{"paper_id":"49ef061b-98de-4b15-816b-3917718d3f2d","body_text":"R E S E A R C H Open Access\nSymptom severity of bipolar disorder\nduring the menopausal transition\nWendy K. Marsh 1*, Bernice Gershenson 2 and Anthony J. Rothschild 3\nAbstract\nBackground: Little is known about the mood symptom experience of women with bipolar disorder during the\nmenopausal transition (MT). Yet times of rapid hormonal decline, such as the postpartum, are associated with\nincreased risk of severe mood episodes in bipolar disorder, and the MT is a time of increased risk for unipolar\ndepression in women with or without a history of depression.\nMethods: Enrollment included 56 women 40 –60 years old diagnosed in the bipolar spectrum who were\nexperiencing menopausal symptoms or were up to 5 years since their final menstrual period. Menopausal\nstages included early menopause, late menopause, or early postmenopause based on standardized criteria.\nObservational, prospective standardized mood symptom and reproductive hormone assessments were\ncompleted periodically. Concurrent menopausal symptoms as well as history of mood exacerbation during\npast reproductive events were assessed.\nResults: Forty-four women were included in the main analysis. The average Montgomery-Asberg Depression\nR a t i n gS c a l e( M A D R S )s c o r ew a s4 . 4 3p o i n t sh i g h e ri nt h el a t et r a n s i t i o n / e a r l yp o s t m e n o p a u s a ls t a g ew o m e n\n(n = 29) compared to the early menopausal stage women ( n = 15) (±SE 2.14; p = 0.039), corresponding to a\nroughly 10 % higher score (range 0 –40) in the late/post stage across all study visits. Results were similar for\nthe Young Mania Rating Scale (YMRS), where the average score was 2.54 points higher in the late/early\npostmenopausal stage women compared to the early menopausal stage women (±SE 1.15; p =0 . 0 2 7 ) , a l s o\nroughly 10 % higher (range 0 –26). Estradiol and follicle-stimulating hormone (FSH) absolute levels as well as\nbetween-visit change in levels were not notably associated with YMRS or MADRS during study observation.\nTotal Greene Climacteric Symptom (menopausal symptom ) score was significantly associated with MADRS but\nnot YMRS. History of mood exacerbation premenstrually and/or postpartum was not significantly associated\nwith YMRS or MADRS severity during the MT.\nConclusions: These results support the theory that times of in creased reproductive hormonal changes, such\nas the late MT and early postmenopause, here compared to early MT, are associated with greater mood symptom\nseverity in bipolar spectrum women. Nonetheless, absolute or change in FSH and estradiol levels were not significantly\nassociated with depression or mood elevation severity.\nKeywords: Bipolar disorder; Menopause; Neuroendocrinology; Estrogen\n* Correspondence: Wendy.Marsh@umassmemorial.org\n1Department of Psychiatry, School of Medicine, University of Massachusetts,\n55 Lake Ave North, S3-314, Worcester, MA 01655, USA\nFull list of author information is available at the end of the article\n© 2015 Marsh et al. Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0\nInternational License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and\nreproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to\nthe Creative Commons license, and indicate if changes were made.\nMarsh et al. International Journal of Bipolar Disorders  (2015) 3:17 \nDOI 10.1186/s40345-015-0035-z\n\nBackground\nDue to the scarcity of studies, little is known about\nthe course of bipolar disorder through the perimeno-\npause, or menopausal transition (MT), and yet every\nwoman with bipolar disorder expects to or has experi-\nenced menopause. Compelling evidence on the increased\nrisk of mood episodes in women with bipolar disorder\n(BD) during other times of reproductive hormonal transi-\ntion, like postpartum (Kendell et al. 1987; Sit et al. 2006),\nindicates the importance of examining mood during the\nmenopausal transition. Given the substantial societal bur-\nden of bipolar disorder (Kleine-Budde et al. 2013), detect-\ning times of risk is critical.\nPsychological symptoms in women without a mood\ndisorder are frequent during the MT and most often\ninclude irritability, tearfulness, anxiety, depression, emo-\ntional lability, low energy and motivation, poor concen-\ntration, and interrupted sleep (Avis et al. 1994). While\nthe role of reproductive hormones in the risk of mood\nsymptoms is unclear, psychological symptoms are postu-\nlated to be linked to the fluctuation, not absolute levels,\nof estradiol (Rubinow et al. 1998). These psychological\nsymptoms overlap with those of functionally impairing\nmood disorders leading one to speculate that the peri-\nmenopause may contribute to depression or mood eleva-\ntion in vulnerable women.\nUnipolar depression risk in the MT has an expanding\nliterature reporting an increase of both major depression\nand depressive symptoms during the perimenopause\n(Maartens et al. 2002; Freeman et al. 2004a; Schmidt\net al. 2004; Bromberger 2006; Cohen et al. 2006; Steinberg\net al. 2008). Compared with premenopausal women,\nwomen with a history of depression are nearly five times\nmore likely to have a diagnosis of major depression in the\nmenopausal transition, whereas women with no his-\ntory of depression are two to four times more likely\nto report depressed mood (Freeman 2010). Within\nthe Stages of Reproductive Aging Workshop + 10\n(Harlow et al. 2012) defined reproductive stages, the\nlate perimenopause (interval of amenorrhea of 60 to\n364 days) (Dennerstein et al. 1993; Schmidt et al.\n2004; Freeman et al. 2006; Bromberger et al. 2007;\nWoods et al. 2008; Freeman 2010; Bromberger et al. 2011)\nand early postmenopause (amenorrhea 1 –6 years)\n(Dennerstein et al. 2004; Bromberger et al. 2007;\nBromberger et al. 2011) are times of heightened risk\nof unipolar depression. Women reporting a history of\npremenstrual or postpartum mood disturbance are at\ngreater risk of depression during the MT than those who\ndo not (Freeman et al. 2004b; Freeman 2010).\nLess is known in bipolar disorder. Postmenopausal\nwomen recalling the MT report severe mood distur-\nbances or worsening depression during the transition\n(Blehar et al. 1998; Freeman et al. 2002). Mood\nelevation is reported less frequently than depression\nin the MT (Blehar et al. 1998; Kennedy et al. 2005).\nOur work reviewing longitudinal standardized clinical\nassessments in menopausal age women with bipolar\ndisorder found a high degree of depression, 68 % over\na 17-month average duration, that was significantly\ngreater than what subjects reported experiencing during\ntheir reproductive years (Marsh et al. 2008). Compared to\na concurrent pooled comparison group of like-aged men\nand reproductive age women with bipolar disorder,\nwomen of menopausal age with bipolar disorder had a sig-\nnificantly greater proportion of clinic visits in the de-\npressed state and significantly lower proportion in the\neuthymic state, with no difference in proportion of visit in\nthe elevated/mixed state (Marsh et al. 2009). When exam-\nining mood by menopausal stage in a large multisite data-\nbase of patients treated for bipolar disorder, the small\nnumber of women transitioning from perimenopause to\npostmenopause had significantly greater depression than\nother female reproductive groups while overall euthymia\nand mood elevation decreased with progressing female re-\nproductive stage (Marsh et al. 2012). Nonetheless, there is\na lack of prospective data on mood experience during this\ncritical time in bipolar disorder.\nThis study prospectively examines mood across the\nMT stages and in association with reproductive hor-\nmones in women with bipolar disorder. We hypothe-\nsized that depression scores would be higher during the\nlate menopausal transition and at times of greater hor-\nmonal changes.\nMethods\nStudy participants\nStudy enrollment occurred from January 2010 to July\n2012, and subjects were recruited from the commu-\nnity as well as from an academic psychiatry clinic.\nEligibility included (a) a diagnosis of bipolar disorder\n(I, II, NOS) based on DSM IV criteria, (b) age 40 –60\nyears with intact uterus and ovaries, (c) a menstrual\nperiod within the last 5 years, if menstruating regu-\nlarly (every 21 –35 days) subject must be experiencing\nmenopausal symptoms, and (d) having a current psy-\nchiatric provider. Menopausal symptoms were defined\nnarrowly as vasomotor symptoms (hot flashes and\nnight sweats) so as not to overlap with mood symp-\ntoms (for example, sleep disturbance and irritability).\nAs study personnel did not assume care of subjects,\nfor subject safety, a psychiatric provider was required.\nWomen using vaginal hormone therapy ( n = 1) or intra-\nuterine device with hormonal progesterone ( n =1 ) w e r e\naccepted; those on oral contraceptives were not. Partici-\npants provided signed, written informed consent prior to\nstudy entry.\nMarsh et al. International Journal of Bipolar Disorders  (2015) 3:17 Page 2 of 9\n\nProcedures\nThis study conducted an observational longitudinal\nevaluation of mood for 4 months during each of the\nfollowing menopausal stage s: late reproductive, early\nmenopausal transition, late menopausal transition,\nand early postmenopause (defined in the “Menopausal\nstatus ” section below) in women with bipolar dis-\norder. The initial interview was scheduled in the\nearly follicular phase on days 2 –6o ft h em e n s t r u a l\ncycle (day 1 being the first day of menstruation) in\nwomen who were menstruating more frequently than\nevery 60 days for standardized hormonal assessments.\nInitial visit included a standardized interview with the\nAffective Disorder Evaluation (ADE), a modified ver-\nsion of the mood and psychosis modules from the Struc-\ntured Clinical Interview for DSM Disorders (SCID)\nintended for routine use by practicing clinicians, which\nwas performed by an ADE trained psychiatrist. The ADE\nprovided diagnoses of DSM IV bipolar I disorder, bipolar\nII disorder, bipolar disorder not otherwise specified,\npsychiatric and medical comorbidities, as well as mood\ndisorder history (Sachs et al. 2003). The ADE includes\nstandardized questions for subject report of the percent of\ntime in the prior 12 months spent experiencing mood ele-\nvation, depression, or anhedonia symptoms. It also asks\nfor patient self-reported endorsement of a history of\npremenstrual or postpartum mood exacerbation. As-\nsessments of current mood state severity with Young\nMania Rating Scale (YMRS) (Young et al. 1978) and\nMontgomery-Asberg Depression Rating Scale (MADRS)\n(Montgomery and Asberg 1979) were performed by\ntwo trained raters who periodically compared score\nresults for inter-rater reliability. Subjects completed\nself-administered questionnaires detailing menstrual\nand medical history and the Greene Climacteric Scale\n(Greene 1976) to assess menopausal symptoms. A blood\nsample was obtained in the early follicular phase of the\nmenstrual cycle to assess follicle-stimulating hormone\n(FSH) and estradiol levels. In women who had not men-\nstruated for 60 days or longer, a random hormonal sample\nwas assessed.\nFollow-up visits included standardized YMRS and\nMADRS mood assessments. If a woman was men-\nstruating with predictabil ity, visits would be sched-\nuled up to 6 weeks apart in the early follicular phase\n(days 2 –6) of her menstrual cycle and hormone\nblood draws performed. If a woman had gone 60 days\nwithout menstruation at intake visit, then mood\nassessments were scheduled once a month and hormone\nblood draws were scheduled every other month re-\ngardless of where a woman was in her menstrual\ncycle. The study was approved by the University of\nMassachusetts Medical School Institutional Review\nBoard.\nSerum hormone measures\nEstradiol and follicle-stimulating hormone were analyzed\nby institutional standard laboratory procedures. Estradiol\nlevels were assessed by Access Estradiol, a competitive\nbinding immune-enzymatic assay. FSH was assayed by\nAccess hFSH which is a sequential two-step immune-\nenzymatic assay.\nMenopausal status\nMenopausal staging was determined using the execu-\ntive summary of Staging of Reproductive Aging in\nWomen (STRAW) designed as a comprehensive basis\nfor assessing reproductive aging in research settings\n(Harlow et al. 2012). Subjects were categorized based on\nSTRAW + 10 stages: (1) second half of the late reproduct-\nive (subtle changes in menstrual cycle characteristics and\nvasomotor symptoms), (2) early perimenopasue (increased\nvariability in menstrual cycle length defined as a persistent\ndifference of 7 or more days in the length of consecutive\ncycles), (3) late perimenopause (no menstrual cycle in\n60 days but menstrual bleeding in the last 12 months),\nand (4) early postmenopause (no menstrual cycle in the\nlast 1 –5 years). Given the small sample size, the repro-\nductive groups were collapsed.\nAs hormonal levels and menstrual cyclicity remain\nrelatively predictable in the earlier stages of the MT\n(Harlow et al. 2012), the late reproductive and early\nMT women were collapsed into one group, called early\ntransition. Hormonal levels show greater variability and\nunpredictability during the later menopausal transition\nstages, including the early postmenopause (Burger et al.\n1999; Burger et al. 2007; Sowers et al. 2008a; Sowers et al.\n2008b; Harlow et al. 2012). In accordance, the late\nperimenopausal and early postmenopausal women were\ncollapsed into one group, called the late and early-post\ntransition group.\nStatistical analysis\nThe two groups analyzed included (a) the early transi-\ntion group, consisting of women in late reproductive\n(n = 4) or early perimenopause ( n =1 7 ) p h a s e , a n d ( b )\nthe late and early-post transition, women in their late\nperimenopause ( n = 20) or early postmenopause ( n =1 5 )\nphase.\nChi-square ( χ2) tests were used to compare categor-\nical variables, and two sample t tests with equal vari-\nance were utilized to test differences in continuous\nvariables by the two groups; Pearson correlation coef-\nficients measured the linear association between two vari-\nables (Sedgwick 2012). For the main outcome measure,\nwe analyzed longitudinal changes across visits for MADRS\nand YMRS sums using generalized estimating equation\n(GEE) methods to control for the correlation within the\nsubjects at the five study time points. Models were run\nMarsh et al. International Journal of Bipolar Disorders  (2015) 3:17 Page 3 of 9\n\nexamining potential associations between MADRS and\nYMRS and menopausal stage categories (early vs. late/post),\nadjusted for study visit (where visit 1 was used as the\nreferent category). We used an autoregressive covariance\nstructure in all of our GEE models. p values are reported\nfrom a z test that a single regression coefficient was equal\nto 0. We also tested the significance of difference among\nthe time effects as a multiple degrees of freedom chi-\nsquare test. SAS software version 9.3 (SAS Institute, Cary,\nNC) was used for the GEE modeling while SAS version\n9.2 was used for the remaining analyses.\nResults\nOf 255 women screened, 99 were eligible and 57 enrolled.\nOne woman was dropped due to active illicit substance\nabuse influencing mood at the time of enrollment. Of the\n56 women who entered the study, 21 % ( n = 12) com-\npleted only the first visit, 32 % ( n = 18) completed two\nvisits, and 66 % ( n = 37) completed all third, fourth, and\nfinal visits. The dropout rate in the early transition group\nwas 43 % and the late and early-post transition group\n31 %. There were no significant differences in demo-\ngraphic characteristics, bipolar disorder characteristics, or\ntreatment approach between the early transition and the\nlate and early-post transition groups with the exception of\nage-dependent variables (age and duration of illness) being\ngreater in the older aged late and early-post transition\ngroup. Current alcohol use not meeting the criteria for de-\npendence or abuse was more common in the early transi-\ntion group. There was no significant difference between\nreproductive groups in other common comorbidities that\nhave been shown to affect mood course in BD including\nrapid cycling status, age at onset (Schurhoff et al. 2000),\ncurrent comorbid anxiety (Simon et al. 2004), and sub-\nstance use (Goldstein et al. 2006) disorders. The sample\ncharacteristics are presented in T able 1.\nMenopausal stage and mood\nOne woman who entered the study in late menopause\ntransitioned to early postmenopause during the study\nobservation. One woman who entered the study in early\nmenopause proceeded to late menopausal transition dur-\ning the study observation; she was analyzed in the early\ntransition group as the diagnosis of change of reproduct-\nive stage is made after the occurrence, as it would in\nclinical practice.\nA total of 44 women were included in this analysis, as\nthey had information from at least two visits in order to\nadjust for the autocorrelation within women. Of these\nwomen, 15 were in the early transition group and 29\nwere in the late and early-post transition (5 of which\nwere over the age of 55).\nResults from the MADRS model showed that average\nMADRS score was 4.43 points higher in the late and\nearly-post transition stage women compared to the early\ntransition women (±SE 2.14; p = 0.039), corresponding\nto a roughly 10 % increase in score (score range 0 –40)\nin the late/post stage across all study visits. The effect of\ntime on the MADRS score was not significant ( χ2 = 9.41;\ndf =4 ; p = 0.052). The mean MADRS score for the early\ntransition women was 12.10 (±SE 1.63), 95 % CI 8.89,\n15.30, and for the late and early-post transition women,\nthe mean was 16.52 (±SE 1.67) and 95 % CI 13.25, 19.80.\nResults were similar for the YMRS model, where aver-\nage score was 2.54 points higher in the late and early-\npost transition women compared to the early transition\nwomen (±SE 1.15; p = 0.027), also a roughly 10 % in-\ncrease (score range 0–26). The effect of time on the YMRS\nscore was not significant ( χ2 =8 . 6 8 ;df =4 ; p =0 . 0 7 ) . T h e\nmean YMRS score for early transition women was 6.06\n(±SE 1.02), 95 % CI 4.07, 8.06, and for late and early-post\ntransition women, the mean was 8.60 (±SE 1.02), 95 % CI\n6.60, 10.60.\nIn contrast, retrospective report of percent of past year\nin depressed/anhedonic or elevated mood state was not\nassociated with menopausal stage (Table 2).\nTable 1 Subject characteristics\nEarly MT\n(n = 21)\nLate MT/early\npostmenopause\n(n = 35)\np value Total\n(n = 56)\nAge, mean ± SD 45.0 ± 3.4 50.6 ± 4.7 <0.001 48.5 ± 5.0\nAge at onset of bipolar\ndisorder, mean ± SD\n14.9 ± 7.1 15.5 ± 8.6 0.82 15.3 ± 8.0\nBipolar diagnosis, N (%) 0.45\nI 11 (53) 17 (48) 28 (50)\nII 7 (33) 16 (45) 23 (41)\nNOS 2 (14) 2 (6) 5 (9)\nDuration of illness,\nmean ± SD\n30.0 ± 6.7 35.1 ± 8.6 0.02 33.2 ± 8.3\nRapid cycling, N (%) 13 (61.9) 24 (75) 0.31 37 (69)\nEthnicity, Caucasian,\nnon-Hispanic, N (%)\n17 (80) 33 (97) 0.18 50 (91)\nMarital status, married/\npartnered, N (%)\n10 (48) 19 (56) 0.55 29 (53)\nEmployment status,\nfull or part time, N (%)\n8 (38) 14 (41) 0.82 22 (40)\nBMI, mean ± SD 28.4 ± 5.7 31.3 ± 7.3 0.14 30.1 ± 6.8\nComorbid anxiety,\nN (%)\n19 (90) 33 (94) 0.60 52 (93)\nComorbid alcohol use,\nN (%)\n8 (38) 2 (6) <0.01 10 (18)\nComorbid substance\nuse, N (%)\n3 (14) 8 (23) 0.43 11 (20)\nNumber of mood\nstabilizers (Li, AED,\nand antipsychotics),\nmean ± SD\n1.28 ± 0.8 1.25 ± 0.7 0.88 1.27 ± 0.7\nMarsh et al. International Journal of Bipolar Disorders  (2015) 3:17 Page 4 of 9\n\nNeuroendocrine results\nMean FSH increased and mean estradiol decreased as\nwould be expected across the progression from early to\nlater menopausal transition to postmenopause (T able 3).\nEstradiol levels varied by 264 pg/ml (19 –283 pg/ml) for\nFSH levels less than 40 IU/L; for FSH levels greater than\nor equal to 40 IU/L, estradiol levels remained low <44 IU/L,\nin accordance with the clinically accepted FSH cutoff\nof 40 IU/L.\nAt initial visit, there was a positive correlation between\nabsolute estradiol level and YMRS ( r(51) = 0.31, p = 0.02)\nacross reproductive groups. This significant association\nwas not born out when estradiol levels and YMRS scores\nwere assessed at follow-up visits. FSH levels were not\nassociated with YMRS score at intake or follow-up visits.\nNeither FSH nor estradiol absolute levels were associ-\nated with MADRS scores at any visit. Likewise, the\namount of change in FSH or estradiol levels from the\nprevious visit to the assessed visit was not associated\nwith MADRS of YMRS score at any follow-up visit\n(Table 4).\nMenopausal symptoms and mood\nMean total Greene Climacteric Symptom rating was\n23.4 (SD ± 10.7) in line with mean of 22.9 of normative\nresults of 40 –55-year-old women visiting a menopausal\nclinic. Greene score was significantly associated with\nconcurrent MADRS score, r(52) = 0.52, p < 0.001, but\nnot YMRS score at baseline visit. On sub-scale analysis,\nMADRS score was associated with Greene Psychological\nsub-scale, r(52) = 0.55, p < 0.001, but not with Greene\nVasomotor subscale (Table 5).\nHistory of reproductive events and mood disruption\nWomen who had given birth (n =3 6 )a n dr e p o r t e dn op o s t -\npartum mood exacerbation tended to have lower, although\nnot statistically significant (t = −1.91, p =0 . 0 7 ) ,m e a nY M R S\nacross visits (7.2 ± 7.0) than those who reported postpartum\nmood exacerbations (YMRS 11 .2 ± 5.5). Otherwise, re-\nported history of a postpartum and or perimenstrual mood\nexacerbation was not associated with mean severity of\nMADRS or YMRS in the MT and early postmenopausal\nyears (T able 6).\nDiscussion\nThe primary finding of this study is significantly higher\ndepression and mood elevation symptom scores in women\nwith bipolar disorder during the late menopausal transi-\ntion and early postmenopause than those of women in the\nlate reproductive years and ea rly menopausal transition.\nA difference of greater than two points on the MADRS\n(4.4 points in this report) has been called clinically\nmeaningful (Kennedy et al. 2006). A YMRS difference of\n2.5 points reported in this study, even in an outpatient\npopulation, while statistically significant, is less likely to be\nclinically significant (Lukasiewicz et al. 2013).\nThe increased level of depressive symptoms in the late\nand early-post transition group is consistent with previ-\nous chart reviews (Marsh et al. 2012) and retrospective\nrecall (Blehar et al. 1998; Freeman et al. 2002) of the MT\nin women with bipolar disorder. It is also in agreement\nwith the greater rates, compared to premenopause, of\ndepression during the late MT (Schmidt et al. 2004;\nFreeman 2010; Freeman et al. 2014) and often early\npostmenopause (Bromberger et al. 2010) found in the\nunipolar depression literature. The statistically signifi-\ncant higher YMRS mood elevation rating in the late and\nearly-post transition group than early transition group is\nin contrast to the decrease in DSM IV mood elevation\nsymptoms reported previously (Marsh et al. 2012) but\ndoes not contradict the retrospective report of worsen-\ning mood elevation or irritability retrospectively reported\nby many postmenopausal women with bipolar disorder\n(Blehar et al. 1998; Freeman et al. 2002). The concomi-\ntant worsening of both depression and mood elevation\nsymptoms in the later stages of the menopausal transi-\ntion may be indicative of worsening of mixed features,\nnow a DSM V specifier for bipolar disorder (APA 2013).\nGiven these early results, future studies are needed, par-\nticularly with a larger sample size to confirm, or refute,\nthe menopausal transition as a time of risk of depression,\nmood elevation or mixed symptoms in the course of bipo-\nlar disorder. Larger studies would further elucidate the\nrole of clinically defined menopausal stage by evaluating\nlate reproductive, early transition, late transition, early\npostmenopause, and late postmenopause. A longer dur-\nation of observation would ideally include observing mood\nTable 2 Bipolar depression and mood elevation reported in the\nprevious year by menopausal stage\nMenopausal stage\nEarly MT\n(n = 21)\nLate MT/early\npostmenopause\n(n = 35)\np value\n% of previous year reported\nin a depressed mood\n36.3 ± 32.1 39.5 ± 26.2 0.69\n% of previous year reported\nin an elevated mood\n10.3 ± 11.6 16.5 ± 17.5 0.16\nTable 3 Reproductive hormone levels by menopausal stage\nFSH (IU/L) Estradiol (pg/ml)\nN Mean ± SD p value Mean ± SD p value\nEarly MT 20 11.6 ± 11.4 <0.0001 74.4 ± 63.8 =0.09\nLate MT 19 53.1 ± 47.6 53.8 ± 69.3\nEarly\npostmenopause\n14 79.9 ± 33.2 27.9 ± 26.3\nMarsh et al. International Journal of Bipolar Disorders  (2015) 3:17 Page 5 of 9\n\nacross menopausal stages within individual subjects. Of\nparticular interest would be the role of the last menstrual\nperiod as a marker in the late transition and early post-\nmenopause (Freeman et al. 2014) and as an easy reference\npoint in the clinical discussion with a patient.\nNeuroendocrine\nThis is the first study to examine reproductive hormones\nin the MT in bipolar disorder. Despite a positive correl-\nation between absolute estradiol level and YMRS at the\ninitial visit, there were no further associations between\nabsolute or change between visits in FSH or estradiol\nlevels with either depression or mood elevation scores.\nWhile potentially a chance association of estradiol and\nmood elevation symptoms, this is the first report on es-\ntradiol levels and mood elevation. This result may indi-\ncate a subgroup of women who are vulnerable to mood\nelevation symptoms during times of high estradiol. No\nstandard testing exists for assessing the degree of vari-\nability in hormonal levels after cessation of predictable\nmenses, thus a potential future definition of variability,\nmay yet reveal associations with mood severity.\nNeurocognitive evidence suggests that estradiol may\nhave antidepressant-like properties in the brain (Kendall\net al. 1982; McEwen et al. 1997), and clinical trial results\nhave reported an antidepressant effect of estradiol ad-\nministration during the MT in unipolar depression\n(Soares et al. 2001; Rasgon et al. 2002). Nonetheless, re-\nsults in unipolar depression during the MT overall do\nnot relay an association of depressive symptoms with\nabsolute FSH or estradiol levels (Schmidt et al. 2004;\nBromberger et al. 2010) with the exception of increased\nFSH being positively associated with depressive symp-\ntoms in one study (Freeman et al. 2006). Reports on es-\ntradiol and FSH variability and risk of depression during\nthe MT have been both positive (Freeman et al. 2004a;\nFreeman et al. 2006; Freeman et al. 2014) and negative\n(Schmidt et al. 2004; Bromberger et al. 2010; Freeman\net al. 2014) and have employed differing ways of asses-\nsing variability.\nMore thorough hormonal assessments should include\ntestosterone which has been found associated with uni-\npolar depression during the MT (Bromberger et al.\n2010), inhibin B, an early indicator of ovarian aging,\nand potentially sex hormone-binding globulin as well as\ndehydroepiandrosterone which have both been assessed\nin association with unipolar depression during the\nmenopausal transition (Bromberger et al. 2010).\nGreene Climacteric Scale and vasomotor symptoms\nThis is the first evaluation of bipolar mood symptoms\nand menopausal symptoms. The Greene Climacteric\nScale overall score was positively associated with concur-\nrent depressive symptoms in women with bipolar dis-\norder during the MT. This was driven by the Greene\npsychological subscale association with MADRS depres-\nsion scores. MADRS depression scores in women with\nbipolar disorder were not associated with concurrent\nvasomotor symptoms reported on the Greene Scale.\nWhile the mechanism behind VMS is not well under-\nstood, an association between vasomotor symptoms and\nunipolar depressive symptoms has often (Cohen et al.\n2006; Freeman et al. 2009) but not always been reported\n(Freeman 2010) .\nThe Greene Scale overall and subscale scores were not\nassociated with mood elevation symptoms. Thus, despite\nthe supposition that hormonal changes during the tran-\nsition are behind VMS and the increased risk of unipolar\ndepression as well as the overlap in menopause and\nmood symptoms (example sleep changes), the associ-\nation of menopausal and mood symptoms did not bear\nout in bipolar disorder. As this is a cross-sectional\nreport, it may be useful to further examine the duration\nof VMS or menopausal symptoms and risk of mood\nsymptoms and assess for the potential of a longitudinal\nassociation.\nTable 4 Bipolar depression and mood elevation score correlation with between-visit change in hormone levels\nVisit N MADRS and change in FSH MADRS and change in estradiol YMRS and change in FSH YMRS and change in estradiol\n1–2 24 r (22) = 0.22p = 0.29 r(22) = 0.22p = 0.30 r(22) = −0.01p = 0.97 r(22) = 0.03p = 0.90\n2–31 6 r(14) = −0.09p = 0.72 r(14) = 0.16p = 0.56 r(14) = −0.14p = 0.6 r(14) = 0.29p = 0.28\n3–4 19 r (17) = −0.20p = 0.41 r(17) = 0.02p = 0.92 r(17) = −0.29p = 0.22 r(17) = 0.34p = 0.16\n4–5 18 r (16) = 0.15p = 0.54 r(16) = 0.01p = 0.97 r(16) = 0.20p = 0.43 r(16) = −0.40p = 0.10\nTable 5 Greene Climacteric Symptom scale overall and subsection\ncorrelation with bipolar depression and mood elevation scores\nGreene MADRS\nN =5 4\nYMRS\nN =5 4\nOverall r(52) = 0.52 r(52) = 0.06\np < 0.0001 p = 0.66\nSubsection\nSomatic r(52) = 0.24 r(52) = 0.19\np= 0.09 p = 0.21\nVasomotor r(52) = 0.19 r(52) = −0.22\np = 0.16 p = 0.23\nPsychological r(52) = 0.55 r(52) = 0.08\np < 0.001 p = 0.59\nMarsh et al. International Journal of Bipolar Disorders  (2015) 3:17 Page 6 of 9\n\nHistory of reproductive-related mood exacerbations\nWomen who reported a mood exacerbation during the\npostpartum period tended to experience more mood\nelevation symptoms during the MT than women who did\nnot have a mood exacerbation in the postpartum. Depres-\nsion scores in the MT did not significantly differ when a\nwoman reported history of postpartum and or perimenstr-\nual mood exacerbation. Likewise, mood elevation symp-\ntoms in the MT did not significantly differ if a woman\nreported a history of perimenstrual mood changes or peri-\nmenstrual and postpartum mood exacerbations.\nA case series reports postpartum mood episodes associ-\nated with an increased risk of perimenopausal mood epi-\nsodes in bipolar disorder (Robertson Blackmore et al.\n2008). Ours and other previous work in bipolar disorder,\nin line with this study results, did not find reporting a\nmenstrual cycle or postpartum mood exacerbation associ-\nated with perimenopausal depression (Payne et al. 2007;\nMarsh et al. 2012). In the unipolar depression literature,\nthe results are mixed. Women, in one study reporting pre-\nmenstrual symptoms, were at greater risk of menopausal\ndepressed mood (Freeman et al. 2004b); however, this was\nnot the case in a cross-sectional report in unipolar depres-\nsion during the perimenopause (Steinberg et al. 2008).\nThe strength of this study is that it is the first to pro-\nspectively evaluate mood during the MT in women with\nbipolar disorder. The assessment of reproductive hormo-\nnal levels and mood in women with bipolar disorder has\nalso not been reported prior. Likewise, VMS in women\nwith bipolar disorder had not been previously re-\nported in relation to mood. The limitations of this\nstudy include a small sample size due to higher than\nanticipated dropout and slow recruitment. The sample\nwas too small to further evaluate early MT, late MT,\nand early postmenopause as separate groups. This\nstudy is also unable to differentiate what part, if any,\nthe worse mood ratings in the older aged late and\nearly post transition group may be due to worsening\ncourse of bipolar disorder with longer duration of the\nmental illness. Life stressors were not assessed.\nConclusions\nIn summary, women with bipolar disorder were found to\nhave significantly more depression and mood elevation\nsymptoms during the late MT and early postmenopause\ncompared to late reproductive age and early MT women.\nOverall, absolute levels and variability in FSH and estra-\ndiol were not associated with concurrent depression or\nmood elevation mood ratings despite an initial positive as-\nsociation of estradiol level and YMRS score at first visit.\nMenopausal symptoms, but not specifically VMS, were\nassociated with concurrent depression but not mood ele-\nvation severity. A history of postpartum mood exacerba-\ntions may be associated with greater mood elevation\nsymptoms during the MT. Further studies of greater\nduration are needed to discern whether the final men-\nstrual period is pivotal in mood pattern in the late MT to\nearly postmenopausal years. Knowing that, greater mood\nTable 6 Difference in bipolar depression and mood elevation scores during the menopausal transition based on history of\nreproductive phase mood exacerbation\nReproductive group Reproductive phase\nmood exacerbation\n(by woman), N (%)\nMean MADRS ± SD t test p value Mean YMRS ± SD t test p value\nPerimenstrual\nNwomen =5 5\nHistory of perimenstrual\nmood exacerbation\n47 (85) 15.4 ± 8.5 t = 0.20 p = 0.85 9.0 ± 6.1 t = −0.59 p = 0.56\nNo history of perimenstrual\nmood exacerbation\n8 (15) 16.1 ± 14.8 7.6 ± 7.2\nPostpartum\nNwomen =4 6\nHistory of postpartum\nmood exacerbation\n23 (64) 14.4 ± 7.9 t = 0.33 p = 0.74 11.2 ± 5.5 t = −1.91 p = 0.07\nNo history of postpartum\nmood exacerbation\n13 (36) 15.5 ± 10.8 7.2 ± 7.0\nPerimenstrual and postpartum\n(Nwomen = 46)\nBoth: history of postpartum\nand perimenstrual mood\nexacerbation\n20 (91) 14.3 ± 8.0 t = 0.97 p = 0.34 10.9 ± 5.8 t = −0.75 p = 0.46\nNeither: no history of postpartum\nor perimenstrual mood exacerbation\n2 (9) 21.0 ± 22.6 7.5 ± 10.6\nMarsh et al. International Journal of Bipolar Disorders  (2015) 3:17 Page 7 of 9\n\nseverity symptoms occurred during times of greater\nhormonal variability during the menopausal transition\n(the late MT and early postmenopause) may offer an\nopportunity to examine novel hormonal approaches to\nmood stability in women with bipolar disorder.\nCompeting interests\nAR: Anthony J Rothschild, MD—no monies have been received from nor\ninvestments held in an organization that may be financially invested in the\nmanuscript. He is a consultant for Allergan, Eli Lilly, GlaxoSmithKline, Omnicare,\nand Pfizer; has received grant/research support from Alkermes, AssureRx,\nCyberonics, National Institute of Mental Health, Janssen, and St. Jude Medical;\nand has received royalties for the Rothschild Scale for Antidepressant\nTachyphylaxis (RSAT)™ and from the American Psychiatric Press, and\nUp-to-Date. WM and BG declare that they have no competing interests.\nAuthors’ contributions\nWM designed the study and wrote the protocol, managed the study,\nperformed literature searches, engaged and directed the statistical analysis, and\ndrafted the manuscript. BG executed the statistical analysis and edited the\nstatistical section. AR consulted on execution of the study and contributed to\nthe final manuscript. All authors read and approved the final manuscript.\nAcknowledgements\nFunding for this study is from KL2RR031981 Clinical Research Scholar Award,\n# UL1TR000161 NIH/NCRR UMass Medical, and School Clinical and\nTranslational Science Award (CTSA). No role was played in influencing the\nanalysis, interpretation, manuscript composition, or submission.\nOur thanks to Joanne Nicholson, PhD (Dartmouth Psychiatric Research\nCenter) and Jean Frazier MD (University of Massachusetts Medical School) for\nthe mentorship and wisdom throughout the study.\nBruce Barton PhD and Aimee Kroll-Desrosiers, MS Quantative Health Services,\nUmass Medical School, are thanked for oversight, supplemental analyses, and\ntargeted editing of methods and results.\nAuthor details\n1Department of Psychiatry, School of Medicine, University of Massachusetts,\n55 Lake Ave North, S3-314, Worcester, MA 01655, USA. 2Department of\nOrthopedics and Physical Rehabilitation, University of Massachusetts, 55 Lave\nAve N - AC7 069, Worcester, MA, USA. 3Department of Psychiatry, University\nof Massachusetts, 55 Lave Ave North, Worcester, MA, USA.\nReceived: 5 February 2015 Accepted: 28 July 2015\nReferences\nAPA. American Psychiatric Association. Diagnostic and statistical manual of\nmental disorders, fifth edition. Arlington VA: American Psychiatric Publishing;\n2013.\nAvis NE, Brambilla D, McKinlay SM, Vass K. A longitudinal analysis of the\nassociation between menopause and depression. Results from the\nMassachusetts Women ’s Health Study. Ann Epidemiol. 1994;4(3):214 –20.\nBlehar MC, DePaulo Jr JR, Gershon ES, Reich T, Simpson SG, Nurnberger Jr JI.\nWomen with bipolar disorder: findings from the NIMH Genetics Initiative\nsample. Psychopharmacol Bull. 1998;34(3):239 –43.\nBromberger JT. The menopausal transition increases the risk of depressive\nsymptoms and depression diagnosis in women without a history of\ndepression. Evid Based Ment Health. 2006;9(4):110.\nBromberger JT, Kravitz HM, Chang YF, Cyranowski JM, Brown C, Matthews KA.\nMajor depression during and after the menopausal transition: Study of\nWomen’s Health Across the Nation (SWAN). Psychol Med. 2011;41(9):1879 –88.\ndoi:10.1017/S003329171100016X.\nBromberger JT, Matthews KA, Schott LL, Brockwell S, Avis NE, Kravitz HM, et al.\nDepressive symptoms during the menopausal transition: the Study of Women’s\nHealth Across the Nation (SWAN). J Affect Disord. 2007;103(1–3):267–72.\ndoi:10.1016/j.jad.2007.01.034.\nBromberger JT, Schott LL, Kravitz HM, Sowers M, Avis NE, Gold EB, et al.\nLongitudinal change in reproductive hormones and depressive symptoms\nacross the menopausal transition: results from the Study of Women ’s Health\nAcross the Nation (SWAN). Arch Gen Psychiatry. 2010;67(6):598 –607.\ndoi:10.1001/archgenpsychiatry.2010.55.\nBurger HG, Dudley EC, Hopper JL, Groome N, Guthrie JR, Green A, et al.\nProspectively measured levels of serum follicle-stimulating hormone,\nestradiol, and the dimeric inhibins during the menopausal transition in a\npopulation-based cohort of women. J Clin Endocrinol Metab.\n1999;84(11):4025–30.\nBurger HG, Hale GE, Robertson DM, Dennerstein L. A review of hormonal\nchanges during the menopausal transition: focus on findings from the\nMelbourne Women’s Midlife Health Project. Hum Reprod Update.\n2007;13(6):559–65. doi:10.1093/humupd/dmm020.\nCohen LS, Soares CN, Vitonis AF, Otto MW, Harlow BL. Risk for new onset of\ndepression during the menopausal transition: the Harvard study of moods\nand cycles. Arch Gen Psychiatry. 2006;63(4):385 –90. doi:10.1001/\narchpsyc.63.4.385.\nDennerstein L, Guthrie JR, Clark M, Lehert P, Henderson VW. A population-based\nstudy of depressed mood in middle-aged. Australian-born women.\nMenopause. 2004;11(5):563 –8.\nDennerstein L, Smith AM, Morse C, Burger H, Green A, Hopper J, et al.\nMenopausal symptoms in Australian women. Med J Aust. 1993;159(4):232 –6.\nFreeman EW. Associations of depression with the transition to menopause.\nMenopause. 2010;17(4):823 –7. doi:10.1097/gme.0b013e3181db9f8b.\nFreeman EW, Sammel MD, Boorman DW, Zhang R. Longitudinal pattern of\ndepressive symptoms around natural menopause. JAMA Psychiatry.\n2014;71(1):36–43. doi:10.1001/jamapsychiatry.2013.2819.\nFreeman EW, Sammel MD, Lin H. Temporal associations of hot flashes and\ndepression in the transition to menopause. Menopause. 2009;16(4):728 –34.\ndoi:10.1097/gme.0b013e3181967e16.\nFreeman EW, Sammel MD, Lin H, Nelson DB. Associations of hormones and\nmenopausal status with depressed mood in women with no history of\ndepression. Arch Gen Psychiatry. 2006;63(4):375 –82.\nFreeman EW, Sammel MD, Liu L, Gracia CR, Nelson DB, Hollander L. Hormones\nand menopausal status as predictors of depression in women in transition to\nmenopause. Arch Gen Psychiatry. 2004a;61(1):62 –70.\nFreeman EW, Sammel MD, Rinaudo PJ, Sheng L. Premenstrual syndrome as a\npredictor of menopausal symptoms. Obstet Gynecol. 2004b;103(5 Pt 1):960–6.\nFreeman MP, Smith KW, Freeman SA, McElroy SL, Kmetz GE, Wright R, et al. The\nimpact of reproductive events on the course of bipolar disorder in women.\nJ Clin Psychiatry. 2002;63(4):284 –7.\nGoldstein BI, Velyvis VP, Parikh SV. The association between moderate alcohol use\nand illness severity in bipolar disorder: a preliminary report. J Clin Psychiatry.\n2006;67(1):102–6.\nGreene JG. A factor analytic study of climacteric symptoms. J Psychosom Res.\n1976;20(5):425–30.\nHarlow SD, Gass M, Hall JE, Lobo R, Maki P, Rebar RW, et al. Executive summary\nof the Stages of Reproductive Aging Workshop + 10: addressing the\nunfinished agenda of staging reproductive aging. Menopause.\n2012;19(4):387–95. doi:10.1097/gme.0b013e31824d8f40.\nKendall DA, Stancel GM, Enna SJ. The influence of sex hormones on\nantidepressant-induced alterations in neurotransmitter receptor binding.\nJ Neurosci. 1982;2(3):354 –60.\nKendell RE, Chalmers JC, Platz C. Epidemiology of puerperal psychoses.\nBr J Psychiatry. 1987;150:662 –73.\nKennedy N, Boydell J, Kalidindi S, Fearon P, Jones PB, van Os J, et al. Gender\ndifferences in incidence and age at onset of mania and bipolar disorder over a\n35-year period in Camberwell, England. Am J Psychiatry. 2005;162(2):257–62.\nKennedy SH, Andersen HF, Lam RW. Efficacy of escitalopram in the treatment of\nmajor depressive disorder compared with conventional selective serotonin\nreuptake inhibitors and venlafaxine XR: a meta-analysis. J Psychiatry Neurosci.\n2006;31(2):122–31.\nKleine-Budde K, Touil E, Moock J, Bramesfeld A, Kawohl W, Rossler W. Cost of\nillness for bipolar disorder: a systematic review of the economic burden.\nBipolar Disord. 2013. doi:10.1111/bdi.12165.\nLukasiewicz M, Gerard S, Besnard A, Falissard B, Perrin E, Sapin H, et al.\nYoung Mania Rating Scale: how to interpret the numbers? Determination\nof a severity threshold and of the minima l clinically significant difference\nin the EMBLEM cohort. Int J Method s Psychiatr Res. 2013;22(1):46 –58.\ndoi:10.1002/mpr.1379.\nMaartens LW, Knottnerus JA, Pop VJ. Menopausal transition and increased\ndepressive symptomatology: a community based prospective study.\nMaturitas. 2002;42(3):195 –200.\nMarsh et al. International Journal of Bipolar Disorders  (2015) 3:17 Page 8 of 9\n\nMarsh WK, Ketter TA, Crawford SL, Johnson JV, Kroll-Desrosiers AR, Rothschild AJ.\nProgression of female reproductive stages associated with bipolar illness\nexacerbation. Bipolar Disord. 2012;14(5):515–26. doi:10.1111/j.1399-5618.2012.01026.x.\nMarsh WK, Ketter TA, Rasgon NL. Increased depressive symptoms in menopausal\nage women with bipolar disorder: age and gender comparison. J Psychiatr\nRes. 2009. doi:10.1016/j.jpsychires.2008.11.003.\nMarsh WK, Templeton A, Ketter TA, Rasgon NL. Increased frequency of depressive\nepisodes during the menopausal transition in women with bipolar disorder:\npreliminary report. J Psychiatr Res. 2008;42(3):247 –51. doi:10.1016/\nj.jpsychires.2006.12.006.\nMcEwen BS, Alves SE, Bulloch K, Weiland NG. Ovarian steroids and the brain:\nimplications for cognition and aging. Neurology. 1997;48(5 Suppl 7):S8 –15.\nMontgomery SA, Asberg M. A new depression scale designed to be sensitive to\nchange. Br J Psychiatry. 1979;134:382 –9.\nPayne JL, Roy PS, Murphy-Eberenz K, Weismann MM, Swartz KL, McInnis MG,\net al. Reproductive cycle-associated mood symptoms in women with major\ndepression and bipolar disorder. J Affect Disord. 2007;99(1 –3):221–9.\ndoi:10.1016/j.jad.2006.08.013.\nRasgon NL, Altshuler LL, Fairbanks LA, Dunkin JJ, Davtyan C, Elman S, et al.\nEstrogen replacement therapy in the treatment of major depressive disorder\nin perimenopausal women. J Clin Psychiatry. 2002;63 Suppl 7:45 –8.\nRobertson Blackmore E, Craddock N, Walters J, Jones I. Is the perimenopause a\ntime of increased risk of recurrence in women with a history of bipolar\naffective postpartum psychosis? A case series. Arch Womens Ment Health.\n2008;11(1):75–8. doi:10.1007/s00737-008-0215-2.\nRubinow DR, Schmidt PJ, Roca CA. Hormone measures in reproductive\nendocrine-related mood disorders: diagnostic issues. Psychopharmacol Bull.\n1998;34(3):289–90.\nSachs GS, Thase ME, Otto MW, Bauer M, Miklowitz D, Wisniewski SR, et al.\nRationale, design, and methods of the systematic treatment enhancement\nprogram for bipolar disorder (STEP-BD). Biol Psychiatry. 2003;53(11):1028 –42.\nSchmidt PJ, Haq N, Rubinow DR. A longitudinal evaluation of the relationship\nbetween reproductive status and mood in perimenopausal women. Am J\nPsychiatry. 2004;161(12):2238–44.\nSchurhoff F, Bellivier F, Jouvent R, Mouren-Simeoni MC, Bouvard M, Allilaire\nJF, et al. Early and late onset bipolar disorders: two different forms of\nmanic-depressive illness? J Affect Disord. 2000;58(3):215 –21.\nSedgwick P. SAS Institute Inc. Cary, NC: SAS Institute Inc.; 2012.\nSimon NM, Otto MW, Wisniewski SR, Fossey M, Sagduyu K, Frank E, et al. Anxiety\ndisorder comorbidity in bipolar disorder patients: data from the first 500\nparticipants in the Systematic Treatment Enhancement Program for Bipolar\nDisorder (STEP-BD). Am J Psychiatry. 2004;161(12):2222 –9. doi:10.1176/\nappi.ajp.161.12.2222.\nSit D, Rothschild AJ, Wisner KL. A review of postpartum psychosis. J Womens\nHealth (Larchmt). 2006;15(4):352 –68. doi:10.1089/jwh.2006.15.352.\nSoares CN, Almeida OP, Joffe H, Cohen LS. Efficacy of estradiol for the treatment\nof depressive disorders in perimenopausal women: a double-blind,\nrandomized, placebo-controlled trial. Arch Gen Psychiatry. 2001;58(6):529 –34.\nSowers MR, Zheng H, McConnell D, Nan B, Harlow S, Randolph Jr JF. Follicle\nstimulating hormone and its rate of change in defining menopause\ntransition stages. J Clin Endocrinol Metab. 2008a;93(10):3958 –64.\ndoi:10.1210/jc.2008-0482.\nSowers MR, Zheng H, McConnell D, Nan B, Harlow SD, Randolph Jr JF.\nEstradiol rates of change in relation to the final menstrual period in a\npopulation-based cohort of women. J Clin Endocrinol Metab.\n2008b;93(10):3847 –52. doi:10.1210/jc.2008-1056.\nSteinberg EM, Rubinow DR, Bartko JJ, Fortinsky PM, Haq N, Thompson K et al. A\ncross-sectional evaluation of perimenopausal depression. J Clin Psychiatry.\n2008:e1–e8. doi:ej07m03790 [pii].\nWoods NF, Smith-DiJulio K, Percival DB, Tao EY, Mariella A, Mitchell S. Depressed\nmood during the menopausal transition and early postmenopause:\nobservations from the Seattle Midlife Women ’s Health Study. Menopause.\n2008;15(2):223–32. doi:10.1097/gme.0b013e3181450fc2.\nYoung RC, Biggs JT, Ziegler VE, Meyer DA. A rating scale for mania: reliability,\nvalidity and sensitivity. Br J Psychiatry. 1978;133:429 –35.\nSubmit your manuscript to a \njournal and benefi t from:\n7 Convenient online submission\n7 Rigorous peer review\n7 Immediate publication on acceptance\n7 Open access: articles freely available online\n7 High visibility within the fi  eld\n7 Retaining the copyright to your article\n    Submit your next manuscript at 7 springeropen.com\nMarsh et al. International Journal of Bipolar Disorders  (2015) 3:17 Page 9 of 9","source_license":"CC0","license_restricted":false}