The purpose of this study was to evaluate the GH response to exercise and the effects of endurance training on this response in early middle-aged men. Seven healthy middle-aged [M; 42.0+/-2.4 (+/-SD) yr old] and five young (Y; 21.2+/-1.1 yr old) competition cyclists were investigated before and after 4 months of intensive endurance training. Subjects performed an exhaustive incremental exercise test (50 watts for 3 min) with gas exchange measurement, and blood samples for lactate, glucose, and GH determinations were drawn before exercise, at the end of the exercise, and in the recovery phase (1, 3, 5, 10, 15, 20, and 30 min). Basal insulin-like growth factor I was also determined. At exhaustion no differences were found in relative maximal heart rate or blood lactate and glucose peaks. On the contrary, the two groups had markedly different GH responses; in fact, the peak GH response to exhaustive exercise was much lower in M than in Y (8.1+/-1.3 vs. 57.1+/-15.5 microg/L; P<0.01). The training, similar in subjects of the same group, increased progressively from 182 to 300 km/week (+64.8%) in M and from 350 to 600 km/week (+71.4%) in Y. After the training, the percent increase in maximal oxygen consumption was similar in the two groups (M, +15.2%; Y, +17.5%), confirming that the efficiency of the training performed was comparable. In neither group did training have any effect on the GH peak response to exercise, confirming the blunted GH response in M compared to Y (6.7+/-1.0 vs. 61.0+/-12.9 microg/L; P<0.01). Similarly, insulin-like growth factor I concentrations were not significantly affected by training. In conclusion, active middle-aged subjects, compared with the young, showed a blunted GH response to a physiological stimulus such as exercise, indicating that the age-related decline in GH secretion appears in early middle age. This response was not modified by training in either early middle-aged or young subjects.