IN-VIVO STUDY OF COLONIC FERMENTATION OF CARBOHYDRATE IN INFANTS

Dietary carbohydrate in the colon is fermented and converted into shor t-chain fatty acids. We studied the fate of carbohydrate that arrives in the colon under circumstances similar to those that occur during an episode of diarrhea and determined whether a quantitative correlation exists among certain indicators of colonic fermentation of carbohydra te arriving in the large bowel. A stable-isotope method was used to es timate carbon scavenging by the colon. Fourteen infants with severe ma lnutrition and history of watery stools and/or increased numbers of st ools in the preceding 20 days were studied. Infants underwent nasoceca l intubation and a 60-min infusion of 0.5 g/kg glucose containing 5 mg /kg of C-13-glucose. Stools were assessed for carbohydrate-fermenting bacteria, acetate, glucose, and C-13 abundance; blood was assessed for acetate; and breath was assessed for hydrogen. Some of the infants el iminated the infusate per anus within 30 min of the infusion (group I; n = 5), while others did so 120 min or more after the infusion (Group II; n = 9). The volume of fecal output after the intracecal infusion differed significantly between group I and group II (57 +/- 13 vs. 24 +/- 4 ml; p < 0.05). Although baseline breath hydrogen (3 +/- 1 vs. 10 +/- 4 parts per million) and plasma acetate (41 +/- 4 vs. 47 +/- 9 mu M) values were not significantly different between groups, breath hyd rogen values at 45 min after start of infusion were significantly lowe r in group I than in group II (3 +/- 1 vs. 19 +/- 5 ppm; p = 0.008); 6 0 min after start of infusion, plasma acetate levels were significantl y lower in group I than in group II (34 +/- 8 vs. 69 +/- 11 mu M; p = 0.02), and the difference remained significant throughout the study. R egression analyses indicated a significant association between time el apsed from start of intracolonic infusion and plasma acetate (p = 0.01 4) and breath hydrogen (p = 0.04) levels for group II. Although we fou nd good correlation between two indicators of colonic fermentation, bl ood acetate and breath hydrogen, these values did not correlate well w ith carbon scavenging or fecal bacterial counts. It is possible that t he way the colon handles the infusate-i.e., water absorption and/or se cretion-may affect carbohydrate fermentation.