechanistic ways. Recent research strongly suggests that OCD in
human beings is the result of genetically controlled dysfunction of genes
involving regulatory systems [11,12]. There are also possible roles for diet-
ary cofactors in aﬀecting subcellular signaling [13]. Such complex regulatory
functions having a genetic heritable basis have also been reported for dogs
with narcolepsy [10] and warrant further investigation in dogs and cats
aﬀected with OCD.

Implicit in the intellectual construct of a truly ‘‘abnormal’’ or

‘‘pathologic’’ behavior is the concept that the behavioral presentation, or
the phenotype, is driven by some integrated anomaly that functions in
concert with numerous levels from the genomic and subcellular to the
neuroanatomic [14–16]. Using this integrated mechanistic paradigm, we can
view truly abnormal behaviors as we would metabolic disorders—as another
form of an organic condition. Unfortunately, there are few hard data that
support the extent to which any speciﬁc underlying pathophysiologic
mechanism or cause is involved in any behavioral condition, so it is as
foolish to assume that every behavioral complaint is purely behavioral as it
is to assume that none are. There is no substitute for paying attention to the
pattern of the pet’s behavior.

The roles for history and observation: collecting the data

A good behavioral history combined with honed observational skills (and

even the most inexpert observer can hone his or her skills by having clients
videotape the problem pet) helps to suggest where diagnostic probability
should be placed. Pattern recognition is key, and a few tips should help
veterinarians with this.

For example, purely behavioral conditions seldom appear quickly, and

the early and more subtle signs are often missed by client and veterinarian
alike. A cat that cries every time it urinates and still uses its litter box but
dislikes entering it since the period when the client noticed the crying is
unlikely to have a purely behavioral problem but is likely to have an

214

K.L. Overall / Vet Clin Small Anim 33 (2003) 213–229

underlying medical condition. The cat that does not cry when it urinates but
does not use the box at all since a new cat was introduced to the household
ﬁts the proﬁle of a cat with a nonmedical behavioral concern. What clues
were present in this simple dichotomous case example that allowed us to
draw these conclusions? The discomfort experienced by the ﬁrst cat may be
perceived by the cat to be associated with the litter box, and, accordingly,
the cat is becoming reluctant to get into the box. The second cat seems to
show no signs of discomfort but is demonstrating the extent to which urine,
scent, and subtle social interactions can be important in modulating up-
heavals in feline groups. The veterinarian should perform a urinalysis
on both cats; in the latter case, one would be doing so to rule out the
complication of a medical contribution. It is unlikely that the urinalysis
will be informative regarding the etiology of the underlying problem. To
perform a urinalysis in the case of the second cat is to be thorough, but the
veterinarian would have a low index of suspicion that the change in
behavior was driven by an underlying organic pathologic change associated
with the urogenital system. This does not mean that behavioral changes
could not aggravate chronic feline lower urinary tract disease (FLUTD),
given what we now know about the complexity of stress and neuromyogenic
responses, nor does it imply the absence of a concomitant medical condition.
In the ﬁrst case, pathologic ﬁndings of the urogenital system warrant a high
index of suspicion. A thorough history would examine the potential
secondary behavioral complications of pain associated with elimination: the
development of anxieties, substrate aversions, and location aversions.
Oddly, tricyclic antidepressants (TCAs) may help both cats, because these
drugs address inﬂammation, the receptors involved in myogenic pain, and
the anxiety associated with pain and social discomﬁture. This also illustrates
how examining only one mechanistic level (eg, the neurochemical one) in a
vacuum could lead to confusion about the entire ‘‘causal’’ process.

A logical structured thought process allows the veterinarian to acquire the

data to pursue tandem treatment of any organic problems and the associated
behavioral changes that may be associated with them (and vice versa).

An organ systems approach to identifying common organic problems
that may best be recognized by their behavioral signs

Adequate extant reviews of medical conditions that can be associated

with behavioral signs have previously been published with relevant
references [14,17,18]. Accordingly, the nonspeciﬁc signs and the systems
involved are summarized here in tabular form. Table 1 summarizes
conditions associated with organ system–speciﬁc nonspeciﬁc signs and po-
tential underlying causes. Table 2 summarizes age associations with certain
medical

/organic conditions that may aﬀect behavior or have a primarily

behavioral presentation.

215

K.L. Overall / Vet Clin Small Anim 33 (2003) 213–229

Table 1
Non-speciﬁc behavioral signs, organ systems potentially involved, and organic pathology
associations

Nonspeciﬁc signs

/system involved

Underlying ‘‘cause’’

Inappropriate elimination

urogenital system complains

Degenerative

/developmental

Vesicourachal diverticula
Ectopic ureter
Lissencephaly
Portosystemic shunting
Hydrocephalus

Acquired

Intraluminal obstruction: urolithiasis, polyps,

blood clots, sloughed tissue

Extraluminal obstruction: tumor, prostatic

disease, stricture, hernia

Metabolic

/endocrine

Diabetes mellitus
Renal glucosuria
Central diabetic insufﬁciency
Nephrogenic diabetic insufﬁciency
Renal insufﬁciency

/failure

Hypercalcemia
Hypokalemia
Hyperadrenocorticism
Hyperthyroidism
Hypoadrenocorticism
Hepatic insufﬁciency
Renal medullary washout
Primary polydipsia

Neurogenic: loss of control

/voluntary

Lesions of the

cerebrum
basal ganglia
thalamus
cerebellum
external urethral sphincter
perineal muscles
pudendal nerve

Upper motor neuron disease: reﬂex dyssynergia,

large amounts of infrequently produced urine,
bladders that are difﬁcult to express because of
the increased urethral tone

Lower motor neuron disease (S

–S

or L

–S

nerve roots and pudendal and pelvic nerves;
associated with bladder atony (ﬂaccid,
neuropathic bladders), large urinary volumes,
and overﬂow; easy expression of the bladder;
may also have fecal obstructions, hind limb

/tail

deﬁcits, loss of anal tone

Inﬂammatory

Prostatitis
Vaginitis
Cystitis

/FLUTD

216

K.L. Overall / Vet Clin Small Anim 33 (2003) 213–229

Table 1 (continued )

Nonspeciﬁc signs

/system involved

Underlying ‘‘cause’’

Urethritis
Enteritis
Colitis

Infectious

Cell-associated herpes virus
Syncytia-forming virus
Feline calicivirus
Pyelonephritis
Pyometra
Parasitemia (primarily concerns involving

defecation)

Complaints involving aggressive

behaviors

Degenerative

/developmental

Lissencephaly
Hydrocephaly
Porencephaly
Congenital portosystemic shunts
Congenital urea cycle enzyme defects
Cerebral hypoxia
Seizure activity
Feline ischemic syndrome
Lysosomal storage diseases
Chronic polioencephalomalacia in the piriform

lobes and hippocampus

Endocrine

/metabolic

HE (possibly associated with changes in

perception, fear, and incoordination)

Hyperthyroidism (remember that in cats with

hyperthyroidism, serum transaminases and ALP
are often increased)

Uremic encephalopathy
high levels of corticotropin (increased aggression)
Low levels of corticotropin (decreased aggression)

Nutritional

Thiamine deﬁciencies (primarily cats)
Taurine deﬁciencies (primarily cats)
High-protein diets (may affect some reactivity,

caution is urged)

Neoplastic

Intracranial neoplasia, including meningiomas
Temporal lobe, limbic system, and hypothalamic

lesions

Lesions in the VMH and PLH in cats

Infectious

Rabies
Toxoplasmosis
Distemper
FIP
Tick-borne conditions
Miscellaneous bacterial and fungal conditions

(continued on next page)

217

K.L. Overall / Vet Clin Small Anim 33 (2003) 213–229

Table 1 (continued )

Nonspeciﬁc signs

/system involved

Underlying ‘‘cause’’

Toxicity

Heavy metal
Psychotropic medication–induced serotonin

syndrome

Organophosphates

Trauma

Cerebral injury
Cerebral vascular disease
Cerebral infarct (cats)
Cuterebra

larval infarcts

Complaints involving depression,

sleepiness, or listlessness

Degenerative

/developmental

Lissencephaly
Portosystemic shunting
Lysosomal storage diseases

Endocrine

/metabolic

HE (can be associated with parenchymal liver

disease, [including cirrhosis, acquired
portosystemic shunts, toxicity, and neoplasia]
congenital portosystemic shunts, and congenital
urea cycle enzyme defects)

Hepatic dysfunction

/failure (postulated to be

associated with increased gamma aminobutyric
acid resorption from the gastrointestinal tract)

Thyroidal illness (primarily canine

hypothyroidism)

Uremic encephalopathy
Hyperkalemia
Hypothyroidism
Hyperadrenocorticism

Neoplastic

Thalamic, subthalamic, midbrain,

and frontal lobe lesions

Intracranial neoplasia (primarily responsible for

stupor and coma)

Pontine

/tegmental lesions (cats)

Toxicity

Heavy metal
Drugs (prescription and human recreational)

(anticonvulsants increase alkaline phosphatase,
aspartate aminotransferase, and glutamyl
transferase and may affect alanine amino-
transferase (glutamic pyruvic transaminase);
glutamyl transferase has been implicated as
a ﬂag for hepatotoxicity and necrosis associated
with atypical diazepam toxicity in cats)

Trauma

Cerebral injury
Cerebral vascular disease
CCT (can be caused by ﬁghts between animals)

218

K.L. Overall / Vet Clin Small Anim 33 (2003) 213–229

Table 1 (continued )

Nonspeciﬁc signs

/system involved

Underlying ‘‘cause’’

Complaints involving changes in

ingestive behaviors

Endocrine

/metabolic

Hepatoencephalopathy
Hepatic dysfunction

/failure

Hyperthyroidism
Diabetes
Hyperadrenocorticism

Neoplastic

Thalamic lesions (polyphagia, polydipsia, pica,

and aphagia)

Lesions in the VMH and PLH in cats

Infectious

Rabies

Complaints involving cognition,

including dementia and inability
to learn

Neoplastic

Frontal lobe and internal capsule lesions

Nutritional

Thiamine deﬁciencies (primarily cats)
Taurine deﬁciencies (primarily cats)

Infectious

Distemper
Rabies
Mycoses
Toxoplasmosis
Prion-associated conditions

Traumatic

Frontal lobe and internal capsule lesions

Complaints involving ritualistic

behaviors

Degenerative

Granulomatous meningoencephalitis
Cauda equina syndrome

Endocrine

/metabolic

Hypocalcemia
Hypomagnesemia
Abnormalities of acid-base balance

Neoplastic

Lesions in the frontal lobe, internal capsule, and

basal nuclei (particularly the caudate nucleus)

Nutritional

Excessively low-protein diet
Food hypersensitivity

Infectious

Rabies

Toxic

Tetanus
Botulism

Complaints involving non-speciﬁc

fear and anxiety

Degenerative

Auditory changes (deafness)
Visual changes (cataracts)
Mobility changes (arthritis)

(continued on next page)

219

K.L. Overall / Vet Clin Small Anim 33 (2003) 213–229

There are three issues that concern us when we are faced with the question
of whether the presenting nonspeciﬁc signs are behavioral or medical:

1. How can we routinely check, by organ system or by clusters on

nonspeciﬁc signs that are largely associated with speciﬁc organ systems,
whether the presenting signs are most consistent with a purely
nonbehavioral diagnosis?

2. How can we routinely perform a similar check by age?
3. How can we rigorously test newer ideas about the effects of other systems

on behavior, given proposed treatments that involve these systems (eg,
the thyroid axis role in nonspeciﬁc signs associated with anxiety)?

All three of these issues involve assessment of probability. Unfortunately,

probability can only be conﬁdently relied on when suﬃcient data are

Table 1 (continued )

Nonspeciﬁc signs

/system involved

Underlying ‘‘cause’’

Endocrine

/metabolic

Hypothyroidism
HPA axis disorders
Disorders involving glucose metabolism
HE
Uremic encephalopathy

Neoplastic

Lesions in the frontal lobe, internal capsule, and

basal nuclei

Infectious

Rabies
Toxoplasmosis
Distemper
FIP
Tick-borne conditions
Miscellaneous bacterial and fungal conditions
Prion-associated conditions

Toxicity

Heavy metal
Psychotropic medication–induced serotonin

syndrome

Organophosphates

Trauma

Cerebral injury
Cerebral vascular disease
CCT (can be caused by ﬁghts between animals)

Complaints involving sexual

behavior

Neoplastic

Temporal lobe, limbic system, and hypothalamic

lesions

Abbreviations

: ALP, alkaline phosphatase; CCT, cranial cerebral trauma; FIP, feline

infectious peritonitis; FLUTD, feline lower urinary tract disease; HE, hepatoencephalopathy;
HPA, hypothalmic pituitary adrenal; PLH, posterolateral hypothalamus; VMH, ventromedial
hypothalamus.

220

K.L. Overall / Vet Clin Small Anim 33 (2003) 213–229

available about the nonspeciﬁc signs associated with aﬃliated medical and
behavioral conditions [16] and when we know how to elicit the relevant data
using a thorough behavioral history designed to ferret out relevant
correlations [14,15,19]. The third issue involves more knowledge of com-
plex interactions than is commonly available. The use of algorithms that
structure thought processes is essential here.

Finally, discussed brieﬂy as examples are just two of the complicating

issues that overlay all concerns for medical diﬀerentials: roles of hormones
like thyroid hormone and roles of sex and gender.

Table 2
Age ranges and associations with organic conditions in which behavioral signs are key

Age group

Common conditions

<9 months of age (youngsters)

Congenital hydrocephalus
Lissencephaly
Lysosomal storage diseases
Distemper

/FIP encephalitis

Viral, fungal, protozoal, and bacterial

encephalitis

Trauma
Toxicity, primarily lead
Hypoglycemia
HE (portosystemic shunt)
Congenital defects and metabolic disease
Thiamine deﬁciencies

9 months (sexual not

social maturity) to 5 years

Distemper

/FIP encephalopathy

Viral, protozoal, or fungal encephalopathies
Steroid-responsive meningoencephalitis
Granulomatous meningoencephalitis
Trauma
Toxicity
Hypoglycemia
HE (acquired hepatopathy

/portocaval shunt)

Other acquired metabolic disease
Acquired epilepsy
Cerebral neoplasia

>5 years

Distemper

/FIP

Steroid-responsive meningoencephalopathy
Granulomatous encephalopathy
Trauma
Toxicity
Hypoglycemia (insulinoma)
HE (acquired hepatopathy)
Other metabolic disease
Acquired epilepsy
Cerebral neoplasia
Watch in the future for prion-associated

conditions

Abbreviations

: FIP, feline infectious peritonitis; HE, hepatoencephalopathy.

221

K.L. Overall / Vet Clin Small Anim 33 (2003) 213–229

The issue of thyroidal function

Hypothyroidism has been postulated to be associated with fear or

aggression in dogs and with sudden shifts in these behaviors. Hypothyroidism
has been reported to account for 1.7% of canine aggressive behaviors [20].
Dogs with hypothyroid-associated aggression may not show the other classic
signs of hypothyroidism, such as a poor hair coat, lethargy, or weight gain.
Aggression is gradual in onset, and triggers can be inconsistent. Appropriate
thyroid screening should be part of a minimum database on an aggressive pet
to rule out thyroid-related diseases; however, caution is urged in over-
interpretation of data, given the relative rareness of the condition. The canine
prevalence of hypothyroidism is thought to be approximately 0.2%.

Hyperthyroidism in cats has been associated with manic behaviors, in-

appetence

/anorexia, and aggression. The feline prevalence of hyperthy-

roidism is likely higher than that of hypothyroidism in dogs, but since its
ﬁrst report as a clinical disorder in 1979, the frequency of diagnosis con-
tinues to increase [21–23].

It is important to realize that the estimated incidence of behavioral

problems is considerably greater than either of these endocrine conditions.
Although the age of onset for hypothyroidism overlaps that for the de-
velopment of most behavioral conditions in dogs (

12–24 months of age)

[19,22,24], the age of onset of hyperthyroidism is greater (middle-aged to
older cats) [23] than that for most behavioral conditions in cats (

24–36

months). These data are important because they show that although
behavioral and thyroidal conditions may be comorbid, the patterns in-
volving age of onset strongly suggest that any rule postulating a causal
association may be overly optimistic. Furthermore, for cats, environmental
factors associated only with learned behaviors (eg, dietary choices and
preferences) seem to play some role in the development of hyperthyroidism.

In human beings, an association between thyroid dysfunction and

depression has been suggested but is diﬃcult to prove. True hypothyroidism
has had clinical depression as one of its nonspeciﬁc signs; however, true
clinical hypothyroidism seems to be a relative rather than absolute state.
True clinical hypothyroidism aﬀects less than 1% of the human population,
but ‘‘subclinical hypothyroidism,’’ a description rather than a diagnosis of a
biochemically measurable degree of relative thyroid failure, aﬀects 5% to
10% of the population, is more common in women and the elderly, and may
be associated with autoimmune conditions or conditions involving neuro-
immunologic regulatory function [25]. The eﬀect of this subclinical disease
on the treatment of depression in human patients may be to retard the
intended outcome of psychopharmacologic treatment as a result of these
complex interactions.

Decreased thyroid activity in clinical depression in human beings may,

paradoxically, be accompanied by an exaggerated thyroid stimulation
response to exogenous thyrotropin-releasing hormone (TRH) in approx-

222

K.L. Overall / Vet Clin Small Anim 33 (2003) 213–229

imately 10% of depressed human patients [26,27]. Thyrotropin response to
TRH may be reduced in up to one third of all depressed human subjects [28].
These patients seem to be nonresponders to traditional antidepressant
medications but can be converted to responders with the addition of
thyroxine. This eﬀect has not been noted for the 90% of the human
population demonstrating a normal TRH response. Transthyretin is the
protein associated with thyroid hormone transport, and it seems to be low in
nonresponder depressed patients with normal peripheral T

concen-

trations. The extent to which transthyretin abnormalities are involved
in this response is unknown, but in a small sample of patients with major
depression who were refractory to treatment with antidepressants,
cerebrospinal ﬂuid (CSF) levels of transthyretin were signiﬁcantly lower
than for neurologic controls. The condition would then involve a defect
primarily affecting a transporter protein, suggesting that the association
between neurotransmitters and depression may be both direct and indirect.

It seems that consistent patterns involving alteration of thyroidal

hormones occur in human patients acutely aﬀected with aﬀective illness
[29,30]. These patterns may be a function of the brain’s ability to regulate
thyroid hormone levels independent of peripheral needs, however. Basal
thyroid-stimulating hormone (TSH) production and the response of TSH
to TRH are inﬂuenced by nonthyroid and non-TRH factors, including
somatostatin, dopamine, and serotonin [29]. This suggests that no single
measurements of thyroidal function can be viewed as an indicator of thyroid
axis dysfunction in aﬀective disorders in human beings. Certainly, such
implications are important for dogs, where age and breed reference ranges
are not fully explored and may vary considerably. Furthermore, correlations
between sex, thyroidal function, and aﬀective illness in human beings may
be spurious and the result of another more complex association. The c-erb-
A family of gene products shares considerable homology at the binding
regions for nuclear T

, glucocorticoid, and estrogen receptors. Correlational

associations may be the result of poorly or incompletely understood
interactions between families of transcriptional factors [30].

That said, although there have been anecdotal suggestions for the use

[17], either alone or in combination, of thyroid supplementation for the
treatment of canine behavioral conditions, there is currently no good
rationale for supplementation with exogenous thyroxin in the absence of
speciﬁc behavioral clinical signs and aberrant levels or a response to
transthyretin in pets with behavioral problems. Such cavalier dispensation
of potent medication is particularly problematic given the wide range of
breed-speciﬁc reference ranges for T

, T

, free T

, and free T

. Resting free

levels, if grossly low, may provide a rough gauge to thyroid function,

because free T

is less likely than total T

to be affected by nonthyroidal

illness or drug therapy [31]. Most accurately measured using dialysis
methods, low free T

concentration combined with an increase in TRF is

consistent with a diagnosis of hypothyroidism [31]. In the absence of clinical

223

K.L. Overall / Vet Clin Small Anim 33 (2003) 213–229

signs of hypothyroidism (obesity, seborrhea, alopecia, weakness, lethargy,
bradycardia, and pyoderma) and low (not low-normal) free T

(or the

equivalent TSH stimulation test result), there is no rational or competent
reason to treat dogs with behavioral or any other diagnoses with thyroxin
[32]. There may be a small population of dogs for which this generalization
might prove false in the future, but they should be rare. These dogs, if they
exist, may have anxiety conditions that are affected by cholecystokinin
(CCK). In male rats, CCK-A (alimentary) receptor stimulation inhibits TSH
secretion at the level of the anterior pituitary [33].

This is not to say that dogs with thyroid disorders do not alter their

behavior; they do, but the behavioral changes are nonspeciﬁc and should
not be confused with a behavioral diagnosis. There may be nonresponders
to antidepressant and antianxiety medications in the canine population that
are analogous to those in the human population. If so, supplementation
with thyroxin in addition to the antidepressant or antianxiety medications
may help. Nevertheless, most single-mechanism (ie, thyroid) hypotheses for
the underlying cause and treatment of complex behavioral phenomena like
aggression and fear are simplistic and invariably wrong. In this case, sup-
plementation is not benign; executed in the absence of a diagnosis, it is
a poor reﬂection on the practice of veterinary medicine. Although the issue
of hormonal interaction in behavioral conditions is a complex one, sup-
plementation with thyroxin is largely a popular bandwagon movement for
which there are no rigorous data. Collection of these data would doubt-
less be enlightening.

In human psychiatry, the current view reﬂects the absence of a speciﬁc

link between any abnormality of the thyroid axis and the pathophysiology
of any aﬀective illness [34]. The presence of a behavioral depression or other
aﬀective disorder alone does not warrant full thyroidal assessment or
treatment unless the patient is female and older. We would do better to
focus on the roles of thyroid hormones in the regulation of neural de-
velopment and other complex links. In the embryonic or developing
hypothyroid brain, morphologic and biochemical alterations of neurons are
comparable to those seen in degenerative illness [35]. This is not surprising,
given that deﬁciencies of thyroid hormone in neonates and children lead
to abnormal brain development with concomitant deﬁcits in behav-
ior, locomotor ability, speech, hearing, and cognition. These functions
may be partially caused by interreliance between thyroid hormones and
acetylcholine, nerve growth factor, and hippocampal function [36]. Accord-
ingly, the reported correlation between cognitive decline, age, and thyroid
function may be a result of complex interactions between regionally
relevant populations of cholinergic neurons, nerve growth factors, and
all factors aﬀecting learning. In such cases, thyroid hormones may be best
viewed as one cog in a complex neuromodulatory scheme. Certainly, given
this, mere supplementation with thyroxine to treat nonspeciﬁc behavioral
signs is unlikely to be either rational or informative.

224

K.L. Overall / Vet Clin Small Anim 33 (2003) 213–229

The issue of sex

The most common eﬀects of chromosomal sex determination that are

relevant to behavioral medicine are those associated with sex steroid
hormonal function. For example, one of the beneﬁcial eﬀects of castration
and removal of testicular testosterone is the decrease in roaming behavior
that subsequently follows. The decreased roaming may serve to decrease the
local canine population, but it may also have indirect eﬀects on longevity for
the neutered male: intact males of almost any species exhibit more ‘‘risk-
taking’’ behavior than do females, usually as a result of the pursuit of mates.
In some species of lizards, such behaviors decrease the amount of foraging
time available to male lizards. In other species, including many birds, males
are more susceptible to predation and parasitism because of their ap-
pearance [37–42]. Of course, testosterone raises reactivity, and combined
with the search for mates, males of many species are at increased risk for
injury and subsequent infection and debility.

The more subtle eﬀects of sex steroid hormones may be immunologic

ones that indirectly aﬀect behavior [43]. Male sex steroid hormones, an-
drogens, and their female analogue, estrogens, have immunoneuromod-
ulatory properties. Although estrogens and androgens tend to have
similar eﬀects on gonadal tissues, androgens may lower several aspects
of immunity, whereas estrogens enhance these same levels [41]. Even in
controlled laboratory settings, male rodents are more susceptible to
infection than are female rodents. This diﬀerence is attributable to dif-
ferential eﬀects of sex steroid hormones; males are more susceptible to
parasitic, bacterial, fungal, and viral infections than are females. In part, this
eﬀect is a result of lowered humoral immune responses associated with
androgens. There is also a role for T helper (Th) cells: females exhibit higher
Th2 responses, which are associated with higher interleukin (IL)-4, IL-5,
IL-6, and IL-10 responses, than do males. In general, estrogens seem to en-
hance both humoral and cell-mediated immune responses. Female rodents
also have increased immunologic tolerance to foreign substances compared
with male rodents. These subtle eﬀects of sex steroid hormones may cause
males and females to seek and respond to risk, social environments, and
social interactions diﬀerently.

High-aﬃnity androgen and estrogen receptors have been found in the

thymus, bone marrow, and spleen of rodents and in human macrophages,
and estrogen receptors are present in the cytosol of circulating lymphocytes.
Again, diﬀerential responses of exposure to infectious agents may directly
and indirectly aﬀect normal and abnormal behaviors. Certainly, male cats
aﬀected with the feline immunodeﬁciency virus experience more lymphocyte
apoptosis (programmed cell death) than do female cats [44].

In terms of basic Mendelian genetics, animals with only one X

chromosome are at increased risk for expression of conditions carried on
sex chromosomes; deleterious recessive alleles are more likely to be

225

K.L. Overall / Vet Clin Small Anim 33 (2003) 213–229

expressed in the heterogametic sex, and many genes contributing to both
neurochemical synthesis and regulation [45] and immunoregulation are
found on the X chromosome [43].

This last category may have the most relevance for veterinary behavioral

medicine and human psychiatry. Alcoholism, drug abuse, antisocial per-
sonality, attention deﬁcit disorder, Tourette’s syndrome, and completed
suicide predominate in male human beings with psychiatric conditions,
whereas depression, anxiety, eating disorders, and attempted suicide are
more common in female human beings with psychiatric conditions [46].
Unfortunately, we do not know to what extent social and cultural mores,
socioeconomic factors, and overt and subtle gender roles are responsible for
this dichotomy in human beings. In contrast, there is good evidence in many
rodent models of anxiety that the behavioral manifestation of anxiety is
dependent on the phase of estrus and the relative roles of estradiol and
progesterone. In veterinary behavioral medicine, most dogs aﬀected with
‘‘dominance’’ or impulse-control aggression are male. When the aﬄicted
patients are female, however, these dogs exhibit a diﬀerent subset of signs
than do aﬄicted male dogs, and they exhibit the associated behaviors
at a younger age than do male dogs [47]. For OCD, male dogs are over-
represented compared with female dogs. Aggressive behaviors, and some
ritualistic ones, are aﬀected by alterations in the neurotransmitters se-
rotonin, norepinephrine, and dopamine. Some of the genes coding for these
neurotransmitters or enzymes that metabolize them are on the X chro-
mosome. Accordingly, deletions in these X chromosome genes may be crit-
ically important for our understanding of the mechanisms underlying many
problematic behaviors [45,48].

It is also possible that in addition to the temporal and cyclic eﬀects of

circulating sex steroid hormones, behaviors may have been shaped by eﬀects
of these hormones early in ontogeny. Organizational structure of many
brain regions and the extent to which lateralization of some functions occurs
are dependent on activational eﬀects of estradiol and testosterone [49,50].
Finally, there seem to be sex-associated diﬀerences in hippocampal
reactivity, which has profound eﬀects for any changes in behavior that
have a basis in learning. Baseline excitability of hippocampal slices from
male mice is higher than for female mice, testosterone has an augmentative
eﬀect on ‘‘kindled’’ or stimulated generalized responses, and both aspects
are decreased by castration [51]. The ﬁndings warrant careful consideration
if neutering and behavior modiﬁcation are commonly recommended tools of
the trade. In this case, neutering may lower reactivity via the excitatory
eﬀect on hippocampal tissue. Although there may be a subsequent slower
eﬀect for routine behavior modiﬁcation, such a decrement may be more
than compensated for by avoiding the reinforcement of impulsive reactive
behaviors.

Those wishing to understand the complex interaction of genomic,

molecular, neurochemical, and neuroanatomic mechanisms that inﬂuence

226

K.L. Overall / Vet Clin Small Anim 33 (2003) 213–229

suites of behaviors would do well to consider these subtle hormonal eﬀects
in addition to the more straightforward eﬀects of genomic sex determi-
nation. This is especially true given that hormonal actions work in concert
with nuclear proteins. These nuclear proteins or cofactors help to regulate
receptor transcription activity at the subcellular level [50]. The relevance for
the pathologic ﬁndings of behavioral conditions that are largely treated
using pharmacologic agents that aﬀect receptor function and transcription
(eg, all TCAs and selective serotonin re-uptake inhibitors [SSRIs]) could
not be greater.

Summary

Boundaries between behavioral conditions and medical diﬀerentials are

likely to blur more rather than less as we learn more about genomic, cellular,
and subcellular eﬀects on common conditions. These changes should lead
to better treatment but may also require a paradigm shift in how we view
behavioral conditions and the mechanisms that contribute to them.

References

[1] Duman RS. Novel therapeutic approaches beyond the serotonin receptor. Biol Psychiatry

1998;44:324–35.

[2] Duman RS, Heninger GR, Nestler EJ. A molecular and cellular theory of depression. Arch

Gen Psychiatry 1997;54:597–606.

[3] Overall KL. Dogs as ‘‘natural’’ models of human psychiatric disorders: assessing validity

and understanding mechanism. Prog Neuropsychopharmacol Biol Psychiatry 2000;24:
727–76.

[4] New JC, Salman MD, King M, Scarlett JM, Kass PH, Hutichison JM. Characteristics of

shelter-relinquished animals and their owners when compared with animals and their
owners in US pet-owning households. J Appl Anim Welfare Sci 2000;3:179–201.

[5] Salman MD, Hutchison J, Ruch-Gallie R, et al. Behavioral reasons for relinquishment of

dogs and cats to 12 shelters. J Appl Anim Welfare Sci 2000;3:93–106.

[6] Scarlett JM, Salman MD, New JG, Kass PH. The role of veterinary practitioners in

reducing dog and cat relinquishments and euthanasias. JAVMA 2002;220:306–11.

[7] Shore ER, Girrens K. Characteristics of animals entering an animal control or humane

society shelter in a midwestern city. J Appl Anim Welfare Sci 2001;4:105–16.

[8] Huang Z-L, Qu W-M, Li W-D, et al. Arousal effect of orexin A depends on activation of

the histaminergic system. Proc Natl Acad Sci USA 2001;98:9965–70.

[9] Hungs M, Fan J, Lin L, et al. Identiﬁcation and functional analysis of mutations in the

hypocretin (orexin) genes of narcoleptic canines. Genome Res 2001;11:531–9.

[10] Mignot E. A commentary on the neurobiology of the hypocretin

/orexin systems.

Neuropsychopharmacology 2001;25(Suppl):S5–15.

[11] Greer JM, Capecchi MR. Hoxb8 is required for normal grooming behavior in mice.

Neuron 2002;33:23–34.

[12] Nestadt G, Lan T, Samuels JF, et al. Complex segregation analysis provides compelling

evidence for a major gene underlying obsessive-compulsive disorder (OCD) and
heterogeneity by gender. Am J Hum Genet 2000;67:1611–6.

227

K.L. Overall / Vet Clin Small Anim 33 (2003) 213–229

[13] Harvey BH, Brink CB, Seedat S, Stein DJ. Deﬁning the neuromolecular action of myo-

inositol: application to obsessive-compulsive disorder. Prog Neuropsychopharmacol Biol
Psychiatry 2002;26:21–32.

[14] Overall KL. Clinical behavioral medicine for small animals. St. Louis: Mosby; 1997. p. 503–9.
[15] Overall KL. Pharmacological treatment in behavioral medicine: the importance of

neurochemistry, molecular biology, and mechanistic hypotheses. Vet J 2001;62:9–23.

[16] Overall KL, Dunham AE, Frank D. Frequency of nonspeciﬁc clinical signs in dogs with

separation anxiety, thunderstorm phobia, and noise phobia, alone or in combination.
JAVMA 2001;219:467–73.

[17] Aronson LP. Systemic causes of aggression and their treatment. In: Dodman NH,

Shuster L, editors. Psychopharmacology of animal behavior disorders. Malden, MA:
Blackwell Science; 1998. p. 64–102.

[18] Reisner I. The pathophysiologic basis of behavior problems. Vet Clin North Am Small

Anim Pract 1991;21:207–24.

[19] Overall KL, Dunham AE. Phenomenology, breed. age of onset, and outcome in dogs and

cats with obsessive-compulsive disorder. JAVMA 2002;221:1445–52.

[20] Beaver BV. Clinical classiﬁcation of canine aggression. Appl Anim Ethol 1983;10:35–43.
[21] Martin KM, Rossing MA, Ryland LM, et al. Evaluation of dietary and environmental risk

factors for hyperthyroidism in cats. JAVMA 2000;217:853–6.

[22] Scarlett JM. Epidemiology of thyroid diseases of dogs and cats. Vet Clin North Am Small

Anim Pract 1994;24:477–86.

[23] Scarlett JM, Moise NS, Rayl J. Feline hyperthyroidism: a descriptive and case-control

study. Prev Vet Med 1998;6:295–309.

[24] Overall KL, Dunham AE. Effects on renal and hepatic values of long-term treatment with

tricyclic antidepressants and selective serotonin re-uptake inhibitors. Proceedings of
AVSAB Annual Symposium of Animal Behavior Research 2002:73–6.

[25] Haggerty JJ, Prange AJ. Borderline hypothyroidism and depression. Annu Rev Med

1995;46:37–46.

[26] Gold MS, Pottash ALC, Extein IL. ‘‘Symptomless’’ autoimmune thyroiditis in depression.

Psychiatr Res 1988;6:261–9.

[27] Nemeroff CB, Simon JS, Haggerty JJ, Evans DL. Antithyroid antibodies in depressed

patients. Am J Psychiatry 1985;142:163–72.

[28] Loosen PT, Prange AJ. Serum thyrotropin response to thyrotropin-releasing hormone in

psychiatric patients: a review. Am J Psychiatry 1982;139:405–16.

[29] Bauer MS, Whybrow PC. Thyroid hormones and the central nervous system in affective

illness: interactions that may have clinical signiﬁcance. Integr Psychiatry 1988;6:75–100.

[30] Whybrow PC. Sex differences in thyroid axis function: relevance to affective disorder and

its treatment. Depression 1995;3:33–42.

[31] Peterson ME, Melian C, Nichols R. Measurement of serum concentrations of free

thyroxine, total thyroxine, and total triiodothyronine in cats with hyperthyroidism and
cats with nonthyroidal disease. JAVMA 2001;218:529–36.

[32] Panciera DL. Hypothyroidism in dogs: 66 cases (1987–1992) JAVMA 1994;204:761–7.
[33] Ma¨nnisto¨ PT, Peuranen E, Harro J, Vasar E. Possible role of cholecystokinin-A receptors

in regulation of thyrotropin [TSH] secretion in male rats. Neuropeptides 1992;23:251–8.

[34] Joffe RT, Sokolov STH. Thyroid hormones, the brain, and affective disorders. Crit Rev

Neurobiol 1994;8:45–63.

[35] Rahaman SO, Ghosh S, Mohanakuma KP, et al. Hypothyroidism in the developing rat

brain is associated with marked oxidative stress and aberrant intraneuronal accumulation
of neuroﬁlaments. Neurosci Res 2001;40:273–9.

[36] Smith JW, Evans AT, Costal B, Smythe JW. Thyroid hormones, brain function and

cognition: a brief review. Neurosci Biobehav Rev 2002;26:45–60.

[37] Forstad I, Kanter AJ. Parasites, bright males, and the immunocompetence handicap. Am

Nat 1992;139:603–22.

228

K.L. Overall / Vet Clin Small Anim 33 (2003) 213–229

[38] Forstad I, Nilssen AC, Halvorsen O, Andersen J. Why do male reindeer (Rangifert

tarandus

) have higher abundance of social and third instar larvae of Hypoderma tarandi

than females? Oikos 1989;55:87–92.

[39] Zuk M. Reproductive strategies and sex differences in disease susceptibility: an

evolutionary viewpoint. Parasitol Today 1990;6:231–3.

[40] Zuk M, Hohnsen TS, Maclarty T. Endocrine-immune interactions, ornaments and mate

choice in red jungle fowl. Proc R Soc Lond B Biol Sci 1995;260:205–10.

[41] Zuk M, McKean KA. Sex differences in parasite infections: patterns and processes. Int

J Parasitol 1996;26:1009–24.

[42] Zuk M, Thornhill R, Lignon JD, Johnson K. Parasites and mate choice in red jungle fowl.

Am Zool 1990;30:235–44.

[43] Klein SL. The effects of hormones on sex differences in infection: from genes to behavior.

Neurosci Biobehav Rev 2000;24:627–38.

[44] Hoffman-Lehmann R, Holznagel E, Lutz H. Females cats have lower rates of apoptosis in

peripheral blood lymphocytes than male cats: correlation with estradiol-17B, but not with
progesterone blood levels. Vet Immunol Immunopathol 1998;65:151–60.

[45] Cases O, Seif I, Grimsby J, et al. Aggressive behavior and altered amounts of brain

serotonin and norepinephrine in mice lacking MAOA. Science 1995;268:1763–6.

[46] Palanza P. Animal models of anxiety and depression: how are females different? Neurosci

Biobehav Rev 2001;25:219–33.

[47] Overall KL, Beebe AD. Dominance aggression in young female dogs: what does this

suggest about the heterogeneity of the disorder? In: Proceeding of the First International
Conference on Veterinary Behavior Medicine. Potter’s Bar, UK: Universities Federation
for Animal Welfare; 1997. p. 58–63.

[48] Andersen SJ, Teicher MH. Sex differences in dopamine receptors and their relevance to

ADHD. Neurosci Biobehav Rev 2000;24:1137–41.

[49] Arnold AP, Breedlove SM. Organizational and activational effects of sex steroids on brain

and behavior: a reanalysis. Horm Behav 1985;19:469–98.

[50] Kawata M, Matsuda K, Nishi M, et al. Intracellular dynamics of steroid hormone

receptor. Neurosci Res 2001;40:197–203.

[51] Smith MD, Jones LS, Wilson MA. Sex differences in hippocampal slice excitability: role of

testosterone. Neuroscience 2002;109:517–30.

229

K.L. Overall / Vet Clin Small Anim 33 (2003) 213–229

Behavioral dermatology

Vint Virga, DVM

Behavioral Medicine for Animals, Veterinary Healing Arts, Inc., PO Box 219,

Newport, RI, 02840–0219, USA

The practice of behavioral dermatology encompasses the management of

any dermatologic condition for which there is a substantive behavioral
or emotional component. It is estimated that emotional factors are
an important consideration in the eﬀective management of at least one
third of human patients presenting for dermatologic conditions [1,2]. It is
reasonable to assume that a similar incidence occurs in companion animal
species. Integral to the clinical incidence of behavioral dermatoses is an
underlying neurophysiologic basis for psychodermatologic conditions. Both
the central nervous system (CNS) and integument are derived from the
embryonic ectoderm. Although the CNS and skin are diﬀerentiated during
fetal development, a substantial number of hormones, neuropeptides, and
receptors are common to both [46]. Physiologic changes that can be
clinically observed, such as pruritus, vascular ﬂushing, and sweating, are
related to underlying actions of psychoneuroendocrinoimmunologic medi-
ators, such as enkephalins, endorphins, substance P, and vasoactive
intestinal peptide, on the CNS, integument, and immune system [4–6].

The skin is the primary organ of tactile stimulation and communication

in the body. Underscoring the integral relation between the integument
and behavior is the ﬁnding that cutaneous contact and stimulation during
postnatal development may substantially inﬂuence cell growth and diﬀer-
entiation, CNS maturation, neurosensory responses, immune function, and
the incidence of behavioral problems (eg, aggression and anxiety-related
behaviors) in a variety of species [6–10]. Meriting further consideration,
a number of forms of psychotherapy (eg, biofeedback, hypnosis, cognitive
therapy) have been used eﬀectively to manage human patients with a num-
ber of dermatologic disorders [11–14]. Together, these ﬁndings present a com-
pelling picture of the intrinsic relation between the brain, emotions, and skin.

Vet Clin Small Anim

33 (2003) 231–251

E-mail address:

vvirga@att.net

0195-5616/03/$ - see front matter

PII: S 0 1 9 5 - 5 6 1 6 ( 0 2 ) 0 0 1 0 2 - X

Using the framework established for human psychodermatology,

behavioral dermatoses in animals may be classiﬁed into four categories
of phenomena: psychophysiologic disorders, primary behavioral disorders,
secondary behavioral disorders, and cutaneous sensory disorders [15,16].
Although these groupings may provide a useful framework for discussion
and clinical management, individual patients may present with features
characteristic of more than one of these categories of psychodermatologic
disease.

Psychophysiologic disorders

Psychophysiologic disorders are primary dermatologic conditions that

can be aﬀected by emotional stress. A substantial number of conditions,
particularly chronic dermatoses (eg, acne, urticaria, atopic dermatitis,
psoriasis, rosacea, seborrheic dermatitis), that meet this criterion have been
identiﬁed in human patients [15–17]. Of primary concern in animals, con-
ditions in this category include atopic dermatitis, chronic inﬂammatory
dermatoses, and acral lick dermatitis (ALD). Although many cases of ALD
respond to standard dermatologic treatment, some refractory cases should
be considered as primary behavioral disorders.

With psychophysiologic disorders, the activation of psychoneuroendo-

crinoimmunologic mediators may contribute to an exacerbation of clinical
signs. By means of vasoactive mediators, emotional stress can precipitate
or perpetuate the ‘‘itch-scratch cycle’’ associated with psychophysiologic
disorders [18]. It has been documented clinically and physiologically that
emotional stress can trigger or exacerbate pruritus by activating the release
of neuropeptides distributed to the skin by means of hypophyseal portal
vessels and peripheral circulation as well as by descending autonomic ﬁbers
[19,20]. Sensory nerves can act not only aﬀerently, conveying sensory input
to the CNS, but eﬀerently in a neurosecretory fashion. Thus, neuropeptides
mediating a behavior along CNS pathways may also contribute sensation
(eg, pruritus, pain) and manifestation of a behavior (eg, scratching, licking,
biting) peripherally [4].

The physiologic sensation of pruritus may share common biochemical

origins with some anxiety states, which supports consideration of neuro-
psychodermatologic etiologies [21]. An association between atopic derma-
titis and emotional reactivity has been well established in human beings.
Reduced coping strategies, irritability, hostility, depression, anxiety, and
aggression are common concomitant behaviors in human atopic patients
[4,13,22]. Excitability and inadequate stress-coping skills, particularly,
correlate with elevated serum IgE levels in atopic patients [22]. Not only
are certain behaviors more likely to manifest in atopic patients, but certain
behavioral characteristics may predispose individuals to atopic dermatitis.
In human beings, a well-documented relation has been demonstrated to
exist between maternal rejection and atopic dermatitis [4].

232

V. Virga / Vet Clin Small Anim 33 (2003) 231–251

Based on psychophysiologic research to date, it is likely that veterinary

patients with clinical atopic disease and other chronic inﬂammatory diseases
may be predisposed to behavioral sequelae, particularly reduced coping strat-
egies and increased reactivity, anxiety, and aggression. Stressors (environ-
mental or social) may contribute to the exacerbation of clinical signs or, in
a quiescent state, may lead to the recurrence of clinical signs. Thus, environ-
mental or social stress may contribute toward a patient reaching the
pruritic threshold in the same manner as speciﬁc antigenic stimuli. Fur-
thermore, chronic stress may exacerbate dermatologic management by ef-
fectively bringing the patient closer to the pruritic threshold. Concurrent
behaviors that may be elicited in response to stress include changes in
appetite, grooming behaviors, elimination patterns, social interaction, and
activity [23]. Potential environmental and social stressors that should be
considered are as follows:

Inadequate mental stimulation
Inadequate aerobic exercise
Inadequate interaction with family or other pets
Limited access to essential resources
Social isolation
Status-related conﬂicts
Territorial-related conﬂicts
Addition or loss of family members or pets
Changes in health status of family members or pets
Changes in daily routine of family members or pets
New home/environment
Changes in physical environment
Boarding
Hospitalization

Primary behavioral disorders

Conditions for which the primary problem is related to behavior in

nature and secondary skin manifestations are self-induced are classiﬁed
as primary behavioral disorders. In human beings, primary psychiatric
disorders presenting as dermatologic complaints include delusional behav-
iors (eg, delusions of parasitosis), obsessive-compulsive disorder (OCD,
including some cases of trichotillomania and neurotic excoriations),
dermatitis artefacta, dysmorphic disorder, and psychogenic pruritus. Pri-
mary behavioral disorders in animals include ALD, compulsive behav-
iors, excessive grooming behaviors (psychogenic alopecia), and psychogenic
pruritus.

Self-injurious behavior (SIB) refers to any volitional behavior that results

in self-mutilation or damage [24]. A diagnosis of SIB in small animals must
meet the criteria of barbering or removal of hair and

/or abrasion;

233

V. Virga / Vet Clin Small Anim 33 (2003) 231–251

petechiation; or ulceration of any body part using the teeth, tongue, claws,
or an external substrate (eg, rubbing against a wall). A condition for a diag-
nosis of SIB is that these behaviors must be demonstrated repeatedly and
consistently in the absence of any primary dermatologic or physiologic
condition [24]. It is important to recognize that a diagnosis of SIB does
not necessitate or imply that the behavior is compulsive, only that it
is volitional.

The essential features of OCD in human beings are recurrent obsessions

or compulsions that are severe enough to be time-consuming or cause
marked distress or signiﬁcant impairment. Obsessions are deﬁned as per-
sistent ideas, thoughts, impulses, or images that are experienced as intru-
sive and inappropriate and that cause marked anxiety or distress. Common
obsessions reported in people include repeated thoughts about contami-
nation (eg, by shaking hands), speciﬁc repeated doubts (eg, having left a
door unlocked), a need to have things in a particular order (eg, intense
distress from disorder or asymmetry), and aggressive or horriﬁc impulses.
These thoughts, impulses, and images exceed simple worries about real-life
problems and are unlikely to be related to a real-life problem. Compulsions
in human beings are deﬁned as repetitive behaviors (eg, hand washing,
checking) or mental acts (eg, praying, counting), the goal of which is to
prevent or reduce anxiety or distress rather than to provide pleasure or
gratiﬁcation [25].

In veterinary behavioral medicine, sequences of movements that serve

no obvious purpose or function and occur repetitively, out of context, at
an excessive frequency or duration exceeding that necessary to achieve a
real or potential goal, and in a relatively unvaried fashion are termed
stereotypic behaviors

[23,26]. In most cases, they are derived from behaviors

that are part of the animal’s normal behavioral repertoire. Most stereotypic
behaviors are not compulsive in nature. To establish a diagnosis of
a compulsive disorder, the behavior must interfere with the patient’s abil-
ity to function normally in its social environment [26,27]. Because any evi-
dence of obsessive behavior in nonhuman species is problematic and
must be inferred, the term compulsive disorder more accurately describes
these behaviors in our companion animal species. As in human medicine,
a compulsive disorder should be considered a manifestation of an anxiety
disorder. In the author’s experience, most patients referred with suspected
compulsive behaviors do not meet the criteria for a compulsive disorder.
Accurate data denoting the incidence of compulsive disorders or the relative
percentage of compulsive SIBs in the general canine or feline population are
not currently available.

Considering these criteria, it is evident that some patients presenting

to the small animal practitioner may meet the conditions for both SIB
and compulsive disorder. Compulsive behaviors associated with dermato-
logic signs are most commonly classiﬁed as grooming compulsive disorders,
although some may be neurotic in origin. In canine patients, these may

234

V. Virga / Vet Clin Small Anim 33 (2003) 231–251

include ALD

/granuloma, ﬂank sucking, tail chewing (which may or may not

be associated with tail chasing), excessive chewing of the feet and

/or nails, and

excessive scratching. Other compulsive behaviors observed in canine pa-
tients may be classiﬁed as hallucinatory (eg, ﬂy

/light chasing, prey search-

ing, staring), locomotor (eg, circling, tail chasing, fence running), eating

drinking (eg, fabric sucking, psychogenic polydipsia, some picas), vocal (eg,
rhythmic barking, barking at food or inanimate objects), or neurotic (eg,
vicious self-biting, spontaneous aggression to human beings) [23,26]. In
feline patients, compulsive behaviors associated with grooming include
psychogenic dermatitis, feline hyperesthesia syndrome, tail sucking, and
excessive chewing of the feet and

/or nails. As with canine patients, other

compulsive behaviors noted in feline patients may be categorized as
hallucinatory (eg, prey chasing or searching, air batting), locomotor (eg,
paw shaking, head shaking, pacing), vocalization (eg, repetitive howling

crying), or neurotic (eg, vicious self-biting, spontaneous aggression to hu-
man beings) [23].

Independent of compulsive disorders, anxiety may lead to the develop-

ment of a variety of other behavioral dermatoses. Anxiety may be deﬁned as
an apprehensive anticipation of future danger or misfortune accompanied
by a feeling of dysphoria and

/or somatic symptoms of tension [24]. Anxieties

may be internally or externally focused and may be in response to real
or perceived stimuli. Anxiety may result from motivational states of conﬂict
(the tendency to simultaneously perform more than one type of activity)
or frustration (engagement in a sequence of behaviors that cannot be
completed because of physical or psychologic obstacles) [23]. In human
psychiatry, anxiety disorders are subcategorized into a number of discrete
diagnoses (eg, OCD, agoraphobia, posttraumatic stress disorder, general-
ized anxiety disorder) [25].

Nonspeciﬁc behaviors directed toward speciﬁc body parts that may be of

psychogenic origin include tail biting, ﬂank sucking, preputial licking, self-
nursing, licking in the anal region, and foot licking. Based on the evidence
to date, these behavioral patterns represent a heterogenous array of under-
lying conditions rather than speciﬁc dermatologic or behavioral diagno-
ses. Attention-seeking, displacement, self-injurious, compulsive, and other
anxiety-related behaviors may lead to the establishment of such behavioral
patterns. Seizure activity involving the amygdala and ventromedial hypo-
thalamus can result in stereotypic self-directed behaviors [24].

Canine acral lick dermatitis (acral lick granuloma)

ALD is characterized by ﬁrm, raised, ulcerative plaques preceded by

erosion of the skin secondary to chronic or intense licking. Most cases
present with single unilateral lesions on the cranial carpus or metacarpus.
Additional sites include the cranial radius, metatarsus, and tibia [28,29].
Secondary bacterial infection is common, may contribute to these lesions

235

V. Virga / Vet Clin Small Anim 33 (2003) 231–251

being intensely pruritic in nature, and may necessitate prolonged anti-
microbial therapy. As noted previously, ALD may be dermatologic or
psychogenic in origin. It has been estimated that psychogenic and idiopathic
ALD combined may comprise as many as 50% of cases [59]. A strong
association seems to exist between licking and anxiety in dogs. It has been
reported that as many as 70% of dogs diagnosed with ALD have concurrent
fear- and

/or anxiety-based conditions (eg, noise phobia, separation anxiety,

anxiety-related aggression) [31]. It is not uncommon to ﬁnd evidence of
salivary staining of the carpi in dogs presented for signs of separation
anxiety. As with overgrooming in cats, ALD may also be associated with
displacement grooming in response to social or environmental stressors.
Other possible psychogenic associations include attention seeking; compul-
sive behavior; and limited opportunity for social interaction, environmental
stimulation, or aerobic activity.

The occurrence and incidence of correlative behaviors to ALD in

feral and wild dogs is not known. Among domestic dogs, certain breeds
seem to be overrepresented (Labrador Retrievers, Great Danes, Doberman
Pinchers, German Shepherds, and some northern breeds), with some evi-
dence of familial inheritance [29,31,32]. This may be reﬂective not only of a
genetic component but of selection pressures placed on these breeds for
selected work duties and social relationships with human beings.

Psychogenic alopecia (excessive grooming)

Psychogenic alopecia is characterized by excessive self-grooming that is

initiated or intensiﬁed by nonorganic causes or persists beyond resolution
of an organic cause [30]. Although cases are more typically identiﬁed in
feline patients, it is possible for psychogenic alopecia to occur in dogs in
the absence of discrete lesions of ALD. The predominant clinical sign is
alopecia, particularly in the area of the medial forelegs, caudal abdomen,
inguinal region, tail, and

/or dorsal lumbar areas, in which medical causes

have been ruled out. Because cats may groom reclusively, excessive lick-
ing, biting, scratching, or rubbing may or may not be observed by the client.
Barbering and

/or frank alopecia may be the only dermatologic sign.

In other cases, self-mutilation with possible secondary bacterial infection
may be evident. Symmetric alopecia of the caudomedial thighs and
ventrum is possible. Licheniﬁcation and hyperpigmentation may develop
in chronic cases. A dermatitic form (atypical neurodermatitis) can develop
from abnormally prolonged or intense grooming behavior. In the der-
matitic form, bright red and elongated oval streaks or plaques may result
[28].

Physical examination reveals (1) short broken hairs that are readily

palpated by stroking the aﬀected area against the normal angle of hair
growth, (2) remaining hairs that do not epilate easily, (3) broken hair shafts
detected by microscopic evaluation, (4) hair regrowth that occurs normally

236

V. Virga / Vet Clin Small Anim 33 (2003) 231–251

and lesions that heal with placement of an Elizabethan collar, and (5) sig-
niﬁcant amounts of hair on fecal examination [28,29].

Licking of the hair and skin, nibbling, biting, facial rubbing with the

forepaws, and scratching may all be observed in cats exhibiting normal
grooming behavior. Although they regularly self-groom, speciﬁc times and
percentages relative to other behaviors are not well-documented in house-
hold cats [31,33]. Studies in farm cats and colony-housed domestic cats
have reported grooming to occupy 15% and 4% of the total time budget,
respectively [34,35]. Beyond such basic purposes as cleansing, removal of
parasites, and thermoregulation, grooming in cats may occur as a displace-
ment behavior (an activity that is performed out of context as a result of
conﬂict or frustration) in response to social or environmental stressors.

Displacement grooming may be rooted in anxiety and may serve to

lower arousal, deﬂect aggression from other individuals, or provide some
distraction for the cat [23,24]. Although the occurrence of such behavior
in feral or wild cat species is not known, incidences of psychogenic alopecia
have been noted in captive wild cats [36]. There is some evidence to date that
psychogenic alopecia occurs in captive wild cats secondary to inadequate
environmental or social enrichment (V. Virga, unpublished data). Similarly,
it is possible that domestic cats with limited environmental or social
stimulation may display excessive grooming. Psychogenic alopecia is re-
ported to be more prevalent in strictly indoor cats [31]. A seasonal inci-
dence, even in indoor cats, can result from changes in environmental and social
stressors (eg, accessibility

/visibility of other cats) [37].

Self-directed attention-seeking behavior

In the author’s experience, a signiﬁcant percentage of cases referred for

evaluation of compulsive or SIBs are ultimately diagnosed as attention-
seeking behaviors. Animals can readily learn that not only disruptive
behaviors (eg, barking, jumping, pawing, nuzzling) but less directly de-
manding behaviors (eg, limb

/foot/preputial licking, chewing, scratching,

sucking, pawing) often eﬀectively get the client’s attention. The clients may
have historically tried a variety of approaches to discourage such behaviors;
such attempts often include varying degrees of physical and verbal cor-
rections, comforting the patient with physical touch and verbal reassur-
ances, banishment with physical and social isolation, and ignoring the
behavior to varying degrees. As the animal persists in the behavior, clients
typically report that they eventually provide some form of attention; in
so doing, the behavior can quite eﬀectively be reinforced as a result of
a variable and high ratio-reward schedule [38]. It is important for the client
to recognize that verbal or physical punishment, particularly if intermittent,
may serve to reinforce the problem behavior.

In establishing a diagnosis of attention-seeking behavior, a careful review

of the history should reveal that the patient demonstrates the problem

237

V. Virga / Vet Clin Small Anim 33 (2003) 231–251

behavior only in the immediate presence or close proximity of the client.
Observation of the patient at the time of consultation should reveal that the
behavior is dramatically reduced or nonexistent when the client (or, in some
cases, all parties, including the clinician) is absent.

Psychogenic pruritus

Pruritus is deﬁned as an unpleasant sensation that provokes the desire to

scratch [39]. Typically, the term pruritus is used synonymously with itching.
Although it is difﬁcult to determine if dogs are scratching because of a
sensation of itching, by deﬁnition, they experience pruritus by virtue of some
stimulus that initiates scratching. It has been well established that human
beings may experience psychogenic pruritus (ie, pruritus in the absence of
primary dermatologic lesions or signiﬁcant metabolic, endocrine, neuro-
logic, or other medical ﬁndings) [17,18,40]. Based on similar psychoneu-
roimmunoendocrinologic mediators and physiologic mechanisms, it is
reasonable to assume that psychogenic pruritus may occur in animals.

A diagnosis of psychogenic pruritus in animals must be made after

exclusion of other potential causes for scratching. In assessing an animal
for pruritus in the absence of dermatologic lesions, it is important to
identify any contexts, environments, social interactions, or temporal as-
sociations with the incidence of pruritus. As noted earlier, scratching (or
other self-directed behaviors) occurring only in the presence of selected
individuals is strongly suggestive of attention-seeking behavior. Scratching
limited to certain speciﬁc contexts or in the presence of selected stimuli
may be a manifestation of displacement behavior secondary to conﬂict or
frustration.

Secondary behavioral disorders

Secondary behavioral disorders may result from dermatologic conditions

that adversely aﬀect the normal behavioral patterns and social function-
ing of the animal. It is well documented that pruritus in human beings
can signiﬁcantly inﬂuence emotional reactivity and sleep patterns [41,42].
Furthermore, anxiety and depression have been well substantiated as
common concomitants of pruritus [19,40,43]. Although secondary psychi-
atric disorders in human beings are frequently associated with emotional
responses secondary to disﬁguring conditions, it is diﬃcult to determine if
animals suﬀer similarly from disﬁgurement. Based on sensory stimuli (eg,
pain, pruritus, hyperesthesia, allodynia) typically associated with a variety
of dermatologic lesions, it is likely that the behavioral patterns and social
functioning of animals may be aﬀected. As with stressful environmental
and social stimuli, sensory stimuli associated with pain or discomfort may
result in reduced coping strategies and increased reactivity, anxiety, or
aggression.

238

V. Virga / Vet Clin Small Anim 33 (2003) 231–251

Cutaneous sensory disorders

Cutaneous sensory disorders are conditions for which the patient

experiences a purely sensory complaint without clinical evidence of a
dermatologic, neurologic, or medical condition. By deﬁnition, the patho-
genesis is typically not identiﬁable and diagnostic tests prove unremarkable.
Humans with cutaneous sensory disorders most commonly report sensa-
tions of stinging, burning, or itching [44]. The persistence of clinical signs in
the absence of signiﬁcant diagnostic tests suggests an underlying neuro-
pathic phenomenon. The pathologic ﬁndings and clinical signs may be
generalized or speciﬁc to certain body parts [44,45].

Cases may present with no diagnosable behavioral problem or with

signiﬁcant behavioral ﬁndings that may or may not be related to the
cutaneous sensory disorder. It is well documented that some human patients
with depression and anxiety disorders perceive pain or itching sensations in
an exaggerated manner. As such, these cases could be considered primary
psychiatric disorders with a secondary neurosensory component; however,
among patients with no diagnosable psychiatric ﬁnding, there is suﬃcient
evidence to consider that these patients have a primary neurosensory
disorder. In the absence of signiﬁcant clinical ﬁndings, the clinician is
dependent on the patient’s description of the phenomena he or she ex-
periences [44]. As such, although it is problematic to conﬁrm a diagnosis in
animals, cutaneous sensory disorder should be carefully considered if the
patient exhibits (1) a response as if experiencing pain to nonnoxious stimuli,
such as touch, mild pressure, mild heat, or mild cold (ie, allodynia); (2) a
markedly exaggerated response to a stimulus that is typically painful (ie,
hyperalgesia); (3) behaviors that subjectively seem to be in response to or
avoidance of an unpleasant stimulus (ie, dysesthesia); or (4) excessive self-
directed behavior (eg, grooming, licking, biting, chewing, rubbing, scratch-
ing, rolling) of an intensity suggesting that the animal may be responding to
a sensory stimulus. The behavior may directed toward speciﬁc body parts or
generalized to multiple foci.

Feline hyperesthesia syndrome

Feline hyperesthesia syndrome refers to a complex of behaviors that may

include (1) behaviors similar to those observed in estrous females (eg,
increased motor activity, rolling, crouching with elevation of the perineal
region, vocalizations); (2) excessive licking, plucking, biting, and

/or chew-

ing, particularly at the tail, ﬂank, anal, or lumbar areas; and (3) rippling
of the skin, muscle spasms, or twitches (especially dorsally), which may be
accompanied by vocalization, running, jumping, possible hallucinations, or
self-directed aggression [23,24]. Aﬀected cats tend to be diﬃcult to distract
from the behavior or, if successfully distracted, remain so for only a short
period. The occurrence of signs may be episodic in nature, accompanied by
variable periods of unremarkable behavior.

239

V. Virga / Vet Clin Small Anim 33 (2003) 231–251

As with feline psychogenic alopecia, environmental and social stressors

have been associated with this disorder. The cues or changes precipitating
the behavior may be exogenous or endogenous [24]. Cats may present with
clinical signs consistent with hyperesthesia or allodynia without evidence of
alopecic or other dermatologic lesions. Review of the behavioral history
may further support a lack of excessive grooming. Such cases support the
hypothesis that this complex of behaviors may represent a number of
discretely diﬀerent phenomena. Although not currently discussed in the
veterinary literature, based on clinical syndromes observed in human
patients, it may be worthwhile to consider hallucinatory, rheumatologic, or
neurogenic origins in future research.

Clinical management

Diagnostic approach

Because of the heterogenous and potentially multifactorial origins of self-

directed behaviors, clinical evaluation should include a thorough clinical
examination (general, dermatologic, and neurologic), complete blood cell
count (CBC), serum chemistry proﬁle, and urinalysis. Any potentially sig-
niﬁcant clinical ﬁndings should be pursued by appropriate laboratory and
diagnostic tests so as to rule out any organic causes before considering
primary behavioral disorders. Patients presenting with clinical signs sug-
gestive of psychogenic alopecia or a cutaneous sensory disorder of the
ventral abdomen should be evaluated for lower urinary tract disease. Pa-
tients presenting with clinical signs suggestive of psychogenic alopecia or a
cutaneous sensory disorder of the perianal region should be evaluated for
anal sac disease or distension.

Psychophysiologic, primary behavioral, secondary behavioral, and

neurosensory disorders should carefully be considered even with signiﬁcant
dermatologic or neurologic ﬁndings. This is particularly true for patients
with a history of being refractory to standard courses of treatment. Behav-
ioral evaluation should incorporate a review of the behavioral history and
direct observation of the patient, including consideration of environ-
mental stimuli, social stimuli, the motivational state of the animal, and
underlying neurophysiologic mechanisms in developing a treatment plan.
Important considerations to consider in the behavioral history include the
following:

Detailed description of the patient’s behavior immediately before,

during, and after eliciting problem behavior

Chronology, incidence, and progression of problem behavior
Ease with which problem behavior may be interrupted and tendency for

return to behavior

Locations, circumstances, and potential eliciting stimuli associated with

the problem behavior

240

V. Virga / Vet Clin Small Anim 33 (2003) 231–251

Review of other problem behaviors
Review of home environment, including all persons and animals in

household

Presence of the client(s), other people, and other animals in relation to

animal when behavior occurs

Responses of the client(s), other people, and other animals in relation to

the problem behavior

Patient’s background, including adoption source, familial history, early

temperament

/behavior of patient, and history of obedience work

Interactions with familiar and unfamiliar household guests
Dietary history, including consideration of who feeds patient and review

of feeding schedule

Daily routine of patient in relation to other human and animal members

of household

Speciﬁc types, amount, and frequency of exercise
Speciﬁc form, duration, and frequency of interaction with client(s) and

other people

Notation of sleeping location and favorite resting places
Review of medical history with notation of any current medications

being administered

Although a videotape of the patient’s behavior does not preclude direct

observation, it can provide valuable clues toward accurate diagnosis, be-
cause the cat can be observed in its home environment. A hypothesis in-
corporating the above, which can account for the patient’s dermatologic and
behavioral manifestations, provides a rational starting point from which
a program of environmental, behavioral, and pharmacologic management
can be developed.

Environmental management

Because the patient’s response to its environment may contribute to the

establishment of behavioral dermatoses, it is important to manipulate the
environment so as to eliminate or reduce exposure to stressors. If this is
not possible, counterconditioning and systematic desensitization should be
used to minimize the response to provocative environmental stimuli. Client
resistance to appropriate environmental changes may be encountered, and
creativity is often needed when presenting the management plan. The client’s
commitment to proposed environmental changes should be assessed before
determining behavioral and pharmacologic management.

Behavioral modification

Counterconditioning and desensitization provide the framework for

behavioral modiﬁcation. Counterconditioning consists of teaching the
patient new behaviors that are incompatible with the problem behavior.

241

V. Virga / Vet Clin Small Anim 33 (2003) 231–251

Because self-directed behaviors are often based on emotional states of stress,
anxiety, arousal, conﬂict, or frustration, behavioral modiﬁcation can serve
to minimize the stress response when the animal is exposed to these stimuli
[15,46]. It is often most eﬀective to select for behaviors that encourage
relaxation [12]. Desensitization consists of reinforcing the selected new be-
haviors while gradually introducing progressively more provocative cir-
cumstances and stimuli. In the author’s experience, clients commonly wish
to progress more rapidly than the patient can be expected to tolerate.
Patience, consistency, and commitment on the part of the client are critical
for success to support performance of the new behaviors in the face of
increasingly provocative stimuli. The willingness of the patient to follow the
leadership of the client can, in some cases, facilitate counterconditioning and
desensitization. Deference to the client can be established through routine
and regular reinforcement of leadership on a daily basis. Withdrawal of
attention is an eﬀective gentle correction for failure of the animal to follow
the client’s leadership. In all phases of counterconditioning and desensiti-
zation, appropriate responses should be supported with encouragement,
aﬀection, and small food rewards as positive reinforcement. Rewarding the
patient at any time when it is not exhibiting the problem behavior and is
relaxed can augment active counterconditioning. Massage therapy, when
the patient is relaxed, can further facilitate relaxation and encourage ap-
propriate interaction between the animal and the client. These techniques
are not limited to application in dogs but can be eﬀectively employed with
cats and other species with appropriate modiﬁcation.

Client responses to the patient, particularly when it is performing the

problem behavior, can be problematic. Despite the history and experience of
the problem behavior, the client should avoid expressing frustration in any
way in the presence of the patient. Doing so may reinforce or exacerbate
any anxiety that the animal may be experiencing. The client should further
avoid oﬀering any measure of comfort (verbal, physical, or emotional) to
the patient when it is exhibiting the problem behavior. Attention-seeking
behaviors are based on the response of the client or, in some cases, other
people. Attempts to distract the behavior or even aversive responses may be
rewarding to some animals and may reinforce the observed behavior.

Interactive activity and opportunities for aerobic exercise can be critical

components of behavioral modiﬁcation. Agility, ﬂy ball, and freestyle ex-
ercises provide dogs the opportunity for interactive aerobic activity with the
clients beyond traditional activities, such as ﬁeld work, sheep herding,
Frisbee tossing, ball retrieving, running

/jogging, hiking, and walking.

Interactive exercise can facilitate desensitization by exposing the dog to a
variety of potentially provocative stimuli while providing something else on
which to focus. Interaction with the dog in such activities also provides
something to which it can look forward, encourages mental and physical
agility, and serves to enhance the relationship between the dog and the
client. Additional opportunities for environmental and social enrichment for

242

V. Virga / Vet Clin Small Anim 33 (2003) 231–251

canine patients may include providing interactive devices and toys on a
variable rotating basis, massage therapy, social grooming, and social play
with other dogs as well as people.

Environmental enrichment for cats may provide opportunities for mental

and physical stimulation, serve as a form of distraction from potentially
provocative stimuli, and oﬀer some degree of control over their physical and
social environment. Opportunities for environmental enrichment for feline
patients may include providing access to elevated sites and window perches
on which to explore and rest; oﬀering substrates to encourage marking by
rubbing, rolling, and scratching; limiting free access to food and distributing
food and treats to encourage exploration; and providing an ample variety
of interactive devices and toys available on a variable rotating basis.
Opportunities for social enrichment for cats may include social play,
massage therapy, social grooming, and facilitating selected and interactive
behaviors through shaping with a secondary reinforcer (eg, clicker).

Pharmacologic support

When behavioral dermatoses are considered, it is likely that a hetero-

geneous array of neurophysiologic mechanisms involving dopaminer-
gic, serotonergic, GABA-ergic, and noradrenergic systems may be
involved in the manifestation of these disorders. Numerous clinical studies
and case reports have explored pharmacologic manipulation of these
neurotransmitter systems in patients with behavioral dermatoses with
varying results. Diﬀerences in responses to pharmacotherapy may be
reﬂective of individual variations in neuroanatomic and neurochemical
function. Therefore, it is important that the clinician consider the underlying
motivational state and possible neurochemical correlates when assigning
behavioral diagnoses and recommending pharmacologic and behavioral
management.

Pharmacologic support with anxiolytics may be necessary to eﬀect a

clinical response, particularly in cases where suﬃcient environmental and
social modiﬁcation is problematic. Even with pharmacologic support,
however, it is worth noting that behavioral and environmental manage-
ment may enhance the eﬃcacy of standard dermatologic treatments and
eﬀectively reduce the dosage regimen and duration of treatment with
medications [12,16]. Pharmacotherapeutic agents should be selected
speciﬁcally to address the motivational state of the patient and a proposed
underlying neurophysiologic mechanism of action.

Tricyclic antidepressants

Amitriptyline (Elavil) and doxepin (Adapin, Sinequan) are tricyclic

antidepressants (TCAs) that are used in human and veterinary medicine as
an anxiolytics. Both exert their primary clinical eﬀects by inhibiting the

243

V. Virga / Vet Clin Small Anim 33 (2003) 231–251

presynaptic reuptake of serotonin and norepinephrine to varying degrees.
Blocking norepinephrine and, to a lesser degree, serotonin reuptake
prolongs the inhibitory action of these neuropeptides, released from the
brain stem by pain-modulating systems, on spinal cord neurons involved in
transmitting pain [47–49]. Amitriptyline has been shown to function as a
partial antagonist of N-methyl-D-aspartate (NMDA) receptors, which may
become activated secondary to peripheral nerve injury or inﬂammation
[48,50]. By virtue of these properties, amitriptyline and doxepin can be
effective analgesics for neuropathic pain. Both compounds have substantial
antihistaminic properties brought about by their ability to block H

and

receptors. Amitriptyline equally affects H

and H

receptors, whereas

doxepin is much more selective for H

receptors [51]. Compared with other

antihistamines, doxepin has an afﬁnity for H

receptors 56 times that of

hydroxyzine and 775 times that of diphenhydramine. These antihistaminic
properties have made these TCAs useful in treating pruritic conditions
refractory to traditional antihistamines [50].

Both amitriptyline and doxepin eﬀectively block muscarinic cholinergic

receptors, which may result in anticholinergic side eﬀects. Additional
reported side eﬀects include increases in appetite, weight gain, sedation
(particularly in cats), gastrointestinal disturbances, anxiety, aggression,
alopecia, pruritus, urticaria, photosensitivity, potential cardiac conduction
disturbances, lowered seizure threshold, and a suggested role in sick
euthyroid syndrome at higher doses [23,51,52]. Contraindications may
include hepatic, renal, or cardiac disease. This class of drugs should not be
administered concurrently with monoamine oxidase inhibitors (MAOIs)
selegiline (Anipryl), or L-tryptophan. Amitriptyline has the distinct dis-
advantages, particularly in cats, of having a particularly bitter taste and a
narrow therapeutic index associated with a high rate of toxicity with over-
dose. By virtue of the mechanism of action of TCAs, it is important to allow
at least 3 to 4 weeks of TCA administration before evaluating the patient’s
clinical response at any given dosage. For patients in which dermatitis,
neuralgia, pain, or pruritus is suspected to be contributory to the patient’s
clinical signs, the antihistaminic, analgesic, and anxiolytic properties of
these compounds can prove to be quite eﬀective in managing clinical signs.

Clomipramine (Clomicalm) is a TCA that is relatively more serotonergic

and less anticholinergic than previously mentioned medications. Clomicalm
has been approved by the US Food and Drug Administration (FDA) for use
in dogs in the management of separation anxiety and may be an eﬀective aid
in the management of other anxiety-related behaviors. Clomipramine is
also the only TCA that has documented eﬃcacy in the management of
compulsive behaviors in both human beings and animals [32,51,53]. As a
TCA, potential side eﬀects and contraindications of clomipramine are similar
to those of other medications in that class [51]. For patients with stress,
anxiety-related, or compulsive behaviors and no neurosensory component,
clomipramine may be an eﬀective means of pharmacologic support.

244

V. Virga / Vet Clin Small Anim 33 (2003) 231–251

Selective serotonin reuptake inhibitors

Fluoxetine (Prozac), paroxetine (Paxil), and sertraline (Zoloft) are

selective serotonin reuptake inhibitors (SSRIs). They share a common
mechanism of action of being serotonergic without substantially aﬀecting
the reuptake of norepinephrine or dopamine. The relatively speciﬁc action
of SSRIs is associated with fewer side eﬀects. Adverse eﬀects reported
with SSRIs include increased or reduced appetite, weight loss, gastrointes-
tinal disturbances, anxiety, aggression, restlessness, sleep disturbances, and
alterations in cardiac conduction [15,54]. Despite the relative speciﬁcity of
SSRIs compared with clomipramine, they seem to be equally eﬀective in
the management of OCD [55,56]. As with TCAs, SSRIs should not
be administered concurrently with MAOIs, selegiline, or L-tryptophan.
Contraindications may include hepatic, renal, or cardiac disease. A min-
imum of 4 to 6 weeks of administration should be allowed so as to ob-
serve and evaluate the clinical eﬀects at any given dosage. Paroxetine is the
most potent SSRI available, but it does have some anticholinergic ef-
fects. In the author’s experience, this is of primary concern in the cat, where
constipation is common. Metabolism of paroxetine is unique in that no
active metabolites are produced. This feature may favor the admin-
istration of paroxetine in elderly patients or animals with liver or kidney
disease.

Additional anxiolytics

The benzodiazepine tranquilizers, including diazepam (Valium), alpra-

zolam (Xanax), lorazepam (Ativan), oxazepam (Serax), and clonazepam
(Klonopin), are eﬀective as short-term anxiolytics and, as such, may be
beneﬁcial for acute or time-limited stress [15,57]. Benzodiazepines potentiate
the action of gamma aminobutyric acid (GABA) and, as such, act not only
as anxiolytics but also as tranquilizers. Unlike phenothiazine tranquilizers,
which are dopamine receptor antagonists, the benzodiazepines are not
neuroleptics. A primary beneﬁt of benzodiazepines is their immediate onset
of action relative to the TCAs and SSRIs. An advantage of alprazolam,
clonazepam, and lorazepam is their sustained clinical eﬀect relative to
diazepam; although variable from patient to patient, their duration of
clinical eﬀect is typically 8 to 12 hours. The benzodiazepines may potentially
interfere with short-term memory and learning [58,59]. Other potential side
eﬀects include increased appetite, physiologic dependency, paradoxic
excitement, anxiety, sleep disturbances, and hepatocellular toxicity in cats
[15,58].

Other psychotropics

Naltrexone (ReVia) is a pure opioid antagonist that may be readily

absorbed via oral administration. Although its clinical applications are
limited, one open-ended trial evaluating the eﬃcacy of naltrexone in the
management of ALD demonstrated that it may substantially reduce or

245

V. Virga / Vet Clin Small Anim 33 (2003) 231–251

Table

Psycho

tropic

medic

ations

comm

only

used

the

man

agement

behavio

ral

derma

toses

Medica

tion

Form

supp

lied

pical

dosage

range

felin

patie

nts

Typical

sage

ran

can

ine

patients

Amitriptyline

(Elav

il)

10-,

25-,

50-

75-,

100-,

150-m

tablet

0.5–1

/kg

12–2

hours

(allow

3–4

wee

for

initial

trial)

1–4

/kg

hours,

begin

ning

1–2

/kg

hour

2–3

eeks,

›

0.5–1

/kg

hours

2–3

wee

maximum

sage

PRN;

clin

ical

respo

nse

ﬂ

0.5–1

/kg

urs

eeks

until

initial

dosage)

(allow

3–4

wee

for

initial

ial)

Doxepin

dapin

Sin

equan)

10-,

25-,

50-,

75-,

100-,

150-m

capsu

les;

10-

/mL

oral

solut

ion;

(4.43%

)

top

ical

cream

0.5–1

/kg

12–2

hours

25–5

per

cat

(all

3–4

wee

for

initial

ial)

1–5

/kg

8–12

hours,

beginnin

/kg

hours

2–3

eeks,

›

0.5–1

/kg

hours

2–3

wee

maximum

sage

PRN;

clinical

response

ﬂ

/kg

hour

eeks

until

initial

dosage

)

(allow

3–4

wee

for

initial

trial)

Clomipramine

(Clo

micalm)

20-,

40-,

80-m

tablets

(sco

red)

0.5–1

/kg

hour

begin

ning

0.5

/kg

3–4

wee

ks,

›

0.5

/kg

hour

3–4

weeks

(allow

4–6

wee

for

initial

ial)

1–3.5

/kg

urs,

ginning

/kg

3–4

eeks,

›

0.5–1

/kg

hour

3–4

weeks

max

imum

dosage

clin

ical

resp

onse

crease

/kg

PRN

hours

weeks

until

initial

dosage

)

(allow

3–4

wee

for

initial

trial)

Fluoxetin

(Pro

zac)

10-,

20-,

40-m

capsules;

10-,

20-m

table

cored);

5-mg

ora

solut

ion

0.5–1

/kg

hour

(allow

4–6

wee

for

initial

trial)

/kg

12–24

urs

(all

4–6

eeks

for

initial

trial)

Paroxetin

(Pa

xil)

10-,

20-,

30-,

40-m

(sco

red);

/mL

ora

solution

0.5–1

/kg

hour

(allow

4–6

wee

for

initial

trial)

/kg

hours

(allow

4–6

weeks

for

initial

trial)

Sertralin

(Zo

loft)

25-,

50-,

100-mg

tablets

(sco

red)

0.5–1

/kg

hour

(allow

4–6

wee

for

initial

trial)

/kg

hours

(allow

4–6

weeks

for

initial

trial)

Diazepam

alium)

1-,

2-,

5-,

10-m

table

ts;

1-mg

/mL

5-mg

/mL

solut

ions

0.2–0

/kg

12–2

hours

0.55–2.2

/kg

PRN

Alprazolam

anax)

0.25-,

0.5-

1-,

2-mg

table

(sco

red)

0.025

–0.2

/kg

12–2

hour

0.05–0.25

/kg

12–2

hour

246

V. Virga / Vet Clin Small Anim 33 (2003) 231–251

Lorazep

(Ativan

)

0.5-,

1-,

2-mg

table

ts;

/mL

solut

ion

0.025–0.2

/kg

12–2

hour

0.025

–0.25

/kg

12–2

hour

Oxazep

(Serax)

10-,

15-,

30-m

capsu

les;

15-m

table

0.2–0.5

/kg

2–24

hour

0.2–1

/kg

12–24

urs

Clonaz

epam

(Klonop

in)

0.5-,

1-,

2-mg

table

cored)

0.025–0.2

/kg

12–2

hour

0.05–

0.25

/kg

12–2

hour

Melato

nin

0.3-,

1-,

3-mg

table

cap

sules

3–6

12–2

hour

3–12

12–24

hour

Wit

few

exc

eptions,

the

applic

ation

psy

chotro

pic

dications

vete

rinary

beha

vioral

medicine

constit

utes

extralab

use.

imp

ortant

not

that

extra

label

use

requir

complian

ith

preme

dication

databas

routine

used

huma

medici

ne.

Hepa

tic

tabolism

and

renal

cle

arance

these

compo

unds

furth

supp

ort

preme

dication

assessm

ent

serum

bio

chemist

ry,

complete

blood

unt,

and

thyroid

fun

ctio

Psycho

tropic

dicat

ions,

catego

ry,

may

aﬀect

thyro

horm

one

ncent

rations,

poten

tiate

card

iac

arrhyt

hmias,

potent

iate

epil

eptiform

seiz

ures,

and

increase

hepa

tic

zyme

activ

ities

parti

cularly

serum

alka

line

phosph

atase

(SAP)

ractitio

ners

are

well

adv

ised

become

familiar

with

the

speciﬁc

indica

tion

contra

indica

tions,

side

eﬀec

ts,

and

pharm

acod

ynamics

the

psy

chotrop

ics

they

wish

use.

Abbr

eviatio

PO,

os;

eve

ry;

ﬂ

decre

ase;

›

increa

se;

PRN,

needed

247

V. Virga / Vet Clin Small Anim 33 (2003) 231–251

eliminate clinical signs in some cases [59]. Potential side eﬀects of naltrexone
can include severe pruritus and gastrointestinal disturbances [3,51,60].

Melatonin is a naturally occurring indole amine hormone produced by

the metabolism of serotonin and secreted by the pineal gland. In both
nocturnal and diurnal animals, the relative levels of serotonin and
melatonin in the pineal gland are inversely related during the daily
photoperiod (ie, during daylight hours, serotonin levels are high and
melatonin levels are low; during nighttime, hours melatonin levels are high
and serotonin levels are low) [61,62]. Beyond applications for some primary
dermatologic conditions, melatonin may be eﬀective as an anxiolytic in
some patients [63–65]. Melatonin has been reported to be serotonergic
(possibly as a result of its derivative and inverse relation with serotonin) and
may also function as a GABA-ergic agonist [64,66]. Potential side eﬀects of
melatonin include gastrointestinal disturbances and somnolence; however,
no data on long-term side eﬀects have been collected on companion animal
species [67].

Table 1 reviews the dosages of more commonly used psychotropics

previously discussed.

References

[1] Koo JYM, Chantal T. Psychodermatology: practical guidelines on pharmacotherapy. Arch

Dermatol 1992;128:381–8.

[2] Savin JA, Cotterill JA. Psychocutaneous disorders. In: Champion RH, Burton JL, Ebling

FJG, editors. Textbook of dermatology. vol. 4. 5th edition. Oxford: Blackwell Scientiﬁc
Publications; 1992. p. 2479–96.

[3] White SD. Naltrexone for treatment of acral lick dermatitis in dogs. JAVMA

1990;196:1073–6.

[4] Panconesi E, Hautman G. Psychophysiology of stress in dermatology. Clin Dermatol

1996;14:399–422.

[5] Pert CB, Ruff MR, Weber RJ, et al. Neuropeptides and their receptors: a psychosomatic

network. J Immunol 1985;135(Suppl):S820–6.

[6] Rassmussen JE. Psychosomatic dermatology. Arch Dermatol 1990;126:90–3.
[7] Field TM, Schanberg SM, Scaﬁd F, et al. Tactile kinesthetic stimulation effects on preterm

neonates. Pediatrics 1986;77:654–8.

[8] Meier GW. Infantile handling and development in Siamese kittens. J Comp Physiol

Psychol 1961;54:284–6.

[9] Paul J, Kuhn CM, Field TM, et al. Positive effect of tactile and kinesthetic or vestibular

stimulation on neuroendocrine and ornithine decarboxylase activity in maternally deprived
rat pups. Life Sci 1986;39:2081–7.

[10] Solomon GF, Levine S, Kraft JK. Early experience and immunity. Nature 1968;220:821–2.
[11] Bar LHJ, Kuypers BRM. Behavioral therapy in dermatological practice. Br J Dermatol

1973;88:591–8.

[12] Fried RG. Nonpharmacologic treatments in psychodermatology. Clin Dermatol 2002;

20:177–85.

[13] Koblenzer CS. Successful treatment of a chronic disabling dermatoses by psychotherapy.

J Am Acad Dermatol 1986;15:390–3.

248

V. Virga / Vet Clin Small Anim 33 (2003) 231–251

[14] Surman OS, Gottleib SK, Hackett TP, et al. Hypnosis in the treatment of warts. Arch Gen

Psychiatry 1973;28:439–41.

[15] Gupta MA, Gupta AK. Psychodermatology: an update. J Am Acad Dermatol

1996;34:1030–46.

[16] Koo JYM, Do JH, Lee CS. Psychodermatology. J Am Acad Dermatol 2000;43:848–53.
[17] Koblenzer CS. Cutaneous manifestations of psychiatric disease that commonly present to

the dermatologist—diagnosis and treatment. Int J Psychiatry Med 1992;22:47–63.

[18] Koo JYM, Lebwohl A. Psychodermatology: the mind and skin connection. Am Fam

Physician 2001;64:1873–8.

[19] Koblenzer CS. Itching and the atopic skin. J Allergy Clin Immunol 1999;104(Suppl):S109–13.
[20] Tobin D, Nabarro G, Baart dela Faille H, et al. Increased number of immunoreactive

nerve ﬁbers in atopic dermatitis. J Allergy Clin Immunol 1992;90:613–22.

[21] Shanley K. Pathophysiology of pruritus. Vet Clin North Am Small Anim Pract

1988;18:971–81.

[22] Scheich G, Florin I, Rudolph R, et al. Personality characteristics and serum IgE level in

patients with atopic dermatitis. J Psychosom Res 1993;37:637–42.

[23] Overall K. Fears, anxieties, and stereotypies. In: Clinical behavioral medicine for small

animals. St. Louis: Mosby-Year Book; 1997. p. 209–50.

[24] Overall K. Self-injurious behavior and obsessive-compulsive disorder in domestic animals.

In: Dodman NH, Shuster L. editors. Psychopharmacology of animal behavior disorders.
Malden (MA): Blackwell Science; 1998. p. 222–52.

[25] American Psychiatric Association. Diagnostic and statistical manual of mental disorders.

4th edition. Washington, DC: American Psychiatric Association; 1994.

[26] Luescher UA, McKeown DB, Halip J. Stereotypic or obsessive-compulsive disorders in

dogs and cats. Vet Clin North Am Small Anim Pract 1991;21:401–13.

[27] Overall K. Terminology in behavioral medicine: necessary and sufﬁcient conditions for

behavioral diagnoses. Newsletter of the American Veterinary Society of Animal Behavior
1995;17:3–7.

[28] Scott DW, Miller WH, Grifﬁn CE. Psychogenic skin diseases. In: Small animal

dermatology. 5th edition. Philadelphia: WB Saunders; 1995. p. 846–58.

[29] Shanley KS, Overall KL. Psychogenic dermatoses. In: Kirk RW, Bonagura JD, Current

veterinary therapy XI: small animal practice. Philadelphia: WB Saunders; 1992.
p. 552–8.

[30] Young MS, Manning TO. Psychogenic dermatoses (dog and cat). Dermatol Rep

1994;3:1–8.

[31] Moon-Fanelli AA, Dodman NH, O’Sullivan RL. Veterinary models of compulsive self-

grooming: parallels with trichotillomania. In: Christenson GA, Stein DJ, Hollander E,
editors. Trichotillomania: new developments. Washington, DC: American Psychiatric
Press; 1999. p. 63–92.

[32] Rappaport JL, Ryland DH, Kriet M. Drug treatment of canine acral lick: an animal model

of obsessive-compulsive disorder. Arch Gen Psychiatry 1992;49:517–21.

[33] Houpt KA. Aggression and social structure. In: Domestic animal behavior for

veterinarians and animal scientists. 3rd edition. Ames (IA): Iowa State University Press;
1998. p. 33–81.

[34] Eckstein RA, Hart BL. The organization and control of grooming in cats. Appl Anim

Behav Sci 2000;68:131–40.

[35] Panaman R. Behaviour and ecology of free-ranging female farm cats. Zeitschrift

Tierpsychologie 1981;56:59–73.

[36] Kenny DE. Use of naltrexone for treatment of psychogenically induced dermatoses in ﬁve

zoo animals. JAVMA 1994;204:1021–3.

[37] Beaver BV. Animal behavior case of the month. JAVMA 1993;203:651–2.
[38] Houpt KA. Learning. In: Domestic animal behavior for veterinarians and animal

scientists. 3rd edition. Ames (IA): Iowa State University Press; 1998. p. 251–85.

249

V. Virga / Vet Clin Small Anim 33 (2003) 231–251

[39] Scott DW, Miller WH, Grifﬁn CG. Structure and function of the skin. In: Muller & Kirk’s

small animal dermatology. 5th edition. Philadelphia: WB Saunders; 1995. p. 1–54.

[40] Gupta MA, Gupta AK, Schork NJ, et al. Depression modulates pruritus perception:

a study of pruritus in psoriasis, atopic dermatitis, and chronic idiopathic urticaria.
Psychosom Med 1994;56:36–40.

[41] Dahl RE, Bernhisel-Broadbent J, Scanlon-Holford S, Sampson HA, Lupo M. Sleep

disturbance in children with atopic dermatitis. Arch Pediatr Adolesc Med 1995;149:
856–860.

[42] Stores G, Burrows A, Crawford C. Physiologic sleep disturbance in children with atopic

dermatitis: a case control study. Pediatr Dermatol 1998;15:264–8.

[43] Ginsburg IH, Prystowsky JH, Kornfeld DS, Wolland H. Role of emotional factors in

adults with atopic dermatitis. Int J Dermatol 1993;32:656–9.

[44] Koo JYM, Gambla C. Cutaneous sensory disorder. Clin Dermatol 1996;14:497–502.
[45] Backonja MM, Gaylor BS. Pain assessment and evaluation in patients who have

neuropathic pain. Neurol Clin 1998;16:775–89.

[46] VanNoorden S, Polak JM, Negri L, et al. Common peptides in the brain, intestine, and

skin: embryology, evolution, and signiﬁcance. J Endocrinol 1997;75:33–4.

[47] Dubner R, Bennett GJ. Endogenous pain control mechanisms: review and hypothesis.

Ann Neurol 1978;4:451–62.

[48] Karlsten R. How do drugs relieve neurogenic pain? Drug Ther 1997;5:398–412.
[49] Max MB, Lynch SA, Muir J, et al. Effects of desipramine, amitriptyline, and ﬂuoxetine

on pain in diabetic neuropathy. N Engl J Med 1992;326:1250–6.

[50] Tennyson H, Levine N. Neurotropic and psychotropic drugs in dermatology. Clin

Dermatol 2001;19:179–97.

[51] Kaplan MD, Sadock BJ. Tricyclics and tetracyclics. In: Pocket handbook of psychiatric

drug treatment. 2nd edition. Baltimore: Williams & Wilkins; 1996. p. 170–83.

[52] Garnis-Jones S. Dermatologic side effects of psychopharmacologic agents. Clin Dermatol

1996;14:503–8.

[53] Overall KL. Use of clomipramine to treat ritualistic stereotypic motor behavior in three

dogs. JAVMA 1994;205:1733–41.

[54] Kaplan MD, Sadock BJ. Serotonin-speciﬁc reuptake inhibitors. In: Pocket handbook

of psychiatric drug treatment. 2nd edition. Baltimore: Williams & Wilkins; 1996.
p. 146–61.

[55] Freeman CPL, Trimble MR, Deakin JFW, et al. Fluvoxamine versus clomipramine in the

treatment of obsessive-compulsive disorder: a multi-center, randomized, double blind,
parallel group comparison. J Clin Psychiatry 1994;55:301–5.

[56] Greist JH, Jefferson JW, Kobak KA, et al. Efﬁcacy and tolerability of serotonin

transport inhibitors in obsessive-compulsive disorder. Arch Gen Psychiatry 1995;52:
53–60.

[57] Koo JYM, Gambla C. A psychopharmacology for dermatological patients. Clin Dermatol

1996;14:509–23.

[58] Overall K. Behavioral pharmacology. In: Clinical behavioral medicine for small animals.

St. Louis: Mosby-Year Book; 1997. p. 293–322.

[59] White SD. Management of acral lick dermatitis (ALD, lick granuloma). In: Proceed-

ings of the Fourth World Congress of Veterinary Dermatology, San Francisco. 2000. San
Francisco: World Congress of Veterinary Dermatology Association. p. 126–28.

[60] Schwartz S. Naltrexone-induced pruritis in a dog with tail-chasing behavior. JAVMA 1993;

202:278–80.

[61] Nelson RJ. The endocrine system. In: An introduction to behavioral endocrinology. 2nd

edition. Sunderland: Sinauer Associates; 2000. p. 35–96.

[62] Nelson RJ. Hormones and mood. In: An Introduction to behavioral endocrinology. 2nd

edition. Sunderland: Sinauer Associates; 2000. p. 627–58.

[63] Aronson L. Animal behavior case of the month. JAVMA 1999;215:22–4.

250

V. Virga / Vet Clin Small Anim 33 (2003) 231–251

[64] Pierreﬁche G, Zerbib R, Laborit H. Anxiolytic activity of melatonin in mice: involve-

ment of benzodiazepine receptors. Res Commun Chem Pathol Pharmacol 1993;82:
131–142.

[65] Van Moffaert M. Psychodermatology: an overview. Psychother Psychosom 1992;58:

125–136.

[66] Mills KC. Serotonin syndrome: a clinical update. Crit Care Clin 1997;13:763–83.
[67] Paradis M. Melatonin in canine alopecias. In: Proceedings of the Fourth World Congress

of Veterinary Dermatology, San Francisco. 2000. p. 52–9.

251

V. Virga / Vet Clin Small Anim 33 (2003) 231–251

Diagnosis and management of compulsive

disorders in dogs and cats

Andrew U. Luescher, DVM,

PhD

Department of Veterinary Clinical Sciences, Purdue University School of Veterinary Medicine,

1248 Lynn Hall, West Lafayette, IN 47907–1248, USA

Repetitive or sustained apparently abnormal behaviors performed out

of context have been described in various species of farm animals and
zoo animals, especially in horses. In horses, these behaviors have often
erroneously been termed stable vices and include cribbing, weaving, stall
walking, stargazing, and aggression to self. There is a vast literature on this
behavioral abnormality in these species [1]. These abnormal behaviors have
always been considered to develop from conﬁnement-induced conﬂict
behaviors and have been linked to speciﬁc husbandry practices [2].

The term conflict usually refers to motivational conﬂict (ie, the conﬂict

resulting from two opposing similarly strong motivations (eg, approach
and withdrawal). Frustration refers to the situation in which an animal is
motivated to perform a behavior but is prevented from doing so. Various
forms of conﬂict behaviors are caused by frustration or conﬂict and have
been studied in a great variety of species [3]. They are mostly derived from
normal behaviors relating to different motivational systems. With repeated
or prolonged conﬂict or frustration, these behaviors become ‘‘emancipated’’
from their original context; become exaggerated, repetitive, or sustained;
and are triggered in a variety of situations by a progressively lower level of
arousal [4].

In companion animals, behaviors such as ‘‘ﬂy snapping,’’ tail chasing,

or wool sucking were more commonly considered to be symptomatic for
seizure disorders, and treatment was usually unsuccessfully attempted
with seizure-controlling drugs. In 1991, we proposed that these abnormal
behaviors in companion animals were homologous to the stereotypic
behavior of livestock and zoo animals [5].

Vet Clin Small Anim

33 (2003) 253–267

E-mail address:

luescher@vet.purdue.edu

0195-5616/03/$ - see front matter

PII: S 0 1 9 5 - 5 6 1 6 ( 0 2 ) 0 0 1 0 0 - 6

Around the same time, researchers at the National Institutes of

Health (NIH) recognized that these behaviors shared similarities with
human obsessive compulsive disorder (OCD) [6]. Since that time, when
exhibited by companion animals, these abnormalities are usually referred
to as OCD [5,7] or compulsive disorder (CD) [8]. Obsessive compulsive
behaviors in people include repetitive behaviors, such as hand washing,
rituals, checking, arranging

/ordering, counting and hoarding, and are ac-

companied by intrusive thoughts, such as concern of contamination; con-
cern for symmetry; fear of harm; aggressive, religious, or sexual thoughts; or
pathologic doubt. Interestingly, the intrusive thoughts (obsessions) and the
associated behaviors (compulsions) do not necessarily correspond. For
instance, checking can be accompanied by aggressive, sexual, religious, or
somatic obsessions [9]. Compulsive behaviors are carried out to reduce
discomfort or to prevent a dreaded event [10].

The extent of the similarities between the human and canine conditions

is not yet known. One similarity is that, overall, the behaviors in compan-
ion animals are amenable to the same pharmacologic treatment as are
obsessions and compulsions in people. Another similarity is the type and
repetitiveness of behavior displayed (actually, the ﬁrst connection between
the canine and human disorders was made because dogs with acral lick
dermatitis appeared similar to people washing their hands repetitively and
excessively).

There are diﬀerences between the human and canine conditions.

Obsessions in animals are not amenable to direct study, and because
obsessive thoughts and the compulsive behavior do not necessarily
correspond even in people, it is diﬃcult to draw conclusions regarding
obsessions from the observed behavior. In human beings, the intrusive
thoughts are experienced as being disturbing. Most patients try to resist
them and are ashamed and disgusted. The behavior is considered senseless
by the patient. All this adds to stress. Also, obsession compulsion disorder
(OCD) patients tend to have an inﬂated sense of responsibility and great
need for reassurance. I ﬁnd it hard to see these in, for example, a tail-chasing
dog. Furthermore, OCD implies a cognitive component [11]. In what is
grouped here as canine CD, the cognitive component and the cognitive
control over the performance of the behavior may vary considerably (see
section on homogeneity of CD).

A lot more work is needed to validate the diagnosis of CD or OCD [12].

As a working deﬁnition of CD, Hewson and Luescher [8] proposed the
following:

Behaviors that are usually brought on by conﬂict, but that are subsequently
shown outside of the original context. The behaviors might share a similar
pathophysiology (e.g. changes in serotonin, dopamine and beta-endorphin
systems). Compulsive behaviors seem abnormal because they are displayed
out of context and are often repetitive, exaggerated or sustained.

254

A.U. Luescher / Vet Clin Small Anim 33 (2003) 253–267

Presenting signs of compulsive disorder

The behaviors performed by dogs and cats with CD could be categorized

as locomotory, oral, aggressive, vocalization, and hallucinatory behaviors
(Figs. 1–6).

In dogs, locomotory behaviors include circling, tail chasing, pacing,

jumping in place, chasing light reﬂections, and freezing; in cats,
locomotory behaviors include freezing, sudden agitation and skin
rippling, ducking, and circling.

Oral behaviors manifest in dogs as leg or foot chewing, self-licking, air

or nose licking, ﬂank sucking, scratching, chewing or licking of objects,
polyphagia, polydipsia, pica, and snapping in the air (ﬂy snapping); in
cats, oral behaviors manifest as overgrooming (‘‘psychogenic dermati-
tis’’), chewing legs or feet, chewing or licking objects, wool sucking or
eating, and pica.

Compulsive behaviors related to aggression in dogs include self-directed

aggression, such as growling or biting the rear end, rear legs, or tail;
attacking the food bowl or other inanimate objects; and possibly un-
predictable aggression to people; in cats, compulsive behaviors in-
clude self-directed aggression, especially attacking the tail.

Vocalization may be compulsive rhythmic barking or whining and

persistent meowing or howling.

Hallucinatory behaviors may be staring at shadows, chasing light

reﬂections, and startling. Dogs that wake up suddenly without any
discernible trigger and jump or are aggressive may suffer from
hallucinatory CD. Cats that avoid imaginary objects, stare at shadows,
or startle without obvious cause may fall into the same category.

Fig. 1. This English Bull Terrier ﬁxated on any object on the ﬂoor. This picture illustrates the
dog’s reaction to the water bowl after approaching it to take a drink. The dog had to be given
water from a bottle to prevent dehydration.

255

A.U. Luescher / Vet Clin Small Anim 33 (2003) 253–267

Fig. 2. A hind-end checking Miniature Schnauzer.

Fig. 3. This Great Dane would ‘‘hide’’ under a curtain and freeze.

Fig. 4. Fixation on an object in a Rottweiler.

256

A.U. Luescher / Vet Clin Small Anim 33 (2003) 253–267

Causes of compulsive disorder

Stress

In agreement with the theory of stereotypic behavior in farm animals, we

consider compulsive behaviors to be conﬂict behaviors caused by environ-
mentally induced conﬂict, frustration, or stress. Therefore, any environ-
mental factor resulting in frustration (eg, no exercise oﬀ property), conﬂict

Fig. 5. Foot chewing in a Dalmatian.

Fig. 6. This Jack Russell Terrier paced the room, always staring at the ceiling. It would walk
into objects that were in its way.

257

A.U. Luescher / Vet Clin Small Anim 33 (2003) 253–267

(eg, inconsistent interaction), or stress (eg, the presence of other stressful
behavioral problems, such as dominance conﬂict with another dog,
separation anxiety, or disease) may contribute to the CD.

Genotype

A genetic predisposition is probably present in any case of CD.

Individuals may be genetically susceptible to development of a compulsive
behavior, or the genotype may determine which, if any, compulsive behavior
an animal develops. Apparent breed predispositions include ﬂank sucking in
Dobermans; spinning or freezing with the head under or between objects,
such as clothes, in Bull Terriers; tail chasing in German Shepherds and
Australian Cattle Dogs [13]; and checking the hind end in Miniature
Schnauzers. Furthermore, large-breed dogs seem to be more likely than
small-breed dogs to develop lick granulomas. In cats, it seems that Siamese
and Burmese breeds are particularly prone to develop wool sucking [5].

Temperament traits, such as genetic fearfulness, are likely to contribute

to the development of a CD as well.

Medical problems

In some cases, a dog may start to lick a lesion or sutures but then also

start to lick other parts of the body. Persistent licking may cause lick
granulomas at sites unrelated to the original lesion. This suggests that the
stress associated with physical lesions or irritations, such as those caused by
allergy, can contribute to the development of CD in an already susceptible
animal and that the irritation can initially direct the compulsive behavior
toward a particular body site.

Any other disease that increases stress and

/or irritability, such as

dermatologic disease or endocrine imbalance, may contribute to CD as well.

Conditioning

Most owners pay attention to their pets when they perform a compulsive

behavior. Therefore, most cases of CD are aggravated by inadvertent
conditioning. Performance of the behavior only in the owner’s presence is
suggestive of a purely conditioned behavior.

Pathophysiology of compulsive disorder

The pathophysiology of CD is not well understood. Most evidence stems

from drug eﬀects on the performance of compulsive behavior. Large doses
of dopaminergic drugs, such as amphetamine or apomorphine, are eﬀective
in inducing stereotyped behavior in animals, whereas the dopamine
antagonist haloperidol results in suppression of spontaneously occurring
stereotyped behavior [14].

258

A.U. Luescher / Vet Clin Small Anim 33 (2003) 253–267

The role that b-endorphins play in the development of compulsive

behavior is not known, but it has been suggested that they play a signiﬁcant
role only in the initial stages of stereotypy development [14]. The
b-endorphin antagonists can be eﬀective in suppressing compulsive behaviors,
but their application in clinical cases is not practical.

Similar to the treatment of human OCD, drugs inhibiting serotonin

reuptake have been found to be eﬀective in the treatment of CD in dogs
[6,15]. The eﬀectiveness of such drugs implies that serotonin is involved
in animal CD. Direct evidence of serotonin involvement has also been
presented [16].

Development of compulsive behavior

If compulsive behaviors in dogs are homologous to stereotypic behaviors

in farm animals, one would expect them to be shown ﬁrst in speciﬁc conﬂict
situations (acute or normal conﬂict behavior) and, with prolonged or
repeated conﬂict, to become generalized to other contexts in which the
animal experiences a high level of arousal. As the number of eliciting
contexts increases, the threshold of arousal needed to elicit the compulsive
behavior decreases and the compulsive behavior becomes more and more
frequent. The compulsive behavior can interfere with normal function and
aﬀect the human-animal bond [17]. This is indeed the pattern for some
(mostly locomotory) compulsive behaviors. Interestingly, in a clinical trial
involving 51 dogs with CD, it was found that this process was reversed
during treatment. When the severity of CD was rated as improving, the
number of contexts in which the behavior was displayed also decreased [15].

Homogeneity of compulsive disorder

Aside from the fact that several of the behaviors described in this article

may be found to have other possibly medical or neurologic reasons in the
future, there is some evidence that CD is not a homogeneous condition and
that there may be diﬀerent classes of compulsive behavior.

Development

The categories of locomotory and oral compulsive behaviors seem to

diﬀer. In general, locomotory compulsive behaviors follow the above-
described development pattern, starting in one context and gradually
generalizing to other contexts in which the animal is agitated. Oral self-
directed behaviors, however, seem to be displayed suddenly without
identiﬁable initial conﬂict and are performed at a constant rate in contexts
with little outside stimulation, (ie, when the animal seems quiet [although its
arousal level may be high]). Owners often describe that it appears as if the dog
had to perform the oral compulsive behavior so as to be able to settle down.

259

A.U. Luescher / Vet Clin Small Anim 33 (2003) 253–267

Neurophysiologic studies also seem to justify this categorization. It was

suggested that oral stereotypic behaviors may involve the mesolimbic
dopaminergic system, whereas locomotory stereotypic behaviors may
involve activation of the nigrostriatal dopaminergic system [18].

Level of cognition involved

Some behaviors, such as ﬂy snapping or spinning, seem to involve little

cognition and seem more akin to tic disorders in human beings. Other CDs,
however, seem to involve a high level of cognition. For example, Miniature
Schnauzers that are hind-end checkers are not simply looking at their hind
end in a repetitive constant fashion. First, they may turn either way. More
importantly, they may get up and check the ﬂoor where they have been
sitting and perhaps even scratch it. This implies that they actually perceive
that there is something wrong. Dogs that chase light reﬂections may wait in
the morning in the appropriate location where they know the rising sun
produces the ﬁrst light reﬂections. Dogs that are ﬁxated on a particular
object may look for that object when it is removed.

Ease of distraction

Some patients can easily be distracted with an innocuous noise, such as

clicking of the tongue. An example was a 2-year-old German Shepherd that
chased its tail. In other cases, the behavior can only be interrupted by
physically interfering with the behavior (eg, by pulling on a leash attached to
a head halter), as was the case with a Border Collie that chewed its front
foot. In some cases, even that is not possible. A Jack Russell Terrier would
attack itself ferociously in response to minute triggers. No noise or visual
distraction worked, and attempts to interfere physically increased the
intensity of the behavior even further. It is unknown if these behaviors
represent diﬀerent behavioral pathologic characteristics or if they simply
diﬀer in severity.

Up to this point, a clinical relevance for these distinctions has not been

established. In a clinical trial investigating the eﬃcacy of clomipramine in
the treatment of CD, there was no diﬀerence in response to drug treatment
between locomotory and oral compulsive behaviors [15]. We also suggest
the same behavioral treatment for all compulsive behaviors.

Clinical approach to compulsive disorder

Diagnosis

A diagnosis of CD is primarily based on a detailed history and on ruling

out other possible causes for the observed behavior. The history includes
information on the life history (ie, source, age obtained) and management
(ie, exercise, conﬁnement, training, feeding, owner-dog interaction) of the

260

A.U. Luescher / Vet Clin Small Anim 33 (2003) 253–267

animal, the disposition or temperament of the animal, and the compulsive
behavior itself. General information on the problem includes the contexts in
which the behavior occurs (triggers of the behavior), a description of the
behavior, and the events that follow the behavior. The ease or diﬃculty with
which the animal can be distracted should be noted as well. Age of the dog at
the onset of the problem and any correlated change as well as previous
attempts by the owner to treat the problem are recorded. One aspect of the
history that is particularly important for the diagnosis is the development of
the problem. This is assessed by comparison of the contexts in which the be-
havior was shown initially and the contexts in which it is shown now and the
change in the intensity of the inciting stimulus needed to trigger the behavior.

To exclude other possible causes for the behavior, a minimal medical

database consisting of a physical examination, including a basic neurologic
examination, complete blood cell count (CBC), chemistry proﬁle, and
urinalysis should be obtained.

The basic neurologic examination [19] includes observation of the animal

(ie, movement, balance); symmetry of face and eye position; and testing of
the menace reﬂex, the eye blink reﬂex, vestibular eye movement, and the
pupillary reﬂex. Sensation on the nose and lower jaw and jaw tone are
assessed. Symmetry of the larynx, pharynx, and tongue is observed, and the
gag reﬂex is tested. The masseter, trapezius, and brachiocephalicus muscles
are palpated to assess atrophy, and the spine is palpated. Hopping and
proprioceptive positioning are tested. Hearing can be tested by making a
noise behind the dog and watching for a reaction. This basic neurologic
examination should determine if neurologic problems are present or not. If
the results are normal, further neurologic testing is not warranted. In some
cases, further tests, such as electroencephalography (EEG), neurologic
imaging and spinal tap, vision and hearing tests, endocrinologic assays,
and dermatologic diagnostics, may be indicated. In cases with suspected
psychogenic polydipsia, a modiﬁed water deprivation test may need to be
performed.

Compulsive behaviors are always displayed outside their natural context,

usually in several diﬀerent contexts, and

/or are excessive. They are often

directed toward unusual target objects and are frequently repetitive or
sustained. The animal is in full consciousness while performing the behavior
and aware of its surroundings (although it may not respond to any stimuli in
the environment in some cases and may even run into furniture, for
example). The behavior can usually be interrupted, and the animal does not
exhibit a postictal phase characteristic of seizures. Performance of the
behavior is not dependent on the owner’s presence (to exclude attention-
getting behaviors). Locomotory compulsive behavior and ﬂy snapping are
typically initially shown in a speciﬁc conﬂict situation and, later, in an
increasing number of situations in which the animal is excited. Self-directed
oral compulsive behaviors are likely to be shown in situations with little
external stimulation.

261

A.U. Luescher / Vet Clin Small Anim 33 (2003) 253–267

Differential diagnosis

The diﬀerential diagnosis has to consider various behavioral, neurologic,

dermatologic, and other medical conditions.

Behavior

Whenever an animal is in a situation of frustration or motivational

conﬂict, it is normal for it to demonstrate (acute) conﬂict behavior. In
contrast to compulsive behavior, conﬂict behavior is shown only in conﬂict
situations and not when an animal reaches a threshold of arousal for other
reasons, such as anticipation of being fed. In contrast to self-directed oral
compulsive behavior, acute conﬂict behavior is not shown in situations in
which there is no outside stimulus inducing a conﬂict.

A speciﬁc conﬂict behavior could have been shown once in an acute

conﬂict and then may have become conditioned by the owner paying
attention to the animal. Such behavior is only shown in the owner’s presence.
We therefore always ask in the history if the problem is also performed when
the animal is alone (eg, outside in the yard or in another room). We also ask
what attention-getting behavior the dog is normally showing.

Neurology

Seizures need to be diﬀerentiated from CD. Circling Bull Terriers were

reported to have an abnormal EEG indicative of seizure activity [20]. Seizures
rarely look like CD, however. Animals performing compulsive behaviors are
aware of their surroundings, can usually be distracted (although sometimes
with diﬃculty) from performing their behavior, and, most importantly, do
not show a postictal phase. As opposed to seizures, animals perform the
compulsive behavior usually when alert and interacting with their environ-
ment. Also, seizures do not follow the pathogenesis typical for CD.

Neurologic disorders, such as forebrain and brain stem lesions, may

cause an animal to walk aimlessly in large circles. Circling in tighter circles
with a head tilt indicates involvement of the vestibular system [19]. In these
cases, circling is related to a balance deﬁcit. Circling has also been attributed
to lumbosacral stenosis or cauda equina syndrome. Hydrocephalus has been
suggested as a cause of circling in Bull Terriers [21].

Sensory neuropathies can induce chewing on the feet among other signs.

Pain sensation in the distal extremities is reduced. The condition can be
hereditary and is then usually apparent in young animals [19]. Tail dock
neuroma may draw a dog’s attention to its hind end. The author is not
aware of any case, however, in which tail-dock neuroma was a proven cause
for circling behavior or aggression to the tail.

Dermatology

Any dermatologic lesion or skin gland disease resulting in itching or pain

can cause licking. Dermatologic conditions that need to be considered

262

A.U. Luescher / Vet Clin Small Anim 33 (2003) 253–267

include staphylococcal furunculosis, dermatophytosis, mycotic or myco-
bacterial granuloma, allergies, and endocrinologic imbalance. Licking, in
turn, can worsen many dermatologic lesions, resulting in an itch-scratch
or itch-lick cycle. Licking may persist even long after the initiation
dermatologic cause has been removed [22]. Dermatologic conditions also
likely impose some stress on the animal, increasing the likelihood of that
animal developing a CD. Thus, there is a mutual inﬂuence between the
dermatologic and behavioral condition, resulting in a vicious cycle. Der-
matologic lesions, such as preexisting wounds or pressure point gran-
ulomas, can also direct compulsive licking toward a particular area.

Treatment

Because a CD derives from conﬂict behavior, an attempt should be made

to identify and remove the cause of conﬂict, frustration, and stress. In cases
where the cause of stress cannot be removed, it may be possible to
desensitize the animal to the stressful situation.

Lack of predictability and control over the environment is an important

stress-inducing factor and may arise from inconsistent owner-animal in-
teraction, lack of training to commands, and thus inconsistent use of com-
mands. The inappropriate use of punishment, an inconsistent routine, and
frustration of motivations, such as the motivation for social interaction, are
additional contributing factors.

Casual interaction should therefore be avoided and replaced with highly

structured interactions in a command-response-reward format. Formal
obedience sessions allow for such consistent interaction with dogs and
establish a habit of using consistent commands in everyday situations. In
cats, we recommend regular quality time at a time of day when it can always
be provided. The owners are advised to play with the cat with toys or to
clicker-train them to do tricks, such as retrieving a ball.

Owners frequently apply punishment (eg, scolding). Because it is

practically impossible to apply owner-related punishment correctly (ie, with
correct timing, at the right intensity, and every time the objectionable
behavior is shown), such punishment becomes unpredictable and thus
stressful. It should therefore not ever be used in aﬀected animals. An
acceptable alternative to punishment is response substitution. If the animal
engages in an inappropriate behavior, it is distracted with a noise, a
command is issued, and the animal is rewarded for obeying the command.

A regular routine increases the predictability of the animal’s environ-

ment. It is particularly important that feeding and exercise become a
consistent daily part of the owner’s routine.

Suﬃcient exercise provided to dogs serves to ﬁll their need for exploration

and social interaction with other dogs, even if just by sniﬃng and leaving
scent marks.

263

A.U. Luescher / Vet Clin Small Anim 33 (2003) 253–267

Rotating toys maintains the animal’s interest in them and may provide an

opportunity to reduce arousal. Particularly attractive toys, such as food-
dispensing toys, can be given at times when the performance of the
compulsive behavior is likely.

In most cases, drug therapy may prove necessary or at least facilitate

treatment. Pharmacologic intervention is most likely achieved with sero-
tonin reuptake inhibitors, although it may take 4 weeks or longer to see an ef-
fect. A clinical trial involving 51 dogs with a variety of compulsive behaviors
has proven the eﬀectiveness of the tricyclic antidepressant clomipramine [15].
A case series suggested eﬀectiveness of clomipramine for tail chasing in Ter-
riers [17]. Clinical trials on cases of acral lick dermatitis have been performed
for clomipramine, ﬂuoxetine, and sertraline [23]. Paroxetine has also been
used clinically, but its eﬀect has not been evaluated. We usually give the drug
until at least 3 weeks after it seems to have had a satisfactory eﬀect and then
wean oﬀ gradually over at least 3 weeks by reducing dose but maintaining dos-
ing frequency. If the behavior reappears during the weaning process, the dose
is increased again and maintained at the eﬀective level for some time before
resuming weaning. Weaning is important so as to avoid a rebound eﬀect.

Clomipramine can be combined with ﬂuoxetine to slow down its

metabolism. Tricyclics other than clomipramine or other anxiolytic drugs,
such as benzodiazepines or buspirone, are unlikely to have any eﬀect on CD
on their own, and their use in drug combination therapy has not been
evaluated in dogs.

b-endorphin antagonists have been used experimentally [21], but most are

injectables and have a short half-life [24], and their use for clinical cases
is not practical. The dopamine antagonist haloperidol has also not been
proven eﬀective in practice, most likely because an appropriate dosing
regimen has not been established.

Drugs, dosages, contraindications, and side eﬀects are summarized in

Table 1.

The main goal of drug treatment is to reduce the frequency of the

compulsive behavior to the point that behavioral modiﬁcation (ie, response
substitution) becomes practical. In dogs, the patient is initially trained with
positive reinforcement to perform a desirable behavior that is incompatible
with (ie, cannot be performed at the same time as) the compulsive behavior.
Whenever the dog cannot be supervised, it is put into a situation where it
cannot perform the compulsive behavior (eg, the dog can be crated if it does
not perform the behavior in the crate). As often as the dog can be closely
supervised (we often recommend keeping the dog on a leash), it is put in the
situation in which it is likely to perform the behavior. Every time the dog
shows any inclination to perform the compulsive behavior, it is distracted (if
necessary, by pulling on a leash connected to a head halter). The command
for the alternate behavior is then given. The dog either performs or is made
to perform the alternate behavior and is then rewarded. The reward can be
progressively delayed so that the dog has to stay in the chosen position for

264

A.U. Luescher / Vet Clin Small Anim 33 (2003) 253–267

increasingly longer times before the reward is given. Some serious cases of
CD have been treated successfully with this behavioral modiﬁcation
technique alone (ie, without the use of drugs).

The distraction is important. If the dog is not distracted before a com-

mand (ie, attention) is given, the treatment attempt could result in aggra-
vation of the problem through inadvertent reinforcement of the behavior.

In cats, we recommend a similar program. The cat is continuously

supervised or placed in a position in which it does not perform the behavior.
Every time the cat is about to perform the compulsive behavior, it is
distracted (startled), and its attention is then reoriented by throwing a toy.
Cats can also be clicker-trained and then treated as described for dogs.

Prognosis

Analysis of a referral case load showed that approximately two thirds of

cases improved to the client’s satisfaction and that outcome was negatively

Table 1
Pharmacologic treatment of compulsive disorder

Drug

Dose rate

Side effects

Contraindication

Clomipramine

(Clomicalm)

C: 2–3

/kg bid

F: 0.5–1

/kg sid

Sedation, urine retention

(cats), change in appetite,
diarrhea, vomiting; also,
lowering of seizure
threshold and arrhythmias;
drug should be given with
food to reduce likelihood of
gastrointestinal upset

Liver disease, history of

seizures, cardiovascular
problems, hyperthyroidism
or use of thyroid
medication, glaucoma;
diabetes mellitus patients
may become diﬃcult to
regulate because of
ﬂuctuation of blood
glucose levels; simultaneous
use of MAO inhibitors;
simultaneous use of thyroid
medication

Fluoxetine

(Prozac)

Paroxetine

(Paxil)

Sertraline

(Zoloft)

1 mg

/kg

sid–bid

1 mg

/kg

sid–bid

C: 1–3 mg

/kg

sid–bid

Sedation, increased anxiety,

animal seems ‘‘withdrawn,’’
loss of appetite; possibly
lowering of seizure threshold

With all serotonin active drugs,

there is the rare possibility of
development of the
serotonin syndrome; in the
one case known to the
author, an affected cat
periodically crossed her
front legs, abducted her hind
legs, and carried her tail in a
stiff (broomstick) position
(Straub tail)

Simultaneous use of MAO

inhibitors; diabetes mellitus
patients may be difﬁcult to
regulate

Abbreviations

: C, canine; F, feline; bid, twice daily; sid, once daily; MAO, monamine oxidase.

265

A.U. Luescher / Vet Clin Small Anim 33 (2003) 253–267

aﬀected by problem duration [25]. It is therefore important to treat CD as
early as possible.

Acknowledgement

The author thanks Dr Dianne Bevier, Purdue University, and Dr

Elizabeth Klopp, University of Illinois, for their help with dermatologic and
neurologic diﬀerentials, respectively.

References

[1] Lawrence AB, Rushen J, editors. Stereotypic animal behavior. Oxon: CAB International;

1993.

[2] Wiepkema PR. Abnormal behaviours in farm animals: ethological implications. Neth

J Zool 1985;35:279–99.

[3] Hinde RA. Animal behavior. 2nd edition. New York: McGraw Hill; 1970. p. 396–421.
[4] Oedberg FO. Behavioral responses to stress in farm animals. In: Wiepkema PR,

van Adrichem PWM, editors. Biology of stress in farm animals: an integrative ap-
proach. Seminar in the CEC programme of coordination of research on animal welfare,
Oosterbeek, Netherlands, April 17–18, 1986. Boston: Martinus Nijhoff; 1987. p. 135–50.

[5] Luescher UA, McKeown DB, Halip J. Stereotypic and obsessive-compulsive disorders in

dogs and cats. Vet Clin North Am Small Anim Pract 1991;21:401–13.

[6] Goldberger E, Rapoport JL. Canine acral lick dermatitis: response to the antiobsessional

drug clomipramine. J Am Anim Hosp Assoc 1990;27:179–82.

[7] Overall KL. Recognition, diagnosis and management of obsessive-compulsive disorders.

Part I. Canine Pract 1992;17:40–4.

[8] Hewson CJ, Luescher UA. Compulsive disorder in dogs. In: Voith VL, Borchelt PL,

editors. Readings in companion animal behavior. Trenton: Veterinary Learning Systems;
1996. p. 153–8.

[9] Summerfeldt L, Richter MA, Anthony MM, Swinson RP. Symptom structure in obsessive

compulsive disorder: a conﬁrmatory analytic study. Behav Res Ther 1999;37:297–311.

[10] Anthony MM, Downie F, Swinson RP. Diagnostic issues and epidemiology in obsessive-

compulsive disorder. In: Swinson PR, Anthony MM, Rachman S, Richter MA, editors.
Obsessive compulsive disorder. Theory, research and treatment. London: The Guilford
Press; 1998. p. 3–32.

[11] Rachman S, Shafran R. Cognitive and behavioral features of obsessive-compulsive

disorder. In: Swinson PR, Anthony MM, Rachman S, Richter MA, editors. Obsessive
compulsive disorder. Theory, research and treatment. London: The Guilford Press; 1997.
p. 51–78.

[12] Hewson CJ. Clomipramine in dogs: pharmacokinetics, neurochemical effects, and efﬁcacy

in compulsive disorder [PhD thesis]. Guelph: Ontario Veterinary College; 1998.

[13] Hartigan PJ. Compulsive tail chasing in the dog: a mini-review. Ir Vet J 2000;53:261–4.
[14] Kennes D, Odberg FO, Bouquet Y, DeRycke PH. Changes in naloxone and haloperidol

effects during the development of captivity induced jumping stereotypy in bank voles.
J Pharmacol 1988;153:19–24.

[15] Hewson CJ, Parent JM, Conlon PD, Luescher UA, Ball RO. Efﬁcacy of clomipramine in

the treatment of canine compulsive disorder: a randomized, placebo-controlled, double
blind clinical trial. JAVMA 1998;213:1760–6.

[16] Vanderbroek I, Odberg FO, Caemaert J. Microdialysis study of the caudate nucleus of

stereotyping and non-stereotyping bank voles. In: Proceedings of the 29th International
Congress of the International Society of Applied Ethology. Potters Bar, England:
Universities Federation for Animal Welfare; 1995. p. 245

266

A.U. Luescher / Vet Clin Small Anim 33 (2003) 253–267

[17] Moon-Fanelli AA, Dodman NH. Description and development of compulsive tail chasing

in terriers and response to clomipramine treatment. JAVMA 1998;212:1252–7.

[18] Cabib S. Neurobiological basis of stereotypies. In: Lawrence AB, Rushen J, editors.

Stereotypic animal behavior: fundamentals and applications to welfare. Wallingford: CAB
International; 1993. p. 119–145.

[19] Oliver JE, Lorenz MD. Neurological history and examination. In: Handbook of veterinary

neurology. 2nd edition. Philadelphia: WB Saunders; 1993. p. 9–10.

[20] Dodman NH, Knowles KE, Shuster L, Moon-Fanelli AA, Tidwell AS, Keen CL. Behavioral

changes associated with suspected complex partial seizures in Bull Terriers. JAVMA
1996;208:688–91.

[21] Dodman NH, Shuster L, White SD, et al. Use of narcotic antagonists to modify stereotypic

self-licking, self-chewing and scratching behavior in dogs. JAVMA 1988;193:815–9.

[22] Reisner I. The pathophysiological basis of behavior problems. Vet Clin North Am Small

Anim Pract 1991;21:207–24.

[23] Rapoport JL, Ryland DH, Kriete M. Drug treatment of canine acral lick: an animal model

of obsessive compulsive disorder. Arch Gen Psychiatry 1992;49:517–21.

[24] Garrett ER, el-Koussi AEA. Pharmacokinetics of morphine and its surrogates. V.

Naltrexone and naltrexone conjugate pharmacokinetics in the dog as a function of dose.
J Pharm Sci 1985;74:50–6.

[25] Luescher AU. Factors affecting the outcome of behavioral treatment. Presented at the

134th Meeting of the American Animal Hospital Association, San Diego, CA, July, 1997.

267

A.U. Luescher / Vet Clin Small Anim 33 (2003) 253–267

Diﬀerential diagnosis and management

of human-directed aggression in cats

Diane Frank, DVM

Joel Dehasse, DVM

Universite´ de Montre´al Faculte´ de Me´decine Ve´te´rinaire, Centre Hospitalier Ve´te´rinaire

Universitaire, CP 5000, Saint-Hyacinthe, Quebec City, Quebec, Canada

Brussels, Belgium

Human-directed aggression in cats occurs for various reasons and is

therefore managed diﬀerently depending on the clinical manifestations and
causes. We suggest a medical model allowing a stepwise progression from
signs to diagnoses and treatments. This approach not only analyzes the
aggressive behaviors shown by the cat but correlates these aggressive
patterns with all the signs presented by the cat without forgetting to consider
the cat within an ecosystem that includes people, the physical environment,
and, occasionally, other pets.

The ﬁrst step is to understand what we observe. Any behavior or conduct

(sequence of acts) that threatens to lead or leads to a physically or
psychologically harmful contact and

/or prejudice, such as a ﬁght, can be

considered aggressive. Aggressive behaviors between socialized individuals
are agonistic behaviors and are means to solve communication problems
related to sharing resources or protecting safety. The sole purpose of the
aggression may be to increase distance between two individuals.

Several didactic classiﬁcations of human-directed aggression by cats are

possible. Ethologists have classiﬁed aggressive behaviors by context (com-
petition, predation, territorial defense, interaction, and oﬀspring defense),
by emotional motivation (irritation and fear), by sensorial motivation
(pain), by cognitive motivation (anticipation), and by consequences (dis-
tancing). When considered together, these descriptions are incongruent. Un-
fortunately, we cannot solve this incongruence. We should keep in mind that
classiﬁcations and deﬁnitions are a matter of consensus and can be changed.

Vet Clin Small Anim

33 (2003) 269–286

* Corresponding author.
E-mail address:

Diane.frank@umontreal.ca (D. Frank).

0195-5616/03/$ - see front matter

PII: S 0 1 9 5 - 5 6 1 6 ( 0 2 ) 0 0 1 3 1 - 6

Theoretically, from the cat’s point of view, a human being can be a

known individual to whom the cat is attached, a friend, a predator (a
threat), or, sometimes, a ‘‘prey’’ object.

Kitten development and aggression

Events during early life can aﬀect the degree to which a cat shows

aggression to human beings. Feral kittens can show aggression toward
human beings at an early stage. At play are two aspects of kitten
development: socialization and learning to modulate their responses
(learning self-control).

Socialization to people

The interspeciﬁc socialization of kittens begins as early as 2 weeks of age

and ends between 7 and 9 weeks [1–5] . During this short period, the kitten
has to learn to relate (communicate, aﬃliate itself, and bond) to diﬀerent
people, a process it is not always able to complete.

Just as socialization gives the kitten resources to relate and communicate

with diﬀerent species, it also reduces the risk of defensive aggression.

As summarized by Karsh and Turner [3], the more a kitten has been

handled, the more it becomes accustomed to people. Kittens that are
handled early (beginning at 3 weeks of age) become signiﬁcantly more
responsive and friendlier to people than kittens that are handled after
7 weeks of age [2,4,5]. In this experiment, handling kittens for 40 minutes
daily produced increased attachment to people when these cats were com-
pared with kittens handled 15 minutes daily. It is interesting to note that
approximately 15% of kittens had a temperament that was ‘‘resistant’’ to
socialization. Handling did not change their behavior to people, signifying a
probable eﬀect of heredity, particularly for excitable and hyperactive
temperaments [5]. Beyond the speciﬁc meaning of the temperaments
described, we should keep in mind that early handling speciﬁcally aﬀects
the cat’s perception of familiar and unfamiliar people [6].

Interspeciﬁc socialization facilitates sociability, but it is not an ultimate

and perfect process. The capacity to communicate socially with people needs
maintenance (ie, repetitive positive contact), or the cat may lose its
sociability to people [1].

Learning to modulate their responses (self-control)

It is assumed that the queen educates her kittens about appropriate

behavior. Two speciﬁc behaviors (corrections) may have an inﬂuence on
teaching kittens self-control: the queen batting on the kitten’s nose and the
queen kicking and scratching the kitten’s belly with her hind limbs. Kittens
bite and scratch during social play and, by mutual feedback, learn to

270

D. Frank, J. Dehasse / Vet Clin Small Anim 33 (2003) 269–286

modulate their actions (bite inhibition and claw sheathing). There are
currently no data on the regulatory inﬂuence of these corrections on
subsequent kitten self-control. Nevertheless, it is interesting to note that
orphaned kittens do not inhibit their bites, scratches, or motor activity as
well as kittens raised by a ‘‘good’’ mother. It is, of course, always important
to deﬁne context when talking about ‘‘self-control.’’ Some situations do not
warrant self-control.

Heredity also plays a role. In people, it has been shown that genetic

predisposition may control or modify the environmental exposure and
inﬂuence. Theoretically, an aggressive kitten could avoid corrections and
the educational inﬂuence of the queen. Each litter mate within a given
environment is thus experiencing diﬀerent environmental inﬂuences. This
process has been called the genetic-environment correlation [7]. Kittens
from a given litter can therefore develop diﬀerent aggressive ‘‘drives,’’ which
cannot be predicted based solely on environment and heredity.

Classiﬁcation of aggressive behaviors

Several patterns of feline aggression towards human beings can be

diﬀerentiated based on the sequence of motor acts, triggering events, and
probable motivations. There is no behavior without previous emotion.
Emotion is deﬁned as an urge, drive, or push to act (from the Latin emotio
[e-, out

þ movere, to move]) or a motivational state, arousal, or action

tendency, such as fear, anger, or joy, accompanied by certain physiologic
and behavioral activities [8]. Hence, all aggressive behaviors are affective.

Predatory behavior

The typical sequence involves stalking ambush; slow walk; crouching

close to the ground; slinking trot alternating with lying down; low-proﬁle
walk to reach the striking range; short run, springing, lunging, and then
pouncing (rodent prey); forepaw catching (bird) or hooking (ﬁsh); nape bite;
and play with the motionless prey. During this entire sequence, the head and
ears are stretched forward, the whiskers are spread, and the tail twitches [9].
The cat remains silent. Sometimes the teeth ‘‘chatter.’’ The captured prey
may be brought to a dining area, eaten on the spot, or left behind. Ingestion
of the prey by household-fed cats is infrequent. Cats frequently hunt for an
average of 10 trials before catching prey. Hunting is high in the hierarchy of
feline behavioral drives. Motivationally, a cat hunts to hunt, not to eat, even
if hunger increases the incentive. Elements of this predatory sequence are
also observed during play.

Does this type of aggression target human beings? Usually, cats hunt

animals their size or smaller; humans do not match this proﬁle. Elements of
the predatory sequence, such as pouncing, swatting, grabbing, and pawing
behaviors, are observed and directed to human body parts, however. These

271

D. Frank, J. Dehasse / Vet Clin Small Anim 33 (2003) 269–286

are typical of both predatory behavior and play and are generally referred to
as inappropriate play or play aggression [10]. The body parts are usually the
ankles, legs, and hands. The head can be a target when cats crouch down
on a high ambush place. Cats may wait and ambush people by hiding under
furniture, at the bottom of stairs, or in other commonly traveled areas [10].
Bites may be controlled or uninhibited. Predatory behavior should be
considered when the entire hunting behavioral sequence is observed. Spe-
ciﬁc hunting behavioral patterns are performed in an organized sequence. Play
behavior, conversely, has a totally disorganized sequence. The other
problem with the term play is that, by deﬁnition, play is not injurious and
does not lead to prolonged mutual avoidance by the players [11].

This displaced predatory behavior can be seen in cats of all ages and is most

often observed when cats have no opportunity to hunt or play aerobically. The
expected duration of hunting activity of free-ranging house or farm cats varies
from 0% to 46% of the 24-hour day, with an average of 14.8%. The average
duration of an excursion is 30 minutes [12]. This means that a normal and
suﬃciently fed cat should still have more than 3 hours of hunting-like activity
per day. Motivation for hunting is higher than for playing. It is therefore no
wonder that this behavior is sometimes expressed without any identiﬁable
stimulus as a vacuum or overﬂow activity (ie, an innate behavior occurring
spontaneously in the absence of the triggering stimulation) [13].

Hunting or play ﬁghting with human body parts is an opportunistic

activity, with people being the only moving interactive ‘‘item’’ inside a
house. As with hunting, these aggressive displays are more often observed
at sunrise and sunset (and this is not correlated with client presence and
activity level), when the cat is hungry (hunger activates hunting), and when
people are moving about (motion activates hunting). Environmental
enrichment is essential to address this clinical manifestation [14]. The cat
needs hunting and playing opportunities to redirect the behaviors on
appropriate targets. The cat should be provided with toys that move, ﬂutter,
bounce, and stimulate chasing, stalking, pouncing, grabbing, and swatting
behaviors [10]. Aversive stimuli, such as water spraying, or disruptive
stimuli, such as throwing a ping-pong ball, may be used to interrupt the
early signs of the aggressive behavior, but cats have to be provided with an
alternative target to hunt or chase.

Predatory behavior directed toward people is seldom the only sign

displayed by a cat. Careful history taking may reveal other signs, such as
mydriasis, vomiting, or diarrhea (or other autonomic signs), hypervigilance
(scanning or startling for trivial stimulus), and bouts of overactivity (see
Appendix). Medical workup is always performed for nonspeciﬁc signs of
organic disease. Even if an organic condition is identiﬁed, together, all these
signs may also point to a problem of anxiety. The French veterinary
behaviorists call this cluster of signs ‘‘l’anxie´te´ du chat en milieu clos,’’
which can be translated as ‘‘anxiety disorder of the cat in conﬁned sur-
roundings,’’ or ‘‘anxie´te´ du chat en appartement,’’ translated as ‘‘anxiety

272

D. Frank, J. Dehasse / Vet Clin Small Anim 33 (2003) 269–286

disorder in apartments’’ [15]. Treatment of this disorder requires not only
redirecting hunting to toys but educating the clients.

Play-related aggression

Play has no evil or aggressive motivation. One of the main diﬀerences

between play and reality, such as hunting or ﬁghting, is the absence of
intense emotions. When emotions like fear and aggression arise, it is no
longer play. Play can be rough and cause harm, namely, if cats lack self-
control (do not modulate their responses) or suﬃcient exercise or have
learned to play inappropriately (operant conditioning). A second diﬀerence
between play and reality is that play is made up of a speciﬁc sequence of
acts in which acts may be repeated, exaggerated, uncompleted, or reordered,
and the sequence itself may be terminated earlier than normal by the
introduction of irrelevant activities [11]. In other words, motor patterns are
similar to adult behavioral patterns but the context, intensity, and
sequencing are altered if they are part of play. Speciﬁc behavioral patterns
are performed in disconnected and varying groupings.

Management of play-related aggression requires instituting appropriate

and regular play opportunities. Clients should never allow their pet to play
with them in an aggressive manner and should withdraw attention when the
cat is not behaving appropriately. Wrestling behaviors, play using human
body parts, and teasing behaviors should all be avoided [10]. Play-related
aggression is a nonspeciﬁc sign seen in ‘‘overactive’’ cats (cats unable to
modulate motor activities) or associated with an environment lacking
adequate stimulation. In the former case, redirecting play toward adequate
toys should be suﬃcient to cure the problem.

Self-defense aggression

The cat has the right to defend itself against real or imaginary dangers

coming toward it. It may show mild to extreme self-defense aggression,
varying from controlled irritable frustration or aversion-induced aggres-
sion to pain-induced aggression, uninhibited fear aggression, or life-saving
antipredatory aggression [8,16]. There is a continuum in defensive ag-
gression. The aggressive reaction is proportional to the danger perceived by
the cat. It means that self-defense aggression involves emotions (eg, fright,
fear, irritation) and cognition, including memory of past experiences,
socialization, and learning.

The danger is also correlated to the proximity of the threat and the ﬂight

(or security) and critical distances of the cat. For example, the ﬂight
distance, deﬁned as the distance at which an animal ﬂees from a threat, is
also a ‘‘security distance,’’ deﬁned as the distance at which an animal oﬀers
proactive aggression (distancing aggression) to increase distance between
itself and the perceived threat [13], This security distance varies from one

273

D. Frank, J. Dehasse / Vet Clin Small Anim 33 (2003) 269–286

individual to another, depending on the threat (eg, known, unknown) and
the mood of that individual. The critical distance is deﬁned as the distance at
which an animal shows fear aggression because escape is not possible. This
fear aggression is also called the critical reaction [13].

Defensive postures include a low crouching posture, ﬂattened ears,

dilated pupils, and piloerection. Vocalizations, such as snarling, growling,
hissing, and spitting, are common. The increased intensity of fear shows in
the change of posture, varying from ventral to lateral recumbency, with all
teeth bared and claws exposed.

Being emotionally and cognitively related, self-defense aggression is

easily conditioned both in the classic Pavlovian way (to a neutral stimulus
associated with a triggering event) and by operant conditioning. For
example, the sight of the owner who had to treat the cat for a painful injury
is enough to trigger fear and fear aggression (classic conditioning). The
withdrawal of the owner in front of the threatening cat is a rewarding
process (operant conditioning).

The cat may easily become sensitized and may generalize the process.

Being aroused, the cat may redirect the aggressive display to a neutral
stimulus interfering at the least appropriate moment.

Self-defense aggression is a nonspeciﬁc sign. As mentioned earlier, it can

be expressed as redirected aggression or as a result of pain, fear, or anxiety.
Self-defense aggression can be managed with desensitization and counter-
conditioning techniques, but symptomatic treatment does not systematically
address underlying causes. It is best to assess the animal presenting this sign
for other signs that also need to be analyzed and correlated. This approach
gives us a more complete picture on which to base our diagnosis and
treatment. Careful history taking allows the veterinarian to determine if the
patient is ill (and requires treatment and

/or medication) or not. It can also

reveal if the environmental conditions meet the ‘‘ethologic’’ and social needs
of the cat. Finally, it should indicate whether or not client behaviors and
expectations are appropriate and realistic for the patient. It is equally
important to question oneself as to why a cat is exhibiting fear or distancing
aggression toward only one person or family member. In some instances, it
may be the human behaviors that are totally inappropriate. Like children,
animals can be silent victims of abuse and violence.

Defensive aggression is the most common form of aggression in cats.

Keeping this point in mind, veterinarians and clients need to have some
empathy for these cats and not lose patience or yell at these animals.
Correctly reading and interpreting the cat’s body language is essential to
prevent defensive aggression properly.

Petting-induced aggression

This type of aggression is distressing to pet owners but not well

understood [10]. Often, a cat tolerates physical interaction for a period of

274

D. Frank, J. Dehasse / Vet Clin Small Anim 33 (2003) 269–286

time and then bites and

/or scratches the person and leaves. Easily enough,

one can consider this aggression as defensive. The assumption made by
clients is that a cat coming within close proximity to a person wants to be
petted. This remains a human perception. Whether or not all cats enjoy
physical contact still needs to be demonstrated. The cat probably enjoys
the proximity as a place to be or to lie down. The display of mild, generally
controlled, aversion-induced aggression may be quite appropriate from
the cat’s standpoint. A cat does not have words to communicate ‘‘stop’’
or ‘‘enough’’ to people. One hypothesis could be that several neurologic
pathways of touch and pain (eg, the neospinothalamic and spinoreticular
pathways) are the same and that repetitive physical contact may induce
arousal, excitement, pain, and, sometimes, static electricity [17]. Pain,
especially when chronic, can be diﬃcult to assess in cats.

This aggression is often quoted as unpredictable, but close observation

shows that the cat is in fact signaling through changes in body posture, such
as stiﬀening, ears ﬂattening, pupil dilation, tail twitching rigidly, and low
growling. It is therefore not unpredictable.

Treating the cat is a bit more complicated than treating just the sign,

because some of these cats are in fact anxious. A cluster of several signs
presented by the patient is often compatible with anxiety. Treatment for
anxiety requires more than just desensitizing and counterconditioning. If
the cat is in pain, desensitizing and counterconditioning are no longer
appropriate recommendations. The importance of looking at the big picture
is again emphasized. If this type of aggression is the only sign presented, the
patient may simply not enjoy close physical contact, and treatment is easily
achieved by respecting the cat’s desire not to be petted.

Offensive aggression

The postural displays and sequence of motor acts of oﬀensive aggression

include an inverted U-shaped tail posture, piloerection, and lateral hops
toward the intruder along with growling and hissing sounds. If this display is
insuﬃcient, the cat charges and ﬁghts. The tail can then be horizontal and
piloerected. Oﬀensive aggression (a cat moving aggressively in the direction of
a potential victim who is not coming in its direction) is not always correlated to
speciﬁc triggering events or stimuli. One author has seen a cat showing this
kind of aggression towards men only. The sequence went from the mock
attack to a real charge followed by scratches and bites. The cat attack could be
disrupted and prevented by the throwing of an object at the beginning of the
behavioral sequence. With known members of the family, the attack could be
truncated at the beginning of the sequence when a sustained gaze from the
oﬀensive cat resulted in the departure (escape) of the potential victim.

Once again, this aggression display is only a sign and is seldom seen

alone, with these cats often exhibiting several signs of anxiety or other
disorders.

275

D. Frank, J. Dehasse / Vet Clin Small Anim 33 (2003) 269–286

Redirected aggression

Redirected aggression is not one type of aggression with a speciﬁc

sequence and triggering context. In fact, any kind of aggression may be
redirected. The victim of the aggressive display has nothing to do with the
triggering event or context. For example, a cat is aroused by the presence of
another cat seen through the window, but the ﬁrst cat cannot attack the
second cat. The ﬁrst cat then redirects the attack toward another more
accessible target, such as the owners. The triggering stimulus can be the sight
of another cat, smells, and

/or sounds. The redirected target is most often a

moving individual in proximity to the cat at the time of arousal.

Operant and classic conditioning can modify redirected aggression. For

example, the cat sees an outdoor cat and redirects the aggression to a new
target (a person or another household animal). The next time it sees the
outdoor cat, it seeks out the other household pet or person and redirects the
aggression toward the ‘‘new’’ target again. The victim’s behavior may also
play a role in perpetuating the cycle of aggression.

Pathologic aggression

This form of aggression is an automated, reﬂexive, conditioned

aggression that is nonadaptive. The sequence is truncated, and the bites
and scratches are uninhibited. The attack is often direct without any
preliminary threat, the aggression is severe, and the bites are serious, caus-
ing harm. Pathologic aggression can be primary or secondary. Primary
pathologic aggression is often observed in cats unable to modulate their
responses, such as orphaned or early-weaned kittens that have been raised
exclusively by people. It can also be observed in cats suﬀering from neu-
rologic problems, such as neoplasia or infection. Never overlook poten-
tial organic causes. Secondary pathologic aggression may have evolved
from another type of aggression through operant conditioning. For
example, a cat was aroused by the smell of roasting chicken and a child
accidentally stepped on the cat’s tail. The cat redirected a self-defensive
(pain) aggression to the child. The cat had to be conﬁned to the kitchen,
where he climbed on walls and windows for several hours. After that
episode, whenever the clients cooked chicken, the cat would seek out the
child and attack him severely. Between these aggressive episodes, the cat was
quite aﬀectionate with the child and even slept with him.

Pathogenesis

With respect to the medical model, signs arise and develop depending on

pathogenic factors. Pathogenesis deals with the origins and development of
disease. In behavioral medicine, the situation is similar. The identiﬁcation of
these factors may lead to solutions and treatments. In this article, we cannot
dwell on these factors but do enumerate a few examples. These are as follows:

276

D. Frank, J. Dehasse / Vet Clin Small Anim 33 (2003) 269–286

Internal factors, such as acute or chronic pain (eg, arthritis),

endocrinopathy (eg, hyperthyroidism), senile dementia, brain tumors,
hepatosis, or portosystemic shunt

Biologic factors, such as viral (rabies, feline immunodeﬁciency virus

[FIV], feline leukemia virus [FeLV], feline infectious peritonitis [FIP]),
bacterial, or parasitic (toxoplasma) infections, and chemical factors,
such as dissociative anesthetics (ketamine)

Multiple psychologic factors, such as sensitization, generalization, (poor

or lack of) imprinting and socialization, or acute (traumatic) and
chronic stress

Environmental factors, such as physical environment (eg, restriction of

space, lack of stimulation or hypostimulating, excess stimulation or
hyperstimulating) or social environment (eg, presence of other animals,
client interactions and expectations, abusive people)

For example, when confronted with predatory aggression (complete

organized hunting sequence), one should think about metabolic distur-
bances (hunger, diabetes, or hyperthyroidism), overactivity, or lack of
stimulation within the environment. Identiﬁcation and treatment or
correction of these factors may improve the animal’s condition.

Several factors may coexist and act in synergy to modify the normal

adaptive behaviors into nonadaptive behaviors. These behaviors may
become nonadaptive in their sequence, frequency, expression, or context
and become signs of illness or pathologic behaviors.

Disorder description (nosography)

A cat often shows more than one sign. One sign is not necessarily

indicative of illness. A cluster of signs may indicate illness. One can treat
each sign or, better still, treat the cat. One can make several symptomatic
diagnoses, or one can gather all the signs within a cluster and call it a speciﬁc
disorder. If a description of disorders is realized, it can be purely descriptive
or it can take into account the etiologies and pathogenesis. In the ﬁrst
scenario, it is simply a conventional labeling so that people can speak the
same language. In the second scenario, therapies can be implemented. We
have listed a few examples.

It is important to understand that the behaviors in each disorder interfere

signiﬁcantly with the animal’s normal routine, occupations, and social
activities. The cluster of behaviors is not accounted for by another disorder
or medical condition and is not a result of the direct eﬀects of a substance,
such as medication. In other words, medical rule-outs have been done for
the nonspeciﬁc signs that are also compatible with ‘‘identiﬁable organic
disease,’’ and careful history taking has identiﬁed any inappropriate client
behaviors. The presence of an organic cause does not rule out ‘‘behavioral
illnesses.’’ Several conditions may occur concurrently in a given patient. The

277

D. Frank, J. Dehasse / Vet Clin Small Anim 33 (2003) 269–286

importance of taking a good and complete history cannot be over-
emphasized.

Generalized anxiety disorder

Generalized anxiety disorders include frequent or excessive fear-like

behaviors and anticipation not triggered by speciﬁc repetitive identiﬁable
stimuli. Several of the following signs are present:

Defensive behaviors, such as freezing (immobility), avoidance, escape,

distancing aggression, or fear aggression

Distress signs, such as vocalizations or clinging to a familiar individual

(eg, the owners)

Increased autonomic hyperreactivity, such as perspiration, myosis, or

mydriasis

Hypervigilance, such as scanning or startling for trivial stimuli
Cautiousness or shyness
Displacement activities, such as fur licking, pacing, polyphagia, or

polydipsia

Increased marking behaviors, such as face rubbing, scratching, or urine

spraying

Increased incidence of house soiling
Modiﬁcation of the frequency, intensity, or social context of feeding

routines, such as eating at night or eating alone

Change in the sleep routine, such as sleeping in a hiding place

The etiology and pathogenesis of this disorder are multifactorial (it is

recommended that the reader refer back to the section on pathogenesis).

Cat anxiety in closed surroundings

1. Signs appear in a cat living in a hypostimulating small-sized environment.

This environment does not allow the cat to distance itself or withdraw as
easily from unwanted situations. Generally, these cats have been raised
in a more stimulating environment, such as a cat adopted from outdoors,
or may be descendants of outdoor cats (importance of heredity).

2. Signs as speciﬁed in generalized anxiety disorder are seen.
3. There is, or has been, an increase in one or several of the following signs:

Predatory (hunting) or play-like behaviors directed to human body

parts, such as ankles, hands, or head (or other cats)

Bouts of hyperactivity, sometimes at speciﬁc hours, such as sunset
Rolling skin syndrome
Defensive aggression triggered by constraint, manipulation, or

combing, for example

Hypervigilance (eg, constantly or periodically scanning, startling for

trivial stimulus)

Autonomic signs, such as salivation, diarrhea, or vomiting

278

D. Frank, J. Dehasse / Vet Clin Small Anim 33 (2003) 269–286

Displacement activities such as licking-related alopecia or hyper-

phagia-related obesity

Sweaty paws and/or anal sacs emptying

‘‘Hyper’’ episode disorder

1. A distinct period of abnormally and persistently ‘‘high mood’’ lasting

several days and alternating with periods of normal mood

2. The high mood period is characterized by objective signs presented

almost daily, most of the day, for at least 1 week as follows:

Decreased need for sleep (hyposomnia)
Increased motor activity (agitation) compared with the normal

expected activity level of that cat, such as moving from place to
place, running, jumping, climbing, or playing (rough)

Hypervigilance (eg, constantly scanning)
Hyperexcitability, on the go, and reacts easily to trivial stimulation
Increased distractibility

3. Other signs are possible, such as the following:

Offensive or defensive aggression triggered by trivial stimuli
Repetitive stereotyped behaviors or even stereotypies, such as tail

chasing

Staring periods lasting longer than 10 seconds
Onset of the high mood episode may be accompanied by mydriasis

Treatment and management

One can treat a sign, but it seems better to treat a cat presenting signs in a

speciﬁc environment. Defensive aggression may be a sign of anxiety. The
cat should be treated for anxiety. Oﬀensive and predatory aggressions may
be signs of overactivity or of an environment lacking adequate stimulation
(stimulus deprivation). If aggression is correlated to overactivity, this dis-
order should be treated accordingly. Environmental enrichment and drugs
that reduce activity and improve learning should be prescribed. Cur-
rently, there are few data on traditional or ‘‘alternative’’ pharmacologic
treatment of aggression in cats. We are hopeful that the new classes of drugs
acting on serotonin neurotransmission can improve conditions like anxiety
and overactivity as well as reduce reactivity and aggression. There are three
main classes of drugs that can be considered for cats aggressive to human
beings: the tricyclics antidepressants, such as amitriptyline and clomipra-
mine; the selective serotonin reuptake inhibitors (SSRIs), such as
ﬂuoxetine, ﬂuvoxamine, sertraline and paroxetine; and the monoamine
oxidase inhibitors, such as selegiline. Selegiline may be the drug of choice to
stabilize mood. None of these drugs are labeled for use in cats. Before
prescribing these psychotropic drugs, all patients should have a physical
examination, laboratory screenings for liver and kidney function, and, in

279

D. Frank, J. Dehasse / Vet Clin Small Anim 33 (2003) 269–286

some cases, electrocardiograms. Signed consent and release forms are
advisable [10]. ‘‘Plug-in’’ cat-appeasing pheromones (Feliway) are available
and are presently under study for feline anxiety. Anecdotally, the
pheromones seem eﬀective for treatment of some fearful cats.

Drug selection

All drugs mentioned previously, except pheromones, have diﬀerent eﬀects

even if they belong to the same class. They are not interchangeable.
Clomipramine and ﬂuvoxamine are more sedative, and ﬂuoxetine can cause
hyporexia or anorexia. Tricyclic antidepressants and paroxetine have
anticholinergic eﬀects and can cause constipation and urine retention.

Dosage can change the sedative eﬀect. Clients should be instructed not to

reduce or increase the dosage of drugs prescribed for the pet without
professional advice.

The drug also has to be selected with respect to the cat. Because we are

dealing with cats that are aggressive to people, one needs to recommend
medications that are easily administered to a recalcitrant cat. The client
should not be put in a risky situation trying to force a pill down an
aggressive patient. These drugs are often bitter. Compliance with treatment
and palatability of the drug is a concern when medicating aggressive cats.
Consider drugs available in chewable forms and liquids, gelatin capsules for
halved tablets, and pharmacies that do specialty compounding for your
clients. Dissolving drugs (eg, ﬂuoxetine) into highly palatable (ﬁsh or meat
ﬂavored) liquids increases compliance.

Dosage and duration of treatment

Tricyclic antidepressants should be given at a dose of 0.3 to 0.5 mg

/kg/d

(the exception being amitriptyline, which can be given at a dose of 1 mg

/kg

twice a day). The anticholinergic eﬀects may cause urine retention,
constipation, tachycardia, and dryness of the mouth and conjunctiva. These
drugs should never be given to patients with heart or glaucoma problems.
SSRIs should be started at a dose of 0.5mg

/kg/d, and dosage may be

increased by increments of 0.1 mg

/kg/d every 3 to 5 days cautiously up to

1 mg

/kg/d only if needed.

The duration of a medical treatment should be for 3 to 6 months. The

duration is set arbitrarily to ensure at least 3 months without any
‘‘aggression’’ (if aggression is the problem for which the client consulted)
before considering a weaning schedule. Discontinuation of the medication
should be gradual, with a weaning period of 1 week per month of treatment.

Therapies

The goal of therapies is to help the cat become better adapted to diﬀerent

situations.

280

D. Frank, J. Dehasse / Vet Clin Small Anim 33 (2003) 269–286

Play therapy

If the cat spends long hours alone and is not fearful of the clients or

anxious, the clients can schedule and initiate a daily interactive play session
lasting at least 15 minutes [18,19]. The cat should beneﬁt from the aerobic
activity and from undivided client attention. In addition, the client can
rotate cat toys every few days so that the cat has new and novel items while
the client is gone. These might include boxes, bags, and feeder toys. Clients
can provide a selection of toys to determine the cat’s individual preferences.
Try a selection of toys ﬁrst, and see which type cat chooses to use most so as
to meet the individual cat’s preferences.

Interrupting predatory or play aggression

If the cat still occasionally ‘‘attacks’’ the client, the client is encouraged

to startle the cat with the mildest stimulus necessary. The purpose is not to
punish or terrify the cat with an aversive stimulus but only to interrupt the
undesirable behavior. Clients can make a high-pitched sound or use a squirt
bottle or a compressed air canister. If the behavior can be interrupted at the
beginning of the sequence (ie, the cat is just ‘‘thinking’’ of pouncing), the
potential for teaching is increased. As soon as the cat stops the undesirable
behavior, it should be redirected to an appropriate item. Tossing a ping-
pong ball or having the cat chase a toy teaches it appropriate behavior.
The more accurately the time and place of these attacks can be predicted,
the greater is the chance of preventing them. Clients can throw a toy ahead
of time as they pass by the cat’s favorite ambush spot. If the cat tolerates
wearing a harness or a quick-release collar, a few bells can be attached, thus
decreasing the owner’s element of surprise at the sight of a cat ﬂying across the
room.

Offensive aggression

Clients need to remain as calm as possible and stand still. They may want

to shield themselves physically with a blanket, a piece of cardboard, or
anything else available in the environment. They should not yell or scream,
because this generally makes the situation worse.

Preventing or interrupting defensive aggression

Clients should learn the cat’s defensive posture body language so that

they can respect the cat’s communication and avoid being attacked when the
cat defends itself.

Preventing or interrupting petting-induced aggression

The treatment of this sign (petting-induced aggression) is twofold [10].

First, the client must learn to anticipate the change to aggressive behavior
and either stop petting the cat or stand up and allow the cat to jump to the
ﬂoor. The client must be vigilant and learn to recognize the impending
signs of aggression and cease all interactions at the ﬁrst indication. Second,

281

D. Frank, J. Dehasse / Vet Clin Small Anim 33 (2003) 269–286

the client can attempt to teach the cat to tolerate increased amounts of
interaction by associating them with something pleasant. The cat must not
be in pain or uncomfortable. The client is instructed to pet the cat for the
predetermined amount of time that does not elicit the inappropriate
response. If the cat stays calm (as evidenced by body posture, facial
expressions, and ear set), it is rewarded with a food treat. Once the cat does
well at that level, the client attempts to increase petting time slowly and
rewards calm behavior with food so that the cat learns to tolerate increased
interactions. Inherent in this approach is learning what type of petting the
cat enjoys most. For some, this might be just scratching around the head
and neck as opposed to stroking down the entire body. An ethical question
does arise from the fact that we are imposing an interaction on an animal
because we can and not because it beneﬁts the animal.

Avoiding punishment

Clients should be discouraged from using any type of physical punish-

ment to correct aggression, because the cat may become more aggressive.
Clients should not try to deter the cat from inappropriate play by trying to
strike or push the cat away. The cat may interpret these human behaviors as
reciprocal play.

Confining or boarding

Conﬁning or boarding the aggressive cat is sometimes recommended to

clients when they have become exasperated or fearful. This is a temporary
measure preventing client frustration and potential injury to the cat or client.
It is important to notify clients that locking up the cat does not teach the cat
appropriate behaviors. It allows both the client and cat to calm down.
Aggressive cats can stay reactive for extended periods (several hours to days).
Interaction with the cat should not be initiated unless the cat is totally relaxed.

Client education

Clients need to understand normal feline behaviors and respect feline

communication. They need to think about the lack of opportunities
available to a cat living strictly indoors as opposed to one living outdoors.
Cats need physical activity, mental stimulation, olfactory information, and
respect. We impose choice of food, litter, housemates (people and other cats
or animals), density of population, and available physical space. Some
indoor animals would prefer to withdraw or avoid certain situations but do
not have that option anymore in our environments. Clients must understand
all this and remember that cats react quickly and defensively and stay
reactive for long periods of time.

Environmental enrichment

Most often, the environment provided for the cat does not respect the

animal’s ethologic needs. This is particularly true for cats living indoors. In

282

D. Frank, J. Dehasse / Vet Clin Small Anim 33 (2003) 269–286

these circumstances, one should provide the cat with aerobic activities and
hiding and observing places, for example These essential requirements have
been detailed by Schroll [14,20]. They include three-dimensional spatial
organizational opportunities that allow scratching, climbing, and jumping;
separate feeding and drinking areas; and hunting and exploration activities,
such as dry food hidden in toys or under furniture. Hiding opportunities,
such as discrete resting places making it possible for the cat ‘‘to see but not
be seen,’’ should also be provided. Giving more details on environment
enrichment is outside the scope of this article.

Surgery

No surgical technique can cure aggression. Some procedures, such as

declawing or removing teeth, may help to decrease danger to people,
particularly to children or the elderly. These cats can still cause severe
bruising, however. These procedures should never be performed as a sole
treatment and should only be considered as last-resort options.

Castration of tomcats or queens only helps if the aggression is displayed

during phases of sexual activity or excitement. Neutering does not address
underlying causes of feline aggression toward people.

Prevention

Selecting a kitten that is well socialized to many diﬀerent people (eg,

babies, children, elderly people, diﬀerent races), keeping the kitten in
contact with people several times a week, rearing it with kindness, and
avoiding aversive interaction are the ﬁrst preventative steps to avoid
aggression toward people.

It is important that clients do not play or wrestle with the cat using their

hands and feet. If the kitten or cat accidentally grabs body parts in its mouth
or claws, the interaction should stop. Withdrawal of attention is negative
punishment (something pleasant is taken away) and is generally eﬀective,
especially if it is also coupled with rewarding appropriate behaviors. An
owner should only play with cats or kittens using toys at a distance, such
as a toy attached to a ﬁshing line and rod.

Summary

Human-directed aggression in cats should be evaluated as a multi-

factorial problem. It results from the combined actions of heredity,
environment, learning, human social requirements (or needs), client
interactions, lack of understanding of normal feline behavior, unrealistic
client expectations, and lack of meeting the cat’s basic ethologic needs.

Managing human-directed aggression in cats encompasses the use of

environmental modiﬁcation, therapies, and, when and if needed, regulatory

283

D. Frank, J. Dehasse / Vet Clin Small Anim 33 (2003) 269–286

drugs so as to increase learning capabilities and adaptation and decrease
danger to the human victims.

Acknowledgments

The authors thank Jenny O’Connor, Sabine Schroll, and Alison Seward

for their valued commentaries during the drafting of this article.

Appendix

Lexicon

Anticipation: a prediction and interpretation of events, stimuli, or

environmental changes before their actual manifestation. In anxiety, the
animal may show fear-like behaviors in anticipation of events that never
occur; these fear-like behaviors seem to be triggered by trivial stimuli or
even stimuli that defy objectivation by owners.

Behavior: a speciﬁc sequence of acts with a typical postural display,

which is divided in four periods, such as the starting, operant, ending, and
refractory periods

Critical distance: the distance at which an animal shows fear aggression

(also called the critical reaction) when escape is not possible

Disorder: a speciﬁc unadaptive pathologic state described by a cluster of

signs

Emotion: an ‘‘impulse to act’’ (from the Latin emotio [e-, out

þ movere,

to move]) or a motivational state, arousal, or action tendency, such as fear,
anger, or joy. It is modiﬁed by the underlying mood, cognition, body state,
or other factors.

Hyper-: an abnormal state of excess
Hypervigilance: an abnormal state of excess vigilance, alertness, or

readiness

Impulsiveness: impetuosity, hastiness, recklessness, or overreactivity
Mood: a preliminary aﬀective state inﬂuencing emotions, cognition, or

behaviors, such as anxiety, depression, or a hyperstate

Overactivity or hyperactivity: a hyperstate expressed as abundant

and inappropriate activity, such as running about, jumping, or climbing
excessively in situations in which it is inappropriate. This may be a sign of
anxiety or hyperactivity disorder or other disorders.

Pathogeny, pathogenesis: the course and development of a disease or

disorder

Play: a nonspeciﬁc sequence of acts in which acts may be repeated,

exaggerated, uncompleted, or reordered and the sequence itself may be
terminated earlier than normal by the introduction of irrelevant activities
[11]. The play behaviors never show the emotional intensity of the normal
behaviors they mimic.

284

D. Frank, J. Dehasse / Vet Clin Small Anim 33 (2003) 269–286

Play ﬁghting: a social play that does not decide which animal wins a

resource, is not injurious, and does not lead to prolonged mutual avoidance
by the players [11].

Predatory or prey-catching, behavior, hunting: a particular sequence of

acts (behavior) leading to the capture of prey or a prey-like item and having
a variable threshold depending on physiologic drives, such as hunger and
movement. In the absence of the speciﬁc trigger, it may express itself as a
vacuum activity.

Security or ﬂight distance: the distance at which an animal oﬀers

proactive aggression to increase distance between itself and the perceived
threat or at which an animal ﬂees from a threat. This security distance varies
from one individual to another depending on the threat (eg, known,
unknown) and the mood of that individual.

Hyperepisode disorder: an abnormal variation of mood between two

temporarily stable states, such as normal and hyper.

Vacuum or overﬂow activity: an innate behavior occurring sponta-

neously in the absence of the triggering stimulation.

References

[1] Dehasse J. Le de´veloppement comportemental du jeune chaton: ontogene`se sensorielle,

affective et relationnelle. Pratique Me´dicale et Chirurgicale de l’Animal de Compagnie
1993;28:469–78.

[2] Karsh EB. The effects of early handling on the development of social bonds between cats

and people. In: Katcher AH, Beck AM, editors. New perspectives on our lives with
companion animals. Philadelphia: University of Pennsylvania Press; 1983. p. 22–8.

[3] Karsh EB, Turner D. The human-cat relationship. In: Turner D, Bateson P, editors.

The domestic cat: the biology of its behaviour. Cambridge: Cambridge University Press;
1988. p. 159–77.

[4] Karsh EB. The effects of early and late handling on the attachment of cats to people. In:

Anderson RK, Hart BL, Hart LA, editors. The pet connection, conference proceedings.
Minneapolis: University of Minnesota Press; 1983.

[5] Karsh EB. Factors inﬂuencing the socialization of cats to people. In: Anderson RK,

Hart BL, Hart LA, editors. The pet connection: its inﬂuence on our health and quality
of life. Minneapolis: University of Minnesota Press; 1984. p. 207–15.

[6] McCune S. The impact of paternity and early socialisation on the development of cats’

behaviour to people and novel objects. Appl Anim Behav Sci 1995;45:109–24.

[7] Plomin R, Defries J, McClearn G, Rutter M. Des genes au comportement. Bruxelles: De

Boeck Universite´; 1999.

[8] Immelmann K, Beer C. A dictionary of ethology. Cambridge, MA: Harvard University

Press; 1989.

[9] Beaver B. Feline behavior: a guide for veterinarians. Philadelphia: WB Saunders; 1992.

[10] Horwitz DF. Feline aggression. In: Houpt KA, editor. Recent advances in companion ani-

mal behavior problems. Ithaca: International Veterinary Information Service; 2001. Avail-
able at: http:

//www.ivis.org/advances/Behavior_Houpt/horwitz/chapter_frm.asp?LA

¼1.

[11] Barrows EM. Animal behavior desk reference. 2nd edition. Boca Raton (FL): CRC Press,

2001. p. 554.

[12] Turner D, Meister O. Hunting behaviour of the domestic cat. In: Turner D, Bateson P,

editors. The domestic cat, the biology of its behaviour. Cambridge: Cambridge University
Press; 1988. p. 111–21.

285

D. Frank, J. Dehasse / Vet Clin Small Anim 33 (2003) 269–286

[13] Heymer A. Ethological dictionary. Berlin: Parey; 1977.
[14] Schroll S. Environment enrichment for indoor cats as prevention and therapy. Practical

advice for quality of life. Proceedings of the 8th ESVCE Meeting, Granada, October 2002.
Paris: Publibook; 2002a.

[15] Dehasse J. Anxieties in cats. In: Mills DS, Heath SE, Harrington LJ, editors. Proceedings

of the First International Conference on Veterinary Behavioural Medicine. Birmingham:
Universities Federation for Animal Welfare; 1998. 114–9.

[16] Moyer KE. Internal impulses to aggression. Trans NY Acad Sci Series II 1988;31:104–14.
[17] Caston J. Psycho-physiologie, tome I. Paris: Ellipses; 1993.
[18] Overall K. Clinical behavioral medicine for small animals. St. Louis (MO): Mosby; 1997.
[19] Voith VL. Play behavior interpreted as aggression or hyperactivity: case histories. Mod Vet

Pract 1980;61:707–9.

[20] Schroll S. Miez, miez—na komm! Niebu¨ll: Videel 2001.

286

D. Frank, J. Dehasse / Vet Clin Small Anim 33 (2003) 269–286

Feline house soiling: elimination and

marking behaviors

Jacqueline C. Neilson, DVM

Animal Behavior Clinic, 809 SE Powell Boulevard, Portland, OR 97202, USA

Deposition of urine and

/or feces in inappropriate locations in the home

continues to be the most common behavioral problem for which cat owners
seek professional counsel [1–3]. House-soiling is also the primary behavioral
reason given for relinquishment of cats to shelters [4]. This is not surprising
considering the facts that the consequences of this behavior can be quite
unpleasant to human beings dwelling in the home and cleaning bills or
replacement of soiled objects can quickly escalate into a signiﬁcant expense.
This article reviews appropriate diagnosis and treatments for elimination in
inappropriate locations.

Diagnosis

Identification of culprit(s)

In multicat households, the ﬁrst challenge in diagnosis may be

identiﬁcation of the culprit(s). Sometimes, owners have only found the
evidence of inappropriate soiling and have never witnessed the actual
deposition of the urine

/feces. More commonly, the owners have witnessed a

small percentage of the actual depositions, perhaps implicating one cat but
not absolving others of guilt. There are several methods that can aid in
identiﬁcation of the culprit(s), but all have their limitations. Segregation of
cats can help to identify participants. If urine

/feces are found inappropri-

ately deposited in the segregated location of one of the cats, that cat can be a
conﬁrmed participant. The actual segregation may inﬂuence the social
dynamics of the household enough to modify the cats’ elimination behavior,
however. In fact, segregation is sometimes used as part of the treatment
program for cats with house-soiling problems [5].

Vet Clin Small Anim

33 (2003) 287–301

E-mail address:

abcjndvm@hevanet.com

0195-5616

/03/$ - see front matter

PII: S 0 1 9 5 - 5 6 1 6 ( 0 2 ) 0 0 1 2 9 - 8

Another method of identifying participants involves video monitoring.

With advances in technology, it is possible for the average owner to set up
some type of surveillance program. Cats that have unpredictable target
spots and infrequent episodes may make this option nonproductive,
however.

Finally, the administration of products to change the properties of the

urine

/feces can aid in the identiﬁcation of culprit(s). Fluorescein dye, given

either orally (50 mg of ﬂuorescein per cat) or subcutaneously (0.3 mL of
Fluorescite Injection 10% per cat) to each possible culprit in a sequential
manner may help to elucidate participants [6]. Compounding pharmacists
can create capsules containing ﬂuorescein for oral administration. Alter-
nately, one can make an oral capsule by using the ﬂuorescein-impregnated
ophthalmic strips generally used in clinical practice for identiﬁcation of
corneal irregularities. To do this, the tip of the strip that contains the
ﬂuorescein (orange-colored part of the strip) is torn off from the white part
of the strip. The orange part is then folded to ﬁt inside an empty gelatin
capsule. A total of six large ﬂuorescein strips (9 mg of ﬂuorescein per strip)
should be folded up into capsule(s) and administered orally. Although urine
from nontreated cats ﬂuoresces a yellow-green color when viewed with a
ﬂuorescent black light in a darkened room, one can distinguish urine from
cats that have received ﬂuorescein because it is a much more vibrant yellow
green when viewed with a ﬂuorescent black light for 24 hours after
administration. It is important to warn owners that if the cat targets light-
colored upholstery or carpet, the ﬂuorescein-treated urine may leave a stain
that is visible to the naked eye and resistant to cleaning. Once again, if there
is a sporadic participant or infrequent soiling, this technique may miss that
cat’s contribution.

To identify the cat depositing feces inappropriately in the home, the

owner can make small shavings of nontoxic crayons using a kitchen cheese
grater. Each cat can have one-half teaspoon of a speciﬁc color of crayon
shavings added to its moist food. For example, Cat 1 would receive green
crayon shavings, Cat 2 would receive purple crayon shavings, and Cat 3
would receive red crayon shavings mixed into canned cat food or a special
moist food treat, such as tuna. The nontoxic crayon shavings should pass
through the intestinal tract intact, and feces can then be examined for the
crayon shavings, identifying the ‘‘owner’’ of that fecal deposit. Using this
example, if the owner ﬁnds feces on the carpet with purple crayon shavings
in it, Cat 2 is a conﬁrmed participant in the problem.

Medical evaluation

When a pet presents with a house-soiling problem, it is widely accepted

that infectious, inﬂammatory, or metabolic changes could contribute to
the problem [7–10]. Any disease that inﬂuences the frequency, urgency or
quantity of urination

/defecation could easily present as inappropriate

288

J.C. Neilson

/ Vet Clin Small Anim 33 (2003) 287–301

elimination. Cystitis, renal failure, diabetes, inﬂammatory bowel disease,
and hyperthyroidism are all examples of diseases that can directly inﬂuence
elimination patterns and may result in inappropriate elimination. Cats that
experience waxing and waning disease processes, such as feline lower
urinary tract disease (FLUTD), may be especially challenging to diagnose
and treat, because the clinical signs are transient in nature [7]. Congenital,
neurogenic, and musculoskeletal causes should also be investigated as po-
tential causes of house soiling. The prevalence of radiographic degen-
erative joint disease (DJD) in cats older than 12 years of age was found at an
astounding incidence of 90% in one study [11]. These cats may have
discomfort in accessing a litter box or positioning in the litter box and may
therefore select an alternative location for elimination. In managing
elimination problems, it is imperative that the veterinarian considers the
entire animal and its health status. It is important to recognize that medical
and behavioral conditions may not be mutually exclusive. A medical
problem may initiate a problem, but after resolution of the medical problem,
the pattern may be sustained as a result of behavioral reasons.

Historical information

A thorough history is important in the diagnosis of any disease process

but is paramount in behavioral medicine. Obtaining an accurate picture of
the nature of the problem (urine, feces, or both), the frequency of the
problem, the duration of the problem, and the location (eg, vertical versus
horizontal deposition, location within the home) of the problem all aid in
the proper diagnosis. Social interactions between people and other animals
are important to gather and interpret with respect to the elimination issue.
Information about litter boxes, including number, locations, size, cleaning
routine, litter type, and box type, is necessary to evaluate the case properly.
House calls can be enlightening in problem elimination cases, because a
great deal of information can be gathered that might have been more
challenging to extract via interview alone.

Behavioral diagnosis

Once medical problems have been investigated and addressed, the

behavioral reasons for the unacceptable elimination should be pursued.
There are two large diagnostic categories that should be considered: mark-
ing behavior and inappropriate elimination (Fig. 1). Within each of these
categories, a more speciﬁc diagnosis may be achieved based on under-
lying motivations [12].

Marking

Marking is done with the intention of communication rather than to void

the bladder

/bowels. Although deposition of urine and deposition of feces

289

J.C. Neilson

/ Vet Clin Small Anim 33 (2003) 287–301

have both been described as methods of marking, urine marking is the
predominant form of marking seen in the domestic cat [13]. The classic
marking cat is one that backs up to a vertical surface with the tail held
vertically and twitching while emitting a spray of urine on the vertical target.
The back feet of the cat may be treading during this activity, and the eyes are
often half closed [5]. Cats may also mark by depositing urine on a horizon-
tal surface from a squatting position. With both vertical and horizontal
marking, a small quantity of urine is typically deposited rather than a full
bladder voiding. Both male and female cats can engage in urine marking,
but male cats are more likely to exhibit the behavior [13]. In one study of
feral cats dwelling in an outdoor location, urine marking tended to occur
along well-traveled paths instead of the territory perimeter, suggesting that
the intent of the mark may not be to deter intruders but to provide in-
dividual and temporal information [13]. In the home, urine marks are
usually found in socially signiﬁcant areas, such as near windows or doors
or on items that have the scent of a particular person and

/or animal. The

substrate under the cat’s feet (eg, carpet, tile, linoleum) while the cat is
depositing the mark does not seem to be important. This is in contrast to
cats with inappropriate elimination, which often exhibit a pattern of similar
substrate choice for deposition of the urine

/feces (Table 1). Sexual status,

social status, territorial disputes, arousal, and anxiety can all inﬂuence the
likelihood of urine marking [5].

Fig. 1. Visual depiction of diagnosis.

Table 1
Characteristics of marking versus inappropriate elimination

Marking

Inappropriate elimination

Posture

Stand or squat

Squat

Quantity of urine

Small

Medium to large

Litter box use

Normal

Reduced or absent

Location

Socially signiﬁcant locations

Acceptable substrates

290

J.C. Neilson

/ Vet Clin Small Anim 33 (2003) 287–301

Elimination outside the litter box

In contrast to marking, elimination is about evacuating the contents

of the bladder and

/or bowels. The cat chooses an elimination spot other

than the litter box for this activity. The natural posture for voiding the
bladder

/bowels is the squat; therefore, the depositions are usually found

on horizontal surfaces. Because the objective is to evacuate the
bladder

/bowels, a large quantity of urine and/or a pile of feces is usually

observed. Aversions or preferences are often the underlying motivation for a
change in elimination spot. Common litter box aversions include inadequate
litter box cleanliness, litter box location, litter box size, litter substrate, and
litter box type. Common preferences include location and substrate pre-
ferences. It is common to ﬁnd that the cat returns to either a certain lo-
cation or a particular type of substrate (eg, always on carpet) for its
inappropriate elimination. Many cats still use the litter box for some
elimination. It is important to recognize that there may be diﬀerent ini-
tiating and maintenance factors for this behavior. For example, a cat may
elect to eliminate on the carpet when the litter box gets dirty (litter box
aversion caused by inadequate cleanliness). Even when the owners clean the
litter box, however, that cat may continue to eliminate on the carpet because
it now has a location and

/or substrate preference.

Treatment

Marking

Once a diagnosis has been established, treatment can be properly

targeted. The ﬁrst treatment consideration for cats that urine mark should
be neutering

/spaying the oﬀending cat. Although neutering/spaying can

signiﬁcantly decrease urine marking, it does not always control this normal
communication behavior. In one study, approximately 90% of male cats
signiﬁcantly reduced or stopped urine marking after neuter surgery [14].
Other surgical interventions are not usually performed to treat urine
marking; however, two techniques, bilateral ischiocavernosus myectomy
and olfactory tractotomy, have been described in the literature as effective at
reducing or controlling urine marking [15,16].

Because the frequency of urine marking increases when the population of

cats increases, restricting the number of cats in a given household or area
may help to prevent or treat the problem [5]. Although population reduction
may help, most owners are unwilling or unable to do this because of pet
attachment and

/or lack of control of the feline population.

If an underlying stressor can be identiﬁed, it should be modiﬁed or

removed. Stressors can vary widely from the addition of a new baby to a
change in the owner’s work schedule. A thorough historical proﬁle can help
to elucidate potential stressors. Agonistic social interactions between
resident or visiting cats is often a contributing factor to marking and

291

J.C. Neilson

/ Vet Clin Small Anim 33 (2003) 287–301

should be addressed. A survey by Borchelt and Voith [17] reported that 70%
of cohabitating cats had an occasional ﬁght and that more than 80%
swatted

/hissed at each other. A simple step to decrease social stress includes

the creation of an ‘‘environment of plenty’’ in multicat households. This
involves providing an abundance of resources so that two cats never are
forced to cross paths to get to a valuable resource. Valuable resources for
cats include single cat–sized resting perches, food, water, and litter boxes. In
households with multiple cats, it is important to use vertical space to help
create a larger relative living space. This is easily done by installing shelv-
ing at different vertical heights or cat trees with multiple perches around
the home (Fig. 2). If there is an offensively aggressive or ‘‘bully’’ cat in the
household, that cat can be ﬁtted with a belled collar to give the other cats
advanced warning as to that cat’s movements, allowing them to avoid the
bully if they so desire. Periodic segregation of cats in a multicat household

Fig. 2. Multiple vertical perches help to create an environment of plenty in multiple cat
households.

292

J.C. Neilson

/ Vet Clin Small Anim 33 (2003) 287–301

may also help to decrease social tension. The facial pheromone Feliway
(Veterinary Product Laboratories, Phoenix, AZ) has been shown to decrease
stress in clinical settings; thus, treatment in the home may also have
beneﬁcial stress-reducing properties [18]. For more serious social conﬂicts,
complete segregation with gradual reintroduction may be necessary. If a cat
that dwells indoors only is getting upset by viewing outdoor cats through the
windows, remotely activated pet deterrent devices like the Scarecrow (a
motion-activated sprinkler, Contech Electronics, Victoria, B.C., Canada)
can be placed in the yard to deter visitor cats. Alternately, the indoor cat’s
view of visitor cats can be blocked by closing shades or putting poster board
on the lower part of the window.

A study examined the eﬀects of environmental management on the

frequency of urine marking [19]. Forty-seven cats exhibiting vertical urine
marking were enrolled in the study. Owners collected baseline frequency
of urine marking for 2 weeks without making any changes in home manage-
ment. Owners were then given instructions to clean urine-marked spots
with an enzymatic cleanser (Anti-Icky-Poo; Mister Max Quality Products,
Lakeside, CA) for 2 weeks. Additional instructions included providing one
litter box per cat plus one additional litter box, scooping the box daily,
and changing the box weekly. The number of urine marks recorded dur-
ing the baseline phase (11.7

^1 marks) was signiﬁcantly higher than the

number of urine marks recorded during the environmental management
phase (9.7

^1.3 marks). This indicates that environmental management

should be considered as part of the treatment for feline urine marking.

Encouraging other forms of marking behavior such as bunting (facial

rubbing) and scratch marking should be part of the treatment plan for the
urine-marking cat (Fig. 3). The facial pheromone Feliway is applied to the
environment on previously urine-marked spots and

/or prominent locations

on a daily basis until bunting is observed. There have been variable results
reported with this treatment, with some of the better results suggesting an
80% to 90% success rate [20]. Even if lower efﬁcacy is more realistic, it is
still a valuable treatment option with no apparent negative side effects.
One necessary caution is to avoid applying Feliway when treating with
cleansers, because the pheromone may be altered by the cleansing product.
To encourage scratching marking, scratch pads or posts should be placed in
prominent locations around the house (Fig. 4).

Drug therapy can be considered for the treatment of urine marking in

cats. Drug therapy has a history of being used to help control urine
marking, and recent studies have furthered our knowledge about the most
successful treatments [21–24]. To date, there is no US Food and Drug
Administration (FDA)–approved drug therapy for the treatment of urine
marking in cats; thus, off-label treatment should be discussed with owners
before instituting therapy [25]. For many years, progestins (megestrol
acetate) were used as a treatment for urine marking in cats. Low treatment
efﬁcacy, serious side effect proﬁles, and better drug options have rendered

293

J.C. Neilson

/ Vet Clin Small Anim 33 (2003) 287–301

progestins nearly obsolete, however [5,32]. Benzodiazepines, although often
effective [26,27], also had problematic side effect proﬁles as well as the
potential for abuse and caused acute hepatic failure in some cats [28].
Serotonin-enhancing medications are currently the most promising class
of drugs in the treatment of urine marking. Several studies have been
conducted to evaluate treatment efﬁcacy of the serotonin-enhancing
medications; however, to date, most of the studies are not designed with a
rigorous double-blind placebo-controlled model. The exception to this was a
double-blind placebo controlled study evaluating the efﬁcacy of ﬂuoxetine
(Prozac) dosed at 1 mg

/kg/d in the treatment of urine-marking behavior in

cats [24]. Seventeen cats completed the study, and there was a signiﬁcant
reduction in the weekly number of vertical sprays in the cats in the drug
group compared with the cats in the placebo group. All the cats receiving the
drug experienced a greater than 90% decline in urine marking. Other clin-
ical trials that have not employed a double-blind placebo-controlled model
have looked at the efﬁcacy of clomipramine (Clomicalm, Anafranil) and

Fig. 3. A cat exhibiting bunting where the facial pheromone Feliway was sprayed.

294

J.C. Neilson

/ Vet Clin Small Anim 33 (2003) 287–301

buspirone (Buspar), and they indicate efﬁcacy at controlling urine marking
[21–23]. Other serotonin-enhancing medications like paroxetine (Paxil) are
also being used to treat urine marking; however, aside from positive
anecdotal reports, data are currently lacking on their efﬁcacy. Side effects
are always a concern when instituting drug therapy, and some of the more
frequently reported side effects are noted in Table 2. It is important to

Fig. 4. Cat using a scratching pad for visual and scent marking.

Table 2
Drug treatments for urine marking

/spraying

Drug brand name

Drug class

Feline dose

Side effects

Buspirone (Buspar)

Azapirone

5–10 mg per

cat BID

Increased intercat

aggression (10%)

Amitriptyline

(Elavil)

Tricyclic

antidepressant

5–10 mg per cat

SID to BID

Sedation,

anticholenergic eﬀects

Clomipramine

(Anafranil

Clomicalm)

Tricyclic

antidepressant

0.5 mg

/kg SID or

2.5–5 mg per
cat SID

Sedation,

anticholinergic eﬀects

Fluoxetine (Prozac)

Selective serotonin

reuptake inhibitor

0.5–1 mg

/kg SID

Inappetence,

constipation,
lethargy

Paroxetine (Paxil)

Selective serotonin

reuptake inhibitor

2.5–5.0 mg per cat

SID to EOD

Lethargy,

inappetence,
constipation,
urinary retention

Diazepam (Valium)

Benzodiazepine

0.2–0.4 mg

/kg

SID to BID

Sedation, acute

hepatic failure

Abbreviations

: BID, twice daily; EOD, every other day; SID, once daily.

Partial list of side eﬀects.

295

J.C. Neilson

/ Vet Clin Small Anim 33 (2003) 287–301

remember that the full extent of side effects may not yet be identiﬁed
because of the lack of large-scale controlled clinical trials.

Complimentary medicine may also oﬀer some treatment options for urine

marking, such as acupuncture or naturopathic remedies. Once again, clin-
ical data are lacking to support eﬃcacy, but future studies may elucidate
treatments.

Elimination outside the litter box

The treatment for elimination outside the litter box focuses on two main

components: identifying and providing the cat with its ‘‘ultimate’’ litter box
and making inappropriate target spots unattractive or unavailable. The
diagnosis should help to direct treatment. For example, if the cat was
diagnosed with house soiling because of inadequate litter box cleanliness,
providing a scrupulously clean litter box is indicated in the treatment plan.
If the issue seems to be a location preference, however, identifying and
providing the cat with a litter box at or near the preferred location would be
the ﬁrst course of action.

Because most cats prefer clean boxes, it is always appropriate to discuss

litter box hygiene with owners. They should oﬀer adequate litter boxes;
a general rule is to oﬀer the same number of boxes as there are cats plus
an additional box. These boxes should be scooped at least once daily and
dumped and washed with soap and water on a routine basis. If the cat is
using a clumping or absorbent litter, washing the box once every 1 to 4
weeks should suﬃce. If the cat is using a nonclumping

/nonabsorbent litter,

washing weekly is suggested. Some cats require more frequent cleaning, and
others are less particular. It is suggested that the owners smell the box for
residual odor after washing and dry the litter box, because plastic retains
odors and needs periodic replacement.

Litter preferences vary between individual cats; however, in studies,

most cats prefer unscented and ﬁnely particulate litter material as is typical
of the clumping type litters compared with other litter options [29]. Large
particulate matter is generally not preferred by cats. To help determine the
attractiveness of the new silica (‘‘pearl’’) litters, a preference study was
conducted on shelter cats [30]. Fifty-four shelter cats were given two novel
litter options (clumping and pearl) for a 12-hour overnight period, and
use was recorded. A total of 74 uses were recorded: 58 (36 urination

/22

defecation) were in clumping litter, 13 (11 urination

/2 defecation) were in

pearl litter, and 3 (1 urination

/2 defecation) were out of the litter box. These

results suggest that most cats prefer a clumping type litter compared with
silica litters for elimination. To determine a particular cat’s preference, a
litter box cafeteria can be set up. The cat is offered a selection of litters, and
preference is identiﬁed by use (Figs. 5–8). In challenging cases where no
commercial litter seems to be accepted by the cat, offering nontraditional
options, such as cloth baby diapers, carpet, newspaper, or sand, can

296

J.C. Neilson

/ Vet Clin Small Anim 33 (2003) 287–301

sometimes result in treatment success. If the cat can be attracted back into
the box with these nontraditional ‘‘litters,’’ if the owner desires, he or she
can try to add a commercial litter to the nontraditional litter gradually.

The litter box cafeteria can also be used to identify type of box preference

(Fig. 9). In most cases, covered litter boxes are discouraged because they
trap odors and are small and owners may be less likely to scoop routinely
(‘‘out of sight is out of mind’’). Box size may be an underappreciated cause
of litter box aversion. Many boxes are considerably smaller than the cat,
perhaps resulting in discomfort or awkwardness in posturing for elimina-
tion (Fig. 10). Large plastic storage boxes or plastic kiddy pools may make
good alternative litter boxes for large-sized cats. Novel box types intermit-
tently hit the market, often designed to reduce the need for owner
scooping

/cleaning. Although they may be a solution for some cats, they

Fig. 5. Cat approaches the litter box cafeteria oﬀering two diﬀerent litter types.

Fig. 6. Cat selects the pearl type litter.

297

J.C. Neilson

/ Vet Clin Small Anim 33 (2003) 287–301

should be fully investigated for potential drawbacks, such as mechanical
problems, size, and noise associated with automatic cleaning machinery.

Inappropriately soiled spots should be well cleaned. One study showed

that enzymatic cleansers did the best job at removing urine odor, at least to
the human nose [31]. Professional cleaning or replacement of soiled items is
often necessary to remove residual urine odor completely.

Making inappropriate target spots unattractive or unavailable is best

instituted after the owner has started to pursue the quest for the ultimate litter
box; otherwise, the cat may just select an alternative inappropriate location.
Ideas for making target spot(s) unattractive include the placement of double-
sided sticky tape, an upside down vinyl runner (nub side up), aluminum foil,
heavily scented potpourri, or ultrasonic deterrent devices at the inappropriate
spot(s). There are unlimited possibilities to make a location unattractive, but
the welfare of the animal should always be a priority.

For households where it is diﬃcult to make chosen locations unavailable

or unattractive, or for cats that seem to have a problem during speciﬁc

Fig. 7. Cat urinating in the pearl type litter.

Fig. 8. Second cat harassing cat that is eliminating in box.

298

J.C. Neilson

/ Vet Clin Small Anim 33 (2003) 287–301

periods, conﬁnement to a smaller area with a litter box cafeteria may be
necessary. With consistent litter box use, freedom can then be gradually
restored.

Drug therapy is generally not indicated for cats that eliminate outside

their litter box. An exception to this may be the cat that is too frightened or
anxious to use the litter box, perhaps as the result of a particularly negative

Fig. 9. Cafeteria for selection of preferred box type. All cats eventually selected to eliminate in
the uncovered box.

Fig. 10. Note relative size of litter box compared with size of cat. Standard litter boxes are too
small for this cat.

299

J.C. Neilson

/ Vet Clin Small Anim 33 (2003) 287–301

experience when using the box or a social conﬂict with other cats in the
household. Exhausting other management options is suggested before
resorting to drug therapy in elimination cases, because many cats respond
without medication. For example, if the cat had a particularly negative
experience in a box, try oﬀering the cat another box with diﬀerent attributes
(eg, diﬀerent style of box, diﬀerent location of box), because this may
overcome the negative box association. The cat with social conﬂicts may
beneﬁt from creating an environment of plenty, because it then does not
have to cross the path of another cat to get to the litter box.

It is important to note that punishment is not a recommended treatment

for elimination problems. Correct application of punishment requires that it
be delivered during or immediately after the behavior occurs and every time
the behavior occurs and that it be eﬃcacious at inhibiting the unwanted
behavior without hurting the animal. Because it is unlikely that any of these
criteria can be met, punishment does not solve the problem and may even
intensify it.

Summary

Although challenging at times, managing cats with elimination issues can

be rewarding. The frustration in management of these cases is often a result
of the lack of a proper diagnosis and random treatment application. A
systematic approach to these cases should help to achieve treatment success.
Also, new information should lead to further advancements in treatment.

References

[1] Beaver BV. Housesoiling by cats: a retrospective study of 120 cases. JAVMA 1989;25:

631–637.

[2] Borchelt PL, Voith VL. Elimination behavior problems in cats. Compend Contin Educ

Pract Vet 1986;8:197–207.

[3] Olm DD, Houpt KA. Feline house-soiling problems. Appl Anim Behav Sci 1988;20:

335–345.

[4] Salman MD, Hutchison J, Ruch-Gallie R, et al. Behavioral reasons for relinquishment of

dogs and cats to 12 shelters. J Appl Anim Welfare Sci 2000;3:93–106.

[5] Overall KL. Feline elimination disorders. In: Clinical behavioral medicine for small

animals. St. Louis: Mosby; 1997. p. 160–94.

[6] Hart BL, Leedy M. Identiﬁcation of source of urine stains in multi-cat households.

JAVMA 1982;180:77–8.

[7] Bufﬁngton TCA, Chew DJ, Kendall MS, et al. Clinical evaluation of cats with

nonobstructive urinary tract diseases. JAVMA 1997;210:46–50.

[8] Chew DJ, Bufﬁngton TCA, Kendall MS, et al. Amitriptyline treatment for severe recur-

rent idiopathic cystitis in cats. JAVMA 1998;213:1282–6.

[9] Dowling PM. Potential therapies for recurrent idiopathic cystitis in cats. Veterinary

Medicine 2000;95:512–5.

[10] Lekcharoensuk C, Osborne CA, Lulich JP. Epidemiologic study of risk factors for lower

urinary tract diseases in cats. JAVMA 2001;218:1429–35.

300

J.C. Neilson

/ Vet Clin Small Anim 33 (2003) 287–301

[11] Hardie EM, Roe SC, Martin FR. Radiographic evidence of degenerative joint disease in

geriatric cats:100 cases (1994–1997). JAVMA 2002;220:628–32.

[12] Halip JW, Vaillancourt JP, Luescher UA. A descriptive study of 189 cats engaging in

inappropriate elimination behaviors. Feline Pract 1988;26:18–21.

[13] Feldman HN. Methods of scent marking in the domestic cat. Can J Zool 1994;72:1093–9.
[14] Hart BL, Barrett RE. Effects of castration on ﬁghting, roaming and urine spraying in

adult male cats. JAVMA 1973;163:290–2.

[15] Hart BL. Olfactory tractotomy for control of objectionable urine spraying and urine

marking in cats. JAVMA 1981;179:231–4.

[16] Komtebedde J, Hauptman J. Bilateral ischiocavernosus myectomy for chronic urine

spraying in castrated male cats. Vet Surg 1990;19:293–6.

[17] Borchelt PL, Voith VL. Aggressive behavior in cats. In: Readings in companion animal

behavior. Trenton: Veterinary Learning Systems; 1996. p. 208–16.

[18] Grifﬁth CA, Steigerwald ES, Bufﬁngton T. Effects of synthetic facial pheromone on

behavior of cats. JAVMA 2000;217:1154–6.

[19] Pryor PA, Hart BL, Bain MJ, Cliff KD. Causes of urine marking in cats and effects of

environmental management on frequency of marking. JAVMA 2001;219:1709–13.

[20] Hunthausen W. Evaluating a feline facial pheromone analogue to control urine spraying.

Veterinary Medicine 2000;95:151–5.

[21] Dehasse J. Feline urine spraying. Appl Anim Behav Sci 1997;52:365–71.
[22] Kroll T, Houpt KA. A comparison of cyproheptadine and clomipramine for the treatment

of spraying cats. In: Overall KA, Mills DS, Heath SE, Horwitz D, editors. Proceedings of
the Third International Congress on Veterinary Behavioural Medicine. Hertfordshire, UK:
Universities Federation for Animal Welfare; 2001. p. 184–5.

[23] Landsberg GM. Effects of clomipramine on cats presented for urine marking. In: Overall

KA, Mills DS, Heath SE, Horwitz D, editors. Proceedings of the Third International
Congress on Veterinary Behavioural Medicine. Herfordshire, UK: Universities Federation
for Animal Welfare; 2001. p. 186–9.

[24] Pryor PA, Hart BL, Cliff KD, et al. Effects of a selective serotonin reuptake inhibitor on

urine spraying behavior in cats. JAVMA 2001;219:1557–61.

[25] Davis LE. Adverse drug reactions. In: Ettinger JS editor. Textbook of veterinary internal

medicine: diseases of the dog and the cat. 3rd edition. Philadelphia: WB Saunders; 1989.
p. 499–510.

[26] Cooper L, Hart BL. Comparison of diazepam with progestins for effectiveness in sup-

pression of urine spraying behavior in cats. JAVMA 1992;200:797–801.

[27] Marder AR. Psychotropic drugs and behavioral therapy. Vet Clin North Am Small Anim

Pract 1991;21:329–52.

[28] Center SA, Elston TH, Rowland PH, et al. Fulminant hepatic failure associated with

oral administration of diazepam in 12 cats. JAVMA 1996;209:618–25.

[29] Borchelt PL. Cat elimination behavior problems. Vet Clin North Am Small Anim Pract

1991;21:257–64.

[30] Neilson JC. Pearl vs. clumping: litter preference in a population of shelter cats [abstract].

In: Abstracts from the American Veterinary Society of Animal Behavior, Boston, 2001.
p. 14

[31] Beaver BV, Terry ML, LaSagna CL. Effectiveness of products in eliminating cat urine

odors from carpet. JAVMA 1989;194:1589–91.

[32] Hart BL. Objectionable urine spraying and urine marking in cats: evaluation of progestin

treatment in gonadectomized males and females. JAVMA 1980;177:529–33.

301

J.C. Neilson

/ Vet Clin Small Anim 33 (2003) 287–301

Diﬀerential diagnosis and management

of human-directed aggression in dogs

Ilana R. Reisner

, DVM, PhD

Department of Clinical Studies and Behavior Clinic, School of Veterinary Medicine,

University of Pennsylvania, 3900 Delancey Street, Philadelphia, PA 19104, USA

Aggression is perhaps the most challenging canine behavioral problem

presented to behavioral specialists and continues to be the most common
reason for referral to veterinary behaviorists [1], at least in part because of
the emotional stress and risks of living with a biting dog. Approximately 4.7
million dog bites occur annually in the United States [2]; in spite of this great
number, bites are largely underreported to public health authorities [3].
Contrary to the notion that biting dogs tend to be feral and unfamiliar, 35%
of reported bites in one large study were attributed to the dog of a friend,
acquaintance, neighbor, or relative, whereas 30% were delivered by the
family’s or victim’s own pet [4]. More than one third of these occurred in
the context of day-to-day activities, such as petting, feeding, and play [4],
further supporting the fact that biting dogs are often family pets in family
situations.

Children are the most frequent targets of reported dog bites; almost half

of victims are children younger than 12 years of age [5]. A survey of 3238
Pennsylvania schoolchildren revealed that by the twelfth grade, a startling
46% had been bitten [6]. Younger children were more likely to be bitten in
the face and by the family dog in their own homes [7,8]. Dog bite injuries to
children can do damage through laceration or life-threatening blunt injury
[9]. Adults, of course, are also victims of dog bites by their own pets but may
be less likely to seek attention from emergency rooms [10]. Dog bites may
occasionally result in fatal injury. Between 1979 and 1998, there were 227
reports of fatal attacks, most of which (75%) were committed by dogs on
their own property [11].

Clearly, biting is not an insigniﬁcant risk in living with a companion dog,

even with a pet whose job description does not include protection. In spite

Vet Clin Small Anim

33 (2003) 303–320

E-mail address:

reisner@vet.upenn.edu

0195-5616/03/$ - see front matter

PII: S 0 1 9 5 - 5 6 1 6 ( 0 2 ) 0 0 1 3 2 - 8

of the statistics, owners seeking help probably represent only a small
percentage of the total number of people whose pets have a history of biting.

Diagnosis

General principles

Aggressive behavior includes threats and unsuccessful attempts to injure

as well as biting. Threatening behavior includes stiﬀening, staring, growling,
baring teeth or snarling, lunging and charging, and aggressive barking. For
example, a dog that growls when it is approached while eating is probably
exhibiting food-related aggression, regardless of whether it lunges at or bites
an interloper.

The dog owner who, perhaps defensively, does not consider his or her pet

to be vicious (or aggressive) may misinterpret even biting, which is often
euphemized as ‘‘nipping.’’ The author had one client report that her dog,
when petted while eating, responded by ‘‘purring.’’ Dog owners who seek
help for behavioral problems tend to be deeply attached to their pets; it is
therefore important to consider the emotional consequences of living with
a biting dog. The clinician might be sensitive to terms like ‘‘aggressive,’’
perhaps explaining that it refers to deliberately threatening or injurious
behavior rather than to the dog’s character.

When a veterinarian is presented with a dog aggressive to people, the

clinical convention is to identify the presumed motivation for biting.
Emphasis should be placed, however, on the concept of presumption; all
that is known with certainty is that the dog has bitten. It is then up to the
clinician to try to determine the reasons for biting so that, ﬁrst, the risk of
future bites can be eliminated or reduced and, second, management of the
aggression can be speciﬁcally targeted to the motivation. A fearful dog, for
example, might be conditioned to behave more calmly in the face of the fear-
inducing stimulus, whereas a territorially defensive dog might learn to
accept the ‘‘trespass’’ of visitors. It is interesting to note that the man-
agement of diﬀerent categories of problem aggression is similar regard-
less of the presumed motivation or diagnosis.

Diagnosis

The clinical diagnosis of problem aggression may be classiﬁed by its

target (eg, family member or stranger) or by its apparent function. Although
target-based categories are more objective, classiﬁcation by function may
be more practical in the clinic, because it can be argued that successful
management of aggression depends more on its reasons than on its choice of
victim. For example, a dog may bite a child because children are unfamiliar
and therefore frightening, or it may bite because the child is at eye level and
is perceived to be competing as an equal for some highly valued resource.

304

I.R. Reisner / Vet Clin Small Anim 33 (2003) 303–320

In each case, the child is the target of aggression; however, whereas
management of the fearful dog involves safe and graduated exposure to
children while the dog is rewarded for acting unafraid, treatment of the
conﬂict-minded dog often involves modiﬁcation of his social status so that
the dog assumes a deferential role (most safely to an adult).

In an eﬀort to identify the most eﬀective preventive or intervention

strategies, categories of aggression have been suggested. Classiﬁcation
schemes have been inconsistent in terms of target, function, or the con-
tribution of disease [12–24], however, and interrater reliability has not been
addressed. Still, categorization is important because of diﬀerences in risk
factors for and prevention or management of biting. The prognosis may
also diﬀer among diﬀerent categories of aggression. Caution or circum-
spection is therefore advised in the diagnosis of aggression, and emphasis
should instead be placed on the safe management of aggression, regardless
of its cause. The following are diagnostic categories of aggressive behavior
as currently used at the Behavior Clinic of the Veterinary Hospital of the
University of Pennsylvania.

Fear-related aggression

Fear-related or defensive aggression is a normal reaction to threatening

stimuli, particularly when the dog feels trapped and cannot escape. Fear-
related aggression may be a particularly diﬃcult problem when the dog is
frightened by subtle or unintentionally threatening stimuli. Although it may
be understandable that a dog would bite when physically punished, it makes
less sense to the uneducated dog owner that bending over the dog may elicit
growling or biting simply in response to the owner’s provocative posture.
Fearfulness and fear-related biting may be seen in dogs of any age, includ-
ing young puppies, and in dogs of either sex. Neutering should have no
signiﬁcant eﬀect on fearfulness, because fear is not a sexually dimorphic
trait. Some breeds may be predisposed to fearful (or ‘‘sensitive’’) behavior,
including but certainly not limited to German Shepherd Dogs, Australian
Shepherds, and Border Collies. (It is interesting to note that these are all
members of the herding group and considered to be among the most
intelligent and trainable breeds.)

There may be an important distinction between the dog responding

with fear to a speciﬁc stimulus and the dog exhibiting anxiety or excessive
‘‘worry’’ in anticipation of some imagined stimulus. The roots of anxiety
may be nonspeciﬁc and unrelated to particular stimuli. Anxious dogs are
predisposed to fear-related aggression when something does make them
uncomfortable; perhaps most important, anxiety may lower the threshold
for aggression of any classiﬁcation. It is often included in the diﬀerential
diagnosis of aggression. The body language associated with fear or anxiety
includes exaggerated watchfulness and ‘‘scanning’’ of the environment,
panting, trembling, pupillary dilatation, eﬀorts to hide or escape,
piloerection, and lip-licking. Dogs may also exhibit anxiety in contexts

305

I.R. Reisner / Vet Clin Small Anim 33 (2003) 303–320

unrelated to aggression, such as separation anxiety or noise phobia (eg,
extreme fear of thunderstorms).

Owner-directed aggression (the behavior formerly known as
dominance-related aggression)

Dominance-related aggression has, by conventional deﬁnition, referred

to aggression directed to human family members in contexts related to
competition over resources or perceived social rank [25,26]. The term has
recently been the subject of debate. Anecdotally, the appearance of such
aggressive behavior is anxious rather than self-conﬁdent; dogs presented to
veterinary behaviorists for owner-directed aggression are frequently quite
nervous and ‘‘needy,’’ chronically seeking attention from their human com-
panions. Furthermore, the contention that such dogs are attempting to dom-
inate their owners may lead to misguided attempts to dominate the dog
in turn when the dog may already be uncomfortable about interactions with
family members.

Owner-directed aggression is most likely to be seen in young adult dogs

as they become behaviorally mature at approximately 1 to 3 years of age
[25]. Although the behavior is seen more in male dogs than in female
dogs [16,20,25,27,28], either sex can be presented. Purebred dogs may be
overrepresented among dogs diagnosed with owner-directed (dominance-
related) aggression [28].

Aﬀected dogs may bite when competing (with familiar people) for highly

valued resources, such as food, a resting place or bed, or a favored family
member. Even dogs that appear unconcerned with their daily kibble may
become reactive around human food or trash. A resting place of contention
may be a spot on the bed or sofa, under a table, or in a crate; there is
wisdom in the adage ‘‘let sleeping dogs lie,’’ because it is not uncommon for
a dog to be reactive when nudged, pushed, petted, hugged, or simply
approached while sleeping or resting. Dogs that threaten one family mem-
ber for approaching (hugging, dancing with, wrestling) another may be
‘‘protecting’’ one member of the social group from another—an alliance
that may change in diﬀerent situations.

More diﬃcult for pet owners to understand are the subtle social or postural

provocations that may present a threat or challenge to dogs. For example, one
dog may assert himself by standing over another; human beings who behave
similarly by bending down to hug a dog may ﬁnd themselves bitten. Assertive
physical manipulations, such as restraining a dog by the collar or toweling his
wet feet, may elicit an argument, but even benign interactions like petting can
result in aggression. A dog may even solicit attention only to appear to change
its mind midstream, biting the hand that strokes its head.

It is important to note that fear and anxiety may play a role in owner-

directed aggression, particularly when interactions are associated with pain
(or the anticipation of pain (eg, when a dog is physically punished or a
painful condition like otitis externa or osteoarthritis is present). At the

306

I.R. Reisner / Vet Clin Small Anim 33 (2003) 303–320

Behavior Clinic of the Veterinary Hospital of the University of Pennsylvania,
the diagnosis of owner-directed or dominance-related aggression is
usually associated with some degree of fearful or anxiety-related behavior.

A survey of dogs that have bitten their owners revealed risk factors that

included aggression over food and sleeping in the owner’s bed in the ﬁrst
2 months of ownership as well as a history of dermatologic disease (the latter
ﬁnding suggesting that chronic discomfort or pain may lead to irritability)
[29]. It was also noted that dogs with a history of dominance-related or
possessive aggression ‘‘had signiﬁcantly higher fear scores, as they were
reported to be afraid of more stimuli by their owners

…and were more likely

to be described by owners as being generally fearful’’ [10], again, an
unexpected characteristic in dogs labeled dominant-aggressive.

Food-related aggression

Food-related aggression may exist in the absence of fear-related or other

owner-directed aggression. Although food-related aggression in puppies
may be a risk factor for future owner-directed or conﬂict-related aggression,
it may also occur independently. Dogs may protect highly valued food items
(including rawhide chews) or even nonfood items, such as toys or shoes.

Territorial aggression

Territorially aggressive dogs protect the home territory of their social

group from unfamiliar people or animals. Like dogs exhibiting owner-
directed aggression, territorially aggressive dogs are typically presented at
behavioral maturity. Dogs of either sex may be seen between 1 and 4 years
of age, although more male dogs than female dogs are presented for biting
unfamiliar people in territorial contexts. Thus, although more male dogs
than female dogs are presented for biting owners in contexts of competition
or social conﬂict and territorial aggression, fear-related aggression does not
seem to have a sex predilection.

Territorial defense may occur in the home or yard, in the car, or even at the

end of a lead being held by the owner. A great many dogs that bite strangers on
their property (‘‘friendly’’ visitors as well as trespassers) seem to be motivated
primarily by fear. Again, as with owner-directed aggression, it is diﬃcult and
perhaps unrealistic to distinguish one function of aggression from another or,
especially, to tease fearful behavior from other drives to bite.

A disturbing number of severe attacks reported to public health au-

thorities are inﬂicted by dogs tied outdoors for long periods [30,31]. It is
unclear whether these dogs are motivated by territorial defense, fear, or
other reasons, but it would make sense that territorial aggression plays at
least some role in these incidents.

Predatory aggression

Predatory behavior is a normal canine trait but has been inhibited or

enhanced through artiﬁcial selection. Predatory behavior may also be a

307

I.R. Reisner / Vet Clin Small Anim 33 (2003) 303–320

learned trait in dogs predisposed to chasing or grabbing small prey.
Aggression obviously becomes a concern when it carries the risk of injurious
or fatal attacks toward people or other animals. Predatory aggression to
people is relatively rare [32], but precautions are advised with newborn infants
and the elderly, each of which group is at high risk for fatal attacks [33].

The role of anxiety

Anxiety contributes to the development and expression of aggressive

behavior. When a detailed behavioral history is possible, dog bites can often
be traced to uneasiness and visible anxiety or to a stressful situation oc-
curring just before the event. Regardless of the functional classiﬁcation of
aggression, anxiety may stem from a lack of predictability in the social or
physical environment or from some other chronic environmental stress;
when challenged, chronically stressed dogs are predisposed to uncertainty,
excitability, and aggression [34]. For example, dogs that bite owners may do
so in contexts related to social dominance, but their motivation may be
based almost entirely on the anticipation of a threat—in other words, on
social anxiety.

Careful questioning of owners and observation of their dogs can help

to identify anxiety-related behavior, such as hypervigilance, environment
scanning, yawning, pupillary dilatation, lip-licking, frequent swallowing,
panting, and trembling. In addition, a detailed behavioral history may
reveal other manifestations of anxiety, such as separation anxiety or noise
phobia [35].

Owners who understand that biting may stem from a state of ‘‘worry’’ are

also less likely to respond with punishment or rough handling, which are
responses likely to increase anxiety and possibly to increase the risk of biting.

Management of aggression

The treatment of behavioral problems in general and of canine aggression

in particular does not tend to follow a ‘‘cookbook’’ format, because each
case is unique and includes a tremendous amount of individually deﬁned
characteristics. Among these characteristics are historical behavioral
information, the medical history of the patient, family makeup, and the
speciﬁc situational details associated with aggression, including identiﬁca-
tion of provocations (however subtle) and other circumstances associated
with the behavior. Individual attention is needed because a ‘‘canned’’
approach to treatment is likely to miss important details. It is important to
consider that behavioral intervention is often the ‘‘last resort’’ for
emotionally stressed owners who are contemplating euthanasia.

The emotional expense of living with an aggressive dog is high. Families

are faced with daily stresses of avoiding bites or, worse, of being bitten. For

308

I.R. Reisner / Vet Clin Small Anim 33 (2003) 303–320

families whose own pets are biting them, there is the additional emotional
conﬂict of facing ‘‘betrayal’’ by a loved family member, one to whom there
is an emotional attachment. Owners of aggressive dogs often feel guilty
that they have somehow contributed to the behavior problem. An addi-
tional stressor is the advice they receive from neighbors, relatives, and the
veterinarian simply to have the dog euthanized.

Recommendations for management of the problem are usually tailored to

ﬁt a particular family’s needs so that the rationale is clear. People may need
time to absorb all the new information being given to them; appointments
are quite long and can easily be overwhelming to owners unfamiliar with
behavioral principles. In some cases, it may be necessary to take the
treatment process one phase at a time so that owners are not faced with
permanent (and perhaps daunting) changes in lifestyle. For example, if told
the dog should be restrained by an indoor lead at all times, owners might
ﬁnd the suggestion easier to accept if it is planned for a brief period, such
as 3 months, at ﬁrst. At that time, a review of results might be quite
encouraging, such that owners are more willing to extend the practice of
leashing the dog indoors.

Although the treatment of aggressive behavior can be rewarding and

eﬀective, there are limitations that should be addressed before getting started.
Most important is the caveat that aggressive behavior cannot be ‘‘cured’’ or
completely resolved. An understanding of this simple fact is critical so that
people continue to be vigilant of and responsible for their dog’s behavior.
Although the behavioral problem is unlikely to be stopped forever, it can
usually be managed so that risks of biting are decreased. The clinician must
identify as many risks as possible both at present and in the future. Family
composition would change, for example, if a baby were to be born or adopted
in the future. Adults may migrate into or out of the home, or the family may
move to a new neighborhood with a diﬀerent range of stimuli. The dog must
also be considered and is rarely behaviorally static; for example, fearful
behavior in the juvenile dog may be expressed as aggression, and reactivity
may occur in new contexts in the mature adult dog.

Safety issues

No matter how carefully designed the treatment plan is, there can be no

relationship to work with if safety issues are not addressed. How can the
clinician help the family to avoid being bitten or to protect neighbors and
visitors from being threatened?

First, it is critical to avoid the circumstances and provocations that have

historically been associated with aggression. This should include avoidance
of speciﬁc targets, especially if they are not family members; for example, if
the dog tends to lunge after running children, it should be kept away from

309

I.R. Reisner / Vet Clin Small Anim 33 (2003) 303–320

parks and other areas where children run. Dogs that are sensitive to human
postures or attitudes should not be subjected to those provocations. Simply
asking visitors to avoid eye contact or other interactions with the dog can
help everyone to relax safely. Handlers of food-aggressive dogs should be
instructed not to remove items from the dog’s mouth. The clinician should
provide a speciﬁc list of such situations.

Second, because so many biting dogs are either anxious or dominantly

aggressive (or both), it is important to discontinue all aversive interactions.
Punishment of any kind should be avoided, including hitting, leash
corrections, ‘‘hanging’’ by holding up the leash, holding the dog by the
scruﬀ, shocking the dog at the moment of aggression (using a electric shock
device), rolling the dog onto its back, and other misguided actions. Any of
these can increase anxiety and is almost certain to result in further biting.
Reacting to an anxious or fearful dog with such a display also guarantees
increased aggression at the next exposure to whatever situation sparked the
aggression in the ﬁrst place.

Because so many dogs are presented for biting unfamiliar people both

in and away from the home, it is critical to outline a plan for dealing
with visitors, children, or anyone who might be unaware of risks or cannot
reliably follow rules established by the owner. Liability issues ring true for
many dog owners, who may more clearly recognize risk when it is described
as potential loss of homeowner’s insurance or the prospect of a bite-related
lawsuit. The clinician can make simple recommendations both orally and in
writing so as to reduce risks. An example is the recommendation that
owners do not rely solely on underground electric fencing for restraint of a
territorially aggressive (or fearful) dog. Instead, if housed outdoors without
supervision, the dog should be enclosed in a secure chain-link kennel run, or
a secure ‘‘visible’’ fence can be constructed outside the perimeter of the
underground fence. The dog should otherwise be kept on a lead and under
active control at all times. A basket-style muzzle can be helpful as additional
insurance against biting when dogs are particularly large, fast, or diﬃcult to
control in high-risk situations.

Medical problems contributing to aggression

Behavioral problems are generally considered to be diagnoses of

exclusion. In practical terms, this would indicate that a primary behavioral
diagnosis depends on the clinician’s reasonable conﬁdence that medical
problems are not the predominant reason for the presenting complaint;
however, not all medical problems must (or realistically can) be ruled out
in many cases. For this reason, it is important to consider physical and
neurologic examination ﬁndings, a minimum database of laboratory tests,
and results of additional tests indicated for individual cases. Animals too

310

I.R. Reisner / Vet Clin Small Anim 33 (2003) 303–320

fractious to be examined thoroughly and in whom there is strong suspicion
of a medical disorder can be sedated or anesthetized for examination and
laboratory workup.

Aggressive behavior is a particularly nonspeciﬁc presenting complaint

and may be a result of acute or chronic pain (eg, osteoarthritis or other
musculoskeletal disease, otitis, cystitis), metabolic or systemic disease (eg,
hyperadrenocorticism, gastrointestinal disease), dermatopathy, neurologic
disease (eg, hydrocephalus in predisposed breeds), sensory impairment,
senile dementia (cognitive dysfunction syndrome), or even iatrogenic causes
like glucocorticoid administration [36].

Rage syndrome

Primarily a lay term, ‘‘rage syndrome’’ [37,38] is used inconsistently to

describe a range of aggressive behaviors, most often sudden aggression to
human beings seen in particular breeds (eg, the English Springer Spaniel
[26,39]). Explosive rage (episodic dyscontrol [39]) in human beings has been
associated with a variety of disorders, including epilepsy [40,41]. In most
cases, however, no speciﬁc etiology has been identiﬁed [40]. A response
to phenobarbital has been reported in isolated cases of aggression in the
veterinary literature [42]; however, it is inconclusive whether a response
to the anticonvulsant is related to seizure activity or whether nonspeciﬁc
sedation plays a role in reducing aggression. Most cases of sudden or
‘‘unprovoked’’ canine aggression seem to be normal responses to subtle pro-
vocations, such as dominance-related conﬂicts [43]. Impulsive aggression,
or impaired impulse control [44], probably comes closest to the clinical
presentation of rage syndrome, but reliable diagnostic criteria have yet to be
identiﬁed.

Behavior modiﬁcation

General principles of behavior modification

Few behavioral problems can be suﬃciently addressed with lasting results

without modiﬁcation of the dog’s behavior through learning. Problems with
roots solely in physiology and with little or no consequential learning may
respond satisfactorily to surgical or chemical (drug therapy) intervention
alone. Most behavioral cases, however, involve at least some learned
component and therefore require learning—modiﬁcation of behavior
through operant (or at times classic) conditioning—as part of their man-
agement.

It is safest to assume that the average dog owner seeking help is relatively

naive about normal or abnormal dog behavior as well as about behavior
modiﬁcation techniques. In the interest of success, owners should be

311

I.R. Reisner / Vet Clin Small Anim 33 (2003) 303–320

counseled about methods of teaching that are both eﬃcient and safe.
General principles of behavior modiﬁcation should include the following
points:

Dogs (and their owners) beneﬁt from consistency and, most important,

predictability. Behavior modiﬁcation should emphasize the predictable
so that anxiety can begin to drop. A great many dogs being presented
with a history of biting are apparently quite anxious.

Communication from owner to dog can probably be improved. At the

least, communication should change so that the owner’s requests are
unambiguous (and, again, predictable). Confused dogs often misbehave
simply because they have not received clear instructions.

Only positive reinforcement (reward) techniques should be used in

the management of aggressive dogs. Aversive tools such as electric
stimulation (shock), prong or training (choke) collars that require
pulling or jerking to work, hitting, and scolding can increase anxiety
and therefore increase the risk of biting; in addition, they are likely to
lead to treatment failure. Punishment or harsh correction of any kind
should be avoided completely; this would include rolling the dog on its
back (also known as the ‘‘alpha roll’’) or assertively grabbing the neck
or cheeks.

Food is a universally accepted form of positive reinforcement. Although

some dogs perform more enthusiastically for toys, tossed objects, or
even praise, the average dog learns new tasks much more readily for
food than for the widely held notion that it should want to please the
instructor

/owner. Owners uncomfortable with this principle can be as-

sured that although the ﬁrst stages of learning do involve continuous
reward, food rewards can (and should) be given only intermittently and
randomly once the task has been learned.

Getting started in behavior modification

Behavior modiﬁcation for canine aggression has a common objective

regardless of the speciﬁc clinical diagnosis. Dogs that bite for reasons related
to social conﬂict, territorial defense, or fear, whether directed to human
beings or to other dogs, all beneﬁt from a standard protocol in which the
previous principles are applied. The basic program helps to teach the dog
a new vocabulary with his human companion—one that can then advance
to more challenging exercises in which speciﬁc (provocative) stimuli are
presented.

The sit-stay exercise developed by Voith as cited by Marder and Reid [45]

is intended to introduce just such a vocabulary. Distinct from obedience
training, the sit-stay exercise helps to teach the dog to relax in a variety of
neutral and then mildly challenging circumstances. Dogs should also be
rewarded for looking up at the owner whenever asked (or called), an exercise

312

I.R. Reisner / Vet Clin Small Anim 33 (2003) 303–320

that can be quite useful in interrupting undesirable behavior. In the face of
some anxiety-inducing stimulus, the owner can call the dog’s name and then
ask it to ‘‘look’’. Having successfully interrupted the dog’s arousal, the
owner can then request a ‘‘sit’’ and follow with praise and food.

Dogs are asked to sit or lie down before getting the things they want; in

this way, they learn quickly that desires and independent decisions must be
deferred to the owner. Deferential behavior can reduce anxiety simply by
taking away the dog’s need to make decisions in distressing circumstances.

A head halter such as the Gentle Leader (Premier Pet Products,

Richmond, VA) is often recommended to help the owner control the dog’s
behavior without the need for physical correction or strength. Aggressive
dogs might beneﬁt from wearing the head collar with an indoor lead at
prescribed times or intervals. An example of its use would be the territorially
defensive dog kept on a lead and head collar when visitors are expected.

Both for the sake of safety and for therapeutic reasons, it is recommended

that a ‘‘safe haven’’ be identiﬁed and used whenever the dog might otherwise
not be adequately supervised. Because bites can happen quickly, visual
supervision alone is not adequate. If the owner is unable to keep the dog on
a lead or if the situation is too potent either for the dog or for people, a
room or sequestered crate can be used to keep everyone calm and safe.
Acclimating the dog to the safe haven may require time and some eﬀort,
including gradually desensitizing the dog to being left alone. To avoid bar-
rier anxiety, baby gates (stacked if necessary) are preferable to shut doors.

A ‘‘paciﬁer’’ or food-ﬁlled toy can be useful for occupying the dog’s at-

tention both while in its safe haven and when kept on a lead for extended peri-
ods. An example is the food-ﬁlled Kong toy (Kong Company, Golden, CO).

Stopping unwanted behavior

Owners of dogs presented with aggressive behavior are advised

repeatedly to avoid punishment and harsh corrections. It is clearly impor-
tant to provide suggestions for what can be done to interrupt aggressive
or anxious behavior and in ways that are not aversive to the dog.

First, simply ‘‘changing the subject’’ can eﬀectively interrupt an un-

desirable behavior. A dog charging the door might quickly be distracted
from it if a box of biscuits is shaken loudly, or a dog visibly frightened or
aroused by any stimulus might welcome the appearance of a leash and the
walk it signiﬁes. Similarly, it may be most eﬀective just to remove the dog
from a stressful or high-risk situation (again, it is advisable to have the dog
on a lead so that collar restraint and ‘‘grabbing’’ can be avoided in tense
moments) by walking him away.

Walking away from a high-risk situation can also be eﬀective when the

human being is the one to leave. This might apply to attention-seeking
behavior associated with aggression or to highly tense situations in which
the owner (or a stranger) is inadvertently provoking an aggressive reaction.

313

I.R. Reisner / Vet Clin Small Anim 33 (2003) 303–320

An example would be the owner who is insistent on cutting the toenails of a
reluctant, fearful, or dominant-aggressive dog. Although owners thus ad-
vised might question the wisdom of ‘‘allowing him to win,’’ leaving unequiv-
ocally prevents biting. Furthermore, the dog cannot be taught eﬀectively
to accept nail trims by the use of force; instead, a gradual reward-based ap-
proach is recommended.

Dogs with a history of food-related or possessive aggression can be

‘‘convinced’’ to relinquish objects by trading a more palatable food item for
the coveted object. With a little more time and eﬀort, they can learn to
‘‘leave it’’ (ie, not to pick up an interesting item) or ‘‘drop it’’ once the item is
in the mouth by use of rewards for relinquishing objects of gradually
increasing palatability or interest. ‘‘Luring’’ a dog away with a food treat
is another useful strategy when necessary. To distinguish the lure from a
‘‘bribe’’ (although the distinction is not especially important, because it
does get the job done), the owner can ask the dog to sit or lie down before
receiving the lure, which has then been transmogriﬁed into a reward.

Finally, the ‘‘time out’’ can be used to reduce anxiety and risks of biting,

because the dog is taken (on a lead) out of the situation to a diﬀerent room
or safe haven. Rather than serving as punishment, which must occur much
more immediately and consistently and must be suﬃciently unpleasant to be
eﬀective, time out probably serves only as a defuser, providing an op-
portunity for the dog to settle and calm down.

Desensitization and counterconditioning (response substitution)

Desensitization is most often paired with a conditioned or learned

response that diﬀers from the unconditioned inappropriate response for
which the dog is being treated. Eﬀective desensitization requires that the
provocative stimulus can be identiﬁed (this might require breaking down
a sequence or group of provocations) and that it can be satisfactorily rep-
licated. It is also helpful if the stimulus can be reduced to a low intensity and
then gradually increased as the dog habituates to each level.

Desensitization and counterconditioning to provocative stimuli require

that the dog has learned basic behavior modiﬁcation tools. The owner can
then begin to control the dog’s reactions by a combination of distraction,
instructions, and predictable (human) behavior. The dog is asked to sit or lie
down and to look up at the owner while the stimulus approaches or passes
by. The exercises should increase in intensity over time, starting with a
neutral stimulus and becoming more challenging only if the dog is receptive
to each step in the program. It is especially important not to move too
quickly through the program, considering the risk that the dog might be
further sensitized (ie, even more reactive) rather than desensitized.

For obvious reasons, it is more diﬃcult to desensitize a dog exhibit-

ing aggressive behavior than it might be to desensitize a fearful but
nonaggressive dog. Desensitization exercises require the gradual approach

314

I.R. Reisner / Vet Clin Small Anim 33 (2003) 303–320

or intensiﬁcation of a stimulus; for the aggressive dog, that stimulus is
likely to be a human being (or the human being’s behavior). To desensitize
safely and eﬀectively, it is critical to keep two points in mind. First, the
reliability of the exercise becomes questionable when an emotional reaction
like aggression is involved. The dog may bite before its brain can process the
consequences or even the true identity of the target [46]. Dogs painstakingly
desensitized to being approached, then gently bumped, more vigorously
bumped, and ﬁnally stepped on by a toddler are still at high risk of biting
the unsteady toddler in the future (when startled). Second, the risks of
biting present a special case of liability. Desensitization exercises should, in
principle, allow the owner and dog to be in certain circumstances that
would have elicited a bite in the past. It is important to keep in mind that
avoidance of provocations or prevention of biting is still the most im-
portant aspect of the management of aggression.

Drug therapy

Drug therapy may be helpful as an adjunct to behavior modiﬁcation and

safety counseling in the management of canine aggression. Indications for
drug therapy include fear or anxiety, particularly because stress can override
the eﬀects of behavior modiﬁcation, and impulsive or explosive (disinhib-
ited) behavior, which may be eﬀectively blunted by serotonergic agents [47].
Duration of drug therapy ranges from relatively short term (eg, 6 months
for cases in which anxiety must be reduced to allow learning to occur) to
longer term therapy (ie, years for animals whose aggressive behavior is not
adequately responsive to behavior modiﬁcation alone).

It is essential that the clinician counsel dog owners about the importance

of safety, preventive strategies, and behavior modiﬁcation, all of which are
critical to success, in addition to drug therapy. Also important, the
extralabel use of psychotropic medication for aggression carries some
liability concerns. Because there is no drug approved for use in aggressive be-
havior, informed consent is particularly important (although it is unlikely
to indemnify the clinician in case of litigation). Although aggressive behav-
ior can be managed, it cannot be cured. Regardless of drug choices or
even behavior modiﬁcation, the risks of biting continue throughout the
dog’s life, which is a concept critical for owners to understand. Some drug
classes may carry additional risks because of the potential for aggres-
sion disinhibition (as may occur with benzodiazepines [48], but this is
potentially true for any drug aﬀecting brain chemistry). Psychotropic
drugs can and do occasionally have unpredictable eﬀects. For example,
ﬂuoxetine may lead to emotional apathy or, in contrast, profound rest-
lessness and agitation in human patients [49]. In the author’s ex-
perience, ﬂuoxetine may also be associated with transient agitation in
dogs. When dealing with potential human injury, such uncertainties must

315

I.R. Reisner / Vet Clin Small Anim 33 (2003) 303–320

be seriously considered and addressed. In spite of these caveats, drug
therapy can help signiﬁcantly in the management of aggressive behavior.

Categories of behavioral drugs

Serotonergic agents

Fluoxetine (Prozac), a selective serotonin reuptake inhibitor (SSRI),

successfully reduced the frequency of growling in one small, single-blind,
crossover study of dominant-aggressive dogs [50]. Fluoxetine, paroxetine
(Paxil), sertraline (Zoloft), and ﬂuvoxamine (Luvox) have been used
anecdotally in the treatment of aggressive behavior, particularly in referral
practices. Clomipramine (Clomicalm), a tricyclic antidepressant (TCA) with
nearly selective serotonin reuptake inhibition, was not more eﬀective than
placebo in a double-blind placebo-controlled study of canine aggression
toward family members [51]. The authors followed its eﬀects for only 6
weeks, however. Another placebo-controlled study on the use of amitripty-
line for aggressive behavior reported a similar lack of eﬀect [52]. Again, this
study was relatively short (4 weeks) and used a relatively small sample of
dogs; a repeated study with more subjects and extending for a longer period
would be informative. There is clearly some beneﬁt, and certainly a rational
basis, in SSRI or TCA therapy for aggression. Further work is needed on
larger numbers of dogs to examine the long-term serotonergic, noradrener-
gic, and behavioral eﬀects of these agents.

Other agents

In addition to SSRIs and TCAs, other agents reported to reduce aggressive

behavior include propranolol [53,54], carbamazepine [54], and lithium [55]. A
favorable response to phenobarbital has been reported in several case reports
of aggressive behavior in dogs [42], although the relation between canine
episodic aggression and seizure disorders remains unvalidated. Phenobarbital
may exert its eﬀects via nonspeciﬁc sedation. Synthetic progestins have been
used successfully in the management of canine dominance-related aggression
[56] but have undesirable long-term eﬀects, such as diabetes mellitus, which
make them a less than ideal choice for chronic use.

Surgical management

Castration

Castration can eﬀectively reduce urine-marking, mounting, and roaming

behavior [57,58]. It would make intuitive sense that only sexually dimorphic
behaviors (eg, dominance-related and territorial aggression but not fear- or
food-related aggression) would be aﬀected by castration. Testosterone may
have an intensifying eﬀect on aggression of all types; in a pair of studies,
intact male dogs were involved in 70% to 76% of reported bites [5,30].

316

I.R. Reisner / Vet Clin Small Anim 33 (2003) 303–320

Hopkins et al [59] reported that castration consistently (in 60% of cases)
reduced only intermale aggression, which is primarily related to dominance
conﬂicts in dogs familiar to each other.

Ovariohysterectomy

Unlike castration, which is likely to reduce aggression (with the exception

of fear-related or food-related aggression) in the male dog, the behavioral
eﬀects of ovariohysterectomy are less clear. One often-cited study on the
eﬀects of ovariohysterectomy on behavior reported that ovariohysterec-
tomized female dogs that had been aggressive before surgery exhibited
signiﬁcantly more aggression toward family members than their sexually
intact counterparts [60]. No signiﬁcant diﬀerences between the ovario-
hysterectomized and intact female dogs were seen in territorial or intra-
speciﬁc aggression. It might be inferred that ovariohysterectomy releases
the expression of otherwise inhibited masculine behaviors [61], such as
owner-directed aggression. More studies are needed on the eﬀects of ovario-
hysterectomy on behavior; in the meantime, such ﬁndings might be
counterbalanced by the insurance that a spayed bitch cannot pass un-
desirable behavior to her oﬀspring.

Summary

Canine aggression directed to human beings is a common presenting

complaint and requires attention to safety issues and behavior modiﬁcation
to minimize the risks of future aggression. Dogs may bite familiar people,
including family members, or unfamiliar people for a variety of reasons.
Anxiety plays an important role in aggression regardless of its target or
circumstances. Eﬀective management of aggression may include education
and safety counseling for owners, lifestyle changes for dogs and owners,
avoidance of provocations when possible, and behavior modiﬁcation to
minimize the risk of future bites. Drug therapy may be indicated to facilitate
behavior modiﬁcation or to reduce reactivity in the dog.

Acknowledgments

The author thanks Alison Seward and Jenny O’Connor, CVT, for as-

sistance in review of the manuscript.

References

[1] Beaver BV. Owner complaints about canine behavior. JAVMA 1994;204:1953–5.
[2] Sacks JJ, Kresnow M, Houston B. Dog bites: how big a problem? Inj Prev 1996;2:52–4.

317

I.R. Reisner / Vet Clin Small Anim 33 (2003) 303–320

[3] Weiss HB, Friedman DI, Coben JH. Incidence of dog bite injuries treated in emergency

departments. JAMA 1998;2711:51–3.

[4] Canadian Hospitals Injury Reporting and Prevention Program (CHIRPP). Health Canada:

injuries associated with dog bites and dog attacks. Ottawa: Health Protection Branch,
Laboratory Centre for Disease Control; 1996.

[5] Wright JC. Canine aggression toward people: bite scenarios and prevention. Vet Clin

North Am Small Anim Pract 1991;21:299–314.

[6] Beck AM, Jones BA. Unreported dog bites in children. Public Health Rep 1985;100:315–21.
[7] Bernardo LM, Gardner MJ, Rosenﬁeld RL, et al. A comparison of dog bite injuries in

younger and older children treated in a pediatric emergency department. Pediatr Emerg
Care 2002;18:247–9.

[8] Karlson TA. The incidence of facial injuries from dog bites. JAMA 1984;251:3265–7.
[9] Calkins CM, Bensard DD, Partrick DA, et al. Life-threatening dog attacks: a devastating

combination of penetrating and blunt injuries. J Pediatr Surg 2001;36:1115–7.

[10] Guy NC, Luescher UA, Dohoo SE, et al. A case series of biting dogs: characteristics of the

dogs, their behaviour, and their victims. Appl Anim Behav Sci 2001;74:43–57.

[11] Sacks JJ, Sinclair L, Gilchrist J, et al. Breeds of dogs involved in fatal human attacks in the

United States between 1979 and 1998. JAVMA 2000;217:836–40.

[12] Beaver BV. Canine social behavior. In: Canine behavior: a guide for veterinarians.

Philadelphia: WB Saunders; 1999. p. 137–99.

[13] Beaver BV. Clinical classiﬁcation of canine aggression. Appl Anim Ethol 1983;10:35–43.
[14] Borchelt PL. Aggressive behavior of dogs kept as companion animals: classiﬁcation and

inﬂuence of sex, reproductive status and breed. Appl Anim Ethol 1983;10:45–61.

[15] Borchelt PL, Voith VL. Aggressive behavior in cats. In: Voith VL, Borchelt PL, editors.

Readings in companion animal behavior. Trenton: Veterinary Learning Systems; 1996.
p. 208–16.

[16] Borchelt PL, Voith VL. Classiﬁcation of animal behavior problems. Vet Clin North Am

Small Anim Pract 1982;12:571–85.

[17] Chapman BL. Feline aggression: classiﬁcation, diagnosis and treatment. Vet Clin North

Am Small Anim Pract 1991;21:315–27.

[18] Crowell-Davis SL, Barry K, Wolfe R. Social behavior and aggressive problems of cats.

Vet Clin North Am Small Anim Pract 1997;27:549–68.

[19] Hart BL. Canine behavior. Canine Pract 1974;1:6–8.
[20] Houpt KA. Aggression in dogs. Compend Contin Educ Pract Vet 1979;1:123–8.
[21] Houpt KA. Aggression and social structure. In: Domestic animal behavior for veterinarians

and animal scientists. 3rd edition. Ames: Iowa State University Press; 1998. p. 61–73.

[22] Landsberg G, Hunthausen W, Ackerman L. Canine aggression. In: Handbook of

behaviour problems of the dog and cat. Oxford: Butterworth-Heinemann; 1997. p. 129–50.

[23] Moyer KE. Kinds of aggression and their physiological basis. Commun Behav Biol

1968;2:65–87.

[24] Overall KL. Canine aggression. In: Clinical behavioral medicine for small animals. St.

Louis: Mosby-Year Book; 1997. p. 88–137.

[25] Borchelt PL, Voith VL. Dominance aggression in dogs. In: Voith VL, Borchelt PL,

editors. Readings in companion animal behavior. Trenton: Veterinary Learning Systems;
1996. p. 230–9.

[26] Reisner IR. Neurophysiologic, familial and environmental correlates of canine dominance-

related aggression [doctoral dissertation]. New York: Cornell University; 1996.

[27] Hart BL, Hart LA. Selecting pet dogs on the basis of cluster analysis of breed behavior

proﬁles and gender. JAVMA 1985;186:1181–5.

[28] Reisner IR, Erb HN, Houpt KA. Risk factors for behavior-related euthanasia among

dominant-aggressive dogs: 110 cases (1989–1992). JAVMA 1994;205:855–63.

[29] Guy NC, Luescher UA, Dohoo SE, et al. Risk factors for dog bites to owners in a general

veterinary caseload. Appl Anim Behav Sci 2001;74:29–42.

318

I.R. Reisner / Vet Clin Small Anim 33 (2003) 303–320

[30] Gershman KA, Sacks JJ, Wright JC. Which dogs bite? A case-control study of risk factors.

Pediatrics 1994;93:913–7.

[31] Wright JC. Severe attacks by dogs: characteristics of the dogs, the victims, and the attack

settings. Public Health Rep 1985;100:55–61.

[32] Borchelt PL, Lockwood R, Beck AM, et al. Attacks by packs of dogs involving predation

on human beings. Public Health Rep 1983;98:57–66.

[33] Lockwood R, Rindy K. Are ‘‘pit bulls’’ different? An analysis of the pit bull terrier

controversy. Anthrozoos 1987;1:2–8.

[34] Beerda B, Schilder MB, van Hooff JA, et al. Chronic stress in dogs subjected to social and

spatial restriction. I. Behavioral responses. Physiol Behav 1999;66:233–42.

[35] Podberscek AL, Serpell JA. Environmental inﬂuences on the expression of aggressive

behaviour in English Cocker Spaniels. Appl Anim Behav Sci 1997;52:215–27.

[36] Klein JF. Adverse psychiatric effects of systemic glucocorticoid therapy. Am Fam

Physician 1992;46:1469–74.

[37] Dunbar I. Rage syndrome reexamined. Purebred Dogs

/American Kennel Club Gazette

1990;July:24–5.

[38] Fisher GT. It’s all the rage. Purebred Dogs

/American Kennel Club Gazette 1990;January:

78–83.

[39] Bach-Y-Rita G, Lion JR, Climent CE, et al. Episodic dyscontrol: a study of 130 violent

patients. Am J Psychiatry 1971;127:49–54.

[40] Elliott FA. Neurological ﬁndings in adult minimal brain dysfunction and the dyscontrol

syndrome. J Nerv Ment Dis 1982;170:680–7.

[41] Mark VH, Ervin FR. Acquired brain disease and violence. In: Violence and the brain. New

York: Harper & Row; 1970. p. 55–68.

[42] Dodman NH, Miczek KA, Knowles K, et al. Phenobarbital-responsive episodic dyscontrol

(rage) in dogs. JAVMA 1992;201:1580–3.

[43] Podberscek AL, Serpell JA. The English Cocker Spaniel: preliminary ﬁndings on aggressive

behavior. Appl Anim Behav Sci 1996;47:75–89.

[44] Reisner IR, Mann JJ, Stanley M, et al. Comparison of cerebrospinal ﬂuid monoamine meta-

bolite levels in dominant-aggressive and non-aggressive dogs. Brain Res 1996;714:57–64.

[45] Marder A, Reid PJ. Treating canine behavior problems: behavior modiﬁcation, obedience,

and agility training. In: Voith VL, Borchelt PL, editors. Readings in companion animal
behavior. Trenton: Veterinary Learning Systems; 1996. p. 56–61.

[46] LeDoux JE. Emotion, memory and the brain. Sci Am 1994;270:50–7.
[47] Cherek DR, Lane SD, Pietras CJ, et al. Effects of chronic paroxetine administration on

measures of aggressive and impulsive responses of adult males with a history of conduct
disorder. Psychopharmacology (Berl) 2002;159:266–74.

[48] Bruhwyler J, Chleide E. Comparative study of the behavioral, neurophysiological, and

motor effects of psychotropic drugs in the dog. Biol Psychiatry 1990;27:1264–78.

[49] Stahl SM. Essential psychopharmacology: neuroscientiﬁc basis and practical applications.

2nd edition. Cambridge: Cambridge University Press; 2000. p. 230–4.

[50] Dodman NH, Donnelly R, Shuster L, et al. Use of ﬂuoxetine to treat dominance

aggression in dogs. JAVMA 1996;209:1585–7.

[51] White MM, Neilson JC, Hart BL, et al. Effects of clomipramine hydrochloride on

dominance-related aggression in dogs. JAVMA 1999;215:1288–91.

[52] Virga V, Houpt KA, Scarlett JM. Efﬁcacy of amitriptyline as a pharmacological adjunct to

behavioral modiﬁcation in the management of aggressive behaviors in dogs. J Am Anim
Hosp Assoc 2001;37:325–30.

[53] Greendyke RM, Kanter DR, Schuster DB, et al. Propranolol treatment of assaultive

patients with organic brain disease. A double-blind crossover, placebo-controlled study.
J Nerv Ment Dis 1986;174:290–4.

[54] Sheard MH. Clinical pharmacology of aggressive behavior. Clin Neuropharmacol

1988;11:483–92.

319

I.R. Reisner / Vet Clin Small Anim 33 (2003) 303–320

[55] Reisner I. Use of lithium for treatment of canine dominance-related aggression [abstract].

Appl Anim Behav Sci 1994;39:183–92.

[56] Hart BL. Progestin therapy for aggressive behavior in male dogs. JAVMA 1981;178:

1070–1.

[57] Hart BL, Hart LA. Canine and feline behavioral therapy. Philadelphia: Lea & Febiger;

1985.

[58] Neilson JC, Eckstein RA, Hart BL. Effects of castration on problem behaviors in male

dogs with reference to age and duration of behavior. JAVMA 1997;211:180–2.

[59] Hopkins SG, Schubert TA, Hart BL. Castration of adult male dogs: effects on roaming,

aggression, urine marking, and mounting. JAVMA 1976;168:1108–10.

[60] O’Farrell V, Peachey E. Behavioural effects of ovariohysterectomy on bitches. J Small

Anim Pract 1990;31:595–8.

[61] Voith VL, Borchelt PL. Diagnosis and treatment of dominance aggression in dogs. Vet

Clin North Am Small Anim Pract 1982;12:655–63.

320

I.R. Reisner / Vet Clin Small Anim 33 (2003) 303–320

Separation anxiety in dogs

The function of homeostasis in its

development and treatment

David Appleby, MSc

a,b,

Jolanda Pluijmakers, AS

The Pet Behaviour Centre, Upper Street, Defford, Worcestershire, WR8 9AB, England

Queens Veterinary School, Cambridge University, Cambridge, England

Anthrozoology Institute, School of Biological Sciences,

University of Southampton, Southampton, England

Separation problems

Problems involving destruction, vocalization, and house soiling by dogs

that occur during the owner’s absence are common [1–5] and constitute a
signiﬁcant proportion of the caseload of the behavioral specialist [6]. Until
relatively recently, the term separation anxiety was used generically to
describe all separation-related problems [7]. There are causes that are
unrelated to anxiety [1,7–9], however, and previous publications have
categorized them [1,2,6,9]. In particular, McCrave [6] produced an in-
ﬂuential paper that identiﬁes differentials for the motivations of the three
most common problem behaviors (Table 1) [10]. As a consequence of this
classiﬁcation, new generic terms were introduced [2], and it is now common
practice to refer to a separation problem, followed by a description of the
perceived motivation, one of which is separation anxiety.

Diagnosing separation anxiety

It is generally recognized that successful treatment of separation anxiety

requires careful consideration of the history of the dog and the present-
ing signs, followed by diagnosis based on empiric evidence [1,3,4,6,11–13].

Vet Clin Small Anim

33 (2003) 321–344

* Corresponding author. The Pet Behaviour Centre, Upper Street, Defford, Worcestershire,

WR8 9AB, England.

E-mail address:

appleby@petbcent.demon.co.uk (D. Appleby).

0195-5616/03/$ - see front matter

PII: S 0 1 9 5 - 5 6 1 6 ( 0 2 ) 0 0 1 0 1 - 8

The process is made diﬃcult by a lack of consensus about how separation
anxiety should be deﬁned, however [14–16].

The symptoms or combinations of symptoms commonly reported dur-

ing owner absence are destruction, house soiling, and vocal behavior
[1,4,6,12,13,18] indicative of distress [17,18]. Less reported but welfare-
signiﬁcant symptoms [10] can also occur, with the possibilities including
withdrawal, inappetence, hyperventilation, salivation, gastrointestinal
disorders (vomiting

/diarrhea) [5], increased and repetitive motor activity

(eg, pacing, circling), and repetitive behaviors (eg, overgrooming or self-
mutilation) [1,3,4,6]. Dogs with the condition can become anxious and
agitated or display depressed behaviors in response to stimuli associated
with the owner’s departure [6,13]. Separation anxiety has been cited as the
reason why some dogs try to prevent owner departure aggressively
[1,6,20,21]. They often have an aggressive behavioral problem independent
of their separation anxiety [4,6], are uninhibited in their relationship with
their owners, and may attempt to control resources (including the owner’s
presence) [4]. These factors may also combine (Fig. 1) [1]. In one recent
study, the median age of dogs at onset of separation anxiety was older than
1.5 years [5]. The signiﬁcance of breed diﬀers between studies; although an
increased prevalence of the problem has been reported in mixed breeds [6], a
study of a general population found no bias [2]. The problem is reported
more often in male dogs than in female dogs [2,3,5,14,20,22,23]. Prolonged
periods without separation from the owner, a prolonged period without the
person to whom the dog is attached, periods of kenneling [24], a house move
with the owner [22,25], and time spent at a shelter [6,24,26] have all been
cited as causes of separation anxiety.

In the broadest deﬁnition of separation anxiety, the condition is

described as problematic behavior motivated by anxiety that occurs exclu-
sively in the owner’s absence or virtual absence [1,12,22]. A more speciﬁc
deﬁnition of separation anxiety [14] requires ongoing attachment to the
maternal or primary caregiver or person to whom this attachment is trans-
ferred after homing [4,9,14,26]. This deﬁnition is borrowed from human
psychology and attachment theory in human and nonhuman apes [26]. It
has been suggested that a resultant bond [27] allows the infant a secure base

Table 1
Diﬀerential diagnoses for separation problems

House soiling

Destruction

Vocalisation

House breaking

Play behavior

Reaction to external stimuli

Submissive

/excitement

Puppy chewing

Socially facilitated

Urine marking

Reaction to arousing stimuli
Overactivity

Play

/aggression

Fear induced

Fear response

Fear induced

Separation anxiety

Data from

McCrave EA. Diagnostic criteria for separation anxiety in the dog. Vet Clin

North Am Small Anim Pract 1991;21:247–55.

322

D. Appleby, J. Pluijmakers / Vet Clin Small Anim 33 (2003) 321–344

from which it can explore its environment [4,27–29] and develop all its
behavior [4]. The potential for an attachment bond to develop is increased if
the owner also forms a strong attachment to the dog, because the owner
responds to, and therefore reinforces, dependent behavior [4].

Hyperattachment

Some authors in the ﬁeld of pet behavior have suggested that hyper-

attachment is a necessary condition for separation anxiety [3,4,6,14,24]. This
has been subdivided into primary and secondary hyperattachment [4,14].
Primary hyperattachment is the continuance of the primary attachment
bond to an individual beyond puberty, which constitutes the speciﬁc de-
ﬁnition of separation anxiety and correlates with a perpetuation of other
characteristics of immaturity [14]. Secondary hyperattachment can develop
at any age and is described as dependency on one or more persons in the
dog’s ‘‘family’’ circle. A dog suﬀering from an emotional disorder, such as
phobia or loss of a primary attachment ﬁgure, may develop this type of
attachment [4,14]. Typical manifestations of hyperattachment are the
organization of all the dog’s activities around the attachment ﬁgure when
that individual is present [4], following the owner from room to room [20],
the owner not able to go to the bathroom without the dog wanting to follow
[4,8], the dog wanting to sleep next to its owner [4], the dog leaning on its
owner [24], the dog constantly wanting to be held [20], and the dog dis-
playing distress if separated from the owner when that individual is at home,
which may involve destruction at the point of access [30]. Such dogs also

Fig. 1. Dependence on the owner’s presence increases that person’s value as a resource, which
the uninhibited dog may try to control by preventing departure. The associated frustration and
anxiety caused by anticipation of failure may result in aggression [5,30,36].

323

D. Appleby, J. Pluijmakers / Vet Clin Small Anim 33 (2003) 321–344

stand out from the normal canine population in respect to their eﬀusive
greeting behavior at the time of the owner’s return [1,6,12,13].

There are arguments against hyperattachment being a necessary

condition for separation anxiety. These include the observations that dogs
that are ‘‘spoiled’’ and encouraged to have a close relationship with their
owners do not necessarily develop separation problems [1,6,12,22]. Several
authors have commented that only some dogs that display separation
anxiety in the broader sense display symptoms of hyperattachment when
the owners are at home [12,13].

Destruction and vocal behavior motivated by separation anxiety are

routine rather than the intermittent behavior that occurs with other moti-
vations [6,30]. Some authors have suggested that destruction orientated
toward doors and windows that give access to the direction by which the
owner left is indicative of separation anxiety [1,6] and barrier frustration
[30], which is consistent with hyperattachment. Destructive behavior
involving items impregnated with the owner’s scent, such as shoes, papers,
bedding, and television controls, also occurs [30] and has been attributed to
disorganization of exploratory behavior related to seeking the owner [4].
When the dog is separated from the owner, vocalization is synonymous with
puppy vocalization during distress and aﬃliative behavior and is generally
higher in pitch, uses repeated subunits, has little variation in tone, and occurs
at a greater rate when compared with vocalization of ‘‘normal’’ dogs [18].

The timing of onset of symptoms when such dogs are left is signiﬁcant,

typically within the ﬁrst 30 minutes and often almost as soon as the dog is
left [6]. They rapidly mount in magnitude and reach a peak within 30
minutes [1,6,9,20,31,32], followed by a gradual adaptation period and a
steady decline in distress from the level of arousal caused by departure or
rearousal caused by external stimulation in addition to an internally con-
trolled 20- to 30-minute cyclic component [33]. Symptoms can persist until
the owner returns, but the dog may recover and relax sooner [1,6].

The opponent-process theory [34] oﬀers a useful construct for under-

standing the adverse separation reactivity in dogs [9]. According to this
theory, a hypothetic neural system regulates emotional arousal and prevents
aﬀective extremes from occurring as the result of attractive or aversive
stimulation. Feelings of well-being and comfort are shadowed by hedoni-
cally opposite eﬀects, such as feelings of contact need. In terms of the
phenomena of dependence, repeated stimulation of these feelings results in
the gradual attenuation of dependent behavior.

When a separation-anxious dog is comforted by social contact and

security, opposing aﬀects are generated and become problematic whenever
the dog is left alone, when it is overwhelmed by loss of security and con-
trol. Repeated stimulation of these processes results in a condition of
perpetual social attention seeking and neediness, and the repeated stimula-
tion of positive social eﬀects strengthens the underlying anxiety or fear.
As a consequence, when the dog is left alone, the aversive emotions reoccur.

324

D. Appleby, J. Pluijmakers / Vet Clin Small Anim 33 (2003) 321–344

The process continues for as long as the dependence is not eﬀectively treated
by broadening the dog’s maintenance set and fear and anxiety are not
reduced through systematic desensitization and counterconditioning [9].

Fear, anxiety, and phobia

A possible explanation for any failure of co-occurrence of hyperattach-

ment and anxiety when separated from the owner is that separation anxiety
describes more than one category of motivation, in which case, another
generic term may be necessary, within which separation anxiety is a
subgroup. Subclassiﬁcation is made diﬃcult by the fact that separation
anxiety is not uncorrelated with fear, but the link is poorly understood [12].
Although anxiety, fear, and phobia are said to be distinct in some way and
may not be driven by identical mechanisms [9], they are probably related at
the neurochemical level [35].

Fear, ‘‘a hypothetical state of the brain, or neuro-endocrine system,

arising under certain conditions and eventuating in certain forms of
behaviour’’ [36], together with anxiety, has often been considered as a
motivator [37]. Fear and anxiety are deﬁned as emotional states that are
caused by the perception of any factual danger (fear state) or possible
danger or nonreward (anxiety state) [36] that threatens the well-being of the
individual and are characterized as a feeling of insecurity [36,37] and distress
[9]. Phobia occurs when fear is not extinguished but remains at the same
high level even though the conditioned stimulus is never paired again
with the noxious unconditioned stimulus [65], because the sensation of
fear becomes the unconditioned stimulus [12].

To alleviate distress in aversive situations that are a threat to homeo-

stasis, animals display an adaptive response to recent or anticipated danger,
which involves two interdependent facets: psychobehavioral changes that
nullify the eﬀects of the trigger and neuroendocrine adjustments necessary
to maintain internal homeostasis [37]. Two main systems are involved: the
autonomic nervous system and the hypothalamic-pituitary-adrenocortical
(HPA) system [38]. Examples of situations that cause a feeling of insecurity
and induce hormonal signs of stress include mother-infant separation and
exposure to novelty [38]. Anticipation of distress requires a predictable
relation between a cue and the stressor [10,35,36,39,40], and response can be
dependent on cues that lack distinctiveness or on patterns or sequences of
events that are diﬃcult to identify [35,40], which can include owner absence
if fear-eliciting stimuli have previously occurred in that context [24]. Control
over the eﬀect of the stressor is associated with lesser signs of distress [38].
Activation of the HPA axis does not seem to occur when the animal is in a
familiar situation in which it has a tried and tested coping strategy available
for dealing with any anticipated challenge in that situation and where the
actions taken are expected to deliver the results [41].

325

D. Appleby, J. Pluijmakers / Vet Clin Small Anim 33 (2003) 321–344

Maintenance stimuli

It seems unlikely that the diﬀerent models for separation anxiety as

described previously could develop without being interpretations of the
same process. Therefore, is it possible that there is a mechanism that sits
comfortably with both schools of thought? Rather than being exclusively
caused by being left unattended as suggested by many authors, separation
anxiety can be deﬁned as ‘‘apprehension due to removal of signiﬁcant
persons or familiar surroundings’’ [42].

In the dog, which is a social pack animal [13], a greater predisposition

for problems associated with owner absence may have been unwittingly
developed as a consequence of selecting for neotenous, aﬀectionate, and
socially dependent behavior [17,26]. Formation and continuation of de-
pendence result from an ongoing conditioning process during which re-
sponse patterns become attached to the cues provided by the social and
nonsocial objects in the animal’s environment [17,43,44]. Therefore, what is
often interpreted as attachment or bonding is actually a high level of
conditioned dependency required for emotional homeostasis, deﬁned as
stability in the normal neurophysiologic states of the organism [42]. The
signiﬁcant factors in the extent to which dependency on any one stimulus
develops are salience, duration of exposure [4,45], and stimulation
[43,46,47]. Removal of an object on which the response system of the
animal has been strongly conditioned to depend for the maintenance of
homeostasis is associated with a signiﬁcant disruption of the animal’s
behavior. The degree of disruption is correlated with the likelihood of be-
havior to reintroduce the maintenance stimulus. This, in turn, decreases
disruption and increases the display of the behavior aiming to achieve
reintroduction of the stimulus on subsequent occasions [43].

What is described as maternal attachment is inevitable because of the

availability and salience of the stimulus, the sensory and cognitive de-
velopment at the time of exposure [17,44], and absence of opportunity to
attach to other social or environmental stimuli because of limited mobility.
The apparent attachment to or dependence on the maternal ﬁgure [5] is not
an aﬀectional bond but a way for the individual to maintain behavioral
organization [43,47], in eﬀect what the puppy needs for the maintenance of a
sense of well-being, or homeostasis of the autonomic nervous system
[38,43,48]. This stability results in the conﬁdence to explore and develop
parasympathetic responses to other stimuli through learning [2]. Exposure
to experiences and learning to cope reduce emotionality when exposed to
novel or challenging stimuli [17,26]. The process results in reduced de-
pendence on the initially narrow and salient stimulus set necessary for the
maintenance of homeostasis and behavioral organization [45] associated
with proximity to the dam, littermates, and nest site [43,47].

Dependence on the dam is reduced as she becomes less salient after wean-

ing, less responsive, and less tolerant [4], which suggests that dependence

326

D. Appleby, J. Pluijmakers / Vet Clin Small Anim 33 (2003) 321–344

on speciﬁc stimuli for behavioral organization is also unlearned [17,45,
49,50]. A puppy may remain dependent on the maternal ﬁgure if the process
is disrupted, however [4]. This can occur if the individual is overprotected or
oversocialized with its owners and not given the opportunity to develop
independence [4,17]. Illness during puppyhood and

/or nursing is similarly

disruptive [26]. Owners can unwittingly encourage dependence through
reinforcement of care-soliciting behaviors and reinforcement of sympathetic
automatic responses to challenging or fear-eliciting stimuli, the probability
of which is predisposed genetically [51] and

/or through stimulus deprivation

in early life [17,26,52,53]. Conversely, a lack of nurturing stimuli or their
premature withdrawal can result in an inability to learn normal social
responses [10].

Development of maintenance stimuli

After 4 weeks of age, attachment to both animate and inanimate stimuli

occurs [44,54,55]. The length of association with an animate or inanimate
object and its relative cue weight determine the development of dependent
behavior [43]. The process is quick; for example, Scott [44,56] found that site
attachment was formed in 20 minutes [54,55,57,58]. Available data support
the concept that animals tend to remain in the presence of objects to which
they have been exposed [37].

Several studies, some in puppies but with implications for latter life, have

been concerned with the eﬀects of environmental and social experience on
behavioral organization and the alleviation of distress. The eﬀect of these
and the potential for inducing dependence can be ranked as follows:

1. Isolation in an unfamiliar and uncomfortable location causes more

distress vocalization than isolation in an unfamiliar but comfortable
location [55].

2. Isolation in an unfamiliar location results in higher levels of distress

vocalization than isolation in a familiar location in puppies [54], but the
opposite is true in older dogs [59]. This could be a result of the fact that
puppies are more dependent on the stimuli associated with their limited
experience and recent reinforcement of contact

/care soliciting. Vocaliza-

tion in a familiar environment in adult dogs could be the consequence
of previous reinforcement, higher expectation, and frustration in that
location.

3. Food [28,60] and toys [60] are less effective in the amelioration of

distress than warmth and comfort [28,60].

4. Isolation and segregation in a familiar location cause more distress

vocalization than retention in the same location with a familiar
conspeciﬁc [54].

5. Food and toys are less effective in the amelioration of distress

vocalization than the presence of a familiar or unfamiliar conspeciﬁc.

327

D. Appleby, J. Pluijmakers / Vet Clin Small Anim 33 (2003) 321–344

6. Familiar or unfamiliar conspeciﬁcs reduce distress vocalizations less

effectively than a human companion. The effect of the latter is pro-
portionate to the level of interaction [59,60]. This is consistent with
research that shows that attachment in children is not dependent on
care giving but on responsiveness to infant behavior and provision
of stimulation [29].

7. Another human example suggests that the presence of an unfami-

liar person beneﬁts conﬁdence less than the presence of a familiar per-
son [61].

We propose that the canine population can be divided into three groups

according to the maintenance stimuli they depend on. These stimuli can
change with time, although the probability of change is dependent on
several factors: (1) the degree of dependence on and salience of existing
stimuli, (2) the availability of existing stimuli, and (3) how these factors com-
pare with the properties of new stimuli. Therefore, movement between groups
can occur in response to events.

Group A: those that do not develop autonomy because of primary

hyperattachment

Group B: those that transfer their dependence to one or more stimuli,

normally social, through need or an increase in the stimuli’s salience
and

/or availability

Group C: those that learn to depend on a range of stimuli without any

narrow set of social or environmental stimuli becoming exceptionally
salient

If homeostasis is disrupted, a dog may try to re-establish it by proximity

to maintenance stimuli, which might be a salient human companion. The
extent to which proximity to salient social maintenance stimuli is displayed
is dependent on a dog’s expectation that it will be left. Animals that are
seldom left or recognize a context in which they are unlikely to be left, such
as after a certain time at night, seem to show less need to stay in proximity,
generally or at speciﬁc times, than those that are left frequently [62]. Con-
versely, the need for proximity seems to increase if the owner’s departures
are unpredictable [13], as is sometimes the case with shift workers. Separ-
ation from emotionally rewarding stimuli is frustrating and has a punishing
eﬀect, and anticipation of it can result in anxiety [36]. In turn, this can lead
to an increase in the vigor or depressed behavior associated with main-
taining proximity at its withdrawal or at anticipation of withdrawal [9,36].
The extent to which these behaviors are displayed is aﬀected by the extent to
which homeostasis is disrupted.

Symptoms of distress when left unattended often start after the owners or

an owner has been at home for a period of time, such as during a holiday
[1,63]. This can be explained as transference from Group C to Group B as a
result of long exposure to the person(s) who, in some instances, also become

328

D. Appleby, J. Pluijmakers / Vet Clin Small Anim 33 (2003) 321–344

more salient (eg, if the owner lies with the dog for long periods of time if
unwell) or more stimulating (eg, through increased activity or stimulation
together).

Maintenance set disruption, fear, and anxiety that a fear-eliciting event

may occur during owner absence do not necessarily result in attempts to
remain in contact with the owner [9,10]. Instead, the dog may seek
maintenance from inanimate stimuli and escape [20] by any door or window
rather than speciﬁcally by the one that would give access to the owner. Dogs
may try to increase homeostasis by digging into locations to hide in or
to gain access to rooms shut oﬀ from them that they associate with
maintenance stimuli, such as the owner’s bedroom, and, occasionally, the
owner may ﬁnd torn carpet or furniture [30,31,62]. The extent to which fear
is expressed when exposed to the same stimuli when the owner(s) is present
may be reduced [64], because the set of maintenance stimuli is more
complete and behavior is more ordered.

Disruption of homeostasis

Disruption of homeostasis can result from internal or external stim-

ulation or both. The potential for disruption of homeostasis increases with
the magnitude of challenge, which is inﬂuenced by the following:

1. The loss of several less salient or one or more major stimuli from the

maintenance set, which leads to behavioral disturbance, disruption of
responses to situations and events, and a feeling of loss of control that
can cause anxiety [37,43,47]. It is well recognized that the removal of
salient social maintenance stimuli is a precursor to fear [65]. The
reintroduction of stimuli, the introduction of stimuli comparable to the
original(s), or the learning of new maintenance stimuli allows return to
homeostasis and reorganization of behavior.

2. The presence of novel stimuli leads to a negative emotional state, which

requires the animal to compare the event with events experienced in the
past [37,45]. Behavioral arousal caused by the exposure to novelty is said
to be similar to the arousal caused by an electric footshock [66]. The
reaction to novelty normally decreases with repeated exposure to an
earlier novel environment, however [38,67].

3. Animals may react fearfully toward a stimulus because of its physical

characteristics (ie, intensity, duration, suddenness) or because it is
associated with a threatening event as a result of learning [36].

The eﬀect of these factors can be combined and can accumulate through

a process of sensitization [67]. The extent to which disruption occurs is
attenuated by the strength of the maintenance provided by the stimuli
present in that context.

329

D. Appleby, J. Pluijmakers / Vet Clin Small Anim 33 (2003) 321–344

Behavioral responses to disruption of homeostasis

The type and magnitude of neuroendocrine arousal and the expression of

behavioral signs associated with disruption of homeostasis are determined
by the following:

1. Psychologic factors [37], such as the composition of the stimulus set the

animal depends on and the state of the neuroendocrine system when
confronted with a challenging stimulus. The potential is inﬂuenced
by both phenotype and any underlying pathologic characteristics that
could play a role [4,10].

2. The amount of control the animal can exert on a challenging stimulus

or threatening environment by the display of suitable behaviors [68].

3. The physical properties of the triggering stimulus [37] (eg, suddenness,

intensity).

The animal’s ability to predict and control a threatening event determines

the neuroendocrine pattern and intensity of emotion experienced [37,68,69].
As long as the animal is only challenged in its control, the medullosympa-
thetic system is dominant [68]. Catecholamines are released in situations that
call for attention and vigilance. The loss of control or the perspective of fail-
ure to meet expectations causes a favorable activation of the HPA axis [68].

The behavioral response to aversive events varies greatly and depends on

whether threat is present (fear state) or anticipated (anxiety state) [37] and on
the intensity of emotion stimulated [70]. Low fear levels enhance activity (eg,
moving around is generally an active behavioral strategy of coping), which
leads to a decrease of the HPA axis arousal [66]. Intermediate levels normally
lead to conﬂict between the expression of fear and activity (eg, exploratory
behavior is reduced). Intense fear disrupts behavior or inhibits it totally
[37,36]. In relation to separation anxiety, destruction and vocalization are
usually said to be attempts to regain contact with the owner by escaping
from conﬁnement and following or by distressed

/relocation vocaliza-

tion [18,23,30]. These behaviors could be interpreted as an attempt to cope
by regaining control and indicative of a low level of arousal. In contrast,
inappropriate defecation and urination may be symptomatic of a higher level
of arousal, generalized anxiety [23], or an intense reaction to a threatening
stimulus [71], and they may occur if the dog ﬁnds it has no control over the
arousing stimulus because of the lack of a successful coping strategy.

Diagnosis

Formulation of treatment depends on whether the dog is classiﬁed as a

member of Group A, B, or C, because diﬀerent treatment rationales are
required and can be more or less essential for establishing or re-establishing
homeostasis and resolution of the animal’s distress from which the problem

330

D. Appleby, J. Pluijmakers / Vet Clin Small Anim 33 (2003) 321–344

arose (Table 2). Classiﬁcation and the magnitude of symptoms also
determine how and the extent to which the treatment program should be
phased and which, if any, drugs are most suitable to support therapy.

All symptoms should be listed, after which the diagnosis has to be further

reﬁned using the following criteria:

Onset [12,35,63], duration, and intensity of the symptoms displayed
Behavior of the dog when the owner is present [3,4,6,13,22]
Departure and greeting behaviors [3,4,6,13,22]
Detailed analysis of the displayed symptoms [6,10]

The listing of the symptoms provides information about the possible

causes of the problem behavior [10] and the level of anxiety. A broad range
of symptoms is indicative of multiple causes and

/or a high level of arousal.

More speciﬁc symptoms of attempting to cope, such as destruction and

/or

vocalization, indicate a lower level of anxiety during which the medullo-
sympathetic system is more likely to be dominant. A single cause, such as
disturbance of the maintenance stimulus set, is more likely to underlie the
behavior. Defecation and urination alone or in combination with other
symptoms indicate a high and more generalized level of arousal during
which the HPA system is dominant and the possible involvement of a fear-
eliciting stimulus (eg, noise phobia).

For members of Group A that have not learned to depend on a broad

stimulus set, the presence of the owner, onto whom maternal dependence
has been transferred, is necessary for emotional homeostasis. Virtual or
actual separation from the owner or its anticipation causes a reduced sense
of control, anxiety, and disruption of behavior. Destruction typically in-
volves attempts to regain contact (eg, at doors and windows that would
give access to the owner). Anxiety during the owner’s absence increases the
potential for fear in response to stimuli causing or associated with threat.
Treatment for anxiety caused by the absence of the owner requires a
reduction of dependence on the owner and increasing dependence on other
stimuli for emotional homeostasis. It is often appropriate for treatment to
be phased, each stage of which is introduced gradually. If a problem of
response to fear stimuli coexists, it should be treated separately, and
consideration should be given to doing so before addressing anxiety caused
by the owner’s absence.

For dogs in Group B, disturbance of homeostasis and the experience of

loss of control can result from (1) the removal of one salient stimulus,
normally social; (2) removal of several less signiﬁcant stimuli from the
maintenance set, normally social; or (3) a change in the need of the animal
to rely on the maintenance set, for example, as a result of feeling threatened
by an aversive or novel stimuli or as the result of the process of ageing. If
disruption results in excessive dependence on a person or persons rather
than on environmental stimuli, which we have argued is likely, anxiety when
the dog is left unattended increases the potential for fear.

331

D. Appleby, J. Pluijmakers / Vet Clin Small Anim 33 (2003) 321–344

Table

Diﬀeren

tial

sympto

(ie,

those

that

are

not

gener

all

three

gro

ups)

Group

Ons

Beha

vior

whe

owner

presen

parture

-greeting

havior

Symp

toms

hen

owner

abs

ent

From

pupp

yhood

The

timin

onset

sym

ptoms

when

left

sign

iﬁcant,

typically

eve

time

within

the

ﬁrst

minute

and

oft

alm

ost

immed

iately

after

actual

virtu

remov

the

speciﬁc

social

imulus

that

the

dog

depe

ndent

Orga

nization

all

activities

around

speciﬁ

social

stimulus

Follo

wing

about

the

use

Phys

ical

ntact

need

(eg,

leanin

owner,

slee

ping

own

er,

wanting

held)

Needy

tention

/aﬀ

ection

and

seeking

havior

Explo

ratory

havior

depend

ent

presen

speciﬁ

soc

ial

stimulus

the

dog

dependen

Distre

sign

(eg,

trembling,

sha

king,

howl

ing,

thdrawal)

whe

eparture

antic

ipated

Possib

attempts

prevent

depa

rture

pressio

appe

aseme

beha

viour

possib

result

antic

ipation

punis

hment

whe

own

returns

Destructio

pically

involve

attempts

regain

contac

with

the

owner

and

orie

ntated

toward

door

and

windows

that

give

acc

ess

the

directio

which

the

speciﬁ

social

stimu

lus

left

Destructive

beha

vior

involving

item

impregnated

with

the

owner’s

scent,

such

shoes

pape

rs,

beddin

and

remote

controls

Vocalization

consis

tent

ith

separa

tion

distre

/relo

cation

Ons

after

remov

one

salie

stimulus,

several

less

signiﬁc

ant

stimuli

chan

the

need

the

animal

rely

the

mai

ntenanc

set

cause

rehom

ing,

ving

hous

left

oth

room

than

norm

ally,

whe

ustrated

becau

viation

norm

patte

rns,

ter

holida

illness,

agin

are

exa

mples

Depend

ency

behavio

norm

ally

directe

toward

one

several

social

stimu

li;

howev

er,

dogs

can

also

depe

ndent

non-soc

ial

stimu

(eg,

certain

location

the

hous

Depend

ency

toward

soc

ial

stimuli

may

increa

unpred

ictabilit

separa

tion

and

frustra

tion

increases

Onset

disp

lay

dependen

behavio

may

occur

conseq

uence

increa

sed

parture

distress

and

excess

ive

greeting

norm

ally

but

not

cessarily

directe

one

soc

ial

stimuli

Possib

attempts

prevent

depa

rture

parture

distress,

agit

ation,

depressio

pressio

appe

aseme

beha

vior

possible

resu

antic

ipation

punis

hment

whe

own

returns

overd

epen

dent

soc

ial

stimuli

destruct

ion

typic

ally

occurs’

result

tempts

rega

acc

ess

the

ind

ividual(s)

Alternat

ively,

the

dog

may

seek

maintenanc

inanim

ate

stimuli

or,

fearful,

escape

any

door

indow

(eg,

tryin

increa

meosta

sis

digg

ing

into

loc

ations

hid

gain

access

room

shut

oﬀ

them

that

they

associa

with

332

D. Appleby, J. Pluijmakers / Vet Clin Small Anim 33 (2003) 321–344

Onl

when

the

dog

left

circumst

ances

where

its

maintenanc

set

inade

quate

increa

sed

salie

nce

the

stimu

lus

mai

ntenan

stimu

li,

such

the

own

er’s

droom).

Vocalization

nsistent

with

sep

aration

distre

/reloca

tion

(may

not

occur

cause

distre

fear

externa

stimuli

relia

nce

nsocial

stimuli)

Defecat

ion

and

urin

ation

alone

mbinat

ion

with

oth

sympto

sug

gest

the

possible

involve

ment

fear-

eliciting

stimulus

(eg,

noi

obia)

Onset

coincides

with

fearfu

obic

exper

ience

noxio

event,

which

may

asso

ciated

with

and

trigge

red

the

absence

the

own

er(s)

inapp

ropriate

depend

ency

havior

Rea

ctio

fearfu

sti

mulus

also

disp

layed

hen

own

er(s)

present

alt

hough

the

extent

which

fear

expre

ssed

whe

the

owner(s)

pre

sent

may

reduc

becau

the

set

mai

ntenanc

stimuli

mplete

and

beha

vior

ord

ered

Distre

signs

can

develop

resulting

from

inc

rease

predict

ability

and

anxiety

owner’s

abs

ence

asso

ciated

with

noxio

stimuli

Defecat

ion

and

urin

ation

alone

mbinat

ion

with

oth

sympto

sug

gest

the

possible

involve

ment

fear

elic

iting

imulus

(eg,

noise

obia)

the

dog

tries

pe,

struction

ran

dom

object

may

cause

resu

tryin

escap

hid

333

D. Appleby, J. Pluijmakers / Vet Clin Small Anim 33 (2003) 321–344

For treatment to be successful, fear-eliciting stimuli that cause or

contribute to the disturbance of homeostasis have to be identiﬁed and
removed or their eﬀect reduced. The balance in the maintenance set has to
be restored by (1) reintroduction of maintenance stimuli, (2) reducing the
dependency on one speciﬁc or several stimuli, (3) increasing the dependence
on alternative stimuli, or (4) a combination of these.

Removal of maintenance stimuli from dogs in Group C does not cause

disturbance of homeostasis, because the breadth of the overall set of social
and environmental stimuli means that the dog has suﬃcient stimuli available
to maintain control of the parasympathetic system. Members of this group
can become fearful or phobic as a result of experiencing a noxious event,
however, which may or may not be associated with and triggered by the
absence of the owner. If the dog tries to cope, destruction of random objects
may be caused as a result of trying to escape or hide. If the level of anxiety is
high, symptoms like defecation and urination are possible. Systematic
desensitization and counterconditioning responses to fear-eliciting stimuli
form an essential part of treatment [12,63]. The level to which the dog’s
response to its stimulus set is disrupted by anticipation and

/or generalization

to other stimuli has to be evaluated and treated if necessary [41].

Treatment

Every case requires a treatment program devised for the animal’s needs,

the owner’s circumstances, and the environment the dog is to be left in.
The rationales discussed here provide the essence of what may have to
be considered for conditioning relaxation that is not disproportionately
dependent on social or nonsocial maintenance stimuli. Furthermore,
treatment should be phased to avoid an unintentional increase in anxiety,
which might otherwise be induced by radical alteration of the dog’s
circumstances and relationship with its owner [63]. Separation distress and
its consequences are likely to continue while treatment is taking eﬀect [3],
and owners should be advised accordingly. Where possible, short-term
management, such as the use of a dog sitter when the dog must be left, can
reduce the potential for this to occur [64].

Reducing the salience of the person(s) on whom the dog is dependent
and developing alternative maintenance stimuli

Ignore attention-seeking behavior

Attention-seeking behavior can be associated with distress during owner

absence [9,12]. Because it is indicative of overdependence and sympathetic
arousal, such behavior should be ignored to avoid unwitting reinforcement
[3,13,23], but vocal or physical rejection should not be used, because

334

D. Appleby, J. Pluijmakers / Vet Clin Small Anim 33 (2003) 321–344

reinforcement results from the attention given. Interaction by touch, voice,
and eye contact should be initiated and concluded by the owner at times
when the dog is relaxed so as to reinforce relaxation and develop in-
dependent behavior [13,23,62,63,65]. In some cases, there is a potential
for dependence to transfer to a new social stimulus. If this occurs, it is
necessary for all family members to control interaction in the same way.
The addition of scheduled regular sessions of attention that the dog can
predict may help it to relax [12,13]. Such sessions may also improve
client compliance because they are in keeping with the perception of
pet ownership. Owners should be warned that attention-seeking behavior
is likely to increase before it decreases, because absence of an expected
response increases the vigor of the behavior and owner response unwittingly
reinforces it, making it likely to reoccur [72].

Reduce physical contact

If a dog tends to remain within a meter of an owner or in physical con-

tact whenever it settles, this is indicative of overdependence [3,6,19,24].
Preventing the dog from sitting on furniture next to the owner or on the
owner’s lap reduces reinforcement of dependent behavior and the contrast
between owner presence and absence [3,23,24,73]. Conversely, attention
given when the dog chooses to lie at a distance from the owner in a relaxed
manner develops independent behavior [64,74].

Divide tasks

If the dog seems to be dependent on a particular individual for activities

that enhance attachment, such as playing, feeding, walking, and training,
these activities should be shared by other members of the household
whenever possible [4,65]. The feeding of gratuitous tidbits as opposed to
rewards should be stopped so as to reduce the salience of the provider [23].

Stimulate independent behavior

Self-rewarding activities when the owner is present can help to develop

independent behavior. Examples include encouraging the dog to lie on its
bed with a chew [74] or to play with toys that cause it to work for food to be
released [5,20,23,63,65] and games that encourage it to search for food or
toys during walks or in the owner’s yard.

Sleeping location

Although sleeping with the owner is not causal [12,13] in cases where the

dog sleeps in the owner’s bedroom because it is distressed when separated
from him

/her, it is advisable that the dog be conditioned to be able to sleep

in another location [23]. This can be achieved by moving it from the owner’s
bed (if it sleeps on it) and onto a bed of its own, which, in turn, is gradually

335

D. Appleby, J. Pluijmakers / Vet Clin Small Anim 33 (2003) 321–344

moved out of the room. Subsequently, a dog gate or child gate can be used
across the open bedroom door, and when the dog is ready, a series of
relocations can be used to move the dog gradually toward where it is
ultimately expected to sleep [23].

Canine companion

If separation anxiety is caused by overdependence on a canine

companion, the treatment principles should be adapted to reduce the
salience of this dog and to encourage the development of alternative
maintenance stimuli. In most cases, owners do not realize that this may have
been an issue until after the demise or permanent departure of the dog onto
which dependence was placed. In these circumstances, overdependence may
have transferred to another social stimulus, such as an owner, before the
problem is presented.

Providing maintenance stimuli during owner absence

The reduction of disproportionate dependence on social and nonso-

cial stimuli cannot be addressed without developing the dog’s capacity
to maintain emotional homeostasis through alternative stimuli, although
stimuli associated with social contact can be used [23].

Relaxation cues associated with maintenance stimuli

Relaxed behavior in the owner’s presence (parasympathetic autonomic

response) can become associated with a visual, auditory, or olfactory
stimulus, which can then be used to trigger relaxation during the owner’s
absence by putting it in place before departure [13,23,63,64,74].

Food items, such as chews and palatable food pieces hidden in a toy,

can generate relaxation and become relaxation cues during owner absence
if they are introduced gradually. As with other relaxation cues that are
purposely developed rather than preexisting, they should be introduced
when the dog is relaxed and the owner is present but not interacting with it.
Subsequently, they can be used in conjunction with systematic desensiti-
zation sessions and then to stimulate relaxation when the dog is left
unattended [5,20,23,62,63]. The item should be removed when access to the
owner is re-established during therapeutic sessions and when the owner
returns home during actual use. Failure to show interest in food items
during separation from the owner is indicative of sympathetic arousal [3,13].

Dog appeasing pheromone (DAP, Ceva, Chesham, Bucks, UK) is a

synthetic version of a secretion from sebaceous glands between the
mammary glands produced during lactation [75] that is atomized by a
plug-in device. It is claimed to have a beneﬁcial eﬀect in the treatment of
separation problems [75], and the indications for its use [75] suggest that it
stimulates relaxation.

336

D. Appleby, J. Pluijmakers / Vet Clin Small Anim 33 (2003) 321–344

Any stimuli, such as the sound of the television or radio, that are

normally associated with relaxation in the owner’s presence should be left
on when the dog is left unattended so as to provide continuity [20,23,63].
Recordings of voices and sounds that occur when members of the household
are at home can also be used for this purpose [20,23]. It is important that
these are also used at times when the owner is present to prevent their
becoming a cue for imminent departure.

An owner on whom the dog is dependent should leave cloths

/blankets

impregnated with his

/her scent in the place where the dog is known to lie

when left alone [20]. Putting items with unwashed laundry for a few hours
before each use refreshes the scent [3].

Changing the environment

Fear and anxiety can be associated with areas of the home in which

the dog has experienced these emotions [20]. Providing an alternative loca-
tion for the dog to settle in during owner absence can make stronger
maintenance stimuli available and

/or reduce the salience of or remove fear-

eliciting stimuli [20,31]. The change of location can be indeﬁnite or until a
positive association with the original location has been developed. The
original location can be adjusted to suit the individual’s needs, for example,
by creating free access to a sound-reducing den to hide in if the dog reacts to
fear-eliciting sound stimuli. Once associated with relaxation when the owner
is present and subsequently during systematic desensitization sessions,
conﬁnement in a crate can be used for some dogs [24]. Abrupt conﬁnement
may increase anxiety, however [20,24,63].

Systematic desensitization to departure cues and
separation from the owner

Desensitization to departure cues

While the dog is in a relaxed state, stimuli associated with owner

departure, such as the sound of car keys or putting a coat and shoes on,
should be performed when the owner is not leaving but is instead
performing activities associated with remaining at home. It is important
that the dog remains relaxed, to which end the level of stimulation should be
increased gradually. Owners can subsequently combine an increasing
number of stimuli [4,5,12,13,24,63–65].

Desensitization to owner absence

Some authors have advocated training the dog to sit or lie at a distance

from the owner in a state of relaxation [5,12,24,63,65], which can be
associated with the relaxation cue discussed previously. Initially, the owner

337

D. Appleby, J. Pluijmakers / Vet Clin Small Anim 33 (2003) 321–344

should only take a few steps away before returning to reward the dog.
Provided that the dog remains relaxed, the distance and duration of absence
are gradually increased; however, they should be increased on a variable
schedule to prevent dog from predicting the owner’s return, which might
otherwise result in its behavior being disrupted if its expectations are not met.

An alternative approach to training that does not risk increasing the

salience of the owner through the interaction involved requires the use of a
child gate or dog gate [23]. Initially, this can be used to prevent the dog from
following the owner from room to room as he

/she moves about the house.

Because the dog can see the owner through the gate, it is less likely to be
distressed by its use than by that of a closed door.

On some of the occasions when the owner stays in a room for an

extended period of time, the gate can be used to keep the dog in an adjacent
room. The technique should be used for variable periods in conjunction with
relaxation. Once it is evident that the dog has been conditioned to relax in
these conditions, the gate can be repositioned so that it retains the dog in an
area further away from the owner. Subsequently, the dividing doors can be
left less ajar, and, ﬁnally, they can be closed.

To achieve optimal progress during the treatment period, the dog should

only be subjected to separations it can tolerate [20,24,63,65]. If longer
separations are inevitable, the relaxation cue should only be used during
therapeutic separations. If feasible, the dog can be placed in a diﬀerent part
of the house during separations that have to occur in the course of everyday
events [20].

Leaving and returning rituals

Owner interaction before departure can reinforce anxiety [4,5,24]. To

avoid this, interaction should be withdrawn approximately 30 minutes before
the owner leaves. The dog should be placed with relaxation cues in a place
where it has learned to be relaxed when separated from the owner when he

she is at home. When it is evident that the dog is relaxed, the owner can leave
but without speaking [5,23,63–65]. Excessive greeting behavior displayed
by the dog when the owner returns should be ignored so as to avoid the
unwitting reinforcement of the associated emotional disturbance. Con-
versely, the owner should respond to and therefore reinforce relaxed greeting
behavior [23,64], such as sitting. It is also important to note that what seems
to be excessive greeting behavior can be appeasement caused by anticipation
of owner aggression carried out as misguided attempts to punish.

Avoiding punishment

Punishment for perceived wrongful behavior during the owner’s absence

is not an eﬀective technique for changing that behavior [1,9,12,24], and the

338

D. Appleby, J. Pluijmakers / Vet Clin Small Anim 33 (2003) 321–344

emotional state caused by anticipation of the owner’s apparently un-
provoked aggression is one of the most common reasons why separation-
related behavior worsens [13,24]. Owners often believe that their dog ‘‘looks
guilty,’’ but this is misinterpretation of a posture motivated by fear [24]. It is
therefore important that owners ignore any damage or soiling found on
their return [3,23,24].

Systematic desensitization and counterconditioning to fear-eliciting stimuli

In cases where distressed behavior occurs because of fear of speciﬁc

stimuli and stimuli that have become associated with them, such as rain on
windows as a predictor of thunder or the owner’s absence if it is associated
with noxious stimuli, the dog’s response has to be altered through systematic
desensitization and

/or counterconditioning. These processes involve either

predisposing relaxation and gradually increasing the level of exposure to the
stimuli or pairing the stimuli with another, such as food, that results in a
response that is incompatible with fear [24,39,64]. The stimuli can be real or
recorded; whichever method is in use, it must always be presented at a level
that is within the dog’s capacity to remain relaxed and increased at a rate
that is compatible with the dog’s continuing to develop an association with
remaining relaxed. For dogs represented in Group B, it may also be
necessary to address overdependence on social stimuli. For dogs in Groups
A and B, the involvement of fear-eliciting stimuli has to be assessed and
treated as necessary.

Drug support for behavioral modiﬁcation

Choice of drug therapy is dependent on the nature of the disturbance

to homeostasis and the nature of action required; therefore, an accurate
diagnosis of anxiety, fear response to threat, or a combination of both is
essential [3,4,13]. Whether drug support is used is dependent on clinical
judgment and the severity of the disturbance. It should always be used as an
adjunct to behavioral therapy as a means of achieving homeostasis more
quickly, thereby increasing the likelihood of behavioral therapy being suc-
cessful, and to prevent further disruption of homeostasis and, in some cases,
block memory of the disruption [74].

Table 3 shows how elements of the treatment program may be ap-

propriate to the three groups and can be phased.

Summary

The model discussed here should help to provide an understanding of the

range of stimuli that each dog needs for the maintenance of emotional
homeostasis and the interplay between these stimuli and events in the dog’s

339

D. Appleby, J. Pluijmakers / Vet Clin Small Anim 33 (2003) 321–344

Table 3
Elements of treatment program

Possible stage of introduction

Group A

Group B

Group C

Reducing the salience of the person(s) on whom

the dog is dependent and developing alternative
maintenance stimuli
Ignore attention-seeking behavior; all

interactions are initiated and concluded
by the owner at times when the dog is
relaxed

Phase 1

Schedule frequent and regular attention

sessions

Phase 1

Reduce physical contact (eg, lying on lap)

Phase 1

Decrease dependency on a particular

individual by dividing tasks

Phase 1

Stop feeding gratuitous tidbits

Phase 1

Stimulate independent behavior by

providing self-rewarding activities

Phase 1

Change sleeping location; if the dog

sleeps in the bedroom, gradually move
it to another location

Phase 3

Providing maintenance stimuli during owner absence

Develop a relaxation cue when the owner

is present associated with maintenance
stimuli

Phase 1 or 2

Provide relaxation cues during systematic

desensitization sessions (eg, chew toy,
television, voice recordings, clothes)

Phase 2

Provide relaxation cues during owner absence

Phase 3

Change the environment; provide an alternative

location to settle in during owner absence
with stronger maintenance stimuli, or adjust
present location to the dog’s need

Phase 1

Dog appeasing pheromone

Phase 1

Remove fear-eliciting stimuli if possible

Phase 1

Systematic desensitization to departure cues and

separation from the owner
Desensitize to departure cues

Phase 2

Systematically desensitize to owner absence

Phase 2

Stop the dog from following throughout the

house

Phase 2

Leaving and returning rituals

Withdraw all interaction 30 minutes before

leaving

Phase 1

Place dog in location it has learned to relax in

Phase 3

Phase 1

Ignore excessive greeting behavior and

reinforce relax greeting behavior

Phase 1

340

D. Appleby, J. Pluijmakers / Vet Clin Small Anim 33 (2003) 321–344

environment. In turn, this should aid in diagnosis and the setting of
appropriate treatment plans.

Acknowledgments

The authors thank Tiny De Keuster and Emmanuel Gaultier for their

help with translation and their comments. They also acknowledge addi-
tional comments provided by John Bradshaw and Emily Blackwell.

References

[1] Borchelt PL, Voith VL. Diagnosis and treatment of separation-related behaviour problems

in dogs. Vet Clin North Am Small Anim Pract 1982;12:625–35.

[2] Bradshaw JWS, McPherson JA, Casey RA, et al. Aetiology of separation related behaviour

in the domestic dog. Vet Rec 2002;151:43–6.

[3] King JN, Simpson BS, Overall KL, et al. Treatment of separation anxiety in dogs with

clomipramine: results from a prospective, randomised, double-blind, placebo-controlled,
parallel-group, multicentre clinical trial. Appl Anim Behav Sci 2000;67:255–75.

[4] Pageat P. General psycho-psychology and nosography of behaviour disorders of dogs. In:

Pathologie du comportement du chien. 2e`me e´dition. Paris: Editions du Point Ve´te´rinaire;
1998. p. 43–112, 265–363.

[5] Takeuchi Y, Houpt KA, Scarlett JM. Evaluation of treatments for separation anxiety in

dogs. JAVMA 2000;217:342–5.

[6] McCrave EA. Diagnostic criteria for separation anxiety in the dog. Vet Clin North Am

Small Anim Pract 1991;21:247–55.

[7] Heath S. Dealing with separation problems in dogs. In: Scientiﬁc Proceedings of the 45th

Annual Congress of the British Small Animal Association. Gloucester, UK: British Small
Animal Veterinary Association; 2002. p. 536–8.

[8] Heath S. Separation-related behaviour problems in pet dogs. Vet Times 2001;31:11–2.
[9] Lindsay RL. Attachment, separation, and related problems. In: Handbook of applied dog

behavior and training, etiology and assessment of behaviour problems. Ames: Iowa State
University Press; 2000. p. 93–129.

Table 3 (continued )

Possible stage of introduction

Group A

Group B

Group C

Systematic desensitization and counter-

conditioning to fear-eliciting stimuli
Identify fear-eliciting stimuli and start

systematic desensitization and
counterconditioning program during owner
presence; this may also necessitate systematic
desensitization and counterconditioning to
departure cues and separation from the
owner if these have become a conditioned
stimulus for fear

Phase 1

Punishment (always inappropriate and should

always be stopped)

Phase 1

Drug support (where appropriate)

Phase 1

341

D. Appleby, J. Pluijmakers / Vet Clin Small Anim 33 (2003) 321–344

[10] Overall KL, Dunham AE, Frank D. Frequency of nonspeciﬁc clinical signs in dogs with

separation anxiety, thunderstorm phobia, and noise phobia, alone or in combination.
JAVMA 2001;219:467–73.

[11] Mill DS, Sheppard G. An evolutionary approach to the analysis, assessment and treatment

of behaviour problems in companion animals. In: Proceedings of the Second World
Meeting on Ethology, 1999. p. 61–8

[12] Overall KL. Fear, anxieties, and stereotypies. In: Clinical behavioral medicine for small

animals. St. Louis: Mosby; 1997. p. 209–50.

[13] Simpson BS. Canine separation anxiety. Compend Contin Educ Pract Vet 2000;22:328–39.
[14] Gaultier E. Separation related behaviour problems: diagnostic criteria identiﬁcation using

a cluster analysis. In: Proceedings of the Third International Congress on Veterinary
Behavioural Medicine. Wheathampstead: Universities Federation for Animal Welfare;
2001. p. 76–82.

[15] Iracka J. French and American approach to small animal behavioural disorders: a few

examples. In: Proceedings of the Second World Meeting on Ethology, 1999. p. 69–75.

[16] Mills DS, Heath SE, Harrington LJ, Workshop B. Dialogue regarding the deﬁnition of

separation anxiety (what is it?) and diagnostic criteria of clinical separation anxiety in the
dog. In: Proceedings of the First International Conference on Veterinary Behavioural
Medicine. Wheathampstead: Universities Federation for Animal Welfare; 1997. p. 181.

[17] Fox M. Vocalizations in wild canids and possible effects of domestication and stages and

periods in development: environmental inﬂuences and domestication. In: The dog: its
domestication and behaviour. New York: Garland, S & PM Press; 1978. p. 69–89, 153–76.

[18] Overall KL, Agulnick L, Dunham AE, et al. Qualitative and quantitative differences

in vocalizations by dogs affected with separation anxiety and unaffected dogs using
sonographic analysis. In: Proceedings of the Second World Meeting on Ethology, 1999.
p. 108–13.

[19] Overall KL, Dunham AE. Disagrees with alternative view of separation anxiety. JAVMA

2001;219:1520.

[20] Beaver BV. Canine social behavior. In: Canine behavior: a guide for veterinarians.

Philadelphia: WB Saunders; 1999. p. 137–81.

[21] Mugford RA. Canine behavioural therapy. In: Serpell J, editor. The domestic dog.

Cambridge: Cambridge University Press; 1995. p. 139–52.

[22] Flannigan G, Dodman NH. Risk factors and behaviors associated with separation anxiety

in dogs. JAVMA 2001;219:460–6.

[23] Podberscek AL, Hsu Y, Serpell JA. Evaluation of clomipramine as an adjunct to

behavioural therapy in the treatment of separation-related problems in dogs. Vet Rec
1999;145:365–9.

[24] Voith VL, Borchelt PL. Separation anxiety in dogs. Compend Contin Educ Pract Vet

1985;7:42–52.

[25] Seksel K, Coyle W, Chaseling S. Stresses associated with moving house with pets.

In: Proceedings of the Third International Congress on Veterinary Behavioural Medicine.
Wheathampstead: Universities Federation for Animal Welfare; 2001. p. 47–9.

[26] Serpell J, Jagoe A. Early experience and the development of behaviour. In: Serpell J, editor.

The domestic dog: its evolution, behaviour and interactions with people. Cambridge:
Cambridge University Press; 1995. p. 79–102.

[27] Bowlby J. Attachment and loss, vol. 1. Attachment. Middlesex: Pequin; 1959, . Cited in:

Gross R, editor. Attachment and separation: the effects of early experience. In: Psychology:
the science of mind and behaviour. London: Hodder & Stoughton; 1996. p. 548–73.

[28] Harlow HF, Zimmerman RR. Affectional responses in the infant monkey. Science

1959;130:421–32.

[29] Schaffer HR, Emmerson PE. The development of social attachments in infancy. In:

Monographs of the Society for Research in Child Development, number 3. 1964. p. 29.
Cited in: Gross R, editor. Attachment and separation: the effect of early experience. In:

342

D. Appleby, J. Pluijmakers / Vet Clin Small Anim 33 (2003) 321–344

Psychology: the science of mind and behaviour, 3rd edition. London: Hodder &
Stoughton; 1996. p. 548–73.

[30] Lindell EM. Diagnosis and treatment of destructive behaviour in dogs. Vet Clin North Am

Small Anim Pract 1997;27:533–47.

[31] Hetts S. Canine escape problems. In: Pet behavior protocols: what to say, what to do, when

to refer. CO: AAHA Press; 1999. p. 119–32.

[32] Hothersall D, Tuber DS. Fears in companion dogs: characteristics and treatment. In:

Keehn JD, editor. Psychopathology in animals. New York: Academic Press; 1979. Cited in:
Borchelt PL, Voith VL. Diagnosis and treatment of separation-related behaviour problems
in dogs. Vet Clin North Am Small Anim Pract 1982;12:625–35.

[33] Damkjer Lund J, Jorgensen MC. Behaviour patterns and time course of activity in dogs

with separation problems. Appl Anim Behav Sci 1999;63:219–36.

[34] Solomon RL. Corbit JD. An opponent-process theory of motivation: 1. Temporal dyna-

mics of affect. Psychol Rev 1974; 81:119–45. Cited in: Lindsay RL. Attachment, separation,
and related problems. In: Handbook of applied dog behavior and training, etiology and
assessment of behaviour problems. Ames: Iowa State University Press; 2000. p. 93–129.

[35] O’Farrell V. Stress, anxiety and arousal. In: Manual of canine behaviour. 2nd edition.

Gloucestershire: British Small Animal Veterinary Association; 1992. p. 37–44.

[36] Gray JA. Fear and frustration. In: The psychology of fear and stress. 2nd edition.

Cambridge: Cambridge University Press; 1987. p. 174–206.

[37] Boissy A. Fear and fearfulness in animals. Q Rev Biol 1995;70:165–91.
[38] McFarland D. Homeostasis and behaviour. In: Animal behaviour: psychobiology, etho-

logy and evolution. 3rd edition. Paciﬁc Grove: Paerson Education Limited; 1999. p. 259–83.

[39] Lieberman DA. Foundations of conditioning. In: Learning, behaviour and cognition. 2nd

edition. Paciﬁc Grove: Brooks Cole; 1992. p. 76–107.

[40] Manning A, Dawkins MS. Learning and memory. In: An introduction to animal

behaviour. 4th edition. Cambridge: Cambridge University Press; 1992. p. 122–49.

[41] Toates F. Stress. Chichester: Wiley; 1995. Cited in: McFarland D. Animal behaviour: psy-

chobiology, ethology and evolution. 3rd edition. Essex: Paerson Education Limited; 1999.

[42] DorlandÕs pocket medical dictionary. Philadelphia: WB Saunders; 1989.
[43] Cairns RB. Attachment behavior of mammals. Psychol Rev 1966;73:409–26.
[44] Scott JP. The phenomenon of attachment in human-nonhuman relationships. In: Davis H,

Balfour D, editors. The inevitable bond: examining scientist-animal interactions. Cam-
bridge: Cambridge University Press; 1992. p. 72–3.

[45] Bateson P. Control of sensitivity to the environment during development. In: Immelmann

K, Barlow GW, Petrovich L, Main M. editors. Behavioural development. Cambridge:
Cambridge University Press; 1981. p. 432–53.

[46] Gross R. Attachment and separation: the effects of early experience. In: Psychology: the

science of mind and behaviour. London: Hodder & Stoughton; 1996. p. 548–73.

[47] Gubernick DJ. Parent and infant attachment in mammals. In: Gubernick DJ, Klopfer PH,

editors. Parental care in mammals. New York: Plenum; 1981. p. 243–305.

[48] Bourdin M. Development of the young dog and behavioral disorders. In: Proceedings of

the Second World Meeting on Ethology, 1999. p. 215–7.

[49] Fox MW. Developmental aspects of some conditioned and learned responses. In:

Integrative development of the brain and behaviour in the dog. Chicago: University of
Chicago Press; 1971. p. 185–225.

[50] Whoolpy JH, Ginsberg BE. Wolf socialization: a study of temperament in a wild social

species. Am Zool 1967;7:357–63.

[51] Murphree OD, Dykman FA. Litter patterns in the offspring of nervous and stable dogs:. 1.

Behavioural tests. J Nerv Ment Dis 1965;141:321–2.

[52] Appleby DL, Bradshaw JWS, Casey RA. Relationship between aggressive and avoidance

behaviour by dogs and their experience in the ﬁrst six months of life. Vet Rec
2002;150:434–8.

343

D. Appleby, J. Pluijmakers / Vet Clin Small Anim 33 (2003) 321–344

[53] Scott JP, Fuller JL. The development of behaviour and the critical period. In: Genetics and

social behaviour in the dog. Chicago: University of Chicago Press; 1965. p. 84–150.

[54] Elliot O, Scott JP. The development of emotional distress reactions to separation in

puppies. J Genet Psychol 1961;99:3–22.

[55] Gurski JC, Davis K, Scott JP. Interaction of separation discomfort with contact comfort

and discomfort in the dog. Dev Psychobiol 1980;13:463–7.

[56] Scott JP. Critical periods in behavioural development: critical periods determine the

direction of social, intellectual and emotional development. Science 1962;138:949–58.

[57] Cairns RB, Johnson DL. The development of interspecies social attachments. Psychonomic

Sci 1965;2:337–8.

[58] Cairns RB, Werboff J. Behavior development in the dog: an interspeciﬁc analysis. Science

1967;158:1070–2.

[59] Tuber DS, Hothersall D, Peters MF. Treatment of fears and phobias in dogs. Vet Clin

North Am Small Anim Pract 1982;12:607–23.

[60] Pettijohn TF, Wong TW, Ebert PD, et al. Alleviation of separation distress in 3 breeds of

young dogs. Dev Psychobiol 1977;10:373–81.

[61] Ainsworth MDS, Blehar MC, Waters E, Wall S. Patterns of attachment: a psychological

study of the strange situation. Hillsdale (NJ): Erlbaum; 1978.

[62] Appleby DL. Separation problems. In: Ain’t misbehavin’. Bristol: Broadcast Books; 1997.

p. 239–60.

[63] Landsberg G, Hunthausen W, Ackerman L. Destructive behaviour. In: Behaviour

problems of the dog and cat. Oxford: Butterworth-Heinemann; 1997. p. 97–106.

[64] Hetts S. Destructive behavior in dogs. In: Pet behavior protocols: what to say, what to do,

when to refer. CO: AHAA Press; 1999. p. 103–17.

[65] O’Farrell V. Problems of emotionality. In: Manual of canine behaviour. 2nd edition.

Gloucestershire: British Small Animal Veterinary Association; 1992. p. 103–4.

[66] Dantzer R. Behavioural, physiological and functional aspects of stereotyped behavior: a

review and a reinterpretation. J Anim Sci 1986;62:1776–86.

[67] Domjan M. Habituation and sensitization. In: The essentials of conditioning and learning.

Austin: Wadsworth/Thompson Learning; 2000. p. 25–40.

[68] Henry JP. Present concept of stress theory. In: Usdin E, Kvetnansky R, Kopin IJ, editors.

Catecholamines and stress: recent advances. New York: Elsevier North Holland; 1980.
p. 557–71.

[69] Weiss JM. Physiological factors in stress and disease. Sci Am 1972;226:104–13. Cited in:

Boissy A. Fear and fearfulness in animals. Q Rev Biol 1995;70:165–91.

[70] Archer J. Behavioural aspects of fear. In: Sluckin W, editors. Fear in animals and man.

New York: Van Nostrand Reinhold; 1979. p. 56–85.

[71] Beaver BV. Canine eliminative behavior. In: Canine behavior: a guide for veterinarians.

Philadelphia: WB Saunders; 1999. p. 267–87.

[72] Lieberman DA. Reinforcement. In: Learning, behaviour and cognition. 2nd edition.

Paciﬁc Grove: Brooks Cole; 1992. 194–246.

[73] Nack RA. Managing separation anxiety in a dog. Vet Med 1999;94:704–16.
[74] Askew HR. Fear problems and separation anxiety. In: Treatment of behaviour problems in

dogs and cats. Blackwell Science. p. 204–27.

[75] Mills DS. Considerations on the use of pheromonotherapy and pharmacotherapy for

the treatment of companion animal behaviour problems. In: Proceedings of the 2002
CABTSG

/AWSELVA Study Day, Birmingham. Oxford: CABTSG-AWSELVA; 2002,

p. 31–3.

344

D. Appleby, J. Pluijmakers / Vet Clin Small Anim 33 (2003) 321–344

Behavior genetics

Yukari Takeuchi, DVM, PhD

Katherine A. Houpt, VMD, PhD

Laboratory of Veterinary Ethology, The University of Tokyo, Yayoi,

Bunkyo-ku, Tokyo 113–8657, Japan

Animal Behavior Clinic, College of Veterinary Medicine, Cornell University,

Ithaca, NY, USA

Human beings have long noticed individuality in animals. To take just the

case of companion animals, it is obvious that even within the same species
and the same breed, behavior is not homogeneous; individual animals may
be excitable, aggressive, docile or nervous. The sum of these behavioral char-
acteristics is called temperament. An understanding of temperament and the
biologic background of individual diﬀerences not only contributes to the
growth of basic neuroscience research but is important within veterinary
medicine because of its great signiﬁcance in our attempt to ﬁnd an ap-
propriate relationship for coexistence between human beings and animals.

It is still unclear whether individuality is determined by heredity or

whether the early environment is more important. As common sayings
like ‘‘the child is father of the man’’ and ‘‘genius displays itself even in
childhood’’ indicate, the argument of heredity versus environment has long
been a matter of debate. Research in the ﬁeld of clinical developmental
psychology has yielded important results about the inﬂuence of the early
environment, whereas, the identiﬁcation of the existence of genes related
to temperament in human beings has drawn much attention recently. This
article brieﬂy discusses the history of behavioral genetics, concentrating on
the study of companion animals, and considers recent research trends.

Classic Mendelian genetic research

The Mendelian laws of heredity were published in 1865 and rediscovered

by the beginning of the twentieth century, when they had come to be
generally accepted. Researchers began to favor analyzing behavioral

Vet Clin Small Anim

33 (2003) 345–363

* Corresponding author.
E-mail address:

aytake@mail.ecc.u-tokyo.ac.jp (Y. Takeuchi).

0195-5616/03/$ - see front matter

PII: S 0 1 9 5 - 5 6 1 6 ( 0 2 ) 0 0 1 1 6 - X

tendencies that could be explained through these laws. In early research,
behavioral tendencies that could be evaluated objectively and easily, such
as hunting style and gun shyness in hunting dogs and style of livestock
management in herding dogs, were at the center of items for analysis [55].
Among studies concerning temperament, Thorne [2] revealed that exces-
sive shyness in Basset Hounds was caused by dominant genes. Later, Dykman
et al [3] developed a nervous strain of pointer, which has been maintained with
a normal strain.

The most famous research concerning the temperament of dogs and

heredity was that by Scott and Fuller [4], begun in 1945 at the Jackson
Laboratory and continuing for 13 years. The results of the research were
published as a book in 1965. The main aim of the research was to try to
evaluate the heritability of special behavioral traits in various breeds of dog
(Basenji, Beagle, American Cocker Spaniel, Shetland Sheep Dog, Wire-
Haired Fox Terrier, and crosses between the breeds) by analyzing those
behavioral traits in dogs raised in the same environment over generations.
When the researchers analyzed their data, they realized that there were
great individual diﬀerences within a single breed, which could be equivalent
to those among breeds, and that a simple Mendelian explanation was
impossible. They did ﬁnd some interesting breed eﬀects. For example, in one
study, puppies were weaned at 3 weeks and raised in pairs in isolation. Their
only contact with human beings was with one handler who either indulged
them (permitted to do anything they wanted with their human handler) or
disciplined them (made to sit and stay, come and, later, heel). At 8 weeks,
they were tested. The test was preventing the puppies from eating from a
dish of meat by slapping them on the rump and shouting ‘‘No!’’ and then
leaving the puppies alone with the meat. Two breeds, the Shetland Sheep
Dogs and the Basenjis, were unaﬀected by type of handling. The Basenjis,
whether disciplined or indulged, ate the meat with a short latency. The
Shelties would not eat whether they had been indulged or disciplined. The
other two breeds, the Beagles and the Wire-Haired Fox Terriers, diﬀered
depending on the way they were handled. The indulged puppies of both
breeds took signiﬁcantly longer to eat the food than the disciplined ones.
Later, when placed in a pen with other pups of the same breed, the indulged
Beagles were wary and shy. This study indicates the complex relation
between nature (breed) and nurture (early handling) [5].

Scott and Fuller [4] also found maternal inﬂuence on a puppy’s

temperament, which was best demonstrated by crossing Basenjis and
Cocker Spaniels. When the dam was a Cocker Spaniel and the sire was a
Basenji, the puppies were friendly to people, but in the reciprocal cross,
where the mother was a Basenji, the puppies were not friendly (Fig. 1). For
a more recent observation of maternal eﬀect, see Wilsson [64].

In the 1970s, the object research shifted to working dogs, such as guide

dogs and military dogs with a special breeding program. Signiﬁcant
heritability was not discerned among the items of analysis in a large-scale

346

Y. Takeuchi, K.A. Houpt / Vet Clin Small Anim 33 (2003) 345–363

survey of guide dogs in America [59], although a similar survey in Australia
reported high heritability in the evaluations of ‘‘success at becoming a guide
dog’’ and ‘‘fear’’ (Table 1) [49]. Mackenzie et al [8] estimated the heritability
for temperament scores of working German Shepherds as 0.51. High
(desirable) scores were associated with poor hip conformation.

We could see more detail in some reviews about the general theory of

classic genetic research, bringing together particulars that stretch across time
and topic [15,18,63]. Although the results of classic Mendelian research on
heredity have contributed greatly to breeding programs for working dogs,
for example, it is not possible to apply them to programs that have to suit
individual temperaments, such as the treatment of behavioral problems and
training programs.

Feline temperament seems to be more easily and consistently described

than canine temperament, perhaps because we demand less of cats. There
have been four studies on the inheritance of temperament in cats. Adamec
et al [10] tested laboratory-raised kittens with strange people and with the
vocalization of an aggressive adult cat. By 10 weeks, the cats demonstrated
behavior that would persist into adulthood; 25% of the cats were fearful.
Cats’ reactions to other people can be equable, hostile, fearful, or sociable.
Turner et al [11] found that litters sired by one of two tom cats and out of
several queens diﬀered in sociability and hostility depending on which tom

Fig. 1. Maternal eﬀect. Black symbols indicate hostile behavior. Gray symbols indicate friendly
behavior. Squares indicate males, and circles indicate females. When a hostile male Basenji (1) is
mated with a friendly female Cocker Spaniel (2), the puppies (3–6) are friendly. When a friendly
male Cocker Spaniel (7) is mated with a hostile female Basenji (8), the puppies (9–12) are hostile.

Table 1
Heritability of various traits

Breed

Trait

Heritability

English Setter

Hunting eagerness

0.22

German Wire-Haired Pointer

Water retrieving

0.32

German Short-Haired Pointer

Tracking

0.48

German Shepherd Dog

Temperament

0.51

Labrador Retriever

Nervousness

0.58

Data from

Houpt KA, Willis MB. Genetics of behaviour. In: Ruvinsky A, Sampson J,

editors. The genetics of the dog. New York: CABI Publishing; 2001. p. 371–400.

347

Y. Takeuchi, K.A. Houpt / Vet Clin Small Anim 33 (2003) 345–363

sired them. This indicates a paternal eﬀect (Fig. 2). A further study using the
same two toms tested the eﬀect of handling and paternity. The kittens were
handled daily from 2 to 12 weeks of age. The tests were approach to a
familiar person or a strange person and approach to a novel object, a box.
The results were that handled kittens of friendly fathers were more likely to
approach and interact with people and more likely to approach and enter
the box. The unhandled kittens of the unfriendly father were least likely to
approach people or the box. This indicates a general tendency to be either
bold or fearful. Handling did not aﬀect approach to the box, indicating that
socialization is speciﬁc to people and not to all stimuli [12]. Reisner et al [13]
also studied the eﬀect of handling and paternity. Thirteen litters sired by
ﬁve toms were tested. Kittens were weaned at 5 weeks and handled for
15 minutes three times a week for 3 weeks (ie, nine handling bouts).
Temperament tests commenced at 8 weeks and were repeated four times.
The temperament test consisted of a friendliness test and a response to
restraint test. In the friendliness test, one of the handlers sat in a circle and
the latency for the kitten to approach and tail posture were measured. The
kitten’s response to restraint for venepuncture was scored in the handling
test. The results indicated no diﬀerences as a result of early handling but a
signiﬁcant litter eﬀect. The oﬀspring of two of the ﬁve toms were friendlier
than those of the other three toms [13]. This study conﬁrms the British
ﬁnding of a paternal eﬀect on kitten temperament. No doubt if the kittens of
one tom and several queens were tested, an even stronger maternal eﬀect
would be found as was the case in dogs.

The complexities of the interaction between genetics and development are

seen in the calico cat cloned through nuclear transplantation, which was
born in 2002. The pattern of colored and white fur of the clone was diﬀerent
from the pattern of the donor cat [14]. Early in embryogenesis, all but one
X chromosome are functionally inactivated through a process called X
chromosome inactivation. The gene encoding orange coat color is X-linked
(ie, on the X chromosome). Black color is encoded by either codominant

Fig. 2. Paternal eﬀect. Black symbols indicate hostile behavior. Gray symbols indicate friendly
behavior. White symbols indicate unknown temperament. Squares indicate males, and circles
indicate females. When a friendly tom (1) is bred to several queens of unknown temperament
(2–4), most of the kittens (5–13) are friendly. When a hostile tom (14) is bred to several queens
(15–17) of unknown temperament, most of the kittens (18–26) are hostile.

348

Y. Takeuchi, K.A. Houpt / Vet Clin Small Anim 33 (2003) 345–363

allele on the X chromosome. In black patches, the X chromosome bearing
the orange allele has been inactivated and the X chromosome bearing the
nonorange allele is active. In patches of orange fur, the X chromosome
bearing the orange allele is activated. The random nature of X chromosome
inactivation is evident—there are relatively large patches of both black and
orange. If chance plays a part in simple pigmentation, it surely may be
involved in emotional reactivity as well.

Genetic research concerning behavioral problems

Dogs have a long history as domestic animals. As a result of ongoing

intensive breeding improvements for various purposes, such as hunting and
working, there are now, if we limit ourselves to the main ones found around
the world, more than 140 breeds. They vary in shape and size from the tiny
Chihuahua to the enormous Great Dane. Just as in looks, their behavioral
characteristics diﬀer greatly according to breed; some dogs are brave,
determined, and aggressive, whereas others are friendly and aﬀectionate to
everyone (Fig. 3). Despite this, scientiﬁc investigations into the relation
between breed and temperament have been sparse. Hart and Hart [15] at
the University of California surveyed obedience judges and veterinarians
as to their opinions about 13 behavioral characteristics of 54 representa-
tive breeds. Unsurprisingly, breeds used as guard dogs like the Doberman
Pinscher and German Shepherd ranked high in the behavioral traits of
territorial defense and watchdog barking, whereas hunting dogs like the
Labrador Retriever and Golden Retriever showed little aggressiveness and
were highly responsive to obedience training. It was shown in this way that
diﬀerences between breeds in terms of behavioral characteristics were
quantiﬁable. The inﬂuence of hereditary factors was indicated for the traits
of excitability and general activity.

The frequency of behavioral problems concerned with temperament, such

as aggressive behavior, level of anxiety, and obsessive-compulsive disorder

Fig. 3. Each breed has characteristic behavioral traits. Most Golden Retrievers are aﬀectionate
to people, whereas most Shibas show aggression toward strangers.

349

Y. Takeuchi, K.A. Houpt / Vet Clin Small Anim 33 (2003) 345–363

(stereotypy), varies according to breed, which suggests a connection with
hereditary factors. When aggressiveness appeared in Bernese Mountain
dogs, it was quickly eradicated, or at least reduced in frequency, by selective
breeding, indicating a heritable cause of the misbehavior [16].

There are some interesting data concerning connections between coat

color and aggressiveness. When queried, cat fanciers and veterinarians state
that tortoiseshell or calico cats are quite aggressive. To investigate this, we
compared the frequency with which cats of various coat colors—black,
orange, and tricolor—are presented to the Cornell University Hospital for
Animals for medical problems and to the Animal Behavior Clinic for
behavioral problems. There was no signiﬁcant diﬀerence in the proportions
of coat colors. There was a signiﬁcant eﬀect, and that was that cats
described by the owner as ‘‘and white’’ (ie, black and white, orange and
white) were less likely to be presented for behavioral problems.

Podberscek et al [17] of Cambridge University conducted a survey among

owners of English Cocker Spaniels concerning the aggressive behavior of
their dogs. Dogs with a coat of a solid color rather than a particolored coat
and dogs with red or golden coats rather than black coats tended to exhibit
aggressive behavior. Houpt and Willis [18] at Cornell University analyzed
data concerning the coat color of Labrador Retrievers. Their results
showed that yellow Labrador Retrievers made up a larger proportion of
dogs presented for aggressive behavior than those presented for medical
problems, whereas aggressive chocolate and black Labrador Retrievers did
not diﬀer in proportion from that of the medical population. Because coat
color is largely genetic, according to Mendelian laws, these data seem to
point to the possibility that tendencies toward aggressive behavior may be
hereditary and coinherited with genes controlling pigmentation. There is
a direct metabolic reason why coat color might be linked to aggression.
DOPA is the precursor of dopamine, a neurotransmitter, and of the pigment
melanin DOPA. Genes that code for the enzyme that synthesizes DOPA and
its products are probably involved in some way with the aggressive behavior
associated with coat color. The relation between coat color and tempera-
ment is a topic that researchers must consider seriously in the future.

It is well known that there are diﬀerences between male and female dogs

regarding the frequency of problem behavior. It has been reported that
dominance-related aggression and territorial aggression are much more
common in male dogs [47,51,52,54,61]. Separation anxiety is also suﬀered
more by male dogs in some reports [53,57,60,62]. Such tendencies are
unrelated to whether the animal has been castrated or spayed, which
indicates that they are not the product of adult sexual hormone level.
Hormones can aﬀect an animal activationally by their presence at the
moment in the blood stream or organizationally by their eﬀect on the fetal
brain during development. The lack of eﬀect of castration on most types of
aggression indicates that it may be the organizational eﬀect of androgens
that has lowered the threshold for aggression in intact and castrated male

350

Y. Takeuchi, K.A. Houpt / Vet Clin Small Anim 33 (2003) 345–363

dogs. In addition to the hormonal eﬀects, some genes related to behavioral
traits (eg, monoamine oxidase [20] and the serotonin 1c receptor [56]) are
located on human sex chromosomes; such genes, if present on sex chro-
mosomes of dogs, may well be connected to problem behavior.

One behavioral problem that does not occur frequently but may fail to be

recognized is canine narcolepsy. This closely resembles human narcolepsy
and cataplexy triggered by a positive emotion, and sudden episodes of
muscle weakness akin to rapid eye movement (REM) sleep-associated
atonia are produced. Narcolepsy in human beings is generally sporadic and
obviously hereditary, whereas in dogs, it has been conﬁrmed that it is
autosomal recessive in transmission with full penetrance. In the 1980s, much
research was carried out regarding genetic linkages by backcrossing such
animals to discover the pathogenic gene. In 1999, Lin et al [27] at Stanford
University reported that by using positional cloning, they had determined
that canine narcolepsy originated from the regional deletion of hypocretin
(orexin) receptor 2. Although at this point, we do not know if there is any
causative relation between the mutation of this gene and human narcolepsy
[25], the concentration of hypocretin in cerebrospinal ﬂuid is low among
human patients [28] and some narcoleptic dogs [9]. This kind of approach
using dogs might provide valuable information applicable to the human
clinical ﬁeld. Narcolepsy seems to be a neurodegenerative disease in which
there is atrophy of orexin cells.

There seem to be genetic diﬀerences in the frequency of some behavioral

problems within breeds. The most obvious one is wool sucking. Several
groups have noted that Siamese cats are most likely to be wool suckers
[29,30]. In fact, they do not suck but rather chew with their molars as they
would chew gristly prey. The cats seek out fabric and prefer loosely woven
wool over woolen suiting material. The latency to chew wool is shortened by
fasting, indicating a dietary basis. Furthermore, the medication clomipra-
mine, which is most helpful in treating canine obsessive-compulsive prob-
lems like tail chasing and lick granulomas, is not eﬀective in reducing the
frequency of wool chewing. If one particular item is available to the cat,
it is likely to eat a few holes in it and then to seek other material. This
indicates that although wool is attractive, it is not meeting the needs of the
cat; thus, the cat seeks something else. There is at least a component of
roughage craving, because exposure to the outdoors during the growing
season or access to a cat garden helps to reduce the frequency of wool
chewing. Other suggested treatments are tough meat and raw chicken wings.

Recent research trends and directions

Temperament is related to the style of emotional cognition, which varies

with the individual. When a dog reacts to stimuli (eg, meeting a person it does
not know), the visual input is sent to the amygdala of the limbic system. This is
the center for biologic value judgments and the source of emotional reactions.

351

Y. Takeuchi, K.A. Houpt / Vet Clin Small Anim 33 (2003) 345–363

Wary dogs (or those with a high level of anxiety) see people they do not know
as a threat; thus, they bark or growl to make the person go away, but if this
does not intimidate, they attempt to ﬂee. At this point, the reaction to
emergencies known as ‘‘ﬁght or ﬂight’’ comes into play; the heartbeat becomes
rapid, the hackles rise, and the blood pressure rises. At the same time, the
unpleasant emotions of fear and anger probably arise. Conversely, in the
same situation, a friendly dog wags its tail happily and greets the person.
The biologic value judgment mechanism in the brain tells the dog it should
approach the person, and pleasant emotions may arise. Thus, the same
stimulus may have diﬀerent behavioral reactions depending on the state of
the emotional cognition. This is temperament and is the basis of the individ-
uality of each animal. There seems to be little diﬀerence among animals in
the area of the brain that governs such emotional reactions, and there is a high
degree of similarity between human beings and dogs.

In recent years, research into the genetic base of temperament in human

beings has drawn a great deal of attention. Research related to the heredity
of personality in monozygotic twins has shown that individual personalities
are determined by hereditary and environmental factors fairly equally [31].
The earliest hereditary factor recognized as related to personality was
genetic polymorphism related to the dopamine D4 receptor [32]. The
dopamine receptor, which is linked to the emotions, is broadly divided into
ﬁve subtypes from D1 through D5; there exists a repetitive element in the
exon 3 domain of the D4 receptor, and this is known to aﬀect the rate of
transmission of information after the dopamine has bound to the receptor.
People with a high repetitive rate tend to be highly novelty seeking.
Extrapolating these results to an investigation of the same area in dogs,
it has been demonstrated clearly that Shiba breed dogs, with a strong
territorial instinct, have a longer repetitive element in the exon 3 domain
than Golden Retrievers (see Fig. 3) [33]. It has also been reported in hu-
man beings that serotonin-related genes are related to levels of anxiety,
aggression, and harm avoidance [1,22,58]. Because dopamine and serotonin
are so intimately involved in changes in mood and the emotions, these
neurotransmitters have been considered as targets of various antianxiety
drugs, antidepressant drugs, and stimulants [23,34].

Such research began within the human psychiatry ﬁeld, but because the

inﬂuence of social, cultural, and environmental factors is so great on the
development of brain functions in human beings and on the development of
the personality, it is not easy to determine causal relations, and this has
been an obstacle to further research. In contrast, when we try to analyze
laboratory mice and rats whose environment can be easily controlled, it
is diﬃcult to make a detailed personality analysis based on behav-
ioral parameters, because there seem to be few individual diﬀerences in
personality traits. Companion animals like cats and dogs, which have a
simpler system than human beings but whose rich individuality can be
described, would seem to be ideal as objects for this kind of research. In

352

Y. Takeuchi, K.A. Houpt / Vet Clin Small Anim 33 (2003) 345–363

our laboratory, we have begun to analyze the various kinds of genetic
polymorphisms related to neurotransmitters, centering on dogs.

Single nucleotide polymorphisms (SNPs) are a class of DNA poly-

morphisms that have proven useful as markers in mapping genes. An
example would be a series of the 3000 base pairs in which a single base pair
can diﬀer between individuals. If a gene that modiﬁed behavior in a fearful
dog is always inherited along with a particular marker, the gene and the
SNP marker are linked. The position on the chromosome of the gene for
fearfulness can be identiﬁed. We have already conﬁrmed several SNPs
accompanying amino acid substitution in the exon domain of some genes
[unpublished data]. Because the brain function that governs emotion is
preserved across a wide range of animals, it may be inferred that it is
possible to apply our research results to other animals. We hope that the
discovery of genes related to temperament is useful in improving the
human–companion animal bond by elimination of traits leading to
dominance, aggression, and separation anxiety and by selective breeding
for desirable characteristics. Knowledge of a dog’s genotype could lead to
screening programs suited to the individual (Table 2).

Research into genes governing behavior has only just begun, but,

already, there are more than 30 genes that are postulated to be related to
temperament, and this number continues to rise (Table 3). Four types of
genetic polymorphisms are so far known to be related to individuality in
human beings: restriction fragment length polymorphism (RFLP, one type
of SNP) [19], variable number of tandem repeat (VNTR) [6,35], micro-
satellite (2–4 tandem repeats) [36], and SNP [7]. Because the microsatellite
is a thousand times more susceptible to mutation across generations than
the SNP, it is not considered suitable as a marker for genes related to
temperament. It is probably not possible to explain animal temperament in
terms of a single gene, as reports on numerous such genes indicate. Rather,
small variations in the coding of multiple genes may constitute tempera-
ment diﬀerences. The structure of the human genome has been made public
by the Human Genome Project [21,37], and genetic research is now moving
toward the postgenome age. Recently developed microarray technology is
able to analyze the expression pattern of a large number of genes at the
same time from a small sample and can also distinguish the SNP [38,39].
Employing such technology, we may in the near future be able to conﬁrm
genetic characteristics of a companion animal from its genes originating in
blood (white blood cell), saliva (cheek swab), or coat hair (radix pili)
(Table 3, Fig. 4).

We also know that the father’s genes inﬂuence the daughter’s future

maternal behavior [40,41]. The Peg1 (Mest) and Peg3 genes, which have
been called the imprinting genes, are the monoallelic genes originating in
the father, and the daughter of a mutant mouse father in which these genes
were not expressed was not able to exhibit suﬃcient maternal behavior when
she became a mother. Because this daughter had few oxytocin cells in

353

Y. Takeuchi, K.A. Houpt / Vet Clin Small Anim 33 (2003) 345–363

the paraventricular nucleus of the hypothalamus, the father’s mutation ob-
viously inﬂuenced the oxytocin neurons in the daughter (Fig. 5).

Nature or nurture?

Even though hereditary factors are important as the base on which

personality is formed, it is impossible to ignore the importance of
environmental inﬂuence during the period after birth. Because the

Table 2
Deﬁnitions

Gene

The fundamental unit of heredity composed of sequences of

hundreds to thousands of nucleotide bases strung
together in giant DNA molecules, the chromosomes.
Genes have two functional units: the coding region, which
is transcribed into RNA and then translated into protein
based on the three-base genetic code, and the promoter
regions, which ﬂank the coding region and determine the
timing, amount, and anatomic localization of gene
expression and protein synthesis. In eukaryotic cells, the
RNA transcribed from the DNA sequence of a gene’s
coding region can be ‘‘cut and pasted’’ during RNA
splicing to produce variations in protein translated from
the underlying genetic code. Exons are the portion of the
gene that is included in the ﬁnal mRNA transcript of a
gene and thus are translated into part of a protein.
Introns are the segments of a gene located between the
exons that are transcribed into RNA but are cut out from
the mRNA and degraded with the nucleus, such that they
are never translated into protein. The function of this
‘‘junk DNA’’ is unknown, although it may contribute to
gene regulation.

Single nucleotide

polymorphism (SNP)

A single necleotide within the DNA sequence of hundreds

or thousands of bases that make up a gene may or may
not have an eﬀect on the protein coded for by that gene,
but the presence of a distinct SNP can be used as a marker
regardless of functional consequences.

Tandem repeats

Tandem repeat regions of the gene are particularly

susceptible to replication mutations, in which the gene is
extended in length through the inadvertent replication of
a short subsequence, or unit, of nucleotides, during
meiosis. Two forms of tandem repeats are commonly
analyzed: variable length tandem repeats (VLTRs) consist
of multiple repeats of DNA units from 9 to 80 base pairs,
and microsatellites are tandem repeats of units of one to
six base pairs. Because the speciﬁc number of repeats at
any VTLR or microsatellite site in the genome may
vary between individuals of subpopulations, the size of
speciﬁc VTLRs or microsatellites may serve as a genetic
marker.

354

Y. Takeuchi, K.A. Houpt / Vet Clin Small Anim 33 (2003) 345–363

impossibility of deciding whether nature or nurture is the deciding factor has
been shown by discussion over many years, both are indispensable in the
formation of personality [42].

From developmental and behavioral research conducted down to the

present, it can be predicted that socialization is closely related to the
development of the central nervous system, particularly the hypothalamus
and the limbic system, which govern emotional response. The period of
socialization, which is determined for each species, is when the dog or cat is
able, without any anxiety or fear, to regard other dogs or cats as well as
other animals, including human beings, as being on his list of acquaintances.

Because cats and dogs are born with immature eyes and ears, their

socialization period cannot begin until they are able to perceive the
environment. The canine socialization period lasts from around 3 to 12 weeks
[48]. The feline socialization period is 2 to 5 weeks [24]. A new understanding of
the inﬂuence of the early environment has been gained through experiments on
rodents. Characteristic maternal behavior of small rodents, such as mice
and rats, comprises lactation, grooming, stimulation for elimination, and
retrieving. It is clear that there is, in particular, a close connection between
grooming in the early stage and later behavioral patterns. On the one hand,
young rats that have received a lot of grooming from the mother tend to be
contented individuals when grown; conversely, young rats that do not receive
much grooming display an aggressive character and a high level of anxiety
later (Fig. 6) [43]. Further, maternal grooming also inﬂuences the development

Table 2 (continued)

Microsatellites

Short tandem repeat sequences of two to six base pairs

Restriction length fragment

polymorphism (RFLP)

DNA is digested with a restriction enzyme (an enzyme that

cleaves DNA at speciﬁc sites) into shorter fragments of
DNA. If there is variation in the DNA sequence between
the cleavage sites, the pattern of fragment lengths diﬀers
between individuals or subpopulations. These patterns
can thus be used to distinguish between the individuals.

Microarrays

Speciﬁc tandem repeats of SNPs can be detected one at a

time by amplifying DNA using polymerase chain
reaction. To detect multiple polymorphisms at the same
time, small quantities of known DNA sequences
(corresponding to polymorphisms of interest) can be
arrayed as small spots on a microscope slide. Thousands
of different polymorphisms can be spotted in a single
microarray. DNA to be analyzed from an individual can
be incubated on the microarray; the presence of a speciﬁc
polymorphism in the individual is detected when the
DNA sticks to the corresponding spot on the array.
Because polymorphisms within different genes and
chromosomes stick to the array in parallel and are
revealed simultaneously, microarrays promise high
throughput characterization and screening of patient
DNA.

355

Y. Takeuchi, K.A. Houpt / Vet Clin Small Anim 33 (2003) 345–363

Table 3
Polymorphic genes positively related to temperament or psychic disease (reports are collected from
MEDLINE database from 1966 to 2000)

Target gene and temperament

References

5-HT transporter

Anxiety (VNTR)

Science 274:1527–31, J Hum Genet 44:15–7

Harm avoidance (VNTR)

Psychiatr Genet 8:41–4, Arch Gen Psychiatry

55:936–40

Novelty seeking (VNTR)

Biol Psychiatry 43:908–12

Depression (VNTR)

Am J Med Genet 81:58–63, Psychol Med

29:1249–54

Prog Neuropsychopharmacol Biol Psychiatry 23:55–65

Alcoholism (VNTR)

Biol Psychiatry 43:908–12, Biol Psychiatry 45:647–51

Smoking (VNTR)

Cancer Epidemiol 8:831–3

5-HT1B receptor

Antisocial alcoholism (SNP)

Arch Gen Psychiatry 55:989–94

Antisocial alcoholism (VNTR)

Arch Gen Psychiatry 55:989–94

5-HT2A receptor

Schizophrenia (SNP)

Lancet 9011:1294–6, J Psychiatr Neurosci 24:141–6

Clozapine response (SNP)

Lancet 8970:281–2, Schizophr Res 32:93–9

Clozapine response in

schizophrenia (SNPa)

Neurosci Lett 217:177–8, Neuropsychopharmacology

19:123–32

Seasonal affective disorder

(SNP)

Mol Psychiatry 4:89–92, J Aﬀect Disord 53:203–10

Anorexia nervosa (SNP)

Neurosci Lett 277:134–6

Bulimia nervosa (SNP)

Neurosci Lett 277:134–6

5-HT2C receptor

Clozapine response (SNPa)

Neuroreport 7:169–72

Bipolar disorder (SNPa)

Neurosci Lett 212:65–7

Reward dependence (SNPa)

Am J Med Genet 74:65–2

5-HT6 receptor

Clozapine response in

schizophrenia (SNP)

Neuroreport 10:1231–3

Alzheimer’s disease (SNP)

Neurosci Lett 276:138–9

5-HT7 receptor

Alcoholism impulsivity (SNPa)

Psychiatr Res 77:139–45

Dopamine transporter

Attention-deﬁcit hyperactivity

(VNTR)

Am J Med Genet 56:993–8, Mol Psychiatry 2:311–3,

Genomics 52:289–97

Bipolar disorder (VNTR)

Am J Med Genet 67:533–40

Novelty seeking (VNTR)

Biol Psychiatry 42:1070–2, Health Psychol 18:7–13

Schizophrenia (VNTR)

Eur Neurol 38(Suppl 1):6–10

Alcoholism (VNTR, SNP)

Psychiatr Genet 7:87–91, Mol Psychiatry 4:552–7

Smoking (VNTR)

Health Psychol 18:7–13

Dopamine D2 receptor

Alcoholism (RFLP)

Psychiatr Genet 7:87–91, Biol Psychiatry 43:40–51

Novelty seeking (RFLP)

Am J Med Genet 81:257–67

Reward dependence (RFLP)

Am J Med Genet 81:257–67

Harm avoidance (RFLP)

Am J Med Genet 81:257–67

Schizophrenia (RFLP, SNP)

Mol Psychiatry 2:239–46, Schizophr Res 40:31–6

Dopamine D3 receptor

Schizophrenia (RFLP)

J Med Genet 29:858–60, Am J Med Genet 67:63–70,

Genomics 52:289–97

Bipolar disorder (RFLP)

Genomics 52:289–97

Novelty seeking (RFLP)

Am J Med Genet 81:192–4, Psychiatr Genet 9:17–21

356

Y. Takeuchi, K.A. Houpt / Vet Clin Small Anim 33 (2003) 345–363

Table 3 (continued)

Target gene and temperament

References

Dopamine D4 receptor

Novelty seeking (VNTR)

Nat Genet 12:78–80, Am J Med Genet 74:501–3, Am J

Med Genet 81:257–67

Schizophrenia (VNTR)

Eur Neurol 38(Suppl 1):6–10

Attention-deﬁcit hyperactivity

(VNTR)

Mol Psychiatry 3:38–41, Mol Psychiatry 3:419–26, Mol

Psychiatry 3:427–30, Am J Psychiatry 156:768–70

Bipolar disorder (VNTR)

Am J Med Genet 88:486–91

Harm avoidance (VNTR)

Am J Med Genet 88:634–41

Dopamine D5 receptor

Autism (SNPa)

Am J Med Genet 81:172–8

GABA-A alpha 6 receptor

Alcoholism (SNPa)

Biol Psychiatry 45:647–51

Cholecystokinin receptor

Alcoholism (SNP)

Alcohol Clin Exp Res 22(Suppl 3):93S–6

Catechol-o-methyltransferase

(COMT)
Schizophrenia (SNPa)

Psychiatr Genet 6:131–3, Psychiatr Res 69:71–7, Neurosci

Lett 243:109–12

Parkinson’s disease (SNPa)

Neurosci Lett 221:202–4, J Neural Transm 104:1313–7

Obsessive-compulsive disorder

(SNPa)

Proc Natl Acad Sci USA 94:4572–5, Psychiatr Genet

7:97–101

Bipolar disorder (SNPa)

Psychiatr Genet 7:97–101, Pharmacogenetics 7:349–53, Mol
Psychiatry 3:342–5, Mol Psychiatry 3:346–9

Attention-deﬁcit hyperactivity

(SNPa)

Am J Med Genet 88:497–502

Dopamine beta-hydroxylase

Drug-treated schizophrenia

(microsatellite)

Biol Psychiatry 41:762–7

Schizophrenia (RFLP)

Schizophr Res 22:77–80

Monoamine oxidase A

Impulsive aggression (SNP)

Science 5133:578–80

Bipolar disorder (microsatellite)

Am J Hum Genet 54:1122–4

Panic disorder (VNTR)

Hum Mol Genet 8:621–4

Antisocial alcoholism

(VNTR, RFLP)

Psychiatr Res 86:67–72, Genomics 55:290–5

Monoamine oxydase B

Parkinson’s disease

(microsatellite, SNP)

Mov Disord 14:219–24, Am J Med Genet 74:154–6

Tryptophan hydroxylase

Impulsive aggression (SNP)

Am J Med Genet 81:13–7

Bipolar disorder (SNP)

Arch Gen Psychiatry 55:33–7

Suicidal behavior, impulsivity

(SNP)

Genomics 52:289–97, Arch Gen Psychiatry 55:593–602

Aggression, anger-related traits

(SNP)

Biol Psychiatry 45:603–14

Tyrosine hydroxylase

Bipolar disorder (RFLP)

Am J Med Genet 74:289–95, Am J Med Genet 81:127–30

Schizophrenia (VNTR)

Eur Arch Psychiatr Clin Neurosci 248:61–3

Deliberation, dutifulness (VNTR)

Psychiatr Res 95:1–8

Polymorphic types are shown in parentheses.
Abbreviations

: GABA, gamma aminobutyric acid; RFLP, restriction fragment length poly-

morphism; SNP, single nucleotide polymorphism; SNPa, SNP accompanying amino acid sub-
stitution; VNTR, variable number of tandem repeats; microsatellite, two to four tandem repeats.

357

Y. Takeuchi, K.A. Houpt / Vet Clin Small Anim 33 (2003) 345–363

Fig. 4. Development of molecular biology may allow us to solve the mystery of individuality
based on genomic polymorphism.

Fig. 5. Paternal imprinting. A mutation for poor maternal behavior is expressed (activated)
only when inherited from the father (1). The daughter (3) of a male carrying the mutation (1)
has poor maternal behavior, but her daughter (5) has normal maternal behavior even though
she is a carrier. Her son (6) carries the mutation; his daughter (8) will have poor maternal
behavior, and her son (9) will be a carrier. Black symbols indicate poor maternal behavior. Gray
symbols indicate carriers. White symbols are normal. Squares indicate males, and circles
indicate females.

358

Y. Takeuchi, K.A. Houpt / Vet Clin Small Anim 33 (2003) 345–363

of learning abilities in young rats [26]. Experiments with foster mothers have
shown that these tendencies are not hereditary [44]. These results indicate
clearly that the degree of care the mother gives its young determines the later
behavioral pattern of the grown rat. Studies of the mechanism by which young
rats show high levels of anxiety focus on the role of the corticotropin-releasing
factor (CRF), a neuropeptide that governs the endocrine system (ie, the
hypothalamus-pituitary-adrenal cortex axis) [45]. Investigations of individu-
als that did not receive much grooming as babies from their mothers show an
increase of CRF mRNA in the paraventricular nucleus of the hypothalamus
and an increase of the CRF receptor in the amygdala. By contrast, individuals
that have received a lot of grooming exhibit an increase in the hippocampus of
the glucocorticoid receptor, giving negative feedback to the CRF such that the
CRF is easily controlled. This shows that early environment is an important
inﬂuence on behavioral patterns and that the care a mother gives her young,
whether good or bad, brings permanent change to the neural mechanism of the
brain. This mechanism also holds true for pigs to some extent [46], and it may
be possible to extrapolate this to companion animals like cats and dogs as well.

It is said that when a certain breed becomes popular, behavioral

problems arise after some time. One factor is hereditary in that unsuitable
breeding has occurred in the interests of consumerism. We must also con-
sider the possibility that the animals, being popular consumer items, are re-
moved from their mother and siblings at too early an age and are injured
psychologically by being placed alone in the window of a pet shop or by

Fig. 6. The level of maternal care aﬀects the later behavioral pattern of the grown rat. Young
rats that have received a lot of grooming from the mother tend to be contented individuals,
whereas young rats that have not received much grooming display an aggressive character and a
high level of anxiety.

359

Y. Takeuchi, K.A. Houpt / Vet Clin Small Anim 33 (2003) 345–363

being transported over a long distance during the important time when the
basis of emotions is being formed [50].

Summary

The inﬂuence of hereditary and environmental factors is indispensable as

the foundation on which the temperament of an animal is formed. Genetic
research on animal temperament has experienced a turning point in recent
years as a result of the development of molecular biology. In the near future,
it may be possible to explain the formation process of animal temperament
as the two ﬁelds share their research. We look forward to applying these
research results to the development of new genetic treatment methods for
problem behavior and training programs suited to the individual.

Acknowledgment

The authors thank Prof Yuji Mori (Veterinary Ethology, The University

of Tokyo) for his valuable comments and generous permission to use his
drawings, T.A. Houpt for Table 2, and E.J. Pollock for Figs. 1, 2, and 5.

References

[1] Lesch KP, Merschdorf U. Impulsivity, aggression, and serotonin: a molecular psycho-

biological perspective. Behav Sci Law 2000;18:581–604.

[2] Thorne FC. The inheritance of shyness in dogs. J Genet Psychol 1944;65:275–9.
[3] Dykman RA, Murphree OD, Peters JE. Like begets like: behavioral tests, classical

autonomic and motor conditioning, and operant conditioning in two strains of pointer
dogs. Ann NY Acad Sci 1969;159:976–1007.

[4] Scott JP, Fuller JL. Genetics and the social behavior of the dog. The classic study. Chicago:

The University of Chicago Press; 1965.

[5] Freedman DG. Constitutional and environmental interactions in rearing of four breeds of

dogs. Science 1958;127:585–6.

[6] Nakamura Y, Koyama K, Matsushima M. VNTR (variable number of tandem repeat) se-

quences as transcriptional, translational, or functional regulators. J Hum Genet 1998;43:
149–152.

[7] Comings DE, Johnson JP, Gonzalez NS, et al. Association between the adrenergic alpha

2A receptor gene (ADRA2A) and measures of irritability, hostility, impulsivity and
memory in normal subjects. Psychiatr Genet 2000;10:39–42.

[8] Mackenzie SA, Oltenacu EAB, Leighton E. Heritability estimate for temperament scores

in German Shepherd Dogs and its genetic correlation with hip dysplasia. Behav Gen 1985;
15:475–82.

[9] Ripley B, Fujiki N, Okura M, et al. Hypocretin levels in sporadic and familial cases of

canine narcolepsy. Neurobiol Dis 2001;8:525–34.

[10] Adamec RE, Stark-Adamec C, Livingston KE. The expression of an early developmentally

emergent defensive bias in the adult domestic cat (Felis catus) in non-predatory situations.
Appl Anim Ethol 1983;10:89–108.

[11] Turner DC, Feaver J, Mendl M, Bateson P. Variation in domestic cat behaviour towards

humans: a paternal effect. Anim Behav 1986;34:1890–2.

360

Y. Takeuchi, K.A. Houpt / Vet Clin Small Anim 33 (2003) 345–363

[12] McCune S. The impact of paternity and early socialisation on the development of cats’

behaviour to people and novel objects. Appl Anim Behav Sci 1995;45:109–24.

[13] Reisner IR, Houpt KA, Quimby KW. Friendliness to humans and defensive aggression in

cats: the inﬂuence of handling and paternity. Physiol Behav 1994;55:1119–24.

[14] Shin T, Kraemer D, Pryor J, et al. A cat cloned by nuclear transplantation. Nature

2002;415:859.

[15] Hart BL, Hart LA. The perfect puppy. New York: WH Freeman and Company; 1988.
[16] Van Der Velden NA, De Weerdt CJ, Brooymans-Schallenberg JHC, et al. An abnormal

behavioural trait in Bernese Mountain Dogs (Berner sennenhund). Tijdschr Diergeneesk
1976;101:403–7.

[17] Podberscek AL, Serpell J. The English Cocker Spaniel: preliminary ﬁndings on aggressive

behaviour. Appl Anim Behav Sci 1996;47:75–89.

[18] Houpt KA, Willis MB. Genetics of behaviour. In: Ruvinsky A, Sampson J, editors. The

genetics of the dog. New York: CABI Publishing; 2001. 371–400.

[19] Blum K, Noble EP, Sheridan PJ, et al. Genetic predisposition in alcoholism: association

of the D2 dopamine receptor TaqI B1 RFLP with severe alcoholics. Alcohol 1993;10:
59–67.

[20] Grimsby J, Chen K, Wang LJ, et al. Human monoamine oxidase A and B genes exhibit

identical exon-intron organization. Proc Natl Acad Sci USA 1991;88:3637–41.

[21] Lander ES, Linton LM, Birren B, et al. Initial sequencing and analysis of the human

genome. Nature 2001;409:860–921.

[22] Lesch KP, Bengel D, Heils A, et al. Association of anxiety-related traits with a poly-

morphism in the serotonin transporter gene regulatory region. Science 1996;274:1527–31.

[23] Petty F, Davis LL, Kabel D, et al. Serotonin dysfunction disorders: a behavioral

neurochemistry perspective. J Clin Psychiatry 1996;57:11–6.

[24] Martin P, Bateson P. Behavioural development in the cat. In: Turner DC, Bateson P,

editors. The domestic cat. The biology of its behaviour. Cambridge: University of
Cambridge Press; 1988, p. 9–22.

[25] Olafsdottir BR, Rye DB, Scammell TE, et al. Polymorphisms in hypocretin

/orexin pathway

genes and narcolepsy. Neurology 2001;57:1896–9.

[26] Liu D, Diorio J, Day JC, et al. Maternal care, hippocampal synaptogenesis and cognitive

development in rats. Nat Neurosci 2000;3:799–806.

[27] Lin L, Faraco J, Li R, et al. The sleep disorder canine narcolepsy is caused by a mutation in

the hypocretin (orexin) receptor 2 gene. Cell 1999;98:365–76.

[28] Nishino S, Ripley B, Overeem S, et al. Low cerebrospinal ﬂuid hypocretin (orexin) and

altered energy homeostasis in human narcolepsy. Ann Neurol 2001;50:381–8.

[29] Bradshaw JWS. The behaviour of the domestic cat. Wallingford, UK: CAB International;

1992.

[30] Houpt KA. Feeding and drinking behavior problems. Vet Clin North Am Small Anim

Pract 1991;21:281–98.

[31] Plomin R. The role of inheritance in behavior. Science 1990;248:183–8.
[32] Benjamin J, Li L, Patterson C, et al. Population and familial association between the D4

dopamine receptor gene and measures of novelty seeking. Nat Genet 1996;12:1–84.

[33] Niimi Y, Inoue-Murayama M, Murayama Y, et al. Allelic variation of the D4 dopamine

receptor polymorphic region in two dog breeds, Golden Retriever and Shiba. Veterinary
Medicine 1999;61:1281–6.

[34] Comings DE, Blum K. Reward deﬁciency syndrome: genetic aspects of behavioral

disorders. Prog Brain Res 2000;126:325–41.

[35] Manuck SB, Flory JD, Ferrell RE, et al. A regulatory polymorphism of the monoamine

oxidase-A gene may be associated with variability in aggression, impulsivity, and central
nervous system serotonergic responsivity. Psychiatry Res 2000;95:9–23.

[36] Jonsson EG, von Gertten C, Gustavsson JP, et al. Androgen receptor trinucleotide repeat

polymorphism and personality traits. Psychiatr Genet 2001;11:19–23.

361

Y. Takeuchi, K.A. Houpt / Vet Clin Small Anim 33 (2003) 345–363

[37] Venter JC, Adams MD, Myers EW, et al. The sequence of the human genome. Science

2001;291:1304–51.

[38] Blohm DH, Guiseppi-Elie A. New developments in microarray technology. Curr Opin

Biotechnol 2001;12:41–7.

[39] Tillib SV, Mirzabekov AD. Advances in the analysis of DNA sequence variations using

oligonucleotide microchip technology. Curr Opin Biotechnol 2001;12:53–8.

[40] Lefebvre L, Viville S, Barton SC, et al. Abnormal maternal behaviour and growth

retardation associated with loss of the imprinted gene Mest. Nat Genet 1998;20:163–9.

[41] Li L, Keverne EB, Aparicio SA, et al. Regulation of maternal behavior and offspring

growth by paternally expressed Peg3. Science 1999;284:330–3.

[42] Meaney MJ. Nature, nurture, and the disunity of knowledge. Ann NY Acad Sci 2001;

935:50–61.

[43] Caldji C, Tannenbaum B, Sharma S, et al. Maternal care during infancy regulates the

development of neural systems mediating the expression of fearfulness in the rat. Proc Natl
Acad Sci USA 1998;95:5335–40.

[44] Francis D, Diorio J, Liu D, et al. Nongenomic transmission across generations of maternal

behavior and stress responses in the rat. Science 1999;286:1155–8.

[45] Liu D, Diorio J, Tannenbaum B, et al. Maternal care, hippocampal glucocorticoid receptors,

and hypothalamic-pituitary-adrenal responses to stress. Science 1997;277:1659–62.

[46] Weaver SA, Aherne FX, Meaney MJ, et al. Neonatal handling permanently alters

hypothalamic-pituitary- adrenal axis function, behaviour, and body weight in boars.
J Endocrinol 2000;164:349–59.

[47] Borchelt PL. Aggressive behavior of dogs kept as companion animals: classiﬁcation and

inﬂuence of sex, reproductive status and breed. Appl Anim Ethol 1983;10:45–61.

[48] Freedman DG, King JA, Elliot E. Critical period in the social development of dogs.

Science 1961;133:1016–7.

[49] Goddard ME, Beilharz RG. Genetic and environmental factors affecting the suitability of

dogs as guide dogs for the blind. Theor Appl Genet 1982;62:97–102.

[50] Jagoe A, Serpel J. Owner characteristics and interactions and the prevalence of canine

behavior problems. Appl Anim Behav Sci 1996;47:31–42.

[51] Landsberg GM. The distribution of canine behaviour cases at three behaviour referral

practices. Veterinary Medicine 1991;86:1011–8.

[52] Line S, Voith VL. Dominance aggression of dogs towards people: behavior proﬁle and

response to treatment. Appl Anim Behav Sci 1986;16:77–83.

[53] Lund JD, Jorgensen MC. Separation anxiety in pet dogs behaviour patterns and time

course of activity. In: Proceedings of the First Conference on Veterinary Behavioral
Medicine. Potters Bar, UK: Universities Federation for Animal Welfare; 1997. p. 133–42.

[54] Lund JD, Agger JF, Vestergaard KS. Reported behaviour problems in pet dogs in

Denmark: age distribution and inﬂuence of breed and gender. Prev Vet Med 1996;28:33–48.

[55] Mackenzie SA, Oltenacu EAB, Houpt KA. Canine behavioral genetics—a review. Appl

Anim Behav Sci 1986;15:365–93.

[56] Milatovich A, Hsieh CL, Bonaminio G, et al. Serotonin receptor 1c gene assigned to X

chromosome in human (band q24) and mouse (bands D-F4). Hum Mol Genet 1992;1:
681–684.

[57] Podberscek AL, Hsu Y, Serpell JA. Evaluation of clomipramine as an adjunct to

behavioural therapy in the treatment of separation-related problems in dogs. Vet Rec
1999;145:365–9.

[58] Ricketts MH, Hamer RM, Sage JI, et al. Association of a serotonin transporter gene

promoter polymorphism with harm avoidance behaviour in an elderly population.
Psychiatr Genet 1998;8:41–4.

[59] Scott JP, Bielfelt SW. Analysis of the puppy testing program. In: Pfaffenberger CJ, Scott

JP, Fuller JL, Ginsburg BE, Bielfelt SW, editors. Guide dogs for the blind: their selection,
development and training. New York: Elsevier; 1976. p. 39–75.

362

Y. Takeuchi, K.A. Houpt / Vet Clin Small Anim 33 (2003) 345–363

[60] Takeuchi Y, Houpt KA, Scarlett JM. Evaluation of treatments for separation anxiety in

dogs. JAVMA 2000;217:342–5.

[61] Takeuchi Y, Ogata N, Houpt KA, et al. Differences in background and outcome of three

behavior problems of dogs. Appl Anim Behav Sci 2001;70:297–308.

[62] Voith VL, Borchelt PL. Separation anxiety in dogs. Compend Contin Educ Pract Vet

1985;7:42–52.

[63] Willis MB. Genetic aspects of dog behaviour with particular reference to working ability.

In: Serpell J, editor. The domestic dog. Its evolution, behaviour, and interactions with
people. New York: Cambridge University Press; 1995. p. 51–64.

[64] Wilsson E. The social interaction between mother and offspring during weaning in German

Shepherd Dogs: individual differences between mothers and their effects on offspring. Appl
Anim Behav Sci 1984

/1985;13:101–12.

363

Y. Takeuchi, K.A. Houpt / Vet Clin Small Anim 33 (2003) 345–363

Pharmacologic management

in veterinary behavioral medicine

Barbara Sherman Simpson, MS, PhD, DVM

Mark G. Papich, DVM, MS

The Veterinary Behavior Clinic, 6045 US Highway 1 North, Southern Pines,

NC 28387–8614, USA

Department of Molecular Biological Sciences, College of Veterinary Medicine,

North Carolina State University, Raleigh, NC, USA

In the past decade, knowledge of pharmacologic management in

veterinary behavioral medicine has paralleled that in human behavioral
medicine. Medications, combined with behavioral therapy, have been used
to manage diﬃcult behavioral problems. Pharmacologically, the focus on
behavioral drugs has been to understand how they can be used to reduce
anxiety, decrease arousal, attenuate repetitive compulsive behaviors, and
help to manage organic states and other disorders [1]. These applications
have led to increased familiarity among veterinarians with speciﬁc drugs and
drug classes, the use of well-known drugs in combination, and the use of less
familiar drugs to achieve improved management of difﬁcult problems. These
strategies require knowledge of how psychotropic drugs act and interact.

Because there are so few drugs currently registered for behavioral

disorders in veterinary medicine, veterinarians have relied on human-label
drugs to treat these problems. Without an approved claim, it has fallen on
veterinary behaviorists, pharmacologists, and other specialists to examine
the relevant data and conduct studies to examine the appropriate published
record on these medications to predict clinical use and response. One of the
most diﬃcult challenges in evaluating the studies published in human
medicine or from experimental animal studies is to interpret the data in light
of the various species diﬀerences that exist in drug metabolism, receptor
sensitivity, and susceptibility to toxicosis. This article is an attempt to con-
dense some of that information and present it in a manner that is useful to

Vet Clin Small Anim

33 (2003) 365–404

* Corresponding author.
E-mail address:

simpson.barbara@earthlink.net (B.S. Simpson).

0195-5616

/03/$ - see front matter

PII: S 0 1 9 5 - 5 6 1 6 ( 0 2 ) 0 0 1 3 0 - 4

veterinarians interested in the pharmacologic management of behavioral
disorders.

Pharmacokinetics

When there is a relation of the plasma or serum drug concentrations to

clinical eﬀect, pharmacokinetics can be helpful to predict pharmacologic
response. This becomes particularly important when we have few controlled
clinical trials of drugs in animals but do have comparative pharmacokinetic
data. Pharmacokinetics is the science that describes the drug’s absorption,
distribution, metabolism, and elimination [2].

Absorption

Bioavailability of a drug depends on both the extent and rate of drug

absorption. Because most behavioral drugs discussed in this article are
administered orally, absorption becomes a critical pharmacokinetic para-
meter. Drug absorption is determined by examining the relative con-
centrations in plasma or serum, because the availability of a drug to the
central nervous system (CNS) cannot be easily measured.

Drugs given orally can be absorbed quickly and avoid signiﬁcant

metabolism, be poorly absorbed because of unfavorable dissolution or
solubility, or be absorbed from the gastrointestinal tract (GIT) and then
undergo ﬁrst-pass metabolism. First-pass metabolism is the process of
intestinal or hepatic metabolism before reaching the systemic circulation.

Many of the behavioral drugs to be discussed in this article are weak

bases, for example, the substituted amines, which are the tricyclic anti-
depressants (TCAs), and other centrally acting drugs. Weak bases are gen-
erally more lipid-soluble in an alkaline environment. The inﬂuence of the
animal’s gastrointestinal physiology on drug absorption and disposition
is well reviewed. Although monogastric animals, such as dogs and cats, se-
crete stomach acid intermittently, the stomach maintains a relatively acidic
environment. The proximal small intestine is considerably more alkaline.
The behavior-modifying drugs, most being weak bases, are generally well
dissolved in the acidic stomach, where hydrophilic solubility is expected to
be high. When they pass into the intestine, they become more lipophilic in
the alkaline environment and are generally well absorbed. The rapid
intestinal metabolism also renders these drugs susceptible to the ﬁrst-pass
metabolic eﬀects of intestinal and hepatic enzymes, however, which reduces
the overall systemic availability.

Obviously, other drugs and factors can inﬂuence these processes. For

example, antisecretory drugs can reduce stomach acidity, food can decrease
stomach emptying, and other drugs can aﬀect intestinal and hepatic drug-
metabolizing enzymes.

366

B.S. Simpson, M.G. Papich

/ Vet Clin Small Anim 33 (2003) 365–404

Metabolism

Drug metabolism is the process whereby drugs are metabolized to active

and inactive metabolites or an inactive drug is metabolized to an active drug
(if administered as a prodrug). For the drugs discussed in this article, the
metabolic fate is determined primarily by hepatic metabolism. Few drugs
used in behavioral therapy are aﬀected much by renal clearance, although
the kidneys may be the ultimate route of elimination for conjugated water-
soluble metabolites.

Many behavioral drugs are substrates for or aﬀect cytochrome P450

(CYP) enzymes, which are microsomal drug-metabolizing enzymes located
primarily in the liver and GIT [2,3]. These enzymes are nonneuronal sites
of neurotransmitter action with potential for important pharmaco-
kinetic drug-drug interactions. CYP enzymes are designated by family and
subfamily, and isoforms are designated by a number and letter sequence. In
human beings, the important CYP enzymes include CYP1A2, CYP2C9-10,
CYP2C19, CYP2D6, and CYP 3A3

/4. The enzymes CYP3A3/4 and

CYP2D6 are responsible in human beings for 50% and 30%, respectively,
of known oxidative drug metabolism. Because these enzymes can be both
induced and inhibited by certain drugs, interactions are possible. Drugs
classiﬁed as antidepressants are inhibitors; therefore, concentrations of
other drugs metabolized by the same CYP enzymes increase. For example,
in human beings, ﬂuoxetine and paroxetine inhibit CYP2D6, which is
important in the oxidative metabolism of the TCAs. When ﬂuoxetine or
paroxetine is used in combination with a TCA, a signiﬁcant increase of the
TCA plasma concentration results, potentially causing toxicity unless the
TCA dose is reduced [2].

One of the problems in veterinary medicine is that the enzymes, and

subsequently their substrates and inhibitors, are not as well characterized as
in human medicine [4]. The enzyme responsible for the greatest proportion
of metabolism in human beings is CYP3A4 oxidase. There are only low
levels of CYP3A4 in dogs and cats, but other enzymes play a larger role (eg,
CYP3A12) [5]. Other enzymes present in dogs are the 1A, 2C, 3A, and 2D
families and subfamilies [4,5]. There are large interspecies differences in
the P450-mediated metabolism in dog and cat microsomes compared with
human microsomes. The variation is in the metabolic activity as well as in
the effect of speciﬁc inhibitors on P450 enzyme activity [4]. Information on
the inhibitory activity of drugs on various enzyme systems in human beings
may not be extrapolated to dogs and cats [5].

The other step in drug metabolism is a biosynthetic reaction called

conjugation. Drug metabolic conjugation is the process whereby the drug
or metabolite is linked with endogenous compounds, such as amino acids,
glucuronic acid, sulfate, glutathione, or acetyl (acetate). These polar con-
jugates are more water soluble and more easily excreted than the parent
compound. The conjugated products are usually inactive, but there are

367

B.S. Simpson, M.G. Papich

/ Vet Clin Small Anim 33 (2003) 365–404

exceptions (eg, the 6-glucuronide metabolite of morphine is more active
than the parent drug).

Just as with the other metabolic reactions, there are tremendous species

diﬀerences in the conjugation reactions. Dogs lack the ability to acetylate
drugs like sulfonamides; cats have a deﬁcient ability to form glucuronide
metabolites with drugs like salicylate and phenols (eg, acetaminophen
metabolites).

Clearance

The rate of hepatic metabolism is measured by hepatic clearance.

Clearance is one of the determinants of elimination half-life (T

), the time

needed to clear 50% of a drug from the plasma [2]. Drugs with short
elimination half-lives must be given more frequently to maintain a
consistent plasma concentration. With repeated dosing, these drugs also
achieve steady state more quickly. Usually, after ﬁve half-lives, a drug
reaches plasma steady state, as long as the dosing schedule or other
metabolic processes remain constant. Once plasma steady state is achieved,
drug concentration in other tissues, such as the brain, is at equilibrium. The
time to reach steady state is relevant to the use of behavioral drugs. Some
drugs, such as diazepam in the dog, have short half-lives (T

less than

1 hour) [6]. Unless administered more frequently than once every ﬁve half-
lives, these drugs never reach a steady state. Conversely, drugs with long
half-lives accumulate with chronic dosing and attain steady state in ap-
proximately ﬁve half-lives. If the half-life is 24 hours or longer, several
days may be necessary before the drug accumulates to a level high enough to
produce a consistent clinical response. This may be one reason why some
antidepressant drugs do not have immediate effects when administered
chronically in animals.

Distribution

The pharmacokinetic term that describes drug distribution is the volume

of distribution, often termed the apparent volume of distribution (VD)
because its physiologic relevance is only apparent. The word does not
attempt to describe a physiologic process; it is only a proportionality
constant relating drug concentration in the plasma or serum to drug dose, as
in the formula VD = [dose]

/[plasma concentration].

The physiologic distribution of drugs is determined by their lipid

solubility and protein binding. The higher the lipophilicity, the greater is
the ability to distribute across biologic lipid membranes as long as plasma
protein binding is not so high as to limit the diﬀusion. Because most behav-
ioral drugs are weak bases, protein binding is expected to be low for these
drugs. This is only assumed, however, because there are few or no published
data documenting the true plasma protein binding for these drugs in dogs
and cats. Tissue protein binding or intracellular trapping of drugs can in-
crease the distribution of drugs from the plasma to tissue compartment.

368

B.S. Simpson, M.G. Papich

/ Vet Clin Small Anim 33 (2003) 365–404

Relevant to the behavioral drugs is the distribution across the blood–

brain barrier (BBB). The BBB consists of unfenestrated capillaries with tight
junctions that prevent large or poorly lipophilic molecules from passing
from the blood to the brain [7,8]. There is also a blood–cerebrospinal ﬂuid
(CSF) barrier, but it makes up a relatively smaller component of the
distribution of drugs to the CNS. The BBB also is composed of membrane
pumps called p-glycoprotein. These are membrane proteins that serve
as pumps to move drugs out of the brain or CSF into the blood for
elimination. Some drugs are good substrates for p-glycoprotein, and other
drugs serve as inhibitors of these pumps [7].

Most of the behavioral drugs are weak bases and unionized and lipo-

philic at physiologic pH. Some may be trapped in the brain un-ionized or
CSF, because it is relatively more acidic than the plasma owing to pH
partitioning. For example, some of the substituted amines can attain higher
concentrations in the CNS than in other tissues in the body.

Neurotransmitter receptors and second messengers

Behavioral drugs act either as stimulators (agonists) or blockers

(antagonists) of neurotransmitter receptors or as inhibitors of regulatory
enzymes [9,10]. Drugs that mimic the action of naturally occurring neu-
rosignals affect the monoamine neurotransmitters serotonin (5-hydroxy-
tryptamine [5-HT]), norepinephrine (NE), and dopamine (DA) as well as
acetylcholine (ACh), glutamate, and gamma-aminobutyric acid (GABA)
receptors, among others. Each of these may have multiple receptor subtypes
with which they interact. Many other substances, such as circulating hor-
mones, pituitary peptides, opioid peptides, and neurokinins, can affect
behavior [10].

The purpose of chemical neurotransmission is to alter the function

of postsynaptic target neurons. This process, in turn, aﬀects gene expres-
sion [10]. The neurotransmitter released from the presynaptic neuron
is considered the ﬁrst messenger. It binds to its receptor, and the bound
neurotransmitter causes manufacture of a second messenger inside the cell
of the postsynaptic neuron. This second messenger, in turn, forms trans-
cription factors that, when activated, bind to regulatory regions of genes.
This process activates RNA polymerase, and the gene transcribes itself
into its mRNA, leading to translation of the corresponding protein, which
can modify behavior [10].

Norepinephrine

NE is derived from the amino acid tyrosine, which is transported from

the blood and into each noradrenergic neuron by means of an active
transport pump [10]. There, tyrosine is acted on by three enzymes, eventually
converting it to DA and then to NE, which is stored in vesicles. NE can be

369

B.S. Simpson, M.G. Papich

/ Vet Clin Small Anim 33 (2003) 365–404

destroyed by monoamine oxidase (MAO), located in mitochondria, and
catechol-O-methyl transferase (COMT), located outside the presynaptic
nerve terminal. There are three postsynaptic receptors for NE: b

, a

, and a

NE has little activity on b

receptors. The a

receptors are also found

presynaptically. Called autoreceptors, they regulate NE release via a negative
feedback system [10].

Most cell bodies for noradrenergic neurons are located in the locus

ceruleus area of the brain stem. This region determines whether attention is
focused on the external environment (eg, response to a threat) or to internal
signals (eg, pain). There are many speciﬁc noradrenergic pathways in the
brain controlling both psychologic and physiologic activities. For example,
projections from the locus ceruleus to the limbic cortex regulate emotions;
projections in cardiovascular centers may control blood pressure.

Dopamine

Like NE, DA is synthesized intraneuronally from the amino acid

tyrosine. DA neurons lack the third enzyme that leads to conversion to NE.
The same enzymes that destroy NE (MAO and COMT) also metabolize
DA. There are at least ﬁve DA receptor subtypes. Best known is the DA2
receptor, which is stimulated by dopaminergic agonists for the treatment of
Parkinson’s disease in human beings and blocked by DA antagonist
antipsychotics. It is not clear to what extent DA

, DA

, and DA

receptors

contribute to the behavioral effects of antipsychotic drugs. When DA re-
ceptors are blocked, as with an antipsychotic drug, ACh activity increases.
This is because DA normally suppresses ACh activity. An increase in ACh
activity can lead to extrapyramidal symptoms as discussed below.

Serotonin

Serotonin is also called 5-HT. Abnormalities in central serotonin

function have been hypothesized to underlie disturbances in mood, anxiety,
satiety, cognition, aggression, and sexual drive [11]. Drugs that enhance se-
rotonin are among the most effective modulators of behavior [12]. For
synthesis of serotonin, the amino acid tryptophan is transported into the
brain from plasma. Two enzymes are involved in the conversion of
tryptophan to 5-HT. Analogous enzymes, transport pumps, and receptors
exist in the 5-HT neuron. Classiﬁcation of 5-HT receptors was reviewed
thoroughly by Hoyer and colleagues [13]. There are two key presynaptic
receptors (5-HT

and 5-HT

) and at least six postsynaptic receptors (5-

, 5-HT

, and 5HT

). As with NE and

DA, presynaptic receptors act as autoreceptors that detect high concen-
trations of 5-HT and inhibit further 5-HT release and slow 5-HT neuronal
impulse ﬂow. Postsynaptic 5-HT receptors regulate 5-HT release from the
presynaptic nerve ending. The 5-HT

, 5-HT

, and 5-HT

receptors are

370

B.S. Simpson, M.G. Papich

/ Vet Clin Small Anim 33 (2003) 365–404

implicated in several serotonin pathways in the CNS [13]. Although some 5-
HT4 receptors are located in the CNS, their action is primarily localized to
the GIT. Serotonergic nuclei are localized to the raphe nucleus of the brain
stem [11]. This area has projections to the frontal cortex, which may regulate
mood, the basal ganglia, which may control movement and compulsive
behaviors, and the limbic area, which may be involved in anxiety and panic.

There is evidence that the serotonin system may exert ‘‘tonic inhibition’’

on the central dopaminergic system [11]. This may explain the unexpected
occurrence of extrapyramidal side effects during therapy with a selective
serotonin reuptake inhibitor (SSRI) [11].

Acetylcholine

ACh is formed in cholinergic neurons from two precursors: choline,

which is derived from dietary sources, and acetyl coenzyme A, which is
synthesized in the neuron. There are two major types of cholinergic
receptors: nicotinic and muscarinic. Each of these is further divided into
numerous receptors subtypes. The muscarinic receptors were originally
classiﬁed as M

(ganglionic) or M

(effector). With advanced technology,

scientists are able to classify ﬁve muscarinic receptor subtypes: M

through

. M

receptors are found on ganglia. M

receptors and M

receptors are

found on smooth muscle and secretory organs, such as those of the GIT. All
ﬁve subtypes are found in the CNS, however. A nonspeciﬁc blocker of
muscarinic receptors is atropine. One of the side effects of some behavior-
modifying drugs (eg, TCAs) is blockade of muscarinic receptors producing
cardiovascular and gastrointestinal side effects.

Gamma-aminobutyric acid

GABA is the major inhibitory neurotransmitter in the CNS and is

localized particularly in the cortex and thalamus [14]. GABA is synthesized
from the amino acid precursor glutamate. Glutamate participates in mul-
tiple metabolic functions. The GABA neuron has a presynaptic trans-
porter similar to those of NE, DA, and 5-HT. There are two subtypes of
GABA: GABA

and GABA

. GABA

subtype receptors are allosteri-

cally modulated by benzodiazepine (BZD) receptors and others.

It is now known that many neurons respond to more than one

neurotransmitter, a process called cotransmission [10]. This may explain
why multiple drugs in combination may be particularly effective and why
some beneﬁcial drugs act on more than one neurotransmitter. At this time,
there is no rational treatment approach based on cotransmission; however, a
strategic multiple drug program may enhance treatment success in the future.

Major drug classes

Behavioral drugs are classiﬁed historically according to their ﬁrst human

clinical application (eg, the antidepressant category) in spite of the fact that

371

B.S. Simpson, M.G. Papich

/ Vet Clin Small Anim 33 (2003) 365–404

the drugs are now commonly used in human and veterinary medicine for
a wider range of behavioral disorders. Traditionally, drugs are further
classiﬁed according to their chemical structure and neurochemical activity.
TCAs (referring to a common chemical structure) and SSRIs (referring to
neurochemical activity) are examples of drugs classiﬁed by their chemical
structure and mechanism of action, respectively. Drugs in the same category
share many characteristics, including mechanism of action and common side
eﬀects. Because many reference sources utilize this traditional and logical
categorization, it is retained in the discussion that follows (Tables 1 and 2).

Antipsychotics

The antipsychotics include a number of structurally dissimilar drugs used

in human beings to treat psychosis, which is typiﬁed by conditions like
schizophrenia, aﬀective disorder, and psychoses associated with organic
mental disorders [144]. Because the conventional antipsychotics produce
neurologic side effects, they are sometimes called neuroleptics. This word is
generally not applied to newer atypical antipsychotics for which neurologic
side effects are less likely. Antipsychotics block central DA receptors,
particularly of the subtype D

. Antipsychotics produce ataraxia, a state

of relative indifference to external stimuli [9]. Most antipsychotics are
metabolized by the CYP enzymes belonging to families 2 and 3; therefore, it
is possible that drug-drug interactions described for people also would be a
concern for animals [144].

Except for acepromazine, antipsychotics are not commonly used in

modern veterinary behavioral medicine for a number of reasons. First, small
animals are rarely diagnosed with ‘‘psychosis.’’ Second, when animals
are given traditional antipsychotics at relatively high doses, they develop
catalepsy, a syndrome with immobility, increased muscle tone, and ab-
normal postures [144]. Most veterinarians are familiar with the effects of
acepromazine on dogs and cats. Third, spontaneous motor activity caused
by DA receptor blockade in the striatum and inactivation of DA neurons in
the substantial nigra may result from the administration of phenothiazines
to animals [144]. Finally, other important side effects of antipsychotic drugs,
summarized below, can be unacceptable.

Antipsychotics can cause extrapyramidal signs (EPSs) as a result of their

eﬀect of inhibiting DA. EPSs are most likely in high-potency antipsychotics
like haloperidol. EPSs documented in human beings include pseudoparkin-
sonism (stiﬀness, tremor, and shuﬄing gait), akathisia (motor restlessness),
and acute dystonic reactions (tightening of facial and neck muscles). The
involuntary muscle movements of EPSs have been confused with seizures.
Antipsychotic drugs may also lower seizure threshold, reduce blood
pressure, and elevate prolactin levels. One subclass of antipsychotics, the
phenothiazines, may inhibit the learning processes necessary for behavioral
modiﬁcation techniques [16].

372

B.S. Simpson, M.G. Papich

/ Vet Clin Small Anim 33 (2003) 365–404

Table

Psycho

tropic

age

nts

used

cats

Drug

class

Drug

name

Dose

cats

References

Benzodiazepine

(BZD)

Diazepam

1–4

per

cat

q12–24

hours

Papich

(2002)

[6]

0.2–0.4

/kg

q12–24

hours

Cooper

and

Hart

(1992)

[35]

BZD

Alprazolam

0.125–0.25

per

cat

q12

hours

arder

(1991)

[144]

BZD

Oxazepam

2.5

per

cat

PRN

for

ppetite

stimulation

Papich

(2002)

[6]

Azaspirone

Buspirone

2.5–5

per

cat

q12–24

hours

Hart

(1993)

[36]

5.0

per

cat

q12

hours

Sawyer

(1999)

[67]

2.5–5.0

per

cat

q8–12

hours

arder

(1991)

[144]

Tricyclic

antidepressant

(TCA)

Amitriptyline

5–10

per

cat

q24

hours

Papich

(2002)

[6]

2.5–5.0

per

cat

q12–24

hours

awyer

(1999)

[67]

0.5–1.0

/kg

q12

hours

Halip

(1998)

[139]

per

cat

qHS

Chew

(1998)

[68]

TCA

Clomipramine

0.5

/kg

q24

hours

DeHasse

(1997)

[83]

1.25–2.5

per

cat

q24

hours

Sawyer

(1999)

[67]

1–5

per

cat

q12–24

hours

Papich

(2002)

[6]

Selective

serotonin

reuptake

inhibitor

(SSRI)

luoxetine

0.5–4.0

per

cat

q24

ours

apich

(2002)

[6]

/kg

q24

hours

Pryor

(2001)

[90]

1–1.5

per

cat

q24

hours

Hartmann

(1995)

[92]

per

cat

q24–72

hours

Romatowski

(1989)

[122]

SSRI

Paroxetine

1.25–2.5

per

cat

q24

hours

Papich

(2002)

[6]

Monoamine

oxidase

inhibitor

(MAOI)

Selegiline

0.5

/kg

q24

hours

Mills

and

Simpson

(2002)

[140]

Anticonvulsant

arbamazepine

q12

hours

Schwartz

(1994)

[110]

Progestogen

hormones

egestrol

acetate

(see

text)

2.5–5

per

cat

q24

ays,

then

1–2

per

week

Papich

(2002)

[6]

per

cat

7–10

days,then

eod

days,

then

per

eek

Hart

(1980)

[121]

/kg

q24

hours

ays,

then

/kg

q24

days,

then

0.5

/kg

q24

ays

Romatowski

(1989)

[122]

See

text

for

spe

cial

nsider

ations

and

side

eﬀects.

All

doses

are

os.

Abbr

eviatio

eod,

every

oth

day

;

PRN,

need

ed;

qHS,

dtime.

373

B.S. Simpson, M.G. Papich

/ Vet Clin Small Anim 33 (2003) 365–404

Table

Psycho

tropic

age

nts

used

Drug

clas

Drug

name

Dose

dogs

Refe

rences

Benzodiaz

epine

(BZD)

Diazepam

0.55–

2.2

/kg

PRN

umb

(2002)

[26]

BZD

Alprazolam

0.02–

0.1

/kg

q8–12

hour

ndsberg

997)

[138]

0.02

/kg

PRN

(with

clomipramine)

Crowe

ll-Dav

(2001

)

[33]

BZD

Clora

zepate

/kg

q12

hours

Papic

002)

[6]

Forre

ster

(1990

)

[24]

BZD

Loraz

epam

0.02–

0.1

/kg

q8–24

hour

ills

and

Simps

(2002

)

[140]

BZD

Oxazep

0.2–1

/kg

q12–24

hour

ndsberg

997)

[138]

Azaspirone

Busp

irone

2.5–1

per

dog

q12

–24

hours

1.0–2.0

/kg

q12

hour

Papic

002)

[6]

Tricyc

lic

antid

epress

ant

(TCA)

Amitr

iptyline

2.2–4

/kg

q12–24

hour

Juarb

e-D

iaz

997)

[52,53]

/kg

q24

hours

keuchi

(2000

)

[63]

0.74–

2.5

/kg

q12

hour

Reic

000)

[58]

1–2

/kg

q12–24

hour

Papic

002)

[6]

TCA

Clomipram

ine

1–3

/kg

q12

urs

Papic

002)

[6]

1–2

/kg

q12

urs

King

(2000

)

[73]

1–2

/kg

q12

urs

Moo

n-Fan

elli

and

Dodman

(1998

)

[77]

1–2

/kg

q12

urs

Sekse

and

Lind

eman

(2001

)

[76]

/kg

q12

hours

wson

(1998

)

[75]

/kg

q24

hours

Rap

oport

(1992

)

[79]

TCA

Imipramine

2–4

/kg

q12–24

hour

Papic

002)

[6]

Selective

sero

tonin

reupta

inhibitor

(SS

RI)

Fluo

xetine

Start

0.5

/kg

q24

hour

increa

1.0

/kg

q24

urs

Papic

002)

[6]

Dodm

(1996)

[86]

/kg

q24

hours

Rap

oport

(1992

)

[79]

per

q24

hour

chan

and

Berk

(1998

)

[88,1

51]

SSRI

Paro

xetine

0.5–1

/kg

q24

urs

Papic

002)

[6]

0.5–2

/kg

q24

urs

lls

and

Simps

(2002

)

[140]

SSRI

Sertralin

3.42

/kg

q24

hours

Rap

oport

(1992

)

[79]

374

B.S. Simpson, M.G. Papich

/ Vet Clin Small Anim 33 (2003) 365–404

2.5

/kg

q24

urs

Dodman

(person

commu

nication,

2000)

Larson

and

Summ

ers

001)

[97]

Monoa

mine

oxidas

inhibitor

(MA

OI)

Selegiline

0.5–1

/kg

rning

Calve

(2000

)

[104]

Atypical

antid

epressan

Trazodo

/kg

q12

day

then

/kg

q12

hour

and

bolus

time

noi

obia

Anticonvulsa

Phenoba

rbita

0.45

/kg

q24

urs

Crowe

ll-Dav

(1989

)

[106]

1.5–2

/kg

q12

hour

Dodm

992)

[108]

/kg

q12

urs

(with

cloraz

epate)

Forre

ster

(1993

)

[43]

2–8

/kg

q12

urs

Papic

(2002

)

[6]

Carbam

azepine

4–8

/kg

q12

urs

Hollan

(1988

)

[111]

Beta

anta

gonist

Propran

olol

2–3

/kg

q12

urs

(with

phenob

arbital)

Walk

(1997

)

[132]

Narcot

anta

gonist

Naltrexo

2.2

/kg

q12–24

hour

White

(1990

)

[114]

2.2

/kg

q12

urs

Papic

(2002

)

[6]

Progesto

gen

horm

ones

Megestro

acet

ate

text)

Males

/kg

hours

day

then,

impro

ved,

/kg

days,

Joby

(1984

)

[117]

Borc

helt

and

Voith

(1986

)

[116]

2.2

/kg

q24

urs

day

then

1.1

/kg

q24

hours

day

then

0.5

/kg

q24

hours

day

Papic

(2002

)

[6]

Hormon

Melaton

0.1

/kg

q8–24

hour

(with

ami

triptyline)

Aro

nson

999)

[16]

See

text

for

specia

consid

eration

and

side

eﬀec

ts.

All

ses

are

per

os.

Abbre

viations

eve

other

day;

PRN,

needed

;

qHS,

bedtim

375

B.S. Simpson, M.G. Papich

/ Vet Clin Small Anim 33 (2003) 365–404

These EPSs should not be confused with another adverse eﬀect of long-term
antipsychotic medication called tardive dyskinesia. Tardive dyskinesia is
characterized by oral-facial, limb, or truncal dyskinesia or twisting postures.
The diﬀerentiation is that EPSs usually occur soon after administration of
the drugs and tardive dyskinesia occurs after prolonged chronic treatment.
The other important diﬀerentiation is that EPSs are believed to be caused by
a deﬁciency of DA and tardive dyskinesia is caused by an excess of DA or
increased DA receptor sensitivity as a result of chronic administration.
Decreasing the dose or withdrawal of the antipsychotic drug worsens tardive
dyskinesia. To our knowledge, tardive dyskinesia has not been described in
veterinary patients.

Historically, one phenothiazine antipsychotic, acepromazine (PromAce),

has been used in the management of veterinary behavioral problems, such as
noise phobia, by reducing animals’ general attendance to environmental
stimuli and producing sedation. The eﬀectiveness of acepromazine as an
oral anxiolytic is often disappointing and causes undesirable side eﬀects [17].
Its intramuscular use may decrease presurgical patient apprehension and
reduce the dose of other drugs used as general anesthetics, however [18].
Because of the sedative and extrapyramidal effects, acepromazine is not
satisfactory for chronic administration. Other agents, such as the BZDs or
antidepressants, are now preferred because of their speciﬁc antianxiety
effects and relatively low range of side effects. A small percentage of com-
panion animals, predominantly cats, administered acepromazine orally ex-
perience spontaneous motor activity.

Phenothiazines have been used to treat compulsive behaviors not

satisfactorily responsive to serotonergic drugs or in combination with
serotonergic agents [19]. DA has been implicated in some forms of
stereotypic behaviors, perhaps because of its effect on serotonin [20].
Dopaminergic agents like apomorphine and amphetamine and the DA
precursor

-dopa can induce stereotypes in animals [19]. Thioridazine

has been used in one case of aberrant motor behavior in a dog [21]. In
2000, proprietary thioridazine (Mellaril) added a label warning that the drug
has been shown to prolong the corrected QT (QT

) interval and has

been associated with arrhythmias and sudden death in human beings. It was
recommended that thioridazine not be used concurrently with ﬂuvox-
amine, ﬂuoxetine, paroxetine, propranolol, pindolol, or any drug that
affects the QT

interval of the electrocardiogram (ECG) or the CYP 2D6

enzymes.

Some newer atypical antipsychotics, such as risperidone (Risperdal), may

prove to be useful in cases of environmental-speciﬁc anxiety (including
veterinary visits) or impulsive explosive behaviors in dogs and cats [22].
At this time, however, there is not sufﬁcient information available to
recommend safe dosage regimens. The side effects associated with the
traditional antipsychotics may not apply to the newer drugs, although
immobility and transient loss of conditioned responses may be observed.

376

B.S. Simpson, M.G. Papich

/ Vet Clin Small Anim 33 (2003) 365–404

At present, their high cost and the lack of published data make them
impractical for use in animals.

Anxiolytics

The anxiolytic drugs include BZDs, azaspirones, barbiturates, and anti-

histamines. Antidepressants (discussed separately below) also have anxi-
olytic properties. Discussed here are the BZD and azapirone classes.

Benzodiazepines

The BZDs constitute a large class of drugs with a long history of safe and

eﬃcacious use in human beings. All drugs in the class act on BZD receptors
in the CNS to facilitate GABA

, an inhibitory neurotransmitter. After

binding to the GABA

receptor, these drugs enhance the GABA-mediated

conductance through ionic channels and stabilize excitable membranes. The
effects of BZDs on behavior may be attributed to potentiation of GABA
pathways that act to regulate release of monoamine neurotransmitters in
the CNS.

Examples of drugs in this class are diazepam, clorazepate, alprazolam,

oxazepam, and lorazepam. Diﬀerences between speciﬁc drugs in this class
are based on their pharmacokinetic properties. BZDs are used in people
primarily for generalized anxiety disorder or panic disorder [14]; they are
used similarly in small animals [23]. Dosing schedule can affect pharmaco-
kinetics. For example, diazepam in humans [14] or clorazepate in dogs [24]
given one time for anxiety has a lower maximal nordiazepam concentration
compared with the same BZD given twice daily for anxiety. Flumazenil
(Romazicon) is a BZD receptor antagonist and inhibits the effects of BZD.
Flumazenil has been use to counteract the adverse effects of large overdoses
of BZD.

After oral administration, behavioral responses to BZDs are immediate,

although conﬂict studies in animals suggest that the anxiolytic eﬀects are
greater after they have been administered for several days [25]. Sedation,
ataxia, muscle relaxation, increased appetite, paradoxic excitation, and mem-
ory deﬁcits may be observed [26,27]. Tolerance to sedation, ataxia, and
muscle relaxation may develop over the ﬁrst few days of therapy [28].
Owners should be cautioned to assist older animals to avoid falls. Animals
should be observed for hyperphagia when given BZDs on a regular basis.
Agitation and restlessness may occur as an idiosyncratic response to
BZDs. Paradoxic reactions of excitement have been observed in dogs. If
these are observed, the drug should be discontinued and another drug
from another class should be selected. Amnesia from BZDs has been
observed in human beings for many years [29]. In animals, memory deﬁcits
and diminished conditioned responses may also be observed (ie, the animal
may seem to ‘‘forget’’ what it has been previously taught). Difﬁculty in
learning new material, such as desensitization protocols, may be observed.

377

B.S. Simpson, M.G. Papich

/ Vet Clin Small Anim 33 (2003) 365–404

At routine doses, BZDs have little if any eﬀect on cardiovascular

and respiratory systems [14]. BZDs may disinhibit behavior. In people,
manifestations of disinhibition include hostility, aggressiveness, rage reac-
tions, paroxysmal excitement, irritability, and behavioral dyscontrol [30].
These effects are also reported in animals [31]; thus, BZDs should be
used with caution in aggressive animals, because bite inhibition may be
lessened. BZDs are controlled substances with the potential for human
substance abuse. Pet owners should be screened before prescribing, and reﬁll
requests should be carefully scrutinized for appropriate dosing.

After daily administration for more than 1 week, a BZD should be

withdrawn gradually to avoid discontinuation syndrome. Discontinuation
syndrome, especially common in high-potency BZDs like alprazolam,
includes nervousness, tremors, or even seizures [27,32]. The longer a BZD is
taken and the higher the dose used, the greater is the likelihood of
withdrawal reactions when it is discontinued, especially abruptly [2]. Signs
may be reversed by administration of the BZD. Discontinuation syndrome
may be avoided by tapering the BZD dose 25% per week for 1 month.
Commonly used BZDs include diazepam, clorazepate, alprazolam, oxaze-
pam, lorazepam, and temazepam.

Among dogs, BZDs are used to treat fears and phobias as well as

generalized anxiety. BZDs may be added to a TCA like clomipramine to
decrease latency to eﬀect and reduce the panic-like states of thunderstorm
phobia [33] and separation anxiety (B.S. Simpson, MS, PhD, DVM,
personal observation). Among cats, BZDs are used for management of urine
spraying and generalized anxiety, such as anxiety associated with changes in
a new home environment.

Diazepam

Diazepam (Valium) is the best known of the BZDs. It has been used for

behavioral disorders and as a sedative, muscle relaxant, anxiolytic,
anticonvulsant, and adjunct for anesthesia. It has a weak base with a low
pK

and a benign taste and is relatively easy to administer directly or by

mixing in moist food. Its high lipophilicity and rapid distribution make it
suitable for the emergency treatment of seizures because it crosses the BBB
quickly. Its high lipophilicity allows it to be absorbed across membranes
quickly, and it is even rapidly and almost completely absorbed from rectal
administration [34]. The vehicle carrier is not suitable for intramuscular
administration, however. As an oral anxiolytic in dogs, particularly for
panic-like states of thunderstorm phobia and separation anxiety, clinicians
anecdotally report the anxiolytic performance of diazepam to be disappoint-
ing. High doses may be required, which are sufﬁcient to produce ataxia but
insufﬁcient to reduce anxiety. Other agents, such as alprazolam, may be more
satisfactory, or a regimen of daily rather than as-needed dosing may be more
effective. In one study of ‘‘socially withdrawn’’ laboratory Beagles, diazepam
caused behavioral improvement for 4 to 6 hours after acute administration.

378

B.S. Simpson, M.G. Papich

/ Vet Clin Small Anim 33 (2003) 365–404

In cats, diazepam has been used for treatment of urine spraying [35,36].

In open trials, efﬁcacy was approximately 55%, although relapse was
common on discontinuation [15,35].

The pharmacokinetics of diazepam are complex but have been examined

in both dogs and cats [14,28,34,37]. Diazepam undergoes metabolism ﬁrst to
a demethylated metabolite, desmethyldiazepam (also called nordiazepam)
and then to oxazepam. Both of these metabolites are active but not as active
or as lipid soluble as diazepam. Desmethyldiazepam is believed to have
anticonvulsant properties that are equal to [38] or about one third [39] of
the potency of diazepam. The pharmacokinetics of diazepam illustrate the
tremendous species differences in clearance and elimination. In people,
diazepam is considered a drug with low hepatic clearance and a long half-
life. The half-life in people is 43 hours (but may range from 24–48 hours),
and systemic clearance is 0.38 mL

/min/kg. In dogs, the half-life is less than

1 hour, and clearance is in excess of liver blood ﬂow at 57 to 60 mL

/min/kg

[34]. Cats are intermediate between dogs and human beings with a T

5.5 hours and a systemic clearance of 4.7 mL

/min/kg [37]. In all species,

the half-lives of metabolites of diazepam are longer than the half-life
of diazepam. For example, the T

of desmethyldiazepam is 51 to 120

hours, 21.3 hours, and 2.2 to 2.8 hours for human beings, cats, and dogs,
respectively [34,37]. These differences show that for long-term treatment,
diazepam is not suitable for dogs, because frequent administration is
necessary to avoid high peaks and low troughs. Its short half-life and ability
to attain therapeutic concentrations make it suitable for short-term use,
however. This large difference in pharmacokinetics among species for
diazepam means that information published for diazepam is not applicable
to dogs to the same degree. For example, P450 enzyme inhibition and other
drug interactions are not likely in dogs because of the already high systemic
clearance. Because liver clearance is dependent primarily on hepatic blood
ﬂow, however, changes in hepatic perfusion drastically affect diazepam
clearance. Clearance is expected to be altered in dogs with congenital or
acquired hepatic vascular shunts.

The most serious adverse reaction associated with diazepam is that of

hepatic necrosis in cats. This idiopathic hepatic necrosis is a rare but often
fatal condition and has been documented in cats given oral diazepam. The
speciﬁc etiology is not known [40–42]. Diazepam undergoes complex
metabolism to intermediate compounds. It is possible that in susceptible
cats, an aberrant metabolite is produced that is responsible for the hepatic
toxicosis. In the reported cats, the reaction occurred within 7 days of oral
administration of generic or proprietary diazepam [41,42]. It is not known
if there is a greater likelihood of producing this reaction from oral
administration compared with injections of diazepam. It is possible that in
susceptible individuals, metabolites responsible for the toxicosis are more
likely to be produced from oral administration because of ﬁrst-pass
metabolism compared with parenteral administration. Other BZDs

379

B.S. Simpson, M.G. Papich

/ Vet Clin Small Anim 33 (2003) 365–404

(lorazepam and oxazepam) are conjugated directly without undergoing
intermediate metabolism. Alprazolam and temazepam seem to have only
one intermediate metabolite (alphahydroxy metabolite) before undergoing
conjugation. These alternatives may be less likely to induce hepatic toxicosis
in cats, and there have been no cases of hepatic toxicosis reported from the
administration of these BZDs in cats.

Clorazepate

The BZD clorazepate (Tranxene) is metabolized in the acidity of the

stomach to its active metabolite before absorption [14,43]. Clorazepate is
used in dogs for treatment of anxiety disorders, particularly thunder-
storm

/noise phobia. Mean peak nordiazepam levels were detected approx-

imately 98 minutes after a single oral dose of clorazepate and 153 minutes
after multiple oral doses [24], suggesting that improved management is
obtained when the drug in given on a twice daily rather than as-needed
basis. The elimination half-life after a single dose (284 minutes) was not
signiﬁcantly different than after multiple doses (355 minutes) [24]. An oral
dose of 2 mg

/kg administered every 12 hours maintains concentrations of

the active metabolite, nordiazepam, in the range considered therapeutic in
human beings [24]. Excessive ataxia and sedation are uncommon [24].
Although available in a sustained-delivery (sd) formulation, one pharma-
cokinetic study in dogs found no difference in the serum disposition
compared with that of regular-release clorazepate [44].

Alprazolam

Alprazolam (Xanax) is a high-potency BZD shown in human beings to be

an eﬀective treatment for panic disorder. It is used in dogs to treat the panic-
like states of separation anxiety, thunderstorm phobia, and other phobias.
In people, it has a more rapid onset of action and shorter elimination half-
life than diazepam, but these comparisons have not been reported for
animals. In human beings [14,45], plasma concentrations vary greatly
among patients administered identical doses of alprazolam. As in people,
higher doses of alprazolam may be required for panic-like states in dogs,
such as thunderstorm phobia and separation anxiety, compared with gen-
eral anxiety. Thus, individual empiric dosing may be required to achieve
treatment success with the fewest side effects [14]. Canine patients receiving
alprazolam at a moderately high dose once a day, as may occur with
separation anxiety or thunderstorm phobia, are at risk for withdrawal-
induced anxiety or tremors before the next day’s dose because of the short
elimination half-life of alprazolam. This may be avoided by administering
the drug twice a day and not skipping doses. To terminate the drug,
alprazolam should be withdrawn slowly, decreasing the dose over weeks.

Oxazepam and lorazepam

As mentioned previously, these drugs are metabolized directly via phase

II conjugation to inactive compounds [14]. Both oxazepam (Serax) and

380

B.S. Simpson, M.G. Papich

/ Vet Clin Small Anim 33 (2003) 365–404

lorazepam (Ativan) have been used by veterinarians as sedatives, anxio-
lytics, and anticonvulsants, but they are not as well known as diazepam.
There are no active metabolites. Because conjugation reactions are usually
preserved even when there is hepatic disease, these drugs are recommended
for individuals with compromised liver function, for aged canine subjects in
which metabolism may be slowed [14], and for cats in which phase II
metabolism may be less likely to trigger idiopathic hepatic necrosis.

In cats, oxazepam has been used as an appetite stimulant [26] and in the

treatment of certain compulsive behaviors and hyperanesthesia [46].
Lorazepam has the advantage of a greater and more prolonged distribution
to the CNS than other BZDs. Lorazepam in healthy dogs has a T

0.88 hours, a systemic clearance less than half that of diazepam at
19.3 mL

/min/kg, and oral availability of 60% (unpublished data). Therefore,

as an oral drug it may be a suitable alternative to diazepam.

Azaspirone

This class of anxiolytics is represented clinically by one drug, buspi-

rone (Buspar). Buspirone was the ﬁrst nonsedating non-BZD anxiolytic
drug to be developed and marketed [47]. Buspirone acts as a full agonist
at presynaptic 5-HT

receptors, with a resulting decrease in serotonin

synthesis and inhibition of neuronal ﬁring. It also acts as a partial agonist at
postsynaptic 5-HT

receptors. In serotonin-deﬁcit states, buspirone acts as

an agonist [47]. Buspirone also has dopaminergic effects.

Buspirone is not a substrate for CYP enzymes, nor does it inhibit them

[47]. It has no interactions with BZDs, and there are no withdrawal con-
cerns after long-term use [10]. In human beings, buspirone is effective for
treatment of generalized anxiety disorder but is ineffective for the control
of panic disorder. Buspirone shows efﬁcacy in certain animal models of
anxiety, such as the conditioned avoidance response.

Because of its short elimination half-life, buspirone must be administered

two or three times per day. Buspirone has a benign taste and may be given
with food. Buspirone produces no sedation, no memory or psychomotor
impairment, and no disinhibition phenomenon. Unlike the BZDs, buspi-
rone produces no immediate behavioral eﬀects. It is only useful when
administered for several weeks. Side eﬀects are uncommon and mild
but may be noted immediately. They include gastrointestinal signs and
alterations in social behavior. Buspirone has no potential for abuse.

In dogs, buspirone does not seem to be particularly therapeutic for the

panic-like condition of thunderstorm phobia or separation anxiety, but it
has been used for generalized anxiety.

In cats, buspirone is used to modulate states of high arousal, including

feline urine spraying. In an open trial, improvement was observed in 55% of
cats, with a 50% relapse rate after the cessation of treatment [36]. It has also

381

B.S. Simpson, M.G. Papich

/ Vet Clin Small Anim 33 (2003) 365–404

been used to reduce anxiety in the ‘‘pariah’’ cat in cases of intercat
aggression within a household [48,49].

Antidepressants

The general category of antidepressants includes a number of classes of

drugs. Depending on the reference consulted, various classiﬁcation systems
have been used to group these drugs. One classiﬁcation lists the TCAs
together. Drugs in this class include imipramine, amitriptyline, doxepin,
and clomipramine. The heterocyclic drugs of the second and third
generation are the newer drugs with which veterinarians are less familiar.
These drugs include amoxapine and maprotiline, which resemble TCAs, and
trazodone and bupropion, which have distinct chemical structures. New
drugs referred to as third-generation heterocyclic antidepressants include
venlafaxine and mirtazapine. The other antidepressants are the SSRIs, the
monoamine oxidase inhibitors (MAOIs), and a few atypical antidepressants.
These classes diﬀer in their mode of action, their side-eﬀect proﬁle, and their
relative eﬃcacy in certain behavioral conditions. Antidepressants also have
important antianxiety properties; at routine doses, they are generally well
tolerated by animals. To some extent, all share a similar action, which is to
alter neurotransmitter (primarily NE and serotonin) at the receptor sites.
The unifying theory that explains their eﬃcacy is called the monoamine
theory of depression, which was described in 1965 by Schildkraut [50], who
stated that depression in people is caused by a deﬁciency of monoamine
neurotransmitters. Initially, this theory focused on NE, but it became clear
later that 5-HT also was involved as an important neurotransmitter. The
relation of the mechanism to the neuronal effects and clinical response is
complex. The action affecting reuptake of neurotransmitters occurs rapidly,
but the clinical response may take days to weeks for maximum effect. This
implies that the increased presence of neurotransmitters NE and 5-HT also
may affect the sensitivity of receptors either presynaptically or postsynap-
tically. A full discussion of the effects of these drugs on synaptic
transmission and neuronal action is beyond the scope of this article.

Tricyclic antidepressants

The TCAs are so called because they have a three-ring nucleus.

Chemically, they resemble phenothiazines, but their actions diﬀer. The
TCAs block the reuptake of serotonin and NE, and among these drugs, they
vary in the extent to which they inhibit one transmitter more than the other.
The drugs most familiar to veterinarians, amitriptyline (Elavil), doxepin
(Sinequan), and imipramine (Tofranil), are tertiary amines and block the
reuptake of both 5-HT and NE. The secondary amines, such as nortriptyline
(Pamelor) and desipramine (Norpramin), are not as well known to
veterinarians. Secondary amines are relatively selective inhibitors of NE,

382

B.S. Simpson, M.G. Papich

/ Vet Clin Small Anim 33 (2003) 365–404

but the tertiary amines inhibit both serotonin and NE. Clomipramine
(Clomicalm, and Anafranil), another tertiary amine, is more selective for
blocking the reuptake of 5-HT. TCAs also have antagonist activity for the
alpha-1 (a

) receptor and also act as antihistamines and anticholinergic

agents. The antagonist activity on these receptors is responsible for some of
the side effects. As with other drugs discussed in this article, there are species
differences in metabolism, elimination, and susceptibility to side effects. This
is discussed in more detail with speciﬁc drugs.

TCAs have a long history of eﬃcacious use in human beings, but in

recent years, they have been largely replaced by the more speciﬁc SSRIs
because of a lower incidence of side eﬀects. The low cost and tolerance of
the TCAs make them particularly useful in small animal behavioral therapy.
In general, the TCAs moderate excessive arousal and reduce anxiety. TCAs
may enhance learning in speciﬁc circumstances [51]. Unlike the BZDs, the
TCAs do not disinhibit behavior. TCAs are used in dogs to manage mild
aggression, canine compulsive disorders, and various anxiety states [52,53].
TCAs may be used in cats to control certain forms of aggression, inap-
propriate urination and spraying, excessive grooming, anxiety states, and
excessive vocalization.

Because of the time to reach pharmacokinetic steady state and the time

for modulation of the receptors aﬀected by TCAs, therapeutic eﬀects may
not be seen for 2 to 4 weeks. Slow discontinuation of any TCA is re-
commended to avoid withdrawal responses.

Metabolism

TCAs are extensively metabolized and are CYP enzyme substrates. In

people, coadministration of drugs that inhibit the CYP enzyme involved in
their metabolism can cause a fourfold increase in serum levels of TCAs.
Drugs of the selective SSRI class (described below) are potent inhibitors of
CYP enzymes [54]. Speciﬁc drug interferences have not been studied in detail
in animals; however, because of the potential for drug-drug interactions in
human beings, caution is advised when prescribing a TCA with other drugs
known to affect drug metabolism.

Adverse reactions

TCAs are complex drugs with eﬀects at numerous receptors, which

predict their side eﬀects. Although not usually of clinical importance in
young normal patients, side eﬀects may be problematic in aged animals or
those at risk of other medical conditions. Onset to side eﬀects may occur
soon after administration of the drug or after chronic use. Side eﬀects
include mild sedation, GIT side eﬀects (especially vomiting), antihistamine
eﬀects, and anticholinergic eﬀects [55]. Anticholinergic effects may include
dry mouth (and consequent increased water consumption), constipation,
and urinary retention. The use of TCAs is contraindicated in cases of
glaucoma or keratoconjunctivitis sicca.

383

B.S. Simpson, M.G. Papich

/ Vet Clin Small Anim 33 (2003) 365–404

A particularly serious adverse eﬀect is possible when TCAs are

administered at high doses as, for example, in an accidental ingestion by a
pet. High concentrations of TCAs can cause a quinidine-like membrane-
stabilizing eﬀect that can lead to fatal cardiac arrhythmias if patients are not
treated promptly [55–57]. If animals receive a high dose (a lethal dose may
be greater than 15 mg

/kg) [56], an ECG should be monitored immediately

for conduction disturbances. If the necessary equipment is available,
blood pressure should be monitored also. Treatment of overdoses
should be initiated promptly, because deaths have occurred within 2 hours.
Contact an animal poison control center for speciﬁc therapy. Treatment
consists of gastric lavage, activated charcoal, a suitable cathartic (not
containing magnesium), and sodium bicarbonate therapy. Antiarrhythmics
that do not affect conduction may be indicated, such as lidocaine, but
other class I antiarrhythmics, such as procainamide and quinidine, are
contraindicated.

The cardiotoxic eﬀect described previously should not be confused with

the eﬀect on the heart produced as an anticholinergic side eﬀect or
adrenergic agonist action by TCAs. These drugs may elevate heart rate in
some individuals as a result of the anticholinergic eﬀects or lower the heart
rate in others as a reﬂex response to the a

-adrenergic effects. Before ad-

ministering a TCA to a patient, it is suggested in patients at risk to conduct a
cardiac assessment (evaluate history, auscultate, ECG if indicated) before ad-
ministration. In a study in which effects on the cardiac ECG were evaluated
in otherwise healthy dogs administered either clomipramine or amitriptyline,
TCAs did not cause ECG abnormalities [58]. In this study, the drugs were
administered at doses recommended for treatment of behavioral problems.

TCAs may also lower seizure threshold and potentiate seizures in

predisposed animals [52]. Agranulocytosis has occasionally been associated
with TCA administration [59].

TCAs may be diﬃcult to administer directly or to disguise in food

because they have a lingering bitter taste. Biting the tablet can induce taste
aversion, future dosing avoidance, and hypersalivation.

Amitriptyline

. Amitriptyline is used in human beings for depression, anxiety

disorders, and certain types of chronic or neuropathic pain [60]. It exerts
active reuptake inhibition on serotonin receptors relative to NE receptors.
It has strong anticholinergic, antihistamine, a

-adrenergic, and analgesic

properties [55,60]. Nortriptyline, also commercially available (Pamelor), is
the active metabolite [60,61].

Amitriptyline (Elavil) has been a useful drug in animals to enhance

behavioral calming and augment a behavioral treatment program [60,62].
Among dogs, amitriptyline is commonly used for treatment of separation
anxiety [63], aggression [60–65], and repetitive self-trauma [46,66]. The time
for maximum effect is 2 to 4 weeks [60]. In cats, amitriptyline has been used
for psychogenic alopecia [67].

384

B.S. Simpson, M.G. Papich

/ Vet Clin Small Anim 33 (2003) 365–404

Reported side eﬀects in cats include weight gain, somnolence, and

decreased grooming. Once daily nighttime dosing is recommended to avoid
excessive daytime sedation [68]. Because there is a generic formulation, it is
popular among veterinarians desiring inexpensive treatments for patients.

Among the most common use in cats is administration for urinary

disorders. It has been used for urine marking and inappropriate urination
secondary to idiopathic interstitial cystitis [68]. Amitriptyline stimulates
b-adrenergic receptors in smooth muscle, including the urinary bladder, to
cause a decrease in smooth muscle excitability and an increase in bladder
capacity [60,68]. For this reason, it has been used to treat feline interstitial
cystitis at an initial dose of 5–10 mg per cat once in the evening [68]. It is not
known if an analgesic effect or behavior-modifying effect plays a role in the
efﬁcacy for treatment of this disease.

Imipramine

. Imipramine (Tofranil) is a TCA with properties similar to

amitriptyline, except that imipramine has more equal aﬃnity for NE and
serotonin receptors. Imipramine has only moderate aﬃnity for H

muscarinic (anticholinergic) receptors. Imipramine has more serotonergic
activity, fewer anticholinergic effects, and modest a-agonist properties. In
human beings, imipramine has an elimination half-life of approximately 12
hours. Pharmacokinetics are not available for dogs, but, empirically, dogs
have been dosed with imipramine twice a day.

Imipramine is successfully used in people to treat panic disorder in adults

and nocturnal enuresis (bedwetting) in children. It is a modestly priced
treatment for narcolepsy in people, dogs, and horses [69]. Imipramine may
be used to treat separation anxiety [15,70], particularly in those cases in
which urine house soiling is problematic (B.S. Simpson, MS, PhD, DVM,
personal observation). After 14 days at a high dose (10 mg

/kg administered

every 24 hours), imipramine improved abnormal ‘‘withdrawn and de-
pressed’’ behavior in population of laboratory Beagles [71].

Clomipramine

. Clomipramine (Clomicalm, veterinary label; Anafranil,

human label) has the most serotonergic activity of all the TCAs. It is
approved in human beings for the treatment of obsessive-compulsive
disorder (OCD) and was the ﬁrst veterinary drug registered to treat
separation anxiety in dogs.

This is another drug for which there are important species diﬀerences in

metabolism that aﬀect clinical use. Dogs metabolize clomipramine more
rapidly than people. The average T

in dogs is 5 hours [72], 7.2 hours after

administration of a single dose and 2.1 to 4 hours after repeated doses [73].
The T

of the desmethylclompramine, an active metabolite, is 2.9 hours

[72], 1.9 hours [74], and 2.2 to 3.8 hours after multiple doses [73]. By
comparison, the half-life in human beings is 24 hours [55] to 33 hours and
has been reported to be as high as 36 to 50 hours. The systemic clearance in
dogs is rapid (23.3 mL

/kg/min) and is approximately equal to hepatic blood

385

B.S. Simpson, M.G. Papich

/ Vet Clin Small Anim 33 (2003) 365–404

ﬂow. The evidence indicates that in dogs, clomipramine is a high-clearance
drug for which hepatic blood ﬂow is the most important factor in clearance.
Changes in enzyme activity (CYP alterations) or protein binding are not
expected to affect clomipramine pharmacokinetics in dogs as in human
beings.

Oral absorption is only 16% to 20% in dogs [72], which is probably

reduced from ﬁrst-pass hepatic clearance. In two pharmacokinetic studies,
repeated-dose administration of clomipramine led to increases in plasma
concentrations of clomipramine and desmethylclomipramine that were
larger than expected from half-life values [72,74]. Differences between
human beings and dogs regarding hepatic metabolism are also important
when examining the proportion of metabolite produced. In dogs, the ratio
of clomipramine to the desmethylclomipramine metabolite is 3:1, whereas in
people, this ratio is only 1:2.5. Clomipramine is believed to act primarily as a
serotonin reuptake inhibitor, but the desmethyl metabolite is probably most
responsible for the anticholinergic side effects. This may explain why there
seem to be fewer anticholinergic side effects in dogs compared with people
from this drug. Alternatively, the more rapid clearance of both the parent
drug and metabolite may account for fewer adverse effects in dogs.

Despite the relatively short half-life in dogs, its extensive distribution and

potency allow it to be given eﬃcaciously every 12 or 24 hours [73].
Clomipramine has shown efﬁcacy for treatment of canine compulsive
disorders [75,76,145], such as tail chasing [77] or acral lick granuloma
[78,79]. Clomipramine (Clomicalm) is approved in the United States for the
treatment of separation anxiety in dogs [73,80], although there is debate
concerning the relative merits of behavioral versus pharmacologic compo-
nents to treatment [81]. In one study, clomipramine was no more effective
than controls for the treatment of dominance-related aggression in dogs
[82]. Clomipramine may be helpful in some cases of noise phobia [76];
adjunctive treatment with a BZD, such as alprazolam, may be necessary
[33].

Clomipramine is used in cats to manage urine spraying [83], hyperesthesia

(B.S. Simpson, MS, PhD, DVM, personal observation), and some feline
compulsive behaviors, including psychogenic alopecia [67].

Selective serotonin reuptake inhibitors

As the name implies, SSRIs are more speciﬁc than the TCAs. SSRIs

block the reuptake of serotonin, making more available in the synapse, with
little eﬀect on the reuptake of NE. SSRIs available in the United States
include ﬂuoxetine, paroxetine, sertraline, ﬂuvoxamine, and citalopram.

Efficacy

Fluoxetine (Prozac), paroxetine (Paxil), sertraline (Zoloft), and ﬂuvox-

amine (Luvox) have been shown to be eﬀective in treatment of human OCD

386

B.S. Simpson, M.G. Papich

/ Vet Clin Small Anim 33 (2003) 365–404

[11]. Higher doses and longer treatments may be required to obtain a
satisfactory response to OCD compared to other behavioral disorders in
humans [11]. All SSRIs are efﬁcacious for the treatment of panic disorder in
human beings [11]. Often, a low dose is used to initiate therapy and is then
titrated upward as necessary [11]. Eating disorders, impulsivity, aggression,
pain syndromes, and premenstrual dysphoria also have been successfully
treated with SSRIs compared with controls [11].

In dogs, SSRIs have been used clinically for management of separation

anxiety, compulsive behaviors [147], and dominance-type aggression. In
cats, SSRIs are used to treat urine spraying, aggression, and compulsive
behaviors, such as psychogenic alopecia and fabric chewing. Because of
concerns about GIT side eﬀects, cats medicated with SSRIs should be
closely monitored for food and water consumption and fecal and urinary
elimination as well as body weight.

There is evidence that individuals exhibiting violent behavior have low

CNS serotonin activity as measured by the CSF concentrations of serotonin
metabolites [84,85]. Therefore, serotonin-enhancing agents have been used
to treat certain forms of aggression, particularly dominance-related
aggression [86].

Adverse effects

SSRIs have an excellent safety record. Side eﬀects vary from agent to

agent, but include gastrointestinal eﬀects and nervous system alterations
ranging from sedation to agitation, irritability, and insomnia. Fluoxetine
and ﬂuvoxamine shorten rapid eye movement (REM) sleep in animal
models [11]. Gastrointestinal effects (occurring in up to 25% of people) are
likely caused by the concentration of serotonin receptors in the GIT. The
absence of cardiovascular side effects distinguishes this group from the
TCAs. Starting a patient at a low dose and then increasing the dose after
1 week may reduce the likelihood of problematic side effects. Onset to action
may be 3 to 4 weeks.

In people, there is great variation in the eﬀect of SSRIs on CYP enzymes

[142]. These diﬀerences have not been reported in animals. In people,
ﬂuoxetine, ﬂuvoxamine, and paroxetine all inhibit one or more CYP
enzymes; sertraline and citalopram do not. Phenobarbital and other
anticonvulsants induce speciﬁc CYP enzymes, principally CYP1A2 and
CYP3A3

/4. Drug interactions have been infrequently reported from

administration of SSRIs. Because of the type and extent to which the
CYP enzymes are responsible for metabolism of SSRIs in animals compared
to humans, it is diﬃcult to predict potential drug interactions.

Fluoxetine

. Fluoxetine (Prozac) is widely used to treat a range of human

behavioral disorders, including depression, generalized anxiety, panic
disorder, OCD, eating disorders, and premenstrual dysphoria.

387

B.S. Simpson, M.G. Papich

/ Vet Clin Small Anim 33 (2003) 365–404

Among dogs, ﬂuoxetine has been used for treatment of dominance-

related aggression [86,87], interdog aggression [31], and acral lick dermatitis
[79,88,151] as well as other compulsive disorders. Fluoxetine has also been
used to treat stereotypical pacing of 22 years’ duration in a captive polar
bear; relapse occurred after discontinuation of treatment [89].

Among cats, ﬂuoxetine has been used for treatment of refractory urine

spraying [90], inappropriate urination [91], psychogenic alopecia [91,92], and
aggression [48,91].

In the past the cost of ﬂuoxetine has been a disadvantage for some pet

owners. There are a variety of dose forms available for a range of animal
sizes, however. It is available in a solid dose form as 10- and 20-mg capsules
that must be divided or compounded for smaller animals or as 10-mg tablets
that can be broken. There is a 4-mg

/mL oral liquid solution, but it has

0.23% alcohol, which some animals (especially cats) may ﬁnd unpalatable.

Pharmacokinetics and metabolism

. Fluoxetine is metabolized to norﬂuox-

etine, its active metabolite. Norﬂuoxetine has an elimination half-life of
4 to16 days in human beings [11]. This long half-life offers protection from
the discontinuation syndromes associated with abrupt interruption or termi-
nation of treatment. It also necessitates a washout period after discontinua-
tion of ﬂuoxetine and initiation of an MAOI like selegiline [11].

Adverse effects

. In people, extrapyramidal side eﬀects are a rare consequence

of therapy with ﬂuoxetine. These are likely caused by inhibition of
dopaminergic transmission [11]. Inappetence is a common side effect [90],
lethargy has been reported [90,92], and vomiting is rare [90].

In animals, side eﬀects listed in some references include sedation,

inappetence, GIT eﬀects, irritability, hyperactivity, or panting [26], but treat-
ment with ﬂuoxetine in dogs for acral lick dermatitis did not increase rating
scales for fearfulness, territorial aggression, dominance aggression, or excit-
ability, suggesting an absence of negative behavioral changes associated with
this treatment [88].

Paroxetine

Like ﬂuoxetine, paroxetine is used in human beings to treat a range of

psychiatric complaints, including depression, social anxiety, and panic
disorder. It has shown eﬃcacy for the management of episodic aggressive
rages in patients with Tourette’s syndrome [93]. Paroxetine is considered by
some to be the ﬁrst choice for treatment of generalized anxiety.

In dogs, paroxetine may be helpful for the treatment of canine aggression

and canine compulsive disorders. In cats, paroxetine may be used for
compulsive behaviors, redirected aggression, and generalized anxiety.

One advantage of paroxetine compared with ﬂuoxetine is that paroxetine

is available 10-, 20-, 30-, and 40-mg tablets that can be more easily divided
than capsules and administered orally to pets.

388

B.S. Simpson, M.G. Papich

/ Vet Clin Small Anim 33 (2003) 365–404

Metabolism

. Compared with ﬂuoxetine, in humans paroxetine has a shorter

elimination half-life and reaches steady state more rapidly. Pharmacoki-
netics have not been reported for animals, however.

Adverse effects

. Unlike ﬂuoxetine, anticholinergic side eﬀects, such as dry

mouth and constipation, are more common with paroxetine. Paroxetine can
cause an idiosyncratic dose-dependent increase in arousal, wakening, and
REM suppression in people [11] and dogs (B.S. Simpson, MS, PhD, DVM,
personal observation). After chronic administration, paroxetine should be
gradually withdrawn over several weeks so as to avoid a discontinuation
reaction, typiﬁed by increased anxiety [94].

Constipation is a common anticholinergic side eﬀect of paroxetine use in

cats. Cats should be carefully monitored for food and water consumption
and urine and feces production during the ﬁrst week of therapy. Reducing
the target dose by half for the ﬁrst week can avoid such side eﬀects (B.S.
Simpson, MS, PhD, DVM, personal observation).

Sertraline

. Sertraline is an SSRI with pharmacokinetic properties in human

beings similar to those of paroxetine. Like other SSRIs, it is used for a range
of behavioral disorders in people, including panic disorder, chronic
depression, compulsive disorders, and anxiety [95]. It is considered by some
to be the ﬁrst choice for treatment of panic disorders in human beings and
may provide some cognitive beneﬁts to elderly patients compared with other
SSRIs. Like other SSRIs, sertraline has been used as a treatment for eating
disorders. In one report, sertraline was used to successfully treat chronic
regurgitation deemed to be a compulsive behavior in a chimpanzee [96].

In dogs, sertraline can be helpful for compulsive behaviors [79],

aggression [97], and anxiety disorders. In dogs, diarrhea is a common side
effect. This may be circumvented by starting the drug at a low dose and
increasing the dose over 2 weeks.

Fluvoxamine

. Fluvoxamine is distinguished by its lack of activity in human

beings at the CYP2D6 isozyme. Therefore, in people, it has been more safely
combined with TCAs than other SSRIs. Fluvoxamine should be tapered oﬀ
gradually so as to avoid discontinuation syndrome [11].

Citalopram

. Citalopram is the SSRI most recently approved by the US

Food and Drug Administration (FDA) after a long history of safe human
administration in Europe [98]. The delay in US approval was, in part, a
reaction to the results of a long-term toxicity study in which 10 dogs given
an extremely high dose of citalopram (8 mg

/kg/d) suffered 50% mortality

because of cardiac effects after 17 to 31 weeks of treatment (Celexa Product
Information, Forest Pharmaceuticals; Warner-Lambert Company, St.
Louis, MO). Based on these ﬁndings suggesting increased cardiac sensitivity
to citalopram in the dog, therapeutic veterinary use is not recommended. To
date, there are no published reports of its use in cats.

389

B.S. Simpson, M.G. Papich

/ Vet Clin Small Anim 33 (2003) 365–404

Other actions of selective serotonin reuptake inhibitors

It has recently been recognized that SSRIs may have eﬀects other than

those on the CNS. The SSRIs have been shown to have antimicrobial activity,
primarily against gram-positive bacteria [143]. They have little activity on
gram-negative bacteria. They also have activity against fungi such as Candida
spp and Aspergillus spp [99]. Despite the attractiveness of this feature of the
SSRIs, they have not been employed as anti-infective drugs at this time. It is
not known if the antimicrobial effects are relevant at clinical dosages.

Atypical antidepressants

The atypical antidepressants include heterocyclic drugs that have been

classiﬁed by some references as second- and third-generation antidepressant
drugs. In other references, they do not fall into distinct categories [12]. These
drugs include trazodone (Desyrel), nefazodone (Serzone), bupropion
(Welbutrin), and mirtazapine (Remeron).

Trazodone

Most notable in this group for clinical use in small animals is trazodone.

Trazodone is a mixed serotonergic agonist

/antagonist. It has no anti-

cholinergic eﬀects, moderate antihistaminergic activity, and is an antagonist
of postsynaptic a

-adrenergic receptors. Trazodone is well absorbed after

oral administration, with peak blood levels in human beings approximately
1 hour after administration on an empty stomach and 2 hours after dosing
when given with food [100]. Empirically, rapid absorption also occurs in
dogs (B.S. Simpson, MS, PhD, DVM, personal observation).

In human beings, the cytochrome isozyme CYP2D6 is involved in the

metabolism of trazodone, and caution is advised when prescribing trazodone
with SSRIs that inhibit CYP2D6 [100]. The wide safe dose range for
trazodone makes this less problematic than with other drugs, such as the
TCAs, however. Trazodone is used in people for treatment of major
depression and to counter the sleep disturbances caused by SSRIs. When
given over weeks, it produces anxiolytic properties similar to diazepam [100].

Trazodone may be used in dogs for mild thunderstorm phobia and as an

adjunct to TCA or SSRI treatment (B.S. Simpson, MS, PhD, DVM,
personal observation). Trazodone may have some limitations, however. In
one study of four ‘‘depressed’’ laboratory Beagles, no eﬀect was noted after
administration of trazodone at a high dose (10 mg

/kg) [71]; its effects seem

insufﬁcient in severe cases of thunderstorm phobia (B.S. Simpson, MS,
PhD, DVM, personal observation). Side effects in dogs include sedation and
GIT symptoms, including vomiting and diarrhea (B.S. Simpson, MS, PhD,
DVM, personal observation), particularly during the ﬁrst few days of
dosing. Starting at a low dose and titrating the dose up over the initial days
and subsequent weeks may be helpful. Priapism, noted as a rare side effect in
human men, has not been observed in neutered male dogs (B.S. Simpson,
MS, PhD, DVM, personal observation).

390

B.S. Simpson, M.G. Papich

/ Vet Clin Small Anim 33 (2003) 365–404

Monoamine oxidase inhibitors

MAOIs are drugs that inhibit the intracellular enzyme MAO. Because

MAO catabolizes intracellular monoamine neurotransmitters, such as
serotonin, DA, NE, and tyramine, MAOIs inhibit this process, causing an
increase in monoamines. There are two MAO subtypes: A, which aﬀects
serotonin, DA, NE, and tyramine, and B, which aﬀects the metabolism of
phenylethylamine as well as DA. Location of the enzymes also characterizes
the diﬀerences between types A and B. Type A is located in the intestine as
well as in the CNS. If inhibited, it can result in a decrease in the metabolism
of compounds from foods that can produce systemic eﬀects.

The nonselective (types A and B) MAOIs are not used to much extent

in veterinary medicine. Only one MAO-B inhibitor, selegiline (

-deprenyl,

Anipryl), is clinically important as a behavioral treatment in animals.
Selegiline is approved by the FDA for treatment of canine cognitive
dysfunction, a disorder of elderly dogs characterized by decreased social
interaction, loss of house training, confusion, and changes in sleep cycle
[101,102]. There is evidence that these behavioral changes have a
histopathologic [103] and metabolic basis. In clinical trials, selegiline
signiﬁcantly improved clinical signs of cognitive dysfunction in treated
dogs compared with controls. Selegiline may also improve learning [51]. In
general, selegiline is given daily for 1 month. Any improvement in clinical
signs dictates continuation of treatment; additional improvement is often
seen in subsequent months [104]. Treatment failure should prompt an
increase in daily dose and an additional trial of medication. As an extralabel
application, selegiline has also been used in geriatric cats diagnosed with
cognitive dysfunction.

Side eﬀects are uncommon, but high doses can cause hyperactivity and

stereotypic behavior in dogs. Selegiline is metabolized to

-amphetamine

and

-methamphetamine in dogs. Phenylethylamine also is increased in the

CNS of treated dogs, which can produce amphetamine-like effects. It is
possible that some of the clinical effects and adverse effects observed at high
doses are related to these compounds.

Drug interactions with selegiline

Because the topical acaricide for animals, amitraz, also is an MAOI, the

use of TCAs and SSRIs should not be used concurrently with MAOIs like
selegiline (

-deprenyl) or amitraz [146].

MAOIs can suﬃciently inhibit catabolism of monoamines so that their

concentration, particularly serotonin, becomes toxic. This may occur when
other antidepressants that inhibit reuptake of serotonin or inhibit MAO are
used concurrently. This can lead to serotonin syndrome, a potentially fatal
condition characterized by hypertension, hyperthermia, restlessness, tremor,
seizures, and altered mental status [11,57,100]. Thus, an MAOI should not
be used with an antidepressant, including TCAs [60]. Drugs from one

391

B.S. Simpson, M.G. Papich

/ Vet Clin Small Anim 33 (2003) 365–404

class should be discontinued 14 days before agents from the other class are
initiated [60].

Because selegiline may aﬀect other monoamine synthesis, there has been

concern about its administration with sympathetic amines used to treat
urinary incontinence in dogs, such as phenylpropanolamine. When this was
studied, however, selegiline had no eﬀect on pulse rate, ECG, or behavior
compared with administration of phenylpropanolamine alone in dogs [105].

Concurrent use of a

agonists, phenothiazines, and opiate analgesics is

discouraged with selegiline. A standard recommendation is to wait at least
ﬁve times the elimination half-life of the SSRI or its active metabolite
(whichever is longer) before administering the next serotonergic agent [11].
A washout period of 1 to 3 weeks is recommended after discontinuation of
an MAOI and initiation of another drug that affects monoamines.

Because selegiline is considered to treat a pathologic condition of aging

dogs, it is not considered in any more detail in this article. At therapeutic
doses, it does not have eﬀects on anxiety or depression and is not used for
this purpose.

Miscellaneous drugs

Anticonvulsants

Some anticonvulsants have been used to manage veterinary behavioral

problems. Some maladaptive behaviors, such as tail chasing and unpro-
voked rage aggression in Bull Terriers, have been postulated to be partial
complex seizures caused, at least in part, by their positive response to
phenobarbital [106–108]. In other cases with similar presentations,
phenobarbital treatment is ineffective [109]. A response may be obtained
with narcotic antagonists [109] or anticompulsive agents [75].

Other applications exist for the behavioral use of anticonvulsants.

Carbamazepine (Tegretol) has been used for the treatment of aggression in
two cats [110] and for treatment of psychomotor seizures in a dog [111].
Because blood dyscrasias may result (at least in human beings), regular
evaluation of the complete blood cell count (CBC) is recommended for
patients given carbamazepine. New anticonvulsants, such as gabapentin
(Neurontin), topiramate (Topamax), lamotrigine (Lamictal), and tiagabine
(Gabitril), show promise for management of anxiety disorders and
other behavioral disturbances, but their current high cost limits their
practical use.

Opiate antagonists

Stereotypic compulsive behaviors, such as self-traumatic licking, tail

chasing, and pacing, are seen in animals conﬁned to zoos and laboratory set-
tings [112]. It has been postulated that such behaviors are ‘‘coping strategies’’
that lead to release of endogenous opiates (endorphins). Some cases of dogs
and other animals have been successfully treated with the oral narcotic
antagonist naltrexone [109,112–114] or other opiate antagonists [113].

392

B.S. Simpson, M.G. Papich

/ Vet Clin Small Anim 33 (2003) 365–404

There is evidence that there may be a developmental component to the

expression of stereotypies [20]. Other client-owned dogs have developed
acral lick dermatitis associated with repetitive licking for unknown reasons
[113,114]. Because the etiology of stereotypic behavior (based on response to
treatment) is uncertain, treatment with anticompulsive [75] or anticonvul-
sant [108] agents should also be considered in such cases.

Progestin hormones

Historically, synthetic progestins have been used to treat a wide range of

behavioral disorders, likely because of their mild sedating eﬀects on steroid
receptors in the CNS [115]. This nonspeciﬁc use has declined with the advent
of the more speciﬁc agents described previously. The use of progestin
hormones for behavioral therapy is now considered a ‘‘last resort’’ therapy
to avoid abandonment or euthanasia of the offending animal.

In castrated and entire male dogs, megestrol acetate (Ovaban) has been

used for the treatment of dominance [116] and intermale aggression,
mounting, urine marking, and tendency to roam [117]. In one study of 123
male dogs, 75% improved with hormone treatment [117]. Side effects
included increased appetite and lethargy. Relapse within 3 months was most
common in dogs that had signs of dominance aggression and marking [117].
Such treatment signiﬁcantly impairs adrenocortical function as measured by
adrenocorticotropic hormone (ACTH) stimulation during treatment [118].

In cats, megestrol acetate may reduce the incidence of aggression

[119,120] and urine spraying, particularly among neutered male cats [121].
The long-acting preparation of medroxyprogesterone acetate (Depo-
Provera) has been used with a similar level of efﬁcacy [121]. The depot
injection is usually administered intramuscularly or subcutaneously once per
month. Side effects include hyperphagia, obesity, hyperglycemia leading
to diabetes, lethargy, mammary gland hyperplasia and adenocarcinoma,
pyometra, and bone marrow suppression [119,122].

Melatonin

Melatonin is an endogenous hormone produced from serotonin in the

pineal gland. At high doses in dogs (1–1.3 mg

/kg administered every

12 hours), melatonin aﬀects endogenous sex hormones but does not aﬀect
prolactin or thyroid concentrations in adult dogs [123]. Melatonin has
been used with amitriptyline to manage thunderstorm phobia [16] because
it impairs psychomotor vigilance [124]. No controlled studies on the be-
havioral effects of melatonin on dogs have been conducted.

Beta blockers

The b-adrenergic antagonist propranolol (Inderal) has been used to

ameliorate the sympathetic symptoms of anxiety in human beings, including
trembling, sweaty palms, and tachycardia, and to treat organically based

393

B.S. Simpson, M.G. Papich

/ Vet Clin Small Anim 33 (2003) 365–404

aggression [125]. No controlled studies have investigated the efﬁcacy of
b blockers for treatment of anxiety disorders in small animals [126].

Pindolol is a b-adrenergic blocker that is also an antagonist and partial

agonist at 5-HT

receptors. Pindolol may disinhibit serotonin neurons and

may serve as a useful adjuvant therapy [10].

Drug combinations

Combined use of antidepressants

To improve overall eﬃcacy or ﬁne-tune medication management, two

or more behavioral drugs may be used concurrently. In such cases, the
interactive eﬀects of the drugs must be considered, particularly with respect
to their eﬀects on drug metabolism. Although these interactions have not
been well documented or studied in animals, based on evidence in human
beings, there is reason to be cautious [55].

Selective serotonin reuptake inhibitors and antipsychotics

Neuroleptic phenothiazines may be used in combination with seroto-

nergic agents to treat compulsive behaviors not satisfactorily responsive
to serotonergic agents alone [19]. In one recent retrospective study of
treatment-resistant OCD in human patients, the antipsychotic quetiapine
added to SSRI therapy (primarily ﬂuoxetine) improved outcome [127];
in another study of depressed patients nonresponsive to SSRIs [128]
and patients with Tourette’s syndrome with episodic rage managed with
paroxetine [93], the addition of risperidone was beneﬁcial. Many pheno-
thiazines are metabolized by the CYP2D6 and CYP3A subfamilies of
enzymes; thus, drug interactions should be considered before initiating
multidrug therapy.

Adjunctive use of benzodiazepines

In dogs, TCAs may be used with BZDs for treatment of thunderstorm

phobia [33] and separation anxiety [63]. When alprazolam is given
concomitantly with ﬂuoxetine, the result is a 30% increase in alprazolam
levels (but no signiﬁcant increases in ﬂuoxetine or norﬂuoxetine plasma
concentrations) as a result of CYP3A inhibition [129]. Therefore,
coadministration may permit a lower dose of alprazolam to be effective.
Fluvoxamine inhibits the CYP3A4 enzyme and can be associated with
increased levels of alprazolam [14]. In fact, the use of a BZD and an SSRI is
a useful strategy with panic disorder in human patients refractory to single
drug therapy [10]. Similar strategies may be helpful in animals, particularly
dogs. Oxazepam has been shown to decrease turnover of serotonin and NE
in people [14]. In one pharmacokinetic study in dogs, clorazepate was used
concurrently with phenobarbital [43]. The amount of the active metabolite

394

B.S. Simpson, M.G. Papich

/ Vet Clin Small Anim 33 (2003) 365–404

nordiazepam in circulation during each dose interval was signiﬁcantly
reduced compared with the administration of clorazepate alone [24].

Adjunctive use of buspirone

Buspirone, the serotonin 1A partial agonist, is used to augment certain

antidepressants, particularly if with SSRI use, intraneuronal serotonin may
have been depleted. Concomitant use may increase the eﬀectiveness of SSRI
action by repleting 5-HT. Buspirone also may act directly on autoreceptors
to inhibit neuronal impulse ﬂow, possibly allowing repletion of 5-HT stores.
Also, buspirone may act at 5-HT

receptors to aid in the targeted de-

sensitization of 5-HT

autoreceptors [10]. Buspirone has been used to

augment SSRI treatment for OCD, with success in some studies [2] but not
in others [130]. In dogs, buspirone has been used with TCAs to treat
separation anxiety [63] and with an SSRI (ﬂuoxetine) to treat a complex case
involving anxiety, aggression, and stereotypic behavior [131].

Adjunctive use of trazodone

Trazodone may be used in combination with antidepressants, particularly

SSRIs, to enhance behavioral calming, decrease agitation, and aid with
sleep. This can be helpful in human beings with depression, OCD, and other
anxiety disorders [10]. It may be similarly effective in dogs (B.S. Simpson,
MS, PhD, DVM, personal observation).

Adjunctive use of beta blockers

In Great Britain, propranolol is used in combination with phenobarbital

to manage phobic behavior [132]. When propranolol is used in combination
with the antipsychotic thioridazine, thioridazine plasma levels increase.

A beta blocker with 5-HT

autoreceptor antagonist properties, pindolol,

may disinhibit serotonin neurons and may serve as a useful adjuvant therapy
to TCAs [10,133]. In one study, however, pindolol was not superior to placebo
for augmenting the effects of paroxetine for social anxiety symptoms [134].

Adjunctive use of phenobarbital

When phenobarbital (5 mg

/kg administered every 12 hours) was

administered with clorazepate (2 mg

/kg administered every 12 hours) for

44 days, the amount of the active metabolite nordiazepam in circulation
during each dose interval was signiﬁcantly reduced compared with that in
dogs receiving clorazepate alone [24,43], likely a result of CYP effects.
Elimination half-life was not signiﬁcantly altered by concurrent admin-
istration of phenobarbital [43]. When paroxetine is coadministered with
phenobarbital, the result is a decreased plasma level of paroxetine. In Great
Britain, phenobarbital is combined with propranolol for treatment of
anxiety disorders [132].

395

B.S. Simpson, M.G. Papich

/ Vet Clin Small Anim 33 (2003) 365–404

Adjunctive use of hormones

Reproductive hormones have been used in postmenopausal women in

combination with a ﬁrst-line SSRI to enhance therapeutic eﬀects [10]. Al-
though controlled clinical trials are not available, clinicians anecdotally report
beneﬁts. Such a strategy may be helpful in spayed female dogs with refractory
behavioral problems, although no published reports are currently available.

Thyroid hormone has been used to accelerate [133] or augment [10] TCA

response, particularly in women. Thyroid supplementation may increase
cortical 5-HT concentrations and desensitize autoinhibitory 5-HT

receptors in the raphe area, resulting in disinhibition of cortical and
hippocampal 5-HT release [133].

Melatonin is produced by the pineal gland according to the light

/dark

cycle, with light acting as a production suppressant. The nocturnal secretion
of melatonin is primarily induced by increased noradrenergic neuro-
transmission, resulting in increased activity of the rate-limiting enzyme that
converts serotonin to melatonin. In recent years, melatonin has become a
popular over-the-counter remedy for insomnia in people and behavioral
calming in dogs, although data on safely, eﬃcacy, or appropriate doses are
not available [10]. Melatonin given with amitriptyline was used in one dog to
improve signs of generalized anxiety disorder [16].

Parasite prophylaxis

Most veterinary behaviorists prescribe psychotropics to patients receiving

heartworm and ﬂea prophylaxis without incident. Clinical trials report
similar ﬁndings [73]. One potential exception is the acaricide amitraz, which
has properties of MAO inhibition. Concurrent use of amitraz products
(eg, PrevenTic collar [Virbac, Fort Worth, TX], Mitaban dip [Pharmacia-
Upjohn Corp., Kalamazoo, MI]) with drugs classiﬁed as antidepressants
is contraindicated.

Transdermal application

The necessity of daily medications over months to intractable animals has

led to interest in transdermal medication application [141]. Transdermal drugs
are applied to the epidermis and then absorbed through the skin to produce
systemic eﬀects. Special vehicles in the form of gels and ointments have been
developed for this purpose. There are no published clinical trials that compare
the plasma concentration, safety, and eﬃcacy of transdermal preparations
compared with their oral counterparts. In the few studies in which transder-
mal gels have been evaluated in cats for administration of other drugs,
the absorption has been low and unpredictable. Therefore, we are reluctant
to recommend this practice for administration of behavioral drugs. Never-
theless, if it is employed as a last resort in animals that cannot be medicated
orally, knowledge of the pharmacokinetics of speciﬁc agents and collabo-

396

B.S. Simpson, M.G. Papich

/ Vet Clin Small Anim 33 (2003) 365–404

ration with an experienced formulating pharmacist are requirements for safe
transdermal formulation. For drugs like amitriptyline and clomipramine in
which approximately 50% is extracted via ﬁrst pass through the liver, the dose
should be reduced signiﬁcantly (by up to 50%). One common preparation is
amitriptyline or clomipramine (5 mg per 0.1 mL) suspended in a transdermal
vehicle. For drugs like buspirone with extensive (up to 95%) ﬁrst-pass he-
patic metabolism or ﬂuoxetine with long elimination half-lives of the active
metabolite, transdermal application may not be safe.

Treatment success

In each case in which psychotropic medication is administered, there

must be a means of documenting treatment response. The ﬁrst step is to
determine target signs that can be documented by the client with regard to
frequency, intensity, and duration. The second step is to document the
occurrence of these signs over time. Treatment response is often deﬁned as a
50% or greater improvement in symptoms.

Clients must also be educated as to probable side eﬀects and duration to

eﬀect. Many behavioral drugs produce side eﬀects within hours or days of
ﬁrst administration, but may require weeks to onset of desired behavioral
eﬀects.

The duration of treatment has not been systematically investigated. One

strategy is to continue treatment for 2 months after a satisfactory treatment
response and then gradually decrease the dose over weeks. If the status quo
is maintained, the drug can be discontinued. If treatment success wanes, the
previous dose should be reinstated for an additional 6 months and the
process repeated.

With the exception of clomipramine (Clomicalm) and selegiline (Anipryl),

the behavioral drugs discussed here are not approved by the FDA for animal
use. The limitations and risks of extralabel prescribing should be explained to
the pet owner [135,136]. An evaluation of the medical and behavioral history,
a physical and neurologic examination, and appropriate laboratory tests
should precede prescribing any psychotropic agent and should be repeated
at reasonable intervals during treatment. The risk associated with treating
animals aggressive to people, especially children, should be carefully con-
sidered [137]. Patients should be monitored at regular intervals.

References

[1] Stein DJ, Dodman NH, Borchelt P, Hollander E. Behavioral disorders in veterinary

practice. Compr Psychiatry 1994;35:25–8.

[2] Janicak PG, Davis JM, Preskorn SH, et al. Pharmacokinetics. In: Principles and practice

of psychopharmacotherapy. 2nd edition. Baltimore: Williams & Wilkins; 1997. p. 61–84.

397

B.S. Simpson, M.G. Papich

/ Vet Clin Small Anim 33 (2003) 365–404

[3] DeVane CL. Principles of pharmacokinetics and pharmacodynamics. In: Schatzberg AF,

Nemeroff CB, editors. The American Psychiatric Press textbook of psychopharmacology.
2nd edition. Washington, DC: American Psychiatric Press; 1998. p. 155–69.

[4] Chauret N, Gauthier A, Martin J, Nicoll-Grifﬁth DA. In vitro comparison of cytochrome

P450-mediated metabolic activities in human, dog, cat, and horse. Drug Metab Dispos
1997;25:1130–6.

[5] Kuroha M, Kuze Y, Shimoda M, Kokue E. In vitro characterization of the inhibitory

effects of ketoconazole on metabolic activities of cytochrome P450 in canine hepatic
microsomes. Am J Vet Res 2002;63:900–5.

[6] Papich MG. Saunders handbook of veterinary drugs. Philadelphia: WB Saunders; 2002.
[7] Jolliet-Riant P, Tillement J-P. Drug transfer across the blood-brain barrier and

improvement in brain delivery. Fundam Clin Pharmacol 1999;13:16–26.

[8] Pardridge WM. Blood-brain barrier biology and methodology. J Neurovirol 1999;5:556–69.
[9] Baldessarini RJ. Drugs and the treatment of psychiatric disorders: psychosis and anxiety.

In: Hardman JG, Limbird LE, editors. Goodman and Gilman’s the pharmacological basis
of therapeutics. 9th edition. New York: McGraw-Hill; 1995. p. 431–46.

[10] Stahl SM. Essential psychopharmacology: neuroscientiﬁc basis and practical applications.

2nd edition. Cambridge: Cambridge University Press; 2000.

[11] Tollefson GD, Rosenbaum JF. Selective serotonin reuptake inhibitors. In: Schatzberg AF,

Nemeroff CB, editors. The American Psychiatric Press textbook of psychopharmacology.
2nd edition. Washington, DC: American Psychiatric Press; 1998. p. 219–35.

[12] Simpson BS, Simpson DM. Behavioral pharmacotherapy. Part 1. Antipsychotics and

antidepressants. Compend Contin Educ Pract Vet 1996a;18:1067–81.

[13] Hoyer D, Clarke DE, Fozard JR, et al. International union of pharmacology classiﬁcation

of receptors for 5-hydroxytryptamine (serotonin). Pharm Rev 1994;46:157–94.

[14] Ballenger JC. Benzodiazepines. In: Schatzberg AF, Nemeroff CB, editors. The American

Psychiatric Press textbook of psychopharmacology. 2nd edition. Washington DC:
American Psychiatric Press; 1998. p. 271–86.

[15] Marder SR. Antipsychotic medications. In: Schatzberg AF, Nemeroff CB, editors. The

American Psychiatric Press textbook of psychopharmacology. 2nd edition. Washington,
DC: American Psychiatric Press; 1998. p. 309–21.

[16] Aronson L. Animal behavior case of the month: extreme fear in a dog. JAVMA 1999;

215:22–4.

[17] Overall KL. Animal behavior case of the month: separation anxiety in a dog. JAVMA

1998a;212:1702–4.

[18] Light GS, Hardie EM, Young MS, et al. Pain and anxiety behaviors of dogs during

intravenous catheterization after premedication with placebo, acepromazine, or oxy-
morphone. Appl Anim Behav Sci 1993;37:331–43.

[19] Goodman WK, McDougle CJ, Price LH, et al. Beyond the serotonin hypothesis: a role for

dopamine in some forms of obsessive compulsive disorder? J Clin Psychiatry 1990;51:36–43.

[20] Kennes D, Odberg FO, Bouquet Y, et al. Changes in naloxone and haloperidol effects

during the development of captivity-induced jumping stereotypy in bank voles. Eur J
Pharmacol 1988;153:19–24.

[21] Jones RD. Use of thioridazine in the treatment of aberrant motor behavior in a dog.

JAVMA 1987;191:89–90.

[22] Chouinard G, Arnott W. Clinical review of risperidone. Can J Psychiatry (Suppl)1993;

38:S89–95.

[23] Simpson BS, Simpson DM. Behavioral pharmacotherapy. Part II. Anxiolytics and mood

stabilizers. Compend Contin Educ Pract Vet 1996b;18:1203–13.

[24] Forrester SD, Brown SA, Lees GE, et al. Disposition of clorazepate in dogs after single-

and multiple-dose oral administration. Am J Vet Res 1990;51:2001–5.

[25] File SE. Animal models for predicting clinical efﬁcacy of anxiolytic drugs: social behavior.

Neuropsychobiology 1985;13:55–62.

398

B.S. Simpson, M.G. Papich

/ Vet Clin Small Anim 33 (2003) 365–404

[26] Plumb DC. Veterinary drug handbook. Ames (IA): Iowa State Press; 2002. 2, 21, 35, 37,

109, 169, 200, 205, 245, 296, 372, 430, 510, 515, 627, 650.

[27] Roy-Byrne PP, Cowley DS. Benzodiazepines in clinical practice: risks and beneﬁts.

Washington, DC: American Psychiatric Press; 1991.

[28] Lo¨scher W, Frey H-H. Pharmacokinetics of diazepam in the dog. Arch Int Pharma-

codynam Ther 1981;254:180–95.

[29] King DJ. Benzodiazepines, amnesia and sedation: theoretical and clinical issues and

controversies. Hum Psychopharmacol 1992;7:79–87.

[30] Dietch JF, Jennings RK. Aggressive dyscontrol in patients treated with benzodiazepines.

J Clin Psychiatry 1988;49:184–8.

[31] Dodman NH. Animal behavior case of the month: interdog intrahousehold aggression.

JAVMA 2000;217:1468–72.

[32] Roy-Byrne PP, Sullivan MD, Cowley DS, Ries RK. Adjunctive treatment of benzo-

diazepine discontinuation syndromes: a review. J Psychiatr Res 1993;27(Suppl 1):143–53.

[33] Crowell-Davis SL, Seibert L, Sung W, et al. Treatment of storm phobia with a

combination of clomipramine, alprazolam, and behavior modiﬁcation: a prospective open
trial. In: Abstracts of the American Veterinary Society of Animal Behavior, Boston, 2001.
Boston: AVSAB, 2001. p. 5–6

[34] Papich MG, Alcorn J. Absorption of diazepam after its rectal administration in dogs.

Am J Vet Res 1995;56:1629–35.

[35] Cooper L, Hart BL. Comparison of diazepam with progestin for effectiveness in

suppression of urine spraying behavior in cats. JAVMA 1992;200:797–801.

[36] Hart BL, Eckstein RA, Powell KL, et al. Effectiveness of buspirone on urine spraying

and inappropriate urination in cats. JAVMA 1993;203:254–8.

[37] Cotler S, Gustafson JH, Colburn WA. Pharmacokinetics of diazepam and nordiazepam

in the cat. J Pharm Sci 1984;73:348–51.

[38] Randall LO, Scheckel CL, Banziger RF. Pharmacology of the metabolites of

chlordiazepoxide and diazepam. Curr Ther Res 1965;149:423–35.

[39] Frey H-H, Lo¨scher W. Anticonvulsant potency of unmetabolized diazepam. Pharmacol-

ogy 1982;25:154–9.

[40] Center SA, Elston TH, Rowland PH, et al. Fulminant hepatic failure associated with oral

administration of diazepam in 11 cats. J Vet Emerg Crit Care 1996;6:618–25.

[41] House J, Banks S, Lynch L, et al. Fulminant hepatic failure associated with oral

administration of diazepam in 12 cats. JAVMA 1996;209:618–25.

[42] Hughes D, Moreau RE, Overall KL, et al. Acute hepatic necrosis and liver failure associated

with benzodiazepine therapy in six cats, 1986–1995. J Vet Emerg Crit Care 1996;6:13–20.

[43] Forrester SD, Wilcke JR, Jacobson JD, et al. Effects of a 44-day administration of

phenobarbital on disposition of clorazepate in dogs. Am J Vet Res 1993;54:1136–8.

[44] Brown SA, Forrester SD. Serum disposition of oral clorazepate from regular-release and

sustained-delivery tablets in dogs. J Vet Pharmacol Ther 1991;14:426–9.

[45] Brandwein J. Benzodiazepines for the treatment of panic disorder and generalized anxiety

disorder: clinical issues and future directions. Can J Psychiatry 1993;38(Suppl):S109–13.

[46] Overall KL. Pharmacologic treatments for behavior problems. Vet Clin North Am Small

Anim Pract 1997a;27:637–65.

[47] Ninan PT, Cole JO, Yonkers KA. Nonbenzodiazepine anxiolytics. In: Schatzberg AF,

Nemeroff CB, editors. The American Psychiatric Press textbook of psychopharmacology.
2nd edition. Washington, DC: American Psychiatric Press; 1998. p. 287–300.

[48] Overall KL. Intercat aggression: why can’t they all just get along? Veterinary Medicine

1999;94:688–93.

[49] Overall KL. Animal behavior case of the month: use of buspirone to treat spraying

associated with intercat aggression. JAVMA 1994;205:694–6.

[50] Maes M, Meltzer HY. The serotonin hypothesis of major depression. In: Meltzer HY,

editor. Psychopharmacology: the 4th generation of progress. New York: Raven Press; 1995.

399

B.S. Simpson, M.G. Papich

/ Vet Clin Small Anim 33 (2003) 365–404

[51] Mills DS, Ledger R. The effect of oral selegiline hydrochloride on learning and training in

the dog: a psychobiological interpretation. Prog Neuropsychopharmacol Biol Psychiatry
2001;25:1597–613.

[52] Juarbe-Diaz SV. Social dynamics and behavior problems in multiple dog households.

Vet Clin North Am Small Anim Pract 1997a;27:497–514.

[53] Juarbe-Diaz SV. Assessment and treatment of excessive barking in the domestic dog.

Vet Clin North Am Small Anim Pract 1997b;27:515–32.

[54] Mealey KL. Clinically signiﬁcant drug interactions. Compend Contin Educ Pract Vet

2001;24:10–22.

[55] Potter WZ, Manji HK, Rudorfer MV. Tricyclics and tetracyclics. In: Schatzberg AF,

Nemeroff CB, editors. The American Psychiatric Press textbook of psychopharmacology.
2nd edition. Washington, DC: American Psychiatric Press; 1998. p. 199–218.

[56] Johnson LR. Tricyclic antidepressant toxicosis. Vet Clin North Am Small Anim Pract

1990;20:393–403.

[57] Wismer TA. Antidepressant drug overdose in dogs. Veterinary Medicine 2000;95:520–5.
[58] Reich MR, Ohad DG, Overall KL, Dunham AE. Electrocardiographic assessment of

antianxiety medication in dogs and correlation with serum drug concentration. JAVMA
2000;216:1571–5.

[59] Alderman CP, Atchison MM, McNeece JI. Concurrent agranulocytosis and hepatitis

secondary to clomipramine therapy. Br J Psychiatry 1993;162:688–9.

[60] Anderson P. Pharm proﬁle: amitriptyline. Compend Contin Educ Pract Vet 2001;23:

433–7.

[61] Boothe DM. Drugs that modify animal behavior. In: Small animal clinical pharmacol-

ogy and therapeutics. Philadelphia: WB Saunders; 2001. p. 457–72.

[62] Hart BL, Cooper LL. Integrating use of psychotropic drugs with environmental

management and behavioral modiﬁcation for treatment of problem behavior in animals.
JAVMA 1996;209:1549–51.

[63] Takeuchi Y, Houpt KA, Scarlett JM. Evaluation of treatments for separation anxiety in

dogs. JAVMA 2000;217:342–5.

[64] Reich MR. Animal behavior case of the month: food aggression in a Cocker Spaniel

puppy. JAVMA 1999;215:1780–2.

[65] Virga V, Houpt KA, Scarlett JM. Efﬁcacy of amitriptyline as a pharmacological adjunct

to behavioral modiﬁcation in the management of aggressive behaviors in dogs. J Am
Anim Hosp Assoc 2001;37:235–330.

[66] Overall KL. Correct diagnosis, correct dose of amitriptyline [letter]. Vet Forum 1998b;

15:10–1.

[67] Sawyer LS, Moon-Fanelli AA, Dodman NH. Psychogenic alopecia in cats: 11 cases

(1993–1996). JAVMA 1999;214:71–4.

[68] Chew DJ, Bufﬁngton CA, Kendall MS, et al. Amitriptyline treatment for severe recurrent

idiopathic cystitis in cats: 15 cases (1994–1996). JAVMA 1998;213:1282–6.

[69] Coleman ES. Canine narcolepsy and the role of the nervous system. Compend Contin

Educ Pract Vet 1999;21:641–50.

[70] Overall KL. Dealing with dogs affected by separation anxiety. Vet Forum 2001;18:40–53.
[71] Iorio LC, Eisenstein N, Brody PE, et al. Effects of selected drugs on spontaneously

occurring abnormal behavior in beagles. Pharmacol Biochem Behav 1983;18:379–82.

[72] King JN, Maurer MP, Altmann BO, et al. Pharmacokinetics of clomipramine in dogs

following single-dose and repeated-dose oral administration. Am J Vet Res 2000a;61:80–5.

[73] King JN, Simpson BS, Overall KL, et al. Treatment of separation anxiety in dogs with

clomipramine: results from a prospective, randomized, double-blind, placebo-controlled,
parallel-group, multicenter clinical trial. Appl Anim Behav Sci 2000b;67:255–75.

[74] Hewson CJ, Conlon PD, Luescher UA, et al. The pharmacokinetics of clomipramine and

desmethylclomipramine in dogs: parameter estimates following a single oral dose and 28
consecutive daily oral doses of clomipramine. J Vet Pharmacol Ther 1998a;21:214–22.

400

B.S. Simpson, M.G. Papich

/ Vet Clin Small Anim 33 (2003) 365–404

[75] Hewson CJ, Luescher A, Parent JM, et al. Efﬁcacy of clomipramine in the treatment of

canine compulsive disorder. JAVMA 1998b;213:1760–6.

[76] Seksel K, Lindeman MJ. Use of clomipramine in treatment of obsessive-compulsive

disorder, separation anxiety and noise phobia in dogs: a preliminary, clinical study. Aust
Vet J 2001;79:252–6.

[77] Moon-Fanelli AA, Dodman NH. Description and development of compulsive tail chasing

in terriers and response to clomipramine treatment. JAVMA 1998;212:1252–7.

[78] Goldberger E, Rapoport JL. Canine acral lick dermatitis: response to the antiobsessional

drug clomipramine. J Am Anim Hosp Assoc 1991;27:179–82.

[79] Rapoport JL, Ryland DH, Kriete M. Drug treatment of canine acral lick: an animal

model of obsessive-compulsive disorder. Arch Gen Psychiatry 1992;49:517–21.

[80] Simpson BS. Canine separation anxiety. Compend Contin Educ Pract Vet 2000;22:328–39.
[81] Podberscek AL, Hsu Y, Serpell JA. Evaluation of clomipramine as an adjunct to

behavioural therapy in the treatment of separation-related problems in dogs. Vet Rec
1999;145:365–9.

[82] White MM, Neilson JC, Hart BL, Cliff KD. Effects of clomipramine hydrochloride on

dominance-related aggression in dogs. JAVMA 1999;215:1288–91.

[83] Dehasse J. Feline urine spraying. Appl Anim Behav Sci 1997;52:365–71.
[84] Mehlman PT, Higley JD, Faucher BA, et al. Low CSF 5-HIAA concentrations of several

aggression and impaired impulse control in nonhuman primates. Am J Psychiatry 1994;
151:1485–91.

[85] Reisner I, Mann JJ, Stanley M, Huang YY, Houpt KA. Comparison of cerebrospinal

ﬂuid monoamine metabolite levels in dominant-aggressive and non-aggressive dogs. Brain
Res 1996;714:57–64.

[86] Dodman NH, Donnelly R, Shuster L, et al. Use of ﬂuoxetine to treat dominance

aggression in dogs. JAVMA 1996a;209:1585–7.

[87] Overall KL. The role of pharmacotherapy in treating dogs with dominance aggression.

Veterinary Medicine 1999a;94:1049–55.

[88] Wynchank D, Berk M. Behavioural changes in dogs with acral lick dermatitis during a

2-month extension phase of ﬂuoxetine treatment. Hum Psychopharmacol Clin Exp
1998b;13:435–7.

[89] Poulsen EMB, Honeyman V, Valentine PA, et al. Use of ﬂuoxetine for the treatment of

stereotypical pacing behavior in a captive polar bear. JAVMA 1996;209:1470–4.

[90] Pryor PA, Hart BL, Cliff KD, Main MJ. Effects of a selective serotonin reuptake inhib-

itor on urine spraying behavior in cats. JAVMA 2001;219:1557–61.

[91] Romatowski J. Two cases of ﬂuoxetine-responsive behavior disorders in cats. Feline Pract

1998;26:14–15.

[92] Hartmann L. Cats as possible obsessive-compulsive disorder and medication models

[letter]. Am J Psychiatry 1995;152:8.

[93] Bruun R, Budman C. Paroxetine treatment of episodic rages associated with Tourette’s

disorder. J Clin Psychiatry 1998;59:581–4.

[94] Michelson D, Fava M, Amsterdam J, et al. Interruption of selective serotonin reuptake

inhibitor treatment. Br J Psychiatry 2000;176:363–8.

[95] Sheikh JI, Longbord P, Clary CM, et al. The efﬁcacy of sertraline in panic disorder:

combined results from two ﬁxed dose studies. Int Clin Psychopharmacol 2000;15:235–42.

[96] Howell SM, Fritz J, Downing S, et al. Treating chronic regurgitation behavior: a case

study. Lab Anim 1997;26:30–3.

[97] Larson ET, Summers CH. Serotonin reverses dominant social status. Behav Brain Res

2001;121:195–202.

[98] Pollock BG. Citalopram: a comprehensive review. Exp Omin Pharmacother 2001;2:681–98.
[99] Lass-Flo¨rl C, Dierich MP, Fuchs D, et al. Antifungal properties of selective serotonin

reuptake inhibitors against Aspergillus species in vitro. J Antimicrob Chemother 2001;
48:775–9.

401

B.S. Simpson, M.G. Papich

/ Vet Clin Small Anim 33 (2003) 365–404

[100] Golden RN, Dawkins K, Nicholas L, et al. Trazodone, nefazodone, bupropion, and

mirtazapine. In: Schatzberg AF, Nemeroff CB, editors. The American Psychiatric Press
textbook of psychopharmacology. 2nd edition. Washington, DC: American Psychiatric
Press; 1998. p. 251–69.

[101] Milgram NW, Ivy GO, Head E, et al. Effect of L-deprenyl on behavior, cognitive

function, and biogenic amines in the dog. Neurochem Res 1993;18:211–9.

[102] Ruehl WW, Bruyette DS, DePaoli A, et al. Canine cognitive dysfunction as a model for

human age-related cognitive decline, dementia, and Alzheimer’s disease: clinical
presentation, cognitive testing, pathology, and response to L-deprenyl therapy. Prog
Brain Res 1995;106:217–25.

[103] Cummings BJ, Head E, Afagh AJ, et al. b-amyloid accumulation correlates with cog-

nitive dysfunction in the aged canine. Neurobiol Learn Mem 1996;66:11–23.

[104] Calves S. Pharm proﬁle: selegiline. Compend Contin Educ Pract Vet 2000;22:204–5, 214.
[105] Cohn LA, Dodam JR, Szladovits B. Effects of selegiline, phenylpropanolamine, or a

combination of both on physiologic and behavioral variables in healthy dogs. Am J Vet
Res 2002;63:827–32.

[106] Crowell-Davis SL, Lappin M, Oliver JE, et al. Stimulus-responsive psychomotor epilepsy

in a Doberman pinscher. J Am Anim Hosp Assoc 1989;25:57–60.

[107] Dodman NH, Knowles KE, Shuster L, et al. Behavioral changes associated with

suspected complex partial seizures in Bull Terriers. JAVMA 1996b;208:688–91.

[108] Dodman NH, Miczed KA, Knowles K, et al. Phenobarbital-responsive episodic

dyscontrol (rage) in dogs. JAVMA 1992;201:1580–3.

[109] Brown SA, Crowell-Davis S, Malcolm T, et al. Naloxone-responsive compulsive tail

chasing in a dog. JAVMA 1987;190:884–6.

[110] Schwartz S. Carbamazepine in the control of aggressive behavior in cats. J Am Anim

Hosp Assoc 1994;30:515–9.

[111] Holland CT. Successful long term treatment of a dog with psychomotor seizures with

carbamazepine. Aust Vet J 1988;65:389–92.

[112] Kenny DE. Use of naltrexone for the treatment of psychogenically induced dermatoses in

ﬁve zoo animals. JAVMA 1994;205:1021–3.

[113] Dodman NH, Shuster L, White SD, et al. Use of narcotic antagonists to modify

stereotypic self-licking, self-chewing, and scratching behavior in dogs. JAVMA 1988;
193:815–9.

[114] White SD. Naltrexone for treatment of acral lick dermatitis in dogs. JAVMA 1990;

196:1073–6.

[115] McEwan BS, Davis PG, Parsons B, et al. The brain as a target for steroid hormone action.

Annu Rev Neurosci 1979;2:65–112.

[116] Borchelt PL, Voith VL. Dominance aggression in dogs. Compend Contin Educ Pract Vet

1986;8:36–44.

[117] Joby R, Jemmett JE, Miller ASH. The control of undesirable behaviour in male dogs

using megestrol acetate. J Small Anim Pract 1984;25:567–72.

[118] van den Broek AHM, O’Farrell V. Suppression of adrenocortical function in dogs

receiving therapeutic doses of megestrol acetate. J Small Anim Pract 1994;35:285–8.

[119] Henik RA, Olson PN, Rosychuk RAW. Progestogen therapy in cats. Compend Contin

Educ Pract Vet 1985;7:132–40.

[120] Marder AR. Diagnosing and treating aggression problems in cats. Veterinary Medicine

1993;88:736–42.

[121] Hart BL. Objectionable urine spraying and urine marking in the cat: evaluation of

progestin treatment in gonadectomized males and females. JAVMA 1980;177:529–33.

[122] Romatowski J. Use of megestrol acetate in cats. JAVMA 1989;194:700–2.
[123] Ashley PF, Frank LA, Schmeitzel LP, et al. Effect of oral melatonin administration on

sex hormone, prolactin, and thyroid hormone concentrations in dogs. JAVMA 1999;
215:1111–5.

402

B.S. Simpson, M.G. Papich

/ Vet Clin Small Anim 33 (2003) 365–404

[124] Graw P, Werth E, Krauchi K, et al. Early morning melatonin administration impairs

psychomotor vigilance. Behav Brain Res 2001;121:167–72.

[125] Williams DT, Mehl R, Yudofsky S, et al. The effect of propranolol on uncontrolled rage

outbursts in children and adolescents with organic brain dysfunction. J Am Acad Child
Psychiatry 1982;21:129–35.

[126] Shull-Selcer E, Stagg W. Advances in the understanding and treatment of noise phobias.

Vet Clin North Am Small Anim Pract 1991;21:353–67.

[127] Mohr N, Vythilingum B, Emsley R, et al. Quetiapine augmentation of serotonin

reuptake inhibitors in obsessive-compulsive disorder. Int Clin Psychopharmacol
2002;17:37–40.

[128] Ostroff RB, Nelson JC. Risperidone augmentation of selective serotonin reuptake

inhibitors in major depression. J Clin Psychiatry 1999;60:256–9.

[129] Lasher TA, Fleishaker JC, Steenwyk RC, et al. Pharmacokinetic pharmacodynamic

evaluation of the combined administration of alprazolam and ﬂuoxetine. Psychopharma-
cology (Berl) 1991;104:323–7.

[130] Grady TA, Pigott TA, L’Heureux F, Hill JL, Bernstein SE, Murphy DL. Double-blind

study of adjuvant buspirone for ﬂuoxetine-treated patients with obsessive-compulsive
behavior. Am J Psychiatry 1993;150:819–21.

[131] Overall KL. Animal behavior case of the month: treatment of a dog with interdog

aggression, stereotypic circling, and fearful avoidance of strangers. JAVMA 1995;
206:629–32.

[132] Walker R, Fisher J, Neville P. The treatment of phobias in the dog. Appl Anim Behav

Sci 1997;52:275–89.

[133] Altshuler LL, Bauer M, Frye MA, et al. Does thyroid supplementation accelerate tricyclic

antidepressant response? A review and meta-analysis of the literature. Am J Psychiatry
2001;158:1617–22.

[134] Stein MB, Sareen J, Hami S, et al. Pindolol potentiation of paroxetine for generalized

social phobia: a double-blind, placebo-controlled, crossover study. Am J Psychiatry
2001;158:1725–7.

[135] Papich MG, Davidson G. Unapproved use of drugs in small animals. In: Bonagura JD,

editor. Kirk’s current veterinary therapy XII: small animal practice. Philadelphia: WB
Saunders; 1995. p. 48–53.

[136] Simpson BS, Voith VL. Extralabel drug use in veterinary behavioral medicine. Compend

Contin Educ Pract Vet 1997;19:329–31.

[137] Baumgardner K. Aggressive dogs cause problems for owners, insurance companies.

DVM Magazine 1997;29:1, 28.

[138] Landsberg G, Hunthausen W, Ackerman L. Handbook of behaviour problems of the

dog and cat. Oxford: Butterworth Heinemann Publishers; 1997.

[139] Halip JW, Vaillancourt JP, Luescher UA. A descriptive study of 189 cats engaging in

inappropriate elimination behaviors. Feline Pract 1998;26:18–21.

[140] Mills DS, Simpson BS. Psychotropic agents. In: Horwitz DF, Mills DS, Heath S, editors.

BSAVA manual of canine and feline behavioral medicine. Gloucester, UK: BSAVA Press;
2002. p. 237–48.

[141] Davidson G. Evaluating transdermal medication forms for veterinary patients, part 1.

Int J Pharm Compounding 2001;5:214–5.

[142] Harvey A, Preskorn SH. Cytochrome P450 enzymes: interpretation of their inter-

action with selective serotonin reuptake inhibitors. J Clin Psychopharmacol 1996;16:
273–85.

[143] Munoz-Bellido JL, Munoz-Criado S, Garcia-Rodriguez JA. Antimicrobial activity of

psychotropic drugs selective serotonin reuptake inhibitors. Int J Antimicrob Agents
2000;14:177–80.

[144] Marder AR. Psychotropic drugs and behavioral therapy. Vet Clin North Am Small Anim

Pract 1991;21:329–42.

403

B.S. Simpson, M.G. Papich

/ Vet Clin Small Anim 33 (2003) 365–404

[145] Overall KL. Use of clomipramine to treat ritualistic stereotypic motor behavior in three

dogs. JAVMA 1994b;205:1733–41.

[146] Simpson BS, Davidson G, Letter to the editor: concerns about concurrent use of a

tricyclic antidepressant and an amitraz tick collar. JAVMA 1996;209:1380–1.

[147] Wynchank D, Berk M. Fluoxetine treatment of acral lick dermatitis in dogs: a placebo-

controlled randomized double blind trial. Depress Anxiety 1998a;8:21–3.

404

B.S. Simpson, M.G. Papich

/ Vet Clin Small Anim 33 (2003) 365–404

Dietary eﬀects on canine and

feline behavior

Katherine A. Houpt, VMD, PhD

Steven Zicker, DVM, PhD

Animal Behavior Clinic, College of Veterinary Medicine, Cornell University,

Ithaca, NY 14853–6401, USA

Hill’s Pet Nutrition, Inc, Topeka, KS, USA

There is considerable interest in the lay press and scientiﬁc publications

about the eﬀects of nutrition on behavior. This particular convergence of
ﬁelds is often diﬃcult to interpret for several reasons: (1) deﬁnitions of
behaviors are often not well characterized, and what constitutes a be-
havioral change is sometimes nebulous or subjective; (2) control of nutrient
intake is often diﬃcult to attain, especially in species that have access to
choice (eg, human beings or free-ranging dogs and cats); and (3) nutritional
inﬂuences may take place months or years from the manifested behavioral
deﬁcit (eg, neonatal iron deﬁciency eﬀects on cognitive performance in
children). Despite these challenges in interpretation of scientiﬁc or
observational data, many studies have been performed that can lend some
insight and credence to the hypothesis linking dietary nutrients and
alterations in behavior. In fact, many diagnoses of abnormal health status
begin as an observation of a behavioral attribute by either the owner or the
veterinarian. Thus, because it is well accepted that nutrition can aﬀect health
status, it is not unreasonable, with appropriate criteria, to deﬁne some nu-
tritional states that might aﬀect behavioral outcomes in pets. Conversely,
anecdotal reports of behavioral alterations attributable to certain nutrients
(hypotheses) may not be supported by well-controlled scientiﬁc experiments.
If appropriate leeway is given in the deﬁnition of behavior, it may be
discovered that there is a greater inﬂuence of nutrition on behavior than
commonly thought.

Failure to support may best be illustrated by a classic example from the

human nutrition arena, wherein the hypothesis states that excessive sugar in

Vet Clin Small Anim

33 (2003) 405–416

* Corresponding author.
E-mail address:

kah3@cornell.edu (K.A. Houpt).

0195-5616/03/$ - see front matter

PII: S 0 1 9 5 - 5 6 1 6 ( 0 2 ) 0 0 1 1 5 - 8

a child’s diet leads to hyperactivity (abnormal behavioral attribute). When
this hypothesis is subjected to appropriate scientiﬁc scrutiny, it does not
seem to be validated [1,2]. The cause for this scientiﬁc conclusion may be
either that the hypothesis is invalid or that the behaviors to be measured are
not easily quantitated and thus lead to a false-negative conclusion.

On the ‘‘pro’’ side of the argument in the veterinary ﬁeld, one might

suggest that physiologic functions, such as urination and defecation
(behavioral attributes), being either increased in frequency or altered in
location (eliminating in the house) may be considered undesirable behavior
by most owners. If the prior suggestion is considered as abnormal behavior,
it is possible that foods with increased water (canned versus dry) content
may alter these patterns either in frequency or volume. To stretch this idea
a bit further, it is known that feline house soiling may be aggravated by
crystalluria, which, in turn, can be inﬂuenced by dietary manipulation [3].
Certainly, one would agree that diabetes mellitus is characterized by the
behavioral attributes of polydipsia and polyuria; as such, a food higher in
ﬁber content may help to regulate glucose absorption and thus may mitigate
some of the adverse behaviors noticed by the owner [3]. In addition, it has
been shown recently that foods can aﬀect the metabolism and pharmaco-
kinetics of drugs that control epilepsy; as such, this may aﬀect a rather
noticeable behavior (seizure) frequency or duration [4]. If all these
possibilities are considered, an exhaustive list of interactions between
behavior and nutrition would be beyond the scope of this behavior. These
interactions of nutrition and behavior are suggested as introductory
statements to keep the reader aware and open to the possible complexities
of interactions and interpretations within these ﬁelds. In this article, we try to
illustrate some classic examples of overt deﬁciencies or excesses as well as
some possible examples of speciﬁc ingredient alterations in adequate but not
excessive combinations that may alter behaviors in dogs and cats. In some
instances, examples from other species may be employed, because insuﬃcient
data are present in the companion animal literature to make conclusions.

Energy balance

Alterations in energy balance may occur at any life stage, which may

result in adverse health. Insuﬃcient or poor-quality food intake may result
in decreased energy balance or protein energy malnutrition (PEM). PEM
has been categorized and subdivided into several categories, such as
marasmus (severe protein and energy intake deﬁcit), kwashiorkor (severe
protein and mild to moderate calorie intake deﬁcit), and short- and long-
term starvation (total calorie deprivation of varying lengths of time).
Conversely, excessive energy intake, without other nutritional deﬁciencies,
may lead to obesity if it is not balanced with increased exercise or energy
demands.

406

K.A. Houpt, S. Zicker / Vet Clin Small Anim 33 (2003) 405–416

Protein energy malnutrition during growth

A lack of calories or speciﬁc nutrients, such as protein, may aﬀect

behaviors, with the group at most risk being young animals. Malnutrition
can be via a direct eﬀect on the animal or via an indirect eﬀect caused by a
malnourished mother.

If a well-nourished mother is producing too little milk, her oﬀspring may

fail to thrive. This may not be apparent to an owner who is not experienced
with neonatal puppies or kittens. A good clue reﬂecting poor milk pro-
duction is that the dam spends more time with her oﬀspring. Apparently,
hungry neonates elicit more or longer nursing bouts than sated ones.
Possible stimuli for increased nursing are the puppy’s or kitten’s vocal-
izations, longer suckling time, or more intense treading or kneading of the
mammary tissue.

Long-term behavioral eﬀects of early malnutrition can be the result of

nutrition or social eﬀects caused by changes in the maternal-oﬀspring re-
lationship. Some of the problem behaviors, such as fear and aggression,
noted in feral cats may be the result of maternal malnutrition as well as lack
of socialization with human beings during the sensitive period from 2 to 7
weeks after birth. Although severe malnutrition leads to a suppression of
play, less severe food restriction of a queen leads her kittens to play more
[5]. Calorie- or protein-deprived dams may display inadequate maternal
behavior. When protein was deﬁcient in the mother’s diet during late
gestation and lactation, one of the main behavioral sequelae was increased
vocalization both when the mother was present and when she was removed.
This probably represents honest signaling by the kittens, whose needs are
greater than control kittens. Despite this, the protein-restricted mothers, in
contrast to the control mothers, did not immediately retrieve or nurse their
kittens after being separated for a few minutes [6]. It has also been suggested
that cannibalism of newborn kittens may possibly be related to some type of
maternal dietary deﬁciency, but this hypothesis has yet to be proven [7].

To control for the eﬀects of malnutrition on lactation and maternal

behavior, kittens whose mothers had been fed 45 g of Purina Cat Chow
(Purina, St. Louis, MO) per day, which is 50% of control intake, through-
out pregnancy were fostered onto nondeprived cats so that postnatal nutri-
tion would not be compromised. The kittens were able to compensate and
eventually reached the same body weight as control kittens, but they were
developmentally delayed on many physical and behavioral milestones. Their
eyes opened later, and they were signiﬁcantly delayed in walking, scratching,
playing, washing, lapping, eating, and climbing. Perhaps most important
from a clinical behaviorist’s perspective, they were delayed in using the litter
pan. As adults, these cats showed marked antisocial behavior, abnormal
posturing, and ﬂuctuating dominance-submissive behavior. The cats dis-
played less exploratory activity and more vocalization, urination, circling,
and clawing when alone in an open ﬁeld test. They spent less time with toys

407

K.A. Houpt, S. Zicker / Vet Clin Small Anim 33 (2003) 405–416

compared with controls. The malnourished cats made more errors in learn-
ing a maze and, particularly, in learning a reversal. What was most interest-
ing was that the second generation (these cats’ kittens) subjected to the same
maternal deprivation was also aﬀected but to a lesser degree [8].

Puppies of malnourished mothers were extremely nervous and resented

being handled as well as exhibiting physical abnormalities. Puppies fed a
protein-deﬁcient corn-based diet were also nervous [7].

Obesity during growth

In several studies, it has been shown that controlled food intake,

compared with ad libitum intake, in large- and giant-breed dogs decreased
the incidence of canine hip dysplasia [9]. This developmental orthopedic ab-
normality may be manifested to owners as an abnormal gait, pain, or
decreased locomotion compared with unaﬀected animals. To treat or
control excessive energy intake in these breeds, a reduction in caloric intake
may be achieved either by calculated dose or by decreasing the
energy density by the addition of noncaloric additives like ﬁber.

Obesity in adult dogs

Few of our pet dogs are malnourished, but many are overweight, with

an estimate of 24% to 30% in adult dogs [10]. Obesity in dogs can exac-
erbate musculoskeletal and cardiovascular problems and increase the risk
of diabetes. The eﬀects of dieting on canine behavior were investigated by
Crowell-Davis et al [11,12] using large-breed dogs (mostly Labrador
Retrievers) of both sexes and small-breed male dogs (Miniature Schnauzers)
group-housed in pens that included an outdoor run. During the ﬁrst day of
caloric restriction, the dogs barked more. The dogs were more active just
before meals when food was restricted. When total daily time budgets were
calculated, those on the greatest caloric restriction were more active on the
ﬁrst few days but later decreased their activity. Large-breed dogs that were
subjected to less caloric restriction showed a slight increase in active be-
haviors as a result of decreased sleep and increased sitting, standing, and
walking. Most interesting was that neither coprophagia nor aggression in-
creased. Owners of obese dogs should be reassured that barking will
decrease if and only if they do not reward the barking with food. They
should also be made aware that weight loss may be slowed, because the dogs
are compensating for less energy intake by lowering energy output.

Obesity in adult cats

The incidence of obesity in cats is approximately 30% [13]. Not many

studies have been performed on the eﬀects of obesity on adult cat behaviors,
much less on the eﬀects of dietary restriction. Obesity in cats is associated
with an increased incidence of urolithiasis and diabetes, however, which may
be manifested as inappropriate or frequent urination. Thus, resolution of

408

K.A. Houpt, S. Zicker / Vet Clin Small Anim 33 (2003) 405–416

these physiologic problems via dietary intervention may result in a return to
normal urination patterns.

Effect on canine aggression

An interesting eﬀect of food on behavior is the motivation of dogs

for diﬀerent foods. A hierarchy can be constructed based on the frequency
with which dogs display aggression over a particular food. Most dogs
aggressively defend rawhides. Next in attraction is any human food, bones,
and toys. A few dogs guard their ordinary ration of food, and fewer still
protect their water dish (Fig. 1).

A standard instruction to owners of aggressive dogs is to obtain

dominance over the dogs by eating before the dog is fed. Forcing the dog
to wait to be fed results in more food begging and general unruliness at
dinnertime. Furthermore, it seems highly unlikely that the dogs equate the
people seated at the table eating salad with a fork and drinking wine from
a glass with a pack of dogs vying for a bite from a carcass. In fact, Jagoe
and Serpell [14] found that dogs that were fed after their owners were more
likely to be aggressive to strangers.

There have been two studies directly testing the hypothesis that higher

protein diets lead to aggression. The ﬁrst compared scores for two types of
aggression and for hyperactivity. The types of aggression were territorial
and dominance. Owners scored the dogs on a 10-point scale, where 0 was no
aggression and 10 was uncontrollable aggression when strangers entered the
house (territorial) or when the dog bit, lunged at, or chased family mem-
bers, becoming worse when disciplined (dominance) in many circumstances.

Fig. 1. Objects guarded by dogs. Percentage of dogs (n

¼ 100) that guarded each type of

article. Table

¼ human food.

409

K.A. Houpt, S. Zicker / Vet Clin Small Anim 33 (2003) 405–416

The diets were 17%, 25%, and 32% protein, which corresponds to 2, 3, or
4 g of protein per kilogram of body weight, and were fed for 2 weeks. Fat
was substituted for protein in the lower protein diets. There was no signiﬁ-
cant eﬀect on dominance aggression and a trend toward decreased ag-
gression when the dogs were fed the higher protein diet. Aggression toward
strangers was signiﬁcantly greater when the dogs were fed the higher protein
diet. Within that group were dogs that were oﬀensively aggressive and
dogs that were defensively (or fearfully) aggressive. The fearful dogs improved
most when fed the lower protein diets. Control dogs were scored for the
same behavior and showed no change in score with diet [15].

In the second study, there were four diets, high (30%) and low (18%)

protein with or without 1.45 g of tryptophan per kilogram of diet added to
the basal diet of 0.15% tryptophan per kilogram. The signiﬁcant eﬀects were
that dominant aggressive dogs fed high-protein diets without tryptophan
were more aggressive than those fed the other three diets and that dogs fed
the low-protein diet plus tryptophan were less territorially aggressive. There
were no dietary eﬀects on fearfulness, hyperactivity, or excitability [16].
Diets formulated for dogs with renal disease and some commercial diets for
normal dogs are lower in protein than the low-protein diets tested and may
be more eﬀective. On the basis of the Dodman et al [15] and DeNapoli et al
[16] studies, lower (18%) protein diets should be recommended for
aggressive dogs. This is in contrast to earlier suggestions that high-protein
diets might improve behavior [17].

The reason why low-protein diets and tryptophan may reduce aggression

is that tryptophan is the precursor of serotonin; it is converted by trypto-
phan hydroxylase to 5-hydroxytryptophan, which, in turn, is converted
to 5-hydroxytryptamine or serotonin. Serotonin is a neurotransmitter as-
sociated with feelings of well-being and satiety. Tryptophan is found in low
concentration (\1%) in most protein sources. It must compete with other
large neutral amino acids for a common blood–brain barrier transporter
mechanism. Increasing dietary tryptophan increases brain serotonin. The
lower the protein level in the diet, the higher is the ratio of tryptophan to
large neutral amino acids and tryptophan transport to the brain.
Tryptophan has been shown to decrease aggression in chickens [18] and
primates [19]. Care must be taken when adding amino acids to a diet,
however, because an imbalance in amino acids and consequent anorexia
could result [20].

Fat

Fat and performance

The responses to a season of quail hunting by English Pointers fed two

diﬀerent diets were compared. The diets diﬀered in protein and fat. Diet A
was 13% fat and 23% protein, and Diet B was 49% protein and 28% fat.

410

K.A. Houpt, S. Zicker / Vet Clin Small Anim 33 (2003) 405–416

The dogs on the higher fat and protein diet (Diet B) lost less weight and
performed better in that they found more birds per hour [21] (Table 1).

As a corollary to this, it seems that the fat, or energy density of the food,

is the more important aspect of endurance performance. A classic ex-
periment by Downey et al [22] showed that a dog’s time to exhaustion was
directly correlated to fat intake but not to protein intake.

Fat and antioxidants

A decline in cognition with age is epitomized by the old saying, ‘‘You

can’t teach an old dog new tricks.’’ Recent ﬁndings of Neilson et al [23] show
that by the age of 11 years, 26% of dogs show one or more of the signs
of cognitive impairment. These signs include disorientation, lack of social
interaction with owners and other pets, loss of house training, nighttime
wakefulness, pacing, or vocalization. By the age of 15 years, 68% of dogs
show more than one of these signs [23]. Disorientation may take the form of
the dog’s trying to leave by the wrong door, getting lost in the yard, or
failing to extricate itself from under a table. Lack of anticipation of walks
and failure to greet owners can be a result of medical problems, such as
arthritis, cataracts, or deafness, but these symptoms also occur in the un-
impaired old dog.

The most serious problems as far as the owner is concerned are failure of

house training and nighttime misbehavior. Dogs may be polyuric secondary
to diabetes, Cushing disease, or renal failure. Any of these medical problems
can result in urination in the house, which the owner interprets as failure of
house training when it is really lack of opportunity to go outdoors.
Incontinence is another urinary tract problem that owners may blame on
failure of house training, because their older bitch leaves puddles of urine
wherever she has been sleeping. Incontinence is a word that lay people (and
some veterinary students) misuse as a synonym for urinating (voluntary
micturition) in the house rather than involuntary loss of urine because of
failure of sphincter control. Some dogs also defecate in the house, and
although this can be a consequence of constipation or other gastrointestinal
problems, this is less likely than in the case of urination. If a dog is both

Table 1
Eﬀect of diet on performance of English Pointers

Diet A

Diet B

Nutrient

Energy (kcal

/g)

4.7

5.9

Fat (%)

12.8

28.3

Protein

22.9

48.7

Performance

Time

103.7

136.1

Distance

15.5

20.4

411

K.A. Houpt, S. Zicker / Vet Clin Small Anim 33 (2003) 405–416

urinating and defecating in the house, it is more likely that it has forgotten
house-training skills.

Some people cannot tolerate urine or feces in their home, and the

oﬀending dog is euthanized. Other owners cover their ﬂoors with plastic or
conﬁne the dog on an easily cleaned surface. Few owners, particularly
working owners, are willing to tolerate sleep deprivation; therefore, the dog
who paces, vocalizes, or destroys things at night is more likely to be
presented to the veterinarian for treatment in the best case or euthanasia in
the worst case.

There are two treatments for cognitive dysfunction: one pharmacologic

and one dietary. The pharmacologic treatment is anipryl. Anipryl (Deprenyl)
or selegiline is a monoamine oxidase inhibitor. By inhibiting monoamine
oxidase, one can increase the amount of dopamine available to the post-
synaptic membrane, because dopamine is inactivated normally by the enzyme
monoamine oxidase as well as by reuptake into the presynaptic neuron.

In laboratory tests, old dogs fed an antioxidant-fortiﬁed food, with or

without social enrichment, displayed signiﬁcantly better problem-solving
tasks than controls after 6 months of intervention. The social enrichment
included walks with people, rotation of toys in the kennel, increased
problem-solving tasks, and a canine cage mate. The old dogs also displayed
cognitive impairment compared with a young dog group. Interestingly, the
food had no eﬀect on the cognitive ability of the young dogs compared with
controls [24].

In addition to the laboratory testing, a behavioral ﬁeld trial was

performed to assess the categories of signs of cognitive dysfunction, which
are (1) disorientation, (2) changes in sleep patterns, (3) changes in activity,
(4) changes in interactions with others, and (5) loss of house training.
Dogs older than 7 years of age that exhibited signs in two or more of
those categories were recruited for the study. Half of the dogs (n

¼ 64)

were fed a commercial dog food, and the other half (n

¼ 61) were fed a

fortiﬁed food (Hill’s Prescription Diet Canine b

/d, Hills Pet Nutrition,

Topeka, KS). The ingredients included vitamins E and C, docosahexanoic
(DHA) and eicosapentaenoic acid (EPA), lipoic acid, and carnitine. At the
end of 60 days, the owners reported that the supplemented dogs were
improved in all ﬁve categories, whereas the owners of the control dogs
reported that the control dogs were improved in two categories. The
supplemented dogs were improved in awareness of their surroundings,
family and animal recognition and interaction, and enthusiasm in greeting.
They circled less and soiled less in the house. They were more agile. The
control dogs and supplemented dogs both exhibited less aimless activity,
vocalized less, and slept more regularly. Overall, the control dogs
improved in 4 of 15 (27%) behaviors, whereas the experimental group
improved in 13 of 15 (87%). This is strong evidence that reducing produc-
tion of free radicals as well as neutralizing existing ones may improve the
behavior of older dogs [25].

412

K.A. Houpt, S. Zicker / Vet Clin Small Anim 33 (2003) 405–416

Vitamins, minerals, and amino acids

One has to read pre-1960 sources to ﬁnd descriptions of classic vitamin

deﬁciencies in dogs. Most of the signs are physical or general behavioral
signs of illness, such as inappetence and apathy [26]. As a more recent
example, cats suﬀering from taurine deﬁciency may develop central retinal
degeneration and cardiomyopathy. The cardiomyopathy produces charac-
teristic signs of malaise and inappetence. Supplementation of taurine to cats
with taurine deﬁciency resulted in improved attitude and appetite as noted
by owners [27].

Calcium

Many of us have been presented with a lactating bitch who is irritable,

hyperthermic, and possibly convulsing. Such a dog is probably suﬀering
from lactation tetany caused by hypocalcemia. Infusion of calcium results
in immediate recovery. There are few moments in practice that are as
rewarding as successful treatment of lactation tetany, because not only is the
bitch apparently completely cured but the owner is quite impressed at the
veterinarian’s skill. Of course, weaning or supplementation of the pups is
necessary, and oral calcium treatment of the bitch should be continued.

Lactation tetany is the most dramatic example of hypocalcemia. Less

profound hypocalcemia may result in irritability. Blood calcium levels
should be determined in aggressive cats or dogs, especially if the animal is
lactating, older, or ill and if the aggression is new or dramatically increased
in severity.

Conversely, excessive calcium intake in large- or giant-breed dogs may

result in the development of orthopedic disease [9]. This may be manifested
by stiﬀ gaits, pain and irritability, or inactivity. This disease risk factor must
be managed in conjunction with excessive energy intake in these same breeds
early after weaning so as to avoid permanent adult manifestations.

Cats with renal failure or those being administered acidifying diets

may display generalized muscle weakness, and in severe cases, persistent
ventroﬂexion of the neck. Muscle pain may be apparent as well. These
behavioral changes may be attributed to potassium depletion and the
resultant hypokalemic polymyopathy that ensues. Behavioral attributes may
be restored to normal with addition of potassium either parenterally or
enterally [28].

Fiber

Fiber is added to canine and feline diets to increase gastrointestinal ﬁll

while lowering caloric intake. Although lower ﬁber (1%–2.4% of total
dietary ﬁber) had no eﬀect on time spent eating a challenge meal [29], after
consuming diets of 12% or 21% ﬁber, dogs ate less of a meal oﬀered 30
minutes later than dogs who had eaten a meal of less than 2% ﬁber [30]. It

413

K.A. Houpt, S. Zicker / Vet Clin Small Anim 33 (2003) 405–416

would be interesting to survey owners of dieting dogs about whether
begging behavior or anticipation of the next meals diﬀered between dogs fed
high- and low-ﬁber diets.

One disadvantage of high-ﬁber diets is that stool volume is increased.

This can be an advantage in treating coprophagia, because dogs are less
likely to eat large soft feces than small hard ones. There are several
commercially available products to deter dogs from eating feces. Their eﬃ-
cacy has not been tested.

Fiber can also be used to treat pica in cats. There are two types of feline

pica: wool chewing by Siamese or other Oriental breeds and ingestion of
plastic, nylon, and a variety of other nonfood items by domestic short-
haired cats. Wool chewing seems to be more dietary than compulsive and
responds to addition of ﬁber such as a high-ﬁber prescription diet, a cat
garden, or access to grass outside. Some behaviorists have suggested meat
or raw chicken wings to give the cat more chewing time. Cats rarely chew on
rawhides or other dog chew toys. The other types of pica are more diﬃcult
to treat and may require the use of tricyclic antidepressants or selective
serotonin reuptake blockers.

Effect of dietary experience on food preferences

The eﬀect of early diet on subsequent food choice is interesting. Studies on

the impact of early diet on later feeding behavior were done long ago by Kuo
[31], who found that cats raised with a rat would not attack it or other rats.
Cats raised from birth as vegetarians (soybeans) or on a diet of mackerel and
rice would not consume novel foods at 6 months of age (but cats raised on
a variety of foods would eat novel foods). Chow-chow puppies were raised
from birth on soybeans or fruits and vegetables or on a variety of dairy and
meat products. The latter group was the only one that would sample novel
foods. This indicates that limited experience with diﬀerent foods can have a
permanent eﬀect on feeding behavior experience [31].

In experiments more relevant to most pet owners, Mugford [32] found

that cats or dogs fed one type of canned food from weaning until 6 months
of age would choose another novel ﬂavor at least initially when oﬀered both
in a two-choice preference test. With experience, over a few days, the ani-
mals might change their preference back to their original diet or maintain
a preference for the novel diet depending on the innate palatability of the
diet [32].

An aspect of diet that should be considered is the cognitive aspect. What

do animals learn to eat and learn to avoid eating? This subject is studied
more intensively in livestock than in companion animals. For example,
phosphorus-deﬁcient cattle select old not fresh bones or mineral phosphorus
[33].

Conditioned taste aversion results in the animal avoiding a food that it

associates with illness, particularly gastrointestinal illness. Unlike most

414

K.A. Houpt, S. Zicker / Vet Clin Small Anim 33 (2003) 405–416

forms of learning, the interval between the unconditioned stimulus, the taste
and smell of food, and the conditioned stimulus (a feeling of illness) can
be quite long, even hours. Use has been made of the phenomenon of
conditioned taste aversion to prevent canid predation on sheep. Pieces of
lamb were baited with lithium chloride, which produces nausea and vomit-
ing. Dogs that eat the bait learn to avoid that variety of meat [34], but it may
not prevent them from chasing live sheep.

We have tried to use the same principle to treat coprophagia in dogs by

placing apomorphine, the emetic, in the conjunctival sac of dogs
immediately after they eat feces; so far, this has not been successful whether
because the animals do not learn because they could safely eat feces so many
times before or because the owners are reluctant to administer the
apomorphine, we do not know.

References

[1] Krummel DA, Seligson FH, Guthrie HA. Hyperactivity: is candy causal? Crit Rev Food

Sci Nutr 1996;36:31–47.

[2] Bellisle F, Blundell JE, Dye L, Fantino M, Fern E, Fletcher RJ, et al. Functional food

science and behavior and psychological functions. Brit J Nutrition 1998;80(Suppl 1):
S173–93.

[3] Osborne CA, Kruger JM, Lulich JP, et al. Feline lower urinary tract diseases. In:

Ettinger SJ, Feldman EC, editors. Textbook of veterinary internal medicine. 5th edition.
Philadelphia: WB Saunders; 2000. p. 1710–47.

[4] Fettman MJ, Phillips RW. Dietary effects on drug metabolism. In: Hand MS, Thatcher

CD, Remillard RL, Roudebush P, editors. Small animal clinical nutrition. 4th edition.
Marceline (MO): Walsworth Publishing; 2000. p. 923–42.

[5] Martin P, Bateson P. The ontogeny of locomotor play behaviour in the domestic cat.

Anim Behav 1985;33:502–10.

[6] Gallo PV, Werboff J, Knox K. Protein restriction during gestation and lactation:

development of attachment behavior in cats. Behav Neural Biol 1980;29:216–23.

[7] Worden AN. Nutritional factors and abnormal behavior. In: Fox MW, editors. Abnormal

behavior in animals. Philadelphia: WB Saunders; 1968. p. 238–260.

[8] Simonson M. Effects of maternal malnourishment, development, and behavior in suc-

cessive generations in the rat and cat. In: Levitsky DA, editor. Malnutrition, environ-
ment, and behavior. New perspectives. Ithaca (NY):

Cornell University Press; 1979.

p. 133–148.

[9] Richardson DC, Zentek J, Hazewinkel HAW, et al. Developmental orthopedic disease of

dogs. In: Hand MS, Thatcher CD, Remillard RL, Roudebush P, editors. Small animal
clinical nutrition. 4th edition. Marceline (MO): Walsworth Publishing; 2000. p. 529–62.

[10] Burkholder WJ, Toll PW. Obesity. In: Hand MS, Thatcher CD, Remillard RL, Roudebush

P, editors. Small animal clinical nutrition. 4th edition. Marceline (MO): Walsworth Pub-
lishing; 2000. p. 401–30.

[11] Crowell-Davis SL, Barry K, Ballam JM, et al. The effect of caloric restriction on the

behavior of pen-housed dogs: transition from unrestricted to restricted diet. Appl Anim
Behav Sci 1995;43:27–41.

[12] Crowell-Davis SL, Barry K, Ballam JM, et al. The effect of caloric restriction on the

behavior of pen-housed dogs: transition from restriction to maintenance diets and long
term effects. Appl Anim Behav Sci 1995;43:43–61.

[13] Scarlett JM, Donoghue S, Saidla J. Overweight cats: prevalence and risk factors. Int

J Obes 1997;18(Suppl 1):S22–8.

415

K.A. Houpt, S. Zicker / Vet Clin Small Anim 33 (2003) 405–416

[14] Jagoe J, Serpell J. Owner characteristics and interactions and the prevalence of canine

behavior problems. Appl Anim Behav Sci 1996;47:31–42.

[15] Dodman NH, Reisner I, Shuster L, et al. Effect of dietary protein content on behavior in

dogs. JAVMA 1996;208:376–9.

[16] DeNapoli JS, Dodman NH, Shuster L, et al. Effect of dietary protein content and

tryptophan supplementation on dominance aggression, territorial aggression, and hyper-
activity in dogs. JAVMA 2000;217:504–508.

[17] Campbell WE. Behavior problems in dogs. Santa Barbara (CA): American Veterinary

Publications; 1975.

[18] Shea MM, Douglass LW, Mench JA. The interaction of dominance status and

supplemental tryptophan on aggression in Gallus domesticus males. Pharmacol Biochem
Behav 1991;38:587–91.

[19] Mehlman PT, Higley JD, Faucher I, et al. Low CSF 5-HIAA concentrations and se-

vere aggression and impaired impulse control in nonhuman primates. Am J Psychiatry
1994;151:1485–91.

[20] Gietzen DW, Duke CM, Hammer VA. Amino acid imbalance, a nutritional model:

serotonin

mediation of aversive responses. Physiol Behav 1991;49:981–5.

[21] Davenport GM, Kelley RL, Altom EK, Lepine AJ. Effect of diet on hunting performance

of English pointers. Vet Ther 2001;2:1–14.

[22] Downey RL, Kronfeld DS, Banta CA. Diet of beagles affects stamina. J Am Anim Hosp

Assoc 1980;6:273–7.

[23] Neilson JC, Hart BL, Cliff KD, et al. Prevalence of behavioral changes associated with age-

related cognitive impairment in dogs. JAVMA 2001;218:1787–91.

[24] Milgram NW, Head E, Cotman CW. The effects on anti-oxidant fortiﬁed food and

cognitive enrichment in dogs. In: Symposium on Brain Aging and Related Behavioral
Changes in Dogs, Orlando, January 11, 2002. Lenexa (KS): Veterinary Health Care
Communications; 2002. p. 31–3

[25] Dodds CE, Zicker SC, Lowry SR, et al. Effects of an investigational food on age-related

behavioral changes in dogs. In: Symposium on Brain Aging and Related Behavioral
Changes in Dogs, Orlando, January 11, 2002. Lenexa (KS): Veterinary Health Care
Communications; 2002. p. 34–6

[26] Mayer K. Nutritional deﬁciencies. In: Hoskins HP, Lacroix JV, Mayer K, editors.

Canine medicine. A text and reference work. Santa Barbara (CA): American Veterinary
Publications; 1953. p. 50–62.

[27] Pion PD, Kittleson MD, Skiles ML, et al. Dilated cardiomyopathy associated with

taurine deﬁciency in the domestic cat: relationship to diet and myocardial taurine content.
In: Lombardinin JB, Schaffer SW, Azuma J, editors. Taurine, nutritional value and
mechanism of action. New York: Plenum Press; 1992. p. 63–73.

[28] Podell M. Neurological manifestations of systemic disease. In: Ettinger SJ, Feldman EC,

editors. Textbook of veterinary internal medicine. 5th edition. Philadelphia: WB Saunders;
2000. p. 548–52.

[29] Butterwick RF, Markwell PJ, Thorne CJ. Effect of level and source of dietary ﬁber on

food intake in the dog. J Nutr 1994;124(Suppl):2695S–70S

[30] Jewell DE, Toll PW. Effects of ﬁber on food intake in dogs. Vet Clin Nutr 1996;3:115–8.
[31] Kuo Z-Y. The effects of food habits of parents on the young. In: The dynamics of behavior

development. An epigenetic view. New York: Random House; 1967. p. 66–68.

[32] Mugford RA. External inﬂuences on the feeding of carnivores. In: Kare MR, Maller O,

editors. The chemical senses and nutrition. New York: Academic Press; 1977. p. 25–50.

[33] Blair-West JR, Denton DA, McKinley MJ, et al. Behavioral and tissue responses to severe

phosphorus depletion in cattle. Am J Physiol (Regul Integr Comp Physiol) 1992;263:
R656–63.

[34] Hansen I, Bakken M, Braastad BO. Failure of LiCl-conditioned taste aversion to prevent

dogs from attacking sheep. Appl Anim Behav Sci 1997;54(2/3):251–6.

416

K.A. Houpt, S. Zicker / Vet Clin Small Anim 33 (2003) 405–416

Behavioral considerations in the

management of working dogs

Walter F. Burghardt, Jr, DVM, PhD

Behavioral Medicine and Military Working Dog Studies,

Department of Defense Military Working Dog Veterinary Service,

1219 Knight Street, Lackland Air Force Base, TX 78236-5519, USA

Arguably, ‘‘a dog is a dog’’; therefore, it should be true that ‘‘canine

behavior is canine behavior.’’ It is readily apparent even from casual
observation of canine behavior that a great degree of variation exists in what
are considered normal and abnormal behaviors as compared between
diﬀerent breeds and even between individual animals of the same breed [1].
The employment of dogs in working or performance settings entails a
different kind of management style and a different intensity of effort than
that required for a house pet, show dog, or laboratory animal. These
differences exist not only in the change in focus from individual patient to
‘‘herd health’’ but in almost every aspect of behavioral management of the
individual.

A focus on working or performance can alter the basic behavioral criteria

by which animals are selected for training or breeding, the ways in which
they are bred and raised, training techniques employed to produce desired
behaviors and eliminate undesired behaviors, the perspective used to
evaluate and deal with performance failure and behavioral problems, and
even the determination of behavioral criteria for continuing employment of
an animal.

Nonetheless, some behaviors that are problematic to pet owners are also

problematic in many working dogs. A candidate retriever that is overly
active, distractible, and inattentive has a behavioral problem as worthy of
attention as a pet dog that cannot learn its obedience commands or
successfully walk on a leash. Even an attack-trained military working dog

Vet Clin Small Anim

33 (2003) 417–446

E-mail address:

Walter.Burghardt@Lackland.af.mil

The views expressed in this article are those of the author and do not reﬂect the oﬃcial

policy or position of the Department of the Army, Department of Defense, or the US
Government.

0195-5616

/03/$ - see front matter

PII: S 0 1 9 5 - 5 6 1 6 ( 0 2 ) 0 0 1 3 3 - X

that bites its handler or another unintended victim is considered to have a
behavioral problem involving an unacceptable display of aggression toward
human beings, as would a house pet that nips at the ankles of strangers that
enter its house. A skilled helper dog that consistently cowers under furniture
during thunderstorms is at least as much of a problem as a pet that does the
same thing.

To be able to discuss the behavioral aspects of working dog management

in suﬃcient depth but without creating an unmanageable volume, we need
to limit our focus to a speciﬁc set of dogs and to choose areas of discussion
that are likely to produce the best information for the reader. Working dogs
can be broadly broken down into several subgroups, based, for instance, on
the American Kennel Club (AKC) groupings [2]. The current AKC breed
groups are sporting, herding, nonsporting, working, terrier, toy, and hound.

The traditional canine jobs reﬂected by these AKC groupings can be

further divided into more speciﬁc tasks that these groups represent. The
sporting group represents a collection of breeds that traditionally participate
in hunting and includes dogs that are used to ﬁnd, ﬂush, chase, and retrieve
game. The herding group includes breeds most commonly used to ‘‘control
the movement of other animals.’’ The working group includes those breeds
that are traditionally used for guarding, pulling, and other physically de-
manding tasks. The hound group includes dogs that are traditionally
employed to track or trail animals or people based on scent cues or to run
fast or for long distances, and the terrier group includes breeds that have
been traditionally employed to hunt small prey and pests.

Some employment opportunities for dogs do not conform to traditional

breed or group deﬁnitions. For instance, retriever and other breeds are
routinely used to locate and assist in the retrieval of lost persons and bodies,
consistent with the ‘‘retriever’’ name. Their skills have also been found to be
useful to assist persons with motor, visual, auditory, and other physical
disabilities, however, and these dogs have become quite popular for tasks
involving specialized substance detection tasks, especially in situations
where public contact requires identiﬁcation and acceptance of a working
dog and an animal with a good temperament and tractability.

Even nonsporting and toy breeds have found employment opportunities.

Their unique characteristics have proven quite useful in tasks that require
access to small areas or, as in the case of the Dalmatian, as the traditional
dog for mingling with, keeping up with, and protecting horses and wagons
in transit.

The largest employers of working dogs are government and helper dog

programs. The US Department of Defense employs approximately 1500
dogs that are used to protect military personnel and equipment from
physical attack and to detect threats to safety and security, such as hidden
explosives and illegal drugs, tasks for which these dogs have displayed
excellent aptitude and eﬃcacy (for historical examples of military working
dog employment and types of tasking, the reader is referred to the article by

418

W.F. Burghardt

/ Vet Clin Small Anim 33 (2003) 417–446

Lemish [3]). Most US military dogs are dually trained, meaning that they are
capable of protection tasks through controlled aggression toward human
beings and proﬁcient in either explosive or drug detection tasks. Some
German and other non-US police departments (eg, the Nordrhein-
Westfalen State program) can employ in excess of 500 dogs [4]. Because
of a greater degree of local organization (municipality versus state), US law
enforcement programs are much smaller, but several large programs do exist
outside of the military, such as the Connecticut State Police program [5].
Other US federal programs include the Transportation Security Admin-
istration (TSA) Explosives Detection Canines, Customs Service Drug
Detector Dogs, Border Patrol Canine Program, and Department of
Agriculture contraband-detecting ‘‘Beagle Brigade’’ [6].

Many helper dog programs exist in the United States and elsewhere in

the world. Large-scale US programs include Seeing Eye, Guide Dog, and
Canine Companions for Independence. These programs focus primarily on
the use of dogs to allow individuals to live more independently by training
and using dogs to perform tasks that the individuals cannot, such as reach-
ing and grasping and even mobility assistance for individuals who have
limited mobility, providing visual feedback and guiding movement for blind
persons, and alerting and orienting to auditory cues for individuals who
have impaired hearing. Dogs are also used as a treatment modality in hos-
pitals and other settings [7,8]. Helper dogs often produce beneﬁts to people
that are not produced by speciﬁc training, such as increasing opportunities
for social interaction in persons conﬁned to wheelchairs [9] and provid-
ing positive emotional support and companionship in hearing-impaired
owners [10].

The organization and philosophy of working dog programs vary

tremendously. Some programs, such as the US Federal Emergency
Management Agency (FEMA) Search and Rescue Dog eﬀort, consist of a
relatively loose-knit network of volunteer people and dogs that may (or may
not) be available for emergency response. These volunteers usually acquire,
train, and maintain their own animals [11]. Other programs, including many
municipal police programs, are organized locally for local requirements and
are loosely afﬁliated with other programs though professional associations
(eg, National Narcotic Detector Dog Association, United States Police
Canine Association). In these programs, the municipality usually acquires
trained or untrained adult dogs on an ‘‘as needed’’ basis to match with a
speciﬁc canine handler. Still other programs, such as the Department of
Defense, are designed to acquire and train large numbers of suitable adult
dogs procured from nongovernment sources while possessing all the
resources and personnel required for training, certiﬁcation, medical care,
and other functions to sustain the program. Finally, some programs, such as
the Australian Customs Service [12] and Seeing Eye [13], operate using
selective breeding models that allow them to produce a reliable supply of
candidate dogs to match their needs.

419

W.F. Burghardt

/ Vet Clin Small Anim 33 (2003) 417–446

This thumbnail sketch of the varieties of governmental and public

working dog programs is far from complete. Program development activities
and resulting changes in strategy are the hallmarks of vibrant programs.
Currently, the US Department of Defense, Transportation Security
Administration, and Customs Service are all in the process of implementing
and evaluating selective breeding models to produce candidate dogs for their
programs [14]. Likewise, diverse programs from around the world are
collaborating on an increasingly frequent basis on matters of veterinary care
(eg, diagnosis and control of hip dysplasia [15,16]) and program manage-
ment [17,18] and in the application of scientiﬁc research to program
execution (eg, use of estrus-inducing drugs [19] or evaluation of tempera-
ment testing in candidate dog selection [20]). An example of a forum for
such innovation and collaboration is the International Working Dog
Breeding Conference, which was held in San Antonio in 1999 and 2001.

For the purposes of the rest of this article, ‘‘working dogs’’ refers most

speciﬁcally to those dogs employed in large-scale governmental and public
programs, where the dogs work in tasks related to protection and controlled
aggression, substance detection, or provision of assistance to persons with
disabilities (especially the US Department of Defense Military Working
Dog Program). Dogs employed for sporting, running or pulling, and
herding tasks and programs for the breeding and development of these dogs
are less intensively addressed.

Behavioral assessment in working dogs

Behavioral assessment in working dogs is a key component in their

employment. Assessment tools of one sort or another have been and are still
being developed and applied to almost every aspect of a working dog’s
behavior. Testing in the working dog world is clearly quite diﬀerent from the
level of behavioral testing used in pet practice. For instance, instruments
have been developed to determine the suitability of an animal for training
[12,20] and to determine the level of proﬁciency in performing critical tasks
(eg, United States Police Canine Association Certiﬁcation Tests). Tests have
also been developed to predict how well an animal comports itself around
other animals and around human beings (eg, Delta Society Pet Partners
Team Evaluation [21]). Other instruments are beginning to be used to
evaluate the presence of potentially heritable behavioral pathologic
characteristics (eg, the use of the lactate response test to identify excessive
fearfulness) [22] or to determine possible limitations of a dog’s learned
behavior or performance [23].

Assessment of working dog behavior usually begins when a juvenile or

adult candidate is considered for training or when puppies that are
purposefully bred for a particular use are evaluated for the presence of the
required behavioral characteristics. These behavioral assessments attempt to

420

W.F. Burghardt

/ Vet Clin Small Anim 33 (2003) 417–446

evaluate stabile behavioral characteristics that might be referred to as
‘‘temperament’’ and ‘‘aptitude.’’

Measures of temperament attempt to predict the likelihood of speciﬁc

social interactions between a dog and people or other animals. In pet
practice, the objective is usually to select animals that display inquisitive and
solicitous interaction with people and to select against animals that are
likely to display ﬂight, ambivalence, apprehension, asociability, or ag-
gression toward others [1]. In working dogs, the social interactions desired
require more deﬁnition and exhibit some difference from those desired for
pet dogs. For instance, words used to deﬁne social interactions in guide dogs
could include calm, observant, and tolerant. Conversely, in military or
police dogs, the deﬁnition might include inquisitive, conﬁdent, and
dominant.

A problem immediately arises in making such temperament assessments.

The terms we often use are qualitative. Also, these terms often possess a high
degree of connotation and anthropomorphism. In other words, a term often
alludes to an unmeasurable, ‘‘internal state’’ of the animal, and such terms
are often bound heavily to our understanding of normal (and abnormal)
human behavior and are also based on personal experience.

An example of such a diﬃculty in the use of behavioral constructs is the

use of a term such as fearful. Most people would agree that it is not desirable
for a pet or working dog to demonstrate a large degree of fearfulness. My
understanding of ‘‘fearful’’ clearly differs from that of my 6-year old son or
my bungee-jumping colleague at work (an illustration of a situation that
might have wide interrater reliability problems). In addition, we have no
assurance that any human perception of ‘‘fear’’ (or a situation that might
reliably cause fear in a person) is reﬂective of a similar state in a dog
experiencing the same stimulus milieu (representing a potential problem in
the validity of a measure).

A full discussion of temperament test development issues is beyond the

scope of the present article, but an excellent review of the primary issues
may be found in an article by Goodloe [24]. To summarize the challenges we
face in attempting to measure internal states or social proclivities in dogs,
the challenge is to produce a test that actually predicts how an animal would
behave in a situation of importance to us and to develop measures that are
repeatable from time to time for a single observer and also repeatable
between different observers.

To get around these potential pitfalls in making predictions about future

behavior, the radical behaviorist would reject the use of temperament and
temperament-related terminology completely, relying only on the measure-
ment of observable desirable and undesirable behaviors and of relations
between stimulus events and these behaviors exclusively. Arguably, this
approach provides the strongest predictive model of future behavior
(indeed, volumes by Skinner and others have been dedicated to the
prediction and control of behavior using only observable behaviors, stimuli,

421

W.F. Burghardt

/ Vet Clin Small Anim 33 (2003) 417–446

and consequences [25]), but this approach also suffers from an inability to
predict future behavior when only slight changes occur in the behavioral
equation, as was illustrated with a bit of humor by the Brelands [26] (who
were Skinner’s students) in their work with applied operant conditioning of
performing animals.

Fortunately, a workable middle ground exists to develop tests to measure

temperament characteristics that could be important in predicting successful
working dog training and behavior. One approach is to identify physiologic
correlates or overt and easily measurable overt behavioral responses that
accompany a less observable behavioral state, such as using the lactate test
described by Overall et al [22] to identify fearful animals.

Another strategy is to measure a number of observable behaviors in

various settings that may or may not be related to a presumed behavioral
characteristic and then to let statistical tools determine relations between the
measured behaviors themselves and their usefulness in predicting other
behaviors in the future. This kind of work initially saw its heyday in the
Jackson Laboratory at Bar Harbor, Maine, in the work of Scott and Fuller
[27] and their colleagues. It is interesting to note that the authors relied
heavily on the use of temperament terminology, but related the terminology
directly to carefully controlled experiments, where they measured observ-
able behavior, deriving mostly quantitative measures, such as response
latency, social interaction counts, and the like. Another interesting ob-
servation is that this line of research recruited investigators such as Anastasi
[28], Pfaffenberger [29], and Cattell [30], whose work would later play a
central role in the development of human test theory and educational and
clinical assessments.

Only recently has there been a sustained resurgence of interest in

subjecting canine performance and temperament tests to factorial analysis
and other statistical methods to ensure the reliability of these instruments
and items in them and to determine whether the measures are valid
predictors of future outcomes. A novel application of factorial design was
recently reported by Serpell and Hsu [31]. These authors prepared a
semantic differential behavior rating scale instrument in the form of a
questionnaire that they administered to volunteer puppy raisers of pro-
spective 1-year-old guide dogs, from which they extracted reliable mea-
sures, identiﬁed 8 common factors among their 40 items, and then
successfully validated these factors against the school’s criteria for accepting
or rejecting candidates. These results are different from strict performance
tests, because the authors relied on puppy-raiser opinions of their animal’s
behavior; they also differ from the work of Hart and Hart [1], because the
Serpell and Hsu study [31] relied on owner observation of individual animals
rather than on human assumptions about a breed-speciﬁc stereotype.
In addition, Serpell and Hsu [31] validated their results against actual
outcomes (accept or reject a candidate dog). In her book entitled Canine
Behavior

, Beaver [32] reviewed the work of Scott and Fuller and their

422

W.F. Burghardt

/ Vet Clin Small Anim 33 (2003) 417–446

attempt to unravel the behavioral genetics of the dog in a section on the
diagnosis of behavioral problems and lamented that, ‘‘Expense alone
probably means a similar effort will never occur again.’’ If recent interest in
selective breeding, quantitative genetics, and assessment continues, Beaver’s
prediction may be proven wrong.

Assessment of aptitude and performance in working dogs may prove to

be more easily addressed than that of temperament, because these constructs
are usually measured by evaluating the performance of a task in a deﬁned
setting, either with or without training. The three signiﬁcant issues in
measuring aptitude for future learning or performance are (1) the need to
ensure that the behavior on the aptitude test actually has a signiﬁcant
predictive relation to performance on a meaningful task later, (2) the results
of the task are not overly aﬀected by learning, and (3) the tests performed
and the measures obtained are repeatable both within and between animals
and observers (either with or without modiﬁcation of the test for age). These
issues are addressed by Hilliard and Burghardt [14] in their article on the
development of performance tests to be used at different ages in growing
candidate working dogs. The authors used a combination of temperament
assessments conducted by trained observers and structured performance
tests, distilling the data using factor analysis and then validating the results
against each candidate’s success or failure in training. The authors con-
cluded that the predictive value of tests rapidly increases until at least 6
months of age in Belgian Malinois Dogs, which is produced by selective
breeding in their program. Beaudet et al [33] also address the value of
behavioral tests and activity level measures performed in puppies and
repeated at various ages in predicting dominance behavior later in life. They
also conclude that the value of the tests increases with age. As tests are
developed further, it would be ideal for a working dog breeding program to
be able to discriminate between successful and unsuccessful candidates at
younger ages rather than at the traditional 12 to 15 months of age, because
raising a puppy becomes progressively more resource-intensive as the puppy
grows and training time is particularly precious. It should be noted that
puppies unsuitable for training usually ﬁnd other forms of employment or
are otherwise adopted as pets.

A ﬁnal issue to address in this section on behavioral assessment in

working dogs is the assessment of limits to performance. The discussion that
follows showcases two unpublished projects performed at the Department
of Defense Military Working Dog Veterinary Service by Hilliard and
Burghardt.

The ﬁrst of these studies asked a clinical question regarding older military

working dogs. A number of authors have published reports describing the
clinical presentation, performance deﬁcits identiﬁed in laboratory settings,
underlying pathologic ﬁndings identiﬁed in the brains of aﬀected animals,
and treatment of behavioral decrements in pet and laboratory dogs, deﬁned
now as canine cognitive dysfunction (CCD) [34]. The question arose as to

423

W.F. Burghardt

/ Vet Clin Small Anim 33 (2003) 417–446

whether CCD could be demonstrated in military working dogs and whether
the disease affected performance in critical tasks. The gold standard for
measuring CCD in dogs behaviorally involved asking the dogs to solve a
task in which they were asked to respond to the correct object in an object
presentation task when the correct and incorrect objects were reversed and
after varying periods of delay. The apparatus used to assess performance is
called the Toronto General Testing Apparatus (TGTA). Performance on the
TGTA tasks is measured in errors to criterion performance; lower numbers
of errors indicate faster acquisition of a task. Fig. 1 represents data provided
from Milgram’s laboratory in Toronto in a group of normal subjects and
data in a group of normal military working dogs at Lackland Air Force
Base in Texas performed as a replication. Both sets of data are fairly similar,
indicating that the military working dogs tested were not greatly dissimilar
in their ability to perform this task compared with laboratory Beagles.

Fig. 2 represents data from two aged working dogs at Lackland (Tara,

aged 10 years, and Toots, aged 12 years) compared with the TGTA normals.
By deﬁnition, Tara is impaired on the task and would be diagnosed with
CCD using this standard. It is interesting that Tara continued to work
superbly in a special project as a substance detection dog well beyond the
time this test was administered. Although the prospect of identifying and
treating CCD in military working dogs seems appealing as a means of
improving the quality of a patient’s life and possibly extending that dog’s
useful service life, the TGTA testing takes months to complete. To make
CCD more identiﬁable to handlers and veterinarians working with military
working dogs, the next logical step is to attempt to identify behavioral and
performance problems in these dogs at an early stage by means of a quick

Fig. 1. Performance expressed as number of errors to criterion for Beagles (Toronto) and
Military Working Dogs (Lackland) plus or minus SD. Toronto General Testing Apparatus
tasks are matching to sample (MTS), reversal (REV), and delayed matching to sample (10, 20,
and 30 seconds). Pooled points are for all delays.

424

W.F. Burghardt

/ Vet Clin Small Anim 33 (2003) 417–446

screening assessment so that maximum beneﬁt might be obtained from
treatment.

The ﬁnal issue addressed involves an unpublished project in which we

attempted to answer some questions regarding the olfactory ability of
military working dogs. Using a pure chemical, we attempted to produce
training and testing objects that reliably produced incrementally smaller
amounts of that pure substance when placed in a cardboard sampling
container. That task in and of itself was not simple and required the services
of an analytical chemist and some rather sophisticated sampling instru-
mentation. Fig. 3 shows the actual amounts of substance available for
detection in sampling boxes for each of the four levels of aid presentation.

The behavioral issue, however, involves the questions that were asked.

The ﬁrst question was, ‘‘What is the smallest available amount that a dog
can reliably detect?’’ The second question was, ‘‘How speciﬁc is the

Fig. 2. Performance expressed as number of errors to criterion for Beagles (Toronto) and two
aged working dogs, Tara (spayed female Labrador Retriever, aged 10 years) and Toots (spayed
female Labrador Retriever, aged 12 years). Legends are the same as for Fig. 1.

Fig. 3. Amount of training substance available in the detection sampling box as measured by
polyurethane ﬁlter cartridge collection and high performance liquid chromatography assay.

425

W.F. Burghardt

/ Vet Clin Small Anim 33 (2003) 417–446

detection response for the target substance?’’ In this project, the dogs were
trained to sample a series of sampling boxes in a square array, with random
re-entry into the array on each trial. This conﬁguration was chosen because
dogs had earlier demonstrated marked place and order biases in their errors.
The dogs were asked to sniﬀ each container and then to sit if they had
encountered the target substance. Varying numbers of containers contained
the target material; other boxes were empty or contained ‘‘vehicle’’ (ﬁlter
paper that normally contained the training substance but without any target
substance on it) or a distracter (the same paper with a diﬀerent substance
placed on it). During training, the dogs reliably learned to detect even the
lowest amount of substance present. During testing with the lowest level of
target substance and with distracters and blanks, an operational problem
was encountered, however. Some animals began to make mistakes, either
ignoring a target (a ‘‘miss’’) or performing their sit response after sampling a
box with no target material present (a ‘‘false alarm’’). A quandary arose
as to whether to reward a dog for all sit responses (whether correct or
incorrect) or to withhold reward during all test trials. A problem exists with
either of these approaches in that the results can be confounded, because
these procedures can produce new learning. If all sit responses were to be
rewarded during the test, an animal would be reinforced for incorrect as
well as correct responses. If results then showed that an animal did not dis-
criminate well between target and nontarget materials, the results might
have occurred because incorrect responses were rewarded (ie, responses in
the presence of the nontarget substance may occur more frequently because
they were being rewarded). Likewise, if reward was withheld during all test
trials, correct responses would be in extinction. Poor performance might
then be attributed to extinction of the previously rewarded correct response.

The solution to this problem was to perform testing in partial extinction.

This was accomplished by decreasing the frequency of rewarding correct
responses from 100% to 25% during the ﬁnal phase of training.
Unrewarded correct trials resulted in no response from the handler and
no reward. During this phase of training, the handler was instructed simply
to wait for a brief period after an incorrect response identiﬁed by the data
recorder and then to move to the next sample location. During testing, the
same lowered density of reward was used, but the rewards were given on
predetermined sit responses independently of whether the response was
correct or incorrect. In this way, correct and incorrect responses had an
equal probability of being rewarded. This procedure also removed any
potential unintentional cues that the handlers might provide if they had been
aware of the correct or incorrect nature of an animal’s response. The results
of the test for discrimination between the test substance, blanks, vehicles,
and distracters are shown in Fig. 4. The results of this project strongly
suggest that the dogs formed an excellent discrimination between the target
materials, empty sampling boxes, vehicle paper, and vehicle paper with an
added distracting substance.

426

W.F. Burghardt

/ Vet Clin Small Anim 33 (2003) 417–446

Taxonomy of behavioral pathologic ﬁndings

Several authors [35,36] have developed fairly in-depth taxonomies of

canine behavioral problems. Most of these inventories are structured in a
functional or ‘‘systems’’ manner. Thus, there are ‘‘ingestive’’ disorders,
sexually dimorphic problems, problems of elimination, fearful behaviors,
aggressive disorders, and the like. Odendaal [37] departs from the systems
approach and suggests a classiﬁcation based on etiology of a behavior,
which is somewhat less dependent on the overt behavioral presentation. In
this etiologic approach, behavioral problems are seen less as speciﬁc ‘‘kinds’’
of behavior and more as the results of genetic predisposition, developmental
differences, normal variation between animals, differing kinds of social
environments and cues, medical disease, variation in the ability to adapt to
change, or some combination of these underlying causative agents.

Taxonomies of one sort or another serve a really useful purpose over and

above providing pigeon holes in which to place a particular animal’s
problem or problems based on cause or appearance. One useful reason to
classify behavioral problems in some functional manner is that it focuses the
veterinary clinician on the general health of a patient and suggests possible
dysfunctions in the physiology underlying the body system associated with
the behavioral problem. For example, a veterinarian evaluating a problem
of fecal house soiling is likely to perform a thorough workup of the
gastrointestinal tract as part of the basic examination because of this focus
[38]. Likewise, a practitioner dealing with a problem of mounting behavior

Fig. 4. Sit responses performed at sampling boxes containing previously trained pure substance
on chromatography paper (Trained), a substance for which the subject had received no previous
training (Distractor) also on chromatography paper, chromatography paper with no additional
substance (Vehicle), or an empty sample box containing no substance or chromatography paper
(Empty). Responding is recorded as sit responses as a percentage of the opportunity to sit at a
station.

427

W.F. Burghardt

/ Vet Clin Small Anim 33 (2003) 417–446

in a presumably neutered male dog is likely to perform a thorough re-
productive tract examination and may expend extra effort to ensure that
the patient does not have normal or retained testicles that could contribute
to the problem [39].

Another reason why a taxonomic approach is useful in evaluating

behavioral problems is that it allows for comparison between clinical pre-
sentations based on the presence or absence of particular behavioral and
medical signs. This approach is often the way that ‘‘syndromes’’ are deﬁned.
The rationale is that similar problems may represent similar underlying
pathologic characteristics and may therefore respond to similar therapy. For
example, fearful behaviors, despite some variance in behavioral topography,
produce a recognizable set of physical signs that are assumed to have a
common underlying physiology. Because of this common physiologic
response, it is reasonable to predict that behavioral problems deﬁned
as fearful are likely to beneﬁt from the use of desensitization

/counter-

ounterconditioning or benzodiazepine anxiolytics [40].

When addressing behavioral problems in working dogs, the need for an

additional system of classiﬁcation arises. Functional or system-related
classiﬁcations still retain their beneﬁt in diagnosis and treatment, but three
new issues arise. First, some common behavioral problems of pet dogs are
not of concern in a working dog. An example would be that house soiling is
not a behavior of concern in an outdoor working dog that is kenneled when
not working. Second, some behavioral problems that occur in working dogs
are not seen in pets. For instance, police dogs might suﬀer from problems
involving underaggression (rather than overaggression, which would be a
problem in a pet dog). Finally, by nature, working dogs are employed for
speciﬁc behavioral characteristics or tasks that they perform. Especially as
the complexity of a working dog program and required tasks increases,
behavioral problems must be viewed in light of whether the problem
interferes with the successful employment of an animal in the tasks and
settings for which it is maintained.

A classiﬁcation scheme for behavioral problems in working dogs that

accounts for these factors reclassiﬁes behavioral problems into four
categories for the purpose of eﬀectively dealing with case management.
The ﬁrst category includes those behavioral characteristics or problems that
interfere with the acquisition of a required task. These kinds of problems
might be called ‘‘learning-related problems.’’ The second category includes
behavioral issues that directly interfere with the performance of a previously
proﬁcient task. These kinds of problems might be considered as ‘‘perform-
ance problems.’’ The third category is composed of behaviors that do not
directly interfere with a required task but that indirectly degrade task
performance because of their presence. These problems might be referred
to as ‘‘disruptive problems.’’ The ﬁnal category is composed of behavioral
problems or characteristics that do not aﬀect the acquisition or performance
of required tasks but constitute a danger to the well-being of the patient,

428

W.F. Burghardt

/ Vet Clin Small Anim 33 (2003) 417–446

other animals, or human beings and might be referred to as ‘‘husbandry
problems.’’

Learning-related problems could run the gamut from medical and

anatomic diﬃculties that produce untoward eﬀects on learning to problems
of aptitude or temperament. Some obvious examples of these kinds of
medical problems with behavioral signs are anosmia, lissencephaly, blind-
ness, and the like [41]. Other less obvious problems might include difﬁculty
in acquiring and maintaining an animal’s attention such that it cannot
successfully practice and be rewarded for correct performance of a be-
havioral task, inability to entice an animal to perform a required be-
havior using available reinforcers, or an apparent lack of aptitude (either
physically or ‘‘intellectually’’) for performing a particular task. For the
purposes of case management, it should be noted that these kinds of
problems are most frequently encountered during animal selection or initial
training and rarely thereafter. In addition, more often than not, these kinds
of problems should result in the elimination of a candidate animal from
consideration or training, because substantive changes in basic temperament
and aptitude can rarely be expected [24].

Performance problems would be expected to occur in dogs that had

successfully acquired and performed a required behavior at some time but
that now have become unable or unwilling to perform that behavior. These
types of behavioral problems might occur in dogs that have temporary or
permanent adverse changes in sensory ability or other physical problems,
such as a detector dog that is unable to ﬁnd a target item by smell because of
a viral rhinitis, a patient demonstrating a decrement in task performance
secondary to the administration of a medication, or an animal no longer
able to meet the physical demands of its tasking as a helper dog because of
peripheral neurodegenerative disease or arthritis. These types of problems
might also occur in animals that display an apparent loss of learned task
performance in association with signs of a central neurodegenerative
disorder, such as CCD [34]. Unlike learning-related problems, performance
problems are more likely to be associated with a medical condition or
behavioral problem that is at least theoretically treatable. Case management
would be focused on identifying the speciﬁc cause of performance failure
through a thorough medical evaluation and a behavioral task analysis and
on resolving or accommodating for the causative condition or conditions.
Behavioral management is often accomplished through a remedial training
plan.

Disruptive problems interrupt the ongoing performance of a required

behavior (rather than aﬀecting the underlying ability to perform a behavior)
and often involve a behavior other than that which is usually related to the
successful performance of the task. A simple example of a disruptive
behavioral problem is the situation where a search and rescue dog becomes
distracted by and begins to chase a wild animal during a search, eﬀectively
preventing successful completion of a search pattern. Another example is

429

W.F. Burghardt

/ Vet Clin Small Anim 33 (2003) 417–446

a drug or explosives detection dog that becomes distracted by a food
item during a search, resulting in an incorrect response. These disrup-
tive behavioral problems also include behaviors that are truly considered
behavioral pathologic ﬁndings and are completely unrelated to the task
at hand. Examples include repetitive tail-chasing in a helper dog and
(apparently) untriggered aggression toward human beings in a detector dog
searching luggage. Rather than task analysis and retraining, these cases
usually beneﬁt more from speciﬁcally addressing the disrupting behavioral
problem and successfully managing it.

Husbandry problems include behaviors like generalized overactivity

intense and persistent enough to interfere with the ability of an animal to
maintain a healthy body weight or a patient that engages in repetitive
spinning when not performing a required task and subsequently injures
itself. Included in this category are also behaviors that are seen as problems
in pets, such as attempts to escape from a location or secondary self-injury
during a thunderstorm or undesired aggressiveness directed toward people
or other animals in social nonwork settings (eg, on walks, while grooming,
during other routine handling). Like disruptive behaviors, husbandry prob-
lems are best managed by directly addressing and managing the problem
behaviors, although husbandry problems often provide the practitioner
with more time for case management, because the problems do not directly
disrupt task performance.

This classiﬁcation and discussion of behavioral problems in working

dogs focuses heavily on problems associated with the acquisition and per-
formance of task-related behaviors and on the safety of people and other
animals exposed to the working dog. The classiﬁcation scheme also focuses
on the well-being of the working dog itself in that behaviors that clearly
cause harm to the working dog are seen as husbandry problems. This
strategy, however, seems to downplay behaviors that might be considered
abnormal in a systems approach if those behaviors do not adversely aﬀect
learning or performance or unless they cause an imminent risk to health or
safety. In some ways, this observation is true. The working dog class-
iﬁcation outlined here would not necessarily identify repetitive behaviors
(eg, spinning or barking) as problems unless they resulted in an inability to
acquire a required behavior, if the repetitive behavior disrupted a required
task as it was being performed, or if the behavior caused injury or illness
in associated people, other animals, or the patient itself. Likewise, an intact
male dog with a strong proclivity to urine-mark repetitively, display ag-
gression toward other male dogs, or mount objects might not be considered
as problematic if that dog was never brought indoors (and there was
nothing to be damaged by mounting or marking) and never encountered
other dogs. Ignoring the presence of behaviors that might be considered
abnormal or unacceptable in pet dogs raises a question as to whether the
strategy, if rigidly applied, might be harmful to the long-term welfare of an
animal.

430

W.F. Burghardt

/ Vet Clin Small Anim 33 (2003) 417–446

In reply to this question, the ﬁrst point to note is that successful working

dog handlers are aware and communicative regarding what they consider
abnormal or unacceptable in their animals and are aware and strong ad-
vocates for the well-being of their animals [42]. Depending on the housing
and employment style (a signiﬁcant number of working dogs live with their
handlers and work indoors and in close proximity to other animals and
people), working dogs may be presented for the same kinds of behavioral
problems as pet dogs, and these problems are often not ones that interfere
with task, life, limb, or property. Secondly, this working dog behavioral
problem classiﬁcation is meant to extend rather than to replace traditional
behavioral taxonomies. It is intended to focus on task-related problems,
potentially as the result of treatable medical condition(s) and then as be-
havioral problems worthy of the application of applied behavioral analysis
and medical and behavioral management and not as the inevitable result of
factors like poor breeding, bad temperament, neglectful or abusive rearing,
fatal training failures, or age-related ‘‘performance-failure,’’ which should
result only in eliminating an animal from a program.

Some behavioral problems of working dogs

Most reviews of behavioral problems in a particular setting or in a

particular species include a summary of the demographics of the population
served and a description of the kinds of behavioral problems commonly
identiﬁed. In a behavioral referral practice serving pet dogs and cats, one
might summarize the most frequently encountered behavioral problems as
those that produce risk to life, limb, or property [43,44].

The Military Working Dog Veterinary Service Behavioral Medicine

Section has a referral population of approximately 1500 military working
dogs located at more than 200 locations around the world. Currently,
Belgian Malinois Dogs represent approximately 50% of the population and
German Shepherd Dogs comprise about 37% of the population (Fig. 5).
The range of ages in the population as of October 1, 2000, was birth to 15
years of age (Fig. 6), and the median age was 8.2 years. Intact male dogs
comprise about 55% of the population, whereas about 23% of the animals
are neutered male dogs (Fig. 7). Virtually all the 18% of the population that
are female are also spayed (intact female dogs are maintained only for the
current breeding program). This proﬁle differs signiﬁcantly from that of
other population bases described previously (eg, Landsberg [43]). In the US
Department of Defense military working dog, only a few breeds are re-
presented, the ratio of male to female dogs does not parallel the typical re-
ferral base (with male dogs being disproportionately represented), and the
proportion of intact versus neutered animals also differs from a typical
veterinary practice in the United States (with intact male dogs and spayed
female dogs being disproportionately represented).

431

W.F. Burghardt

/ Vet Clin Small Anim 33 (2003) 417–446

In a recent review of the medical records of 927 military working dogs

that were humanely euthanized between 1993 and 1996 [45], the attending
veterinarian cited a behavioral problem (overwhelmingly, inappropriate
aggression) as the primary reason for euthanasia in 18 cases or 2% of the
total (see Fig. 2). During the same period, however, another 130 dogs (14%)
were euthanized for what the authors extracted as ‘‘geriatric decline’’ and
another 8 (0.9%) were euthanized for ‘‘brain diseases,’’ whereas a primary
reason for the decision to euthanize the dog could not be adequately
determined from the medical record for 42 additional cases (4.5%). It is not

Fig. 5. Breed distribution of US military working dogs as of October 2000 expressed as
percentage of total population.

Fig. 6. Age distribution of US military working dogs as of October 2000 expressed as
percentage of total population. Whole year ages (eg, 3.0 years) are midpoints of each year group
(eg, 2.50–3.49 years).

432

W.F. Burghardt

/ Vet Clin Small Anim 33 (2003) 417–446

unreasonable to speculate that a number of the geriatric decline (and
possibly brain disease) cases might have had one or more behavioral signs as
a signiﬁcant part of their presenting complaint. It should be re-emphasized
that these data are on the primary reason for spontaneous death or the
decision to perform humane euthanasia and are not measures of incidence
of any of the kinds of conditions listed. Also, additional reasons used in the
decision to consider humane euthanasia as the best course of treatment for
any patient (eg, comorbidity of neoplasia and some signiﬁcant behavioral
problem) were not evaluated in the cited paper.

In an unpublished preliminary evaluation of a series of 60 behavioral

consultations and referrals in military working dogs presented at the
Department of Defense Military Working Dog Veterinary Service
Behavioral Medicine Section, 80 individual behavioral complaints (an
average of 1.3 behavioral complaints per patient) were recorded on the
master problem lists (Fig. 8).

The most commonly presented behavioral complaints were problems

involving aggression, representing over 30% of all the complaints. In this
group of problems, most patients exhibited excessive aggression in
inappropriate settings or toward unacceptable targets, such as the dog’s
own handler. These behavioral problems involving excessive or inappro-
priate aggression usually met the criteria for classiﬁcation as disruptive
problems (indirectly interfering with task performance), although some,
such as aggressiveness directed at the handler when kenneling, could better
be classiﬁed as husbandry problems. Only two complaints in this series were
presented for a problem of insuﬃcient aggression, classiﬁed as learning-
related problems because they prevented acquisition of a required controlled
aggression task.

Fig. 7. Sex and reproductive status for US military working dogs as of October 2000 expressed
as percentage of total population. Note that 0.2% of female dogs were intact and that sex and
reproductive status were unavailable in the database for 4.0% of the population.

433

W.F. Burghardt

/ Vet Clin Small Anim 33 (2003) 417–446

Twenty-ﬁve percent of the complaints involved patients displaying

repetitive behaviors that were either disruptive or constituted husbandry
problems. The most common repetitive behavior was repetitive spinning
(rapidly ambulating in a tight circle). Of patients displaying spinning
behavior, most were Belgian Malinois Dogs, and most of these patients were
presented because of self-trauma caused by the behavior. Only a few of the
patients displayed spinning that was disruptive of task performance. Other
problem repetitive behaviors were repetitive licking and chewing and biting
of self, with or without secondary trauma. One unusual case involved a
patient that repetitively and persistently grasped its prepuce or scrotum in its
teeth, vocalized quietly, and stayed in one position for long periods.
Fortunately, this was a patient that had not yet been purchased for the
program, and it was returned to its owner without further diagnosis. It
might be noted that one article describes as a valuable characteristic the
increased energy and stamina of Belgian Malinois Dogs for US military
service [46] compared with the German Shepard Dog, which had previously
been the breed of choice.

Fig. 8. Summary of clinical presentations for 80 behavioral problems identiﬁed in 60 military
working dogs randomly selected by behavioral case record. Aggression = inappropriate
aggression directed toward human beings; repetitive = repetitive behaviors; escape

agitation = situationally related agitation and attempts to escape; detection = failure to
detect a previously ‘‘known’’ substance; obj release = failure to release a reward object on
command; overactivity = excessive activity interfering with task performance.

434

W.F. Burghardt

/ Vet Clin Small Anim 33 (2003) 417–446

Another 12% of the complaints represented situations in which animals

became aroused or agitated and attempted to retreat, escape from, or avoid
a situation or stimulus. Identiﬁable situations and stimuli producing this
escape and avoidance behavior were speciﬁc kinds of ﬂooring surfaces
(particularly smooth surfaces), loud noises (especially gunﬁre and the
operation of small arms without discharge), thunderstorm-related cues,
industrial noise, and the presence of certain individuals. Dogs displaying this
avoidance behavior when exposed to gunﬁre are particularly interesting,
because all military working dogs are initially procured only if they display a
neutral response to weapons ﬁre. These dogs are also extensively exposed to
weapons ﬁre during maintenance training so as to maintain their neutrality
in behavioral response. When escape and avoidance behavior to gunﬁre
developed in these cases, the behavior was extremely disruptive to task
performance and often resulted in the patient being eliminated from further
use as a military working dog. Likewise, avoidance of people, surfaces,
industrial noise, and storms usually developed as a problem in adult dogs
that had not previously shown avoidance responding in similar situations.

Approximately 9% of the behavioral complaints involved a speciﬁc dis-

ruption of a patient’s performance in substance detection. These problems
were reported both as learning-related problems in which a patient could
not progress successfully in initial training and as later failure of patients
previously proﬁcient in detection work. Detection problems were described
both for patients that did not respond with suﬃcient reliability in the
presence of a substance for which they were trained to respond (misses) and
for mistakes in which patients responded in the absence of a target substance
(false alarms). Observation of these patients’ detection performance re-
vealed that some animals appeared to rely too heavily on cues provided
(unconsciously) by the handler during training, dissociation between
substance identiﬁcation and responding to obtain reward (anticipatory res-
ponding), fatigue and associated panting during task performance (resulting
in less olfactory exposure to olfactory cues), and inattention to task (in-
creased distractibility). Anosmia was not demonstrated in any of the cases.

Approximately 8% of the problems presented involved a problem unique

to dogs trained to perform detection and controlled aggression tasks. This
problem involved the unwillingness of patients to release or relinquish an
object on command. These ‘‘release’’ problems in some ways look like object
guarding in pet dogs, but they are usually limited to situations where the dog
is holding a reward object (reward toy or bite sleeve) in its mouth, and the
possessive behaviors are usually of much greater intensity than problems
seen in pets. Trainers and handlers often call these presentations ‘‘out’’
problems (the command to release a toy or a bite is often ‘‘out’’). Dogs that
do not release a bite sleeve or suit are called ‘‘wrap-happy’’ (the bite sleeve is
often called a ‘‘wrap’’). These problems can present with or without an
aggressive component (cases where the patient displays aggression in an
attempt to avoid relinquishing an object were not included in the aggressive

435

W.F. Burghardt

/ Vet Clin Small Anim 33 (2003) 417–446

behavioral problems discussed previously). These complaints most often
arise during initial training, so they may be classiﬁed as learning-related
problems. Some patients, however, were presented after successful training
and varying amounts of successful task performance as having performance
problems. Anecdotally, some of the problems involving relinquishment of
reward objects during substance detection training or employment were
addressed and resolved easily by switching the patient’s reward to a food
object that could be quickly consumed. Relinquishment problems involving
bite equipment were often extremely diﬃcult to address by the time a dog
was presented for treatment. Interestingly, one of the key factors used to
select a candidate adult animal for military service is the animal’s pursuit of
and interaction with toys.

Finally, 4% of the complaints were for overactivity and distractibility,

often in apparently successfully trained animals, and were presented as
performance problems. Overactivity was occasionally identiﬁed in animals
during initial training as a learning-related problem. Animals presented
for overactivity often had diﬃculty in maintaining their focus on task
performance and became distracted by people and events that occurred
around them. When presented during training, some of these patients were
described as ‘‘slow learners,’’ and the overactivity problem often coexisted
with problems of repetitive behavior. In an ongoing series of cases not
included in this survey, dogs showing signs of overactivity were often
presented initially for failure to maintain optimal body condition scores
despite apparently good health and adequate ingestion of a performance diet.

Not included in this reviewed series of cases but identiﬁed with some

regularity as problems in military working dogs are separation-related
behavioral problems, CCD, forging and lunging on the leash and poor
obedience compliance, unacceptable urination and defecation in work
settings, urine-marking in work settings, and excessive interdog aggression.

Approach to behavioral therapy

In examining some aspects of behavioral therapy in working dogs, it may

be illuminating to identify some of the challenges. Behavioral therapy in
working dogs can be anything from routine to challenging. Many problems
common to working dogs and pet dogs are managed in much the same way
in both settings, often with good outcomes. The fact that working dogs
support organizations with speciﬁc missions and the presence of critical
working behaviors in the individual patient add several dimensions to the
behavioral management of working dogs.

One challenge in working dog behavioral therapy involves the

fundamental decision whether to diagnose and treat a behavioral problem.
Although pet owners in some ways perform a cost-beneﬁt analysis when
deciding on their pet’s spectrum of medical and behavioral care (eg, ‘‘price-

436

W.F. Burghardt

/ Vet Clin Small Anim 33 (2003) 417–446

shopping’’ for veterinary care), the average owner usually makes decisions
on care primarily on the basis of the welfare and comfort of the pet [47,48].
In an organized working dog program (especially in large-scale programs),
however, the comfort and welfare of the individual animal may at times
compete with the needs of the program to produce and maintain a sufﬁcient
number of animals that are capable of successfully performing the critical
tasks associated with their employment. Thus, when considering treatment
options for a working dog, additional emphasis is placed on determining the
likelihood of successfully managing (ideally, curing) a behavioral problem
so that it does not prevent successful completion of a patient’s normal
duties. Time lost from work (for both the dog and its handler) while initially
treating a behavioral problem may also become a critical concern, as might
the amount of time and other costs of managing an animal with a behavioral
problem requiring long-term management.

These management decisions are not made lightly in any large-scale

program. The initial investment in candidate animals is often quite signiﬁcant,
regardless of whether candidate dogs are purchased or produced through a
breeding program. Training time is extensive and costly, and the need for an
unanticipated replacement of an animal may cost the user the services of a dog
for weeks or even months while awaiting a trained replacement.

Both in private practice and in working dog programs, when dogs pose

a signiﬁcant safety risk to themselves or others, humane euthanasia is
sometimes contemplated. In pet practice, some surveys [49,50] suggest that
owners are willing to tolerate many less severe behavioral problems either
without any behavioral treatment or after one or more unsuccessful at-
tempts to manage a problem. In contrast, military working dogs with less
severe behavioral problems may be removed from active service (with or
without attempts at behavioral treatment) but retained by the program to
help train military dog handlers. These dogs may also be reallocated from
the military to other governmental canine programs, where they might be
able to perform successfully despite the presence of a mild to moderate
behavioral fault, or they may be granted to a civilian law enforcement
agency for use. Recently, legislation (Public Law 106-446) has been enacted
that allows the US military to adopt ‘‘excess’’ military working dogs out to
individuals (previously, aggression-trained military working dogs were
considered to be unadoptable). This adoption process now provides another
method to ‘‘retire’’ military dogs with mild to moderate behavioral problems
that do not have an impact on the safety or welfare of the patients
themselves or others.

Working dog handlers are often extremely involved in their dog’s care

and highly motivated to assist [42]. Nonetheless, treatment failure for
behavioral problems in working dogs probably approaches rates for pet
dogs experienced in other settings, although data on outcomes from any
therapeutic setting are fairly limited. It is likely that the single most
important factor resulting in failure of a behavioral plan in working dogs is

437

W.F. Burghardt

/ Vet Clin Small Anim 33 (2003) 417–446

that of poor compliance with the plan. Factors contributing to this cause
of treatment failure likely include issues similar to those experienced in pet
behavioral therapy, such as a perception that the prognosis for successful
resolution or management of the problem is low; too much time is required
to execute the behavioral plan; the plan is excessively disruptive of normal
activities; poor understanding of methods to be applied as part of the
treatment plan; excessive cost of medications, equipment, or supplies
required to follow the treatment plan; inadequate follow-up in support of
the treatment plan; or inability to observe signiﬁcant positive behavioral
change in the patient within a reasonable period of time.

Especially in working dogs (but also applicable to pet behavioral

therapy), the initial assessment of a behavioral problem may lack a clear-cut
identiﬁcation of the speciﬁc unacceptable behaviors that are present or the
required behaviors that cannot be obtained from the patient. This step is
important with working dogs, because a lack of detail makes it diﬃcult
to identify speciﬁc alternative behaviors with which to replace speciﬁc
unacceptable behaviors, leaves the behavioral plan without speciﬁc steps to
reduce the production of speciﬁc unacceptable behaviors and increase the
probability of desired behaviors through reinforced practice (and other
means), and leaves the team without clear and measurable intermediate
criteria for determining if the therapy is successful. Particularly when
working with skilled trainers and handlers, especially when attempting to
manage a case through an attending veterinarian at a physically remote
location, the better deﬁned a behavioral plan can be made (not necessarily
more complicated), the more likely it is that treatment compliance can be
obtained. Anecdotally, treatment compliance can often be enhanced with
regularly scheduled recheck appointments, especially if structured evalua-
tion of behavioral progress can be incorporated into the recheck.

As mentioned several times previously, treating behavioral problems in

working dogs takes on several dimensions not usually experienced in pet
dogs. For example, diagnosis and treatment for a behavioral problem may
not be pursued because of a limitation in time or other resources required to
address a problem. Even in an ideal situation, where treating a behavioral
problem in a working dog may be seen as highly desirable, special
accommodations may need to be made to allow the animal to continue
working during treatment. Alternatively, plans may need to be made to
allow an animal to be out of training or duty for all or part of the time
treatment is attempted to ensure that performance on critical tasks is not
compromised. This is an important issue, because working dogs may be
charged with protecting human life (eg, police dogs, Seeing Eye dogs, bomb-
detector dogs, search and rescue dogs) and failure in a task may be fatal.
In addition, legal issues may exist in one or more aspects of a dog’s
employment. For example, courts have entertained arguments suggesting
that the presence of any sort of ongoing treatment (especially medications)
in a drug detection dog might compromise the ability of the animal to

438

W.F. Burghardt

/ Vet Clin Small Anim 33 (2003) 417–446

perform its detection task accurately, thus weakening the value of the dog
team’s ‘‘testimony’’ when a detection response is made (Dan Craig, DVM,
MA, San Antonio, TX, personal communication, 2002).

Drug therapy in working dogs

One particular consideration in the behavioral treatment of working dogs

is the use of psychoactive drugs in an animal that needs to work during
treatment. There are several problems that arise in this setting. The ﬁrst
potential problem is that using a medication in a working dog may have
direct adverse eﬀects on the physical performance of a required behavior.
Examples of this problem include the obvious example of the potential
adverse eﬀects of a muscle relaxant on the performance of an agility task
[51]. Other drugs may blunt sensory abilities. Examples include the effects of
adrenergic and noradrenergic drugs on vision [52] and olfactory degrada-
tion by some corticosteroids (eg, Kavaliers and Ossenkopp [53]). Some
medications may have undesirable effects on learning, memory, and per-
formance. For example, benzodiazepines may exert an amnesic effect and
disrupt learning [54], selegiline may increase errors and trials to criterion on
certain tasks in normal dogs [55], and methylphenidate can increase re-
action time and errors on some tasks and in some animals [56].

There are two issues regarding the use of drugs in working dogs that

go well beyond concerns of pet owners. First, there has been almost
no systematic work done to evaluate the eﬀects of most medications
on ‘‘normal’’ canine maintenance or social behaviors or on the learning,
memory, and accurate performance of working dog tasks. Much of the
human and laboratory animal psychopharmacology research of the past
50 years [57] suggests that most psychotropics (and many other drugs not
necessarily considered psychotropic medications) produce dose-dependent
effects on the production of a wide variety of learned and species-typical
behaviors maintained by diverse schedules of reinforcement and cued by
diverse intrinsic and extrinsic stimuli. Coupled with the fact that most drugs
used in canine behavioral therapy are still considered extralabel uses,
signiﬁcant questions arise regarding temporary or permanent task-related
disruption that may occur when psychotropics are used in working dogs.

The second issue that arises is the possibility of state-dependent learning.

This phenomenon was ﬁrst described experimentally in a setting where
animal or human subjects were trained to perform a novel task successfully
when under the inﬂuence of a drug (eg, Jackson [54]). These subjects, when
tested later under the inﬂuence of the drug, were able to perform the task
successfully. When retested with no drug present, however, the subjects
showed a marked degradation in their ability to perform the task. When a
benzodiazepine is used as the drug during training, the degradation in
performance has traditionally been called an ‘‘amnesic’’ effect, implying

439

W.F. Burghardt

/ Vet Clin Small Anim 33 (2003) 417–446

that the subjects correctly learned the required behavior when under
the inﬂuence of the drug but forgot how to perform the task when no
drug was present (this serves as a model for a clinical amnesic effect in
human beings and has been used to explain recurrence of symptoms in
benzodiazepine-treated patients after withdrawal of their medication). The
basic ﬁnding, however, has been reproduced using other drugs and tasks.
Some of the drugs producing state-dependent learning do not have obvious
sedative, hypnotic, or amnesic clinical effects. In these models, the drug
seems to have acquired the characteristics of a discriminative stimulus for
the performance of the behavior learned under its inﬂuence. Regardless of
whether a drug serves as a discriminative stimulus or has direct action on
learning, memory, or performance, the possibility of this phenomenon
occurring in the context of task training or performance raises another red
ﬂag that demands further evaluation of psychotropics in situations involving
skilled canine tasks.

Nevertheless, psychotropic drugs can be and are used in working dogs

provided that some precautions are exercised. First, it should be considered
unwise to administer a psychotropic drug (or other medications with po-
tential sensory, neurologic, or behavioral side eﬀects) acutely to a dog re-
quired to perform a critical task with safety concerns. As much as possible,
behavioral medications should be administered in a nonwork setting, used
to obtain desired eﬀects, tapered to a minimum eﬀective dose as soon as is
possible, and withdrawn according to currently recommended schedules as
soon as they are no longer critical to therapy.

Sample cases

Several applications of pharmacotherapy in case management involving

working dogs may be useful in illustrating how behavioral drugs might be
applied and their eﬀects assessed in these patients.

The ﬁrst case involved a Ori, a 3-year-old intact male German Shepherd

Dog that was experiencing diﬃculty in progressing in substance detection
training because of excessive distractibility triggered by the presence of
people during training trials. The patient also engaged in repetitive spinning
in the home cage and had diﬃculty in maintaining body condition despite
adequate nutrition. The patient had no active major medical problems. The
patient’s presentation and clinical assessment indicated moderate over-
activity and distractibility categorized as learning-related problems and
moderate intermittent repetitive spinning in the home kennel and poor body
condition (3

/9) categorized as husbandry problems. The undesirable

behaviors included excessive movement at rest, breaking a search pattern
when distracted by people, and repetitive spinning in the home kennel.
Desired behaviors were relaxed stand, sit, and down with minimal physical
restraint; completion of search pattern from start to ﬁnish with required

440

W.F. Burghardt

/ Vet Clin Small Anim 33 (2003) 417–446

sensitivity and speciﬁcity; and patient engaging in sedentary recreational
behaviors in kennel. A 3-day trial using 0, 0.25, and 0.5 mg

/lb of

methylphenidate [58] was conducted to assess whether the patient would
beneﬁt from this drug (Fig. 9). Results of the assessment suggested that the
patient might beneﬁt from methylphenidate at a dose of 0.25 mg

/lb. At that

dose (but not at 0 or 0.5 mg

/lb), the patient displayed both a decrease in

heart rate of 10 beats per minute (bpm) and a decrease in overall activity
(‘‘attitude’’) as rated by a masked evaluator. A trial course of the 40 mg of
methylphenidate administered by mouth twice daily was prescribed for a
period of 14 days. Because there were no safety contraindications, the
patient was continued in training with the recommendation that the handler

Fig. 9. Methylphenidate response test data from a single patient. Heart rate and attitude rating
measured before treatment and at 30, 60, and 90 minutes after treatment with 0, 0.25, and 0.5
mg of methylphenidate given orally. Attitude is rated as sedate (0), calm (1), active (2), or
agitated (3).

441

W.F. Burghardt

/ Vet Clin Small Anim 33 (2003) 417–446

return to shorter search patterns, attempt to complete these patterns both
with and without distractions, and when successful, gradually increase the
length of the pattern to that required for certiﬁcation. The handler was
contacted frequently by the veterinary behavioral technician to determine
whether the patient was performing acceptably. Reassessment was planned
if the trainer continued to experience problems. The patient was provided a
Buster Cube in the home kennel as an item of environmental enrichment and
to provide a more sedate alternative behavior to spinning. The veterinary
behavioral technician manipulated the toy in the presence of the patient and
made the toy available with differing contents on different days to make it
more attractive to the patient. The patient’s weight was monitored weekly.
Spinning was measured as the amount of time engaged in the activity,
measured for 10 minutes twice each week at the same time of day.

After 2 weeks of medication trial, the trainer reported that the patient

was progressing successfully in training and that the patient would stand,
sit, and down without excessive movement. The patient had gained 3 lb.
Although the patient interacted with the Buster Cube infrequently, spinning
had almost completely resolved by the second assessment.

At 4 weeks, the patient’s body condition score was 4 of 9 and its weight

was within the normal weight range. Training had continued to progress
well, and the patient stood, sat, and ‘‘downed’’ to the trainer’s satisfaction.
Caretakers noted that the patient would engage in spinning for a few
seconds when returned to its kennel from a walk or grooming but then
would immediately settle.

At 8 weeks, the patient was noted to be spinning for the entire 10-minute

measurement period. A check noted that the patient had run out of
medication 3 days earlier and that the prescription had not been reﬁlled. The
trainer did not report any training problems, and the dog was readying for a
certiﬁcation test. The dog’s weight had stabilized within the ideal range, and
the patient had a body condition score of 4 of 9. The medications were
reﬁlled, and the spinning disappeared on re-evaluation the next day.

At 16 weeks, the patient had certiﬁed and was transferred to operational

use. The attending veterinarian was advised to maintain medication,
monitor the patient’s body condition score once weekly, and review the
animal’s performance with the kennel master and the dog’s handler at least
monthly. The veterinarian was advised to contact the Behavior Medicine
Section if any questions or problems arose and to consider quarterly ‘‘drug
vacations’’ to assess whether continued therapy was indicated. At last
follow-up, the patient was performing well and was still on 0.25 mg

/lb of

methylphenidate twice daily.

A second case involved Bruno, a 2-year-old intact male German

Shepherd Dog in training, with a presentation similar to the previous case.
In addition, this patient stool soiled its home kennel daily and demonstrated
tremor associated with elevated activity and vigilance as well as some
apprehension (freezing and retreat) during training sessions. This patient

442

W.F. Burghardt

/ Vet Clin Small Anim 33 (2003) 417–446

was tested with methylphenidate in a similar manner as in the previous case
but with equivocal results. A 3-week trial of clomipramine (0.5 mg

/kg

administered orally twice daily) indicated that the patient might beneﬁt from
the drug because its weight began to stabilize and the trainer indicated that
the patient was somewhat more tractable.

Unfortunately, this patient was also extremely diﬃcult to medicate.

Conventional therapy with oral medications in large kennels of military
working dogs involves hiding the drug in a small amount of canned food,
but this patient did not eat canned food at all (even with 12-hour
deprivation). This patient was also resistant to head handling, making oral
dosing impossible. Because twice-daily or even daily administration of a
medication was not possible, an alternative was sought. Fluoxetine has
recently been approved for human use on a once-weekly basis (Prozac
Weekly). The product monograph indicates that the eﬀectiveness is not
based on release of the drug over an extended period of time. Rather, the
drug is given at a much higher dose than on the daily regimen (80 mg versus
20 mg), and beneﬁcial eﬀects are a result of blood levels of norﬂuoxetine, a
metabolite with a relatively long half-life.

Adequate pharmacokinetics were not available for the dog [59], so

we evaluated the levels of ﬂuoxetine and norﬂuoxetine in the serum of
this patient for a week after the administration of a single 80-mg dose
(0.5 mg

/kg) of ﬂuoxetine (Fig. 10). Fluoxetine levels rose to slightly over

40 ng

/mL on day 1 but decreased to nondetectable levels thereafter.

Norﬂuoxetine rose to almost 100 ng

/mL on day 1 and then decreased much

more slowly. An estimation of its half-life in the dog from these data is

Fig. 10. Serum levels (in nanograms per milliliter) of ﬂuoxetine and norﬂuoxetine measured
before and at daily intervals after the oral administration of 80 mg (1 mg

/lb) of ﬂuoxetine in a 3-

year-old intact male Belgian Malinois military working dog. The lines connecting the daily
sample values have been smoothed by computer algorithm.

443

W.F. Burghardt

/ Vet Clin Small Anim 33 (2003) 417–446

approximately 2 to 3 days. Anecdotally, beneﬁcial behavioral effects were
noted in this patient on days 1 through 3 (but not on day 4), suggesting that
the effective dose of norﬂuoxetine was in excess of 40 ng

/mL in this patient.

Unfortunately, treatment was not successful in this patient, because even

weekly treatments were next to impossible. The patient was eliminated from
training and was transferred to another use.

References

[1] Hart BL, Hart LA. Selecting pet dogs on the basis of cluster analysis of breed behavior

proﬁles and gender. JAVMA 1985;186:1181–5.

[2] American Kennel Club. The complete dog book. 19th edition. New York: Wiley; 1998.
[3] Lemish MG. War dogs: a history of loyalty and heroism. Herndon (VA): Batsford Brassey;

1999.

[4] Maciejewski A, Bonke G. Outcomes in a well-established police service dog breeding

program. In: Proceedings of the International Working Dog Breeding Conference,
September 11–13, 2001, San Antonio, TX. San Antonio (TX): International Working Dog
Breeding Association; 2001.

[5] Chapman SG. Police dogs in North America. Springﬁeld (IL): Charles Thomas; 1990.

p. 24.

[6] The Department of the Treasury. Law enforcement canine training. Washington, DC:

Department of the Treasury; 1993.

[7] Harris M, Gellin M. Pet therapy for the homebound elderly. Caring 1990;9:48–51.
[8] Brodie SJ, Biley FC. An exploration of the potential beneﬁts of pet-facilitated therapy.

J Clin Nurs 1999;8:329–37.

[9] Eddy J, Hart LA, Boltz RP. The effects of service dogs on social acknowledgments of

people in wheelchairs. J Psychol 1988;122:39–45.

[10] Hart LA, Zasloff RL, Benfatto AM. The pleasures and problems of hearing dog

ownership. Psychol Rep 1995;77:969–70.

[11] Hebard C. Use of search and rescue dogs. JAVMA 1992;203:999–1001.
[12] Champness KA. Development of a breeding program for drug detector dogs—based on

studies of a population of guide dogs [doctoral dissertation]. Melbourne: University of
Melbourne; 1996.

[13] Leighton EA. Genetics of canine hip dysplasia. JAVMA 1997;210:1417–8.
[14] Hilliard S, Burghardt WF. Development and validation of behavioral testing instruments

for longitudinal study of military working puppies. In: Proceedings of the International
Working Dog Breeding Conference, San Antonio, TX. San Antonio (TX): International
Working Dog Breeding Association; 2001.

[15] Todhunter R. Physical and genetic methods to prevent hip dysplasia. In: Proceedings of the

International Working Dog Breeding Conference, San Antonio, TX. San Antonio (TX):
International Working Dog Breeding Association; 2001.

[16] Mann KA. Comparison of OFA, PennHIP, and other radiographic evaluations of canine

hip dysplasia. In: Proceedings of the International Working Dog Breeding Conference,
San Antonio, TX. San Antonio (TX): International Working Dog Breeding Association;
2001.

[17] Vandeloo J. Breeding and developing detector dogs, the Australian Customs Service so far.

In: Proceedings of the International Working Dog Breeding Conference, San Antonio, TX.
San Antonio (TX): International Working Dog Breeding Association; 2001.

[18] Dean S. Observations from a large scale breeding programme. In: Proceedings of the

International Working Dog Breeding Conference, San Antonio, TX. San Antonio, (TX):
International Working Dog Breeding Association; 2001.

444

W.F. Burghardt

/ Vet Clin Small Anim 33 (2003) 417–446

[19] Gobello C. Estrus induction with dopaminergic agonists in the bitch. In: Proceedings of the

International Working Dog Breeding Conference, San Antonio, TX. San Antonio (TX):
International Working Dog Breeding Association; 2001.

[20] Slabbert JM. Breeding, socializing and evaluating juvenile police dogs in a closed breeding

system. In: Proceedings of the International Working Dog Breeding Conference,
San Antonio, TX. San Antonio (TX): International Working Dog Breeding Association;
2001.

[21] Frederickson MA. Temperament testing procedures for animals involved in nursing home,

school, and hospital visiting programs through Delta Society Pet Partners [abstract].
Applied Anim Behav Sci 1993;37:83.

[22] Overall KL, Dunham AE, Acland G. Responses of genetically fearful dogs to the lactate

test: assessment of the test as a provocative index and application in mechanistic diagnoses.
World Congress on Psychiatric Genetics, Monterey, CA. Mol Psychiatry 1994;4(Suppl 1):
S125.

[23] Hilliard S, Burghardt WF. Accuracy, sensitivity, and speciﬁcity of substance detection

responding in military working dogs [abstract]. Presented at the Annual Meeting of the
Southwestern Psychological Association, April 18–20, 2002. Corpus Christi, 2002.

[24] Goodloe LP. Issues in description and measurement of temperament in companion dogs.

In: Voith VL, Borchelt PL, editors. Readings in companion animal behavior. Trenton (NJ):
Veterinary Learning Systems; 1996. p. 32–9.

[25] Skinner BF. The behavior of organisms. Behavior of organisms. Acton (MA): Copley

Publishing; 1999.

[26] Breland K, Breland M. The misbehavior of organisms. Am Psychol 1961;16:681–4.
[27] Scott JP, Fuller JL. Genetics and social behavior of the dog. Chicago: University of

Chicago Press; 1965. p. 15–28.

[28] Anastasi A, Fuller JL, Scott JP, Schmitt JR. A factor analysis of the performance of dogs

on certain learning tests. Zoologica 1955;40:33–46.

[29] Pfaffenberger CJ. The knowledge of dog behavior. New York: Howell Books; 1963.
[30] Cattell RB, Korth B. The isolation of temperament dimensions in dogs. Behav Biol

1973;9:15–30.

[31] Serpell JA, Hsu Y. Development and validation of a novel method for evaluating behavior

and temperament in guide dogs. Appl Anim Behav Sci 2001;72:347–64.

[32] Beaver BV. Canine behavior: a guide for veterinarians. Philadelphia: WB Saunders; 1999.

p. 17–8.

[33] Beaudet R, Chalifoux A, Dallaire A. Predictive value of activity level and behavioral

evaluation on future dominance in puppies. Appl Anim Behav Sci 1994;40:273–84.

[34] Ruehl WW, Bruyette DS, DePaoli A, et al. Canine cognitive dysfunction as a model for

human age-related cognitive decline, dementia, and Alzheimer’s disease: clinical
presentation, cognitive testing, pathology and response to l-deprenyl therapy. Prog Brain
Res 1995;106:217–25.

[35] Moyer KE. Kinds of aggression and their physiological basis. Communications Behav Biol

1968;2:65–87.

[36] Borchelt PL, Voith VL. Classiﬁcation of animal behavior problems. Vet Clin North Am

Small Anim Pract 1982;12:571–85.

[37] Odendaal JS. A diagnostic classiﬁcation of problem behavior in dogs and cats. Vet Clin

North Am Small Anim Pract 1997;27:427–43.

[38] Overall KL. Clinical behavioral medicine for small animals. St. Louis: Mosby; 1997.

p. 195–208.

[39] Hopkins SG, Schubert TA, Hart BL. Castration of adult male dogs: effects on roaming,

aggression, urine marking, and mounting. JAVMA 1976;168:1108–10.

[40] Shull-Selcer EA, Stagg W. Advances in the understanding and treatment of noise phobias.

Vet Clin North Am Small Anim Pract 1991;21:353–69.

[41] Chrisman CL. Problems in small animal neurology. Philadelphia: Lea & Febiger; 1991.

445

W.F. Burghardt

/ Vet Clin Small Anim 33 (2003) 417–446

[42] Hart LA, Zasloff RL, Bryson S, Christensen SL. The role of police dogs as companions

and working partners. Psychol Rep 2000;86:190–202.

[43] Landsberg GM. The distribution of canine behavior cases at three behavior referral

practices. Veterinary Medicine 1991;86:1011–18.

[44] Beaver BB. Canine behavior: a guide for veterinarians. Philadelphia: WB Saunders; 1999.

p. 1–42.

[45] Moore GE, Burkman KD, Carter MN, Peterson MR. Causes of death or reasons for

euthanasia in military working dogs: 927 cases (1993–1996). JAVMA 2001;219:209–13.

[46] Jennings PB. Veterinary care of the Belgian Malinois military working dog. Mil Med

1991;156:36–8.

[47] Medical Management, Inc. Setting fees. Presented at the Annual Meeting of the American

Veterinary Medical Association, Washington, DC, July 21–24, 1980. St. Louis, 1980.

[48] Burghardt WF. Behavioral medicine as a part of a comprehensive small animal medical

program. Vet Clin North Am Small Anim Pract 1991;21:343–52.

[49] Vacalopoulos A, Anderson RK. Canine behavior problems reported by clients in a study

of veterinary hospitals. Appl Anim Behav Sci 1993;37:84.

[50] Adams GJ. The prevalence of behavioural problems in domestic dogs: a survey of 105 dog

owners. Aust Vet Pract 1989;19:135.

[51] Maxmen JS. Psychotropic drugs fast facts. New York: WW Norton; 1991. p. 177–216.
[52] Adams HR. Adrenergic and antiadrenergic drugs. In: Booth NH, McDonald LE, editors.

Veterinary pharmacology and therapeutics. Ames: The Iowa State University Press; 1982.
p. 89–112.

[53] Kavaliers M, Ossenkopp KP. Corticosterone rapidly reduces male odor preferences in

female mice. Neuroreport 2001;12:2999–3002.

[54] Jackson A. State-dependent effects of atypical benzodiazepine receptor agonists.

Psychopharmacology (Berl) 1995;119:399–404.

[55] Mills D, Ledger R. The effects of oral selegiline hydrochloride on learning and training in

the dog: a psychobiological interpretation. Prog Neuropsychopharmacol Biol Psychiatry
2001;25:1597–613.

[56] Risner ME, Jones BE. The effects of psychomotor stimulants on single-spatial alternation

behavior in dogs. Psychopharmacology (Berl) 1979;63:137–41.

[57] Boulton AA, Baker GB, Greenshaw AJ, editors. Psychopharmacology. Totowa (NJ):

Humana Press; 1989.

[58] Luescher UA. Hyperkinesis in dogs: six case reports. Can Vet J 1993;34:368–70.
[59] Pohland RC, Bernhard NR. Postmortem serum and tissue redistribution of ﬂuoxetine and

norﬂuoxetine in dogs following oral administration of ﬂuoxetine hydrochloride (Prozac).
J Forensic Sci 1997;42:812–6.

446

W.F. Burghardt

/ Vet Clin Small Anim 33 (2003) 417–446

Index

Note: Page numbers of article titles are in boldface type.

Absorption

in behavioral medicine, 366

Acepromazine

in behavioral medicine, 373

Acetylcholine

in behavioral medicine, 371

Acral lick granuloma, 235–236

Aggression

dietary eﬀects on, 409–410
dominance-related

in dogs, 306–307

fear-related

in dogs, 305–306

food-related

in dogs, 307

human-directed

in cats, 269–286
classiﬁcation of, 271–277
disorder description,

277–279

nosography, 277–279
oﬀensive aggression, 275
pathogenesis of, 276–277
pathologic aggression, 276
petting-induced aggression,

274–275

play-related aggression, 273
predatory behavior,

271–273

prevention of, 283
redirected aggression, 276
self-defense aggression,

273–274

treatment of, 279–283
in dogs, 303–320
anxiety and, 308
diagnosis of, 304–308
dominance-related aggres-

sion, 306–307

fear-related aggression,

305–306

food-related aggression, 307
management of, 308–309
medical problems contri-

buting to, 310–311

owner-directed aggression,

306–307

predatory aggression,

307–308

rage syndrome, 311
safety issues in, 309–310
territorial aggression, 307
treatment of

behavior modiﬁcation

in, 311–315

drugs in, 315–316
surgical, 316–317

in kittens

development of, 270–271

oﬀensive

in cats, 275

owner-directed

in dogs, 306–307

pathologic

in cats, 276

petting-induced

in cats, 274–275

play-related

in cats, 273

predatory

in dogs, 307–308

redirected

in cats, 276

self-defense

in cats, 273–274

territorial

in dogs, 307

Alopecia

psychogenic, 236–237

Alprazolam

in behavioral dermatology

management, 245

in behavioral medicine, 381

Amino acids

eﬀects on cat and dog behavior, 412–413

Amitriptyline

in behavioral medicine

adverse reactions to, 385–386
dermatologic, 243–244

Vet Clin Small Anim

33 (2002) 447–453

0195-5616/03/$ - see front matter

doi:10.1016/S0195-5616(02)00135-3

Anticonvulsant(s)

in behavioral medicine, 393

Antidepressant(s)

atypical

in behavioral medicine, 391

combined use of

in behavioral medicine, 395

in behavioral medicine, 383–395

SSRIs, 387–391. See also

Selective serotonin
reuptake inhibitors
(SSRIs), in behavioral
medicine

tricyclic. See also speciﬁc drugs

in behavioral medicine, 383–387
adverse reactions to, 384–387
clomipramine, 386–387
dermatologic, 243–244
imipramine, 386
metabolism of, 384

Antipsychotic(s)

in behavioral medicine, 372–378
SSRIs with

in behavioral medicine, 395

Anxiety

in dogs, 325

human-directed aggression

resulting from, 308

separation

in dogs, 321–344. See also

Separation anxiety, in dogs

Anxiolytic(s)

in behavioral medicine, 378

Attention-seeking behavior

in dogs

ignoring of, 334–335

self-directed, 237–238

Azapirone

in behavioral medicine, 2183

Behavior(s)

attention-seeking. See

Attention-seeking behavior

compulsive

in dogs and cats
development of, 259

elimination

feline, 287–301

marking

feline, 287–301

predatory

in cats, 271–273

self-injurious, 233–234
stereotypic, 234

Behavior disorders

primary, 233–238

Behavior genetics, 345–363

classic Mendelian genetic research,

345–349

nature vs. nurture in, 354–360
research concerning behavioral

problems, 349–351

research directions in, 351–354
research trends in, 351–354

Behavior modiﬁcation

for human-directed aggression

in dogs, 311–315

Behavioral assessment

in working dogs, 422–428

Behavioral dermatology, 231–251

clinical management of, 240–248

behavioral modiﬁcation in,

241–243

diagnostic approach to, 240–241
environmental management in,

241

pharmacologic support in,

243–248

SSRIs in, 245

cutaneous sensory disorders, 239–240
described, 231–232
in dogs

primary behavior disorders,

233–238

primary behavioral disorders,

233–238

psychophysiologic disorders, 232–233
secondary behavioral disorders, 238

Behavioral disorders

secondary, 238

Behavioral medicine

absorption in, 366
acepromazine in, 373
acetylcholine in, 371
alprazolam in, 381
anticonvulsants in, 393
antidepressants in, 383–395

combined use of, 395

antipsychotics in, 372–378
anxiolytics in, 378
azapirone in, 382–383
benzodiazepines in, 378–379
beta-blockers in, 394–395
clearance in, 368
clinical management of, 213–229
clorazepate in, 381
complexity of, 213
data collection in, 214–215
diazepam in, 379–381
distribution in, 368–369

448

Index / Vet Clin Small Anim 33 (2003) 447–453

dopamine in, 370
drug combinations in, 395–397
GABA in, 371
history in, 214–215
lorazepam in, 381–382
MAOIs in, 392–393
medical diﬀerentials with potential

behavioral manifestations,
213–229

melatonin in, 394
metabolism in, 367–368
neurotransmitter receptors in,

369–371

norepinephrine in, 369–370
observation in, 214–215
opiate antagonists in, 393–394
organ systems approach to, 215–221
oxazepam in, 381–382
pharmacokinetics in, 366–369
pharmacologic management in,

365–405

drug classes in, 371–383

phenothiazines in, 373
progestin hormones in, 394
risperidone in, 373, 378
second messengers in, 369–371
serotonin in, 370–371
sex issues in, 225–227
SSRIs in

antipsychotics with, 395

thyroid function in, 222–224
transdermal application in, 397–398
treatment of

success of, 398

Behavioral modiﬁcation

for separation anxiety in dogs, 339
in behavioral dermatology, 241–243

Behavioral problems

research concerning, 349–351

Benzodiazepine(s)

in behavioral dermatology

management, 245

in behavioral medicine, 378–379

adjunctive use of, 395–396

Beta-blockers

in behavioral medicine, 394–395

adjunctive use of, 396

Busipirone

in behavioral medicine

adjunctive use of, 396

Calcium

eﬀects on cat and dog behavior, 413

Canine acral lick dermatitis, 235–236

Castration

for human-directed aggression in

dogs, 316–317

Cat(s)

behaviors of

dietary eﬀects on, 405–417.

See also Diet, effects
on cat and dog behavior

compulsive disorders in, 253–267.

See also Compulsive disorders,
in dogs and cats

elimination and marking behaviors

of, 287–301

feline hyperesthesia syndrome,

239–240

house soiling by, 287–301. See also

House soiling, feline

human-directed aggression in,

269–286. See also Aggression,
human-directed, in cats

obesity in, 408–409

Citalopram

in behavioral medicine

adverse eﬀects of, 390

Classic Mendelian genetic research,

345–349

Clearance

in behavioral medicine, 368

Clomipramine

in behavioral medicine, 386–387

dermatologic, 244

Clonazepam

in behavioral dermatology

management, 245

Clorazepate

in behavioral medicine, 381

Compulsive disorders

in dogs and cats, 253–267

causes of, 257–258
clinical approach to, 260–263
development of, 259
historical background of,

253–254

homogeneity of, 259–260
pathophysiology of, 258–259
presenting signs of, 255–256
prognosis of, 265–266
treatment of, 263–265

Cutaneous sensory disorders, 239–240

Dermatitis

canine acral lick, 235–236

449

Index / Vet Clin Small Anim 33 (2003) 447–453

Dermatology

behavioral, 231–251. See also

Behavioral dermatology

Diazepam

in behavioral medicine, 379–381

dermatologic, 245

Diet

eﬀects on cat and dog behavior,

405–417

aggression, 409–410
amino acids, 412–413
energy balance, 406–410
fat, 410–412
ﬁber, 413–415
minerals, 412–413
vitamins, 412–413

Distribution

in behavioral medicine, 368–369

Dog(s)

behaviors of

dietary eﬀects on, 405–417.

See also Diet, effects on
cat and dog behavior

compulsive disorders in, 253–267.

See also Compulsive disorders,
in dogs and cats

human-directed aggression in,

303–320. See also Aggression,
human-directed, in dogs

obesity in, 408
separation anxiety in, 321–344. See

also Separation anxiety, in dogs

working

management of

behavioral considerations

in, 419–448. See also
Working dogs,
management of,
behavioral
considerations in

Dominance-related aggression

in dogs, 306–307

Dopamine

in behavioral medicine, 370

Doxepin

in behavioral dermatology

management, 244

Drug(s)

for behavioral modiﬁcation for

separation anxiety in dogs, 339

for human-directed aggression in

dogs, 315–316

for working dogs, 441–442
in behavioral medicine, 365–405

dermatologic, 243–248

Elimination behaviors

feline, 287–301

Energy balance

dietary eﬀects on, 406–410

Environmental management

in behavioral dermatology, 241

Fat(s)

eﬀects on cat and dog behavior,

410–412

Fear

in dogs, 325

Fear-related aggression

in dogs, 305–306

Feline hyperesthesia syndrome,

239–240

Fiber

eﬀects on cat and dog behavior,

413–415

Fluoxetine

for human-directed aggression

in dogs, 316

in behavioral medicine

adverse eﬀects of, 388–389
dermatologic, 245

Fluvoxamine

in behavioral medicine

adverse eﬀects of, 390

Food-related aggression

in dogs, 307

GABA. See Gamma-aminobutyric acid

(GABA)

Gamma-aminobutyric acid (GABA)

in behavioral medicine, 371

Genetic(s)

behavior, 345–363. See also

Behavior genetics

Granuloma(s)

acral lick, 235–236

Grooming

excessive, 236–237

Growth

obesity during, 408
protein energy malnutrition during,

407–408

450

Index / Vet Clin Small Anim 33 (2003) 447–453

Homeostasis

disruption of

separation anxiety in dogs

and, 329

Hormone(s)

in behavioral medicine

adjunctive use of, 397

progestin

in behavioral medicine, 394

House soiling

feline, 287–301

diagnosis of, 287–291
treatment of, 291–300

Hyperattachment

in dogs, 323–325

Imipramine

in behavioral medicine, 386

Kitten(s)

development of

aggression in, 270–271

Lorazepam

in behavioral medicine, 381–382

dermatologic, 245

Malnutrition

protein energy

during growth, 407–408

MAOIs. See Monoamine oxidase

inhibitors (MAOIs)

Marking behaviors

feline, 287–301

Melatonin

in behavioral medicine, 394

dermatologic, 248

Mendelian genetic research, 345–349

Metabolism

in behavioral medicine, 367–368

Mineral(s)

eﬀects on cat and dog behavior,

412–413

Monoamine oxidase inhibitors (MAOIs)

in behavioral medicine, 392–393

Naltrexone

in behavioral dermatology

management, 245, 248

Neurotransmitter receptors

in behavioral medicine, 369–371

Norepinephrine

in behavioral medicine, 369–370

Nosography

in cats, 277–279

Obesity

during growth, 408
in adult cats, 408–409
in adult dogs, 408

Obsessive-compulsive disorders (OCDs),

233, 234, 235

Oﬀensive aggression

in cats, 275

Opiate antagonists

in behavioral medicine, 393–394

Ovariohysterectomy

for human-directed aggression

in dogs, 317

Owner-directed aggression

in dogs, 306–307

Oxazepam

in behavioral medicine, 381–382

dermatologic, 245

Parasite(s)

prevention of, 397

Paroxetine

in behavioral medicine

adverse eﬀects of, 389–390

Pathologic aggression

in cats, 276

Petting-induced aggression

in cats, 274–275

Phenobarbital

in behavioral medicine

adjunctive use of, 396

Phenothiazine(s)

in behavioral medicine, 373

Pheromone(s)

cat and dog

current research in, 187–211

451

Index / Vet Clin Small Anim 33 (2003) 447–453

Pheromone(s) (continued)

perception of, 188–192
pheromonotherapy in, 201–208
structures secreting, 193–201

Pheromonotherapy, 201–208

Phobia(s)

in dogs, 325

Play-related aggression

in cats, 273

Predatory aggression

in dogs, 307–308

Predatory behavior

in cats, 271–273

Primary behavior disorders, 233–238

Progestin hormones

in behavioral medicine, 394

Protein energy malnutrition

during growth, 407–408

Pruritis

psychogenic, 238

Psychogenic alopecia, 236–237

Psychogenic pruritis, 238

Psychophysiologic disorders, 232–233

Rage syndrome, 311

Redirected aggression

in cats, 276

Risperidone

in behavioral medicine, 373, 378

Safety issues

human-directed aggression in dogs

and, 309–310

Second messengers

in behavioral medicine, 369–371

Secondary behavioral disorders, 238

Selective serotonin reuptake inhibitors

(SSRIs)

antipsychotics with

in behavioral medicine, 395

for human-directed aggression

in dogs, 316

in behavioral medicine, 387–391

adverse eﬀects of, 388–391
dermatologic, 245
eﬃcacy of, 387–388

Selegiline

drug interactions with, 392–393

Self-defense aggression

in cats, 273–274

Self-directed attention-seeking behavior,

237–238

Self-injurious behavior (SIB), 233–234

Sensory disorders

cutaneous, 239–240

Separation anxiety

in dogs, 321–344

avoidance of punishment and,

338–339

behavioral modiﬁcation for

drug support in, 339

counterconditioning to fear-

eliciting stimiuli for, 339

departure cues and, 337
diagnosis of, 321–323, 330–334
disruption of homeostasis and,

329

environmental change for, 337
hyperattachment and, 323–325
leaving and returning rituals

and, 338

maintenance stimuli related to,

326–327

development of, 327–329
owner absence, 336–338
owner absence and, 336–338
systemic desensitization for,

339

treatment of, 14–16

Serotonin

in behavioral medicine, 370–371

Sertraline

in behavioral medicine

adverse eﬀects of, 390

Sex issues

in behavioral medicine, 225–227

SSRIs. See Selective serotonin reuptake

inhibitors (SSRIs)

Stereotypic behaviors, 234

Territorial aggression

in dogs, 307

Thyroid function

in behavioral medicine, 222–224

Trazodone

in behavioral medicine, 391

adjunctive use of, 396

452

Index / Vet Clin Small Anim 33 (2003) 447–453

Vitamin(s)

eﬀects on cat and dog behavior,

412–413

Working dogs

behavioral assessment in, 422–428

behavioral pathologic ﬁndings in

taxonomy of, 429–433

behavioral problems of, 433–438
management of

behavioral considerations in,

419–448

approach to, 438–441
case examples, 442–446
drug therapy in, 441–442

453

Index / Vet Clin Small Anim 33 (2003) 447–453

Document Outline

00 Preface
01 Current research in canine and feline pheromones
- Current research in canine and feline pheromones
02 Medical differentials with potential behavioral manifestations
- Medical differentials with potential behavioral manifestations
03 Behavioral dermatology
- Behavioral dermatology
04 Diagnosis and management of compulsive disorders in dogs and cats
- Diagnosis and management of compulsive disorders in dogs and cats
05 Differential diagnosis and management of human-directed aggression in cats
- Differential diagnosis and management of human-directed aggression in cats
06 Feline house soiling_elimination and marking behaviors
- Feline house soiling: elimination and marking behaviors
07 Differential diagnosis and management of human-directed aggression in dogs
- Differential diagnosis and management of human-directed aggression in dogs
08 Separation anxiety in dogs_The function of homeostasis in its development and treatment
- Separation anxiety in dogs The function of homeostasis in its development and treatment
09 Behavior genetics
- Behavior genetics
10 Pharmacologic management in veterinary behavioral medicine
- Pharmacologic management in veterinary behavioral medicine
11 Dietary effects on canine and feline behavior
- Dietary effects on canine and feline behavior
12 Behavioral considerations in the management of working dogs
- Behavioral considerations in the management of working dogs
13 Index

Wyszukiwarka

Podobne podstrony:
Dell'Osso Epidemiologic and clinical updates on impulse control disorder
2008 5 SEP Practical Applications and New Perspectives in Veterinary Behavior
2013 3 MAY Clinical Veterinary Dentistry
6 63 76 Influence of Surface Heat Treatment on Thermal Fatique Behaviour
40 549 563 On the Precipitation Behaviour in Maraging Steels
On the transformation behavior
Report on the Sexual Behavior o Robert F Young(1)
A millennium update on pediatric diarrheal illness
Haisch Update on an Electromagnetic Basis for Inertia, Gravitation, Principle of Equivalence, Spin
Post feeding larval behaviour in the blowfle Calliphora vicinaEffects on post mortem interval estima
2009 2 MAR Veterinary Public Health
a relational perspective on turnover examining structural, attitudinal and behavioral predictors
Clinical trials on onabotulinumtoxinA for the treatment
2007 2 MAR Clinical Pathology and Diagnostic Techniques
49 687 706 Tempering Effect on Cyclic Behaviour of a Martensitic Tool Steel

więcej podobnych podstron

2003 2 MAR Update on clinical veterinary behavior

Document Outline