Relationship Between Dissociative and Medically Unexplained Symptoms in Men and Women Reporting Childhood Sexual Abuse

; Maynes

and Feinauer

). Empirical studies comparing levels

of dissociation between samples of traumatized and non-
traumatized individuals show that traumatic experiences
are associated with higher level of dissociation (Putnam
and Carlson

). Nijenhuis et al. (

) found that

patients with dissociative disorders reported exposure to
more traumatic events, particularly childhood sexual abuse
(CSA). In addition, individuals diagnosed with dissocia-
tive identity disorder (formerly multiple personality dis-
order) have been shown to report high rates of CSA. For
example, Putnam et al. (

) found that 83% of patients

with multiple personality disorder reported a history
of CSA.

Briere and Runtz (

) found higher rates of dissoci-

ation and somatization in a non-clinical sample of women
who reported being sexually abused than in those without a
sexual abuse history. In a similar study, Zlotnick et al.
(

1996

) utilized the Dissociative Experiences Scale (DES)

scale (Bernstein and Putnam

) to assess dissociative

J Fam Viol (2008) 23:569

–575

DOI 10.1007/s10896-008-9175-6

S. N. Gold (

S. A. Ketchman

I. Zucker

M. A. Cott

A. H. Sellers
Center for Psychological Studies, Nova Southeastern University,
3301 College Avenue,
Fort Lauderdale, FL 33314, USA
e-mail: gold@nova.edu

symptoms and the Somatization Scale of the Symptom
Checklist 90-Revised (SCL-90-R; Derogatis

) to assess

somatization symptoms. The sample included 74 inpatient
women who had reported experiencing sexual abuse before
the age of 16. Their findings indicated a significant
difference in the rates of somatization and dissociation
between participants who reported a history of sexual abuse
and those who did not. A study by Atlas et al. (

1995

) found

that male and female adolescents who reported physical,
sexual, or both types of abuse showed more elevated levels
of dissociation than a comparison group. In addition,
somatization scores were shown to be associated with
dissociation scores (Pearson r=0.34, p=0.03, one tailed).
Saxe et al. (

) examined somatization in patients with

dissociative disorders in an inpatient psychiatric hospital.
They compared a high dissociative group with a low
dissociative group (based on DES scores) and found that
64% (n=9) of the participants in the high dissociative group
also met criteria for a somatization disorde. In contrast,
none of the patients in the low dissociative group met
criteria for somatization disorder.

Pribor et al. (

) examined whether participants who

met criteria for the somatoform disorder termed Briquet

’s

syndrome (see Feighner et al.

1972

, for a description of this

disorder) reported sexual or physical abuse. Their sample
consisted of female participants attending an outpatient
medical clinic at a university. They found that 90% of
women diagnosed with Briquet

’s syndrome reported some

type of abuse during childhood or adulthood.

The current study investigates the relationship be-

tween the dissociative and somatic symptoms in a
clinical sample of adults reporting CSA. The purpose
of this study is to explore whether medically unexplained
somatic symptoms are associated with dissociative
symptoms in adult CSA survivors. More specifically,
we sought to test the hypothesis that somatization is an
integral component of dissociation by assessing the
magnitude of relationship between these two variables.
It should be noted, however, that in contrast to the studies
by Nijenhuis et al. (e.g.

), which specifically investi-

gated somatoform dissociative symptoms (e.g. anesthesia,
motor inhibitions, etc.), this study examined medically
unexplained somatic symptoms in general. In addition, it
has been shown that on the whole women report more
somatic symptoms than men (Klonoff et al.

2000

; Myers

et al.

1984

; Russo and Sobel

1981

; Shek

1989

). However,

there is minimal research concerning somatic symptoms in
male victims of childhood sexual abuse, or comparing
their magnitude in men and women survivors. In an
attempt to rectify this situation, therefore, this study exam-
ines the relationship of somatization and dissociation in
male as well as female victims of CSA and compares them
to each other.

Materials and Method

Participants

Participants were 216 women and 35 men entering an
outpatient treatment program for CSA survivors within a
university-based community mental health center. All
participants were over the age of 16, reported having been
sexually abused before age 18, and evidenced psycholog-
ical difficulties consistent with an abuse history.

Demographics

Participants ranged in age from 17 to 58, with women
having a mean age of 33.99 (SD=9.57) and men having a
mean age of 35.23 (SD =9.10). The majority of the sample
identified themselves as Caucasian (80%), with the remain-
der reporting their ethnic backgrounds as Hispanic (9%),
African American (3%), or other (8%). The average level of
education was 12.79 years (SD =2.41) for men and 12.43
years (SD=2.46) for women. Of the total number of
participants, 43% reported being unemployed, 37% were
employed full-time, 19% worked part-time, and two
participants were retired. An annual income of less than
$10,000 was reported by 41%, 29% reported an income
between $10,000 and $19,999, 17% reported an income
between $20,000 and $29,000, and 11% reported an
income of $30,000 or more. No significant differences
were identified between men and women on any of these
demographic variables.

Participants

’ reported age of onset of abuse ranged from

1 to 17 years, with a mean age of 6.44 (SD=5.18). Average
number of perpetrators was 2.93 (SD= 2.31), a mean
duration of abuse by the first perpetrator of 5.18 years
(SD=5.23). Men and women did not differ on any of these
abuse characteristics.

Materials

The Dissociative Experiences Scale The DES is a widely
used screening instrument designed to evaluate dissociation
(Bernstein and Putnam

). The scale consists of 28

items that describe dissociative experiences ranging from
the normal to the pathological. Participants are asked to rate
the percentage of time from 0 to 100 that each dissociative
experience occurs. Extensive research on the scale has
demonstrated that it has good internal consistency, test

–

retest reliability, and criterion, construct, and discriminate
validities (Bernstein and Putnam

; Carlson et al.

;

Frischholz et al.

1990

Multiple investigations conducted on the DES factor struc-

ture using both clinical and non-clinical samples have pro-
duced three-factor solutions constituting the following realms

570

J Fam Viol (2008) 23:569

–575

of dissociation: (a) amnesia, (b) absorption/imaginative
experiences, and (c) derealization/ depersonalization (Carlson
et al.

; Ross et al.

1991

; Sanders and Green

). Using

our clinical sample of CSA survivors these three factors were
replicated and used in the present study (Gold et al.

1997

Further, an eight-item subscale of the DES derived through
taxometric method (DES-T) has been developed to identify
pathological dissociation (Waller et al.

1996

Symptom Checklist-90-Revised The SCL-90-R is a self-
report checklist consisting of 90 symptoms. Participants are
asked to rate on a five point scale (0=not at all, 4=
extremely) the frequency with which symptoms occurred in
the last week. The SCL-90-R has demonstrated adequate
validity and reliability (Derogatis

; Kaplan and Miner

). An additional dissociation subscale consisting of

five items was developed for the SCL-90-R format (Briere
and Runtz

). Research revealed that the subscale was

adequately reliable (

α=0.76) and differentiated levels of

dissociation among abused and non-abused outpatients
(Briere and Runtz

1987

Minnesota Multiphasic Personality Inventory-2 The Min-
nesota Multiphasic Personality Inventory-2 (MMPI-2;
Hathaway and McKinley

1989

) was administered in its

entirety to all participants. More specifically, the Harris

–

Lingoes

“somatic complaints” subscale of the Conversion

Hysteria (Hy) scale was examined. The test

–retest reliabil-

ity and the internal consistency for all the scales have been
shown to be adequate for both men and women (Graham

; Hathaway and McKinley

1989

). The HEA content

scale demonstrated both convergent and discriminant
validity in studies conducted on chronic pain patients
(Butcher et al.

1990

). However, the validity of the other

scales appears more questionable. The Harris

–Lingoes

subscales lack empirical evidence to substantiate their
validity, while previous research on the basic clinical scales
has indicated that these scales do not specifically measure
the syndromes after which they are named (Graham

Procedure

At the time of their initial intake evaluation, participants
completed the DES, SCL-90R, and MMPI-2. Consent to
participate in research and demographic information was
also obtained at that time.

Data Analysis

DES total score was derived by averaging responses to all
completed items. Factor scores were obtained by summing
the items that loaded on each factor and dividing by the

appropriate number of items in each subscale (Gold et al.

1997

). DES-T scores were obtained by averaging responses

to eight DES items that assess pathological dissociation
(see Waller et al.

1996

for taxon items). SCL-90-R and

MMPI-2 scale scores were obtained through adherence to
standard scoring instructions.

Pearson product

–moment correlations were computed

for each gender on all measures (DES total score, DES-T,
DES amnesia factor, DES absorption factor, DES dereal-
ization/depersonalization, SCL-90-R somatization scale,
SCL-90-R dissociation scale, Hypochondriasis (Hs) scale,
Conversion

–Hysteria (Hy) scale, Harris–Lingoes Somatic

Complaints subscale, and Health Concerns (HEA) content
scale). Correlations between the DES total score, DES-T,
and all MMPI-2 basic clinical scales were also computed
for both men and women. As a consequence of missing
data on each measure, sample size for the correlations
ranged from 107 to 210 for the women and from 23 to 33
for the men. Due to the large number of analyses
conducted, the alpha level for all analyses was set at 0.01.

Results

Since analyses were conducted separately on men and
women, the means and standard deviations of all measures
are reported by gender in Table

. A significant difference

between men and women was found on the SCL-90-R
somatization scale (t=

−4.066, df=240, p<0.01), with men

reporting statistically significant less somatic complaints
than women. No other significant differences were found
on any of the measures.

Correlations between all the measures are reported in

Table

. In the table, correlations for men are shown above

the diagonal of the correlation matrix while the correlations
for the women are below the diagonal. Significant
correlations (ranging from r=0.191 to 0.470) were found
between dissociative and somatization scores on all the
measures for the women in the sample. No significant
correlations between dissociation and somatization scores
were found for the men.

As seen in Table

, the correlations between the dis-

sociation and somatization scales were significant for the
women, but their magnitudes were only moderate. Further-
more, among women, correlations of the DES total score
and DES-T with all ten MMPI-2 basic clinical scales except
Mf achieved significance. Conversely, the same correlations
were not significant among men. However, tests of dif-
ferences between dependent correlations (Steiger

1980

)

among both men and women revealed that the correlation
of DES-T with MMPI-2 scale 8 was significantly higher
than the correlations of DES-T with MMPI-2 scales 1 and 3
(p<0.01). Additionally, among women only, the correlation

J Fam Viol (2008) 23:569

–575

571

of the DES-T with scale 8 was significantly higher than that
of DES-T with HEA, and the correlation of DES-T with
scale 7 was significantly higher than that of DES-T with
scale 3. Tests of differences between independent correla-
tions (i.e., men vs. women; Hays

1997

) found no significant

differences, except for the correlations of MMPI-2 scale 1
with DES total score, DES-T, and DES factor 1 (p<0.01).

Discussion

Our findings do not support the hypothesis that medically
unexplained somatic symptoms in general are an integral
component of dissociation. Somatization and dissociation
scale scores were related, but not at a great enough
magnitude to suggest that they are part of the same
syndrome. In fact, we found that the Sc scale was more
strongly related to dissociation than any of the MMPI-2
somatization scales in both men and women. Additionally,
in women the Pt scale was more strongly related to
dissociation than any of the MMPI-2 somatization scales.
These findings are in contrast to those of Nijenhuis et al.
(

), which reported that psychological dissociation and

somatoform dissociation are highly integrated among
sexual abuse survivors. This difference in findings between
our study and theirs highlights the importance of the
distinction between somatization in general and somato-
form dissociation in particular.

With the exception of one of the measures of somatiza-

tion, the SCL-90-R somatization subscale, no significant
differences were found between men and women CSA
survivors in symptom levels. This is consistent with the
findings of previous studies that suggest that gender
differences in symptom levels found in other populations
are absent or much less pronounced in clinical samples of
CSA survivors (Briere et al.

; Gold et al.

1999

; Hunter

1991

). This finding underscores that the results of this study

are only generalizable to outpatient clinical samples of adult
childhood sexual abuse survivors. However, one interesting
implication of this observation is that it contradicts the
historical presumption that somatization is more character-
istic of women than of men.

To investigate why most of the correlation coefficients

did differ between genders, t-tests were conducted on
simulated correlations of various magnitudes with different
sample sizes. We found that correlations of a magnitude of
0.4 or less (i.e., similar to those produced by the women in
this study) do not achieve significance if the sample size is
comparable to that of the men in our study. Conversely,
many of the correlations of a magnitude of 0.18 or higher
(i.e., similar to those produced by the men in this study) do
not achieve significance if the sample size is comparable to
that of the women in our study. Since both genders

’

coefficients were small to moderate, differences between
these relatively low magnitude correlations are neither
statistically significant nor clinically meaningful.

Table 1 Means and standard
deviations and comparisons
between genders on selected
variables

Briquet

’s disorder is character-

ized as a polysymptomatic dis-
order that begins early in life,
affects women predominantly
or exclusively, and is associated
with chronic and recurrent
somatic complaints, often
described dramatically and
without medical explanation.
Conversion sx are also very
common in patients with this
disorder. DSM-III-R listed this
condition as

“somatization

disorder

”—focusing on the

medical complaints
*p<0.01

Variable

Males

Females

Mean

DES

20.94

18.05

212

24.72

18.92

−1.10

DES Fact. 1

9.47

15.67

212

10.43

16.11

−0.33

DES Fact. 2

26.11

20.66

212

31.60

23.05

−1.32

DES Fact. 3

22.32

19.62

212

26.89

21.30

−1.19

DES Taxon

14.16

18.02

212

17.53

19.03

−0.98

SCL Diss.

1.32

0.92

203

1.60

1.01

−1.55

SCL Som.

0.85

0.79

203

1.58

0.93

−4.35*

MMPI Hs

62.77

15.40

139

69.50

14.32

−2.17

MMPI Hy

61.85

14.33

139

69.55

14.00

−2.56

MMPI Hy4

63.61

15.55

110

70.93

16.15

−1.99

MMPI HEA

67.04

13.51

110

71.78

14.74

−1.42

MMPI L

49.85

10.01

139

51.86

10.02

−0.94

MMPI F

83.35

27.77

139

83.19

24.78

0.03

MMPI K

41.89

9.51

139

41.07

8.75

0.43

MMPI D

76.31

17.12

139

76.52

14.43

−0.07

MMPI Pd

72.61

13.87

139

75.51

12.95

−1.03

MMPI Mf

60.81

9.54

139

49.88

9.96

5.17*

MMPI Pa

70.58

21.83

139

74.19

16.72

−0.96

MMPI Pt

76.42

16.13

139

74.68

15.51

0.52

MMPI Sc

79.85

21.17

139

80.07

18.59

−0.06

MMPI Ma

56.92

13.99

139

57.61

11.60

−0.27

MMPI Si

65.96

11.96

139

62.25

11.74

1.47

572

J Fam Viol (2008) 23:569

–575

able

Correlations

between

selected

variables

for

male

(upper

triangle)

and

female

(lower

triangle)

survivors

childhood

sexual

abuse

DES

DES1

DES2

DES3

DEST

SCLD

SCLS

HL9

HEA

DES

–

0.89*

0.94*

0.95*

0.79*

0.28

−

0.17

−

0.12

0.1

0.00

−

0.25

0.31

−

0.46

−

0.22

−

0.10

0.31

0.08

0.27

0.16

DES1

0.84*

–

0.79*

0.80*

0.91*

0.72*

0.19

−

0.30

−

0.24

0.03

−

0.08

−

0.09

0.18

−

0.32

−

0.28

−

0.29

−

0.05

0.23

−

0.06

0.10

0.21

0.1

DES2

0.94*

0.71*

–

0.80*

0.85*

0.77*

0.40

−

0.09

−

0.05

0.08

0.02

−

0.25

0.26

−

0.37

−

0.15

0.03

0.13

0.22

0.13

0.27

0.25

0.1

DES3

0.93*

0.70*

0.79*

–

0.92*

0.73*

0.16

−

0.1

−

0.07

0.19

0.07

−

0.29

0.37

−

0.54*

−

0.18

−

0.1

0.13

0.41

0.09

0.31

0.23

0.19

DEST

0.93*

0.83*

0.78*

0.93*

–

0.74*

0.19

−

0.19

−

0.18

0.17

0.03

−

0.21

0.35

−

0.47

−

0.17

−

0.15

0.09

0.35

0.1

0.29

0.20

0.21

SCLD

0.68*

0.52*

0.63*

0.66*

0.68*

–

0.41*

0.03

0.10

0.16

0.12

−

0.13

0.17

−

0.36

−

0.06

−

0.15

0.17

0.21

0.06

0.23

0.13

SCLS

0.46*

0.39*

0.46*

0.40*

0.56*

–

0.46

0.32

0.41

0.46

0.07

−

0.01

−

0.20

0.01

−

0.02

0.27

−

0.03

0.19

0.20

0.33

−

0.06

0.40*

0.36*

0.37*

0.33*

0.34*

0.41*

0.71*

–

0.84*

0.76*

0.91*

0.08

0.49*

−

0.06

0.60*

0.49*

0.23

0.39

0.61*

0.59*

0.13

0.04

0.23*

0.24*

0.22*

0.17

0.18

0.31*

0.60*

0.85*

–

0.63*

0.72*

−

0.05

0.48

−

0.10

0.64*

0.53*

0.13

0.41

0.56*

0.54*

0.07

0.01

HL9

0.44*

0.40*

0.41*

0.38*

0.36*

0.45*

0.71*

0.84*

0.78*

–

0.91*

−

0.14

0.60

0.46

0.44

0.21

0.14

0.44

0.57*

0.63*

0.15

0.22

HEA

0.44*

0.40*

0.41*

0.38*

0.43*

0.73*

0.91*

0.76*

0.90*

–

−

0.04

0.63*

−

0.26

0.49

0.33

0.22

0.49

0.60*

0.70*

0.23

0.10

−

0.09

0.02

−

0.12

−

0.1

−

0.07

−

0.13

−

0.08

−

0.00

0.03

−

0.03

–

−

0.31

0.44

−

0.01

−

0.28

0.22

−

0.16

−

0.24

−

0.29

−

0.30

−

0.02

0.54*

0.45*

0.50*

0.49*

0.51*

0.58*

0.42*

0.60*

0.64*

−

0.18

–

−

0.56*

0.52*

−

0.07

0.86*

0.73*

0.91*

0.37

0.40

−

0.39*

−

0.23*

−

0.39*

−

0.39*

−

0.34*

−

0.42*

−

0.38*

−

0.15

−

0.05

−

0.37*

−

0.35*

−

0.47*

–

−

0.19

−

0.24

0.13

−

0.53*

−

0.39

−

0.58*

−

0.43

−

0.53*

0.30*

0.22*

0.34*

0.22*

0.21*

0.41*

0.59*

0.64*

0.62*

0.60*

0.61*

−

0.12

0.53*

−

0.30*

–

0.72*

−

0.09

0.33

0.75*

0.63*

−

0.19

0.58*

0.23*

0.17

0.24*

0.19

0.17

0.23*

0.37*

0.45*

0.41*

0.37*

0.38*

−

0.22*

0.53*

−

0.17

0.60*

–

−

0.23

0.37

0.76*

0.64*

0.14

0.35

0.14

0.16

0.09

0.13

0.12

0.08

0.10

0.05

0.09

0.12

−

0.03

0.37*

−

0.17

0.08

0.14

–

−

0.03

−

0.12

−

0.06

−

0.14

−

0.03

0.39*

0.31*

0.36*

0.31*

0.37*

0.50*

0.44*

0.49*

0.51*

−

0.12

0.70*

−

0.34*

0.55*

0.53*

0.33*

–

0.55*

0.78*

0.35

0.39

0.50*

0.40*

0.49*

0.43*

0.55*

0.57*

0.66*

0.61*

0.65*

−

0.18

0.65*

−

0.26*

0.80*

0.61*

0.12

0.62*

–

0.86*

0.20

0.53*

0.60*

0.51*

0.57*

0.54*

0.53*

0.58*

0.63*

0.69*

0.57*

0.71*

−

0.19

0.84*

−

0.39*

0.66*

0.64*

0.27*

0.73*

0.86*

–

0.46

0.47

0.38*

0.30*

0.34*

0.38*

0.34*

0.26*

0.17

0.24*

0.1

0.19

0.24*

−

0.13

0.34*

−

0.26*

−

0.02

0.27*

0.17

0.23*

0.40*

–

−

0.18

0.36*

0.31*

0.38*

0.27*

0.32*

0.42*

0.44*

0.34*

0.24*

0.48*

0.45*

−

0.12

0.53*

−

0.45*

0.67*

0.36*

0.1

0.44*

0.66*

0.58*

−

0.14

–

0.01

J Fam Viol (2008) 23:569

–575

573

Because we used the MMPI-2 to measure somatization

and other symptom patterns, our findings must be inter-
preted with considerable caution. Since early days of the
development of the MMPI, it has been well known that
the basic scales of the instrument do not measure the
diagnostic patterns for which they are named (Graham