The association of endogenous hormone concentrations and bone mineral density measures in pre- and perimenopausal women of four ethnic groups: SWAN.

We evaluated bone mineral density (BMD), hormone concentrations and menstrual cycle status to test the hypothesis that greater variations in reproductive hormones and menstrual bleeding patterns in mid-aged women might engender an environment permissive for less bone. We studied 2336 women, aged 42-52 years, from the Study of Women's Health Across the Nation (SWAN) who self-identified as African-American (28.2%), Caucasian (49.9%), Japanese (10.5%) or Chinese (11.4%). Outcome measures were lumbar spine, femoral neck and total hip BMD by dual-energy X-ray densitometry (DXA). Explanatory variables were estradiol, testosterone, sex hormone binding globulin (SHBG) and follicle stimulating hormone (FSH) from serum collected in the early follicular phase of the menstrual cycle or menstrual status [premenopausal (menses in the 3 months prior to study entry without change in regularity) or early perimenopause (menstrual bleeding in the 3 months prior to study entry but some change in the regularity of cycles)]. Total testosterone and estradiol concentrations were indexed to SHBG for the Free Androgen Index (FAI) and the Free Estradiol Index (FEI). Serum logFSH concentrations were inversely correlated with BMD (r = -10 for lumbar spine [95% confidence interval (CI): -0.13, -0.06] and r = -0.08 for femoral neck (95% CI: -0.11, -0.05). Lumbar spine BMD values were approximately 0.5% lower for each successive FSH quartile. There were no significant associations of BMD with serum estradiol, total testosterone, FEI or FAI, respectively, after adjusting for covariates. BMD tended to be lower (p values = 0.009 to 0.06, depending upon the skeletal site) in women classified as perimenopausal versus premenopausal, after adjusting for covariates. Serum FSH but not serum estradiol, testosterone or SHBG were significantly associated with BMD in a multiethnic population of women classified as pre- versus perimenopausal, supporting the hypothesis that alterations in hormone environment are associated with BMD differences prior to the final menstrual period.