Experimentally induced prolonged hyperglycemia increases insulin release in humans, and in animals has been demonstrated to increase vagal efferent activity. The objective of the present experiment was to determine whether in humans, the compensatory increase in insulin release in response to short-term mild hyperglycemia is mediated by an induction of vagal efferent activity. Lean male subjects (n = 11; body mass index, 23.6 +/- 0.8 kg/m(2)) underwent a frequently sampled iv glucose tolerance test (FSIGT) to determine B cell function and insulin sensitivity. Subjects were then tested under four conditions over 4 months. Subjects were infused for 48 h with either glucose (15% dextrose at 200 mg/m(2).min) or saline (50 ml/h). Three hours after termination of the infusion, an FSIGT was administered in the presence of either saline or atropine (0.4 mg/m(2) bolus: 0.4 mg/m(2).h). Glucose (117 +/- 14 vs. 98 +/- 5 mg/dl) and insulin (49.5 +/- 10 vs. 23 +/- 5 muU/ml) levels were significantly elevated during the 48-h glucose infusion compared with those during saline treatment. Forty-eight-hour glucose infusions increased insulin and C-peptide levels during the FSIGT. When the FSIGT was conducted in the presence of atropine after glucose infusion, C-peptide levels were significantly attenuated during the period of endogenous insulin secretion (0-20 min; 31.8 +/- 13 vs. 39.2 +/- 11.9, atropine vs. no atropine) and exogenous insulin administration [20-40 min; 18.8 +/- 10.8 vs. 31.6 +/- 12.9., atropine vs. no atropine; F(3,9) = 4.99; P < 0.026]. A significant negative correlation was found between the repression of C-peptide by muscarinic blockade and the magnitude of the C-peptide response to the glucose infusion (r = 0.60; P < 0.045). Insulin AUC was not significantly altered by the presence of muscarinic blockade. In summary, we found that prolonged mild hyperglycemia results in a compensatory increase in C-peptide secretion, which is partially mediated by an induction in vagal efferent activity.