Short-term alterations in carbohydrate energy intake in humans. Striking effects on hepatic glucose production, de novo lipogenesis, lipolysis, and whole-body fuel selection.

Short-term alterations in dietary carbohydrate (CHO) energy are known to alter whole-body fuel selection in humans, but the metabolic mechanisms remain unknown. We used stable isotope-mass spectrometric methods with indirect calorimetry in normal subjects to quantify the metabolic response to six dietary phases (5 d each), ranging from 50% surplus CHO (+50% CHO) to 50% deficient CHO (-50% CHO), and 50% surplus fat (+50% fat). Fasting hepatic glucose production (HGP) varied by > 40% from deficient to surplus CHO diets (1.78 +/- 0.08 vs 2.43 +/- 0.09 mg/kg per min, P < 0.01). Increased HGP on surplus CHO occurred despite significantly higher serum insulin concentrations. Lipolysis correlated inversely with CHO intake as did the proportion of whole-body lipolytic flux oxidized. Fractional de novo hepatic lipogenesis (DNL) increased more than 10-fold on surplus CHO and was unmeasurable on deficient CHO diets; thus, the preceding 5-d CHO intake could be inferred from DNL. Nevertheless, absolute hepatic DNL accounted for < 5g fatty acids synthesized per day even on +50% CHO. Whole-body CHO oxidation increased sixfold and fat oxidation decreased > 90% on surplus CHO diets. CHO oxidation was highly correlated with HGP (r2= 0.60). HGP could account for 85% of fasting CHO oxidation on +25% CHO and 67% on +50% CHO diets. Some oxidation of intracellular CHO stores was therefore also occurring. +50% fat diet had no effects on HGP, DNL, or fuel selection. We conclude that altered CHO intake alters HGP specifically and in a dose-dependent manner, that HGP may mediate the effects of CHO on whole-body fuel selection both by providing substrate and by altering serum insulin concentrations, that altered lipolysis and tissue oxidation efficiency contribute to changes in fat oxidation, and that surplus CHO is not substantially converted by the liver to fat as it spares fat oxidation, but that fractional DNL may nevertheless be a qualitative marker of recent CHO intake.