Abstract 3497: Low Serum Estradiol and Cardiovascular Disease Mortality in Older Men: The Rancho Bernardo Study
Context: The traditional view that the male excess in cardiovascular disease (CVD) mortality is due to harmful effects of testosterone has been challenged. In older men, low, rather than high, testosterone predicts atherosclerotic disease progression, and serum estradiol, not testosterone, is prospectively associated with CVD events.
Objective: To determine the association of endogenous sex hormones with cardiovascular disease mortality in older community-dwelling men.
Methods: Prospective population-based study of 858 men aged 50 to 91 years (median 73) who had serum estradiol and testosterone measurements at baseline (1984 – 87) and were followed for mortality through July, 2006.
Results: Median (95% interval) hormone levels were 300 (175, 505) ng/dl for testosterone and 20 (11, 32) pg/ml for estradiol. During the 22 year (average 12.6) follow-up, 559 deaths occurred; 281 (50%) were due to CVD. Age-adjusted Cox regression analyses using hormone quintiles showed testosterone was not related to CVD mortality, but suggested a threshold effect for lower levels of estradiol (P < .001). The risk of CVD death was 1.77 (95% CI 1.32, 2.38) for men in the lowest estradiol quintile (< 15 pg/ml) compared to those with higher levels (P < .001), adjusting for age, BMI, current smoking, physical activity and alcohol intake. This association was independent of the metabolic syndrome, diabetes and prevalent cardiovascular disease. Additional adjustment for lipids, lipoproteins, blood pressure, insulin resistance, CRP, IL-6, leptin, adiponectin or androgen deficiency (total testosterone < 250 ng/dl) had minimal effect on results. Exclusion of deaths that occurred during the first 2 years (to minimize the effect of occult disease) or adjustment for a combination of health status markers also had no effect. Men with low estradiol had lower BMI, smaller waist girth, and were more likely to have lost 10 pounds in the previous 10 years (all P < .001); adjusting for these factors did not influence results.
Conclusions: Low serum estradiol in older men is associated with increased risk of cardiovascular death, independent of age, adiposity, and lifestyle choices. This relationship does not appear to be due to pre-existing disease and is not related to androgen deficiency.