The glucose stimulus-response curve of the beta-cell in physically trained humans, assessed by hyperglycemic clamps

In order to examine the effect of habitual exercise on beta-cell responses over a wide range of plasma glucose levels, plasma insulin and C-peptide responses to 2 1/2-hour hyperglycemic clamps at 7.5, 10, and 15 mmol/L glucose were assessed in six trained athletes and six age- and weight-matched sedentary controls. Athletes were significantly fitter than controls (estimated maximal oxygen uptake [VO2 max] mean 44 v 30 mL.kg-1.min-1, P less than .05) and were more sensitive to insulin as assessed by dividing the mean glucose infusion rate over the last 20 minutes of the clamp by the steady-state plasma insulin (mean 0.44 v 0.19 mg.min-1.kg-1.nmol-1. L, respectively, P less than .01). Plasma C-peptide responses were lower in the athletes, both fasting (geometric mean 0.28 v 0.62 nmol/L, P less than .05), and at the end of all clamps (at 7.5, 10, and 15 mmol/L plasma glucose, respectively, 0.65 v 1.43, 1.25 v 2.85, and 2.40 v 4.46 nmol/L each, P less than .05). First-phase plasma C-peptide responses were lower in the athletes at the 10 and 15 mmol clamp levels. The slope of the glucose-C-peptide stimulus-response curve was approximately linear over the range examined, the slope being significantly shallower in athletes than controls for both first phase (P less than .01) and second phase (P less than .01). Plasma insulin responses were similar to C-peptide responses. The attenuation of beta-cell responsiveness over a wide glucose range may be an adaptation to the enhanced peripheral insulin sensitivity seen in athletes.