Low Serum Levels of Dehydroepiandrosterone Sulfate Predict All-Cause and Cardiovascular Mortality in Elderly Swedish Men

Context: The age-related decline in dehydroepiandrosterone (DHEA) levels is thought to be of importance for general and vascular aging. However, data on the association between DHEA and mortality are conflicting. Objectives: We tested the hypothesis that low serum DHEA and DHEA sulfate (DHEA-S) levels predict all-cause and cardiovascular disease (CVD) death in elderly men. Design, Setting, and Participants: We used gas/liquid chromatography-mass spectrometry to an-alyzebaseline levels ofDHEAandDHEA-S in theprospectivepopulation-basedMrOSSweden study (2644men,aged69–81yr).Mortalitydatawereobtained fromcentral registers andanalyzedusing Cox proportional hazards regressions. Main Outcome Measures: All-cause and CVD mortality by serum DHEA(-S) levels. Results: During a mean 4.5-yr follow-up, 328 deaths occurred. Low levels of DHEA-S (quartile 1 vs. quartiles 2–4), predicteddeath fromall causes [hazard ratio (HR) 1.54, 95%confidence interval (CI) 1.21–1.96; adjusted for traditional cardiovascular risk factors], from CVD (n 123 deaths; HR 1.61, 95 % CI 1.10–2.37) and ischemic heart disease (n 73; HR 1.67, 95 % CI 1.02–2.74) but not cancer. Analyses with DHEA gave similar results. The association between low DHEA-S and CVD death remained after adjustment for C-reactive protein and circulating estradiol and testosterone levels. When stratified by the median age of 75.4 yr, the mortality prediction by low DHEA-S was more pronounced among younger (age adjusted HR for CVD death 2.64, 95 % CI 1.37–5.09) than olde