Sex hormones, insulin, lipids, and prevalent ischemic heart disease

The relations between estradiol, testosterone, insulin, lipids, and prevalent ischemic heart disease were examined using the cross-sectional data from the Caerphilly Heart Disease Study, a cohort of 2,512 men (aged 45-59 years) surveyed between 1978 and 1982. Endogenous levels of estradiol were associated directly with high density lipoprotein (HDL) cholesterol (r = 0.106, p less than 0.001), but this relation was removed after adjustment for testosterone and insulin levels. Estradiol was not associated with prevalent ischemic heart disease. Endogenous levels of testosterone were associated directly with HDL cholesterol (r = 0.148, p less than 0.001) and inversely with triglyceride (r = -0.217, p less than 0.001). Persons with prevalent ischemic heart disease had significantly lower testosterone levels than persons without ischemic heart disease (mean levels 20.9 vs. 22.0 nmol/liter, p less than 0.01). These relations were confounded by associations with insulin. The associations between testosterone and the lipids persist after adjusting for body mass index, age, and insulin. The association between testosterone and prevalent ischemic heart disease was reduced after adjusting for insulin and/or triglyceride levels. The results suggest that insulin and testosterone may have an interdependent regulatory effect on lipid metabolism. The effect of testosterone on ischemic heart disease appears to be primarily mediated through its association with insulin. Future work on sex hormones and ischemic heart disease will need to account for the effects of insulin.