Familial aggregation and possible major gene effects were evaluated for the baseline serum dehydroepiandrosterone sulfate (DHEAS) level and the change in DHEAS in response to a 20-week exercise training program in a sample of 481 individuals from 99 Caucasian families who were sedentary at baseline and who participated in the HERITAGE Family Study. Baseline DHEAS levels were not normally distributed, and were therefore logarithmically transformed and adjusted for the effects of age and sex prior to genetic analysis. The DHEAS response to training was computed as the simple difference, post-training minus baseline, and was adjusted for the baseline DHEAS level, age, and sex. Maximal (genetic and familial environmental) heritabilities (using a familial correlation model) reached 58% and 30% for the baseline and the response to training, respectively. Our estimate for the baseline is generally in agreement with previous reports, suggesting that the magnitude of the familial effect underlying this phenotype in these sedentary families is similar to that in the general population. However, segregation analysis showed no evidence for a multifactorial familial component in data for either the baseline or the response to training. Rather, a major additive gene controlling the baseline was found. For the response to training in the complete sample, transmission of the major effect from parents to offspring was ambiguous, but in a subset of 56 'responsive' families (with at least 1 family member whose response to training was greater than 1 standard deviation) this major effect was Mendelian in nature. The putative major genes accounted for 50% and 33% of the variance for the baseline and the response to training, respectively. The novel finding in this study is that the baseline DHEAS level and the change in DHEAS in response to training may be influenced by major gene effects. (C) 2000 by W.B. Saunders Company.