Intestinal transport and fermentation of resistant starch evaluated by the hydrogen breath test.

OBJECTIVE: To study fermentability of different samples of resistant starch (RS), compared to one another and to lactulose, and to study the effect on gastric emptying of addition of RS to test meal. Finally to study if adaptation to RS results in a measurable change in fermentation pattern, (H2/CH4 production). Sources of RS: Raw potato starch (RPS), 58% RS; corn flakes (CF), 5% RS; hylon VII high amylomaize starch, extrusion cooked and cooled (HAS) 30% RS; highly retrograded hylon VII high amylomaize starch (HRA) 89% RS.

DESIGN: (1) Fermentation: seven healthy volunteers ingested in randomized order 50 g RPS, 100 g CF, 75 g HAS, 25 g HRA. End-expiratory H2/CH4 was measured every 30 min for 12 to 22 hours post-ingestion as a measure of fermentation. A dose-response study of RPS, 5, 10, 25, 50, 75 and 100 g was performed. (2) Adaptation: In five 3-week periods seven volunteers added daily to their usual diet 50 g of either RPS, HAS, oat bran, wheat bran or common maize starch. The polysaccharides were administered in randomized order. The test periods were separated by 1 week's wash out. Basic end-expiratory H2/CH4 was measured once a week prior to and during the test periods. (3) Gastric emptying: The rate of increase in blood glucose was measured after test meals consisting of 50 and 100 g of RPS, 50 g HAS and 50 g glucose dissolved in a gel, alone, and mixed with 25 g of RPS. As controls we chose wheat bran and oat bran.

RESULTS: (1) We found that RPS is fermentable, although the cumulated excessive H2 production after 50 g RPS corresponding to 29 g RS was clearly less than after 10 g lactulose. The time from ingestion of RPS to a sustained increase in end-expiratory H2 (apparent transit time; 5-11 h) was longer than lactulose (1-4 h), indicating either a slow passage through the small intestine or a slow fermentation rate. 100 g of corn flakes (4.6 g RS) resulted in a measurable increase in H2 production, equivalent to 10-20 g RPS, whereas neither of the two samples of hylon VII high amylomaize resulted in any significant increase in H2 production. The dose-response study with RPS showed that even 5 g of RPS resulted in a measurable increase in end-expiratory H2, and increasing doses from 5 g to 100 g resulted in a seemingly exponential increase in H2 production. (2) 3 weeks' daily administration of HAS resulted in a slightly elevated increase in basic end-expiratory H2, although the increase did not reach statistical significance. RPS resulted in a sustained increase in basic end-expiratory H2. Both RS samples increased measurable end-expiratory CH4 in volunteers with measurable CH4 production after a lactulose load, but 3 weeks' daily challenge with these slowly fermentable substrates did not increase measurable CH4 in volunteers, who prior to the study only produced CH4 intermittently. (3) The rate of increase in blood sugar was unaffected by addition of RS or non-starch-polysaccharides to the test meal, indicating that addition of the polysaccharides does not affect gastric emptying.

CONCLUSIONS: A fraction of RPS is resistant to digestion in the small intestine, and it is fermentable by the colonic microbial flora. RS from CF, HAS and RPS give very different H2 responses, either due to differences in digestion patterns or fermentation patterns. Short-term adaptation (3 weeks) to HAS or RPS does not change the H2/CH4 response. RS does not affect gastric emptying of a test meal consisting of glucose dissolved in a gel.