Estradiol rates of change in relation to the final menstrual period in a population-based cohort of women

Context/Objective: The aim was to characterize rates of change in serum estradiol (E2) levels across the menopausal transition and into early postmenopause. Setting/Participants: We studied the Michigan Bone Health and Metabolism Study cohort of 629 women with median age of 38 yr (interquartile range, 7) at the 1992-1993 baseline with annual assessment of E2 levels over the subsequent 15-yr period. Design/Main Outcome Measures: The purpose was to describe patterns of acceleration/deceleration in (log)E2 rates of change before and after the final menstrual period (FMP) using nonparametric and piecewise regression modeling. Results: Between -10 to -2 yr to the FMP, mean fitted serum E2 population values were relatively stable. The 95% confidence bands around the slight increase in E2 rate of change 5 yr prior to the FMP included the value of no change. The fitted population mean E2 value declined 67% from 64.5 pg/ml (SE = 3.6) to 21 pg/ml (SE = 1.2) in the 4 yr between -2 < FMP < +2. A second significant mean E2 rate of change was identified from 6-8 yr after FMP. Fitted population mean E2 values declined 18% from 18.1 pg/ml (SE = 1.3) at FMP = 6 to 14.8 pg/ml (SE = 1.3) at FMP = 8. In nonobese women, the mean E2 percent decline was 42% from FMP = 6 to FMP = 8, whereas in obese women, the mean E2 percent decline over this time was 31%. Conclusions: Population mean serum E2 levels were sustained until approximately 2 yr prior to the FMP. In the ensuing 4-yr period, E2 levels declined 67%. A secondary E2 decline, commencing about 6 yr after the FMP, was observed in nonobese but not obese women.