Dual sugar permeability testing in diarrheal disease

Objective: To assess the validity of the use of a blood specimen for the sugar permeability test because of the high failure rate of 5-hour urine collection in young children with diarrhea. Study design: Simultaneous 5-hour urine collections and timed blood tests were taken after ingestion of an isotonic solution of lactulose (L) and L-rhamnose (R) in 24 children with acute gastroenteritis and 25 children without diarrhea in a control group. Sugars were measured with highperformance liquid chromatography, and the percent of recovered sugars was expressed as an L-R ratio. Results: With acute gastroenteritis the geometric mean L-R ratios (95% confidence intervals) were 12.4 (9.3 to 16.3) in urine and 9.4 (6.7 to 13.1) in blood compared with 6.7 (5.0 to 8.8) and 5.9 (4.4 to 7.8), respectively, in the control group. The level of agreement (kappa) among normal, intermediate, and high ratios for blood and urine was 0.71 (0.51 to 0.92). The failure rate of L-R tests was significantly reduced with a blood specimen (urine 37% vs blood 10%; P < .0001). Conclusions: Intestinal permeability testing on a blood specimen is a valid alternative to urine collection in young children and has a significantly lower test failure rate.