Serum D-xylose absorption tests: reproducibility and diagnostic usefulness in food-induced enterocolitis

We evaluated the use of the D-xylose absorption test as a marker of intestinal mucosal damage following single-dose oral food challenges in infants suspected of food-protein-induced-enterocolitis. The absorption tests were performed before any challenge, and again after each of four separate food challenges. The response to challenge was judged by five objective criteria (diarrhea, polymorphonuclear leukocytosis, fecal blood, leukocytes, and eosinophils). A significant decrease in the group mean for serum xylose was seen following positive challenges. The 1-h serum xylose level prior to any challenge was 42 +/- 9.5, following negative challenges it was 46 +/- 15, and following positive challenges it was 28 +/- 4.4. However, there was not a good individual correlation between the challenge response and xylose absorption, so intrasubject variability of the xylose absorption test was examined in six infants who had at least three tests performed before a positive challenge response was encountered. The average standard deviation for these repeated tests was 8.2 mg/dl, and the coefficient of variability was 12.6%. This would suggest that, in utilizing this test to assess mucosal damage following oral food challenges or gluten reintroduction, a decrease of 16 mg/dl from the prechallenge values would have to be seen before the difference would exceed two standard deviations. This degree of change was not seen in any patient following a single oral food challenge.