Table 1.

Cross-Sectional and Longitudinal Studies of Depression and Menopause^a

Authors, Year, Setting	Study Type	n at Baseline (%)b	Follow-Up, y	Measure	Results	Limitations
McKinlay and Jefferys, 1974, United Kingdom40	Cross-sectional	Premenopausal: 134 (21)	∼1	Postal questionnaire	Depression most frequent symptom all groups; hot flashes and night sweats peak during perimenopause	8 postmenopausal groupings requiring age at last menses, possible recall errors, self-report
		Perimenopausal: 234 (37)
		Postmenopausal: 270 (42)
Ballinger, 1975, United Kingdom21	Cross-sectional	Premenopausal: 228 (45)	∼1	Postal questionnaire	Preponderance of “psychiatric cases” in perimenopausal group and women aged 45–49 y	Self-report, smallest number in perimenopausal group
		Perimenopausal: 81 (16)
		Postmenopausal: 193 (38)
Bungay et al, 1980, United Kingdom7	Cross-sectional	806 women stratified in 5-y age groups, aged 30–64 y	∼ 1	Postal questionnaire	Peaks of prevalence of psychiatric symptoms just before mean age of menopause	Self-report, no clear indication of association of chronological age and menopausal age
Hunter and Whitehead, 1989, United Kingdom46	Cross-sectional	Premenopausal: 248	Not clear	Postal questionnaire	Depressed mood was significantly increased in perimenopausal and postmenopausal women compared with premenopausal women; distress greatest among younger postmenopausal women	Population sample was volunteer based
		Perimenopausal: 351
		Postmenopausal: 761
Hunter, 1990, United Kingdom80	Longitudinal	Premenopausal: 6 (13)	3	WHQ	Depressed mood significantly increased between premenopause and perimenopause or postmenopause	Small sample size
		Perimenopausal: 31 (66)
		Postmenopausal: 10 (21)
Matthews et al, 1990, United States62	Longitudinal	Premenopausal: 541	2.5	BDI	No significant changes in depressive symptoms from premenopause to postmenopause	Short follow-up
Kaufert et al, 1992, United States22	Longitudinal	Total: 469	3	CES-D	Natural menopause does not appear to increase odds of depression	Initial sample of women not random
Koster and Davidsen, 1993, Denmark47	Retrospective, longitudinal	Premenopausal: 205 (39)	4	Postal questionnaire	Depression increased slightly during perimenopause	Study population recruited from metropolitan suburb areas in Denmark
		Perimenopausal: 67 (13)
		Postmenopausal: 51 (10)
Avis et al, 1994, United States23	Longitudinal	Premenopausal: 485 (21)	5	CES-D	Onset of menopause not significantly associated with increased risk of depression; significant increased risk of depression associated with perimenopause vs postmenopause (in model that excluded menopausal symptoms)	Symptoms and depression measured by self-report
		Perimenopausal: 1549 (66)
		Postmenopausal: 240 (10)
		Surgical: 78 (3)
Collins and Landgren, 1994, Sweden81	Cross-sectional	Premenopausal: 967 (73)	Not clear	MMSI	Small but significant differences between premenopausal and postmenopausal women regarding negative mood	Self-report, few perimenopausal subjects included
		Perimenopausal: 79 (6)
		Postmenopausal: 278 (21)
Bromberger et al, 2001, United States24	Cross-sectional,longitudinal	Premenopausal: 4483 (43)	5	12-item symptom questionnaire	Highest rates of psychological distress in early perimenopause, lowest in premenopause and postmenopause; odds of distress varied by ethnic group	No hormonal data to validate menopausal status; checklist used was not a validated instrument
		Perimenopause:
		Early: 3534 (34)
		Late: 609 (6)
		Postmenopause: 1748 (17)
Avis et al, 2001, United States48	Prospective, observational	Premenopausal: 129	Not clear	CES-D	CES-D score not significantly associated with premenopause, perimenopause, or postmenopause	Symptoms of depression measured by self- report
		Perimenopausal: 99
		Postmenopausal: 64
Maartens et al, 2002, Netherlands49	Cross-sectional, longitudinal	Premenopausal: 475 (23)	2.8–4.7	EDS	Transition from perimenopause to postmenopause significantly associated with increased EDS score	Measured depressive symptoms rather than assessing diagnosis of depression
		Perimenopausal: 982 (47)
		Postmenopausal: 646 (31)
Bromberger et al, 2003, United States30	Community- based, cross- sectional, longitudinal	Premenopausal: 1688 (53)	Not clear	Symptom questionnaire	Early perimenopause associated with increased odds of dysphoric mood, irritability, and nervousness compared with premenopause	Overall measure of dysphoric mood was not validated; no prospective menstrual diary data
		Perimenopausal: 1473 (47)
Freeman et al, 2004, United States25	Longitudinal	Premenopausal: 436	4	CES-D	Increased risk of depression during transition to menopause, which decreased during postmenopause; risk of depression decreased with rapidly increasing FSH levels	Less than 5% of patients reached menopausal status during the 4-year study; only measured hormone levels during follicular phase; patient sample limited to African American and white women
Schmidt et al, 2004, United States39	Longitudinal	Premenopausal: 29	5	SCID	Events related to the late and early perimenopause as well as early postmenopause may be associated with an increased susceptibility to develop depression in some women	Small sample size; inability to prospectively confirm incidence of depression during other periods of life
Travers et al, 2005, Australia41	Longitudinal	Premenopausal: 88	Not clear	GCS	Premenopausal and perimenopausal women had higher depression scores vs postmenopausal women; women in earlypostmenopause had higher depression scores vs women in intermediate or late postmenopause	In this normative study, population evaluated was derived from a geographical sector of a large Australian city, leaving some racial and ethnic groups underrepresented
		Perimenopausal: 34
		Postmenopausal: 314
Cohen et al, 2006, United States28	Longitudinal	Premenopausal: 460	3, then 6	CES-D	In women with no history of depression, those who enter the menopausal transition earlier have significant risk for first-onset depression	Prospective assessment of depression not based on structured clinical interviews; at end of initial 3-y period, follow-up interval was 59–92 mo; result is imprecise temporal relationship between menopausal transition and new onset of depression
Freeman et al, 2006, United States29	Longitudinal	Premenopausal: 231	8	CES-D	Transition to menopause strongly associated with new onset of depressed mood among women with no history of depression, especially with greater hormonal flux	Fixed intervals of follow-up assessments may lack precision
Bromberger et al, 2007, United States26	Longitudinal	Premenopausal: 3,302	5	CES-D	Change in menopausal status associated with increased risk of significant depressive symptoms, independent of relevant demographic, psychosocial, behavioral, and health factors	Measured current depressive symptoms rather than diagnosis of clinical depression
Freeman et al, 2007, United States50	Longitudinal	Premenopausal: 367	9	Symptom questionnaire	Depressed mood higher through late premenopause and early and late perimenopause; decreased postmenopause	Findings based on participants' self-reporting of symptom occurrence and severity rather than diagnosis of symptoms
		Perimenopausal: 37
Woods et al, 2008, United States 82	Longitudinal	Total: 302	15	CES-D	Age, late menopause, and hot flashes were significantly associated with depressed mood	Symptoms of depression measured by self-report

^aData from Burt et al.⁷⁹

^bPercentages were not provided for all groups.

Abbreviations: BDI=Beck Depression Inventory, CES-D=Center for Epidemiologic Studies-Depression scale, EDS=Edinburgh Depression Scale, FSH=follicle-stimulating hormone, GCS=Greene Climacteric Scale, MMSI=Modified Menopause Symptom Inventory, SCID=Structured Clinical Interview for DSM-IV, WHQ=Women's Health Questionnaire.