GH induces lipolysis in vivo, increasing plasma free fatty acid (FFA) levels; in turn, FFA are able to reduce GH release, and acipimox, a nicotinic acid analog able to block lipolysis, enhances in normal subjects the GH response to GHRH. Obesity and old age are characterized by a blunted GH response to several stimuli, including GHRH; reports also indicate high plasma FFA levels in obesity and sometimes in the elderly. The aim of this study was to evaluate the possible role of FFA in GH release in obese and elderly subjects. According to a randomized, single blind, cross-over protocol, six healthy subjects, six obese subjects, and six elderly subjects received on 2 different days, with a 1-week interval, placebo or acipimox (250 mg, orally) at 0700 and 1100 h; GHRH [GHRH-(1-44)NH2; 50 micrograms in healthy subjects and in elderly subjects, 100 micrograms in obese subjects] was injected iv at 1300 h, and blood samples for evaluation of plasma FFA, blood glucose, serum insulin (IRI), and serum GH levels were taken from 1200 to 1500 h. Plasma FFA levels were always lower (P < 0.05) after acipimox than after placebo (0.03 +/- 0.01 vs. 0.13 +/- 0.02 g/L in healthy subjects, 0.09 +/- 0.01 vs. 0.27 +/- 0.02 g/L in obese, 0.02 +/- 0.005 vs. 0.17 +/- 0.01 g/L in elderly subjects); serum IRI levels were also lower (P < 0.05) after acipimox than after placebo in the three groups of subjects (16 +/- 3 vs. 30 +/- 5, 120 +/- 30 vs. 181 +/- 32, and 21 +/- 3 vs. 49 +/- 9 pmol/L); both FFA (P < 0.05) and IRI levels (P < 0.05) were higher in obese than in healthy or elderly subjects after placebo and acipimox. Blood glucose levels were not different in the three groups of subjects after either placebo or acipimox. The integrated GH response to GHRH-(GH delta area) was always greater (P < 0.05) after acipimox than after placebo (4677 +/- 633 vs. 1599 +/- 373 in healthy, 1469 +/- 230 vs. 343 +/- 114 in obese, 2304 +/- 759 vs. 325 +/- 133 micrograms/L.120 min in elderly subjects); after both placebo and acipimox, the GH delta area was greater (P < 0.05) in healthy subjects than in obese or elderly subjects. The GH delta area of elderly and obese subjects after acipimox was not different from the GH delta area of healthy subjects after placebo. Changes in GH delta areas were not significantly related to changes in FFA or IRI induced by acipimox; in contrast, absolute values of FFA and IRI as well as basal GH levels were all significantly related to the GH delta area. At multiple regression analysis, FFA was the only significant predictor of GH delta area. These data indicate that acute pharmacological reduction of plasma FFA levels restores the blunted GH response to GHRH commonly observed in obese and elderly subjects: however, when lipolysis is blocked to a similar extent, healthy subjects still show a higher GH delta area than obese or elderly subjects. As FFA are the best predictor of the GH delta area, we suggest that in obesity, the blunted GH release is due to high FFA levels, whereas in the elderly there might be an abnormal sensitivity to normal FFA levels.