RESUMEN
CONTEXT: Older men have lower T levels, but whether differences in circulating T or its metabolites dihydrotestosterone (DHT) or estradiol (E2) contribute to cardiovascular disease remains controversial. OBJECTIVE: We tested the hypothesis that plasma T, DHT, and E2 are differentially associated with the incidence of myocardial infarction (MI) and stroke in older men. PARTICIPANTS AND METHODS: Plasma total T, DHT, and E2 were assayed using liquid chromatography-mass spectrometry in early-morning samples from 3690 community-dwelling men aged 70-89 years. Outcomes of the first hospital admission or death due to MI or stroke were ascertained by data linkage. RESULTS: Mean follow-up was 6.6 years. Incident MI occurred in 344, stroke in 300, and neither in 3046 men. In a multivariate analysis adjusting for age and other risk factors, T, DHT, and E2 were not associated with incident MI [fully adjusted hazard ratio (HR) for T in quartile (Q) 4 vs Q1: 0.92, 95% confidence interval (CI) 0.66-1.28; DHT: 0.83, 95% CI 0.59-1.15; E2: 0.84, 95% CI 0.62-1.15]. Higher T or DHT was associated with a lower incidence of stroke (T: Q4: Q1 fully adjusted HR 0.56, 95% CI 0.39-0.81, P = .002; DHT: 0.57, 95% CI 0.40-0.81, P = .002). E2 was not associated with stroke (HR 0.76, 95% CI 0.54-1.08, P = .123). CONCLUSIONS: Higher plasma T or DHT is a biomarker for reduced risk of stroke but not MI. Androgen exposure may influence outcomes after rather than the incidence of MI, whereas androgens but not E2 are independent predictors of stroke risk. Randomized clinical trials are needed to clarify the impact of modifying T or DHT on the risk of stroke in aging men.