4 Our-—

‘4‘:

 

‘rr- ._. w‘- ._.__

351%

{L
W: 1:
I

:‘3’
flu
gypl’fg 7-,...

 

' 7‘..’j,3.-,". .v V,
‘ - “"7!“ 35.3%:
"3““?
z j gfifflqfi-
u
n :
A «2'2: ’
4! ‘- v.

u ‘“ ' ‘ m. - 1 ‘
AT“??? ,’S&,‘,_ V 52'34‘ .3 V. g. V \ V I; ‘P‘ ‘§)_m“’)9
1 ' ‘7. z s 14;“ ;

‘ ‘ . {12.7, a .
I . . l ‘Uc ‘ .
W51} .

J'V
~ , 5:733.“ ~
’ ”1' W- 4 ,.. .,.; . . i... , . ,.
'mnfln - ' 3" ‘ ' ‘ M" -.*~ . . o _. ~. - .
‘w'ég9§;gxt ‘u‘ .I .y " - r) '1'”; ." '15.? hi 1': ’M‘ ”(‘1 ' ' , '. ‘ c ‘. v ‘5 ’ 3‘ . ‘; .HV Fout‘ vtk'fi. . r
H" S -_' 3w” . . .1! 2:, 4,90%“, H ‘9’ ,fl ...,_ {k ‘ . ’_ _ . u ._ ‘
. .‘njnfl \' 1:. :3 fl Q":- 4!‘ , _- , 3 _ > '_ i- " : ‘ ‘ y u ' t ,‘C
. ' ‘, 4"“; I v 9‘ ‘ - “3;? 3\ x1 7‘
4.3:“ fit'fl: 1::~ I:
‘-~-'~~r"-'<~i§s'::§5§‘:» _
1i 1.. ,4: 5.9'
“L"; ‘8 .
. (V131,-

v
N' - -‘ .
. u. :1
a WW“ m I"
' ‘1 \ V .T
x

v . . m ‘ . Edi, 5Q
‘ ; b . v . ‘ . _ 7 ,4“ fl»... ‘5.
‘ v ~ 1‘ ‘. . w v; g ‘
‘1}:"J4 , ". . ' - "3-" ‘ ' . a ' fi‘ydg'yrh
L : »" '. - ‘” J _ ,- -- ’ “' 5:9 .
. V“, : . "a
‘1.
‘N

. kfg‘!‘;?'\.:.
. ~.4fi$?§:g{:‘h:~
xmifiw.‘ ‘ ..
31W: A:

‘ a: J’:
é &»~ - Jul-Ta ”IT. “3‘ k\‘
I '21“: “Mwfi V: r w?"
:W'fi: {913; 1”,: “@224: .
:1.*:.f:£..ev'<g2:~;g§3:zgg-Nfiz§. 1": '1?
“J’s” 4:?">:i"'v :11. 1 4.:

1:» ,
"1
'3:

#319 ~ V _:
- ‘ 1 £51.
at 51:3 fia‘l’ffix HLI'
: 53%.): _ .
M3,"_ 3: .
, 3°?” 1.:
{"53“, L 1 ..
.::'=:‘1:::;

,
‘11-“.5" ilkmfu
‘ wk - .

.
v w”): “ u'fl
Ms *1 1 .
Jir‘ér‘igdg'igdfi'fi
W“ F "35"“ _ 1‘

. v .
$4.,“ ‘5 ,.
»\ it“s. .' )’f‘..
If?" '4." ' ‘2‘"er
n,- _ ,,.. ’ -.
.‘ d; jg;- :34;
"0.. . 5: .
3‘ 31%"! ."
’f’ .; ,

J
7.? :2
A“ 1,

,3 .
:fi’v'
A. '-”1-’
fl

N! w .-

"" M" ~ .ka ; , ;
nun,“ ,o‘ u . ' I ~ ‘
’31 .I'J.‘ ' 4- . A. ’ ‘ as» ,-;;;
«3?; ; . If,“ ‘ .1' j , J34}.

1E???»

a

' ' u
. , , ,
. 7:311:31!
N. ‘3 H
w; ..
99:.“ “b
.

“L:
.u

 

 

I I
s .
,:
M."

z :62: L s ‘80
IHIIIHIIHIIIIIHIHHHIHIIHIII!IIIIHIHJHHJIIHUIHII

300793 5046

This is to certify that the

dissertation entitled

The Relationship Between
Somatization Disorder
And Concomitant Psychiatric Syndromes

presented by
Patrice Elizabeth Gerard

has been accepted towards fulfillment
of the requirements for

Doctor ofi Philosophy degree in Counseling Psychology

)

31:”de [\x/FKUOCO
J Major pififessor

‘\

Date Visa. 5; mac)

MS U is an Affirmative Action/Equal Opportunity Institution

 

 

 

 

F’ —'\

LIBRARY
Michigan State
University

 

 

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.

DATE DUE DATE DUE DATE DUE

 

 

W4

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MSU Is An Affirmative Action/Equal Opportunity lnditution
extrempma-pd

 

THE RELATIONSHIP BETWEEN
SOMATIZATION DISORDER
AND CONCOMITANT PSYCHIATRIC SYNDROMES
BY

Patrice Elizabeth Gerard

A DISSERTATION

submitted to

Michigan State University
in the partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Counseling, Educational
Psychology and Special Education

1990

ABSTRACT

THE RELATIONSHIP BETWEEN SOMATIZATION DISORDER
AND CONCOMITANT PSYCHIATRIC SYNDROMES

BY

Patrice Elizabeth Gerard

The relationship between somatization disorder and
additional Axis I and Axis II diagnoses was investigated in
a sample of 25 patients, referred because they were thought
to suffer from somatization disorder. Each individual was
assigned multi-axial DSM—III-R diagnoses using both the
Structured Interview for DSM-III-R (SCID) and the Millon

Clinical Inventory, second edition (MCMI-II).

As hypothesized, those individuals receiving a
somatization disorder diagnosis evidenced a higher rate of
personality disorders (Axis II syndromes) than expected in
the overall p0pulation. In addition, the hypothesis
predicting that the somatization disorder group would
evidence at least one other Axis I diagnosis was supported.
The hypothesis predicting that antisocial and histrionic
personality disorders would be the most frequently occurring
personality disorders among the somatization disorder group
was not supported. Instead, dependent, borderline, and
passive-aggressive were the most frequently assigned
personality disorders. The significance of these findings
was obscured by the low number of somatization disorder

diagnoses assigned in the study.

Patrice Elizabeth Gerard

Post hoc analyses demonstrated that the entire sample
was noteworthy in terms of number of assigned Axis I and
Axis II diagnoses. The subjects receiving somatoform
diagnoses were indistinguishable from the rest of the
sample in terms of these Axis I and Axis II disorders.
Analysis of the instruments used in the study (SCID and
MCMI-II) indicated a low level of concordance between the
two, despite the fact that they were both designed to

measure DSM-III-R clinical syndromes.

Patrice Elizabeth Gerard

Post hoc analyses demonstrated that the entire sample
was noteworthy in terms of number of assigned Axis I and
Axis II diagnoses. The subjects receiving somatoform
diagnoses were indistinguishable from the rest of the
sample in terms of these Axis I and Axis II disorders.
Analysis of the instruments used in the study (SCID and
MCMI-II) indicated a low level of concordance between the
two, despite the fact that they were both designed to

measure DSM-III-R clinical syndromes.

To my husband, Kenny

iv

ACKNOWLEDGEMENTS

I wish to express my appreciation to Dr. John Powell,
who served as chairman of my dissertation committee. His

gentle support and kindly direction was invaluable.

I also want to thank the members of my committee, Dr.
Steve Geiger for his statistical expertise, Dr. Richard
Johnson for his overall support, and Dr. Richard Prawat for
his creativity and humor. Finally, special thanks go to Dr.
Lesley Jones who drove me "mad" with her critiques, but
helped me produce a polished product. Her energy and
enthusiasm was boundless, and made my task seem more

achievable.

To my typist, editor, and friend, Tammy WOten, thank
you for standing by me through the entire process. Your

loyalty and maturity are appreciated.

My largest debt of gratitude goes to my husband, Kenny.
Without your unwavering love and support, this undertaking
could never have been completed. I look forward to getting

to know you again.

And, finally, a special thank you to my father, Dr. Roy
Gerard, who has always been a model for perseverance and
achievement, despite the odds.

V

TABLE OF CONTENTS
Page
LIST OF TABLES ...................................... viii
LIST OF APPENDICES .................................. x

CHAPTER I: Introduction .........................
Historical Perspective .........................
Prevalence and Gender Distribution .............
Statement of the Problem .......................
Need for the Study ............................. 15
Theory ......................................... 18
Summary ........................................ 21
Purpose of the Study ........................... 21
Research Questions ............................. 22

CDNINH

CHAPTER II: Review of the Literature ............. 23
Somatization Disorder and Personality Patterns . 23
Somatization Disorder and Hysterical/Histri-
onic Personality Patterns ............... 24
Somatization Disorder and Antisocial
Personality Patterns .................... 31
An Overview of Personality Disorders ........... 38
Millon's Theory of Personality ................. 45
Research Employing the MCMI and the MCMI-II .... 49
Structured Clinical Interview for DSM—III-R
History and Evolution ...................... 54
Summary ........................................ 56

CHAPTER III: Methodology .......................... 58
Sampling Procedures ............................ 58
Procedures for Data Collection ................. 59
Measures ....................................... 60

Millon Clinical Multiaxial Inventory-II .... 6O
MCMI-II Revision Rationale ............. 63

Norms .................................. 64
Reliability ............................ 65
Internal Validity ...................... 68
External Validity ...................... 70
Structured Clinical Interview for DSM-III-R 74
Reliability ............................ 77
Validity ............................... 78
Hypotheses ..................................... 79
Design ......................................... 81
Data Analysis .................................. 81

CHAPTER IV: Analysis ............................... 83
Major Hypotheses ............................... 83
Post Hoc Analyses .............................. 87
Summary ........................................ 104

vi

CHAPTER V:

Summary and Conclusions .................

Review of the Study ............................
Conclusions Regarding Major Hypotheses .........
Post Hoc Analyses ..............................
Sampling Limitations ...........................
Measurement Limitations ........................
Clinical Implications ..........................
Future Research Directions ....................

APPENDICES

REFERENCES

vii

108
108
109
118
124
125
127
128

133
144

Table

1.11
1.12

3.14

3.15

3.17
4.11

4.12

LIST OF TABLES

Briquet's Syndrome OOOOOOOOOOOOOOOOOCOOOOOOO

Lifetime prevalence of additional psychia-
tric syndromes of 78 female psychiatric
outpatients with Briquet's Syndrome ........

Ontological evolution of personality
disorders OOOOOOOOOOOOOIO'OOCOOOOOOOOOOOOOOOO

Prevalence of DSM-III personality disorders
in eight Studies 0......OOOOOOCOOOOOOOOOOOOO

Scales of the Millon Clinical Multiaxial
Inventory-II O0.000000000000000000000000..

Demographic and other features of the
nomative population OOOOOOOOOOOOOOOOOOOOOOO

MCMI-II scale stability coefficients for
outpatient and inpatient populations at
different phases in the course of their
disorders ..................................

Internal consistency estimates of MCMI-II

scales 0.0.0.0...0......OOOOOOOOOOOOOCOOOOOO

MCMI-II classification efficiency of MCMI-II
two highest DSM-III-R diagnosed Axis II
disorders 0.00000COOOOOOOOOOOOOOOOOO00......

Comparison of MCMI-I and MCMI-II two highest
(Axis II) or BR>74 (Axis I) sensitivity
and PPP measures of DSM-III disorders ......

SCID mOdules O...OOOOOOCOOOOOOOOOOOOOOO0....

Distribution of Axis II diagnoses
(personality disorders) for subjects
diagnosed with somatization disorder .......

Distribution of Axis I diagnoses for
subjects receiving a somatization disorder
diagnOSis O0......COO...OOOOOOOOCOOOOOOOOOOO

Distribution of Axis I diagnoses for the
entire sample O...0.0...OOOOOOOOOOOOOIOOOOO

Frequency distribution of Axis I diagnoses
for the entire sample .....................

viii

Page

13

40

44

62

66

67

69

71

73

76

86

88

89

91

Table Page

4.15 Distribution of Axis II diagnoses (person-
ality disorders) for the entire sample .... 92

4.16 Frequency distribution of Axis II diagnoses
for the entire sample ......OOOOOOOOOOOOOIO 93

4.17 T-tests comparing the number of Axis I and
Axis II diagnoses for those diagnosed
with and without somatization disorder .... 95

4.18 Distribution of Axis I diagnoses for
subjects receiving a general somatoform
diagnOSj-S ......OOOOOOOOOOCCC......OOOOOOOO 96

4.19 Distribution of Axis II diaganoses (person-
ality disorders) for subjects receiving
a general somatoform diagnosis ............ 98

4.20 T-tests comparing the number of Axis I and
Axis II diagnoses for those diagnosed with
and without general somatoform disorder ... 99

4.21 Distribution of Axis II diagnoses (person—
ality disorders) as measured by the SCID .. 101

4.22 Distribution of Axis II diagnoses (person-
ality disorders) as measured by the MCMI—II 102

4.23 Distribution of Axis II diagnoses (person-
ality disorders) on the SCID compared
With the Millon .........OOOOOOOOOOOOOOOOO 103

5.11 Distribution of somatoform diagnoses by
referral source ......OOOOOOOOO......OOOOOO 112

5.12 Axis II (personality disorder) frequencies

for the MCMI-II based on two different
cutting points, compared with the SCID .... 123

ix

LIST OF APPENDICES

Page
APPENDIX A
DSM—III-R CRITERIA FOR SOMATIZATION DISORDER ... 133
APPENDIX B
HEALTH CARE PROVIDER LETTER .................... 136
APPENDIX C
DEMOGRAPHIC AND OTHER IDENTIFYING FEATURES OF
THE SAMPLE 0.0.0.0....0.000000000000000......... 139
APPENDIX D
INFORMING AND PARTICIPATION REQUEST ............ 140
APPENDIX E
DEPARTMENTAL AND PARTICIPATION REQUEST ......... 142
APPENDIX F
TAPING, INFORMING AND PARTICIPATION REQUEST .... 143

CHAPTER I

Introduction

 

Somatization, a tendency to experience and communicate
somatic distress in response to psychosocial stressors, is a
significant health care problem that straddles the
borderland between psychiatry and medicine (Lipowski, 1988).
Because of their recurrent, multiple and unexplained
physical symptoms, somatizing patients have long posed
diagnostic and therapeutic challenges to health care
providers (Smith, 1985).

Somatization disorder is one of a group of disorders
termed "Somatoform disorders" in the Diagnostic and
Statistical Manual III (DSM-III) and Diagnostic and
Statistical Manual III, Revised (DSM-III-R) nomenclature
(American Psychiatric Association, 1980 and 1987). The
essential feature of this grouping is the presentation of
physical symptoms suggesting physical disorders for which
there are no demonstrable organic findings (Lloyd, 1986).
The sub-categories of Somatoform disorder are:

1. Body Dysmorphic disorder

2. Conversion disorder

3. Hypochondriasis

4. Somatization disorder

5. Somatoform Pain disorder

6. Undifferentiated Somatoform disorder

7. Somatoform disorder not otherwise specified.

1

2

The concept of somatization disorder has evolved over
the past three decades and incorporates some prior uses of
the terms "Hysteria" and "Briquet's syndrome" (Smith,
Monson, and Livingston, 1985). The primary feature of
somatization disorder is the presence of recurrent and
multiple somatic complaints of several years duration, for
which medical attention has been sought, but which do not
derive from a specific physical disorder. In order to
fulfill the criteria for the diagnosis of somatization
disorder, the patient must have at least 13 of 25 symptoms.
These symptoms must begin before the age of 30. In
addition, they must require the patient to take medication
other than aspirin, to change his or her pattern of living,

or to consult a physician.

Historical Perspective

 

Syndromes resembling what is termed "somatization
disorder" in DSM-III-R (American Psychiatric Association,
1987) have been noted since early medical and psychological
documentation (Martin, 1988). Controversy has always
existed regarding the implications of this diagnosis and its
relationship to personality variables.

Pre-Hippocrates Egyptians maintained that "wandering
of the uterus" resulted in numerous physical symptoms
(veith, 1965). Hippocrates embraced this notion and
developed it, suggesting a uterine etiology for many

symptoms to which the adjective "hysterical" was attached

3
(Martin, 1988). Over time the uterine theory fell out of
favor but the term "hysteria" endured. The first
systematic evaluation of hysteria was reported by Briquet in

his 1859 Treatise on Hysteria. Despite the movement away

 

from the previously noted anatomical hypothesis, hysteria
became linked with personality variables distinctly separate
from a tendency toward dramatic and excessive somatic
complaints (Martin, 1988). Mai and Merskey (1980) noted
Briquet's emphasis on personality traits such as lability of
mood, intense reactive emotionality, over-sensitivity and
liveliness. These traits were thought to be antecedents to
somatic symptoms. Briquet maintained that "brain neurosis”
and poor physical health also were etiological factors in
the development of "hysteria"; however, attention to
personality variables came to supersede the somatic facet of
the originally described disorder (Martin, 1988).

In the past four decades several studies have
attempted to refine and clarify the diagnosis of ”hysteria."
Purtell, Robins, and Cohen (1951) used the term "hysteria"
to denote a disorder of dramatic nature, with an onset early
in life and marked by many medically unexplained complaints.
The diagnosis was limited to women. Despite reference to
the friendly, verbose and dramatic manner of these patients,
personality factors were not part of the diagnostic
criteria.

Chodoff and Lyons (1958) emphasized that hysteria

denoted "a pattern of behavior exhibited by certain

4
individuals said to be hysterical personalities or
characters" (p. 1401). They also linked hysteria with
conversion reactions, anxiety neurosis and "a particular
psycho-pathological pattern" (p. 1402).

DSM-II (American Psychiatric Association, 1968) used
the term "hysterical neurosis" to describe what had
previously been referred to as "hysteria," thus
distinguishing it from hysterical personality which was
characterized by "excitability, emotional instability,
over-reactivity and self-dramatization." Hysterical
neurosis was marked by "psychogenic loss or disorder of
function." Anxiety was noted as the main symptom.

Prior to the publishing of DSM-II (American
Psychiatric Association, 1968), researchers in the field
were seeking to re-establish a medical complaint-based
definition of hysteria (Martin, 1988). Perley and Guze
(1962) refined the concept of hysteria by requiring at least
29 unexplained symptoms in at least nine of 10 symptom
groups. Personality factors were only described as they
pertained to the manner in which the patient reported his or
her medical history. The culmination of this movement
occurred in 1970 when Guze suggested calling "hysteria"
"Briquet's syndrome" in order to isolate it from other
concepts of hysteria. He also intended to avoid the
pejorative associations derived from the hysterical
personality concept. The diagnosis required a complicated

and dramatic medical history beginning before age 30 that

5
consisted of the presence of a minimum of 25 of 59 medically
unexplained symptoms in at least nine of 10 symptom groups
(see Table 1.11).

In 1980, the Task Force on Nomenclature and Statistics
of the American Psychiatric Association separated hysteria
from conversion, psychogenic pain and hypochondriacal
disorder and called it "somatization disorder" (American
Psychiatric Association, 1980). The diagnostic criteria
which comprised Briquet's syndrome were revised as follows:

1. The total symptom list was reduced from 60 to 37
symptoms and all depressive symptoms were
eliminated.

2. Ten symptom groups were reduced to seven and the
requirement that the symptoms be distributed over
groups was dropped.

3. The total number of symptoms required for a
positive diagnosis were reduced from 25 to 14

symptoms in women and 12 symptoms in men.

The current DSM-III-R diagnosis of somatization
disorder (see Appendix A) further reduces the required
number of symptoms to 13 for both men and women. In
addition, the presence of the following seven symptoms

constitutes a screening index for the disorder:

6
Table 1.11

Briquet's Syndrome, Perley and Guze (1962, p. 421)

Group 1 Group 6
Headaches Abdominal pain
Sickly Vomiting

Group 2 Group 7
Blindness Dysmenorrhea
Paralysis Menstrual irregularity
Anesthesia Amenorrhea
Aphonia Excessive bleeding
Fits or convulsions
Unconsciousness Group 8
Amnesia Sexual indifference
Deafness Frigidity
Hallucinations Dyspareunia

Urinary retention
Trouble walking
Other conversion symptoms

Group 3

Fatigue

Lump in throat
Fainting spells
Visual blurring
Weakness
Dysuria

Group 4

Breathing difficulty
Palpitation

Anxiety attacks
Chest pain

Dizziness

Group 5

Anorexia

weight loss

Marked fluctuations in weight
Nausea

Abdominal bloating

Food intolerances

Diarrhea

Constipation

Other sexual difficul-
ties

Vomiting during 9
months of pregnancy
or hospitalized for
hyperemesis gravi-
darum

Group 9

Back pain

Joint pain

Extremity pain

Burning pains of
sexual organs, mouth
or rectum

Other bodily parts

Group 10

Nervousness

Fears

Depressed feelings

Need to quit working
or inability to
carry on regular
duties because of
feeling sick

Crying easily

Feeling life was
hopeless

Thinking a good deal
about dying

Wanting to die

Thinking suicide

Suicide attempts

7
1. vomiting (other than during pregnancy)
2. pain in extremities
3. shortness of breath when not exerting oneself
4. amnesia
5. difficulty swallowing
6. burning sensation in sexual organs or rectum
(other than during intercourse)
7. painful menstruation
The presence of two or more of these items suggests a high
likelihood of the disorder. As in DSM—III, anxiety and

depressive symptoms are omitted.

Prevalence and Gender Distribution

Estimates of the prevalence of somatization disorder
among women vary from 0.2 to 2% depending on sample
demographic variables (WOodruff, Clayton, and Guze, 1971;
Robins, Helzer, and weissman, 1984). Somatization disorder
is reported to occur far less frequently in men (Smith et
al., 1985). However, Lipowski (1988) cautions that both men
and women can exhibit it. Studies addressing the prevalence
and gender distribution of somatization in general are
consistent with these findings. Slavney and Teitelbaum
(1985) found that patients with psychogenic pain disorder
were equally likely to be male or female. In a study of 139
outpatients referred by the medical and surgical services of
a general hospital for evaluation by a psychiatry unit,

Leon, Saiz-Ruiz, Chinchilla, and Morales (1987) found

8
somatization by females was only slightly greater than
males, and not reaching statistical significance. The
traditional notion of women somatizing more than men has
become so prevalent that even physician's view women as
potential somatizers (Armitage, Schneiderman, and Bass,
1979). Despite the fact that somatization disorder is not a
prevalent condition in the population at large, reviews
suggest that somatization disorder occurs with enough
frequency in family practice settings for the physician to
make the diagnosis on an almost daily basis (Kaplan, Lipkin,
and Gordon, 1988; Like, Rogers, and McGoldrick, 1988). In
fact, de Gruy, Columbia, and Dickinson (1987) have estimated
the prevalence rate for somatization disorder in the primary

care setting may be as high as 5%.

Statement of the Problem

Few descriptive studies have addressed the somatizing
patient yet somatization disorder has a significant impact on
both the patient's quality of life and the primary care
physician's time, energy and morale (Rasmussen and Avant,
1989). The course of somatization is usually chronic,
unrelenting and disabling. Only 31% of patients have been
found to recover at 15-year follow—up assessments (Coryell
and Norton, 1981). Misdiagnosis and the absence of
effective treatment regimes result in unnecessary physician
visits, surgery and medical testing (Katon, 1985). The

failure to accurately diagnose somatization disorder may

9
result in the prescription of unnecessary medications with
possible complications and adverse reactions (Rasmussen and
Avant, 1989). Perley and Guze (1962) note that drug and
alcohol abuse frequently accompany somatization disorder.
Interestingly, patients with somatization disorder are also
subject to an increased number of physical illnesses (Oxman,
Harrigan, and Kues, 1983).

Patients with somatization disorder also suffer
psychological and social consequences as a result of their
illness. Katon, Ries, and Kleinman (1984) state that
somatization disorder can impact family homeostasis,
vocational adjustment, social relations and the patient's
c0ping mechanisms. Zoccalillo and Cloninger (1986) note
that somatization disorder is almost always associated with
chronic and pervasive social disability. They cite a high
incidence of marital discord and divorce noted by
researchers in the field (Robins, 1966; Guze, 1968; and
Guze, 1976). In addition, Zoccalillo and Cloninger have
found that women with somatization disorder are
significantly more likely to abuse or neglect their
children, and do poorly at work.

Patients who continually present with multiple physical
symptoms but no apparent disease take up an excessive
amount of the primary care physician's time and energy.
Regier, Goldberg, and Taublo (1978) and Rankin and Oktay
(1979), in studies sponsored by the National Institute of

Mental Health, demonstrated that 50 to 60% of patients with

10
mental illness are exclusively treated in a primary care
setting. Of that percentage, 25% present with somatic
complaints devoid of organic etiology (de Gruy et al., 1987;
Katon, 1985). Smith (1985) reports that the somatizing
patient accounts for at least 40% of the medical outpatient
population. Recent studies, as previously noted, have put
the prevalence rate for somatization disorder in primary
care settings as high as 5%. Rasmussen and Avant (1989)
note that such a prevalence rate would rank somatization
disorder as the fourth most common disorder encountered in
family practice.

The first major obstacle encountered by the primary
care physician in dealing with somatization disorder is
making the diagnosis (Brown and Smith, 1988). Despite its
reported prevalence in primary care settings, medical
practitioners often fail to recognize this diagnostic
entity. Oxman et a1. (1983) note that reluctance to use
DSM-III terminology often results in the physician's failure
to even consider the disorder. They underscore that
prevention and treatment of somatization disorder
will be tedious unless the primary care physician is given
training in practical and non-stigmatizing means of
diagnosing this disorder. These authors call for increased
psychiatric consultation as well as a comprehensive
biopsychosocial assessment.

Katon et al. (1984) also call attention to the primary

care physician's difficulty in diagnosing somatization and

11
other somatoform disorders. They note that primary care
physicians are taught psychiatric nosology in psychiatric
settings where, for the most part, patients are aware of
psychological precipitants and symptoms. These physicians
usually have less training in diagnosing emotional illness
in patients who present somatically. Katon et a1. decry the
lack of research aimed at describing the prevalence,
incidence and phenomenology of DSM—III type diagnoses within
primary care settings.

The differential diagnosis of multiple unexplained
physical problems can be taxing and time consuming.
Rasmussen and Avant (1989) underscore the need to rule out
organic or physical disorders that also present with
multiple, vague somatic symptoms, such as multiple
sclerosis, systemic lupis erythematosus and
hyperparathyroidism.

Somatization disorder as seen in primary care settings
involves not only physical complaints but also a wide array
of concomitant psychiatric diagnoses. Rasmussen and Avant
(1989) note the following comorbidities:

1. affective disorders (e.g. masked depression)

2. psychosis (e.g. Schizophrenia with multiple

somatic delusions)

3. anxiety disorders (e.g. generalized anxiety

disorders)

4. other Somatoform disorders (Conversion disorder,

Hypochondriasis, Psychogenic pain)

12

5. personality disorders (e.g. Histrionic

personality disorder)

6. adjustment disorders (e.g. Adjustment disorder

with physical complaints)

7. factitious disorders (e.g. Factitious disorder with

physical symptoms)

Liskow, Othmer, Penick, DeSouza, and Gabrielle (1986b)
studied 78 patients diagnosed as having Briquet's syndrome.
Seventy-seven of 78 fulfilled inclusive criteria for one or
more other psychiatric syndromes (see Table 1.12). Thus
even after a physical disorder is ruled out, the
differential diagnosis of somatization disorder may be
complex and tedious due to considerable symptom overlap and
similarity among other psychiatric disorders and
somatization disorder.

Rasmussen and Avant (1989) note that physicians
experience considerable frustration in their attempts to
treat the patient who presents with multiple physical
symptoms but no discernible physical or psychiatric disease.
This type of patient may elicit negative feelings in the
physician who is apt to feel depleted by the patient's
excessive demands for emotional support and medical
attention, especially when positive outcomes are minimal or
nonexistant.

Lichstein (1986) states that in addition to thwarting the
well-intentioned therapeutic efforts of the physicians,
somatizers are less than enjoyable patients because they

challenge the physician's tolerance of diagnostic uncertainty.

13
Table 1.12
Lifetime Prevalence of Additional Psychiatric Syndromes of

78 Female Psychiatric Outpatients with Briquet's Syndrome,
Liskow et. al. (1986b, p. 627).

 

 

Psychiatric Syndrome N %
Organic mental disorder 1 1.3
Alcohol dependence 13 16.7
Drug dependence 18 23.1
Mania 31 39.7
Major depressive disorder 68 87.2
SchizOphrenia 21 26.9
Antisocial personality 13 16.7
Anorexia nervosa S 6.4
Obsessive-compulsive disorder 21 26.9
Phobic disorder 30 38.5
Panic disorder 35 44.9
Mental retardation 1 1.3
Homosexuality 3 3.8

Transsexualism 0 O

14
Katon et a1. (1984) suggest that the medical system may
reinforce somatization. They offer that physicians, by
virtue of their highly technological training,
preferentially look for and treat somatic complaints. Thus,
somatizing patterns can be reinforced by medical concerns
and substantial work-ups.

Smith (1985) posits that poor physician relationships
with somatizing patients are the rule rather than the
exception. Somatizers are not "good patients." They do not
present with a clear cut and treatable organic disease and
they make frequent demands on their physician. In view of
the fact that a positive physician-patient relationship is
crucial to the successful management of somatization
disorder, research aimed at more accurate recognition and
effective treatment of somatizing patients should be a high
priority (Katon, 1985).

Patients who suffer from somatization disorder not only
tax the physician's time and energy, they consume an
inordinate proportion of the health care dollar (Katon et
al., 1984). The per capita expenditure for health care of
patients with somatization disorder is up to nine times the
average per capita amount (Smith et al., 1985). Of
particular significance is this population's excessive use
of inpatient services. Smith et al. (1985) report that
patients with somatization disorder spend an average of 7.6
days in the hospital per year, per person, although the

majority of this hospitalization is unnecessary. In 1981,

15
Cummings and Vanden Bos stressed that the somatizing
patient's overuse of health care services could eventually

lead to bankruptcy in the health care financing system.

Need for the Study

Research aimed at examining the relationship between
somatization disorder and personality disorders is
needed in order to further refine and clarify the diagnosis
of somatization disorder, per se, and to gain increased
understanding of this group of patients with the hOpe of
thereby developing viable avenues of treatment.

Zoccalillo and Cloninger (1986) found the DSM-III
diagnosis of somatization disorder problematic. They argue
that reliance on physical symptoms alone, the renaming of
the disorder, as well as removal of all psychological
symptoms from the diagnostic criteria, obscures the many
other symptoms and problems associated with somatization
disorder. They hypothesize that patients with somatization
disorder will present with many and varied psychological
symptoms and will frequently meet inclusion criteria for
other psychiatric diagnoses, including personality
disorders.

Liskow et al. (1986b) also highlight the need for
studies examining the relationship between somatization
disorder and concurrently occurring psychiatric disorders.

They are especially interested in how these disorders may

16
interact to affect the course, prognosis and treatment of
either syndrome.

Katon et a1. (1984) call for more studies to ascertain
the psychiatric characteristics of patients with
somatization disorder. These authors use Goldberg and
Blackwell's (1971) term the "hidden psychiatric morbidity"
to refer to the substantial numbers of patients whose
emotional problems are unrecognized by primary care
physicians.

Liskow, Penick, Powell, Haegle, and Campbell (1986a)
suggest that somatization disorder may be a heterogenous
entity. While there may be a homogenous subset of patients
with this disorder evidencing stable characteristics, there
also may exist a subset who suffer from additional
psychiatric disorders that are more amenable to treatment.
Liskow and his associates call for additional research to
determine if such a subset exists. They caution that until
such studies are conducted, it may be premature to suggest
that the diagnosis and treatment of somatization disorder be
made routinely by the non-psychiatrically trained physician.

Martin (1988) states that the basis of effective
treatment is most often founded upon accurate diagnosis. As
previously noted, the co-existence of a wide array of
concomitant psychiatric diagnoses with somatization disorder
appears to have not only confounded the diagnostic process,
but may have impeded efforts to treat patients presenting

with somatization disorder. A search of the literature

17
reveals only two controlled treatment studies despite the
documented poor recovery rate cited by Coryell and Norton
(1981) of 31 percent. Neither of these studies addressed
the presence of the co-morbidities found to be associated
with somatization disorder. Scallet, Cloninger, and Othmer
(1976) conducted a double—blind trial of electrosleep with
patients diagnosed as having Briquet's syndrome
(somatization disorder). The results suggest no advantages
for electrosleep over relaxation taught by autogenic
training. Smith, Monson, and Ray (1986) addressed the role
of psychiatric consultation in the treatment of somatization
disorder. They demonstrate that patients with somatization
disorder can receive more appropriate care if collaborative
efforts are made between psychiatrists and primary care
physicians. Quarterly health care charges declined by 53%
in the treatment group after a psychiatric consultation,
while functional health status was maintained.

An examination of the personality variables associated
with somatization disorder may aid in the treatment of the
disorder. Kaminsky and Slavney (1976) state that in view of
the association between personality and abiding response
patterns, the presence of a personality disorder may give
somatization disorder both its stability and its
recalcitrance to treatment. They suggest that the
examination of the genesis and treatment of this condition
should be shifted to a characterological approach. Lipowski

(1988) offers that a patient's personality traits and early

learning experiences constitute the predisposition to the

 

18
development and persistence of somatization. Kahana and
Bibring (1964) believe that the diagnosis of personality
types can aid in the medical management of patients who
somatize. They suggest that personality type
characteristics predict behavior management problems. As
such, they suggest that personality type diagnoses be
routinely assigned in order to assist in the treatment and
management of the patient. Fishbain, Goldberg, Meagher,
Steele, and Rosomoff (1986) state that patients with
personality disorders have a tendency to overreact to
disease in the characteristic pattern of their personality
disorder diagnosis. Research targeted at identifying these
patients will not only assist in predicting their
overresponse to their illness, it may also enable the
clinician to better tailor treatment to the patient's

individual personality constellation.

Theory

While a wide range of etiological factors have been
proposed to account for somatization, those underlying the
association of personality variables with somatization
disorder are seldom noted in the literature. In addressing
the association between somatization and a vast array of
psychiatric disorders, Escobar (1987) proposes that
somatization may in itself be a personality trait. Liskow
et al. (1986a) note that predisposing factors underlying

somatization include genetic, developmental, learning,

19
personality and sociocultural variables. Ford (1987) also
suggests a genetic predisposition toward somatization. He
notes that a significant proportion of women who are
diagnosed with somatization disorder evidence antisocial and
hysterical personality traits and/or disorders. He also
emphasizes the significantly high percentage of their male
relatives evidencing sociOpathy or alcoholism. In view of
this marked familial incidence, he suggests that
somatization disorder and male sociopathy are essentially
the same disorder, with the symptomatic expression being
determined by gender. Research exploring this interesting
hypothesis will be reviewed in Chapter Two.

In 1986, Cloninger proposed a biosocial theory
of personality which could be relevant to somatization
disorder. In this theory, he ties somatization and
personality variables together, linking both to an
underlying physiology.

Personality traits are seen as basic response patterns
with underlying neurobiological bases. Somatic anxiety is
defined as "global uneasiness or alarm without specific
premonitory cues, frequent body pains due to low sedation
thresholds and slow fatigability" (p. 170). Somatic anxiety
is related to the trait of high novelty seeking which,
accompanied by low harm avoidance and high reward
dependence, are related to second order traits of
impulsiveness, gullibility or emotional vulnerability and

narcissism. This pattern, according to Cloninger,

20
approximates histrionic personality. Neurobiologically
speaking, high novelty seeking is associated with high
serotonergic activity and reward dependence is associated
with low basal noradrenerigic activity.

Cloninger in a similar fashion links the antisocial
personality with somatization disorder. He notes that both
antisocial personality and somatization disorder have
associated neuropsychological deficits that are similar to
those observed in patients with bilateral frontal lobe
lesions and/or lesions of the septum and hippocampus. He,
like Ford (1987), suggests that these disorders are
different expressions of the same underlying genetic
predisposition.

Lipowski (1988) stresses that somatization may develOp
as a c0ping strategy enabling the individual to deal with
psychological needs and conflicts, feelings of guilt and
anger, and low self-esteem. Such a conceptualization
dovetails with Millon's (1988) notion that personality,
ideally, represents a defense system against external
psychological stressors. He compares Axis I disorders to
break-down symptoms while Axis II disorders represent a line
of defense. Such a model provides an intriguing way to view
the association between somatization disorder and
personality variables. The exact role personality variables
play in somatization disorder is unclear at this time.
However, they do appear to represent a significant etiologic

dimension.

21
Summary

Somatization disorder, a syndrome characterized by the
presence of multiple physical symptoms devoid of discernable
organic etiology, poses a significant health care problem.
Few descriptive studies have addressed this syndrome despite
the significant impact that somatization disorder has on the
patient's quality of life, the health care provider's time,
energy and morale, and the health care system's financial
status. Pertinent research highlights the heterogenous
nature of somatization disorder and suggests that
individuals diagnosed with this syndrome will frequently
meet inclusion criteria for other psychiatric diagnoses,

including personality disorders.

Purpose of the Study

 

The purpose of this study was to examine the
relationship between somatization disorder and concomitant
psychiatric diagnoses. The overall and relative incidence
of Axis I diagnoses were assessed by the Millon Clinical
Multiaxial Inventory, Second Edition (MCMI-II) and the
Structured Clinical Interview for DSM-III-R (SCID).

The study specifically examined the association between
somatization disorder and personality disorders (as measured
by the Millon Clinical Multiaxial Inventory, Second Edition
(MCMI-II) and the Structured Clinical Interview for

DSM-III-R (SCID)) in a population of patients suspected of

22

suffering from somatization disorder. More specifically,

this study addressed the following research questions:

Research Questions

 

Of those patients referred with somatization problems,
how many meet DSM-III-R criteria for somatization
disorder?

What other Axis I diagnoses occur concomitantly with
patients diagnosed as having somatization disorder?
What is the incidence and distribution of these Axis I
diagnoses?

What personality disorders are associated with
somatization disorder?

Is the incidence of personality disorders in the
research sample higher than would be expected in the
overall population?

What is the distribution of these personality disorders

within the sample?

CHAPTER II

REVIEW OF THE LITERATURE

In this chapter, four areas of relevant theory and
research are reviewed. First, research addressing the
association between somatization disorder and personality
patterns is presented. Next, an overview of personality
disorders is set forth. Millon's theory of personality is
examined next, and validation research employing the
operationalization of his theory is briefly highlighted.
Finally, the history and evolution of the Structured

Clinical Interview for DSM-III-R (SCID) is examined.

Somatization Disorder and Personality Patterns

 

Patients who somatize frequently evidence other
psychiatric disorders. While a body of research has
addressed the overall presence of these additional
syndromes, less work has been done in determining the role
personality variables play in the expression of
somatization. In this section pertinent research addressing
the relationship between somatization disorder and
personality patterns is reviewed. Studies involving
somatization disorders diagnostic precursors (Briquet's
syndrome and hysteria), as well as those covering
somatization from a general perspective will be included. A
review of the literature revealed that for the most part
these studies addressed the hysterical/histrionic

23

24
personality and/or the antisocial personality. Therefore,

the literature will be presented accordingly.

Somatization Disorder and Hysterical/Histrionic Personality
Patterns

A histrionic personality style is frequently associated
with somatization disorder and, according to some
researchers, is what defines the syndrome (Woodruff,
Goodwin, and Guze, 1974; Ford, 1983). In 1951, Purtell,
Robins, and Cohen studied 50 women diagnosed as having
hysteria. They note that in contrast to the 50 women in the
control group, the hysterics described their symptoms in a
vague, exaggerated and colorful way. In 1975, Kimble,
Williams, and Agsas compared Briquet's syndrome with the
definition of hysterical personality set forth in DSM-II
(American Psychiatric Association, 1968). They found close
agreement between the diagnoses of hysterical personality
and Briquet's syndrome. Of the 10 subjects diagnosed as
hysterical personalities, nine met the criteria for
Briquet's syndrome. Conversely, of the 11 subjects who met
the criteria for Briquet's syndrome, nine were clinically
diagnosed as hysterical personalities. The authors
concluded that the DSM-II diagnosis of hysterical
personality was very similar to Briquet's syndrome. They
offer that this finding suggests that the two diagnoses
describe different phenomenological aspects of the same

psychiatric illness.

25

Kaminsky and Slavney (1976) described a case of
Briquet's syndrome in a male patient and discussed the
relationship between that disorder and personality features.
They concluded that Briquet's syndrome is a set of behaviors
which stems from personality traits. They equated their
subject's dependent, emotionally labile, excitable and
attention-seeking personality style with the diagnosis of
hysterical personality disorder. In addition, they noted
the presence of obsessions and ruminations in the overall
style of personality evidenced by this patient.

Liskow, Clayton, and Woodruff (1977) were also
interested in studying the relationship between hysterical
personality and Briquet's syndrome. They seriously
questioned the movement in the field to separate Briquet's
syndrome from all other concepts related to hysteria (i.e.
hysterical personality; hysterical neurosis, conversion
type; and hysterical neurosis, dissociative type), and noted
that such a proposal was under consideration for DSM-III.
They cited the Kimble (1975) study as evidence supporting the
relationship between Briquet's syndrome and hysterical
personality disorder. In order to broaden this comparison,
they compared the MMPI scores of 29 inpatients diagnosed as
having hysterical personality and 21 outpatients diagnosed
as having Briquet's syndrome. The two groups of patients
differed significantly on the MMPI lie scale and on the MMPI
hypochondriasis scale, but not on any of the other MMPI

scales. They note that this finding indicates the

26
possibility that patients diagnosed as having hysterical
personalities are similar to or perhaps indistinguishable
from those diagnosed as having Briquet's syndrome. They
cite Kimble and associates (1975) finding as lending their
hypothesis additional credence.

In 1980, DSM-III (American Psychiatric Association)
changed the name of Briquet's syndrome to somatization
disorder. Thus, as noted in Chapter One, all psychological
symptoms were removed from the diagnostic criteria, and the
syndrome was separated from all other entities related to
hysteria. Despite this attempt to place the original
diagnosis of "hysteria" completely within the medical realm,
many researchers in the field continued to note the
association between somatization disorder (Briquet's
syndrome) and hysterical personality.

Kaminsky and Slavney (1983) cited previous studies
noting the presence of obsessional features in patients
diagnosed with a disorder similar to somatization disorder,
namely hypochondriasis (Pilowsky, 1970; Kenyon, 1976). They
too had found obsessional traits in a patient studied
earlier (1976), who also presented with Briquet's syndrome
and hysterical personality patterns. In line with previous
research cited, they posited that patients with somatization
disorder would not only resemble those with hysterical
personality disorder, they would also be more obsessional
than the latter group. This hypothesis was tested by

comparing patients meeting criteria for somatization disorder

27

(N=20) and patients diagnosed as having hysterical
personality disorder (N=20). These patients were assessed
using clinical interviews, mental status exams, and
personality inventories (the Eysenck Personality Inventory,
the Hysteroid-Obsessoid Questionnaire, and the
Lazare-Klerman-Armour Personality Inventory). The results
of this study confirmed past findings that many patients
with the diagnosis of somatization disorder have prominent
hysterical traits. Consistent with the aforenoted
hypothesis, the somatization disorder patients evidenced
more obsessional features than did the patients with
hysterical personality disorder. The authors suggest that a
combination of obsessional and hysterical traits of
personality may constitute an important ingredient in the
complaining behavior symptomatic of somatization disorder.

Katon et a1. (1984) reviewed the five Axis DSM-III
diagnoses of 100 consecutive somatizing patients. The
authors defined somatization as "an idiom of distress in
which patients with psychosocial and emotional problems
articulate their distress primarily through physical
symptomatology" (p. 305). A comparison of the somatizing
patients was made to the control group (N=161) of
nonsomatizing patients. The most prevalent diagnosis on
Axis I was major depressive disorder. Somatizing patients
received this diagnosis 2.5 times more often than the
nonsomatizing group. The authors also noted that personality

disorders or maladaptive traits were twice as common (51%) in

28
somatizers as in nonsomatizers (27%). The most prevalent
personality disorder was histrionic, with a 12% incidence in
the somatization group. Borderline personality disorder was
diagnosed in 6% of the somatizers. The remaining
personality diagnoses noted for the somatizing group were
histrionic and borderline traits. While the authors note that
these findings were theoretically expected in view of prior
studies, they stress that personality disorders are among the
least studied and poorly validated aspects of DSM-III
diagnostic categories. As such they call for future
follow-up studies to validate their findings.

The relationship between DSM-III Axis I and Axis II
diagnoses was also explored by Koenigsberg, Kaplan, Gilmore,
and Cooper (1985) in a review of 2,462 medical center
patients. They found that personality disorders were most
often associated with substance use disorders, anxiety, and
somatoform disorders. The somatoform group was not broken
down into diagnostic categories. However, of the 27
patients found to be somatizers, eight received personality
disorder diagnoses. Three were noted to be dependent
personality disorders, three mixed personality disorders,
and two histrionic.

In an interesting study conducted by Slavney and
Teitelbaum (1985), the judgement of physicians regarding
histrionic traits in patients they referred for psychiatric
evaluation of medically unexplained somatic complaints was

assessed. The purpose of the study was to compare the

29
physician's assessments with the psychiatrist's diagnoses.
The authors found that histrionic traits as rated by
referring physicians did not correlate with the
psychiatrist's diagnoses of somatoform disorders. Of the
100 patients evaluated, nine were diagnosed as having
somatization disorder, eight hypochondriasis, and three
atypical somatoform disorder. Personality disorders were
noted in four patients, two of which were classified as
histrionic personality disorders. In considering reasons
for this lack of association, the authors note that the
referring physicians and the consultant psychiatrists faced
different tasks: the former to rate traits, the latter to
make diagnoses. They also suggest that patients referred for
psychiatric consultation in general may be more readily
assessed by their referring physicians as having histrionic
traits than patients not so referred.

In 1985, Slavney and Teitelbaum examined the DSM-III
diagnoses and demographic characteristics of 100 patients
referred for evaluation of medically unexplained symptoms
suggesting physical disorders. Eight were diagnosed as
having somatization disorder, and an additional four were
assigned personality disorder diagnoses (two histrionic and
two mixed). The authors underscore that 77% of these
referrals were women, and that seven out of eight of those
diagnosed with somatization disorder were female. They

account for this female prevalence rate by noting the

30
historical linkages between women and the concept of
hysteria.

Somatization, hypochondriasis, and hysteria were
investigated by Snyder and Pitts (1986) as possible
associated features of the borderline personality disorder.
Inpatients with DSM-III borderline personality diagnoses
were compared with controls assigned a dysthymia diagnosis,
using the MMPI and the Hamilton Depression Scale. Although
the hysteria-obvious and hypochondriasis scales of the MMPI,
and the Hamilton Depression Scale item measuring
hypochondriasis were elevated in the borderline group, there
were no significant between group differences noted.

Large (1987) assessed 50 consecutive patients with
chronic pain in terms of DSM-III nosology. Four were found
to have Axis I somatization disorder diagnoses. Of these
four patients all received personality disorder diagnoses.
One was rated histrionic, one dependent and two were given
mixed personality disorder diagnoses.

In 1987, Leon et al. studied 131 patients referred for
psychiatric consultation. They divided these patients into
somatizing (N=56) and nonsomatizing (N=75) groups according
to somatization diagnostic criteria established by Bridges
and Goldberg (1985). A statistically significant difference
was noted between the groups on the dimension of personality
disorder and pathologic personality traits. Histrionic
personality disorder or personality trait was significantly

more frequent in the somatizing group. The authors relate

31
these findings to Katon's (1984) study suggesting that
somatizers have more hysterical/histrionic traits. The
following section of Chapter Two highlights research
addressing the relationship between somatization disorder

and antisocial personality patterns.

Somatization Disorder and Antisocial Personality Patterns

Evidence supporting an association between sociopathy
and somatization disorder has been mounting for nearly a
century (Lilienfield, VanValkenburg, Larntz, and Akiskal,
1986). This movement often parallels and at times overlaps
with the similar movement previously noted linking
somatization disorder with hysterical/histrionic personality
variables. In an historic review of the literature,
Lilienfield et a1. note that in 1884 Moravesik reported a
high incidence of criminal histories among hysterical
subjects. They also cite Kraepelin's (1915) report of a
frequent co-occurrance of hysterical and antisocial symptoms
within individuals. Research within the last four decades
has substantiated these early observations.

In 1952, Robins and associates noted that in a sample
group of 44 men diagnosed with hysteria, six had
psychopathic personalities. None of the control subjects
were diagnosed as such. In 1966, L.N. Robins found further
evidence indicating a link between the two conditions. He
found that of 76 girls referred to a child guidance clinic

for antisocial behavior, 20 met criteria for hysteria as

32
adults. There were no cases noted among the 100 normal
control subjects. It is interesting to note that as adults
the 20 women showed little or no antisocial behavior.

In 1970, Cloninger and Guze interviewed 66 female
felons to determine what psychiatric illnesses are
characteristic of the adult antisocial female. In addition,
they were interested in studying the association between
antisocial behavior and hysteria. All of the women received
at least one psychiatric diagnosis. The most frequent
diagnosis found was sociOpathy (65%). Thirty-nine percent
of the women received a diagnosis of hysteria. The authors
note sociopathy or hysteria were found in 80% of the
subjects. They stress that this high prevalence of hysteria
represented 20 times that seen in the general population.
The results of this study, according to these researchers,
confirms the existence of a significant association between
sociopathy and hysteria. Cloninger and Guze further posit
that hysteria and sociopathy may share a common etiology.
They reason that since hysteria is characteristically a
disorder inflicting women, and is rarely seen in men, and
since sociopathy is chiefly noted in men, although it may
occur in women, ”depending upon the sex of the individual,
the same etiologic and pathogenetic factors may lead to
different, although sometimes overlapping clinical pictures"
(p. 310).

In 1975, Guze reported on a review of the literature he

undertook to help clarify the development of both hysteria and

 

33
sociopathy. He noted that Mendelson et a1. (1971) found
that hyperactive children appear to carry an increased risk
of delinquency and sociopathy. The results of two studies
of the families of hyperactive children (Morrison and
Stewart, 1971; Cantwell, 1972) were reported. Guze notes
that each of these studies reports an increased incidence of
sociopathy in the fathers and hysteria in the mothers of
hyperactive children as compared with parents of control
children. Interestingly, Guze observes that similar
findings were not reported in a study of adoptive parents of
hyperactive children (Morrison and Stewart, 1973).

Guze (1975) continued to review studies linking
hysteria with sociopathy (Arkonac and Guze, 1963; WOerner
and Guze, 1968). He noted that 20% of first-degree female
relatives of female hysterics also may suffer from hysteria.
In addition, male relatives evidence an increased incidence
of alcoholism and sociopathy. Conversely, first-degree
female relatives of male sociopaths evidence an increased
prevalence of hysteria (Guze et al., 1967). Finally, Guze,
in his review, calls attention to a pattern of mating
between hysterical women and sociopathic men. Sociopathy is
increased in husbands of hysterical women and hysteria is
more prevalent in the wives of sociopathic men (WOerner and
Guze, 1968; Guze, Goodwin, and Crane, 1970). He summarizes
by stating that Briquet's syndrome (hysteria) follows a

predictable course and clusters in some families. He notes

the strong association between sociopathy and Briquet's

34
syndrome and suggests that the two disorders share similar
pathogenetic factors.

Liskow et al. (1986a) reported on a six year study in
which 1,437 female outpatients at a psychiatric clinic were
administered a structured diagnostic interview. Of the 78
females who met the criteria for Briquet's syndrome, 77 also
fulfilled inclusive diagnostic criteria for one or more
additional psychiatric syndromes. (See Table 1.13, Chapter
One.) Antisocial personality occurred in 16.7% of these
patients. The authors note that this rate is consistent
with associations noted in prior studies. Obsessive-
compulsive disorder was observed in 26.9% of the Briquet's
syndrome patients. This finding is interesting in light of
Kaminsky and Slavney's (1976, 1983) observations of
obsessional features in somatizing patients.

Liskow et al. (1986b) compared 16 psychiatric inpatients
diagnosed with Briquet's syndrome with 32 age and sex-matched
psychiatric inpatient controls. These groups were compared
by using a structured diagnostic interview (Psychiatric
Diagnostic Interview) and the MMPI. Among the individual
syndromes, only antisocial personality occurred with
statistically greater prevalence in the Briquet's group as
compared with the control group; however, the study
indicated that the Briquet's group had several additional
psychiatric syndromes and more of these syndromes than the
control group. The MMPI results show that the Briquet's

group can be clearly distinguished from the age and

35

sex-matched controls by their greater elevation on seven of
13 basic MMPI scales (F, 1, 3, 6, 7, 8 and 9). The authors
note that female psychiatric inpatients with MMPI profiles
on which scale eight plus three or more other clinical
scales was greater than 70, were found to have Briquet's
syndrome with an accuracy of 75%. They suggest that the
MMPI may be a useful screening instrument for Briquet's
syndrome.

Cloninger, Martin, Guze, and Clayton (1986) evaluated
277 females and 129 males and their families for the
presence of somatization disorder using DSM-III and
Briquet's syndrome criteria. Results differed depending on
gender. In women, the prevalence for somatization disorder
was 22%. The DSM-III criteria for somatization disorder
showed high concordance with the criteria for Briquet's
syndrome in this pOpulation. Most striking, however, is the
fact that these women had nearly three times as many female
relatives with the same diagnosis as the control group.
There was no excess of somatization disorder in the male
relatives of female somatizers. The males evaluated were
most noteworthy for anxiety symptoms. Only a minority had
somatization disorder, and the prevalence of somatization
disorder in their relatives was not noteworthy. The
findings suggest that somatization has a different clinical
picture and different familial antecedents in men than in
women. An additional finding in this study was that 13.2%

of the men diagnosed with somatization disorder were also

36
diagnosed as having antisocial personality disorder. Only
3.6% of those women found to have somatization disorder
received this diagnosis.

Zoccalillo and Cloninger (1986) examined 50 outpatient
women with somatization disorder and 25 control women with
major depression disorder. They found that 60% of the
somatizers had at least one other psychiatric diagnosis
including histrionic and antisocial personality traits. In
addition to these statistically significant differences, the
somatization disorder group was significantly more likely to
abuse or neglect their children and do poorly at work. The
authors note that these findings of excessive psychological
symptomatology and social disability among women with
somatization disorder are important in terms of managing,
classifying, and finding the cause of the disorder.

This review of the literature thus far suggests a
strong association between personality disorders and
somatization disorder. While several personality disorders
have been reported to co-exist with somatization disorder,
most studies have underscored the strong association between
antisocial and/or histrionic personality and somatization
disorder. The possibility that these are not three distinct
illnesses, but, rather, different but often co-occurring
stages or endpoints of a shared pathogenesis has been noted.
In an attempt to lend support to this theory, Lilienfeld et

a1. (1986) examined the association of antisocial

37
personality disorder, somatization disorder, and histrionic
personality disorder, both within individuals and within
families, in 250 patients. They conducted diagnostic
interviews in a variety of inpatient, outpatient and
consultation settings. The results strongly confirm
previous reports of an association between antisocial
personality disorder and somatization disorder within

2=13.49, df=1, p<.001). This

individuals (rq=.55; x
relationship was statistically significant for both men
(rq=.56; x2=4.84, df=1, p<.03) and women (rq=.66; x2=10.74,
df=1, p<.001). In addition, a significantly greater
proportion of subjects with somatization disorder than
subjects without somatization disorder reported having at
least one first-degree relative with antisocial personality
(rq=.63; x2=10.69, df=1, p<.001).

Somatization disorder and histrionic personality
disorders were also significantly related within individuals
(rq=.60; x2=l7.06, df=1, p<.001). In addition, a
significant association within individuals was found between
antisocial personality disorder and histrionic personality
disorder (rq=.81; x2=52.86, df=1, p<.001). Fitting these
data to a log-linear model, these authors found that
histrionic personality disorder was more strongly correlated
with antisocial personality disorder than was somatization

disorder. However, the relationship between histrionic

personality disorder and prevalence of antisocial

38
personality disorder among first-degree relatives was not
significant.

The authors summarize their results by contending that
histrionic personality disorder, antisocial personality
disorder and somatization disorder are not three separate
illnesses, but rather, different, but often co-existing
stages or endpoints of a shared pathogenesis. They feel
that the extensive overlap among the three syndromes
suggests that present classification schemes may be
attributing separate diagnoses to diverse manifestations of
a single disease. Further studies are needed to expand and
test this provocative hypothesis. This review of the
literature addressing the role personality disorders play in
the expression of somatization disorder has yielded mixed
findings. However, certain trends are evident.
Specifically, pertinent research suggests a strong
association between antisocial and/or histrionic personality
disorders and somatization disorders. The next section of
Chapter Two puts forth an overview of the personality

disorder diagnosis and its prevalence in clinical settings.

An Overview of Personality Disorders
The provision in DSM-III (American Psychiatric
.Association, 1980) for a separate axis devoted to the
diagnosis of personality disorders has resulted in increased
clinical and research interest in this domain. Prior to

DSM-III, personality diagnoses were often overlooked in the

39
presence of more clinically prominent syndromes. Goldsmith,
Jacobsberg, and Bell (1989) note that the multiaxial
evaluation format set forth in DSM-III underscores the
contribution and interrelationship of long-term trait
factors in the development of psychiatric symptoms and
disorders. Germane to this study, Millon (1981) stressed
that the diagnosis of personality disorders on Axis II is
especially important because it may be both in and of itself
an indication for treatment. Specifically, a personality
disorder diagnosis may alter the presentation, course and
treatment of either an Axis I clinical syndrome (e.g.,
somatization disorder) or of a psychiatric symptom.

Table 2.11 (Blashfield and McElroy, 1989) schematizes
the development of personality disorders from a clinical
perspective from DSM-I to DSM-III-R. DSM-I (American
Psychiatric Association, 1952) set forth five categories of
personality disorder:

1. personality pattern disturbance

2. personality trait disturbance

3. sociopathic personality disturbance

4. special symptom reaction

5. transient situational personality disorders

Personality pattern disturbances were severe disorders
impacting the entire personality structure, while personality
trait disturbances were limited to isolated, rigid and
maladaptive personality variables. The sociOpathic category

referred to individuals expressing deviance in many forms

40

mcuunomonuuamm
ofiunannn
AmnHHHuzmo. xHoznmm<

.ouo .cowuommu mmouum mmouu
unmouOmwo
huaaoaomumm floccuuuauam unmfimcmue

 

unavao>¢
acmpcomoo o>am0amxm .oum .mocnnusamao mawcumma
coauowom Eouoexm Hmaooom
.muHHHIZmoc cacmnunm snaaonoofla
o>amH5oEoo m>wmmmmno Ho noduua>mo Hmsxmm
AHHHuzmo. w>wmasm800 Hufloommmo
w>ammmummmuo>ammmm HHflOOnduad HMAUOmHucd
O mmewoqo oucunusuuao Suwaucomnmm oasuMQOHUOm
o>auasmeoo o>dnu0mno o>HuHaQEou
oaumammaouoz o>amnoummm
ocaaumpuom 0>ummoummolm>annum o>wnmoummulo>ammuo
unoccommp
HddUOmausd o>aumouommno>wmmum
cacoHuumam HdOfiuouuam manuumcs adducowuoem
m mnemoqu ousmnusunqa venue Suwamcomumm
pecuwnom pa0nanom vacuaaom
Huo590udnom owahau0a0>o oqexcuoHomu
ofiocmncm oaocouum oaocmumm
muuaompuaH wuaavmpmcH
d mmamDAo oucunusumwo cumupum muqaunOmumm
mIHHHuzmoxHHanzmo iJmlemo lemo

 

 

.xoMH .n .mmmH. mouamoz can nflmamnnnam .numuuonao

HH.N magma

 

 

Sudan:0muom mo cowusao>m aaoamoHouco

41

including antisocial behavior, alcoholism, and drug addiction.
Disorders such as learning disability and enuresis were subsumed
under special symptom reaction. Finally, the transient category
is the diagnostic forebear of what is now noted as posttraumatic
stress disorder.

DSM-II (American Psychiatric Association, 1968) sought
to narrow the personality disorder headings. The
situational symptom reactions and the transient personality
disorders along with disorders of alcoholism and substance
abuse were placed elsewhere in the classification. The
differentiation among personality trait and personality
pattern was dropped. Instead, nine different personality
disorders were grouped under one heading.

DSM-III (American Psychiatric Association, 1980)
returned to the utilization of subheadings. Eleven
personality disorders were organized into three clusters.
Cluster A referred to the "odd” or strange personality.

Cluster B was noteworthy for being "dramatic," or emotional
and the most likely to be seen in clinical practice.
Cluster C was called the "anxious" group.

The chief difference between the DSM-III and DSM-III-R
classifications involves the diagnostic criteria utilized to
define the personality disorders. Specifically, definitions
of dependent, histrionic, paranoid and passive-aggressive
personalities have been altered. The addition of

self-defeating and sadistic personality

42
disorders is the main categorical difference between DSM-III
and DSM-III-R. The controversy surrounding the inclusion of
these new personality disorders is reflected in their being
placed in an appendix to the DSM-III-R rather than in the
personality disorders section (Blashfield and McElroy,
1989).

Despite the many changes that have been made,
inconsistencies in the diagnosis of personality disorders
remain. Zimmerman and Coryell (1989) suggest that unlike
Axis I disorders, Axis II disorders are based more on
clinical consensus than research-generated operational
criteria. In particular, they note that almost all research
on DSM-III and DSM-III-R personality disorders has been
conducted on patient samples. In order to obtain a more
accurate estimate of the prevalence of personality disorders
in the community they assessed 797 first-degree relatives of
normal controls and patients with a variety of psychiatric
diagnoses. Slightly more than one-sixth of the sample
received a personality disorder diagnosis (n=143, 17.9%).
The most prevalent diagnoses were mixed (n=29, 3.6%),
passive-aggressive (n=26, 3.3%), antisocial (n=26, 3.3%),
histrionic (n=24, 3.0%), and schizotypal (n=23, 2.9%). The
authors conclude that while the overall rate of personality
disorders in the community may be between 10 and 20%, the
rates of specific disorders are very low.

Widiger and Rogers (1989) reviewed eight studies in an

attempt to clarify prevalence estimates of personality

43
disorders (see Table 2.12). While these rates varied as a
function of population and interview-format, some
consistencies emerged, particularly in terms of the rank
order of the prevalence estimates. The most prevalent
personality disorder, both in inpatient and outpatient
settings, was borderline personality disorder. These
authors suggest that across studies this diagnosis is seen
in 1 to 2% of the community, 8% of all outpatients, and 15%
of all inpatients. They offer that the next most common
personality disorders tend to be schizotypal and histrionic,
followed by dependent, avoidant, antisocial,
passive-aggressive, paranoid, and narcissistic.

In summary, the diagnosis of personality disorder has
undergone considerable modification from the original DSM-I
(American Psychiatric Association, 1952) designated
criterion to the present DSM-III-R (American Psychiatric
Association, 1987) conceptualization. Personality disorders
are prevalent in both inpatient and outpatient clinical
settings. In addition, they are likely to be encountered in
nonpsychiatric medical settings. Widiger and Rogers (1989)
suggest that their presence should be evaluated in every
patient, including those in primary-care medical settings,
because these personality variables can considerably
influence the diagnosis and treatment of psychiatric

disorders.

44

 

ad. NH. on. mm. on. «o. oo. no. on. oo. «H. oucosuooaa no .eooa..du um «cannons

mm. no. me. on. HH. me. .no. em. oo. no. on. ouaoauaoas on leoodc.dn uo nooooox
oucoauaacd

«a. no. no. HH. «o. a". «o. oo. oo. so. no. Sauces and Loom”..ao um Hooon
, nunodvam

so. no. «a. on. oo. an. an. oo. 5“. so. so. none saunas Hon loom”. souoz

oo. no. no. no. mo. mg. on. «a. on. do. no. auaouuaooq oo .noos..an no uoooonoo

«o. «o. no. no. no. oo. «H. «o. 4o. so. no. ouaoauaouao ooo lmooflc.aa uo nnnx

no. mo. om. do. on. «a. on. no. nu. no. no. ouaoquaouoo on .vooflo.fln no noocnnn

mo. no. No. oo. «o. a". on. a“. «a. no. so. uncouuoooa no” looms. anno

can moo omo o>¢ maz you new oaa sum can 2mm ouaonuao woooo
wouNOuMmlhuaaa=0nuom‘ «0 .oz

 

..mma .o .mmoo. nuomom

cad uom«o«3 .aouonum a£m«m ad unannouao audmocOnuom HHHazmo mo oocon>mum

NH.N OHAGB

45
The next section of Chapter Two will examine a specific
theory of personality; that of T. Millon. Research using
Millon's theory and its operationalization, the Millon
Clinical Multiaxial Inventory, will be highlighted.

Millon's Theory of Personality

 

The guiding theoretical system for the development of
the MCMI-II is based on Millon's theory of personality
(Millon, 1981). This theory suggests 10 basic styles of
personality functioning that can be constructed logically
from a 5x2 matrix consisting of two basic dimensions.
According to Millon, the first dimension is based upon the
individual's perception of sources of reinforcement
(positive or negative). The manner in which one obtains
comfort and satisfaction in life constitutes positive
reinforcements, while the avoidance of emotional pain is
considered negative reinforcements. Those who experience few
rewards in life, whether from self or others, are referred
to as "detached" types. A "dependent" individual gauges
his/her satisfaction and emotional pain by how others react
to or feel about him or her. Those individuals who measure
their satisfaction in life according to their own values and
desires are described as ”independent"; finally, those who
experience conflict over whether to rely on others versus
their own needs are considered ”ambivalent."

The second dimension of this matrix describes the
coping mechanisms employed to deal with the various types of

reinforcement (active versus passive). Those individuals

46
who are attentive and alert, and who intervene with and
manipulate the circumstances of their environment are
considered ”active.” ”Passive" individuals initiate very
little and evidence a resigned attitude while environmental
events take their course.

Combining the five sources of primary reinforcements
together with the two coping patterns results in 10 basic
personality styles: active and passive detached, active and
passive discordant, active and passive dependent, active and
passive independent, and active and passive ambivalent.

Millon sees personality styles as deeply etched and
pervasive characteristics of one's functioning level. While
these characteristics may impact every day behavior, they
are out of the individual's awareness. Under conditions of
persistent adversity, a maladaptive style of functioning may
decompensate acquiring features deserving of the designation
moderate or marked severity. Millon views these advanced
stages of personality pathology in terms of a slow
deterioration of the personality structure. Despite noted
alterations in psychic cohesion, social competence and
emotional control, the individual continues to display the
major personality characteristic that have become entrenched
in the personality structure.

Millon describes these 13 personality patterns along
two dimensions, functional processes and structural
attributes. Functional processes include behavioral,

interpersonal, cognitive and defense mechanistic dimensions.

47
Structural attributes address the individual's mood,
self-image, internalizations and intrapsychic organization.
For example, the histrionic personality is seen in Millon's
system as behaviorally affected, interpersonally
flirtatious, cognitively flighty and employing
dissociative-type defensive mechanisms. He or she is likely
to evidence a fickle mood, a sociable self-image, shallow
internalizations and disjoined intrapsychic organization.
The antisocial personality, on the other hand, would present
as behaviorally impulsive, interpersonally irresponsible,
cognitively deviant and using acting-out as a primary
defense mechanism. Structurally speaking, the antisocial
personality would be described in terms of callous mood and
autonomous self-image. He or she would harbor rebellious
internalizations and have an unbounded intrapsychic
organization.

In contrast to the personality disorders, Millon sees
the clinical syndrome disorders comprising Axis I as
extensions or distortions of the individual's basis
personality traits. These syndromes are seen as transient
states that caricature or accentuate the basis personality
style. Millon cautions that regardless of how distinctive
these Axis I syndromes appear to be, they take on meaning
and significance only in the context of the patient's
personality. Most of the clinical syndromes Millon
addresses are of the reactive kind and of substantially

briefer duration than the personality disorders. They

48

usually designate states in which an active pathological
process is clearly manifested. Millon underscores the
existence of several possible covariations between Axis I
syndromes and Axis II personality styles. Indeed, these
interrelationships are held to be crucial to his theory of
personality. Six scales are put forth as representing
disorders of moderate severity:

1. Scale A: Anxiety

2. Scale H: Somatoform

3. Scale N: Bipolar - Manic

4. Scale D: Dysthymia

5. Scale B: Alcohol Dependence

6. Scale T: Drug Dependence
The final three scales reflect Axis I disorders of marked
severity:

7. Scale SS: Thought Disorder

8. Scale CC: Major Depression

9. Scale PP: Delusional Disorder

An important aspect of Millon's theory is its attempt to
coordinate directly with the American Psychiatric
Association's official diagnostic system and its syndromal
categories. Millon stresses that his theory has been
developed to be consonant with the nosological format and
conceptual terminology of DSM-III and DSM-III-R. Millon's
theory of personality has been operationalized in his
personality assessment instruments, the MCMI and the

LflnHI-II. Millon (1987) sees the development of the MCMI-II

49
as an attempt to refine and strengthen both the theoretical
logic and the research data underlying the instrument, thus
resulting in not only a sounder psychometric tool, but a
more useful clinical assessment inventory. Moreover, Millon
notes that these revisions fortify even further the
coordination between the MCMI—II and the syndromes of
DSM-III-R, especially those of Axis II. Additional data

regarding the MCMI-II will be reviewed in Chapter Three.

Research Employing the MCMI and the MCMI-II

 

As previously noted, Millon's theory of personality has
been operationalized in his inventories the MCMI and the
MCMI-II. It was Millon's position that, consistent with the
DSM-III multiaxial approach to clinical evaluation, these
instruments assess and differentiate between basic
maladaptive personality characteristics and various clinical
symptom syndromes. Because of its relatively recent
commercial availability, little has been published regarding
the MCMI-11's validity as an instrument which can adequately
perform this distinction among patients with somatization
disorder. Indeed, a search of the literature revealed no
studies which employed either the MCMI or the MCMI-II in
research addressing patients who somatize. Of theoretical
interest, in particular, is the notion of the co-existence
of personality disorders and somatization disorder. In the

absence of empirical tests of this question, this section

50
highlights studies undertaken by independent investigators
with the purpose of examining the validity of MCMI-I.

Craig, Verinis, and Wexler (1985) discovered that the
Axis II personality configuration that typified 106
alcoholics in an inpatient VA unit was notably different
from that of 100 opiate abusers seen at a similar VA unit.
Alcoholics exhibited significantly higher scores on the
Dependent, Avoidant, Schizotypal, and Borderline scales,
whereas opiate addicts scored appreciably higher on
Narcissistic, Histrionic, and Antisocial scales.

McMahon, Flynn, and Davidson (1985) examined the
stability of the basic and pathologic personality and
symptom scales of the MCMI-I in three distinct clinical
samples. Consistent with Millon's theoretical perspective,
higher stability estimates were found among basic
personality scales in comparison with symptom scales. The
Histrionic scale had consistently high stability
coefficients (£é.80). High moderate range stability
coefficients (between £§.70 and 5:.80) were found for the
Avoidant and Compulsive scales. The Schizoid, Narcissistic,
Antisocial, Schizotypal, and Hypomanic scales has stability
coefficients in the low moderate range (£3.60 to 5;.70).
The subjects in these samples were (a) 96 inpatient
alcoholics who completed the MCMI-I at intake and 30 to 45
days into treatment, and (b) 33 drug abusers being treated

in either an inpatient or (c) an outpatient program who

51
completed the MCMI-I upon intake and one month and three
months after admission to treatment.

In addition to alcohol and substance abuse studies,
researchers have used the MCMI—I as a foundation for
identifying and clarifying the clinical characteristics of
posttraumatic stress disorder (PTSD). Male veterans
subjected to war experiences were contrasted with controls
having other psychiatric disabilities. Robert, Ryan,
McEntyre, McFarland, and Lips (1985) found that the PTSD
patients emerged with an 8,2 (passive-aggressive/avoidant)
type in contrast to controls. High scores on Scale C
(Borderline) were also noted in the PTSD group. Axis I
scale scores for Anxiety and Dysthymia also were notably
high, as were the Alcohol and Drug Abuse scales.

A recent study (Bryer, Miller, Nelson, and Kroll,
unpublished study) used the MCMI-I to identify and to
differentiate psychiatric symptoms among adult survivors of
childhood abuse. Mean base rate (BR) scores were obtained
for 66 persons receiving treatment in a short-term inpatient
psychiatric hospital. Of the 66 patients, 27 had
experienced no abuse, 14 had been subjected to sexual abuse
alone, 10 suffered physical abuse only, and 15 suffered both
sexual and physical abuse. In line with the previously
noted PTSD male veterans, those abused evidenced scales
reaching pathological levels (BR>75). In particular,
statistical and near statistical significance was seen on

the 2,8 and C profile (Avoidant, Passive-aggressive, and

52
Borderline). Millon (1987) notes the similarity between
the two groups and suggests that the long-term psychological
effects of profound abuse are much the same for both men and
women, despite differences in the nature of the abuse.

A comprehensive analysis of sex offender subtypes using
the MCMI-I was reported by Bard and Knight (1987). Subjects
(N=101) judged to be "sexually dangerous persons" and given
indeterminate sentences as a function of their repetitive,
violent, and/or compulsive sexual offenses were assessed.
Focusing on the basic personality scales alone, the overall
mean profile for the group was 6,5,4 (Antisocial
[aggressive], Narcissistic, and Histrionic). Using a
cluster analysis, four distinctive groups emerged. One
subgroup, distinguished by its 2,1,3 profile (Avoidant,
Schizoid, and Dependent) was composed of rapists and child
molesters with low IQ's. The second subgroup, composed of
mainly rapists presented with a 5,6,4 configuration
(Narcissistic, Antisocial, and Histrionic). The third
grouping achieved a 6,8 code (Antisocial and
Passive-aggressive). They were identified as aggressing
towards others, as well as being alcohol and drug abusers.
The final subset evidenced no MCMI=I elevations. The
authors report that this group had a less troubled family
history, appeared to be compensating for feelings of
masculine inadequacy, and acted out with premeditation, not

impulsivity.

53
The final study to be reviewed focuses on the
assessment and treatment of eating disorders using the
MCMI-I. Tracy, Norman, and Weisberg (1987) examined
similarities and differences between bulimics and
anorectics. Their findings suggest that both disorders are
characterized by high levels of anxiety, dysthymia and
somatization, but arising in different personality types.
Anorectics evidence significantly higher Schizoid, Avoidant,
and Schizotypal scales, suggested by their marked tendency
toward social withdrawal. Bulimics, in contrast, exhibit
appreciably elevated scores on the Histrionic and
Narcissistic scales. These researchers summarize their
findings in terms of Millon's theory of personality:
This would seem to indicate an important difference
between these two groups of eating disorder patients;
anorectics seeming to renounce need-satisfaction and
interpersonal attachment, bulimic patients in contrast
continuing to actively strive for need-satisfaction,
wavering between doing so via an actively dependent
style and one of more narcissistic self-absorption.
This is intuitively consistent with the clinical
paradigm of these two patterns; anorectics tending
toward constriction and self-deprivation, bulimics
displaying phases of externalized need-satisfaction
through dyscontrolled binging followed by the
reassertion of control through purging (p. 196).
The following section of Chapter Two will trace the
history and evolution of the Structured Clinical Interview
for DSM-III-R (SCID) (Spitzer, Williams, Gibbon, and First,

1990).

54
Structured Clinical Interview for DSM-III-R

History and Evolution

 

The advent of DSM-III in 1980 (American Psychiatric
Association) heralded a major advance in making psychiatric
diagnoses. Both this publication and its successor,
DSM-III-R (American Psychiatric Association, 1987) included
specific diagnostic criteria for virtually all recognized
mental disorders. Before 1980, multiple sets of diagnostic
criteria existed along with structured interviews designed
to render diagnoses specific to each particular system.
Noteworthy among these numerous systems were the Feighner
Criteria (Feighner, Robins, and Guze, 1972) and the Research
Diagnostic Criteria (Spitzer, Endicott, and Robins, 1978).

While these various nosological systems attempted to
standarize the criteria underlying specific psychiatric
diagnoses, the "process" of diagnostic assessment remained
unstandarized. In response to this need for more
consistency in making psychiatric diagnoses, Spitzer and his
associates began developing the Structured Clinical
Interview for DSM-III (SCID) (Spitzer et al., 1990). In
1983, the National Institute of Mental Health (NIMH) also
recognized the need for a clinical instrument capable of
making DSM-III diagnoses. The NIMH Epidemiologic Catchment
Area Program was launched; a study examining the
epidemiology of psychiatric disorders (Regier, Myers, and
Kramer, 1984). In this project, the Diagnostic Interview

Schedule (DIS) was the underlying assessment tool (Robins,

55
Helzer, Croughan, and Ratcliff, 1981). During this same
time period, the NIMH awarded a contract to the Biometrics
Research Department at the New York Psychiatric Institute to
field test the SCID (Spitzer et al., 1990), resulting in
further refining of this instrument.

The SCID represents an evolution in diagnostic
instruments in that it incorporates several features not
present in previous instruments. For example, when
comparing the SCID with the DIS, one notes that the DIS was
capable of identifying only antisocial personality
disorders, whereas the SCID can identify all the DSM-III-R
personality disorders. The uses of the SCID are more
diverse than previously designed assessment instruments.

The SCID can be used by clinicians to confirm and document a
suspected DSM-III-R disorder or to screen for all Axis I and
II disorders. In addition, the SCID lends itself to
research with features such as an overview section that
allows the patient to describe the deve10pment of his or her
current episode of illness, and a modular design that
enables researchers to eliminate examination of major
diagnostic classes that are irrelevant to their study
(Spitzer et al., 1990). Further details regarding the SCID
are presented in Chapter Three. As previously noted with
the MCMI-II, the SCID's relatively recent commercial
availability has resulted in the absence of pertinent

available research.

56
§EEEE£X

This chapter has reviewed pertinent research and theory
in the area of both personality disorders and somatization
disorder. Studies addressing the association between
somatization disorder and personality disorders suggested a
strong relationship between the two syndromes. Histrionic
and antisocial were the most frequently noted personality
disorders in these associations and were thought to play a
key role in the expression of somatization disorder. A
review of the literature on personality disorders traced
their ontological evolution and underscored the absence of
research-generated operational criteria in terms of their
definition and prevalence. Studies examining the prevalence
of personality disorders in the community were reviewed. An
overall prevalence rate of between 10 and 20% (Zimmerman and
Coryell, 1989) was noted. Millon's (1981) theory of
personality was specifically addressed, as this theory has
been Operationalized in his inventory the MCMI-II (1987).
This instrument was designed to assess and differentiate
between basic and maladaptive personality characteristics
and various clinical syndromes, in a fashion consistent with
the DSM-III-R multiaxial approach to clinical evaluation
(American Psychiatric Association, 1987). Finally, the
Structured Clinical Interview for DSM-III-R (SCID) (Spitzer
et al., 1990) was reviewed. Like the MCMI-II, this
instrument was designed to assess Axis I and Axis II

clinical syndromes consistent with DSM-III-R diagnostic

57
criteria. As such, both tools appeared to be promising
instruments in assessing the relationship between

somatization disorder and concomitant psychiatric syndromes.

CHAPTER III

METHODOLOGY
The purpose of this chapter is to delineate the plan of
operation used for the study. The following sections will

be included:

1. Sampling Procedures
2. Procedures for Collecting Data
3. Instruments

a) Millon Multiaxial Clinical Inventory - Second
Edition (MCMI-II)
b) Structured Clinical Interview for DSM-III-R (SCID)
4. Statistical Hypotheses
5. Design

6. Data Analysis

Sampling Procedures

 

The sample for this study was drawn from a population
of health care patients seen in local clinics and hospitals.
Health care providers in these facilities were provided the
DSM-III-R criteria for somatization disorder (see Appendix
A), the abbreviated diagnostic checklist appearing on Page
6, and a summary of the research project (see Appendix B).
They were then asked to refer patients to participate in the
research project who met the salient criteria.

Specifically, subjects referred to the study were required
to be between the ages of 18 and 55, and have no history of

58

59
formal thought disorder. Over a period of nine months,
health care providers referred 31 patients. Of these 31
patients, 25 completed both the SCID and the MCMI-II.
Demographic and other identifying features of the sample

appear in Appendix C.

Procedures for Data Collection

Patients referred by health care providers were
contacted by phone and invited to participate in the study.
Patients were provided with a general statement of the
purpose of the project (see Appendix D). Subjects were
reassured as to the confidentiality of their responses. An
informed consent form was obtained from each subject (see
Appendix E) in accordance with the Ethical Principles in the
Conduct of Research with Human Participants (American
Psychiatric Association, 1982) and the human research
committee standards of Michigan State University. Subjects
were also informed that the interview segment of the study
would be audio-taped, and the appropriate consent forms were
obtained (see Appendix F).

After informed consent was obtained, patients were
administered the Structured Clinical Interview for DSM-III-R
(SCID) by one of two psychiatrists collaborating on the
study. Based on this interview the psychiatrists assigned a
multiaxial DSM-III-R diagnosis to each subject. In
addition, either prior to or following the structured

interview, patients completed the Millon Clinical Multiaxial

60
Inventory - Second Edition (MCMI-II). This personality
inventory yielded DSM-III-R Axis I and Axis II diagnoses.
Both the SCID and the MCMI-II will be described in the
instruments section of this chapter.

Upon completion of data analyses, the referring health
care provider was provided a written summary of the
individual subject's diagnostic profile. Subjects who
desired additional feedback regarding their assessment were

requested to contact their referring health care provider.

Measures

 

The following section will provide a description of the
two measures used in this study: the Millon Clinical
Multiaxial Inventory - II (MCMI-II) and the Structured

Clinical Interview for DSM-III-R (SCID).

Millon Clinical Multiaxial Inventory - II

The Millon Clinical Multiaxial Inventory (MCMI-II) is a
175-item, true-false inventory, developed by Theodore Millon
(1987). The MCMI-II replaces the original MCMI. Designed
specifically to coordinate with DSM—III-R categories of
personality disorders and clinical syndromes, the MCMI-II
provides a measure of 22 personality disorders and clinical
syndromes in individuals 17 years of age and older who are
undergoing psychological or psychiatric assessment or
treatment. These 22 clinical scales are divided into four

main sections: basic clinical personality pattern scales

61
(1—8B), severe or pathological personality disorder scales
(S, C, P), moderate clinical syndromes (A, H, N, D, B, T),
and severe clinical syndromes scales (SS, CC, PP). In
addition, three correction scales have been added (X, Y, Z).
The 22 clinical scales and three modifier indices are listed
in Table 3.11.

The MCMI-II, as noted in Chapter Two, was developed on
Millon's theory of personality and psychopathology (Millon,
1981). Millon, a contributor to DSM-III-R, based the
MCMI-II on the DSM—III—R differentiation between enduring
personality patterns (Axis II) and more acute symptomatology
(Axis I). This study attempted to clarify the diagnosis of
somatization disorder by examining the relationship between
personality disorders and the syndrome in question.

Millon's (1981) conceptualization of personality and
clinical syndromes was highlighted in Chapter Two. In
addition to personality and clinical syndrome scales, the
MCMI-II also offers three correction scales. Scale x
(Disclosure) is designed to assess whether an individual is
frank and self-revealing or reticent and secretive. The
Desirability Gauge (Scale Y) seeks to identify the degree to
which test results reflect an individual's tendency to
appear socially attractive, morally virtuous and/or
emotionally well-composed. The Debasement Measure (Scale 2)

reflects tendencies opposite those of Scale Y, although both

62

Table 3.11

Scales of the Millon Clinical Multiaxial Inventory - II,

Millon (1987).

Modifier Indices

 

 

Name

 

Disclosure
Desirability

Debasement

Severe Personality Pathology

 

 

 

 

m
X
Y
2
Personality Pattern
§gglg Name
1 Schizoid
2 Avoidant
3 Dependent
4 Histrionic
5 Narcissistic
6A Antisocial
6B Aggressive/Sadistic
7 Compulsive
8A Passive-Aggressive
8B Self—Defeating
Clinical Syndrome
S2313 Name
A Anxiety Disorder
H Somatoform Disorder
N Bipolar: Manic Disorder
D Dysthymic Disorder
P Alcohol Dependence
T Drug Dependence

£219. £292
S Schizotypal
C Borderline
P Paranoid

Severe Syndrome

Scale Name

 

SS Thought Disorder
CC Major Depression

PP Delusional Disorder

63
indices may be high when an individual is unusually
self-disclosing (Scale X).

In the MCMI-II, raw scores are transformed into base
rate (BR) scores, a determination based on known personality
and syndrome data. Base rate scores of 74 were fixed for
all scales as the cutting line above which scale percentages
would correspond to the clinically judged "presence" of the
personality or clinical syndrome (Millon, 1987). Scores
above 84 represent severe manifestations of personality or
clinical disorders, while scores greater than 74 are

considered to be in the pathologic range.

MCMI-II Revision Rationale

 

Millon (1987) notes several factors underlying the
decision to undertake a revision of MCMI-I. First, he calls
attention to significant changes that took place in the
theoretical base that served as both the impetus and the
foundation of the instrument. Specifically, he noted the
theoretically based derivation of two ”new" personality
disorders (aggressive/sadistic and self-defeating) and the
need to represent them by formulating corresponding MCMI
scales. Second, Millon calls attention to changes in the
theoretical characterization of several established
personality disorders and clinical syndromes (borderline,
antisocial, major depression) which required parallel
modifications and refinements in items making up their

corresponding MCMI-I scales. Third, Millon sought to

64
strengthen parallels between his inventory and the new
DSM—III-R. Finally, he sought to enhance scale validities
and reduce spurious scale overlapping.

The MCMI and the MCMI-II differ in the following ways:

1. The MCMI-II contains 45 new or revised test items
that replace items in the MCMI.

2. The MCMI—II employs a new item-weighting scoring
method.

3. The MCMI-II contains five new scales; three
correction scales and two personality pattern
scales.

4. The MCMI-II Profile report sets forth a more
comprehensive clinical syndrome (Axis I) analysis
and a more detailed prognostic and treatment
implications section.

Ema

The normative population (N=1,292) consisted of two
combined "representative" normative populations. The first
was obtained by a random selection of clinicians who
routinely use the MCMI. These clinicians (N=519) were asked
to administer both the MCMI-I and the MCMI-II to one or two
of their patients, as well as to complete an Axis I and Axis
II diagnosis employing DSM-III-R criteria. Eight hundred
twenty-five cases were generated from this process.

The second patient sample was derived in similar

fashion, except that the clinicians involved were known to

have had considerable experience with the MCMI-I, as well as

65
having attended at least one of the Millon's graduate
courses or workshops on personality disorders. In this
manner, 467 cases were generated by 93 different clinicians.
Demographic and other features of the normative population

are detailed in Table 3.12.

Reliability

In an attempt to gauge the test-retest stability of the
MCMI-II, non-clinical subjects, psychiatric outpatients, and
psychiatric inpatients were administered the MCMI-II twice,
at times separated by three to five week intervals. Table
3.13 reports the correlation coefficients for these groups.
The data are separated according to clinical status and
setting, and when in the patient's treatment process the two
administrations took place. The nonclinical segment
achieved fairly consistent and high coefficients among all
scales. Millon (1987) notes that their high coefficients,
as compared to the clinical group, are likely to be a
function of the ease with which "normals" are able to
respond consistently to items reflecting extremes in social
behaviors and emotionality. In addition, he points out the
nonclinical group were not exposed to interventions designed
to modify their psychological state.

The second most "reliable" group were outpatients who
were tested at two phases during ongoing treatment. The
lowest reliability coefficients were evidenced by the two

inpatient groups. They were evaluated first at intake, and

66
Table 3.12

Demographic and Other Features of the Normative Population,
Millon (1987, p. 106)*.

 

 

 

 

 

 

 

(N=1,292)
Sex Distribution Settings/Status
Male 49.8% Outpatient 81.9%
Female 50.2% Inpatient 9.9%
Correctional 2.4%
Age Range College Center 2.1%
18-25 15.5% Other 3.8%
26-35 25.8%
36-45 35.7% Patient's Stated "Majgr Problems"
46-55 16.9% MaritaI7Family 30.8%
56 and higher 6.1% Job/School/Work 17.1%
Self-Confidence 16.8%
Marital Status Moodiness 10.9%
Never married 30.1% Ill/Tired 6.0%
First marriage 31.2% Loneliness 5.4%
Remarried 13.5% Sexual problem 4.3%
Separated 6.6% Alcohol 3.8%
Divorced 15.1% Antisocial behavior 2.7%
Widowed 1.1% Drugs 2.2%
Cohabiting 1.6%
Other .8% Duration of Recent Episode
<1 week 20.4%
Religion 1-4 weeks 21.7%
Protestant 43.7% 1-3 months 12.7%
Catholic 28.0% 3-12 months 11.8%
Jewish 5.9% Periodic 1-3 years 5.1%
Other 22.4% Continuous 1-3 years 6.2%
Periodic 3-7 years 2.6%
Ethnic/Race Continuous 3-7 years 2.3%
White 87.7% >7 years 5.5%
Black 6.9% Cannot classify 11.7%
Hispanic 4.3%
Other 1.1%

* Note: Reproduced
by permission of
National Computer
Systems, Inc.

67

.UCH

.wEmumxm nousaeou

Hocofiumz mo cofimmaeuoa

 

 

 

 

up owospouomm ”oaoz «
mm. om. mm. as. mm. Hm. umUHOmao Hmco«msHmo mm
mm. hm. mm. as. mm. mm. Godomoumoo Hohaz 00
mm. av. 44. me. mo. om. uoouonao ugoooza mm
an. mm. mp. om. mo. mm. ucmpcmomo mono 9
oo. mm. on. no. me. mm. ucoocoooo Honooaa m
an. mo. pm. me. me. me. nflexrumxo 0
en. Ho. so. vo. an. on. noon: unnflooom z
mv. mo. mm. up. mm. mm. Eu0woumeom :
mm. no. mm. on. ow. om. xumfixcm <
mm. mm. «m. mm. mm. em. oflocmumm m
cm. mo. Hm. me. mm. an. onwaumpuom 0
mm. om. no. «m. mm. mm. Homxuonnom m
me. we. mm. mm. om. mm. ocaummmoouuamm mm
mo. mm. «o. be. mo. mm. o>finnoumo<uo>onnno as
me. on. me. mm. on. ma. o>finaooeoo A
me. me. on. om. me. am. oaumanam\m>«mmmuom< mm
mm. om. mm. mm. me. mm. HoHUOmaucm <o
he. on. me. on. on. no. oflumfinmfiounz m
om. me. vs. mm. Hm. om. cacoauumflm o
on. me. mm. mm. as. mm. ucwocmmmo m
on. an. we. mm. on. mm. acmofio>< m
we. on. me. me. we. vm. vacufinom H
“Gauze nevuzv. Ammazv «mmnzv Aemuzv aflouzv manom HHquoz
mmmcoofiz mmunno mmmndoaz mongoofiz mongoofiz coaumHooom
0» umao on 0» on on Hnoocaaocoz
wmmnmmfiz oxmuca mxmucH ommnmofiz mxmucH

mucmfiummcH mucmaummuso

oonunonomno oouunnnomno

.msomcmoouwuom msoocmooumuom

 

.«Amma .o .nmmfiv codaaz .mumpHOmao sauce uo mmusou wnu ca mommco ucmuoumao
an mcoHumHsooo ucmaumacH new ucmauomuso now mucofioflwwmou Sufiaanmum mamom HHIHZOZ

ma.m manna

68
then either at a midphase of treatment or prior to
discharge. Notable in these groups, as well as in the
intake-midphase outpatient and midphase inpatient groups, is
the clear difference in reliabilities between the basic
personality disorder scales (l-BB) and those of the clinical
syndrome scales (A-PP). This finding is consistent with
reliability studies of MCMI-I (McMahon, Flynn, and Davidson,
1985; Piersma, 1986), as well as with Millon's (1981)
theoretical expectancy that personality traits and disorder
gauges are more stable than those of clinical states and

syndromes.

Internal Validity

Millon (1987) points out that each MCMI-II scale was
formulated not to be "factorially pure," but to combine the
numerous diverse and overlapping symptoms which comprise
"true" personality and clinical disorders. He sought to
achieve an optimal level of internal consistency within
scales while still preserving high levels of trait
representativeness and discriminant power. Each primary
item on the MCMI-II was selected as a function of its high
point biserial correlation with its substantive scale, thus
these scales should evidence internal consistency on
homogeneity measures. Table 3.14 summarizes data addressing
this issue with Kuder-Richardson Formula 20. The median KR
coefficients for all clinical scales is .90, with a range of

.81 to .95. Millon notes that despite the potential

69
Table 3.14

Internal Consistency Estimates of MCMI-II Scales, Millon
(1987, p. 129)*.

 

 

KR-ZO
Scale (N=825)
1 Schizoid .86
2 Avoidant .93
3 Dependent .88
4 Histrionic .90
5 Narcissistic .87
6A Antisocial .88
6B Aggressive/Sadistic .86
7 Compulsive .91
8A Passive-Aggressive .93
8B Self-Defeating .90
S Schizotypal .93
C Borderline .92
P Paranoid .90
A Anxiety .94
H Somatoform .92
N Bipolar: Manic .84
D Dysthymia .95
B Alcohol Dependent .84
T Drug Dependent .87
SS Thought Disorder .86
CC Major Depression .90
PP Delusional Disorder .81

* Note: Reproduced by
permission of National
Computer Systems, Inc.

70
complications of enhanced levels of scale heterogeneity,

these internal consistency figures are more than adequate.

External Validity

Studies undertaken by independent investigators to
examine the validity of the MCMI-I were highlighted in
Chapter Two. In order to assess the classification
efficiency of the MCMI-II, diagnostic judgements were
obtained on 703 patients. Clinicians using DSM-III-R
criteria assigned these patients diagnoses on Axis I and
Axis II. Table 3.15 outlines the efficiency data for the
MCMI-II's personality disorder scales, using the highest two
(primary and secondary) Axis II diagnoses made by these
clinicians as the criteria. For example, the antisocial
personality diagnosis was first or second in 9% of all cases
(prevalence); MCMI-I scores on this scale ranked first or
second highest among the personality disorder scales, 71%
of the time (sensitivity); of those who were not diagnosed
by clinicians as antisocial, 98% obtained their two highest
MCMI-II scores on personality scales other than the
antisocial (specificity): the MCMI-II antisocial scale's
prevalence percent correctly identified 80% of those
diagnosed antisocial by clinicians using DSM-III-R criteria
(positive predictive power); conversely, this prevalence
percent correctly identified 97% of those who were not
diagnosed as antisocial (negative predictive power); in

total, this prevalence percent paralleled clinician's

above u.um

zozmuHH onmmpmpomnpou manhunoaow.om znzHIHH 6:0 mwououn oszHHHIw Opponenoa wxum HH
ceoondmno .znqou. zmpo and moanwo ooavwaoov. zhwpoa Auooq. w. you...

 

 

 

 

mauwnn<o zoounw<o o<onnHH1
mnooponh<o muonuonp<o Unuoaounwo
wno<mpo=oo mosnpnh<wnw mucounwouww mason mason mason

women A». .a. .a» “a. .av .av
mowenoeo on an on up o9 mu
rcowmaan up um mm an on up
omvmnooan um go up an nu ma
meannnoapo no no on mm on we
znno»mo»nn»o pm a. no an on on
>bnpuoo»ow on a» om no . nu ma
ronnoaaw<oxmuouunuo we «a on a» we on
noapcpuu<o pm «a . nu ma om mo
vnnup<aurnanouuu<o no a» on a. on no
mopmnoomounusn nu mm on no mm up
mosenonwpnw om mg on no on ma
wonaonpwno Hm q» ow am om mo
monono»o on we mm mm on mu

» zono. wounoncnon ow
ponamuuuo: om znnwoanp
noapcnonu mkunoao. H30.

IL

72
judgements regarding antisocial diagnoses in 96% of all
cases (overall diagnostic power).

Millon (1987) also compared the MCMI-I and MCMI-II on
two of the aforenoted efficiency statistics, that of
sensitivity and positive predictive power. He offers that
using the ratio of positive predictive power to prevalence
is useful for evaluating both the specific disorders in
which the MCMI-II shows diagnostic improvement relative to
MCMI-I, and in gauging the overall and relative extent to
which its cutting lines prove superior to that
of chance alone. To illustrate, Table 3.16 demonstrates
that the MCMI-II is superior to MCMI-I in 10 of 11 Axis II
disorders judged by clinicians to be the primary and
secondary diagnosis. The MCMI-II proves superior to MCMI-I,
when considering personality disorders given the principal
diagnosis in nine of 11 comparable Axis II categories. In
addition, by examining Table 3.16 one can calculate the
average diagnostic efficiency in "predicting" among the top
two Axis II disorders for the MCMI-I to be that of 3.63 to
1. However, this level of efficiency improves to 5.04 to 1
for the MCMI-II. Combining all diagnoses (Axis I and II),
Millon states that the MCMI-I's diagnostic accuracy is 4.97
times greater than chance; for the MCMI-II, it is 6.44 times

greater than chance.

73

.ocH .naaunmm Housmaoo
Hoseauuz mo cowuomauwm

 

an couscoumom «ouoz c
o~.~H o~.o Ho on on «n m m noouooHo anoHnoHoo
NH.m no.» me no no Hv oH o ooHonoHooo Honor
om.~H om.» on no em on v o amouonHo newness
MH.o mv.o No Hp «A on «H HH ooooocoooo mouo
mH.o Hn.m No no no on mH pH oocooooooo HonooH¢
«A.H mo.~ on on Ho Ho ow Ho nHaHnunHo
me.mH om.o mm mm «v on v a nun»: ”unHomHm
oo.v mo.~ mm Hm pm «v mH oH suououasom
mH.~ ~m.~ on on on no on «m Huonca
m~.oH mm.v no on on H» v mH oHocnuno
ov.v mm.n mm as Nb be m” cm oanHOpuom
om.HH oo.v no on on on o AH HoohuoNHnom
om.m oo.m vo me «A on oH v~ o>Hnnouou<no>Hnnnm
no.4 mH.~ no pm me on mH on o>HnHooeoo
oo.o oo.v on Ho He mo o mH HnHoooHu:<
oo.v om.o mo mm on on mH o oHuannHounz
mm.m mm.~ mm me me me 5H hw owcowuumam
«o.m oH.~ on on on on no em unooooooo
oe.m oo.~ on on on on Hm n~ ucooHo><
mo.o oH.o He no No no o HH oHoNHsom
HHquoz HnHzoz HHquoz Huonz HHquox Hquoz HHquoz HnHzo: oHnow
Oauom uozom >uw>wuaocom moccao>mum
ouco~n>oum\omm 0>wu0aomum
o>mummom

 

.«Hohfi .Q .bmmav :OHsz .ouoouonwo HHHIZmo mo monommmz mom can

HHH>Hancom HH on<. «exam no HHH one. unosmHm 039 HHquoz can Hquoz mo conHunosoo

mH.m manna

74

Structured Clinical Interview for DSM-III-R

 

The Structured Clinical Interview for DSM-III-R (SCID)
is a semistructured interview for determining Axis I and
Axis II disorders using DSM-III-R criteria (Spitzer et al.,
1990). It is intended for administration by trained mental
health professionals who are familiar with DSM-III-R
criteria. The subjects of the SCID may be virtually anyone,
including previously identified psychiatric patients. The
language and design of the SCID are most appropriately
suited for use with adult subjects, although it may also be
used for adolescents with minor modifications.

Two standard versions of the SCID are available for
determining whether Axis I or II disorders are present. The
SCID-P (Patient Version) is designed for use with subjects
who are identified as psychiatric patients. It contains the
following modules: SCID-P Summary Score Sheet, SCID-P
Overview, Mood Syndromes, Psychotic and Associated Symptoms,
Psychotic Disorders (Differential Diagnosis), Mood
Disorders, Psychoactive Substance Use Disorders, Anxiety
Disorders, Somatoform Disorders, Eating Disorders, and
Adjustment Disorder. For settings in which psychotic
disorders are likely to be rare or for studies in which
patients with psychotic disorders are being screened out, an
abridged version of the SCID-P (SCID-P with psychotic
screen) is available. This version replaces the B and C
modules with a combined B/C module that includes only

screening questions about psychotic symptoms. The SCID-NP

THE;+,~_

75
(nonpatient version) is for use in studies in which the
subjects are not identified as psychiatric patients. The
diagnostic modules of the SCID-NP are the same as those of
the SCID-P (with psychotic screen), with the only difference
in the two versions seen in the introductory overview
section. The SCID-NP with a psychotic screen was used in
this study. The SCID-II is for evaluating the eleven
personality disorders that appear in the main body of
DSM-III-R; it also tests for the presence of Self-defeating
Personality Disorder.

Table 3.17 shows the Axis I and II disorders that are
included in the SCID modules.

Generally, the Axis I SCID is administered in a single
sitting and takes from 60 to 90 minutes. The SCID-II is
usually administered following the Axis I SCID. There is a
SCID-II Screen (Personality Questionnaire) that can be
administered that will shorten the time that it takes to
evaluate the large number of diagnostic criteria. Each item
in the personality questionnaire corresponds to a question
in the SCID-II and the items in the personality
questionnaire set a threshold for a positive response that
is considerably lower than that of the corresponding item of
the SCID-II. Subsequently, the personality questionnaire
acts as a screening device with intentionally high rates of
false positives and few false negatives. Items circled
"yes" by the subject can be probed by the interviewer if a

sufficient number of items are answered affirmatively to

76

Table 3.17

SCID Modules, Spitzer et a1.

PSYCHOTIC DISORDERS (C)
Schizophrenia
Schizophreniform Disorder
Schizoaffective Disorder
Delusional Disorders
Brief Reactive Psychosis
Psychotic Disorder NOS

MOOD DISORDERS (D)

Bipolar Disorder

Major Depression

Dysthymia (past 2 years)

Other Bipolar Disorder (includes
Bipolar Disorder NOS and
Cyclothymia)

Depressive Disorder

Superimposed on Chronic
Psychotic Disorder (diagnosed
as Depressive Disorder NOS in
DSM-III-R)

ADJUSTMENT DISORDER (I)

SUBSTANCE USE DISORDERS (E)

Alcohol

Sedatives, Hypnotics, and
Anxiolytics

Cannabis

Stimulants

Opiods

Cocaine

Hallucinogens/PCP

Polysubstance

Other Substances

(1990).

ANXIETY DISORDERS (F)

Panic Disorder

Agoraphobia with Panic
Disorder

Social Phobia

Simple Phobia

Obsessive Compulsive
Disorder (past 6
months)

SOMATOFORM DISORDERS (G)
(current only)
Somatization Disorder
Somatoform Pain Disorder
Undifferentiated Somato-
form Disorder
Hypochondriasis

EATING DISORDERS (H)
Anorexia Nervosa
Bulimia Nervosa

PERSONALITY DISORDERS
(SCID-II)

Narcissistic

Avoidant

Dependent

Obsessive Compulsive

Passive Aggressive

Self-defeating

Paranoid

Schizotypal

Schizoid

Histrionic

Borderline

Antisocial

NOS

77
make a diagnosis of a particular disorder. Items circled
"no” are generally not probed unless there is reason to
believe that the respondent is not accurate in reporting his
or her symptoms or if the number of SCID-II items answered
”yes" is within one item of the diagnostic threshold for a
particular disorder. The SCID-II screen (Personality

Questionnaire) was used as part of the interview procedure.

Reliability

The reliability of the SCID is somewhat difficult to
assess. Spitzer and his associates (1990) note that the
reliability of this instrument is greatly a function of the
particular circumstances in which it is utilized due to the
fact that the SCID is not a fully-structured interview and
that it requires the clinical judgement of the interviewer.
Citing an unpublished study, Spitzer et al. (1990) state
that data was obtained on 506 pairs of interviews at six
sites, using an earlier version of the Axis I SCID.
Subjects were selected on a random basis and the
interviewers had no access to chart data or to treatment
staff. The resulting kappas varied by diagnosis and by
site, but were reported to range between 0.62 and 0.71,
rates also reported for the NIMH Diagnostic Interview
Schedule (Robins, Helzer, Crougham, and Ratcliff, 1981), and
the Schedule for Affective Disorders and Schizophrenia
(Endicott and Spitzer, 1978). The Kappas for the SCID-II on

226 subjects were not reported in the latest manual (Spitzer

78
et al., 1990). The authors state only that they are similar
to test-retest Kappas for other personality assessment
instruments such as Structured Interview for DSM-III
Personality Disorders (SID-P) (Stangl, Pfohl, and Zimmerman,

1985), (Kappa = .70).

Validity
The validity of the SCID is also difficult to

determine. Spitzer et a1. (1990) note that "procedural
validity" is the congruence between the diagnoses made by
the assessment and some hypothetical "gold standard." They
suggest that such a gold standard is not available for
psychiatric diagnoses. Spitzer (1983) suggests the use of a
”LEAD" standard that could be utilized to assess the
procedural validity of structured diagnostic interviews.
The formula includes "longitudinal" assessment (L),
performed by "expert" diagnosticians (E), implementing all
data (AD) that can be gleaned. The author cautions that
despite the appeal of the "LEAD" standard, its
implementation is problematic. A review of the literature
reveals only one study implementing Spitzer's "LEAD"
standard theory. Skodal, Rosnick, Kellman, Oldham, and
Hyles (1988) employed this "LEAD" standard to compare the
procedural validity of the SCID-II with an early version of
the Personality Disorders Examination. The results
evidenced comparable but low congruence with the "LEAD"

standard.

79

Hypotheses

 

This investigation has highlighted the association
between psychiatric syndromes and somatization disorder. An
especially salient trend has emerged; namely, the important
role personality disorders appear to play in the expression
of somatization disorder. Pertinent research underscores
the high prevalence of these personality disorders among
patients who are diagnosed with somatization disorder, as
compared with expected prevalence rates in the overall
pOpulation. For the purpose of this investigation, this
overall prevalence rate of personality disorders will be
estimated at 15%, consistent with estimates set forth in
germane research (Zimmerman and Coryell, 1989). In
addition to this particular prevalence rate of personality
disorders, pertinent studies suggest a significant
relationship between antisocial and/or histrionic
personality disorders, and somatization disorder. In order
to empirically examine these findings, subjects suspected of
suffering from somatization disorder were referred by health
care providers. Psychiatric profiles were obtained by
administering the Structured Clinical Interview for
DSM-III-R (SCID) and the Millon Clinical Multiaxial

Inventory (MCMI-II). The following hypotheses were tested:

Hypothesis I
Patients diagnosed with somatization disorder, as

measured by the SCID, will evidence a higher rate of

80
personality disorders as measured by the SCID and/or the
MCMI-II, than would be expected in the overall population

(15%) .

Hypothesis II

 

Among those subjects diagnosed with somatization
disorder as measured by the SCID, the most frequently
occurring personality disorder diagnoses, as measured by the

SCID and/or the MCMI-II will be antisocial and histrionic.

Hypothesis III

 

Those subjects receiving a diagnosis of somatization
disorder,as measured by the SCID, will evidence at least one
other Axis I diagnosis as measured by the SCID and/or the
MCMI-II.

In addition to these formal hypotheses, additional
research questions were addressed. The number of patients
referred meeting DSM-III-R criteria for somatization
disorder was assessed as part of Hypothesis I. The
incidence and distribution of personality disorders
associated with somatization disorder was presented as part
of the analysis associated with Hypothesis II. Finally, the
incidence and distribution of Axis I diagnoses associated
with somatization disorder was offered as part of the

analysis undertaken with Hypothesis III.

81
Design

The general design of this study was somewhat
descriptive. The purpose was to explore what personality
disorders, as measured by the MCMI-II and/or the SCID, were
associated with somatization disorder. In addition, a
further purpose was to investigate what other DSM-III-R,
Axis I diagnoses were concomitant with somatization
disorder. It was hoped that this research would generate
new information that would have useful implications for both

the diagnosis and treatment of somatization disorder.

Data Analysis

 

Descriptive statistics (means, standard deviations,
frequencies, etc.) were run on all the variables studied.
The variables that were statistically analyzed in the study
included:

1. Personality disorders: categorical variables.

The exact criteria for group membership was
defined as those who fulfilled DSM-III-R criteria
for personality disorders as measured by the
MCMI-II (BR 3 74) and/or the SCID (categorical
diagnosis).

2. Axis I diagnoses: categorical variables. The

exact criteria for group membership was defined as
those who fulfilled DSM-III-R criteria for Axis I
diagnoses as measured by the MCMI-II and/or the

SCID; the somatization disorder diagnosis was

82
measured by the SCID alone because the MCMI-II
does not assign the diagnosis.

3. MCMI-II clinical scale scores: continuous
variables base rate scores of 74 or above
corresponded to the presence of a clinical
syndrome.

4. SCID diagnoses: categorical variables were
measured in DSM-III-R diagnostic categories.

Specifically, Hypothesis I was examined by using the

Test for Significance of a Proportion (Bruning and Kintz,
1977) and frequency distributions. Hypotheses II and III
were investigated using frequency distributions. Unplanned
post hoc analyses were also undertaken and involved the use
of T-tests, frequency distributions, the Test for
Significance of a Proportion, and the Kappa statistic

(Cohen, 1960).

CHAPTER IV

ANALYSIS

In chapter IV, the results of the data regarding the
major hypotheses and the post hoc tests will be presented.
The following sections are included:

1. Major hypotheses.

2. Post hoc comparisons.

3. Summary.

MAJOR HYPOTHESES

 

A preliminary pool of 25 patients were identified as
possible subjects. All completed both a SCID and a MCMI-II.
Identical diagnoses obtained on both the SCID and the
MCMI-II were counted only one time. Only those subjects who
met diagnostic criteria for somatization disorder were

included in the analysis of the major hypotheses.

A surprising finding in this study was the occurrence
of only three cases of somatization disorder. It was
expected that the majority of patients referred to the study
would meet diagnostic criteria for the somatization disorder
diagnosis. Instead, only three patients were diagnosed with
this disorder. The major hypotheses under investigation
were designed to examine research questions pertinent to the
somatization disorder population. The low frequency of

83

84
somatization disorder diagnoses obtained precluded
meaningful statistical analysis of the major hypotheses.
The major hypotheses were presented in order to maintain
consistency. However, the salient investigations in this

study were undertaken as post hoc analyses.

Hypothesis I
Subjects diagnosed with somatization
disorder as measured by the SCID
will evidence a higher rate of
personality disorders as measured by
the SCID and/or the MCMI-II than
would be expected in the overall

population (15%).

The first hypothesis stating that those receiving a
somatization disorder diagnosis would evidence a higher than
expected rate of personality disorders (15% estimated in
overall population) was confirmed. Three subjects received
a somatization disorder diagnosis. Of the three, all (100%)
received at least one personality disorder diagnosis. Thus,
a statistically significant difference emerges between the
somatization disorder group and the overall population in
terms of rate of personality disorder diagnoses. Using the
Test for Significance of a Proportion, a z score of 11.97
was obtained (p = .0003). Nineteen personality disorder

diagnoses were obtained by these three subjects. The mean

85

score for the personality disorder diagnosis was 6.33 (see

Table 4.11).

Hypothesis II

 

Among those subjects with
somatization disorder as measured by
the SCID, the most frequently
occurring personality disorder
diagnoses, as measured by the SCID
and/or the MCMI-II, will be

antisocial and histrionic.

This hypothesis was not supported. The most frequently

occurring personality disorders among the three subjects

diagnosed with somatization disorder were dependent (3),

borderline (3) and passive aggressive (3). The distribution

of all Axis II diagnoses for the subjects diagnosed with

somatization disorder is found in Table 4.11.

Hypothesis III

 

Those subjects receiving a diagnosis
of somatization disorder, as
measured by the SCID, will evidence
at least one other Axis I diagnosis
as measured by the SCID and/or the

MCMI-II.

86
Table 4.11
Distribution of Axis II diagnoses (personality disorders)

for subjects diagnosed with somatization disorder as
measured by the SCID and the MCMI-II.

 

 

Diagnosis Frequency % Cum. %
Schizotypal 1 5.2 5.2
Obsessive Compulsive 1 5.2 10.4
Histrionic 2 10.6 21.0
Dependent 3 15.8 36.8
Antisocial 1 5.2 42.0
Narcissistic 1 5.2 47.2
Avoidant 2 10.6 57.8
Borderline 3 15.8 73.6
Passive Aggressive 3 15.8 89.4
Not Otherwise Specified 2 10.6 100.0

- Self-Defeating (2)

- Aggressive Sadistic (0)

 

 

19 100.0

6.33

XI
ll

sd. = 1.52

87

Hypothesis III was supported. All three subjects
received at least one other Axis I diagnosis. Twenty Axis I
diagnoses were assigned to these three subjects. The mean
score for the Axis I diagnoses was 6.66 (see Table 4.12).
The small number of subjects receiving a somatization
disorder diagnosis (n = 3) obscures any possible
interpretation of findings reached under the main

hypotheses.

POST HOC ANALYSES

The low frequency of somatization disorder diagnoses (n
= 3) required analysis of the data from additional

perspectives. Four pertinent areas were examined:

1. Analysis of the entire sample.

2. Analysis of those subjects diagnosed with
somatization disorder.

3. Analysis of those subjects diagnosed with
somatoform disorder in general.

4. Comparison of the SCID and the MCMI-II.

The first set of analyses concerned the entire sample
(N = 25). Table 4.13 depicts the distribution of Axis I
diagnoses for this group. The most frequently occurring
diagnosis was major depression (15) which accounted for

13.3% of the distribution of Axis I disorders. The mean

88
Table 4.12
Distribution of Axis I diagnoses for subjects receiving a

somatization disorder diagnosis as measured by the SCID and
the MCMI-II.

 

 

Diagnosis Frequency % Cum. %
Major Depression 3 15.0 15.0
Bipolar Manic Disorder 1 5.0 20.0
Generalized Anxiety Disorder 3 15.0 35.0
Social Phobia 1 5.0 40.0
Simple Phobia 2 10.0 50.0
Dysthymia 3 15.0 65.0
Somatization Disorder 3 15.0 80.0
Alcohol Dependence 1 5.0 85.0
Cocaine Dependence 1 5.0 90.0
Polysubstance Dependence l 5.0 95.0
Somatoform Pain Disorder 1 5.0 100.0

20 100.0
N = 3
35 = 6.66

sd. = 3.05

 

 

 

89
Table 4.13

Distribution of Axis I diagnoses for the entire sample as
measured by the SCID and the MCMI-II.

 

 

 

 

 

Diagnosis Frequency % Cum. %
Organic Mood Syndrome 3 2.7 2.7
Major Depression 15 13.3 16.0
Bipolar Manic Disorder 3 2.7 18.7
Generalized Anxiety Disorder 12 10.6 29.3
Panic Disorder 8 7.0 36.3
Social Phobia 8 7.0 43.3
Simple Phobia 6 5.3 48.6
Obsessive Compulsive Disorder 1 .9 49.5
Dysthymia 14 12.4 61.9
Undifferentiated Somatoform

Disorder 11 9.7 71.6
Somatization Disorder 3 2.7 74.3
Alcohol Dependence 9 8.0 82.3
Opiod Dependence 1 .9 83.2
Cocaine Dependence 1 .9 84.1
Cannabis Dependence 1 .9 85.0
Polysubstance Dependence 7 6.1 91.1
Sedative, Hypnotic Abuse 1 .9 92.0
Anorexia Nervosa 1 .9 92.9
Bulimia Nervosa 2 1.8 94.7
Somatoform Pain Disorder 5 4.4 99.1
Thought Disorder 1 .9 100.0

113 100.0
N = 25
I = 4.52

sd.

2.56

nu

1]

T.‘

pi

CC

90
number of diagnoses per subject for the entire sample was
4.52. Table 4.14 provides a frequency distribution for the
113 diagnoses on Axis I assigned to the entire sample. From
this Table we can see that all the subjects in the entire
sample received at least one Axis I diagnosis. In addition
it is seen that 24 out of 25 subjects received two or more

Axis I diagnoses.

The distribution of Axis II diagnoses (personality
disorders) for the entire sample (N = 25) is presented in
Table 4.15. The most frequently occurring diagnosis was
passive-aggressive (12), which accounted for almost 13% of
the distribution of Axis II diagnoses. Table 4.16 provides
the frequency distribution of these 95 Axis II diagnoses.
Two subjects, according to Table 4.16, failed to receive a
personality disorder diagnosis. Twenty-three (92%) received
at least one personality disorder diagnosis. A
statistically significant difference emerges between the
sample rate of personality disorders (92%) and the estimated
rate in the overall pOpulation (15%). Using the Test for
Significance of a Proportion, a Z score of 10.8 was noted

(p = .0003).

The second set of post hoc analyses was aimed at
examining the relationship between somatization disorder and
the entire sample. The first analysis consisted of

comparing the number of Axis I diagnoses for the

91
Table 4.14

Frequency distribution of Axis I diagnoses for the entire
sample as measured by the SCID and the MCMI-II.

 

 

 

 

 

# of Diagnoses Frequency % Cum. %
1.00 1 4.0 4.0
2.00 5 20.0 24.0
3.00 5 20.0 44.0
4.00 4 16.0 60.0
5.00 2 8.0 68.0
6.00 2 8.0 76.0
7.00 2 8.0 84.0
8.00 2 8.0 92.0
9.00 0

10.00 2 8.0 100.0
25 100.0

N = 113 diagnoses
SE = 4.56
sd. = 2.56

92
Table 4.15

Distribution of Axis II diagnoses (personality disorders)

for the entire sample as measured by the SCID and the

 

 

MCMI-II.

Diagnosis Frequency % Cum.
Paranoid 3 3.2 3.2
Schizoid 3 3.2 6.4
Schizotypal 2 2.1 8.5
Obsessive Compulsive 8 8.4 16.9
Histrionic 11 11.6 28.5
Dependent 11 11.6 40.1
Antisocial 6 6.3 46.4
Narcissistic 7 7.4 53.8
Avoidant 11 11.6 65.4
Borderline 8 8.4 73.8
Passive Aggressive 12 12.6 86.4
Not Otherwise Specified 13 13.6 100.0

- Self-defeating (7)
- Aggressive Sadistic (6)
95 100.0
N = 25
SE = 3.80

sd. = 2.44

 

93
Table 4.16

Frequency distribution of Axis II diagnoses for the entire
sample as measured by the SCID and the MCMI-II.

 

 

 

 

 

# of Diagnoses Frequency % Cum. %
.00 2 8.0 8.0
1.00 4 16.0 24.0
2.00 3 12.0 36.0
3.00 2 8.0 44.0
4.00 3 12.0 56.0
5.00 5 20.0 76.0
6.00 2 8.0 84.0
7.00 2 8.0 92.0
8.00 2 8.0 100.0
25 100.0
N = 95 diagnoses
I = 3.80
sd. = 2.44

94
somatization disorder group (n = 3) with the number of Axis
I diagnoses obtained by the rest of the sample (n = 22).
Comparison of the two groups indicated that there was no
significant difference (t = 1.56, p = .06, Table 4.17).
Thus, those diagnosed with somatization disorder did not
obtain significantly more Axis I diagnoses than the rest of
the sample. Next, the number of Axis II diagnoses for the
somatization disorder group (n = 3) were compared with the
number of Axis II diagnoses received by the rest of the
sample (n = 23). A statistically significant difference was
noted (t = 2.03, p = .02, Table 4.17), suggesting that those
diagnosed with somatization disorder obtained significantly

more diagnoses on Axis II than the rest of the sample.

The third set of post hoc analyses examined the broad
category of somatoform disorder. Although only three out of
25 subjects received a diagnosis of somatization disorder,
15 out of 25 received a diagnosis in the broader
classification of the disorder. The distribution of these
somatoform diagnoses is presented along with the remaining
Axis I diagnoses for this group in Table 4.18. Out of the
15 patients who received a diagnosis in the general
somatoform family, eight received the diagnosis of
undifferentiated somatoform disorder, three received both
the diagnosis of undifferentiated somatoform disorder and
somatoform pain disorder, three received the diagnosis of

somatization disorder alone, one received a diagnosis of

95
Table 4.17
T-tests comparing the number of Axis I and Axis II diagnoses

as measured by the SCID and the MCMI-II, for those diagnosed
with and without somatization disorder.

 

 

Number of Degrees of 1-Tail
Axis Cases Mean sd. t-value Freedom Prob.
3 6.66 3.05
I 1.56 23 .06
22 4.27 2.43
3 6.33 1.52
II 2.03 23 .02
22 3.45 2.36

n = 3 (somatization disorder)

n = 22 (remaining subjects)

96

Table 4.18

Distribution of Axis I diagnoses for subjects receiving a
general somatoform diagnosis as measured by the SCID and the

 

 

 

 

 

MCMI-II.

Diagnosis Frequency % Cum.
Organic Mood Syndrome l 1.7 1.7
Major Depression 7 11.8 13.5
Bipolar Manic Disorder 3 5.1 18.6
Generalized Anxiety Disorder 6 10.1 28.7
Panic Disorder 3 5.1 33.8
Social Phobia 2 3.4 37.2
Simple Phobia 2 3.4 40.6
Dysthymia 8 13.5 54.1
Undifferentiated Somatoform

Disorder 11 18.7 72.8
Somatization Disorder 3 5.1 77.9
Alcohol Dependence 3 5.1 83.0
Cocaine Dependence 1 1.7 84.7
Polysubstance Dependence 2 3.4 88.1
Sedative, Hypnotic Abuse 1 1.7 89.8
Anorexia Nervosa 1 1.7 91.5
Somatoform Pain Disorder 5 8.5 100.0

59 100.0
N = 15
I = 3.93

sd. = 2.25

97
both somatization disorder and somatoform pain disorder, and
one received the diagnosis of somatoform pain disorder
alone. For the purpose of analysis, this group of 15
subjects was categorized as "general somatoform disorder."
Upon examination of Table 4.18, one notes that dysthymia

(8) was the most frequently occurring Axis I diagnosis.

The distribution of Axis II diagnoses for those
receiving a general somatoform diagnosis is presented in
Table 4.19. Fourteen out of 15 received a personality
disorder diagnosis (93%). The most frequently occurring
diagnosis for the 15 subjects given a general somatoform
diagnosis was dependent (9), which accounted for 17% of the

distribution of Axis II diagnoses.

T-tests comparing the number of Axis I and Axis II
diagnoses for this general somatoform disorder group (n =
15) and the remainder of the sample (n = 10) are depicted in
Table 4.20. Comparison of the two groups indicated that
there was no significant difference between the general
somatoform disorder group and the remainder of the sample
when considering number of Axis I diagnoses (t = -1.54, p =
.07). Similar results were noted when comparing the two
groups on the number of Axis II diagnoses (t = -.83, p =
.20). Thus, number of diagnoses does not appear to
distinguish the general somatoform group from the rest of

the sample.

98
Table 4.19
Distribution of Axis II diagnoses (personality disorders)

for subjects receiving a general somatoform diagnosis as
measured by the SCID and the MCMI-II.

 

 

Diagnosis Frequency % Cum. %
Paranoid 1 1.9 1.9
Schizoid 2 3.9 5.8
Schizotypal 1 1.9 7.7
Obsessive Compulsive 5 9.6 17.3
Histrionic 7 13.5 30.8
Dependent 9 17.3 48.1
Antisocial 2 3.9 52.0
Narcissistic 4 7.7 59.7
Avoidant 5 9.6 69.3
Borderline 5 9.6 78.9
Passive Aggressive 5 9.6 88.5
Not Otherwise Specified 6 11.5 100.0

- Self-defeating (3)

- Aggressive Sadistic (3)

 

‘52— 100.0
N = 15 subjects
I = 3.46
sd. = 2.38

99
Table 4.20
T-tests comparing the number of Axis I and Axis II diagnoses

for those diagnosed with and without general somatoform
disorder as measured by the SCID and the MCMI-II.

 

 

Number of Degrees of l-tail
Axis Cases Mean sd. t—value Freedom prob.
15 3.93 2.25
I -1.54 23 .07
10 5.50 2.83
15 3.46 2.38
II -.83 23 .20
10 4.30 2.58
n = 15 (general somatoform disorder)
n = 10 (remaining subjects)

100

The final set of post hoc analyses examined the
relationship between the SCID and the MCMI-II. As
previously noted, the sample for this analysis consisted of
the 25 subjects who completed both a SCID and a MCMI-II.
One would not expect a strong correlation between the SCID
and the MCMI-II on Axis I. As noted in Chapter 1, the SCID
assesses all possible Axis I syndromes. The MCMI-II
measures only nine. However, since both instruments measure
all 12 personality disorders as defined by DSM-III-R (1987),
a strong degree of correlation should exist between these

two instruments on Axis II diagnoses.

The distribution of Axis II diagnoses as measured by
the SCID is presented in Table 4.21. The SCID assigned 29
personality disorder diagnoses. The most frequently
occurring category was borderline. The mean number of
diagnoses assigned per person was 1.16. In contrast, the
MCMI-II assigned 85 personality disorder diagnoses (Table
4.22). The most frequently evidenced categories were
passive aggressive and avoidant. The mean number of

diagnoses for each individual on the MCMI-II was 3.4.

In order to facilitate further comparison of the SCID
and the MCMI-II in terms of assignment of Axis II diagnoses,
the Kappa statistic was employed (Table 4.23). The Kappa is
a statistic for measuring agreement on nominal categories

such as diagnosis. It provides an index of interrater

101
Table 4.21

Distribution of Axis II diagnoses (personality disorders) as
measured by the SCID.

 

 

Diagnosis Frequency % Cum. %
Paranoid 1 3.5 3.5
Schizoid 2 6.8 10.3
Schizotypal 1 3.5 13.8
Obsessive Compulsive 1 3.5 17.3
Histrionic 3 10.4 27.7
Dependent 2 6.8 34.5
Antisocial 2 6.8 41.3
Narcissistic 3 10.4 51.7
Avoidant 4 13.9 65.6
Borderline 5 17.2 82.8
Passive Aggressive 3 10.4 93.2
Not Otherwise Specified 2 6.8 100.0

- Self-defeating (2)

- Aggressive Sadistic (0)

 

 

29 100.0

1.16

xl
u

sd. = 1.24

102
Table 4.22

Distribution of Axis II diagnoses (personality disorders) as
measured by the MCMI-II.

 

 

Diagnosis Frequency % Cum. %
Paranoid 2 2.4 2.4
Schizoid 3 3.6 6.0
Schizotypal 2 2.4 8.4
Obsessive Compulsive 7 8.3 16.7
Histrionic 9 10.6 27.3
Dependent 10 11.2 38.5
Antisocial 6 7.1 45.6
Narcissistic 5 5.9 51.5
Avoidant 11 13.0 64.5
Borderline 6 7.1 71.6
Passive Aggressive 11 13.0 84.6
Not Otherwise Specified 13 15.4 100.0

- Self-defeating (7)
- Aggressive Sadistic (6)
85 100.0

N = 25
32 = 3.4
sd. = 3.72

 

 

103
Table 4.23
Distribution of Axis II diagnoses (personality disorders) on

the SCID compared with the MCMI-II with corresponding
indices of agreement (Kappas).

 

 

 

 

SCID MCMI-II

Diagnosis Frequency Frequency Kappa
Paranoid 1 2 K = 0
Schizoid 2 3 K = .58
Schizotypal 1 2 K = .66
Obsessive Compulsive 1 7 K = .05

Histrionic 3 9 K = 0
Dependent 2 10 K = .04
Antisocial 2 6 K = .44
Narcissistic 3 5 K = .14
Avoidant 4 11 K = .39
Borderline 5 6 K = .43
Passive Aggressive 3 11 K = .13
Not Otherwise Specified , 2 13 K = .18
— Self-defeating (2) (7) K = .20

- Aggressive Sadistic (0) (6) K = 0
29 85 *K = .23

N = 25

* overall Kappa

104
reliability that contrasts rate of agreement with rate
expected by chance (Cohen, 1960). Kappa values vary from
negative values for less than chance agreement, through "0"
for chance agreement to "+1.0" for perfect agreement.
Examination of Table 4.23 reveals that the highest level of
agreement between the SCID and the MCMI-II was on the
assignment of the schizotypal personality disorder
diagnosis (K = .66). Kappas on all 12 individual
personality disorders were provided. The overall index of
interrater reliability for the SCID compared with the
MCMI-II was K = .23, indicating a level of agreement only

slightly better than chance.

SUMMARY

The major hypotheses presented in this chapter were
designed to assess the relationship between somatization
disorder and other Axis I and Axis II diagnoses. Of primary
interest was the prediction that those diagnosed with
somatization disorder would evidence a higher rate of
personality disorders than would be expected in the overall
population. Based on the salient literature, the most
frequently occurring of these personality disorders were
predicted to be antisocial and histrionic. In addition to
personality disorders, those diagnosed with somatization
disorder were also expected to receive at least one

additional Axis I diagnosis. Two out of the three major

105
hypotheses were supported. All of the individuals (n = 3)
receiving a somatization disorder diagnosis received at
least one personality disorder diagnosis (E = 6.33, Table
4.11). These individuals also were assigned at least one
additional Axis I diagnosis (I = 6.66, Table 4.12).
However, the expectation that within the somatization
disorder group the most frequently occurring personality
disorders would be antisocial and histrionic was not
supported. The most frequently assigned personality
disorders were instead dependent (3), borderline (3), and
passive aggressive (3) (Table 4.11). The findings under the
major hypotheses are inconclusive due to the low frequency
(n = 3) of individuals receiving a somatization disorder

diagnosis.

The aforenoted low frequency of somatization disorder
diagnoses necessitated additional examination of the data.
The first set of post hoc analyses investigated the
distribution of diagnoses (Axis I and Axis II) for the
entire sample (N = 25). The entire sample averaged 4.52
Axis I diagnoses (Table 4.13) and 3.80 Axis II diagnoses
(Table 4.15). The most frequently occurring Axis I
diagnosis was major depression (15). The Axis II diagnosis

most often noted was passive aggressive (12).

The second set of post hoc analyses investigated the

relationship between the somatization disorder group and the

106

remainder of the sample. There were no significant
differences noted in terms of number of Axis I diagnoses (t
= 1.56, p = .06, Table 4.17). However, when considering
number of Axis II diagnoses, a statistically significant
difference was noted (t = 2.03, p = .02, Table 4.17). This
suggested that those diagnosed with somatization disorder
obtained significantly more Axis II diagnoses than the rest
of the sample. The strength of this relationship is
questionable in view of the low frequency (n = 3) of

somatization disorder diagnoses.

The third set of post hoc analyses examined the broad
category of somatoform disorders (n = 15). The most
frequently occurring Axis I diagnosis was dysthymia (8)
(Table 4.18). Fourteen out of 15 of the general somatoform
disorder group (93%) received at least one personality
disorder diagnosis. The most frequently occurring of these
Axis II diagnoses was dependent (9) (Table 4.19). There
were no significant differences noted in terms of number of
Axis I diagnoses (t = -1.54, p = .07, Table 4.20) or number
of Axis II diagnoses (t = -.83, p = .20, Table 4.20) between

this group and the overall sample.

The final set of post hoc analyses investigated the
relationship between the SCID and the MCMI-II. Particular
emphasis was placed on the level of agreement noted between

these two instruments in terms of assignment of Axis II

107
diagnoses. The MCMI-II was found to assign more personality
disorders (85) as compared with the SCID (29). Use of the
Kappa statistic provided an index of interrater reliability
between the two instruments both in terms of individual Axis
II diagnoses and overall level of agreement. When
considering individual Axis II diagnoses, the highest Kappa
obtained was on the assignment of the schizotypal
personality disorder (K = .66, Table 4.23). The overall
rate of agreement between the SCID and the MCMI-II was only

slightly better than chance (K = .23, Table 4.23).

CHAPTER V

SUMMARY AND CONCLUSIONS

The purpose of Chapter V is to present a more in-depth
understanding of the results and limitations of the study,
and to look at clinical applications and future research

directions. The chapter is organized as follows:

1. Review of the study.
2. Conclusions regarding major hypotheses.
3. Conclusions regarding post hoc analyses.
4. Limitations.

a. Sampling limitations.

b. Measurement limitations.
5. Clinical implications.

6. Future research directions.

Review of the Study

 

This study represents a preliminary attempt to
investigate the relationship between somatization disorder
and other Axis I and Axis II diagnoses. Pertinent
literature highlights the heterogeneous nature of
somatization disorder and suggests that individuals
diagnosed with this syndrome will frequently meet inclusion
criteria for other psychiatric diagnoses. As discussed in

108

109
Chapter 1 (pages 15-17), investigating the relationship
between somatization disorder and concurrently occurring
psychiatric disorders is essential in understanding how
these disorders may interact to affect the course, prognosis
and treatment of somatization disorder. Of particular
interest in this study is the association between
somatization disorder and personality disorders. The review
of the literature in Chapter 2 (pages 22-29) suggests a
strong relationship between antisocial and/or histrionic
personality disorders and somatization disorder. It was
hoped that the investigation of patients referred with a
diagnosis of somatization disorder would result in further
support for these preliminary findings as well as further
clarification of the diagnosis of somatization disorder per

86.

Conclusions Regarding Major Hypotheses

 

Hypothesis I

 

The first hypothesis predicting that those receiving a
somatization disorder diagnosis would evidence a higher rate
of personality disorders than expected in the overall
population (15%) was supported. All three subjects
receiving a somatization disorder diagnosis also received at
least one personality disorder diagnosis (Table 4.11). The
significance of this finding is questionable in view of the

low frequency of somatization disorder cases (n = 3).

110
However, it is interesting to note that these three subjects
received a total of 19 personality disorder diagnoses (§'=
6.33, Table 4.11). This high rate of personality disorder
diagnoses is also evidenced by all subjects in the sample
(92%). As will be discussed later in this chapter, this
finding is consistent with the literature suggesting that
those who somatize evidence higher than expected rates of

personality disorders.

Hypothesis II

 

The second hypothesis which predicted that antisocial
and histrionic personality disorders would be the most
frequently occurring personality disorders among the
somatization disorder group was not supported. Dependent,
borderline, and passive aggressive were the most frequently
assigned personality disorders. As was the case with
Hypothesis I, the low frequency (n = 3) of subjects
diagnosed with somatization disorder limits the significance
of this finding. However, as will be discussed, this
finding appears somewhat consistent across the entire sample

of patients referred to the study.

Hypothesis III

The third hypothesis predicting that those diagnosed

with somatization disorder would evidence at least one other

111
Axis I diagnosis was supported. These three subjects were
assigned a total of 20 Axis I diagnoses (§'= 6.66, Table
4.12). This high rate of Axis I disorders was also noted
for the entire sample. This finding will be more fully

discussed in the post hoc analysis section of this chapter.

An obvious question emerging from the analyses of the
major hypotheses is, "Why were only three cases of
somatization disorder found?" One possible reason for
obtaining only three cases of somatization disorder may
relate to inappropriate referrals. Table 5.11 examines the
referral sources and the associated somatoform diagnoses.
Three categories of health care providers are represented.
Outpatient primary physicians referred 14 patients, 12 of
which received a diagnosis in the somatoform category.
Outpatient psychologists provided four subjects, three of
which received a somatoform diagnosis. Finally, inpatient
substance abuse physicians referred 7 patients, of whom four
were given a somatoform diagnosis. Further examination
reveals that 86% of the patients referred by primary care
physicians received a diagnosis in the somatoform family,
but only one patient out of the 14 referred met diagnostic
criteria for somatization disorder. Outpatient
psychologists evidenced a 75% rate of accuracy in the
general somatoform category, but none of their referrals
received a somatization disorder diagnosis. Inpatient

substance abuse physicians evidenced a 57% rate of

112

Table 5.11

Distribution of somatoform diagnoses by referral source.

 

 

Inpatient
Outpatient Substance
Primary Outpatient Abuse

Physician Psychologist Physician
Somatization 1 - 2
Disorder
(n=3)
Undifferentiated 7 3 1
Somatoform
Disorder
(n = 12)
Somatoform 4 - 1
Pain Disorder
(n=5)

12 (14)* 3 (4)* 4 (7) *

N = 15

* = number of total referrals

113
diagnosing general somatoform disorders. In addition, 29%
of their referrals (n = 2) were assigned a somatization

disorder diagnosis.

The reasons underlying this inconsistency in obtaining
the desired referrals are unclear. Lack of exposure to the
diagnosis of somatization disorder is one possibility. As
noted in Chapter One (pages 7-8), somatization disorder is
not a prevalent condition in the population at large.
Researchers place the overall rate at approximately 2%
(Woodruff et al., 1971; Robins et al., 1984). However,
reviews suggest that somatization disorder occurs with
enough frequency in family practice settings for the
physician to make the diagnosis on an almost daily basis
(Kaplan et al., 1988; Like et al., 1988). The prevalence
rate of this disorder in inpatient and outpatient mental

health settings is not readily available in the literature.

Another possible factor underlying the low referral
rate of patients with somatization disorder may lie in the
diagnostic process. Brown and Smith (1988) note that
despite the reported prevalence of somatization disorder in
primary care settings, medical practitioners often fail to
recognize this diagnostic entity. Oxman et al. (1983)
attribute this failure to reluctance to use DSM-III
terminology. Katon et al. (1984) suggest that the primary

care physician lacks sufficient training in diagnosing

114
emotional illness in patients who present somatically. It
is interesting to note that the lack of somatization
disorder diagnoses occurred across referral sources.
Outpatient psychologists, for example, would be expected to
be familiar with DSM-III terminology, and have extensive

training in diagnosing emotional disorders.

The aforenoted problems in the referral process may
relate to qualities in the patient that led to, or hindered,
being considered as a possible referral. The patients
referred to participate in this study were noteworthy in
terms of their high levels of depression and anxiety. It is
conceivable that these patients may have been easier to
approach than those with a true somatization disorder
diagnosis. The review of the literature in Chapter Two
revealed a significant association between somatization
disorder and antisocial personality disorders (pages 31-38),
yet antisocial personality disorders were not found to be
highly represented in the sample (6.3%, Table 4.15).

Perhaps those patients with antisocial personality disorders
may have been seen by their physician as being less
cooperative and less amenable to possible referral.
Qualities of the patient, therefore, may relate to the
difficulty encountered in securing referrals of somatization

disorder patients.

115

Problems with the screening index for somatization
disorder (Chapter One, page 7), may have contributed to the
low referral rate of patients diagnosed with somatization
disorder. This study did not designate a process whereby
information leading to a particular referral could be
documented. In order to obtain more information about what
led to a particular referral, a random sample (N = 6) of
patients receiving a general somatoform diagnosis was
selected, and their referring parties were contacted and
asked to provide information regarding what led to a
referral. This particular sample included two patients
referred by primary care physicians, three patients referred
by psychologists, and one patient referred by an inpatient

physician.

Two primary care physicians were contacted. These
individuals both referred patients they perceived as
”difficult." Both stated that symptoms off the screening
index led to the referrals (vomiting, pain in the
extremities; vomiting, menstrual pain), as well as overall
anxiety and depression. Both physicians feel that they had
exhausted all organic possibilities for their patients'
difficulties. It is noteworthy that neither of these
physicians had considered obtaining a psychiatric consult as

part of their overall treatment approach.

116

Two psychologists who had referred patients to the
study were also contacted. One stated that his patient was
referred because of excessive somatization. Vomiting and
shortness of breath were two symptoms noted off the
screening index. This patient was not described as
"difficult," but rather "lonely and depressed." The
remaining psychologist referred two patients whom he
characterized as "less than enjoyable." Multiple,
unexplained physical symptoms were noted, but this referring
party did not remember if he had used the screening index or

not as part of the referral process.

Finally, one inpatient substance abuse physician was
contacted. He had referred a patient whom he thought had
every one of the seven symptoms noted on the screening
index. This individual had a long history of multiple

hospitalizations for surgical procedures.

A total of six patients were reviewed. Only one in the
aforenoted survey received a somatization disorder diagnosis
(referred by inpatient substance abuse physician; thought to
have all seven symptoms off the screening index). All of
the remaining five received a diagnosis in the broad
category of somatoform disorders. Four out of these five,
however, did appear to have at least two symptoms off the
screening index which, according to the index, suggests a

high likelihood of somatization disorder (American

117
Psychiatric Association, 1987). This survey suggests that
the screening index may not be sensitive enough to
accurately detect this diagnosis, especially when only two
items are considered to indicate somatization disorder, and
thus may have contributed to some of the difficulty
encountered in obtaining somatization disorder referrals.
As previously noted, documentation from each referring
party, regarding the use of the screening index, would have
provided a more meaningful basis for judging its overall

sensitivity to the diagnosis in question.

At this point, several factors which may have related
to the overall referral difficulties have been reviewed.
Lack of exposure to the diagnosis of somatization disorder,
diagnostic difficulty due to training and orientation
limitations, qualities of the patient, and problems with the
screening index were possible factors considered in
examining this low referral rate. One additional factor
related to the difficulty noted in securing somatization
disorder referrals may lie in the diagnosis of somatization
disorder itself. Does the diagnosis exist? Is there really
an homogenous group of individuals with the particular
syndrome, or, rather, are excessive somatizers, a
heterogenous group with frequently overlapping
symptomatology? The following post hoc analyses provides
,more detailed information on both the entire sample and the

subsets receiving somatoform diagnoses. Examination of such

118
a sample is essential because all subjects were referred on

the basis of a supposed somatization disorder diagnosis.

Post Hoc Analyses

 

The first set of post hoc analyses examined the entire
sample. These subjects were noteworthy as a group in terms
of their overall incidence of psychopathology. On Axis I,
all 25 subjects received at least 1 diagnosis (Table 4.14).
In addition, 23 out of 25 received at least one personality
disorder diagnosis (Table 4.16). Consistent with the
salient literature in the field, depression was the most
frequently occurring Axis I diagnosis. Passive-aggressive
personality disorder was the most frequently noted Axis II
diagnosis in the sample. These findings suggest that the
patients referred to this study evidence high levels of
psychopathology. In an attempt to determine how much of
this pathology is accounted for by "somatization" those with
somatoform disorders were compared with the remainder of the

sample.

The second set of post hoc analyses compared the
somatization disorder group with the rest of sample. There
were no significant differences observed in terms of Axis I
diagnoses (t = 1.56, p = .06, Table 4.17). However, the
somatization disorder group obtained significantly more Axis

II diagnoses than the rest of the sample (t = 2.03, p = .02,

119
Table 4.17). The meaningfulness of this finding was
obscured by the low frequency (n = 3) of somatization

disorder diagnoses.

In an attempt to further clarify the relationship
between somatization and other Axis I and Axis II diagnoses,
the three types of somatoform diagnoses assigned in this

study were collapsed into one general somatoform category

(n 15). Dysthymia and dependent personality disorder were
the most frequently occurring Axis I and Axis II disorders.
While the same general pattern of high numbers of Axis I and
Axis II diagnoses was noted, there were no significant
differences seen in terms of number of Axis I diagnoses (t =
-1.57, p = .07, Table 4.20), or number of Axis II diagnoses
(t = -.83, p = .20, Table 4.20) between this group and the

overall sample.

These findings suggest that in terms of Axis I and Axis
II disorders, the patient with a somatoform diagnosis is
indistinguishable from the patient "suspected" of having a
somatoform diagnosis. Consistent with literature in this
field, these individuals are noteworthy for their overall
excessive psychopathology. Depression, anxiety and
characterological problems are most frequently noted,
whether or not a somatization disorder or somatoform
disorder diagnosis is assigned. Two issues emerge from

these observations. First, did the presence of additional

120
Axis I and Axis II disorders contribute more to a referral
than somatization problems per se? As was previously
discussed, these patients may have "stood out" because of
such symptoms as depression, anxiety, etc. Second, does the
current DSM-III-R adequately distinguish between various
groups of somatoform disorders? Findings from this study
suggest that the somatization disorder diagnosis itself is
problematic. In addition, those identified as excessive
somatizers do not appear to fall into homogenous categories,
but rather appear to be a heterogenous group with

overlapping symptomatology.

The final set of post hoc analyses examined the
relationship between the SCID and the MCMI-II. A strong
degree of correlation was expected between these two
instruments on the assignment of Axis II diagnoses. Both
tools were specifically designed to assess for the presence of
all 12 personality disorders and both instruments were
formulated to be consistent with DSM-III-R (1987) diagnostic
criteria. Comparison of the SCID and the MCMI-II showed a
disappointingly low level of correlation. The SCID
diagnosed a total of 29 personality disorders (E = 1.16,
Table 4.21), whereas the MCMI-II assigned 85 personality
disorders (i = 3.4, Table 4.22). The Kappa statistic was
used to obtain an overall measure of diagnostic agreement
between the SCID and the MCMI-II, as well as to provide

individual indices for separate diagnoses (Table 4.23). The

121
overall level of diagnostic concordance between these two
instruments was found to be only slightly higher than chance
(K = .23). Examination of this data suggested that perhaps
the MCMI-II was over inclusive in its assignment of Axis II
(personality disorder) diagnoses. Millon (1987) made a
distinction between the "presence" of a personality disorder
(BR, or cutoff point, ;74), and the most "prominent"
syndrome (BR or cutoff, point ;85). As noted in Chapter
three, the cutoff point used for this study was BR ; 74
which, according to Millon's (1987) theory, indicates the
presence of a personality disorder. For the purpose of
further analysis, Axis II frequencies were computed for the
MCMI-II using instead the BR ; 84 cutoff point. Table 5.12
contrasts these frequencies with the original MCMI-II
frequencies, and the frequency of personality disorders
assigned by the SCID. Examination of this table reveals
that even after employing a higher cutoff score, the MCMI-II
still diagnosed significantly more personality disorders
(t = 5.42, p = .001) than the SCID. The reasons for this
inconsistency are unclear. One possible factor may be
related to how the instruments are given. The MCMI-II is a
self-administered questionnaire, whereas the SCID is a
semi-structured interview administered in this study by a
psychiatrist. It is conceivable that individuals may be
more comfortable admitting to difficulties when they are not

directly being interviewed.

122

The inclusion of two instruments both designed to
measure DSM-III-R personality disorders was a strategy
designed to increase diagnostic accuracy. The wide
disparity noted between the SCID and the MCMI-II was not
expected. The effect of this measurement discordance on the
findings obtained in this study is to obscure the strength
of the relationship between somatization disorders and
personality disorders. However, even when using the most
conservative personality disorder estimate (SCID = 29), 12
out of 25 patients (48%) still are shown to evidence at
least one personality disorder. This percentage rate is
significantly higher than the estimated occurrence of
personality disorders in the overall p0pulation (15%) (Z =
4.17, p = .0001). When considering the amount of
personality disorders evidenced by those subjects who
received a diagnosis in the broad category of somatoform
disorders (n = 15), as measured by the SCID, a statistically
significant difference still emerges. Six out of 15 (40%)
of those with a general somatoform disorder diagnosis were
assigned personality disorders which is a statistically
significant difference when compared with the estimate in

the overall population (15%) (Z = 2.14, p = .01).

In conclusion, this particular sample evidences
significantly higher rates of personality disorders than
estimated in the overall pOpulation regardless of which

instrument is used. However, the strength of this

123

Table 5.12

Axis II (personality disorder) frequencies for the MCMI-II
based on two different cutting points, compared with the
SCID*.

 

 

Subject MCMI-II (3 75) MCMI-II (3' 85) SCID
1.* 3 1 o
2. 3 3 1
3. 7 4 1
4. 5 3 o
s.* 2 1 o
6.* 2 o 5
7. 6 3 1
8.* 6 5 o
9. 5 5 o
10.* 6 5 3
11.* 2 o 1
12.* 7 4 5
13.* 1 o o
14. 2 o o
15.* 2 1 o
16.* o o 1
17.* 4 3 1
18.* o o o
19. 1 1 1
20.* 5 3 o
21.* 1 o o
22. o o o
23. 7 5 6
24. 4 3 3
25.* 4 3 0

6‘5— ‘53- 29'

* Subjects who are starred received a diagnosis in the broad
category of somatoform disorder.

124
association is difficult to assess in view of the lack of

concordance between the SCID and the MCMI—II.

Sampling Limitations

 

The subjects in this study were all referred by health
care providers in the greater Lansing, Michigan area. Due
to the infrequency of the somatization disorder diagnosis,
there was no attempt made to randomly select individuals
with the diagnosis of somatization disorder from a larger
group of individuals with the same diagnosis. All of the
available somatization disorder patients who agreed to
participate in the study and met selection criteria were
included. As previously noted, these individuals were
obtained from a variety of settings and referred by
physicians, and psychologists. No attempt was made to
measure or control for demographic variation, other than the

pre-determined selection criteria.

The sampling limitations of this study have external
validity implications (Campbell and Stanley, 1963). The
degree to which the results can be generalized is restricted
to subjects with like somatization issues and receiving
treatment in similar health care settings. Specifically,
the findings in this study are generalizable to individuals
fitting the demographic profile as described in Appendix C.

As indicated previously, the fact that only three subjects

125
were classified as having somatization disorder further
limits generalizing the findings, even to the individuals

fitting the criteria used in this study.

Measurement Limitations

 

A possible measurement error lies in the use of two
psychiatrists to administer the SCID. However, measurement
error due to inconsistencies in SCID administration appears
minimal in view of the fact that the SCID can be
administered by a lay person, and the two examiners in the

study were fourth year psychiatry residents.

Another possible source of measurement error regards
the use of both the SCID and the MCMI-II to diagnose some
Axis I variables and all Axis II variables. Earlier in this
chapter it was noted that the two instruments did not
evidence a strong relationship in terms of assigning
diagnoses. The MCMI-II diagnosed personality disorders at a
much higher rate than the SCID. In addition, the two
instruments were found to evidence an extremely low level of
overall agreement on the assignment of Axis II disorders.
The effect of this inconsistency was to obscure the strength
of the relationship observed between somatization disorders
and personality disorders. A high level of concordance was
not expected between the SCID and the MCMI-II in terms of

assigning Axis I diagnoses. Thus, investigations comparing

126

the two inventories on this dimension were not implemented.

All of the design issues discussed in this section may
have affected accurate measurement of the variables of
interest. These limitations, along with the sampling
limitations noted in the previous section, limit both the
degree to which the results can be deemed significant and
the degree to which they can be generalized to the larger
pOpulation of individuals with somatoform diagnoses.
Specifically, the low number of somatization disorder
referrals (n = 3), as well as the small overall sample (n =
25), limits the overall significance of the findings. It
does appear that those evidencing excessive patterns of
somatization and/or somatoform diagnoses also demonstrate
significant levels of overall psychOpathology. The strength
of these associations was obscured by the sampling and

measurement limitations previously discussed.

Clinical Implications

The results of this study have clinical implications
for the physician and the mental health clinician. First,
the high level of psychopathology found in the individuals
referred to this study underscores the importance of
psychiatric consultation in the diagnosis and treatment of
somatoform disorders. As previously noted, those assigned a

somatoform diagnosis were for the most part

127
indistinguishable from the rest of the sample in terms of
Axis I and Axis II disorders. Psychiatric consultation in
the primary care setting would assist the physician in
making an accurate diagnosis. In addition, a psychiatric
consult would contribute to the overall treatment and
management of the patient with excessive somatization

patterns.

A related clinical implication which emerges from this
investigation is the need to use caution when using the
DSM-III-R to diagnose and distinguish between various
somatoform disorders. Results of this and similar studies
suggest that patients who display patterns of somatization
cannot be neatly assigned to pre-existing diagnostic
categories. Somatization disorder, and related somatoform
disorders, appear to be heterogenous entities which
frequently overlap both among themselves and with other Axis
I and Axis II disorders. For example, the current DSM-III-R
diagnostic criteria for somatization disorder make no
mention of psychological symptoms despite research

indicating the presence of such additional syndromes.

Findings in this investigation suggest that clinicians
exercise caution in the use of the SCID and the MCMI-II,
both in clinical and research settings. The low level of
agreement observed between these two instruments in the

assessment of personality disorders is disturbing,

128
especially considering that both tools were designed to
assess DSM-III-R clinical syndromes. Results obtained from
this investigation suggest that the MCMI-II over-diagnoses
personality disorders. Findings regarding the SCID are
unclear, but it intuitively appears to provide more
realistic assessments. However, the updated manual for the
SCID (Spitzer et al., 1990) provides no specific reliability
statistics other than reference to Kappas obtained by
similar instruments. In conclusion, both instruments appear
to need further refinement in order to be used with
confidence in either clinical or research settings. The
aforenoted clinical implications are directly related to
this investigation. Additional implications, important to
both the clinician and the researcher, are detailed in the

following section.

Future Research Directions

 

Within the present design, certain modifications could
possibly lead to a clearer understanding of the relationship
between somatization disorder and associated psychiatric
syndromes. First of all, the use of a pre-test to screen
for the presence of the somatization disorder diagnosis
appears to be necessary in order to insure obtaining the
desired sample. In addition, the use of a control group
would enable the researcher to more directly examine the

diagnosis in question, as well as provide a more reliable

129
means of determining which psychiatric syndromes are
significantly associated with somatization disorder.
Finally, a method for determining the diagnostic criteria
which resulted in a patient referral would help shed light
on the accuracy of the seven symptom checklist in screening

for the presence of somatization disorder.

Additional directions for future research emerge from
this investigation. One key issue previously referred to
involves the clinical usefulness of the somatization
disorder diagnosis, as well as the remaining somatoform
diagnoses. Research assessing the reliability and validity
of these diagnostic categories is needed. Lloyd (1986)
cautions that to have any clinical usefulness, a diagnosis
must convey information regarding the course of the
condition and preferably its response to treatment. The
somatoform diagnoses do not appear to be fundamentally
distinct categories, nor do these diagnoses carry
implication as to the type of treatment the patient is

likely to receive.

In a similar vein, future research is needed in order
to determine whether the somatoform diagnoses and their
associated psychiatric syndromes are simply manifestations
of a single disorder. Research cited in Chapter Two
(Lilienfield et al., 1986), for example, suggests that

somatization disorder in a woman with underlying histrionic

130
personality disorder is equivalent to malingering in a male
with antisocial personality disorder. This possibility also
raises the question of whether or not there may be a gender
bias in the present conceptualization of somatization and
other somatoform disorders. Future studies are needed in
order to ascertain whether somatization disorder is a
syndrome most often seen in women, and infrequently seen in
men, as suggested by the majority of researchers in the
field. Of the 25 subjects referred to this study, only six

(24%) were male.

Results of this study, and other investigations of
somatization disorder and associated psychiatric syndromes,
underscore the relationship between somatization and
depressive syndromes. Additional research is needed in
order to better understand this association. Is
somatization disorder and other somatoform syndromes really
depression in disguise? Are somatic manifestations core
symptoms of depression? Or is it possible that depressive
syndromes are more clearly specified and thus easier to
diagnose than other psychiatric syndromes? Future studies
are necessary in order to more clearly understand the role
depression plays in the manifestation of somatoform

syndromes and visa versa.

Future research regarding treatment alternatives for

somatization disorder and other somatoform syndromes is

131
needed. As noted in Chapter One (pages 16-17), only two
treatment studies have been cited to date in the pertinent
literature on somatization disorder. In order to provide
more viable avenues of treatment, more attention must be
given to those factors which may underlie the development of
somatoform disorders. For example, cultural factors may
play an important role in the presentation of somatization
disorder and other somatoform syndromes. These disorders
appear to be most often documented in the primary care
setting. In this investigation, 15 out of 25 patients (56%)
were referred by primary care physicians. This referral
rate may reflect the stigma of psychiatric illness present
in our culture. Lloyd (1985) notes that excessive
somatizers evidence an impaired ability to verbalize
emotional distress, and thus may express their feelings
through bodily sensations. Along these lines, Escobar
(1987) offers that in order to clarify the role psychiatric
syndromes play in the development of somatization syndromes,
studies are needed distinguishing between four dimensions of
somatization:

1. Somatization as a primary disorder, as in the case
of the DSM-III-R (1987) classification.

2. Somatization as an associated disorder, as when
full criteria are met for other diagnoses, such as
major depression.

3. Somatization as a masked disorder, as in masked

depression or depressive equivalents.

132

4. Somatization as a trait.
In this study, many of the findings were consistent with
Escobar's (1987) model, however the distinctions between the
four dimensions described are not clear. Further research
is needed in order to distinguish between somatization
as a primary disorder, somatization as an associated or
masked syndrome, or somatization as possibly a trait or
predisposition, similar in expression to a personality

disorder.

Certain trends have emerged, however, as a result of
this investigation. First, individuals who manifest
excessive patterns of somatization, also evidence additional
psychiatric syndromes. Second, existing DSM-III-R (1987)
categories of somatoform disorders do not appear to reliably
distinguish between somatization disorder and other
somatoform syndromes, and the corresponding diagnostic
criteria do not take into account the presence of additional
psychiatric disorders. Third, the SCID and the MCMI-II
evidence a surprisingly low level of agreement in the
assignment of DSM-III-R Axis I and Axis II disorders.
Finally, additional research is necessary in order to
further understand the relationship between excessive
somatization patterns, somatoform disorders and concomitant
psychiatric syndromes, in order to provide more accurate

diagnosis and viable avenues of treatment.

APPENDICES

133

APPENDIX A

DSM-III-R DIAGNOSTIC CRITERIA FOR SOMATIZATION DISORDER

A. A history of many physical complaints or a belief that
one is sickly, beginning before the age of 30 and
persisting for several years.

B. At least 13 symptoms from the list below. To count a
symptom as significant, the following criteria must be
met:

1. No organic pathology or pathophysiologic mechanism
(for example, a physical disorder or the effects
of injury, medication, drugs, or alcohol) to
account for the symptom or, when there is related
organic pathology, the complaint or resulting
social or occupational impairment is grossly in
excess if what would be expected from the physical
findings.

2. Has not occurred only during a panic attack.

3. Has caused the person to take medicine (other than
over-the-counter pain medication), see a doctor,

or alter lifestyle.

Symptom list:
Gastrointestinal symptoms
1. vomiting (other than during pregnancy)*

2. abdominal pain (other than when menstruating)

134
3. nausea (other than motion sickness)

4. bloating (glassy)

5. diarrhea
6. intolerance of (gets sick from) several different
foods

Pain symptoms
7. pain in extremities (*)
8. back pain
9. joint pain
10. pain during urination
11. other pain (excluding headaches)
Cardiopulmonary symptoms
12. shortness of breath when not exerting oneself*
13. palpitations
14. chest pain
15. dizziness
Conversion of pseudoneurological symptoms
16. amnesia*
17. difficulty swallowing*
18. loss of voice
19. deafness
20. double vision
21. blurred vision
22. blindness
23. fainting or loss of consciousness
24. seizure or convulsion

25. trouble walking

135

26. paralysis or muscle weakness

27. urinary retention or difficulty urinating
Sexual symptoms for the major part of the person's life
after opportunities for sexual activity

28. burning sensation in sexual organs or rectum

(other than during intercourse)*

29. sexual indifference

30. pain during intercourse

31. impotence
Female reproductive symptoms judged by the person to occur
more frequently or severely than in most women

32. painful menstruation

33. irregular menstrual periods

34. excessive menstrual bleeding

35. vomiting throughout pregnancy

 

*Note: The seven items following by asterisks may be used
to screen for the disorder; the presence of two or more of
these seven items suggests a high likelihood of the

disorder.

136

APPENDIX B

Dear Health Care Provider:

Dr. Jonathon Henry and I are conducting a study sponsored by
the university looking at the diagnosis of somatization

disorder.

The purpose of this study is to examine the relationship
between somatization disorder and co-existing psychiatric
syndromes. The study will specifically examine the
association between somatization disorder and personality
disorders, although concomitant Axis I diagnoses will also

be assessed.

The structured clinical interview for DSM-III-R (SCID) will
be used to determine the presence of somatization disorder,
as well as other Axis I diagnoses and Axis II personality
disorder diagnoses. In addition, the Millon Multiaxial
Clinical Inventory, second edition (MCMI-II), will be used
to further assess general Axis I psychiatric syndromes, and

personality disorders on Axis II.

Subject participation will involve 1) participating in a
structured clinical interview (SCID), administered by one
of two university-affiliated psychiatrists, and 2)

completing an MCMI-II inventory under the supervision of

137
this examiner, for approximately 30 minutes. The referring
physician will receive a written summary of their individual
referral's diagnostic profile. The potential benefit to the
subject lies in the added dimension, such an evaluation will
provide, to the overall approach to treatment of their
particular disorder. In addition, the referring health care
provider will have pertinent data about the patient which
will not only help clarify the diagnostic picture, but may

assist in the treatment and management of the patient.

For your convenience, a seven symptom screening index is
included. The presence of two or more of these seven items
strongly suggest a high likelihood of the presence of
somatization disorder, and is sufficient to make a referral

to our study.

Subjects referred must be between the ages of 18 and 55, and
have no history of formal thought disorder. A referral of
even one or two patients will be immensely appreciated and

have a significant impact on the research being conducted.

Thank you for your cooperation. If you have any further

questions, please do not hesitate to contact us.

138

Sincerely,

Patrice Gerard, M.A.

Doctoral Candidate, Counseling Psychology
6424 West Lake Drive

Haslett, MI 48840

Phone: 646-6622, ext. 6111 (Work)

339-0829 (Home)

Jonathon Henry, M.D.
Resident In Psychiatry (Fourth Year)
5510 Timberlane, Apt. D-3
East Lansing, MI 48823
Phone: 353-3070 (WOrk)
332-8755 (Home)

Angela Miller, M.D.

Sex:Distribution

 

.Male 6
Fenedra 19
18—25 5
26-35 6
36-45 9
46-55 5

:Marital Status
Never:married
First:marriage
liamarried
Separated
rfiAxnxxzi
Wfldowed

 

Nth§Qfl

Religion

Protestant 1
Catholic

Jewish

Other

\Ot-Ic-H

Ethnic/Race
White 25

.139

APPENDIX C
Demographic and other identifying features of the sample (N=25).

(24%)
(76%)

(20%)
(24%)
(36%)
(20%)

(28%)
(28%)
(16%)
(4%)
(16%)
(8%)

(44%)
(16%)
(4%)

(36%)

(100%)

Setting/Status
(Outpatient Primary Care
Outpatienthental Health
Inpatient Substance Abuse

 

Referral Source

 

Primary Care Physician
Outpatient Psychologist
Inpatient Physician

Patient's Stated “Major Problemsfif

H
\lhab

 

Marital/family
Job/Sohoollwork
Self-confidence
.Moodiness
Ill/Tired
Loneliness
Sexual Problems
Alcohol Problems
Drug Problems
Antisocial
Other

QHNWHnmemWLfl

(56%)
(16%)
(28%)

(56%)
(16%)
(28%)

(20%)
(12%)
(32%)
(12%)
(24%)
(16%)
(4%)

(12%)
(12%)
(4%)

(24%)

* Subjects were allowed to endorse as many major problem as applicable.

140
APPENDIX D

INFORMING AND PARTICIPATION REQUEST

My name is Patrice Gerard and I am a doctoral student in
Counseling Psychology at Michigan State University. I am
currently conducting a research project that is being

sponsored by the university.

In the field of psychology, we are beginning to realize that
the human organism must be viewed as a totality or, in other
words, as both a biological and social being. In order to
understand the total person, therefore, we must look at both
the physical and psychological dimensions. The purpose of
this study is to examine the role an individual's
personality style plays in the expression of physical

illness.

People who agree to participate in this study will be asked
to engage in a structured interview conducted by a
physician, and to fill out a personality inventory
administered by a psychologist. The interview will last 45
minutes to one hour, and the personality questionnaire will

take about 30 minutes to fill out.

Your responses to both procedures will be kept strictly

confidential. Your name will not appear on either measure.

141
Your health care provider will reeceive a summary of your

interview and questionnaire results.

Receiving information about yourself and your personality
style may be uncomfortable for you. If you experience these
feelings, it would be helpful if you would discuss them with
your health care provider. While this evaluation is
designed to help you and your health care provider better
understand how your personality style impacts your
experience with physical illness, it will not answer all
your questions about either your personality style or your

physical concerns.

This research is not part of the usual treatment provided by
your health care provider. If, at any time, you choose to

withdraw from this study, you are free to do so.

I sincerely thank you for your cooperation and appreciate
your participation. If you have any further questions

regarding the study, please feel free to contact me.

Patrice Gerard, M.A.
St. Lawrence Dimondale Center
4000 N. Michigan Rd.
Dimondale, MI 48821

(517) 646-6622 ext. 6111

142
APPENDIX E
MICHIGAN STATE UNIVERSITY
Department of Psychology

DEPARTMENTAL RESEARCH CONSENT FORM

I have freely consented to take part in a scientific
study being conducted by: Patrice Gerard, M.A.
under the supervision of: John Powell! Ph.D.
Academic Title: Professor of Counseling Psychology

The study has been explained to me and I understand the
explanation that has been given and what my participa—
tion will involve.

I am aware that responding to the instruments in this
research might lead to negative and unpleasant emotions.

I understand that I am free not to participate at all
or to discontinue my participation in this study at any
time without penalty or effect on the services to which
I am entitled.

I understand that the results of the study will be
treated in strict confidence and that I will remain
anonymous to all but the principal investigator.

I understand that my participation in the study does
not guarantee any beneficial results to me.

I understand that involvement in this study is not a
part of the usual services I am receiving from my
health care provider.

I understand that results of my evaluation will be made
available to me at my request through my referring
health care provider.

I understand that, at my request, I can receive

additional explanation of the study after my
participation is completed.

Signed:

 

Date:

 

143

APPENDIX F
TAPING, INFORMING AND PARTICIPATION REQUEST

I understand that the interview segment of the study I
have agreed to participate in will be audio-taped.

I understand that these tapes will not include my name,
but will be identified by a code number.

I understand that the only persons who will have access
to the taped interviews will be the psychiatrist
conducting the interview, the psychiatrist supervising
him, and the psychologist undertaking the study.

I understand that these tapes will be erased upon
completion of the study.

Signed:
Date:

 

 

144

References

 

American Psychiatric Association (1952). Diagnostic and
statistical manual of mental disorders. Washington,
D.C.: author.

American Psychiatric Association (1968). Diagnostic and
statistical manual of mental disorders (2nd ed.).
Washington, D.C.: author.

American Psychiatric Association (1980). Diagnostic and
statistical manual of mental disorders (3rd ed.).
Washington, D.C.: author.

 

American Psychiatric Association (1982). Ethical principles
in the conduct of research with human participants.
Washington, D.C.: author.

American Psychiatric Association (1987). Diagnostic and
statistical manual of mental disorders (3rd ed. rev.).
Washington, D.C.: author.

Arkonac, O., and Guze, S.B. (1963). A family study of
hysteria. New England Journal of Medicine, 268,
239-424.

Armitage, K.L., Schneiderman, F., and Bass, R.A. (1979).
Response of physicians to medical complaints in men and
women. Journalgof the American Medical Association,
241, 2186-2187.

Bard, L.A., and Knight, R.A. (1987). Sex offender subtying
and the MCMI. In C. Green (Ed.), Conference on the
Millon Clinical Inventories (pp. 133-137).
Minneapolis, Minn.: National Computer Systems.

Blashfield, R.K., and McElroy, R.A. (1989). Ontology of
personality disorder categories. Psychiatric Annals,
12 (3), 126-131.

Bridges, K.W., and Goldberg, D. (1985). Somatic
presentation of DSM-III psychiatric disorders in
primary care. Journal of Psychosomatics, 22, 563-569.

Brown, F.W., and Smith, G.R. (1988). Somatization disorder
in general medical settings. Psychiatric Annals, 18
(6), 353—356.

Bruning, J.L. and Kintz, B.L. (1977). Compututational
handbook of statistics. Glenview, Illinois: Scott,
Foresman and Company.

145

Bryer, J.B., Miller, J., Nelson, 8., and Krol, P. (1987).
Adult psychiatric symptoms, diagnosis, and medications
as indicators of childhood abuse. Unpublished
manuscript, Charles Rives Hospital, Willesley, M.A.

 

 

Campbell, D.T. and Stanley, J.D. (1963). Experimental and
quasi-experimental designs for research. Chicago:
Rand McNally and Company.

 

Cantwell, D.P. (1972). Psychiatric illness in the families
of hyperactive children. Archives of General
Psychiatry, 21, 414-418.

 

Chodoff, P. and Lyons, H. (1958). Hysteria, the hysterical
personality, and "hysterical conversion." American
Journal of Psychiatry, 21, 1401-1405.

Cloninger, C.R. (1986). A unified biosocial theory of
personality and its role in the development of anxiety
states. Psychiatric Developments, 3, 167-226.

 

Cloninger, C.R., and Guze, S.B. (1970). Psychiatric illness
and female criminality: the role of sociopathy and
hysteria in the antisocial woman. American Journal of
Psychiatry, 127 (3), 303-310.

 

Cloninger, C.R., Martin, R.L., Guze, S.B., and Clayton, P.J.
(1986). A prospective follow-up and family study of
somatization in men and women. American Journal of
Psychiatry, 143, 873-878.

 

 

Cohen, J.A. (1960). A coefficient of agreement for nominal
scales. Educational and Psychological Measurement, 32,
37-46.

 

Collyer, J. (1979). Psychosomatic illness in a solo
practice. Psychosomatics, 22, 762-767.

 

Coryell, W., Norton, 8.6. (1981). Briquet's syndrome
(somatization disorder) and primary depression:
Comparison of background and outcome. Comprehensive
Psychiatry, 33, 249-256.

 

 

Craig, R.J. Verinis, J.S., and wexler, S. (1985).
Personality characteristics of drug addicts and
alcoholics on the Millon Clinical Multiaxial Inventory.
Journal of Personality Assessment, 42, 156-60.

Cummings, N.A., and Vanden Bos, G.R. (1981). The
twenty-year Kaiser-Permanente experience with
psychotherapy and medical utilization: Implications
for national health policy and national health
insurance. Health Policy Quarterly, 1, 159-175.

 

146

deGruy, F., Columbia, L., and Dickinson, P. (1987).
Somatization disorder in family practice. Journal of
Family Practice, 25, 45-51.

 

 

DeSouza, C., and Othmer, E. (1984). Somatization disorder
and Briquet's syndrome. Archives of General
Psychiatry, 41, 334-336.

 

 

Endicott, J. and Spitzer, B.L. (1978). A diagnostic
interview: the schedule for affective disorders and

schizophrenia. Archives of General Psychiatry, 35,
773-782.

Escobar, J.I. (1987). Cross-cultural aspects of the
somatization trait. Hospital and Community Psychiatry,
2g, 174-1800

Feighner, J.P., Robins, E., and Guze, S.B. (1972).
Diagnostic criteria for use in psychiatric research.
Archives of General Psychiatry, 36, 56-63.

Fishbain, D.A., Goldberg, M., Meagher, R., Steele, R., and
Rosomoff, H. (1986). Male and female chronic pain
patients categorized by DSM-III psychiatric diagnostic
criteria. Pain, 26, 181-197.

Ford, C.V. (1983). The somatizing disorders: Illness as a
way of life. New York: Elsevier Biomedical.

Ford, C.V. (1987). Handbook of psychiatric differential
diagnosis. Littleton, Mass: PSG Publishers.

Frances, A., and Widiger, T. (1986). The classification of
personality disorders: An overview of problems and
solutions. In A. Frances and R. Hales (Eds.),
Psychiatric Update (pp. 240-257). Washington, D.C.:
American Psychiatric Press.

Goldsmith, S.J., Jacobsberg, L.B., and Bell, R. (1989).
Personality disorder assessment. Psychiatric Annals,
12 (3), 139-142.

Guze, S.B. (1966). Criminality and psychiatric disorder.
New York: Oxford University Press.

Guze, S.B. (1970). The role of follow-up studies: Their
contribution to diagnostic classification as applied to
hysteria. Seminars in Psychiatry, 2' 392-402.

Guze, S.B. (1975). The validity and the significance of the
clinical diagnosis of hysteria (Briquet's syndrome).
American Journal of Psychiatry, 33, 138-41.

147

Guze, S.B., Cloninger, C.R., Martin, R.L., and Clayton, P.J.
(1986). A follow-up and family study of Briquet's
syndrome. British Journal of Psychiatry, 149, 17-23.

 

Guze, S.B., Goodwin, D.W., and Crane, J.B. (1970). A
psychiatric study of the wives of convicted felons: An
example of assorted mating. American Journal of
Psychiatry, 126, 1773-1776.

 

 

Guze, S.B., Wolfgram, E.D., and McKinney, J.K. (1967).
Psychiatric illness in the families of convicted
criminals: A study of 519 first-degree relatives.
Diseases of the Nervous System, 28, 651-659.

 

Guze, S.B., Woodruff, R.A., and Clayton, P.J. (1971).
Hysteria and antisocial personality: Further evidence
of an association. American Journal of Psychiatry,
121, 957-960.

 

Hankin, J., and Oktay, J.S. (1979). Mental disorder and
primary medical care: An analytic review of the
literature. (DHEW Publication N. ADM 78-661, Series
D). Rockville, MD: National Institute of Mental
Health.

 

 

Hyler, S.E., and Sussman, N. (1984). Somatoform disorders:
Before and after DSM-III. Hospital and Community
Psychiatry, 35 (5), 469-477.

 

Kahana, R.J., and Bibring, G.L. (1964). Personality types
in medical management. In N.E. Zinberg (Ed.),
Psychiatry and medical practice in a general hospital
(pp. 108-123). New York: International University
Press.

 

Kaminsky, M.J., and Slavney, P.R. (1976). Methodology and
personality in Briquet's syndrome: A reappraisal.
American Journal of Psychiatry, 133 (1), 85-88.

Kaminsky, M.J., and Slavney, P.R. (1983). Hysterical and
obsessional features in patients with Briquet's
syndrome (somatization disorder). Psychological
Medicine, l2! 111-120.

Kaplan, C., Lipkin, M.J., and Gordon, G.R. (1988).
Somatization in primary care. Patients with
unexplained and vexing medical complaints. Journal of
General Internal Medicine, 3, 177-190.

 

Katon, W. (1985). Somatization in primary care. The
Journal of Family Practice, 3; (4), 257-258.

 

148

Katon, W., Ries, R.K., and Kleinman, A. (1984). The
prevalence of somatization in primary care.
Comprehensive Psychiatry, 33 (2), 208-215.

Kenyon, F.E. (1976). Hypochondriacal states. British
Journal of Psychiatry, 129, 1-14.

 

Kimble, R., Williams, J.G., and Agsas, S. (1975). A
comparison of two methods of diagnosing hysteria.
American Journal of Psychiatry, 132, 1197-1199.

 

Koenigsberg, H.W., Kaplan, R.D., Gilmore, M.M., and Cooper,
A.M. (1985). The relationship between syndrome and
personality disorder in DSM-III: Experience with 2,462
patients. American Journal of Psychiatry, 333 (2),
207-212.

 

Large, R.G. (1986). DSM-III diagnoses in chronic pain. The
Journal of Nervous and Mental Disease, 174 (5), 295-303.

 

Leon, J. de, Saiz-Ruiz, J., Chinchilla, A., and Morales, P.
(1987). Why do some psychiatric patients somatize?
Acta Psychiatrica Scandinavica, 13, 203-209.

 

Lichstein, P.R. (1986). Caring for the patients with
multiple somatic complaints. Southern Medical Journal,
13 (3), 310-314.

Like, R.C., Rogers, J., and McGoldrick, M. (1988). Reading
and interpreting genograms: A systematic approach.
Journal of Family Practice, 33, 407-412.

Lilienfield, S.O., Van Valkenburg, C., Larntz, R., and
Akiskal, H.S. (1986). The relationship of histrionic
personality disorder to antisocial personality and
somatization disorder. American Journal of Psychiatry,
333 (6), 718-721.

Lipowski, Z.J. (1988). Somatization: The concept and its
clinical applications. The American Journal of
Psychiatry, 145 (11), 1358-1368.

 

 

Liskow, B., Clayton, F., and WOodruff, R. (1977). Briquet's
syndrome, hysterical personality and the MMPI.
American Journal of Psychiatry, 33, 1137-1139.

Liskow, B., Othmer, E., Penick, E.C., DeSouza, C., and
Gabrielle, W. (1986b). Inpatients with Briquet's
syndrome: Presence of additional psychiatric syndromes
and MMPI results. Comprehensive Psychiatry, 31,(5),
461-470.

149

Liskow, B., Penick, E.C., Powell, B.J., Haefle, W.F., and
Campbell, J.L. (1986a). Is Briquet's syndrome a
heterogeneous disorder? American Journal of
Psychiatry, 143 (5), 626-629.

 

Lloyd, G.C. (1986). Psychiatric syndromes with a somatic
presentation. Journal of Psychosomatic Research, 33
(2), 113-120.

Loranger, A.W., Susman, V.L., and Oldham, J.H. (1987). The
personality disorder examination: a preliminary
report. Journal of Personality Disorders, 3, 1-13.

Martin, B.L. (1988). Problems in the diagnosis of
somatization disorder: Effects on research and
clinical practice. Psyghiatric Annals, 33 (6),
357-362.

Mai, P.M., and Merskey, H. (1980). Briquet's treatise on
hysteria: A synOpsis and commentary. Archives of
General Psychiatry, 31, 1401-1405.

 

 

McMahon, R.C., Flynn, P.M., and Davidson, R.S. (1985).
Stability of the personality and symptom scales of
the Millon Clinical Multiaxial Inventory. Journal of
Personality Assessment, 33 (3), 231-234.

 

Mendelson, W., Johnson, N., and Stewart, M.A. (1971).
Hyperactive children as teenagers: A follow-up study.
Journal of Nervous and Mental Disease, 153, 273-279.

Millon, T. (1981). Disorders of personality: DSM-III, Axis
33. New York: Wiley.

Millon, T. (1987). Millon Clinical Multiaxial Inventory-II
(2nd Ed.). Minnesota: National Computer Systems.

Monson, R.A., and Smith, G.R., Jr. (1983). Somatization
disorder in primary care. New England Journal of
Medicine, 308, 1464-1465.

Morrison, J.R., and Stewart, M.A. (1971). A family study of
the hyperactive child syndrome. Biological Psychiatry,
3, 189-195.

Oxman, T.E., Harrigan, J., and Kues, J. (1983). Diagnostic
patterns of family physicians for somatoform,
depressive and anxiety disorders. The Journal of
Family Practice, 31_(3), 439-446.

Perley, M.J., and Guze, S.B. (1962). Hysteria - the
stability and usefulness of clinical criteria. New
England Journal of Medicine, 226, 421-426.

150

Piersma, B.L. (1986). The stability of the Millon Clinical
Multiaxial Inventory for psychiatry inpatients.
Journal of Personality Assessment, 33, 193-197.

 

Pilowsky, I. (1970). Primary and secondary hypochrondriasm.
Acta Psychiatrica Scandinavica, 33, 273-285.

 

Purtell, J., Robins, B., and Cohen, M. (1951). Observations
on clinical aspects of hysteria: A quantitative study
of 50 hysteria patients and 156 control subjects.
Journal of the American Medical Association, 333,
902-909.

 

Rasmussen, N.E., and Avant, R.F. (1989). Somatization
disorder in family practice. American Family
Physician: Practical Therapeutics, 33 (2), 206-214.

 

Regier, D.A., Goldberg, I.D., and Taublo, C.H. (1978). The
de facto US mental health services system. Archives of
General Psychiatry, 685-693.

 

 

Regier, D.A., Myers, J.K., and Kramer, M. (1984). The NIMH
epidemiologic catchment area program. Archives of
General Psychiatry, 33, 934-940.

 

 

Robert, J.A., Ryan, J.J., McEntyre, W.L., McFarland, R.S.,
and Lips, O.J. (1985). MCMI characteristics of
DSM-III: Posttraumatic stress disorder in Vietnam
veterans. Journal of PersonalityyAssessment,‘33,
226-230.

 

Robins, E., Purtell, J.J., and Cohen, M.E. (1952).
"Hysteria” in men. New England Journal of Medicine,
246, 667-685.

 

Robins, L.N. (1966). Deviant children grow up: A
sociological and psychiatric study of sociopathic
personality. Baltimore: Williams and Wilkins.

 

 

 

Robins, L.N., Helzer, J.B., Croughan, J. and Ratcliff, R.S.
(1981). National Institute of Mental Health diagnostic
interview schedule. Its history, characteristics, and
validity. Archives of General Psychiatry, 33, 381-389.

Robins, L.N., Helzer, J.B., and Weissman, M.M. (1984).
Lifetime prevalence of specific psychiatric disorders
in three sites. Archives of General Psychiatry, 33,
948-958.

 

Scallet, A., Cloninger, C.R., and Othmer, E. (1976). The
management of chronic hysteria: A review and double-
blind trial of electrosleep and other relaxation
methods. Diseases of the Nervous System, 31, 347-353.

151

Skodol, A.E., Rosnick, L., Hellman, D., Oldham, J.M., and
Hyler, S.B. (1988). Validating structured DSM-III-R
personality disorder assessments with longitudinal
data. American Journal of Psychiatry, 145, 1297-1299.

 

Slavney, P.R., and Teitelbaum, M.L. (1985). Patients with
medically unexplained symptoms: DSM-III diagnoses and
demographic characteristics. General Hospital
Psychiatry, 3, 21-25.

 

 

Smith, G.R., Monson, R.A., and Livingston, R.L. (1985).
Somatization disorder in men. General Hospital
Psychiatry, 1, 4-8.

 

Smith, G.R., Monson, R.A., and Ray, D.C. (1986).
Psychiatric consultation in somatization disorder.
New England Journal of Medicine, 314, 1407-1413.

Smith, R.C. (1985). A clinical approach to the somatizing
patient. The Journal of Family Practice, 33 (4),

Snyder, S., and Pitts, W.M. (1986). Characterizing
somatization, hypochondriasis and hysteria in the
borderline personality disorder. Acta Psychiatrica
Scandinavica, 33, 307-314.

 

 

Spitzer, R.L. (1983). Psychiatric diagnosis: are
clinicians still necessary? Comprehensive Psychiatry,
33, 399-411.

Spitzer, R.L., Endicott, J., and Robin, E. (1978). Research
diagnostic criteria: Rationale and reliability.
Archives of General Psychiatry, 33, 773-782.

Spitzer, R.L. and Fleiss, J.L. (1974). A re-analysis of the
reliability of psychiatric diagnosis. British Journal
23 Psychiatry, 125, 341-347.

 

Spitzer, R.L., Williams, J.B.W., Gibbon, M., and First, M.
(1990). User's guidg for the structuredjclinical
3nterview for DSM-III-R. washington, D.C.: American
Psychiatric Press.

Stangl, D., Pfohl, B., and Zimmerman, M. (1985). A
structured interview for the DSM-III personality
disorders: a preliminary report. Archives of General
Psychiatry, 33, 591-596.

Tracey, H.M., Norman, D.K., and Wbisberg, L.J. (1987).
Anorexia and bulimia: A comparison of MCMI results.
In C. Green (Ed.), Conference on the Millon Clinical
Inventorigg (pp. 195-197). Minneapolis, Minn.:
National Computer Systems.

 

152

Veith, I. (1965). Hysteria: The history of a disease.
Chicago: The University of Chicago Press.

Widiger, T.A., and Rogers, J.H. (1989). Prevalence and
comorbidity of personality disorders. Psychiatric

 

Woerner, P.I., and Guze, S.B. (1968). A family and marital
study of hysteria. British Journal of Psychiatry, 114,
161-168.

Woodruff, R.A., Clayton, P.J., and Guze, S.B. (1971).
Hysteria: Studies of diagnosis, outcome and
prevalence. Journal of the American Medical
Association, 215, 425-428.

Woodruff, R.A., Goodwin, D.W., and Guze, S.B. (1974).

Psychiatric diagnoses. New York: Oxford University
Press.

Zimmerman, M., and Coryell, W. (1989). DSM-III personality
disorder diagnoses in a nonpatient sample. Psychiatric
Annals, 33 (3), 682-689.

 

Zoccolillo, M., and Cloninger, C.R. (1986). Somatization
disorder: Psychologic symptoms, social desirability
and diagnosis. Comprehensive Psychiatry, 33 (1),
65-73.

"Ii(lmnjllmr