A chemiluminescence-based GH assay with 30- to 100-fold increased sensitivity recently disclosed combined basal and pulsatile GH secretion in men. However, how age, sex steroid hormones, and obesity singly and jointly influence the basal vs. pulsatile modes of GH release is not known. We used the foregoing assay (detection threshold, 0.002-0.005 microgram/L) and high sensitivity and specificity (> or = 90% each) deconvolution analysis to quantitate basal and pulsatile GH secretion from 24-h serum GH concentration profiles in 26 healthy lean and obese men, whose ages spanned 18-63 yr and whose percentage body fat ranged from 12-47%. Concentrations of serum insulin-like growth factor I (IGF-I), IGF-I-binding protein-1 (IGFBP-1), and IGFBP-3 were related to specific measures of basal or pulsatile GH release. We observed that mean (24-h) serum GH concentrations embraced a 140-fold range from 0.013-1.8 micrograms/L and were related negatively to age (r = -0.50; P < 0.01), percentage body fat (r = -0.620; P < 0.01), and their interaction (r = -0.610; P < 0.01). In contrast, testosterone was a robustly positive statistical determinant of mean serum GH values (r = 0.628; P = 0.0006). Stepwise multivariate regression analysis disclosed that percentage body fat alone and jointly with the serum testosterone concentration controlled, respectively, 38% and 50% of the total variability in GH levels (P = 0.0013 and P = 0.0008). As assessed by deconvolution analysis, GH secretory burst mass was negatively related to percentage body fat (r = -0.621; P < 0.01) and positively to serum testosterone (r = 0.529; P = 0.0054). The calculated half-life of GH correlated positively with serum estradiol (r = 0.447; P = 0.032), and negatively with percentage body fat (r = -0.437; P = 0.048). Basal GH secretion rates were negatively related to serum estradiol (r = -0.485; P = 0.016). In contrast, GH secretory burst frequency and duration were unrelated to age, percentage body fat, or sex steroids. The fraction of total GH secreted in bursts was negatively correlated with the body mass index (r = -0.540; P < 0.01). Serum IGF-I concentrations were positively related to total pulsatile GH secretion (r = 0.690; P = 0.0011) and negatively to age (r = -0.597; P = 0.007) and percentage body fat (r = -0.611; P = 0.009).(ABSTRACT TRUNCATED AT 400 WORDS)