Gastric bypass and duodenal and gastric feeding: a comment to Hansen et al.

IN THE PAPER BY Hansen et al. published in American Journal of Physiology Gastrointestinal and Liver Physiology (2), results from a study on gastric bypass (RYGB)-operated patients before, and 1 and 6 wk after surgery are reported. Postoperative feeding was done via two different routes: a jejunal (oral food ingestion) route using the gastric bypass and a gastric route via a gastrostomy catheter. This is a very interesting model, in which administration of nutrients via the gastric route mimics preoperative feeding, at a time when oral intake of nutrients is delivered directly to the jejunum via the gastric bypass. It is thus possible to study the effect of bypassing duodenum and part of jejunum acutely. Nine patients, five with Type 2 diabetes and four with normal glucose tolerance, were studied with a mixed meal administered over a 20-min period. Glucose peak levels, glucose area under the curve (AUC), insulin peak levels, insulin AUC, peak intact glucagon-like peptide-1 (GLP-1), and intact GLP-1 AUC did not differ significantly irrespective of the route of nutrient administration. The authors conclude that exclusion of the foregut does not influence glucose tolerance and insulin resistance and that GLP-1 secretion does not influence insulin secretion after RYGB. These conclusions deviate from those of other investigators including ourselves using the same model, who found oral feeding after RYGB to result in radically different glucose profiles and increased insulin and GLP-1 peak levels and incremental AUCs compared with gastric feeding. Also insulin and GLP-1 peak levels were found to correlate closely on the day of oral but not gastric feeding (1, 4). How can these discrepancies be explained? Hansen et al. ascribe the differences to one subject suffering from postprandial hypoglycemia (4), thus representing a physiological outlier, and their measuring intact GLP-1 (Millipore, Luminex) rather than total GLP-1. We would like to point toward two more probable causes for these differences. In the report from Hansen and coworkers blood was sampled before the meal, after 20 min (end of meal), and thereafter with 60-min intervals. In our two case studies, glucose, insulin, and total GLP-1 levels peaked within the first hour after the end of the meal following oral intake of nutrients. Since peak levels were very