Journal of Affectroe Disorders, 22 (1991) 49-53 0 1991 Elsevier Science Publishers B.V. 0165-0327/91/$03.50 ADONIS 0165032791000904
49
JAD 00802
Menstrual
cycle phase and psychiatric
Steven D. Targum I**,Kevin P. Caputo
admissions
2 and Susan K. Ball *
’ Hahnemunn University School of Medicine and ’ Crozer-Chester Medical Center, Upland PA 19013, U.S.A. (Received 1 October 1990) (Accepted 12 February 1991)
Summary Several surveys have demonstrated increased psychiatric admissions during the para-menstrual phases of the menstrual cycle (4-5 days before and during the onset of menses). We assessed menstrual cycle phase in 51 carefully diagnosed women at the time of emergency psychiatric admission and contrasted their cycle phase distribution with 113 normal hospital staff members assessed at random upon arrival at work. Consistent with other studies, 47% of psychiatric admissions occurred during the para-menstrual phase in contrast to 22% of staff controls (x2 = 9.27; df= 1; P = 0.002). Within the group of psychiatric patients, 33.3% of admissions occurred within 4 days of the onset of menses (x2 = 12.45; df = 6; P = 0.052). There were no significant phase differences found between major depressive and schizophrenic patients, between acutely suicidal and non-suicidal patients, and no significant correlation was noted with depression rating scales. Thus, it appears that menstrual cycle entrainment and associated late luteal phase biological changes may have additive effects which are sufficient to exacerbate the expression of psychiatric disorder in vulnerable patients, independent of their diagnosis.
Key words:
Psychiatric admissions; Menstrual cycle
Introduction The late luteal and menstrual phases of the menstrual cycle have been associated with self-destructive behavior as well as acute psychiatric disturbance in several studies (Wetzel and McClure, 1972; Dalton, 1977; Rubinow and Roy-Byrne,
Address for correspondence: Steven D. Targum, M.D., Crozer-Chester Medical Center, 1 Medical Center Blvd., Suite 202, Upland, PA 19013, U.S.A.
1984; Blumenthal and Nadelson, 1988). Dalton (1959) reported a relationship between menstrual cycle phase and admission to English psychiatric hospitals. Forty-six percent of 276 emergency psychiatric admissions occurred during the paramenstrual phases (the 4 days preceding and 4 days after the onset of menses) (Dalton, 1959). Janowsky and colleagues (1969) replicated these findings in a study conducted at Harbor General Hospital in Los Angeles. Twenty-four of 44 women (55%) were admitted within the same para-menstrual time interval as reported by Dalton. Similarly,
50
Jacobs and Charles (1970) reported a 47% psychiatric admission rate during the para-menstrual interval, and Abramowitz et al. (1982) reported 47.8% for 115 psychiatric admissions. These latter investigators found that 69.2% of 39 depressed women in their sample had been admitted to the hospital during the para-menstrual intervals, and that 41% were admitted on two specific days (the day before the onset of menstrual flow and the first day of flow). It is conceivable that the additive effects of premenstrual biological change in patients who are predisposed to psychiatric disorder may be sufficient to exacerbate acute illness and lead to hospitalization in these vulnerable individuals (Zola et al., 1979). Some studies have reported an association between cyclical recurrence of psychiatric disorder during the late luteal or menstrual phases of the menstrual cycle in support of this concept (Rubinow and Schmidt, 1987). Altematively, some might argue that sociocultural factors and biases influence the behavior of women during different phases of the menstrual cycle and may contribute to the timing of hospitalization. Finally, the menstrual cycle itself may be powerful enough to entrain a variety of episodic disorders by serving as a zeitgeber (synchronizer) of expression of that disorder (Rubinow and Schmidt, 1987). These previous studies have been either uncontrolled suhreys or anecdotal reports which have presumed a normal distribution of behaviors across the menstrual cycle in drawing their conclusions. Further, these studies did not employ DSM-IIIR criteria to ascertain differences within a psychiatric population. Presently, we will report on a prospective, controlled study of menstrual cycle phase distribution in women admitted to a psychiatric hospital compared to the cycle distribution of a group of hospital staff controls. The following questions have been asked. (1) Is there a relationship between menstrual cycle phases and psychiatric admission vs. merely showing up for work in a general hospital? (2) Is there a relationship between a depressive diagnosis and/or acute suicide attempts leading to hospitalization and distinct phases of the menstrual cycle amongst patients who have been admitted to a psychiatric hospital?
Methods
Between December, 1989 and August, 1990 female emergency admissions to the psychiatric inpatient units of the Crozer-Chester Medical Center were screened relative to their menstrual history and birth control methods as part of the psychiatric and physical assessments which occur as part of the admission process. A comparison group of hospital staff were also ascertained by asking them to complete a menstrual rating form while on the job. The questionnaires were distributed on the nursing floors by a research assistant who collected completed forms over a 3-day period. It was our assumption that work attendance would not be influenced by the current menstrual phase of the employee, and that this comparison group could serve as a putative normal distribution. Patients who were pregnant, menopausal, post-hysterectomy, taking birth control pills, or whose menstrual cycle lengths were less than 24 days or more than 35 days were excluded from this study. Fifty-one female admissions to the Inpatient Psychiatric Unit and 113 hospital staff met the above criteria. Psychiatric patients were between the ages of 18 and 45 years (mean age = 30.8 + 7.1 years) and the comparison group ranged in age between 20 and 45 years (mean age = 32.1 + 6.7 years). There was no significant difference between the ages in the two groups. Patients were diagnosed according to DSM-IIIR criteria. In addition to diagnoses, patients were categorized according to their suicidal status: acutely suicidal (the reason for the present admission), suicidal ideation on admission, and non-suicidal on admission. All patients completed the Zung Depression Rating Scale (Zung, 1965). The menstrual cycle was standardized into seven equal time intervals based upon the average cycle length of the individual interviewed. This approach was utilized by Mandell and Mandell (1967) in order to permit analysis of menstrual phase distributions between subjects with different cycle lengths. Thus, an individual with a 2%day average cycle length will have seven 4-day cycle phases, whereas an individual with a 35-day average cycle length will have seven 5-day cycle phases. Statistical analysis of the data included chi square with Yates’ correction for continuity for all
frequency comparisons including menstrual cycle phase distribution between patients and controls, between diagnostic groups of patients (major depressive disorder, schizophrenia, schizoaffective disorder), and between suicidal status categories. Results As shown in Fig. 1, menstrual cycle phases were unevenly distributed within the psychiatric patients admitted to the inpatient psychiatric units. Twenty-four of the 51 psychiatric patients (47%) were admitted during the para-menstrual phase interval of the menstrual cycle in contrast to only 25 of the 113 hospital staff (22%) who were interviewed at work (x2 = 9.27; df = 1; P = 0.002). Seventeen admitted psychiatric patients (33.3%) in contrast to only 14 controls (12.4%) were in the late luteal phase of their menstrual cycle at the time of ascertainment (x2 = 8.74; df = 1; P = 0.003). The x2 distribution within the group of psychiatric patients fell just short of statistical significance (x2 = 12.45; df = 6; P = 0.052). Fig. 2 reveals the percent distribution of menstrual cycle phases between the psychiatric patients and hospital staff controls. There is a marked decrease in psychiatric admissions during the early to mid-luteal phase of the menstrual
rI Fig.
1. Standardized menstrual cycle phase distribution psychiatric admissions and hospital staff controls.
in
r
le”rtnulCycl*
Phases Divided mo s*wn
0 ~~~~~~~~~~~~~~~~ tn=113~
Fig. 2. Menstrual atric patient
m
Equrl Time l”kNal,
Psychiabic
p,tieml _
b=511
-A
cycle phase percent distribution in psychiadmissions and hospital staff controls.
cycle (days 15-21), and a marked increase in the late luteal phase (days 25-28). Menstrual cycle phases were not significantly different between psychiatric patients with different diagnoses. Eight of 20 patients with major depressive disorder (40%), six of 12 schizophrenic patients (SOW), and four of seven schizoaffective disorder patients (57%) were admitted during the para-menstrual phases of the menstrual cycle. Suicidal status was not related to the menstrual cycle phase of the admitted patient. Eight of 14 patients admitted because of acute suicide attempts (57%) nine of 20 patients with suicidal ideation (45%), and seven of 17 non-suicidal psychiatric admissions (40%) were in the para-menstrual phases of the menstrual cycle (x2 = 0.3; df = 2; P = NS). Fig. 3 reveals the percent distribution of menstrual cycle phase within the psychiatric group based upon their suicidal status. The mean Zung Depression Scale score on admission was 62.7 + 12.9. There was no significant correlation between Zung Depression Scale scores and menstrual cycle phases (r = -0.1; P = NS). Five patients were readmitted during the 9month ascertainment period of the study. There was no correspondence between the time of each admission and the patient’s phase of the menstrual cycle. Four women were admitted twice. Each of
52
EqUaI Time
‘m AfYtelyS”iCidaIPatiem* (“Z(4)
q Suicidal
Ideation
,n=m, 0 NOR*“iCdaI Patients (“=I,,
Fig. 3. Menstrual cycle percent distribution and suicidal at the time of psychiatric admission.
I status
these four women were menstruating (interval 1) at the time of one of their admissions, but were in early to mid-cycle phases (intervals 2, 3, or 4) during the other admission. One woman was admitted three times in 9 months and was in the late luteal phase (interval 7) twice and mid-luteal phase (interval 5) on one admission. Menstrual cycle phases were not correlated with the patient’s age (r = 0.0; P = NS) or with the number of hospitalizations (r = 0.2; P = NS). Discussion The present study has confirmed an increased rate of emergency psychiatric admissions occurring within the para-menstrual phases of the menstrual cycle. The late luteal and menstrual phase of the cycle revealed a significantly higher rate of admissions in contrast to the mid-follicular and early to mid-luteal phases of the menstrual cycle of psychiatric patients. The data provide additional support for the concept that premenstrual biological and psychological changes contribute to the overt manifestation of psychiatric symptoms in vulnerable individuals. The data do not suggest that premenstrual changes are sufficient to cause psychiatric symptoms. In fact, there were no significant differences between the rate of
admissions for depressed vs. schizophrenic patients in the present study. These data differ from those of Abramowitz and colleagues (1982) who reported a 69.2% admission rate in the paramenstrual phases for depressed patients in contrast to only 36.8% among schizophrenics, but is in agreement with the study of Dalton (1959). Further, menstrual phase was not related to suicidal status among these admissions to the psychiatric hospital. Finally, there was no correlation noted between depression ratings on the Zung Depression Scale completed on admission and menstrual cycle phases (r = - 0.1; P = NS). Some limitations of this study must be noted. First, the psychiatric population was very heterogeneous with respect to medications at the time of admission, ethnicity, socioeconomic status, parity, as well as diagnosis. Secondly, it can be argued that the comparison group was not an adequate control. Our presumption that going to work bears a negligible relationship to menstrual cycle phase may not be accurate, although there was no significant difference between phases noted in our sample. A more adequate comparison group might have been emergency medical admissions occurring within the same period of ascertainment. Given these limitations, it still appears that the para-menstrual phase of the menstrual cycle (primarily late luteal phase) serves as a non-specific stressor which contributes to the overt manifestation of psychiatric disorder in vulnerable patients. What are the factors which contribute to the exacerbation of psychiatric symptomatology during the para-menstrual phases of the menstrual cycle? Extrinsic factors like changes in therapeutic compliance, expectancy biases, and interpersonal conflicts arising from temperamental shifts (e.g., irritability) may influence the symptomatic threshold in predisposed individuals. An accumulation of these extrinsic psychosocial factors in conjunction with the direct biological changes of the late luteal phase (e.g., ketosteroid fluctuations) may be sufficient to precipitate the overt manifestations of psychiatric disorder in vulnerable individuals leading to hospitalization. The consistent findings of para-menstrual exacerbation of psychiatric symptoms has therapeutic implications. Pharmacologic strategies geared toward minimizing the irritability and ten-
53
sion of the late luteal phase may be indicated in those psychiatric patients who demonstrate a pattern of late luteal affective mood changes. The menstrual cycle phase should be considered during medication adjustments, particularly when anxiolytics are used. The cyclical addition of anxiolytics in the late luteal phase may also be warranted in some patients (Harrison et al., 1990). Future studies might examine the effect of steady-state medication on premenstrual exacerbation of symptoms and the rate of psychiatric admission. For instance, it would be interesting to examine the effect of depot neuroleptics on the psychiatric admission rate of schizophrenic patients during different phases of the menstrual cycle. Similarly, lithium prophylaxis or long-term antidepressant use with measurable blood levels could be examined to determine its effect on readmission rate relative to phases of the menstrual cycle. In conclusion, clinicians must consider the para-menstrual phases of the menstrual cycle as an independent, non-specific stressful factor affecting psychiatric symptomatology in vulnerable patients.
References Abramowitz, ES., Baker, A.H. and Fleischer, S.F. (1982) Onset of depressive psychiatric crises and the menstrual cycle. Am. J. Psychiatry 139, 475-478. Blumenthal, S.J. and Nadelson, C.C. (1988) Late luteal phase dysphoric disorder (premenstrual syndromes): clinical implications. J. Clin. Psychiatry 49, 469-474. Dalton, K. (1959) Menstruation and acute psychiatric illness. Br. Med. J. 1, 1488149. Dalton, K. (1977) The Premenstrual Syndrome and Progesterone Therapy. Heinemann Medical, London. Harrison, W.M., Endicott, .I. and Nee, J. (1990) Treatment of premenstrual dysphoria with alprazolam. Arch. Gen. Psychiatry 47, 210-275. Jacobs, T.J. and Charles, E. (1970) Correlation of psychiatric symptomatology and the menstrual cycle in an outpatient population. Am. J. Psychiatry 126, 1504-1508. Janowsky. D.S., Gomey, R., Castelnuovo-Tedesco. P. and Stone, C.B. (1969) Premenstrual-menstrual increases in psychiatric hospital admission rates. Am. J. Obstet. Gynecol. 103, 189-191. Mandell, A.J. and Mandell, M.P. (1967) Suicide and the menstrual cycle. J. Am. Med. Ass. 200, 792-793. Rubinow, D.R. and Roy-Byrne, P.P. (1984) Premenstrual syndromes: overview from a methodologic perspective. Am. J. Psychiatry 141, 163-172. Rubinow, D.R. and Schmidt, P.J. (1987) Mood disorders and the menstrual cycle. J. Reprod. Med. 32, 389-394. Wetzel, R.D. and McClure Jr. J.N. (1972) Suicide and the menstrual cycle: a review. Compr. Psychiatry 13, 369-374. Zola, P., Meyerson. A.T., Reznikoff, M., Thornton, J.C. and Concool, B.M. (1979) Menstrual symptomatology and psychiatric admission. J. Psychosom. Res. 23, 241-245. Zung, W.W.K. (1965) A self-rating depression scale. Arch. Gen. Psychiatry 12, 63-70.
49
JAD 00802
Menstrual
cycle phase and psychiatric
Steven D. Targum I**,Kevin P. Caputo
admissions
2 and Susan K. Ball *
’ Hahnemunn University School of Medicine and ’ Crozer-Chester Medical Center, Upland PA 19013, U.S.A. (Received 1 October 1990) (Accepted 12 February 1991)
Summary Several surveys have demonstrated increased psychiatric admissions during the para-menstrual phases of the menstrual cycle (4-5 days before and during the onset of menses). We assessed menstrual cycle phase in 51 carefully diagnosed women at the time of emergency psychiatric admission and contrasted their cycle phase distribution with 113 normal hospital staff members assessed at random upon arrival at work. Consistent with other studies, 47% of psychiatric admissions occurred during the para-menstrual phase in contrast to 22% of staff controls (x2 = 9.27; df= 1; P = 0.002). Within the group of psychiatric patients, 33.3% of admissions occurred within 4 days of the onset of menses (x2 = 12.45; df = 6; P = 0.052). There were no significant phase differences found between major depressive and schizophrenic patients, between acutely suicidal and non-suicidal patients, and no significant correlation was noted with depression rating scales. Thus, it appears that menstrual cycle entrainment and associated late luteal phase biological changes may have additive effects which are sufficient to exacerbate the expression of psychiatric disorder in vulnerable patients, independent of their diagnosis.
Key words:
Psychiatric admissions; Menstrual cycle
Introduction The late luteal and menstrual phases of the menstrual cycle have been associated with self-destructive behavior as well as acute psychiatric disturbance in several studies (Wetzel and McClure, 1972; Dalton, 1977; Rubinow and Roy-Byrne,
Address for correspondence: Steven D. Targum, M.D., Crozer-Chester Medical Center, 1 Medical Center Blvd., Suite 202, Upland, PA 19013, U.S.A.
1984; Blumenthal and Nadelson, 1988). Dalton (1959) reported a relationship between menstrual cycle phase and admission to English psychiatric hospitals. Forty-six percent of 276 emergency psychiatric admissions occurred during the paramenstrual phases (the 4 days preceding and 4 days after the onset of menses) (Dalton, 1959). Janowsky and colleagues (1969) replicated these findings in a study conducted at Harbor General Hospital in Los Angeles. Twenty-four of 44 women (55%) were admitted within the same para-menstrual time interval as reported by Dalton. Similarly,
50
Jacobs and Charles (1970) reported a 47% psychiatric admission rate during the para-menstrual interval, and Abramowitz et al. (1982) reported 47.8% for 115 psychiatric admissions. These latter investigators found that 69.2% of 39 depressed women in their sample had been admitted to the hospital during the para-menstrual intervals, and that 41% were admitted on two specific days (the day before the onset of menstrual flow and the first day of flow). It is conceivable that the additive effects of premenstrual biological change in patients who are predisposed to psychiatric disorder may be sufficient to exacerbate acute illness and lead to hospitalization in these vulnerable individuals (Zola et al., 1979). Some studies have reported an association between cyclical recurrence of psychiatric disorder during the late luteal or menstrual phases of the menstrual cycle in support of this concept (Rubinow and Schmidt, 1987). Altematively, some might argue that sociocultural factors and biases influence the behavior of women during different phases of the menstrual cycle and may contribute to the timing of hospitalization. Finally, the menstrual cycle itself may be powerful enough to entrain a variety of episodic disorders by serving as a zeitgeber (synchronizer) of expression of that disorder (Rubinow and Schmidt, 1987). These previous studies have been either uncontrolled suhreys or anecdotal reports which have presumed a normal distribution of behaviors across the menstrual cycle in drawing their conclusions. Further, these studies did not employ DSM-IIIR criteria to ascertain differences within a psychiatric population. Presently, we will report on a prospective, controlled study of menstrual cycle phase distribution in women admitted to a psychiatric hospital compared to the cycle distribution of a group of hospital staff controls. The following questions have been asked. (1) Is there a relationship between menstrual cycle phases and psychiatric admission vs. merely showing up for work in a general hospital? (2) Is there a relationship between a depressive diagnosis and/or acute suicide attempts leading to hospitalization and distinct phases of the menstrual cycle amongst patients who have been admitted to a psychiatric hospital?
Methods
Between December, 1989 and August, 1990 female emergency admissions to the psychiatric inpatient units of the Crozer-Chester Medical Center were screened relative to their menstrual history and birth control methods as part of the psychiatric and physical assessments which occur as part of the admission process. A comparison group of hospital staff were also ascertained by asking them to complete a menstrual rating form while on the job. The questionnaires were distributed on the nursing floors by a research assistant who collected completed forms over a 3-day period. It was our assumption that work attendance would not be influenced by the current menstrual phase of the employee, and that this comparison group could serve as a putative normal distribution. Patients who were pregnant, menopausal, post-hysterectomy, taking birth control pills, or whose menstrual cycle lengths were less than 24 days or more than 35 days were excluded from this study. Fifty-one female admissions to the Inpatient Psychiatric Unit and 113 hospital staff met the above criteria. Psychiatric patients were between the ages of 18 and 45 years (mean age = 30.8 + 7.1 years) and the comparison group ranged in age between 20 and 45 years (mean age = 32.1 + 6.7 years). There was no significant difference between the ages in the two groups. Patients were diagnosed according to DSM-IIIR criteria. In addition to diagnoses, patients were categorized according to their suicidal status: acutely suicidal (the reason for the present admission), suicidal ideation on admission, and non-suicidal on admission. All patients completed the Zung Depression Rating Scale (Zung, 1965). The menstrual cycle was standardized into seven equal time intervals based upon the average cycle length of the individual interviewed. This approach was utilized by Mandell and Mandell (1967) in order to permit analysis of menstrual phase distributions between subjects with different cycle lengths. Thus, an individual with a 2%day average cycle length will have seven 4-day cycle phases, whereas an individual with a 35-day average cycle length will have seven 5-day cycle phases. Statistical analysis of the data included chi square with Yates’ correction for continuity for all
frequency comparisons including menstrual cycle phase distribution between patients and controls, between diagnostic groups of patients (major depressive disorder, schizophrenia, schizoaffective disorder), and between suicidal status categories. Results As shown in Fig. 1, menstrual cycle phases were unevenly distributed within the psychiatric patients admitted to the inpatient psychiatric units. Twenty-four of the 51 psychiatric patients (47%) were admitted during the para-menstrual phase interval of the menstrual cycle in contrast to only 25 of the 113 hospital staff (22%) who were interviewed at work (x2 = 9.27; df = 1; P = 0.002). Seventeen admitted psychiatric patients (33.3%) in contrast to only 14 controls (12.4%) were in the late luteal phase of their menstrual cycle at the time of ascertainment (x2 = 8.74; df = 1; P = 0.003). The x2 distribution within the group of psychiatric patients fell just short of statistical significance (x2 = 12.45; df = 6; P = 0.052). Fig. 2 reveals the percent distribution of menstrual cycle phases between the psychiatric patients and hospital staff controls. There is a marked decrease in psychiatric admissions during the early to mid-luteal phase of the menstrual
rI Fig.
1. Standardized menstrual cycle phase distribution psychiatric admissions and hospital staff controls.
in
r
le”rtnulCycl*
Phases Divided mo s*wn
0 ~~~~~~~~~~~~~~~~ tn=113~
Fig. 2. Menstrual atric patient
m
Equrl Time l”kNal,
Psychiabic
p,tieml _
b=511
-A
cycle phase percent distribution in psychiadmissions and hospital staff controls.
cycle (days 15-21), and a marked increase in the late luteal phase (days 25-28). Menstrual cycle phases were not significantly different between psychiatric patients with different diagnoses. Eight of 20 patients with major depressive disorder (40%), six of 12 schizophrenic patients (SOW), and four of seven schizoaffective disorder patients (57%) were admitted during the para-menstrual phases of the menstrual cycle. Suicidal status was not related to the menstrual cycle phase of the admitted patient. Eight of 14 patients admitted because of acute suicide attempts (57%) nine of 20 patients with suicidal ideation (45%), and seven of 17 non-suicidal psychiatric admissions (40%) were in the para-menstrual phases of the menstrual cycle (x2 = 0.3; df = 2; P = NS). Fig. 3 reveals the percent distribution of menstrual cycle phase within the psychiatric group based upon their suicidal status. The mean Zung Depression Scale score on admission was 62.7 + 12.9. There was no significant correlation between Zung Depression Scale scores and menstrual cycle phases (r = -0.1; P = NS). Five patients were readmitted during the 9month ascertainment period of the study. There was no correspondence between the time of each admission and the patient’s phase of the menstrual cycle. Four women were admitted twice. Each of
52
EqUaI Time
‘m AfYtelyS”iCidaIPatiem* (“Z(4)
q Suicidal
Ideation
,n=m, 0 NOR*“iCdaI Patients (“=I,,
Fig. 3. Menstrual cycle percent distribution and suicidal at the time of psychiatric admission.
I status
these four women were menstruating (interval 1) at the time of one of their admissions, but were in early to mid-cycle phases (intervals 2, 3, or 4) during the other admission. One woman was admitted three times in 9 months and was in the late luteal phase (interval 7) twice and mid-luteal phase (interval 5) on one admission. Menstrual cycle phases were not correlated with the patient’s age (r = 0.0; P = NS) or with the number of hospitalizations (r = 0.2; P = NS). Discussion The present study has confirmed an increased rate of emergency psychiatric admissions occurring within the para-menstrual phases of the menstrual cycle. The late luteal and menstrual phase of the cycle revealed a significantly higher rate of admissions in contrast to the mid-follicular and early to mid-luteal phases of the menstrual cycle of psychiatric patients. The data provide additional support for the concept that premenstrual biological and psychological changes contribute to the overt manifestation of psychiatric symptoms in vulnerable individuals. The data do not suggest that premenstrual changes are sufficient to cause psychiatric symptoms. In fact, there were no significant differences between the rate of
admissions for depressed vs. schizophrenic patients in the present study. These data differ from those of Abramowitz and colleagues (1982) who reported a 69.2% admission rate in the paramenstrual phases for depressed patients in contrast to only 36.8% among schizophrenics, but is in agreement with the study of Dalton (1959). Further, menstrual phase was not related to suicidal status among these admissions to the psychiatric hospital. Finally, there was no correlation noted between depression ratings on the Zung Depression Scale completed on admission and menstrual cycle phases (r = - 0.1; P = NS). Some limitations of this study must be noted. First, the psychiatric population was very heterogeneous with respect to medications at the time of admission, ethnicity, socioeconomic status, parity, as well as diagnosis. Secondly, it can be argued that the comparison group was not an adequate control. Our presumption that going to work bears a negligible relationship to menstrual cycle phase may not be accurate, although there was no significant difference between phases noted in our sample. A more adequate comparison group might have been emergency medical admissions occurring within the same period of ascertainment. Given these limitations, it still appears that the para-menstrual phase of the menstrual cycle (primarily late luteal phase) serves as a non-specific stressor which contributes to the overt manifestation of psychiatric disorder in vulnerable patients. What are the factors which contribute to the exacerbation of psychiatric symptomatology during the para-menstrual phases of the menstrual cycle? Extrinsic factors like changes in therapeutic compliance, expectancy biases, and interpersonal conflicts arising from temperamental shifts (e.g., irritability) may influence the symptomatic threshold in predisposed individuals. An accumulation of these extrinsic psychosocial factors in conjunction with the direct biological changes of the late luteal phase (e.g., ketosteroid fluctuations) may be sufficient to precipitate the overt manifestations of psychiatric disorder in vulnerable individuals leading to hospitalization. The consistent findings of para-menstrual exacerbation of psychiatric symptoms has therapeutic implications. Pharmacologic strategies geared toward minimizing the irritability and ten-
53
sion of the late luteal phase may be indicated in those psychiatric patients who demonstrate a pattern of late luteal affective mood changes. The menstrual cycle phase should be considered during medication adjustments, particularly when anxiolytics are used. The cyclical addition of anxiolytics in the late luteal phase may also be warranted in some patients (Harrison et al., 1990). Future studies might examine the effect of steady-state medication on premenstrual exacerbation of symptoms and the rate of psychiatric admission. For instance, it would be interesting to examine the effect of depot neuroleptics on the psychiatric admission rate of schizophrenic patients during different phases of the menstrual cycle. Similarly, lithium prophylaxis or long-term antidepressant use with measurable blood levels could be examined to determine its effect on readmission rate relative to phases of the menstrual cycle. In conclusion, clinicians must consider the para-menstrual phases of the menstrual cycle as an independent, non-specific stressful factor affecting psychiatric symptomatology in vulnerable patients.
References Abramowitz, ES., Baker, A.H. and Fleischer, S.F. (1982) Onset of depressive psychiatric crises and the menstrual cycle. Am. J. Psychiatry 139, 475-478. Blumenthal, S.J. and Nadelson, C.C. (1988) Late luteal phase dysphoric disorder (premenstrual syndromes): clinical implications. J. Clin. Psychiatry 49, 469-474. Dalton, K. (1959) Menstruation and acute psychiatric illness. Br. Med. J. 1, 1488149. Dalton, K. (1977) The Premenstrual Syndrome and Progesterone Therapy. Heinemann Medical, London. Harrison, W.M., Endicott, .I. and Nee, J. (1990) Treatment of premenstrual dysphoria with alprazolam. Arch. Gen. Psychiatry 47, 210-275. Jacobs, T.J. and Charles, E. (1970) Correlation of psychiatric symptomatology and the menstrual cycle in an outpatient population. Am. J. Psychiatry 126, 1504-1508. Janowsky. D.S., Gomey, R., Castelnuovo-Tedesco. P. and Stone, C.B. (1969) Premenstrual-menstrual increases in psychiatric hospital admission rates. Am. J. Obstet. Gynecol. 103, 189-191. Mandell, A.J. and Mandell, M.P. (1967) Suicide and the menstrual cycle. J. Am. Med. Ass. 200, 792-793. Rubinow, D.R. and Roy-Byrne, P.P. (1984) Premenstrual syndromes: overview from a methodologic perspective. Am. J. Psychiatry 141, 163-172. Rubinow, D.R. and Schmidt, P.J. (1987) Mood disorders and the menstrual cycle. J. Reprod. Med. 32, 389-394. Wetzel, R.D. and McClure Jr. J.N. (1972) Suicide and the menstrual cycle: a review. Compr. Psychiatry 13, 369-374. Zola, P., Meyerson. A.T., Reznikoff, M., Thornton, J.C. and Concool, B.M. (1979) Menstrual symptomatology and psychiatric admission. J. Psychosom. Res. 23, 241-245. Zung, W.W.K. (1965) A self-rating depression scale. Arch. Gen. Psychiatry 12, 63-70.
