Celiac disease in children with persistent diarrhea and failure to thrive

To determine the frequency of celiac disease (CD) in children with failure to thrive (FTT) and/ or persistent diarrhea (PD) not responding to conventional therapy.

Cross-sectional study.

Gastroenterology Unit of Sindh Institute of Urology and Transplantation (SIUT), from January 2002 to January 2004.

Forty nine children and adolescents with PD (defined as diarrhea greater than 14 days duration) and / or FTT (based on anthropometrical indicators, i.e. weight for length below the 5th centile) were included in the study. Demographic data, weaning practices, breast feeding and family history of CD were documented. Laboratory workup included tissue transglutaminase antibody IgA (tTGA), IgA and IgG antigliadin antibodies (IgA AGA and IgG AGA), HLA typing, upper gastrointestinal endoscopy (EGD) with distal duodenal biopsy (where allowed by parents) and anthropometric data. CD diagnosis was in accordance with the "Guidelines for the diagnosis of and treatment of CD for children" by NASPGHAN. Mann-Whitney U-test, Chi-square test and Fisher's exact test were used for analysis as applicable.

Forty nine patients (25 [51%] males) with FTT and or PD were included. The mean (+/-SD) age, height, weight and BMI were: 10.1+/-6.2 years (range, 1-20 years), 107.0 +/- 31.7 cm (45-180 cm), 19.04 +/- 12.5 kg (3-68 kg) and 16.12+/-12.58 kg/cm2 respectively. FTT was present in 30 (61%) patients. Thirty four (69%) children had a history of PD, 38 (77%) had a significant history of weight loss and 32 (65%) children had short stature. Majority of the children (88%) were breast fed. Weaning was started at 6 months of age in 40% and included mixed diet according to age. Protuberant abdomen was present in 26 (53%). Elevated tTGA was significantly (p < 0.001) more frequent in patients with CD. HLA-DQ2 and DQ8 haplotypes were positive in 18 (60%) of these patients. Thirty (61%) patients were positive for CD based on Marsh criteria. CD diagnosis was supported by positive tTGA and/ or AGA tests. Majority i.e. 28/30 patients had Marsh stage 3 disease, while 22/30 patients with CD also had PD. Four CD patients were found to suffer from protein calorie malnutrition (PCM).

Children who fail to thrive with or without PD may have CD. A positive tTGA test in these children is a useful aid in making a CD diagnosis. Majority of CD positive children were found to carry HLA DQ2 or DQ8 genes. Thus, HLA typing could be used to determine genetic disposition to CD. Majority of FTT had Marsh stage 3 changes on histopathology.