Issues in Mental Health Nursing, 27:3–23, 2006
Copyright

c

Taylor & Francis Inc.

ISSN: 0161-2840 print / 1096-4673 online
DOI: 10.1080/01612840500312753

NO MAN’S LAND: GENDER BIAS AND SOCIAL
CONSTRUCTIVISM IN THE DIAGNOSIS OF
BORDERLINE PERSONALITY DISORDER

Pamela Bjorklund, Ph.C., RN, CS, PMHNP-BC

University of Minnesota, Minneapolis, Minnesota, USA and
The College of St. Scholastica, Duluth, Minnesota, USA

The literature on borderline personality disorder (BPD),
including its epidemiology, biology, phenomenology, causes,
correlates, consequences, costs, treatments, and outcomes is
vast. Thousands of articles and books have been published.
Because the true prevalence of BPD by sex (gender) in the
general population is still unknown, the important question
of why women, rather than men, are more frequently
diagnosed with BPD remains largely unanswered—despite
current evidence for the origin of personality disorder in
genetics and neurobiology, and despite recent suggestions
that biased sampling is the most likely explanation for
gender bias in the diagnosis of BPD. This paper reviews
selected literature on (a) the epidemiology of BPD,
(b) gender bias in the diagnosis of BPD, and (c) the social
construction of diagnosis, particularly the diagnostic entity
labeled “Borderline Personality Disorder.” It attempts a
synthesis of diverse, multidisciplinary literature to address
the question of why women outnumber men by a ratio of 3:1
in the diagnosis of BPD. It rests on assumptions that (a) to
varying degrees sociocultural factors inevitably play a role
in the expression of disease conditions, and that (b)
personality disorders, including BPD, have cultural
histories. It also rests on the belief, for which there is
considerable scholarly support, that the phenomenon called
BPD has multiple, complex, interactive, biological,
psychological, and constructed sociocultural determinants.

Address correspondence to Pamela Bjorklund, The College of St. Scholastica, 1200 Kenwood

Avenue, Duluth, MN 55811-4199. E-mail: [e-mail]

3

4

P. Bjorklund

Nurses must understand the phenomenon at this level of
complexity to provide appropriate care.

FROM THE INSIDE OUT

‘It’ has a name, but the condition is characterized by suffering of

such quality and magnitude that it cannot be clearly articulated. Its ori-
gins are largely preverbal (Adler & Buie, 1979; Gunderson, 1984, 2001;
Kernberg, 1975, 1996; Masterson, 1976, 1988, 2000). It is the sort of
suffering from which suicide seems the only escape. Indeed, the im-
pulse to suicide or self-harm is one of its defining characteristics. It is an
unparalleled poverty of soul and self that leaves its possessor bereft of
identity, alone, and empty in the midst of crowds and plenty, self-hating
to the point of viewing self-extermination as deserved. The annihilatory
threat of abandonment fuels frantic attempts to connect—usually with
those guaranteed to leave. Relationships are doomed. There is no sat-
isfactory distance from others. Distance is too close or too far. There
is no satisfactory level of environmental or interpersonal stimulation.
Stimulation is too much or too little. Here, there can be no Goldilocks.
Nothing is “just right,” and nothing soothes for long—not food, drink,
company, sex, or spending. Desperate attempts to relieve the distress
drive others to distraction. However, when others hate them as much as
they hate themselves, at least they are no longer alone. Someone finally
feels what they feel.

Paradoxically, its pain is deep and dark, both transient and unrelent-

ing, but purposeful and meaningful. Such pain can provide form and
substance to an easily fragmented, insubstantial, and ultimately illusory
core self, thus constituting an identity of sorts—but not really and not
for long. Self-destructive behavior patterns—evidence of this fragmen-
tary, false self—offer protection from feeling “bad” at the expense of a
meaningful and fulfilling life (Masterson, 1988). Sooner or later, life is
so chaotic it seems there is no life left worth living. Most live on anyway
as struggling survivors—their very existence a tribute to the tenacity
of the human spirit, although everyday life feels far from triumphal. A
fortunate few actually get helpful treatment and forge that elusive “life
worth living.” Some die.

FROM THE OUTSIDE IN

‘It’ is called borderline personality disorder (BPD). It is one of ten

personality disorders included in the Diagnostic and Statistical Manual

Bias in Diagnosing Borderline Personality Disorder

5

of Mental Disorders (DSM-IV-TR; American Psychiatric Association
[APA], 2000), which is a taxonomy of psychiatric disorders supported
by an extensive empirical foundation and pervasively used as a guide to
the clinical practice of psychiatry in the United States. Its curious name
reflects the initial confusion of both clients and caregivers who struggled
to understand life stories that were perplexing, paradoxical, inconsistent,
contradictory, and dominated to some degree by psychological difficul-
ties that confront us all (Masterson, 1988). In fact, the themes evident
within these struggles reflected social trends and psychological themes
in the culture of the United States at large, including a growing sense of
social isolation; the dissolution of social structures that lent coherence
to self-identity (along with the emergence of social customs that ag-
gravated instead of remediated problematic parent-child relationships)
(Millon, 2000); emphasis on the self to the exclusion of others; fear
of abandonment; difficulties with intimacy and relationships, and with
creativity and self-assertion (Masterson): “To cover [the] confusion, we
diagnosed these patients as ‘borderline’ because they were sicker than
the neurotic but not sick enough to be classified psychotic. They were
on the border, somewhere in between. Although this waste basket term
seemed appropriately descriptive, it really said more about our ignorance
on the subject than about what was wrong with the patient” (p. viii).

According to DSM-IV-TR, the essential feature of BPD is a pervasive

pattern of impulsivity and instability of interpersonal relationships, self-
image, and affect that begins by early adulthood and is present in a variety
of contexts. Five of nine criteria must be met for diagnosis, including
“frantic” efforts to avoid abandonment; a pattern of unstable and intense
interpersonal relationships characterized by alternating idealization and
devaluation; identity disturbance (i.e., markedly and persistently unsta-
ble sense of self); impulsivity in at least two potentially self-damaging
areas (e.g., sex, spending, binge-eating, or substance abuse); recur-
rent suicidal and/or self-mutilating behavior; affective instability due
to marked reactivity of mood; chronic feelings of emptiness; inappro-
priate, intense anger; and transient, stress-related paranoid ideation or
dissociative symptoms (APA, 2000, p. 710). It is associated with sig-
nificant psychiatric mortality and morbidity. Approximately 10% of in-
dividuals diagnosed with BPD will eventually commit suicide (Paris,
1993; APA). Not long ago, people with BPD were considered untreat-
able. Today, the research on psychotherapeutic efficacy (APA, 2003;
Fonaghy & Roth, 1996; Koerner & Dimeff, 2000; Nathan & Gorman,
1998) points to two approaches—psychoanalytic/psychodynamic ther-
apy and dialectical behavior therapy (DBT)—as most helpful in treating
BPD. The literature on both is extensive and growing and is not being

6

P. Bjorklund

reviewed here. Suffice it to say there is now ample documentation to
support real hope that persons with borderline personality disorder can
have lives worth living, learn to share those lives with others in ways
that are healthy, straightforward expressions of their deepest needs and
desires, and in so doing find fulfillment and meaning (Masterson, 1988).

BACKGROUND

Curiously, approximately 70–77% of those diagnosed with BPD are

female (APA, 2000; Swartz, Blazer, George, & Winfield, 1990; Widiger
& Weissman, 1991), which begs the question of how culture affects both
the prevalence and manifestation of the condition. Clearly, judgments
about personality functioning must take into account an individual’s
ethnic, cultural, and social background, as well as the gender ideologies
that shape behavior (APA; Sargent, 2003). Medical anthropologists have
addressed such topics as the culture of biomedicine, the cultural shap-
ing of psychiatric classification, the multiple meanings of body changes
in health and illness, power differentials in doctor-patient relationships,
and cultural variation in conceptions of normality and abnormality (Sar-
gent). Yet few reliable cross-cultural studies of severe personality dis-
order exist (Akhtar, 1995), and there is scant anthropological literature
that directly focuses on either BPD or the self-injurious behavior that
characterizes it (Sargent). Similarly, sociologists have examined sex bias
in psychiatry generally (Busfield, 1989); gender differences in rates of
mental illness as evidence for the social organization of knowledge in
patriarchal societies (Smith, 1990); the ethno- (e.g., cultural) psychol-
ogy that underlies psychiatric classification (Gaines, 1992); the nature of
diagnosis and illness as socially constructed (Brown, 1995); the growth
of psychiatric diagnosis as a function of the remedicalization of the
discipline of psychiatry (Rogler, 1997); and problems with the reliabil-
ity and validity of diagnoses of personality disorder generally (Pilgrim,
2001). But rarely have sociologists focused specifically on BPD, with
the exception of Wirth-Cauchon (2001), who has presented a social con-
structivist account of the borderline diagnosis as the medicalization of
the self-destructive feelings and behaviors of women “that lie at the ex-
treme end of a range of responses to gender contradictions and violence
in late modern society” (p. 211).

PURPOSE

The literature on BPD, including its epidemiology, biology, phe-

nomenology, causes, correlates, consequences, costs, treatments, and

Bias in Diagnosing Borderline Personality Disorder

7

outcomes is vast. Thousands of articles and books have been published.
Because the true prevalence of BPD by sex in the general popula-
tion is still unknown, the important question of why women, rather
than men, are more frequently diagnosed with BPD largely remains
unanswered—despite current evidence for the origin of personality dis-
order in genetics and neurobiology (Cloninger, 2004; Siever, 2003;
Skodol, Siever, Livesley, Gunderson, Pfohl, & Widiger, 2002; Teicher,
Andersen, Polcari, Anderson, Navalta, & Kim, 2003), and despite recent
suggestions that biased sampling is the most likely explanation for gen-
der (sex) bias in the diagnosis of BPD (Skodol & Bender, 2003). This
paper reviews selected literature on the epidemiology of BPD, gender
bias in the diagnosis of BPD, and the social construction of diagnosis,
particularly the diagnostic entity labeled “Borderline Personality Dis-
order.” It attempts a synthesis of diverse, multidisciplinary literature to
address the question of why women outnumber men by a ratio of 3:1 in
the diagnosis of BPD. It rests on assumptions that to varying degrees,
sociocultural factors inevitably play a role in the expression of disease
conditions and that personality disorders, including BPD, have cultural
histories. It also rests on the belief, for which there is considerable schol-
arly support, that the phenomenon called “Borderline Personality Disor-
der” has multiple, complex, interactive, biological, psychological, and
constructed sociocultural determinants.

LITERATURE REVIEW

Epidemiology of BPD

Etiology

The etiology of BPD has been the focus of clinical interest for

approximately 30 years. The literature related to its determinants is vo-
luminous and was systematically reviewed by Zanarini and Frankenburg
(1997) and Zanarini (2000), who discussed six main conceptualizations
of the term “Borderline;” outlined the seminal theories of its pathogene-
sis; reviewed previously studied etiological factors; organized the litera-
ture into first-, second-, and third-generation etiological studies; and pro-
posed a multifactorial model of the complex etiology of BPD consistent
with their conclusion that individuals follow “a unique pathway to the de-
velopment of BPD that is a painful variation on an unfortunate but famil-
iar theme” (Zanarini & Frankenburg, p. 93). They describe (a) Kernberg’s
(1975) conception of the term “borderline” as a level of personality dis-
organization descriptive of the most serious form of character pathology;

8

P. Bjorklund

(b) Gunderson’s (1984) conceptualization of “borderline” as a specific
personality disorder distinguishable from other DSM-III and DSM-III-
R personality disorder diagnoses; (c) “borderline” as a schizophrenia
spectrum disorder characterized by fleeting psychotic symptomatol-
ogy; (d) “borderline” as an affective spectrum disorder characterized
by chronic dysphoria and affective dysregulation; (e) “borderline” as
an impulse spectrum disorder that shares with other DSM-IV impulse
control disorders a propensity for potentially self-damaging acting-out;
and (f) “borderline” as a chronic form of post-traumatic stress disorder
(PTSD) evidenced by the near-universal finding of trauma (broadly de-
fined) in the early backgrounds of those diagnosed with BPD (Zanarini
& Frankenburg).

Zanarini (2000) noted that attempts to explain the development of

BPD initially came from the psychoanalytic community and focused
on the theories of Kernberg (1975), who proposed that an excess of
early aggression due to either innate temperament or intense environ-
mental frustration resulted in the child’s developmental failure to in-
tegrate disparate images of self and other into a more realistic self-
concept; Adler and Buie (1979), who suggested that problems in early
care-giving resulted in the child’s failure to develop a constant view
of self and others that afforded comfort and sustenance during times
of stress; and Masterson (1971, 1988), who suggested that fear of
abandonment resulting from the withdrawal of environmental support
for (innate and inevitable) maturation (e.g., a parent’s distress at her
child’s autonomous strivings) is the essential feature of borderline psy-
chopathology. As a result, the first-generation studies of the pathogen-
esis of BPD focused on the subtle childhood phenomena described
by these theories, including parental separation or loss and disturbed
parental involvement (Zanarini). Second-generation studies of environ-
mental precursors built on the methodological weaknesses of the first
and discovered that histories of childhood physical and sexual abuse
are common in those diagnosed with BPD, physical abuse is reported
just as frequently in control groups, and sexual abuse is consistently
and significantly reported more often by those diagnosed with BPD
than by those diagnosed with depression or other personality disorders
(Zanarini). Third-generation studies all shared conceptual and method-
ological features, the most important of which were a focus on a range
of difficult childhood experiences, including but not limited to sexual
abuse; the exploration of the parameters of sexual abuse; and the use
of multivariate analyses in the determination of significant findings.
Important findings included that BPD was discriminated more by non-
caregiver sexual abuse than caregiver sexual abuse; those diagnosed with

Bias in Diagnosing Borderline Personality Disorder

9

BPD reported more severe forms of sexual abuse; and sexual abuse typ-
ically took place in an environment of bi-parental neglect and abuse
(Zanarini).

Given that most victims of sexual abuse are female, these etiological

findings may partially explain the gendered nature of BPD. What seems
clear, based on the research literature to date, is that a multifactorial
model of the etiology of BPD best captures its complexity: “This model
suggests that BPD symptomatology and its comorbid manifestations are
the end product of a complex admixture of innate temperament, difficult
childhood experiences, and relatively subtle forms of neurologic and
biochemical dysfunction (which may be sequelae of these childhood
experiences or [of] innate vulnerabilities)” (Zanarini, 2000, pp. 98–99).
What this model neglects is any suggestion of the socially constructed
aspects of diagnosis and illness (Brown, 1995).

Prevalence

In addition to etiology, the concept of prevalence (i.e., the rate of

cases during a particular time frame) is important to an epidemiological
understanding of a disorder. The epidemiological data on the prevalence
and incidence (i.e., the rate of new cases in a population) of border-
line personality disorder is very limited. Widiger and Weissman (1991)
reviewed the existing epidemiological literature on BPD and found a
prevalence rate between 0.2 and 1.8% in the general population, along
with a prevalence rate of 15% among psychiatric inpatients and 50%
among those psychiatric inpatients with a diagnosis of personality dis-
order. Their meta-analysis suggested that 76% of patients with BPD are
female. The DSM-IV-TR (APA, 2000) also reports a 3:1 sex ratio for
BPD, but it finds a higher prevalence (20%) of BPD in psychiatric inpa-
tients as well as a higher prevalence (2%) in the general community. In
addition, it reports that in the United States, BPD is seen in about 10% of
all clients treated in outpatient mental health settings—which is higher
than the 8% of all outpatients reported by Widiger and Sanderson (1997).

It is not clear if the actual prevalence of BPD is increasing over time.

The high prevalence of BPD compared to other personality disorders
in clinical samples—27% of outpatients with a diagnosis of personality
disorder (Widiger & Sanderson, 1997)—is striking given that BPD is far
from the most common personality disorder found in community sam-
ples. Samuels, Eaton, Bienveu, Brown, Costa, & Nestadt (2002) who
assessed the frequency of DSM-IV and ICD-10 personality disorders in
a community sample between 1997 and 1999 found, for example, that
BPD has a weighted prevalence of 0.5% compared to a weighted preva-
lence of 0.9% for obsessive-compulsive personality disorder, 1.8% for

10

P. Bjorklund

avoidant personality disorder, and 4.1% for antisocial personality disor-
der. Using ICD-10 diagnostic criteria, emotionally unstable (borderline)
personality disorder was uncommon with a weighted prevalence of 0.1%.
It is possible that the over-representation of females in clinical samples,
the 3:1 gender ratio in BPD, and the peculiarly dangerous features of the
disorder (e.g., impulsivity and suicidal and/or self-mutilating behavior)
account for this disparity. The true prevalence of BPD by gender is un-
known. A difference in the prevalence rates of BPD between men and
women can be determined accurately only from community samples;
and to date, there has been only one representative population-based
study in Norway (Torgersen, Kringlen, & Cramer, 2001). It found no
difference in the prevalence of BPD by gender.

It is notable that BPD is most frequently reported in North America,

Europe, and the United Kingdom. There are few reports of BPD from
developing countries (Pinto, Dhavale, Nair, Patil, & Dewan, 2000). Un-
able to mount a full-scale epidemiological study to assess the prevalence
of BPD in India, Pinto et al. studied a population of suicide attempters
in whom BPD was most likely to be diagnosed. They found that 17.3%,
equally divided between men and women, met DSM-IV criteria for BPD.
Unable to estimate its incidence or prevalence, the authors nevertheless
concluded that BPD does, in fact, exist cross-culturally. Others have
found evidence to suggest that personality disorders are universal in
their prevalence (Akhtar, 1995). However, the frequency with which in-
dividuals with such disorders come to clinical attention has cultural vari-
ants. For example, only one to three percent of psychiatric outpatients
in Ethiopia and India are diagnosed with personality disorders while
the corresponding figure for British outpatients is 32% (Khandelwal &
Workneh, 1988). Akhtar suggested that variations in diagnostic practices
and in the help-seeking patterns of the three societies, as well as varying
levels of affluence, optimism, religiosity, psychological awareness, and
medical orientation in the society, were more likely to account for this
finding than actual differences in the incidence of BPD, for which no
data have ever been reported (Widiger & Weissman, 1991). Akhtar also
suggested that a particular culture might facilitate a specific phenotypal
outcome of personality disorder:

Cultures where child rearing typically is suppressive of affects (espe-

cially those related to aggression) and discouraging of individuation fa-
vor schizoid. . . . phenotypal outcomes. . . . [while] cultures that allow freer
discharge of affect and encourage assertion and individuation favor nar-
cissistic, borderline, and paranoid phenomenologies. Even within a par-
ticular form of severe personality disorder, actual symptomatology might

Bias in Diagnosing Borderline Personality Disorder

11

be affected by culturally held value systems. For instance, a grandiose as-
ceticism. . . . is a more frequent accompaniment of narcissistic personality
disorder in the East and a materialistically acquisitive picture more com-
mon in the West. . . . [E]asy access to alcohol, drugs, perverse sexual out-
lets, pornography, and firearms in Western. . . . society tends to render the
overt manifestations of borderline, histrionic, and antisocial personali-
ties dangerous and more flagrant. Less permissive societies yield more
muted manifestations of these disorders. Japanese borderline patients, for
instance, are less often drug and alcohol dependent than their American
counterparts (Akhtar, pp. 27–28).

Even within a particular culture, subgroup differences exist. Hence,

BPD is diagnosed less often in African Americans than in whites (Akhtar,
1995). It would seem that cultural differences, gender ideologies and
stereotypes, diagnostic biases, referral artifacts—indeed, the cultural
construction of psychiatry itself—to some degree all render the diag-
nostic entity called BPD obscure, if not suspect as something other than
social construction.

Gender Bias in BPD

Akhtar (1995) has suggested that cultural factors might underlie the

differences in the seemingly gender-related manifestations and preva-
lence of severe personality disorders. For example, the much lower
prevalence of antisocial personality disorder among women might be
related to the more intense social control of women’s behavior. As so-
cietal attitudes shifted and women were allowed greater freedom and
access to all means of self-expression, the prevalence of antisocial per-
sonality disorder among females increased. Sargent (2003) speculated
that BPD and the bodily self-injury that characterizes it occur in the
context of such gender ideologies, perhaps in class-based, industrialized
societies where the female body is highly commodified. Certainly, there
are gendered expectations concerning both the emotionality and rela-
tional context of BPD. Women are simply thought to be more emotional,
relationally-defined, and relationship-dependent than men. Clearly, gen-
der stereotypes affect women and their experiences (and expression) of
mental illness (Wright & Owen, 2001).

Horsfall (2001) argued that gendered assumptions are embedded in

psychiatric knowledge and that BPD is essentially a gendered construct
arising from a psychiatric classification system that is itself a social
construction. She described the term “gender” as “introduced into com-
mon parlance. . . . to conceptually separate female and male biological
factors (sex) from social and culturally derived behaviors, expressions

12

P. Bjorklund

and roles. . . . pertaining to being a woman or a man in a specific society.
These social differences relate to notions of femininity and masculinity”
(p. 421). She correctly noted that the term “gender” is loosely and
inconsistently used in the health and illness literature (often with refer-
ence to matters of reproduction, sexual orientation, or sexual activity)
as a substitute for “sex” Because gender is a social space and the term
references diverse social factors related to being male or female—and
psychiatric epidemiology reveals both gender and sex differences in the
etiology and prevalence of mental illness—the conflation of terms is
perhaps understandable.

The gendered nature of BPD has aroused much curiosity, and the

body of literature on gender (sex) bias in personality disorder generally
(and BPD particularly) is large. This paper focuses on two relatively
recent discussions of the issue. Widiger (1998) has systematically ex-
amined the ways in which the differential sex prevalence rates for the
DSM-IV personality disorders, three of which (borderline, dependent,
and histrionic) are diagnosed more often in females than males (APA,
2000), could reflect sex bias. He differentiated among, and described
support for, each of these different forms of sex bias, including biased
diagnostic constructs, biased thresholds for diagnosis, biased population
sampling, biased application of diagnostic criteria, biased instruments
of assessment, and biased diagnostic criteria.

Bias in diagnostic constructs (and diagnostic criteria) occurs when

the constructs are themselves sexist characterizations of females (or the
criteria involve gender-related behaviors that can complicate equal ap-
plication of diagnostic criteria to males and females). Widiger (1998)
reviewed data that indicated the DSM personality disorders do include
gender-related traits. However, this does not imply that the DSM-IV
construct is not valid or reliable. For example, in a rigorous study of the
prototype validity of the BPD construct, Johansen, Karterud, Pedersen,
and Falkum (2004) found high validity for the borderline construct as
a prototypical diagnostic category—as opposed to a classical diagnos-
tic category, which requires both necessary and sufficient criteria and
homogeneous membership. Prototype categories have dimensional as-
pects and heterogenous membership, which means that members of a
prototype category can be more or less like the prototype and that it is
therefore valid to say that a person has many or few borderline char-
acteristics (criteria) or that he does or does not meet criteria for BPD.
The implication of this “prototype constructivist position” is that “true
BPD” does not exist as such in the real world and that “a person who
fulfils all nine BPD criteria is no truer borderline than a person with a
certain combination of five criteria. All that can be said with certainty

Bias in Diagnosing Borderline Personality Disorder

13

is that the first person complies with the full prototype, while the other
does not” (Johansen et al., p. 290). This study confirmed the prototype
nature of the DSM-IV borderline construct. It concluded that the BPD
diagnostic category has high validity as a prototype and that only minor
revisions to the DSM-IV criterion hierarchy are required. The authors do
not explain, of course, why females outnumber males by a ratio of 3:1
within this (valid) prototype category.

Clearly, even if personality disorders exist with differential sex preva-

lence rates consistent with the normative differences between men and
women, diagnosis would still be biased if different thresholds are used
for male-typed (e.g., antisocial) versus female-typed (e.g., borderline
or histrionic) personality disorders, or if diagnostic criteria in clinical
practice are differentially applied (Widiger, 1998). Widiger reviewed ev-
idence for both types of bias, as well as evidence for bias in the diagnostic
instruments used in BPD research. Sampling bias presents a particularly
potent source of gender bias in BPD. It is possible that the perception of
differential sex prevalence rates in personality disorder is an artifact of
the higher rate of females in clinical settings: “If 75% of the persons at a
clinic are female, then one would expect, by chance alone, that 75% of
the persons with any particular disorder will also be female. Therefore,
it might not be particularly meaningful to find that 75% of the persons
with a. . . [borderline] personality disorder are female” (Widiger, p. 105).
Nevertheless, the simple fact that more females than males are present
at any particular clinic does not imply that there are more females than
males in every disorder, including BPD, diagnosed at that clinic (Widiger
& Weissman, 1991). Ultimately, the differential sex prevalence rate of
BPD can only be determined accurately from studies that use probability
samples of community populations, of which there has been only one to
date (Skodol & Bender, 2003; Torgersen et al., 2001).

Skodol and Bender (2003) reviewed Widiger’s (1998) discussion of

possible sources of sex bias in diagnoses of personality disorder and
completed an updated, comprehensive review of the extant empirical re-
search on the pronounced 3:1 gender (sex) ratio in BPD. Conceding that
the true prevalence of BPD is unknown, they framed the essential ques-
tion as whether the higher rate of BPD in females is due to sampling or
diagnostic bias, or due to biological or sociocultural differences between
the sexes. They cited five empirical studies that used semi-structured di-
agnostic interviews to test for gender differences in DSM III-R and DSM
IV personality disorders, only one of which found that the rate of BPD
differed by gender. They also cited one study that showed BPD was one
of several personality disorders that occurred more often among men.
They speculated that the elevated base rate of women in clinical settings

14

P. Bjorklund

may be the reason why clinicians perceive more women to have BPD.
Ultimately, they concluded that the modest empirical support for diag-
nostic biases of various kinds would not account for the wide difference
in prevalence between males and females and that the differential gen-
der prevalence of BPD in clinical settings is largely a function of biased
sampling. They left a wide opening for the possibility that biological
and/or sociocultural factors may account for the gender difference if, in
fact, the “true prevalence” of BPD in the community is eventually found
to differ by gender (Skodol & Bender).

Social Constructivism and BPD

Social Construction of Diagnosis

While asserting that health and illness are often more affected by

political, economic, and cultural factors than by biomedical ones, Brown
(1995) took a social constructivist approach to diagnosis and illness that
left room for the importance of both biological and sociocultural factors
in the social discovery of disorders. He maintained that we must take
a critical look at the biomedical world (and the biological processes
and disease taxonomies so essential to it) and not view that world as
a mere epiphenomenon: “It is quite possible to believe that biomedical
components are important, while still emphasizing social forces as well
as people’s interactive definition-making. . . . We are, after all, talking
about phenomena which occur in people’s bodies” (p. 37).

Brown (1995) noted the centrality of the notion of the social construc-

tion of disease and illness as an organizing theme in medical sociology.
He described the social construction of a disorder as a process of ex-
amining how social forces shape our understanding of it: “We explore
the effects of class, race, gender, language, technology, culture, the po-
litical economy, and institutional and professional structures and norms
in shaping the knowledge base which produces our assumptions about
the prevalence, incidence, treatment, and meaning of disease” (p. 34). In
addition, he systematically examined three competing versions of this
social process. Briefly, the first argues that social problems, for exam-
ple, the problematic behaviors of a person diagnosed with BPD, are
matters of social definition. According to a strict social constructionist
approach, BPD is not “real” in any objective sense but rather is created
by the purposive action of “social labelers and problem finders” (p. 35).
Rather than focusing on either the degree of “reality” in phenomena, or
how specific social structures (e.g., gender or culture) shape the meaning
of phenomena, the second version of social constructivism focuses on

Bias in Diagnosing Borderline Personality Disorder

15

signifiers. In order to show how knowledge of a (psychiatric) category
is created, it deconstructs the language and symbols that represent an
entity like borderline personality—for example, the words “border” or
“wilderness” or “No man’s land.” The third version of social construc-
tivism argues that the production of scientific “facts” (e.g., the BPD
diagnostic category) is the result of “science in action” (p. 36) —a con-
cept that conveys the ongoing, everyday activity of scientists engaged
in efforts to research, disseminate, and promote their work in public
and official venues (e.g., APA work groups or annual meetings). This
accumulation of “evidence” creates scientific “facts-on-the-ground,” so
to speak, including, for example, that BPD not only exists but has (at
least some) biological basis.

According to Brown’s (1995) framework, the social construction of

BPD would involve multiple social forces that combined in various ways
to create, modify, and re-create the phenomenon: “Rather than a given
biomedical fact, we have a set of understandings, relationships, and ac-
tions that are constantly in flux” (p. 37). Brown’s framework includes:
a real condition (disorder or disease entity) that is either accepted or
not accepted as a biomedical category, and a biomedical definition that
is either applied or not applied. The ways these four elements inter-
act describe the manner in which a condition is socially constructed:
(a) there are routinely defined conditions that are usually accepted and
to which biomedical definitions are applied; (b) there are medicalized
definitions in cases where a condition is not accepted (or is consid-
ered non-medical) but a biomedical definition is nevertheless applied;
(c) there are contested definitions in situations where generally accepted
conditions have no widely applied biomedical definition; and (d) there
are potentially medicalized definitions for conditions not yet medically
accepted and to which biomedical definitions have not yet been ap-
plied (e.g., genetic predispositions to disease) (Brown, 1995). In essence,
Brown noted the complex and changing nature of diagnosis. Since so-
cial construction is ongoing, the definition of a condition can shift, or
a condition may have multiple definitions. There are some diagnoses,
like BPD, that can be variously defined and constructed at a single point
in time. Brown’s framework is helpful in understanding, for example,
some of the tension between biological psychiatry, which construes BPD
as a routinely defined condition; and psychoanalysis, which views it as
medicalized.

Social Organization of Psychiatric Knowledge

In a strict approach to social construction, even the statistics that pro-

vide evidence for the prevalence of BPD, or the gender bias associated

16

P. Bjorklund

with it, are suspect; for the statistics themselves are socially constructed
(Brown, 1995). Smith (1990), for example, discoursed on the social
organization of objectified forms of knowledge as integral to the power
relations of contemporary societies. She investigated the statistics on
gender differences in rates of mental illness in Canada and the United
States as an example of this form of socially organized, objectified
knowledge concealing power relationships in patriarchal societies. She
critically analyzed the procedures used to create statistics that pointed to
gender differences in rates of mental illness, and she explored the social
relations and embedded power differentials that produce such statistics.
Moving through and behind possible explanations why the Canadian
figures do not show the same kind of relationship between gender (sex)
and mental illness as do those in the United States, Smith argued that
the process of becoming mentally ill is one in which psychiatric organi-
zations participate: “Put simply, in terms of the statistical information,
this means that when you seem to be counting people becoming men-
tally ill you are in fact also counting what psychiatric agencies do. The
two aspects can’t be taken apart. The figures can’t be decontaminated”
(p. 117). With respect to BPD and the statistics on its prevalence, one
cannot therefore separate the numbers from the actions of an institution
(i.e., psychiatry) that “acts to control people who have come to be seen
as breaching, disrupting, or disorganizing the everyday/everynight and
taken-for-granted accomplishments of a recognizable world” (Smith,
p. 133).

Cultural Construction of Psychiatric Classification

Numbers are meaningless without categories in which to place them,

and cases of mental illness cannot be identified without psychiatric clas-
sifications to define their borders. Presumably, the ongoing process of
psychiatric classification reflects advancing professional knowledge of
mental illness. However, Gaines (1992) argued that psychiatric classifi-
cation is culturally constructed and expresses an underlying ethnopsy-
chology (i.e., cultural psychology) of the ideal self. His constructivist
interpretation of psychiatric classification revealed “the nosological en-
terprise” as a discourse that “embodies and expresses the central concep-
tions of [this] ethnopsychology” (p. 3). His anthropological perspective
is not essentially Foucaultian, for he did not argue that language (dis-
cursive space that locates political power) creates the observed. Rather,
he asserted that “medical classificatory systems” are attempts to articu-
late during particular historical times “a particular culture’s unconscious
ethnopsychological assumptions” (p. 4) about the ideal self, Other, and
the modes of experience and activity that constitute abnormality. Gaines’

Bias in Diagnosing Borderline Personality Disorder

17

fascinating analysis of the voice, or vantage point of DSM classification
showed that the ideal self is gender-(male), ethnic-(German Protestant),
and age-specific (adult), as well as essentially biological. His essay also
showed that current conceptions of mental disorder as biologically based,
including BPD, are not new. Rather, they are rooted in history and cul-
ture, especially the history and culture of science.

Cultural History of Personality Disorder

In an attempt to provide an anthropological perspective on the border-

line classification, Sargent (2003) noted the dearth of ethnopsychiatric
literature that bears directly on the personality disorders. She noted an
important exception in Nuckolls’ (1992) account of the cultural his-
tory of antisocial and histrionic personality disorder. Nuckolls argued
that these personality disorders have cultural histories that “represent in
extreme form values and attitudes strongly congruent with familiar cul-
tural stereotypes: The ‘independent’ male and the ‘dependent’ female”
(p. 37). Nuckolls attempted to interpret the vast disparity in numbers
of men and women diagnosed as “antisocial” and “histrionic” in terms
of Weber’s (1958) analysis of Western capitalism—wherein the ideal
self is constructed as a combination of religious moralism and worldly
materialism:

‘Moralism’ refers to a constellation of values centered on faith in God
and in one’s own ‘election’ to spiritual salvation. ‘Materialism’ refers to a
belief that materially successful work in a God-given ‘calling’ is indica-
tive of spiritual election. Combined, these values created a personality
strongly predisposed to hard work, professionalism, investment, and de-
ferred gratification—the very qualities Weber believed were necessary to
the formation of western-style capitalism (Nuckolls, pp. 37–38).

Nuckolls (1992) proposed that contradictions in moral and materialist

values (intolerable within the self) were partially resolved by keeping
them separated and projecting them outward onto delegate social groups
such that women became delegates for values associated with moralism
and men became delegates for values associated with materialism. Char-
acteristic behavioral styles then emerged: Women were prone to “spiritu-
ality, illness and hysterical display,” and men were prone to “worldliness,
aggression, and shrewd calculation” (p. 46). The masculine and femi-
nine behaviors characteristic of men and women diagnosed respectively
as antisocial and histrionic followed as a logical extension of this psy-
chological and social process. Although Nuckolls thus historicized the
emergence of certain gender-specific behavioral styles, as well as the
evolution of psychiatric diagnoses that classify extreme manifestations

18

P. Bjorklund

of these (gendered) behavioral styles, Sargent (2003) concluded that the
intersection of gender and the behaviors characteristic of those diagnosed
with BPD remains an intriguing but under-explored phenomenon: “It ap-
pears that the [cultural] history of this constellation of symptoms and
behaviors [is yet] to be written” (p. 26). She did not take into account the
work of Wirth-Cauchon (2001) whose scholarly treatise on the social
construction of BPD provides, in essence, another cultural history of a
gendered “disorder.”

Social Construction of Borderline Personality Disorder

Wirth-Cauchon (2001) asserted that cultural, historical, and linguistic

analyses provide a framework for understanding BPD as a phenomenon
with cultural significance: “The symptoms . . . grouped together under
the borderline label—fragmented or unstable identity, feelings of empti-
ness or numbness, depersonalization, self-mutilation—may be meaning-
fully understood as exaggerated or extreme forms of some of the cultural
contradictions of gender in late modern society, as fault lines of a cul-
tural order in which the contradictions are visible in the moment of
breakdown of the feminine subject” (p. 30), which is not to say she does
not view the suffering and instability of the borderline subject as real.
She acknowledged the distress as “real” but considered the diagnosis
fictive, even though it has become a fact of psychiatric classification;
and she drew upon feminist analyses of gender and subjectivity as a
perspective from which to view borderline symptomatology in a socio-
cultural context—“as a response to women’s position at cultural borders”
(p. 168). Women positioned at this border live in “borderland territory,”
which is a “place of struggle over meanings” (p. 168). In essence, Wirth-
Cauchon argued that the borderline—as metaphor, diagnosis, and/or de-
scriptor for certain kinds of selves is a “site of contention, controversy,
and struggle over boundaries” (p. 3). It demarcates the boundaries of the
self, the boundaries of madness (where the “borderline” is the bound-
ary between the “psychotic” and the “normal”), the boundaries between
categories of disorder, and the boundaries around psychiatry’s limits.
Whereas Skodol and Bender (2003), for example, left little room for
the socially constructed aspects of BPD as illuminating in the search
to understand the gender differences in the diagnosis, Wirth-Cauchon
left little room for the empirical evidence that explains these gender dif-
ferences as anything other than “psychiatric text,” as discursive space
that shows where women cross the boundaries of psychiatrically con-
structed notions of normal femininity. Indeed, if Wirth-Cauchon’s social
constructivist position was strictly construed, it could be justifiably crit-
icized for “[exemplifying] a style of thought that directly undercuts its

Bias in Diagnosing Borderline Personality Disorder

19

aims. . . [in stripping women] of their power to ask for help” (Luhrmann,
2002, p. 259).

CONCLUSION

Reviewing the extant empirical research on epidemiology, etiology,

and gender (sex) bias in BPD presents one approach to understanding
why women are more often diagnosed with BPD than men. While it
provides valuable leads to further investigation (e.g., cross-cultural epi-
demiological studies of BPD prevalence in community samples, contin-
ued investigation into possible biological and sociocultural etiologies of
BPD, and especially continued investigation into the differential effects
of trauma across gender as an etiological factor in BPD), at present this
approach fails to fully answer the question. We do not yet know for sure
why females outnumber males by a ratio of 3:1 in the diagnosis of BPD.
Until we can determine the true prevalence of BPD in community sam-
ples, eliminate sampling bias in epidemiological research on BPD, and
thus rule out being female as a predisposing factor for BPD, we cannot
also rule out the likelihood that the gendered nature of societies—which
can range in many cases from invisibly to blatantly sexist and unsafe for
at least some vulnerable women—contributes to the labeling of females
as “borderline” more often than males. Examining the sociological and
anthropological literature on the processes whereby (and the contexts in
which) new diagnostic categories are “discovered” and applied (to some
persons but not others) presents another route to understanding the phe-
nomenon. This approach can mitigate against further medicalization of
the “borderline condition.” It also can mitigate against a narrow focus
in terms of policy, resource allocation, and clinical management on bi-
ological causes and cures or skills-based symptom-management at the
expense of normalizing approaches that also focus on the narrative con-
struction and reconstruction of the meaning of being borderline in social
contexts.

The clinical work nurses and their “borderline” patients do in con-

structing and reconstructing such meaning is essential to patient recovery
and to the eventual integration and consolidation of a healthy identity
other than that of “patient” or “borderline.” For example, when we ask a
“mentally ill” woman diagnosed with BPD to take responsibility for her
behavior, for her self—for a life worth living, in essence—what are we
asking her to do? How is she to construe her task(s), and how does her
diagnosis impinge on this construal? In part, the answers to these ques-
tions depend on how a society constructs the meaning, consequences,
and normative obligations that attach to the borderline diagnosis. In other

20

P. Bjorklund

words, our understanding of our responsibilities (and hence our identi-
ties) and how we are to take them depends to some degree on how others
signal who one is (or is not), what one is (or is not) supposed to do, and
who one can (or cannot) become (Bjorklund, 2005). The meaning of
being borderline in this society cannot help but enter into any treatment
relationship. It matters greatly to patients diagnosed as borderline how
their mental health providers see them, that is, whether they construe
the “borderline” diagnosis in terms of frontal lobe dysfunction, skills
deficits, sex bias in diagnostic criteria and/or research, or in terms of a
certain gendered mode of being sick in early 21st century industrialized
societies—or all of the above. Interventions are shaped by these social
constructions.

While it by no means provides an exhaustive review of the available

literature, this paper begins to address sociological and anthropological
issues that can lead to a more nuanced understanding of the gendered,
historical, and cultural nature of the borderline diagnosis, as well as
to policies and treatment approaches that depathologize the borderline
condition. It attempts a complex integration of sociological and anthro-
pological literature with that on the epidemiology and etiology of BPD.
In doing so, it departs from empirical terrain and journeys to more the-
oretical turf. Roughly, it travels in ever smaller concentric circles—like
the pond ripples of a stone’s throw in reverse—from a broad discussion of
the nature of social construction, to the nature of psychiatric knowledge
as socially organized, to the culturally constructed nature of psychiatric
classification, and on to examples of the cultural construction of two
personality disorder categories, ending narrowly with a discussion of
the social construction of the borderline personality disorder diagnosis
in particular. It arrives at the conclusion that only a thoroughly mul-
tidisciplinary, methodologically diverse, and empirically-theoretically
integrated approach to understanding the gender differences in the bor-
derline diagnosis is credible with phenomena as complex as BPD. It is
clear that nurses and other mental health providers must understand
the phenomenon at this level of complexity to provide appropriate
care.

REFERENCES

Adler, G., & Buie, D. (1979). Aloneness and borderline psychopathology: The possible

relevance of child developmental issues. International Journal of Pschoanalysis, 60,
83–96.

Akhtar, S. (1995). Quest for answers: A primer of understanding and treating severe

personality disorders. Northvale, NJ: Jason Aronson.

Bias in Diagnosing Borderline Personality Disorder

21

American Psychiatric Association. (2000). Diagnostic and statistical manual of mental

disorders (4th ed., text rev.). Washington, DC.

American Psychiatric Association. (2003). Practice guideline for the treatment of pa-

tients with borderline personality disorder. Available: http://www.psych.org/clin
res/borderline.index.cfm

Bjorklund, P. (2005). Taking responsibility: Towards an understanding of morality in

practice; A critical review of the empirical and (selected) philosophical literature
on the nature and social organization of responsibility. Manuscript submitted for
publication.

Brown, P. (1995). Naming and framing: The social construction of diagnosis and illness.

Journal of Health and Social Behavior, 35(Extra Issue), 34–52.

Busfield, J. (1989). Sexism and psychiatry. Sociology, 23(3), 343–364.
Cloninger, C. R. (2004). Psychobiology and treatment of borderline personality disorder.

Acta Neuropsychiatrica, 14(2), 60–65.

Fonaghy, P., & Roth, A. (1996). What works for whom? A critical review of psychotherapy

research. New York: Guilford.

Gaines, A. D. (1992). From DSM-I to III-R; Voices of self, mastery and the other: A

cultural constructivist reading of U.S. psychiatric classification. Social Science and
Medicine, 35(1), 3–24.

Gunderson, J. G. (1984). Borderline personality disorder. Washington, DC: American

Psychiatric Press.

Gunderson, J. G. (2001). Borderline personality disorder: A clinical guide. Washington,

DC: American Psychiatric Press.

Horsfall, J. (2001). Gender and mental illness: An Australian overview. Issues in Mental

Health Nursing, 22, 421–438.

Johansen, M., Karterud, S., Pedersen, G., & Falkum, E. (2004). An investigation of the

prototype validity of the borderline DSM-IV construct. Acta Psychiatrica Scandinav-
ica, 109, 289–298.

Kernberg, O. F. (1975). Borderline conditions and pathological narcissism. Northvale,

NJ: Jason Aronson.

Kernberg, O. F. (1996). A psychoanalytic theory of personality disorders. In J. F. Clarkin

& M. F. Lenzenweer (Eds.), Major theories of personality disorder (pp. 106–140).
New York: Guilford.

Khandelwal, S. K., & Workneh, F. (1988). Psychiatric outpatients in a general

hospital of Ethiopia. The International Journal of Social Psychiatry, 34, 230–
235.

Koerner, K., & Dimeff, L. A. (2000). Further data on dialectical behavior therapy.

Clinical Psychology: Science and Practice, 7(1), 104–112.

Luhrmann, T. M. (2002). Women and borderline personality disorder (book review).

American Journal of Sociology, 108(1), 259–261.

Masterson, J. F. (1976). Psychotherapy of the borderline adult: A developmental ap-

proach. New York: Brunner/Mazel.

Masterson, J. F. (1988). The search for the real self: Unmasking the personality disorders

of our age. New York: Free Press.

Masterson, J. F. (2000). The personality disorders: A new look at the developmental self

and object relations approach. Phoenix, AZ: Zeig, Tucker, & Co.

22

P. Bjorklund

Millon, T. (2000). Sociocultural conceptions of the borderline personality. The Psychi-

atric Clinics of North America, 23(1), 123–136.

Nathan, P. E., & Gorman, J. M. (Eds.). (1998). A guide to treatments that work. New York:

Oxford University Press.

Nuckolls, C. (1992). Toward a cultural history of the personality disorders. Social Science

and Medicine, 35, 37–49.

Paris, J. (1993). The treatment of borderline personality disorder in light of research on

its long-term outcome. Canadian Journal of Psychiatry, 38(1), S28–S34.

Pilgrim, D. (2001). Disordered personalities and disordered concepts. Journal of Mental

Health, 10(3), 253–265.

Pinto, C., Dhavale, H. S., Nair, S., Patil, B., & Dewan, M. (2000). Borderline personality

disorder exists in India. The Journal of Nervous and Mental Disease, 188(6), 386–
388.

Rogler, L. H. (1997). Making sense of historical changes in the Diagnostic and Statis-

tical Manual of Mental Disorders: Five propositions. Journal of Health and Social
Behavior, 38(1), 9–20.

Samuels, J., Eaton, W. W., Bienvenu III, O. J., Brown, C. H., Costa, P. T., & Nestadt, G.

(2002). Prevalence and correlates of personality disorders in a community sample.
British Journal of Psychiatry, 180, 536–542.

Sargent, C. (2003). Gender, body, meaning: Anthropological perspectives on self-injury

and borderline personality disorder. Philosophy, Psychiatry, & Psychology, 10(1),
25–27.

Siever, L. (2003, November 19). New perspectives on neurobiology of personality dis-

orders. Grand rounds presentation sponsored by the Department of Psychiatry, Uni-
versity of Minnesota, Minneapolis, MN.

Skodol, A. B., & Bender, D. S. (2003). Why are women diagnosed borderline more than

men? Psychiatric Quarterly, 74(4), 349–360.

Skodol, A. E., Siever, L. J., Livesley, W. J., Gunderson, J. G., Pfohl, B., & Widiger, T. A.

(2002). The borderline diagnosis II: Biology, genetics, and clinical course. Biological
Psychiatry, 51(12), 951–963.

Smith, D. E. (1990). The statistics on women and mental illness: The relations of ruling

they conceal. In D. E. Smith (Ed.), The conceptual practices of power: A feminist
sociology of knowledge (pp. 107–138). Boston: Northeastern University Press.

Swartz, M., Blazer, D., George, L., & Winfield, I. (1990). Estimating the prevalence of

borderline personality disorder in the community. Journal of Personality Disorders,
4, 257–272.

Teicher, M. H., Andersen, S. L., Polcari, A., Anderson, C. M., Navalta, C. P., & Kim,

D. M. (2003). The neurobiological consequences of early stress and childhood mal-
treatment. Neuroscience and Biobehavioral Reviews, 27, 33–44.

Torgersen, S., Kringlen, E., & Cramer, V. (2001). The prevalence of personality disorders

in a community sample. Archives of General Psychiatry, 58, 590–596.

Weber, M. (1958). The Protestant ethic and the spirit of capitalism. New York: Scribners.
Widiger, T. A. (1998). Invited essay: Sex biases in the diagnosis of personality disorders.

Journal of Personality Disorders, 12(2), 95–118.

Widiger, T., & Sanderson, C. (1997). Borderline personality disorder. In A. Tasman, J.

Kay, & J. Lieberman (Eds.), Psychiatry (pp. 1302–304). Philadelphia: W.B. Saunders.

Bias in Diagnosing Borderline Personality Disorder

23

Widiger, T., & Weissman, M. (1991). Epidemiology of borderline personality disorder.

Hospital and Community Psychiatry, 42, 1015–1021.

Wirth-Cauchon, J. (2001). Women and borderline personality disorder. New Brunswick,

NJ: Rutgers University Press.

Wright, N., & Owen, S. (2001). Feminist conceptualizations of women’s madness: A

review of the literature. Journal of Advanced Nursing, 36(1), 143–150.

Zanarini, M. C. (2000). Childhood experiences associated with the development of

borderline personality disorder. The Psychiatric Clinics of North America, 23(1),
89–101.

Zanarini, M. D., & Frankenburg, F. R. (1997). Pathways to the development of borderline

personality disorder. Journal of Personality Disorders, 11(1), 93–104.