Comparisons of Males and Females With DSM-III Dependent Personality Disorder

Abstract. To determine whether DSM-III dependent personality disorder (PD) differed in males and females, 30 females and 11 males with this diagnosis were selected from a psychiatric outpatient population. Standardized measures of Axis I, Axis II, and family history were used. There were no differences in age or in the prevalence of Axis I or Axis II disorders in males and females, indicating that females were probably not misdiagnosed as having dependent PD. However, relatives of males had significantly more major depressive disorder and DSM-III anxious personality disorder cluster, while relatives of females had significantly more panic disorder. This may indicate different predisposing factors to dependent PD in males and females.

Dependent personality disorder (PD) was placed in DSM-III (American Psychiatric Association, 1980) because experts in the field thought that it was a valid disorder worthy of a preliminary attempt at definition. Little empirical evidence was available to guide this attempt. The inclusion of dependent PD in DSM-III has spurred some research in the field. There is now evidence that dependent personality traits or orientation may be related to risk for physical (Greenberg and Bornstein, 1988a) or psychological disorders (Greenberg and Bornstein, 19886); that different prototypes and categorizagions of dependent personality and attachment patterns might be empirically derived (Pilkonis, 1988; West and Sheldon, 1988); and that DSM-III dependent PD might run in families (Reich, 1989a). Thus, there is some preliminary work supporting the decision to include dependent PD in DSM-III .

One worry clinicians and researchers had about dependent PD was that it might be sex biased. Although it is reasonable for some valid disorders to be more prevalent in one sex than another (e.g., histrionic PD in females and antisocial PD in males), there was concern that normally socialized female traits might be misinterpreted as pathological. To my knowledge, there is only one study with a large number of subjects that investigated the gender prevalence of dependent PD, and that study did not reveal a gender difference in prevalence (Reich, 1987). However, there is a report indicating that gender may be an important intervening variable in how dependency traits relate to depressive symptoms (Smith et al., 1988).

The present study is the first to use a standardized DSM-III personality disorder measure and to compare males and females with dependent PD on clinical, demographic, and family history variables.

Methods

Population. Patients in this study were drawn from two groups. The first group consisted of panic disorder patients who responded to advertisements recruiting subjects for a treatment trial. Axis I disorders in this group were diagnosed by a board-certified psychiatrist on the basis of the Structured Clinical Interview for DSM-III (SCID; Spitzer and Williams, 1983). All patients were required to meet DSM-III-R criteria for panic disorder (American Psychiatric Association, 1987) and to be having at least one panic attack a week. Patients were excluded if they had schizophrenia, mental retardation, organic brain syndrome, mania, obsessive-compulsive disorder, drug or alcohol abuse in the last year, or major depression that dominated or preceded their panic disorder symptoms.

The second group was drawn from a study of randomly selected new intakes to a psychiatric outpatient clinic. Patients with psychotic symptoms, organic brain syndrome, and mental retardation were excluded. Patients’ Axis I diagnoses were determined by a Master’s level interviewer using the Schedule for Affective Disorders and Schizophrenia-Lifetime Version (SADS-L; Endicott and Spitzer, 1978). The SADS-L uses Research Diagnostic Criteria (RDC; Spitzer et al., 1978).

Although different procedures were used to measure Axis I disorders, this is not so much of a problem as it may seem. The major thrust of this study is Axis II not Axis I disorders. Also, practically all Axis I disorders were excluded except depression and panic disorders, and the RDC and DSM-III do not show major differences in these areas. In addition, analysis of the different population groups separately did not indicate any differences in outcome for the two groups.

Instruments. Personality disorders in patients were determined by the Personality Diagnostic Questionnaire (PDQ; Hyler et al., 1983), a 152-item, self-administered, true/false instrument measuring all 11 DSM-III personality disorders (and some self-defeating PD criteria also). Test-retest reliability (kappa) for psychiatric outpatients at 1 month is> 0.56 for paranoid, schizotypal, antisocial borderline, avoidant, and compulsive personality disorders (Hurt et al., 1984). Preliminary work with 2-month test-retest on the PDQ gave kappas of 0.80 for the anxious personality disorder cluster, 0.75 for dependent personality, and 0.50 for avoidant personality (Reich, 19896).

The choice of a self-report instrument could be criticized on the grounds of lesser reliability or validity than an interview instrument. However, my test-retest reliabilities were performed on panic disorder patients before and after effective treatment—a fairly stringent test of reliability. Moreover, the PDQ has now been used in a number of studies which appeared to give results in line with previous studies or concurrently used personality measures (Pfohl et al., 1987; Reich et al., 19876, 1989a; Reich, 1988a; Reich and Troughton, 1988). Although there is one report stating that the PDQ does not have high agreement with clinicians (Hyler et al., 1989), standardized instruments were developed to overcome poor clinician agreement. Unfortunately, not even carefully developed DSM-III personality instruments agree well with each other (Reich et al., 1987a; Widiger et al., 1988). Data are now available indicating that the PDQ may agree as well with interview Axis II measures as they agree with each other. Kappas of agreement for dependent PD between the PDQ and SCID-II, PDQ and Personality Diagnostic Examination (PDE), and SCID-II and PDE were 0.62, 0.56, and 0.60, respectively. (The SCID-II and PDE are standardized DSM-III Axis II interview instruments.) When the PDQ diagnosis for dependent PD was compared to a “lead standard,” predictive power positive was 0.57 and predictive power negative was 0.97 (Skodol et al., 1988; S. Hyler, personal communication). Although these values are not perfect, they are as high as those of other standardized Axis II measures.

Family history measures included the Family History-Research Diagnostic Criteria (FHRDC; Andreasen et al., 1977) and the Family History for DSM-III Anxiety and Personality Disorders (FHPD; Reich et al., 1985). The FHRDC is an accepted standard in the field and is applied to RDC affective and schizophrenic disorders. The FHPD is designed to supplement the FHRDC and to measure DSM-III anxiety and personality disorder clusters. Published studies indicate that it has reasonable validity (Reich, 1988a, 1988c; Reich and Yates, 1988).

Procedures. The PDQ and other measures were given to the panic disorder patients after they had been free of medication for 1 week. (Panic patients were gradually tapered off their antianxiety medications.) For the outpatients, the measures were taken within 1 week of intake, usually before any treatment had begun. (Most of these patients had not begun treatment when tested; the others had been in treatment at most 2 or 3 days when the PDQ was administered.) Patients were asked to respond to the questionnaire “as they usually were” and not as they were during the acute illness. The instruments were administered and scored by Master’s level research assistants, whose only duties were to explain how to take tests and, if needed, to answer specific questions. Patients were diagnosed as having a personality disorder on the PDQ if they met all the personality criteria necessary for a DSM-III diagnosis and scored > 2 on the PDQ impairment distress scale.

Analyses. Patients with dependent PD (with impairment) on the PDQ were selected from the total population. These patients were then divided into two groups based on gender. The two groups were compared to each other on age, Axis I disorders and other relevant diagnostic and functional symptoms, Axis II disorders, and family history results. Fisher’s exact test was used for statistical analysis, and an a level of 0.01 was chosen (indicating that only one positive result out of 100 would be spurious). Since there are fewer than 45 comparisons made in this report, the 0.01 level seemed to be a reasonable margin of safety.

For family history results where a trend ip < 0.10) was shown, a second calculation was performed. This was based on the knowledge that in general family history methods have, at best, half the sensitivity of direct interview methods (Andreasen et al., 1977; Cohn, 1988). Extrapolating what the results would be if the sensitivity were doubled (simulating a direct interview condition) gives the extrapolated values. This method has previously been used by Baron et al. (1985).

Since some of the comparisons had a modest sample size, the power of the key comparisons was computed and reported. Power calculations performed assume an a value of 0.05, onesided with female < male, and use formulas from a medical statistics text (Colton, 1974).

Results

The sample contained 11 males and 30 females who met PDQ criteria for dependent PD. (This is higher than but not significantly different from our overall population female:male ratio of approximately 2:1, power = 0.90.) Mean age was 37.5 (SD 13.2) years for males and 35.2 (SD 10.4) for females (not significantly different, power = 0.62). The mean Global Assessment Scale score for those for whom it was available (6 males and 12 females) was 58.3 (11.5) for males and 53.7 (6.9) for females (NS, power = 0.81). For those for whom information was available (8 males and 18 females), the males had a larger, but not significantly different, social network as measured by the McFarlane scale (McFarlane et al., 1981); 7 (2.8) vs. 4.9 (4.0), power > 0.90. Table 1 provides lifetime history of Axis I disorders in the sample. As can be seen, there are no significant differences. There are also no significant differences in current Axis I diagnoses...

Download PDF to continue reading article, footnotes, tables, and references.

James H. Reich, MD, MPH, is a board-certified Forensic Psychiatrist with extensive civil psychiatry experience who has done hundreds of evaluations. His services are available for civil and some criminal law cases. Clients are assured of his personal dedication to each case. He does high quality research, expert evaluations, writes a solid report, and will testify well. Dr. Reich has been deposed over 50 times.

©Copyright - All Rights Reserved

DO NOT REPRODUCE WITHOUT WRITTEN PERMISSION BY AUTHOR.