Psychoneuroendocrinology,Vol. 17. No. 2/3, pp. 189-194, 1992
0306-4530192 $5.00+0.00 ©1992 Pergamon Press l.~d.
Printed in Great Britain
PSYCHOLOGICAL ASPECTS OF PREMENSTRUAL SYNDROME II: UTILITY OF STANDARDIZED MEASURES P. A. KEENAN, L. A. LINDAMER, and S. K. JONG Department of Psychiatry, Harper Hospital, Wayne State University, Detroit, Michigan, U.S.A. (Received 27 August 1990; in finalform 1 May 1991)
SUMMARY Depressed mood is a salient feature of Premenstrual Syndrome (PMS). Fourteen women with prospectively documented PMS and ten without PMS completed the short form of the Beck Depression Inventory (BDI) and the Zung Self-Rating Scale for Depression (Zung-D) during the follicular and late luteal phases of two consecutive menstrual cycles. The short form of the BDI assesses the more cognitive symptoms of depression, while the Zung-D identifies primarily vegetative symptoms. The short form of the BDI was sensitive to cyclic changes in the PMS women, while the Zung-D was not. The BDI items uniquely endorsed by the PMS women during the late luteal phase were pessimism, sense of failure, dissatisfaction, guilt, self-dislike, and indecision. The premenstrual dysphoria experienced by PMS women thus appears to be more cognitive than vegetative in nature. Finally, differential utility of standardized mood measures to detect premenstrual depression is suggested. The BDI proved to be the more sensitive measure.
INTRODUCTION THERE IS A QUESTIONwhether the depression associated with premenstrual syndrome (PMS) can be considered a model of major depression and, further, if traditional depression assessment measures are appropriate to detect cyclic depressive changes in women with PMS. Women with PMS report cyclic variations of mood, with increased negative mood during the late luteal phase being the most frequently occurring symptom (Frank, 1931; Endicott et al., 1981; Freem a n et al., 1985; Rubinow et al., 1986). Women with late luteal phase negative mood also report increased appetite and hypersomnia, similar to that seen frequently in atypical depression and less commonly in major depression. Accordingly, some authors suggest that PMS dysphoria is a model for atypical affective disorders (Halbreich et al., 1983). Others, however, have described premenstrual depression as negative affect, dysphoria, or sadness rather than as a syndrome of depression (McMillan & Pihl, 1987). Further, premenstrual negative affect emerged as a distinct entity in an extensive epidemiologic study of depressive symptoms (Blazer et al., 1988). Most studies o f p r e m e n s t r u a l d y s p h o r i a have e m p l o y e d questionnaires specifically Address correspondence and reprint requests to: Dr. P. A. Keenan, Department of Psychiatry, Harper Hospital, 3990 John R, Detroit M148201, USA. 189
190
p.A. KEENANet al.
developed to assess the majority of the symptoms of PMS. These include the Menstrual Distress Questionnaire (Moos, 1968), the Prospective Record of the Impact and Severity of Premenstrual Symptomatology (PRISM) (Reid, 1988), and the Premenstrual Assessment Form (Halbreich et al., 1985). Many investigators have used a visual analog scale (Rubinow et al., 1984), while others have used mood adjective check lists (McNair & Lorr, 1964). The Hamilton Depression Scale (Hamilton, 1967) also has been used, but this requires administration by the investigator in an interview format. Self-administered depression questionnaires are rarely used in PMS research, and their usefulness in this population has not been systematically evaluated. The Beck Depression Inventory (BDI) (Beck et al., 1961) and the Zung Self-Rating Scale for Depression (Zung-D) (Zung, 1965) are two self-administered questionnaires frequently used in studies of depression. There is a degree of redundancy between these measures (Davies et al., 1975; Tanaka-Matsumi & Kameoka, 1986), but there are also basic differences between the two. The BDI, the short form in particular (Beck & Beamesdorfer, 1974), emphasizes the cognitive aspects of depression (Steur et al., 1984), whereas the Zung-D (Zung, 1965) assesses both psychological and physiological concomicants of depression equally, and includes many somatic symptoms which the BDI does not. Therefore we hypothesized that the BDI would be more useful than the Zung-D in detecting changes in depressive symptoms over the menstrual cycle in women with PMS. We report here the relative usefulness of the two measures with this population. A secondary aim was to clarify the relationship between commonly recognized mood disorders and the dysphoria associated with Premenstrual Syndrome. SUBJECTS AND METHODS Subjects and Diagnostic Instruments
The subject pool was comprised of 14 women with PMS and 10 women without PMS. Subjects and procedure are described fully in a companion study (Keenan et al., 1992). Diagnosis of PMS was made using the PRISM (Reid, 1988). The PRISM was designed primarily as a clinical tool to allow rapid, visual inspection of the pattern, severity, and lifestyle impact of 23 physical and psychological symptoms commonly associated with PMS. The short form of the BDI utilizes 13 of the original 21 items of the BDI. This form differs from the original scale in that it emphasizes the cognitive aspects of depression. The Zung-D (Zung, 1965) is a 20-item scale measuring the frequency of depressive symptomatology. The validity and reliability of the Zung-D and the BDI have been documented for clinical populations.
Statistical Analyses Group (PMS vs. non-PMS) and session (follicular vs. late luteal phase) main and interaction effects were assessed by two-way repeated measures analysis of variance (ANOVA). Scoreson the Zung-D and BDI served as dependent measures. The relationship between the BDI and Zung-D scores were assessed by one-tailed Pearson correlations for each of the two-cycle phases.
RESULTS Prospective record keeping resulted in classification of 14 women in the PMS group (mean a g e + S D = 3 4 + 3 yr,) and 10 in the non-PMS group (mean age+SD =31+6.2 yr). Mean total PRISM scores (+SD) for the PMS group were 1.1 +2.08 for the follicular phase and 30.0+ 16.45 for the late luteal phase. Mean scores (+SD) for the non-PMS women were 4.3+5.2 for the follicular phase and 6.6 + 8.7 for the late luteal phase. Follicular phase testing took place an average of 8 days after the start of menstruation for both groups (PMS group, SD = 0.9; nonPMS group, SD = 1.5). Late luteal phase testing occurred an average of 3 days premenstrually for the PMS group (SD = 1.1) and 2 days premenstrually for the non-PMS group (SD = 0.9).
ASSESSINGDEPRESSIONIN PMS RESEARCH
191
Means and standard deviations for the measures of mood are listed in the table. Significant session differences were noted for State Anxiety (F [1,21] = 11.02, p < 0.01) and Trait Anxiety (F[1,22]=9.15, p
0306-4530192 $5.00+0.00 ©1992 Pergamon Press l.~d.
Printed in Great Britain
PSYCHOLOGICAL ASPECTS OF PREMENSTRUAL SYNDROME II: UTILITY OF STANDARDIZED MEASURES P. A. KEENAN, L. A. LINDAMER, and S. K. JONG Department of Psychiatry, Harper Hospital, Wayne State University, Detroit, Michigan, U.S.A. (Received 27 August 1990; in finalform 1 May 1991)
SUMMARY Depressed mood is a salient feature of Premenstrual Syndrome (PMS). Fourteen women with prospectively documented PMS and ten without PMS completed the short form of the Beck Depression Inventory (BDI) and the Zung Self-Rating Scale for Depression (Zung-D) during the follicular and late luteal phases of two consecutive menstrual cycles. The short form of the BDI assesses the more cognitive symptoms of depression, while the Zung-D identifies primarily vegetative symptoms. The short form of the BDI was sensitive to cyclic changes in the PMS women, while the Zung-D was not. The BDI items uniquely endorsed by the PMS women during the late luteal phase were pessimism, sense of failure, dissatisfaction, guilt, self-dislike, and indecision. The premenstrual dysphoria experienced by PMS women thus appears to be more cognitive than vegetative in nature. Finally, differential utility of standardized mood measures to detect premenstrual depression is suggested. The BDI proved to be the more sensitive measure.
INTRODUCTION THERE IS A QUESTIONwhether the depression associated with premenstrual syndrome (PMS) can be considered a model of major depression and, further, if traditional depression assessment measures are appropriate to detect cyclic depressive changes in women with PMS. Women with PMS report cyclic variations of mood, with increased negative mood during the late luteal phase being the most frequently occurring symptom (Frank, 1931; Endicott et al., 1981; Freem a n et al., 1985; Rubinow et al., 1986). Women with late luteal phase negative mood also report increased appetite and hypersomnia, similar to that seen frequently in atypical depression and less commonly in major depression. Accordingly, some authors suggest that PMS dysphoria is a model for atypical affective disorders (Halbreich et al., 1983). Others, however, have described premenstrual depression as negative affect, dysphoria, or sadness rather than as a syndrome of depression (McMillan & Pihl, 1987). Further, premenstrual negative affect emerged as a distinct entity in an extensive epidemiologic study of depressive symptoms (Blazer et al., 1988). Most studies o f p r e m e n s t r u a l d y s p h o r i a have e m p l o y e d questionnaires specifically Address correspondence and reprint requests to: Dr. P. A. Keenan, Department of Psychiatry, Harper Hospital, 3990 John R, Detroit M148201, USA. 189
190
p.A. KEENANet al.
developed to assess the majority of the symptoms of PMS. These include the Menstrual Distress Questionnaire (Moos, 1968), the Prospective Record of the Impact and Severity of Premenstrual Symptomatology (PRISM) (Reid, 1988), and the Premenstrual Assessment Form (Halbreich et al., 1985). Many investigators have used a visual analog scale (Rubinow et al., 1984), while others have used mood adjective check lists (McNair & Lorr, 1964). The Hamilton Depression Scale (Hamilton, 1967) also has been used, but this requires administration by the investigator in an interview format. Self-administered depression questionnaires are rarely used in PMS research, and their usefulness in this population has not been systematically evaluated. The Beck Depression Inventory (BDI) (Beck et al., 1961) and the Zung Self-Rating Scale for Depression (Zung-D) (Zung, 1965) are two self-administered questionnaires frequently used in studies of depression. There is a degree of redundancy between these measures (Davies et al., 1975; Tanaka-Matsumi & Kameoka, 1986), but there are also basic differences between the two. The BDI, the short form in particular (Beck & Beamesdorfer, 1974), emphasizes the cognitive aspects of depression (Steur et al., 1984), whereas the Zung-D (Zung, 1965) assesses both psychological and physiological concomicants of depression equally, and includes many somatic symptoms which the BDI does not. Therefore we hypothesized that the BDI would be more useful than the Zung-D in detecting changes in depressive symptoms over the menstrual cycle in women with PMS. We report here the relative usefulness of the two measures with this population. A secondary aim was to clarify the relationship between commonly recognized mood disorders and the dysphoria associated with Premenstrual Syndrome. SUBJECTS AND METHODS Subjects and Diagnostic Instruments
The subject pool was comprised of 14 women with PMS and 10 women without PMS. Subjects and procedure are described fully in a companion study (Keenan et al., 1992). Diagnosis of PMS was made using the PRISM (Reid, 1988). The PRISM was designed primarily as a clinical tool to allow rapid, visual inspection of the pattern, severity, and lifestyle impact of 23 physical and psychological symptoms commonly associated with PMS. The short form of the BDI utilizes 13 of the original 21 items of the BDI. This form differs from the original scale in that it emphasizes the cognitive aspects of depression. The Zung-D (Zung, 1965) is a 20-item scale measuring the frequency of depressive symptomatology. The validity and reliability of the Zung-D and the BDI have been documented for clinical populations.
Statistical Analyses Group (PMS vs. non-PMS) and session (follicular vs. late luteal phase) main and interaction effects were assessed by two-way repeated measures analysis of variance (ANOVA). Scoreson the Zung-D and BDI served as dependent measures. The relationship between the BDI and Zung-D scores were assessed by one-tailed Pearson correlations for each of the two-cycle phases.
RESULTS Prospective record keeping resulted in classification of 14 women in the PMS group (mean a g e + S D = 3 4 + 3 yr,) and 10 in the non-PMS group (mean age+SD =31+6.2 yr). Mean total PRISM scores (+SD) for the PMS group were 1.1 +2.08 for the follicular phase and 30.0+ 16.45 for the late luteal phase. Mean scores (+SD) for the non-PMS women were 4.3+5.2 for the follicular phase and 6.6 + 8.7 for the late luteal phase. Follicular phase testing took place an average of 8 days after the start of menstruation for both groups (PMS group, SD = 0.9; nonPMS group, SD = 1.5). Late luteal phase testing occurred an average of 3 days premenstrually for the PMS group (SD = 1.1) and 2 days premenstrually for the non-PMS group (SD = 0.9).
ASSESSINGDEPRESSIONIN PMS RESEARCH
191
Means and standard deviations for the measures of mood are listed in the table. Significant session differences were noted for State Anxiety (F [1,21] = 11.02, p < 0.01) and Trait Anxiety (F[1,22]=9.15, p
