Endogenous sex hormone changes in postmenopausal women in the diabetes prevention program

Abstract

Context: Whether endogenous sex hormones (ESH) [SHBG, estradiol, testosterone, and dehydroepiandrosterone (DHEA)] are altered by intensive lifestyle modification (ILS) or metformin and whether such changes affect glucose levels among dysglycemic postmenopausal women is unclear.

Objectives: Our objective was to examine intervention impact on ESH and associations with fasting plasma glucose (FPG) and 2-h glucose changes among postmenopausal glucose-intolerant women.

Design: We performed a secondary analysis of a randomized controlled trial.

Participants: Participants included postmenopausal, overweight, glucose-intolerant women not using exogenous estrogen (n = 382) who participated in the Diabetes Prevention Program.

Interventions: Interventions included ILS with the goals of weight reduction of at least 7% of initial weight and 150 min/wk of moderate intensity exercise or metformin or placebo administered 850 mg twice a day.

Main outcome measures: Intervention-related changes in ESH and associations of changes in ESH and glucose levels were evaluated.

Results: ILS significantly increased SHBG and decreased DHEA before and after adjustment for changes in waist circumference and fasting insulin. ILS did not alter estradiol or testosterone. Metformin did not change any ESH. ILS-induced increases in SHBG and declines in DHEA were associated with decreases in FPG and 2-h glucose, and declines in estradiol were associated with decreases in FPG, before and after adjustment for age, FSH, race/ethnicity, changes in waist circumference, and 1/fasting insulin.

Conclusions: Among postmenopausal glucose-intolerant women not using estrogen, ILS increased SHBG levels and lowered DHEA levels. These changes were associated with lower glucose independent of adiposity and insulin. Metformin effects upon ESH were not significant.

Fig. 1.

Unadjusted changes in SHBG and sex hormone levels by randomization assignment, means, and 95% confidence intervals. A, Change in SHBG levels between year 1 and baseline; B, change in DHEA levels between year 1 and baseline; C, change in E2 levels between year 1 and baseline; D, change in T levels between year 1 and baseline. *, P < 0.01 compared with placebo.

Fig. 2.

Change in sex hormone level by ILS and metformin treatment relative to placebo. Values are β-coefficients (95% confidence intervals) with sex hormone change as the dependent variable and randomization arm as the independent variable. A, Association between randomization arm and change in SHBG; B, association between randomization arm and change in DHEA; C, association between randomization arm and change in E2; and D, association between randomization arm and change in T. All associations are adjusted for covariates (age, race/ethnicity, FSH, changes in waist circumference, 1/fasting insulin, and baseline level of hormone.) *, P < 0.01 compared with placebo.