Endogenous sex hormones and cardiovascular disease incidence in men.

BACKGROUND: Data suggest that endogenous sex hormones (testosterone, dehydroepiandrosterone sulfate [DHEA-S], and estradiol) influence cardiovascular disease (CVD) risk factors and vascular function. Yet, prospective studies relating sex hormones to CVD incidence in men have yielded inconsistent results. OBJECTIVE: To examine the association of circulating sex hormone levels and CVD risk in men. DESIGN: Prospective cohort study. SETTING: Community-based study in Framingham, Massachusetts. PARTICIPANTS: 2084 middle-aged white men without CVD at baseline. MEASUREMENTS: The authors used multivariable Cox regression to relate baseline levels of testosterone, DHEA-S, and estradiol to the incidence of CVD (coronary, cerebrovascular, or peripheral vascular disease or heart failure) during 10 years of follow-up. RESULTS: During follow-up, 386 men (18.5%) experienced a first CVD event. After adjustment for baseline standard CVD risk factors, higher estradiol level was associated with lower risk for CVD (hazard ratio per SD increment in log estradiol, 0.90 [95% CI, 0.82 to 0.99]; P = 0.035). The authors observed effect modification by age: Higher estradiol levels were associated with lower CVD risk in older (median age >56 years) men (hazard ratio per SD increment, 0.86 [CI, 0.78 to 0.96]; P = 0.005) but not in younger (median age < or =56 years) men (hazard ratio per SD increment, 1.11 [CI, 0.89 to 1.38]; P = 0.36). The association of higher estradiol level with lower CVD incidence remained robust in time-dependent Cox models (updating standard CVD risk factors during follow-up). Serum testosterone and DHEA-S levels were not statistically significantly associated with incident CVD. LIMITATIONS: Sex hormone levels were measured only at baseline, and the findings may not be generalizable to women and nonwhite people. CONCLUSIONS: In the community-based sample, a higher serum estradiol level was associated with lower risk for CVD events in older men. The findings are consistent with the hypothesis that endogenous estrogen has vasculoprotective influences in men.

Duke Scholars

Author Michael J Pencina Biostatistics & Bioinformatics, Division of Biostatistics