We measured serum reproductive hormone concentrations in a community-based, multiethnic population of premenopausal and early perimenopausal women to determine whether there are ethnic differences in hormones that can be explained by host factors. We studied 2930 participants in the Study of Women's Health Across the Nation who were aged 42-52 yr and self-identified as African-American (27.6%), Caucasian (47.1%), Chinese (7.4%), Hispanic (8.8%), or Japanese (9.0%) at 7 clinical sites. Outcome measures from this baseline assessment of a longitudinal study were serum estradiol (E2), FSH, testosterone (T), dehydroepiandrosterone sulfate, and SHBG concentrations and calculated estimates of free steroid availability, free testosterone index, and free E2 index from serum collected primarily in the early follicular phase of a spontaneous menstrual cycle. The primary explanatory variables were race/ethnicity, menopausal status, age, body mass index, day of the cycle, smoking, alcohol use, and physical activity. Chinese women had lower unadjusted E2 and SHBG levels, and Hispanic women had lower unadjusted T levels than other ethnic groups. Unadjusted serum FSH levels did not differ by race/ethnicity. E2 levels adjusted for host characteristics, particularly body size, did not differ by race/ethnicity. Adjusted FSH levels were higher, and adjusted T levels were lower in African-American and Hispanic women. Serum E2 and FSH concentrations were highly variable. Serum FSH levels, but no other hormone concentrations, were positively correlated with menopausal status. Serum dehydroepiandrosterone sulfate levels were negatively correlated with age, but not menopausal status. All hormone concentrations were significantly correlated with body mass index. We conclude that serum sex steroid, FSH, and SHBG levels vary by ethnicity, but are highly confounded by ethnic disparities in body size. to reproductive senescence in women is characterized by a progressive loss of ovarian response to pituitary stimulation as ovarian follicles lose the ability to develop, mature, and ovulate reproductively competent oocytes. Endocrinologically, this process of reproductive aging is characterized by a progressive rise in serum FSH levels associated with a decrease in serum estradiol (E2) levels (1), although current evidence suggests stage-specific variation in this general pattern (2). In contrast, serum dehydroepiandrosterone sulfate (DHEAS) levels decrease steadily with age without any known relationship to the loss of ovarian folliculogenesis (3, 4), leading to consideration of DHEAS as a marker of somatic aging. Once again, evidence is emerging that menopause stage-specific data are at variance with that pattern (5).No study, whether cross-sectional (6) or longitudinal (1,7,8), has investigated ethnic differences in reproductive hormones in midlife women. However, ethnic variation in hormone concentrations has been observed in menopausal women, as serum E2 levels are lower in Chinese women than in Caucasian women (9). In addition, host facto...

Studies of menstrual cycle length in large populations demonstrated that there is a striking increase in the variability of intermenstrual intervals just before menopause. The changes in serum concentrations of luteinizing hormone (LH), follicle-stimulating hormone (FSH), estradiol (E2), and progesterone (P) during menstrual cycles in a group of perimenopausal women were compared with the findings in young normal women. In 8 women, 46-56 years old with regular cycles, cycle length was shorter and the mean E2 concentration was lower than in younger women. There was a striking increase in FSH concentration throughout the cycle while LH remained in the normal range. In 2 women, 14 cycles of variable length were studied during 2 years of the menopausal transition. In some instances, hormonal changes associated with follicular maturation and corpus luteum function occurred in the presence of high, menopausal levels of LH and FSH with a diminished secretion of E2 and P. In others vaginal bleeding occurred during a fall in serum E2 with no associated rise in P. Cycles of variable length during the menopausal transition may be due either to irregular maturation of residual follicles with diminished responsiveness to gonadotropin stimulation, or to anovulatory vaginal bleeding that may follow estrogen withdrawal without evidence of corpus luteum function. The observation of elevated FSH concentrations and normal LH levels in perimenopausal women emphasizes the complexity of the hypothalamic-pituitary-ovarian regulatory system and suggests that LH and FSH are modulated independently at the level of the pituitary.

A B S T R A C T The changes in serum levels of luteinizing hormone (LH), follicle-stimulating hormone (FSH), estradiol, and progesterone that occur both early and late in reproductive life were characterized and compared with findings in young, normal women and in patients with certain menstrual disorders.A total of 50 complete menstrual cycles in 37 women were examined. Five distinct patterns of hormonal regulation were found, three of which are reported here: (a) A long follicular phase and delayed follicular maturation in young women with long, unpredictable intermenstrual intervals from menarche; (b) a short follicular phase with increasing age and in short cycles in perimenopausal women; and (c) true anovulatory vaginal bleeding in long cycles in perimenopausal women.The short cycles before and during the menopausal transition were found to have lower E2 levels and high FSH concentrations throughout, while LH remained in the normal range. During long cycles in perimenopausal women, concentrations of LH and FSH were in the menopausal range. However, follicular maturation was observed months after high levels of gonadotropins were attained.These studies permit the characterization of the menstrual history of the normal woman in terms of the hormonal changes that occur and provide a basis for the definition of several disorders of follicular maturation.

The relation of the reproductive endocrine system to impotence in older men was examined by measuring the concentrations of testosterone (T), bioavailable testosterone (BT), LH, and PRL and body mass index (BMI) in 57 young controls (YC), 50 healthy potent older controls attending a health fair (HF), and 267 impotent patients (SD). The SD and HF had markedly reduced mean T and BT values compared to YC. When adjusted for age and BMI there was no difference in BT between potent and impotent older men. The percent BT was much higher in YC than in the older groups. While the percent BT rose significantly with increased T in YC, it was inversely related to T in the older subjects, suggesting that increased sex hormone-binding globulin binding was a primary event leading to a low BT. Forty-eight percent of HF and 39% of SD were hypogonadal, as defined by a mean BT of 2.5 SD or more below the mean of YC (less than or equal to 2.3 nmol/L). Ninety percent of these had LH values in the normal range, suggesting hypothalamic-pituitary dysfunction. Thirty-four SD and six each of YC and older control volunteers (OC) underwent GnRH testing. Older subjects showed impaired responsiveness to GnRH compared to YC. A low basal LH level correlated very highly with hyporesponsiveness to GnRH. Thus, secondary hypogonadism and impotence are two common, independently distributed conditions of older men.