Numerous symptoms are commonly attributed to menopause, but their associations with ovarian aging as indicated by menopausal stage are not well understood. Although there is considerable evidence that vasomotor symptoms are associated with menopausal stage, a recent report from the National Institutes of Health concluded that there is little information about whether other symptoms such as mood changes, poor sleep, irritability, and concentration difficulties are associated with menopausal stage.1 Increased information about whether menopause-related symptoms are associated with reproductive hormone changes and menopausal stage (as a proxy for ovarian aging) is important for effective treatment of these symptoms in midlife women.
In our population-based cohort of late reproductive-age women, we have assessed the women’s reports of common menopausal symptoms for 9 years as they transition to menopause. We previously reported that three of the most frequently reported symptoms—hot flushes, musculoskeletal pain, and depressed mood—were clearly and significantly associated with menopausal stage and reproductive hormone changes.2 The prevalence and severity of these symptoms increased in the menopausal transition, independent of age and other known risk factors. We also reported that three other frequently reported symptoms—poor sleep, decreased libido, and vaginal dryness—were not associated with stages in the transition to menopause.2
The purpose of this study was to examine the five remaining symptoms that are assessed in the cohort using a validated menopause symptom questionnaire3: headache, irritability, mood swings, anxiety, and concentration difficulties. The specific aims were to 1) estimate the association of each symptom with menopausal stage as defined by menstrual bleeding patterns; 2) estimate the association of each symptom with levels and fluctuations of reproductive hormones; and 3) estimate the independent associations of menopausal stage, reproductive hormones, and other risk factors with each symptom in multivariable models.
MATERIALS AND METHODS
Data are from the Penn Ovarian Aging Study, a population-based cohort identified by random-digit dialing to households in Philadelphia County, Pennsylvania, during a 16-month period in 1996–1997 and described in previous reports.4,5 At enrollment in the cohort, the participants were ages 35 to 47 years, all were premenopausal as defined by regular menstrual cycles in normal range (22–35 days) for the past three cycles and had an intact uterus and at least one ovary. Exclusion criteria for cohort enrollment included current use of hormonal or psychotropic medications, including hormonal contraception and hormone therapies; pregnancy or breastfeeding; serious health problems known to compromise ovarian function (eg, diabetes mellitus, liver disease, and breast or endometrial cancer); and alcohol or drug abuse within the past year. The Institutional Review Board of the University of Pennsylvania approved the study, and written informed consent was obtained from all enrolled women.
The women who provided symptom reports between Period 2 (when the symptom list was first included in the interview questionnaire) and Period 11, a time span of approximately 9 years, and had no current diagnosis of mood disorder were included in the present study (n=404). Mood disorders were diagnosed at each assessment period using the Primary Care Evaluation of Mental Disorders or its self-report version, the Patient Health Questionnaire.6,7Diagnostic and Statistical Manual of Mental Disorders, 4th Edition criteria were applied for diagnosis of major depressive disorder, other depression syndromes, anxiety syndromes, and somatoform disorder. Observations of pregnancy, breastfeeding, and hormone use were censored at the times of their occurrence. Observations of hysterectomy or cancer treatment were set to missing from the time of occurrence forward.
Baseline data were compared between the participants who continued throughout the study and the women who discontinued during the 9-year interval. These comparisons indicated that there were no significant baseline differences in the study variables between women who continued to the end point and the discontinuers. We previously examined demographic and hormone characteristics of the sample through phase 1 (the first 5 years of the study) and found no substantial differences in racial group or any of the study characteristics compared between the active and inactive participants.8 Only 36 women (8%) discontinued in the 5 years after that analysis, and we believe that our previous observation, that dropout was spread equally across study characteristics, remains valid.
The data were collected in 10 assessment periods. Periods 2–6 were at 9-month intervals, Periods 7–10 were at annual intervals, and Period 11 was approximately 2 years after the previous assessment, for a possible total of 9 years per woman. Each assessment period included two visits, scheduled in the first 6 days of two consecutive menstrual cycles or 1 month apart in noncycling women to obtain blood samples for the hormone measures (yielding a possible maximum of 20 blood samples per participant).
Trained research interviewers obtained all data in individual in-person interviews at the participants’ homes. The study was explained to the participants as a general women’s health study. Information on general health status and behaviors, menstrual cycle dates, premenstrual syndrome (PMS), reproductive history, current medications, menopausal symptoms, diagnosis of depressive disorders, demographics, anthropometric measures, and blood samples for hormone assays were collected at each assessment period.
Each of the study symptoms was queried using a validated symptom list that was imbedded in the structured interview questionnaire.3 The participants were asked to indicate whether each of the symptoms occurred in the past month, the frequency of its occurrence, and its severity rated on a four-point scale from 0 (none) to 3 (severe). The symptoms included in this report are headache, irritability, mood swings, anxiety, and concentration difficulties. The remaining symptoms that were queried in the symptom list were previously reported.2
Menopausal stage was identified at each assessment period using the menstrual dates at each study visit (visits were conducted within 6 days of bleeding) and the two previous menstrual dates obtained at each visit. Other confirmatory data were obtained from the daily symptom diaries that participants recorded for one menstrual cycle at each assessment period, (the diary date was used in cases of disagreement), the reported number of menstrual periods between assessments, cycle length, and number of bleeding days.
The definitions of menopausal stage were from the consensus statement from the Stages of Reproductive Aging Workshop.9 The following categories were used in this study: premenopausal, regular menstrual cycles in the 22–35–day range; late premenopausal, a change in cycle length of 7 days or more in either direction from the participant’s personal baseline at enrollment in the cohort and observed for at least one cycle in the study; early transition, changes in cycle length of 7 days or more in either direction from the participant’s personal baseline at enrollment in the cohort and observed for at least two cycles in the study or two months amenorrhea; late transition, 3–11 months amenorrhea; postmenopausal, 12 months or more amenorrhea excluding hysterectomy.
Nonfasting blood samples for hormone assays were collected at each study visit. Assays of estradiol (E2), follicle-stimulating hormone (FSH), and total testosterone were conducted in the General Clinical Research Center using commercially available kits (Diagnostic Products, Los Angeles, CA). The coefficient of variation for assays in this study was 4.11% for E2, 2.55% for FSH, and 11.60% for total testosterone. Over 98% of the samples from women with identifiable menstrual cycles were collected between days 1 and 6 of the cycle, when these hormones are at low levels in the menstrual cycle. The mean cycle day of the blood sampling in this report was 3.5.
Other potential risk factors for the symptoms were selected because of their significance in previous studies of menopausal symptoms10–12 and the goals of this study: age, race, history of depression, current PMS self-reported at each assessment period, current smoking (yes, no), and perceived stress, a validated questionnaire of Cohen et at.13 The data were obtained from the structured interview questionnaire administered at each assessment period. The perceived stress scale was completed by the participants at each assessment period; a higher total score indicates greater stress. The variable for “severe PMS” (yes, no) was constructed from the subject reports at each assessment period using two interview questions: “Several days to 2 weeks before your period begins, do you experience irritability, mood swings or other emotional distress?” and “How much do these symptoms interfere with your daily activities?” The PMS group included only those participants who responded with the maximal interference rating of 4, indicating that the symptoms resulted in severe interference with usual activity. Participants with a current diagnosis of mood disorder (described above) were excluded.
Estimates of statistical power to detect a change in the prevalence of symptoms in the menopausal transition compared with premenopausal status were conducted. For 80% power with 0.05 alpha, assuming within-subject correlations of 0.23 to 0.30, power is sufficient to detect a decrease of 13% or more in the prevalence of the symptom in postmenopausal stage compared with premenopausal status for symptoms with a premenopausal prevalence of 18% or higher. For symptoms with a premenopausal prevalence of 34% or higher (eg, headache), power is sufficient to detect a decrease in prevalence of 10% between postmenopause and premenopausal status.
The statistical analyses presented in Tables 2 and 3 used generalized linear regression models for repeated measures. All available data for each participant were included in the repeated measures models; premenopausal observations were used as the reference group. We first estimated the unadjusted association of each study variable with the severity of each symptom (moderate or severe compared with none or mild).14 The outcome of symptom severity was dichotomized to examine symptom levels that are likely to be reported in a clinical setting. The same generalized linear regression models for repeated measures were used to estimate the association of menopausal stage with the prevalence of moderate or severe symptoms. These extensions of linear regression address among-women associations, while taking into account the longitudinal nature of the data. All variables except race (which had no significant associations in unadjusted analysis) were then examined simultaneously in multivariable models to estimate their independent contributions to symptom severity after adjusting for all other variables in the model. The variance estimates for the Wald statistics of the true regression coefficients were adjusted for the repeated observations from each participant using generalized estimates equations.15 The same set of covariates was evaluated for each symptom. All covariates were retained in the final models.
The hormone measures were transformed to the natural log values to minimize the effect of their skewed distribution. Each hormone was evaluated separately in the multivariable models because of the biologic correlation of the hormones. For each patient, the average (mean) of the two measurements from each assessment period and the standard deviation about the mean level were calculated for each hormone. This approach was used to provide a measure of the woman’s hormone fluctuations apart from the measured levels. In cases where two hormone values were not obtained in an assessment period, the single value was used for the mean, and the standard deviation was set at missing for that period. Of the 2,578 observations for each hormone, 12% were single observations in an assessment period. In the tables, the associations between the hormones and symptoms are expressed as odds ratios (ORs) with 95% confidence intervals. For hormone levels, an OR represents the increased odds of reporting the symptom for a 1-unit increase in the log-transformed hormone value, or a 1-unit increase in the hormone standard deviation (SD). The odds ratio associated with the perceived stress score is reported for a 5-unit change in the score.
Comparisons of baseline variables between continuers and subsequent dropouts used two-sample t or Pearson χ2 tests of association as appropriate for the distributions of the data. All analyses were conducted using SAS 9 (SAS Institute, Cary, NC). Statistical tests were two-tailed. To adjust for multiple symptom outcomes, a Bonferroni adjustment made to the type 1, alpha error rate results in significance level of P≤.01.
Table 1 shows the study variables at baseline and at end point 9 years later. At the baseline of this study (Period 2), the mean age of the cohort was 42.2 (SD 3.5) years. Ninety-one percent were premenopausal or late premenopausal, 7% were in the early transition and 2% in the late transition stage. At endpoint, 5% were premenopausal or late premenopausal, 38% were in the early transition, 22% in the late transition and 35% were postmenopausal.
Table 2 shows the percentage of women reporting each study symptom as moderate or severe at each menopausal stage. Reports of headache were significantly associated with menopausal stage and decreased in the late transition and postmenopausal stages (P=.002). Thirty-four percent of the premenopausal women but only 24% of postmenopausal women reported headache.
Table 3 shows the univariable associations of each study variable with each symptom. As above, headache was significantly associated with menopausal stage. Headache was 40% less likely postmenopause compared with premenopausal status (P=.003). There were consistent decreases in the likelihood (odds ratios) of irritability, mood swings, and anxiety, but the associations with menopausal stage were not statistically significant.
The associations of the repeated point-in-time average measures of the hormones with symptoms were consistent with the association of menopausal stage with symptoms. Mean FSH levels were inversely associated with mood swings (P=.005) and approached significance for irritability (P=.017). Mean testosterone levels were inversely associated with irritability (P=.008). The odds of reporting irritability or mood swings decreased with each unit increase in these hormones, indicating that the symptoms decreased as hormone levels increased around menopause.
We estimated the independent associations of menopausal stage, reproductive hormones, and other risk factors of the study, age, race, history of depression, PMS, perceived stress, and current smoking, with the severity of each symptom in multivariable models. Each variable was adjusted for all other variables in the model. Estimates were obtained for each symptom using the same set of covariates (Table 3). In the multivariable models, none of the symptoms were significantly associated with menopausal stage or hormone levels at the conservative level of P<.01 after adjusting for all other variables in the model.
The strongest covariate with these symptoms was self-reported PMS, which was associated with all symptoms after adjusting for all other variables in the model (P<.001). This is depicted graphically in Figure 1, which shows that women who met the study definition for PMS were significantly more likely to endorse the symptoms at each menopausal stage compared with women with no PMS. There was no interaction between PMS and menopausal stage, indicating that the incremental differences in symptoms between the PMS and no PMS groups were consistent across the menopausal transition. Women who reported PMS during the study were about four times more likely to report mood swings and irritability, over three times more likely to report concentration difficulties, and nearly twice as likely to report anxiety and headache compared with the no PMS group.
Perceived stress also had a strong independent contribution to the severity of each symptom in the adjusted models: irritability (P<.001), mood swings P<.001), anxiety (P<.001), concentration difficulties (P<.001), and headache (P=.018). A history of depression was strongly associated with anxiety, mood swings, and concentration difficulties in univariable analysis (P<.001) and remained associated with anxiety after adjusting for all other variables in the model (P=.002). Women with a history of depression were nearly twice as likely to report anxiety compared with those with no history of depression.
Age had no significant association with any symptom in multivariable analysis. None of the symptoms differed between the African-American and white women in the cohort in either unadjusted or multivariable analysis. Current smoking was a significant correlate of mood swings in unadjusted analysis (P<.001) but had no significant associations in the multivariable models.
This study examined common menopause-related symptoms to determine whether they were associated with menopausal stage as defined by menstrual bleeding patterns and reproductive hormone levels. Headache was associated with menopausal stage and significantly decreased after menopause. The likelihood of irritability, anxiety, and mood swings diminished in the menopausal transition, but the association with menopausal stage was not significant.
There is considerable evidence that headache, particularly migraine, is linked with female hormones, with up to 69% of women with migraine reporting menstrually related migraine.16 Other studies have shown that the prevalence of migraine increases before menopause and decreases after natural menopause,17 and that women with menstrual migraine experience complete relief with menopause.16,18 Our findings add support to evidence that the prevalence of migraine headaches decreases around menopause.
The few associations of these symptoms with the mean point-in-time levels of reproductive hormones were consistent with the association between symptoms and menopausal stage. The associations with hormones were inverse, which indicates that the symptoms were decreasing as hormone levels increased in the menopausal transition. A possible explanation for these associations is that irritability and mood swings are considered cardinal symptoms of PMS,19 which diminishes with the decline of ovarian activity.20 After adjusting for PMS in the multivariable models, the association of these symptoms with hormones was no longer significant.
Women who reported they had PMS were significantly more likely to report the study symptoms. We previously found that PMS decreased in the transition to menopause, but the women who reported PMS at baseline were at greater risk of menopausal hot flushes, depressed mood, poor sleep, and decreased libido.21 Similar findings were reported by Stewart and Boydell22 and the Melbourne Women’s Health Study, which concluded that women with histories of PMS were more likely to report mood symptoms in the menopausal transition.23,24 The present findings support these observations. The women reporting PMS were approximately four times more likely to report irritability and mood swings, three times more likely to report concentration difficulties, and nearly twice as likely to report headache and anxiety. These associations between symptoms and self-reported PMS remained robust after adjusting for other risk factors, including perceived stress and history of depression. Furthermore, there was no interaction between women with and those without PMS, indicating that the difference in symptom severity between the two groups remained consistent across the menopausal transition.
Our measure of menopausal stage is based on bleeding patterns that are observable and clinically relevant. Menopausal stage is distinct but complementary to reproductive hormone measures, as we have previously reported.25,26 Although the results of these two measures were not identical in this study, they were consistent, with both measures indicating that the symptoms were more severe at the outset than at the end of the menopausal transition. The findings suggest that these symptoms, which are commonly linked with menopause, diminish rather than worsen with the physiologic changes of the menopausal transition.
The strong independent contributions of psychosocial factors to the symptoms underscores their multifactorial context. Perceived stress was strongly correlated with all symptoms in the study. For example, women with higher stress scores were over 40% more likely to report irritability, mood swings, and anxiety compared with those with lower stress scores. A history of depression was a strong correlate of mood swings, anxiety, and concentration difficulties in unadjusted analysis. After adjusting for all other variables, women with a history of depression were nearly twice as likely to report anxiety. These data provide further evidence that women who previously experienced mood disorder are at a greater risk of mood and behavioral symptoms around menopause.5,26
This study did not address treatment. However, it is possible that hormone therapies stabilize the changing hormonal milieu and potentially alleviate these symptoms. When vasomotor symptoms are present in conjunction with these symptoms, hormone therapy remains the most effective treatment for hot flushes, although the extent to which other concomitant symptoms improve with hormone therapy is unclear.27 Recent studies suggest that estrogen therapy, particularly with transdermal delivery, is a therapeutic option for mood symptoms in perimenopausal women, but further studies to determine the safety are needed.28 The findings of strong associations of the symptoms with PMS suggests that serotonergic antidepressants, which have consistently demonstrated efficacy and safety for PMS or premenstrual dysphoric disorder,29 may also be effective for women in the menopausal transition. New oral contraceptive formulations that have demonstrated efficacy for PMS also may be beneficial, particularly for women who want contraception.30,31 It is also possible that symptoms such as anxiety, mood changes, and concentration difficulties reflect depression in some women. Extensive data support the efficacy and safety of serotonergic antidepressants for treating depressed mood and anxiety.
Other study limitations should be considered. The outcomes presented here are based on participants’ self-reports of the occurrence and severity of discrete symptoms as they occurred within the previous month at each of 10 assessment periods. Such self-reports are analogous to symptoms or problems that are presented in the clinical office visit, although our studies also show strong correlations between the self-reported symptoms and in-depth assessment and diagnosis of depression, poor sleep, and anxiety.5,32 The risk factors were selected and based on evidence in previous studies and the design of this cohort, but other risk factors for menopausal symptoms may well exist. The hormone measures were obtained in the early follicular phase of the menstrual cycle and are not comparable to measurements obtained at other times in the cycle. Variation in sampling during the menstrual cycle is a potential source of bias, although more than 98% of the samples from cycling women were collected within the first 6 days of the cycle, when differences are lowest. Had bias occurred due to variation in cycle day, it would be nondifferential with respect to the outcome and bias the results toward no finding. Premenstrual syndrome was determined by self-report and was not a diagnosis confirmed by daily symptom diaries. However, it should be noted that the prevalence of PMS was a low 8%. This may be due in part to excluding all participants who met diagnostic criteria for a mood disorder evaluated at each assessment period and including only severe interference with functioning in the construction of the PMS variable. A Bonferroni adjustment was made to adjust for multiple symptom outcomes, but it is noted that this standard method of adjustment may be too conservative for outcomes that are positively correlated.33 Our findings are from a population-based cohort of urban, generally healthy, African-American and white women and may not be generalizable to other geographic or racial groups.
The strengths of this study are its longitudinal examination of the symptoms for a 9-year period in the transition to menopause. The power of the study was sufficient to detect a decrease of 10–13% in the prevalence of the symptom postmenopause compared with premenopausal status. Menopausal stages were based on careful assessment of bleeding patterns at each assessment period. The hormone measures were obtained within the first 6 days of two consecutive menstrual cycles in each assessment period and were concurrent with the symptom measures. All women were premenopausal at enrollment in the cohort, which allows us to capture the early stages of the transition to menopause and assess the independent contributions of both age and menopausal stage to symptoms. The sampling was stratified to the obtain equal numbers of African-American and white women for the evaluation of racial associations. Strong follow-up and participation rates over the 9-year interval provide a wealth of longitudinally measured data to estimate the associations with changing hormones and bleeding patterns in the menopausal transition.
In summary, these data show that headache significantly decreased in the menopausal transition. Irritability and mood swings also decreased in the menopausal transition as reproductive hormone levels increased. The findings indicate that these symptoms that are commonly linked with menopause diminish rather than increase with the physiologic changes of the menopausal transition. Further studies of the extent of common etiologic pathways of symptoms in the menopausal transition are needed to increase understanding of their associations with physiologic and psychosocial factors and to identify improved treatments.
1. National Institutes of Health. National Institutes of Health State-of-the-Science Conference statement: management of menopause-related symptoms. Ann Intern Med 2005;142:1003–13.
2. Freeman EW, Sammel MD, Lin H, Gracia CR, Pien GW, Nelson DB, et al. Symptoms associated with menopausal transition and reproductive hormones in midlife women. Obstet Gynecol 2007;110:230–40.
3. Freeman EW, Sammel MD, Liu L, Martin P. Psychometric properties of a menopausal symptom list. Menopause 2003;10:258–65.
4. Freeman EW, Sammel MD, Gracia CR, Kapoor S, Lin H, Liu L, et al. Follicular phase hormone levels and menstrual bleeding status in the approach to menopause. Fertil Steril 2005;83:383–92.
5. Freeman EW, Sammel MD, Liu L, Gracia CR, Nelson DB, Hollander L. Hormones and menopausal status as predictors of depression in women in transition to menopause. Arch Gen Psychiatry 2004;6:62–70.
6. Spitzer RL, Williams JB, Kroenke K, Linzer M, deGruy FV 3rd, Hahn SR, et al. Utility of a new procedure for diagnosing mental disorders in primary care. The PRIME-MD 1000 study. JAMA 1994;272:1749–56.
7. Spitzer RL, Kroenke K, Williams JB. Validation and utility of a self-report version of PRIME-MD: the PHQ primary care study. Primary Care Evaluation of Mental Disorders. Patient Health Questionnaire. JAMA 1999;282:1737–44.
8. Nelson DB, Sammel MD, Freeman EW, Liu L, Langan E, Gracia CR. Predicting participation in prospective studies of ovarian aging. Menopause 2004;11:543–8.
9. Soules MR, Sherman S, Parrott E, Rebar R, Santoro N, Utian W, et al. Executive summary: Stages of Reproductive Aging Workshop (STRAW). Fertil Steril 2001;76:874–8.
10. Harlow BL, Wise LA, Otto MW, Soares CN, Cohen LS. Depression and its influence on reproductive endocrine and menstrual cycle markers associated with perimenopause: the Harvard Study of Moods and Cycles. Arch Gen Psychiatry 2003;60:29–36.
11. Bromberger JT, Assmann SF, Avis NE, Schocken M, Kravitz HM, Cordal A. Persistent mood symptoms in a multiethnic community cohort of pre- and perimenopausal women. Am J Epidemiol 2003;158:347–56.
12. Santoro N, Lasley B, McConnell D, Allsworth J, Crawford S, Gold EB, et al. Body size and ethnicity are associated with menstrual cycle alterations in women in the early menopausal transition: The Study of Women’s Health across the Nation (SWAN) Daily Hormone Study. J Clin Endocrinol Metab 2004;89:2622–31.
13. Cohen S, Kamarck T, Mermelstein R. A global measure of perceived stress. J Health Soc Behav 1983;24:385–96.
14. McCullagh P, Nelder JA. Generalized linear models. 2nd ed. New York (NY):Chapman & Hall; 1989.
15. Liang KY, Zeger SL. Longitudinal data analysis using generalized linear models. Biometrika 1986;73:13–22.
16. Facchinetti F, La Marca A. Headaches, migraines and premenstrual syndrome. Gynaecol Forum 2006;11:10–13.
17. Wang SJ, Fuh JL, Lu SR, Juang KD, Wang PH. Migraine prevalence during menopausal transition. Headache 2003;43:470–8.
18. Nappi RE, Cagnacci A, Granella F, Piccinini F, Polatti F, Facchinetti F. Course of primary headaches during hormone replacement therapy. Maturitas 2001;38:157–63.
19. Eriksson E, Andersch B, Ho HP, Landen M, Sundblad C. Diagnosis and treatment of premenstrual dysphoria. J Clin Psychiatry 2002;63 suppl:16–23.
20. Schmidt PJ, Nieman LK, Danaceau MA, Adams LF, Rubinow DR. Differential behavioral effects of gonadal steroids in women with and in those without premenstrual syndrome. N Engl J Med 1998;338:209–16.
21. Freeman EW, Sammel MD, Rinaudo PJ, Sheng L. Premenstrual syndrome as a predictor of menopausal symptoms. Obstet Gynecol 2004;103:960–6.
22. Stewart DE, Boydell KM. Psychologic distress during menopause: associations across the reproductive cycle. Int J Psychiatry Med 1993;23:157–62.
23. Guthrie JR, Dennerstein L, Hopper JL, Burger HG. Hot flushes, menstrual status, and hormone levels in a population-based sample of midlife women. Obstet Gynecol 1996;88:437–42.
24. Dennerstein L, Lehert P, Burger H, Dudley E. Mood and the menopausal transition. J Nerv Ment Dis 1999;187:685–91.
25. Gracia CR, Sammel MD, Freeman EW, Lin H, Langan E, Kapoor S, et al. Defining menopause status: creation of a new definition to identify the early changes of the menopausal transition. Menopause 2005;12:128–35.
26. Freeman EW, Sammel MD, Lin H, Nelson DB. Associations of hormones and menopausal status with depressed mood in women with no history of depression. Arch Gen Psychiatry 2006;63:375–82.
27. North American Menopause Society. Estrogen and progestogen use in peri- and postmenopausal women: March 2007 position statement of The North American Menopause Society. Menopause 2007;14:168–82.
28. Cohen LS, Soares CN, Joffe H. Diagnosis and management of mood disorders during the menopausal transition. Am J Med 2005;118 suppl:93–7.
29. Steiner M, Pearlstein T, Cohen LS, Endicott J, Kornstein SG, Roberts C, et al. Expert guidelines for the treatment of severe PMS, PMDD, and comorbidities: the role of SSRIs. J Womens Health (Larchmt) 2006;15:57–69.
30. Rapkin AJ. New treatment approaches for premenstrual disorders. Am J Manag Care 2005;11:S480–91.
31. Yonkers KA, Brown C, Pearlstein TB, Foegh M, Sampson-Landers C, Rapkin A. Efficacy of a new low-dose oral contraceptive with drospirenone in premenstrual dysphoric disorder. Obstet Gynecol 2005;106:492–501.
32. Freeman EW, Sammel MD, Gracia CR, Kapoor S, Ferdousi T. The role of anxiety and hormonal changes in menopausal hot flashes. Menopause 2005;12:258–66.
33. Savitz DA, Olshan AF. Multiple comparisons and related issues in the interpretation of epidemiologic data. Am J Epidemiol 1995;142:904–8.