Predictors and clinical significance of declining plasma dehydroepiandrosterone sulfate in old age.

Dehydroepiandrosterone sulfate (DHEAS) was measured in random persons of three age cohorts (75, 80 and 85 years, N=271) at five-year intervals in order to find out predictors and significance of declining DHEAS in old age. The mean values decreased from 2.88 micromol/L to 2.39 micromol/L in men (p<0.001), and from 1.93 micromol/L to 1.73 micromol/L in women (p<0.05) at entry. Strong correlations were found between the baseline levels of DHEAS and those measured after five years both in men (r=0.727, p<0.001) and women (r=0.605, p<0.001), and the changes in DHEAS were associated with DHEAS levels at entry (r=-0.418, p<0.05). Baseline DHEAS was higher (2.47 micromol/L vs 2.05 micromol/L, p<0.05) and the decline more pronounced (-0.50 micromol/L vs 0.20 micromol/L, p<0.05) in the healthy subjects than in those suffering from diseases at entry, but the percentage five-year decline was similar (-6.5% and -5.2%) in both groups. The five-year decline in DHEAS was predicted neither by the baseline levels of risk indicators, e.g., serum lipids, body mass index, electrocardiographic, nor echocardiographic findings at entry. The age-and gender-adjusted baseline levels of DHEAS predicted neither mortality nor cognitive decline with 5- and 10-year follow-up periods. The 5-year decline in DHEAS was significant (p<0.05) in the subjects who died or developed cognitive decline during the subsequent 5-year follow-up. However, the changes did not differ significantly from those with favorable prognosis. The data indicate that the decline in DHEAS is primarily a gender-specific aging phenomenon, and only partly a consequence of actual diseases and frailty.