Effects of fasting on plasma glucose and prolonged tracer measurement of hepatic glucose output in NIDDM.

We studied the measurement of hepatic glucose output (HGO) with prolonged [3-3H]glucose infusion in 14 patients with non-insulin-dependent diabetes mellitus (NIDDM). Over the course of 10.5 h, plasma glucose concentration fell with fasting by one-third, from 234 +/- 21 to 152 +/- 12 mg/dl, and HGO fell from 2.35 +/- 0.18 to 1.36 +/- 0.07 mg . kg-1 . min-1 (P less than .001). In the basal state, HGO and glucose were significantly correlated (r = 0.68, P = .03), and in individual patients, HGO and glucose were closely correlated as both fell with fasting (mean r = 0.79, P less than .01). Plasma [3-3H]glucose radioactivity approached a steady state only 5-6 h after initiation of the primed continuous infusion, and a 20% overestimate of HGO was demonstrated by not allowing sufficient time for tracer labeling of the glucose pool. Assumption of steady-state instead of non-steady-state kinetics in using Steele's equations to calculate glucose turnover resulted in a 9-24% overestimate of HGO. Stimulation of glycogenolysis by glucagon injection demonstrated no incorporation of [3-3H]glucose in hepatic glycogen during the prolonged tracer infusion. In a separate study, plasma glucose was maintained at fasting levels (207 +/- 17 mg/dl) for 8 h with the glucose-clamp technique. Total glucose turnover rates remained constant during this prolonged tracer infusion. However, HGO fell to 30% of the basal value simply by maintaining fasting hyperglycemia in the presence of basal insulin levels.(ABSTRACT TRUNCATED AT 250 WORDS)