Multicenter study on season of birth and celiac disease: evidence for a new theoretical model of pathogenesis.

OBJECTIVE: To investigate whether season of birth is associated with celiac disease (CD). STUDY DESIGN: We performed a medical record review of 1964 patients with biopsy-proven CD at 3 teaching hospitals (2 pediatric centers and 1 adult center) between 2000 and 2010. The first positive small intestinal biopsy result defined age of diagnosis. The observed proportions of births in each season (spring [March-May], summer [June-August], fall [September-November], and winter [December-February]) were compared with the expected proportions using binomial probability tests. RESULTS: The mean age at diagnosis was 9.8 ± 5.0 years in the 2 pediatric centers and 43.6 ± 15.8 years in the adult center. The cohort was predominately female (69%). Overall, more patients were born in spring (27%) than in any other season: summer (25%), fall (25%), and winter (23%). In patients diagnosed before age 15 years, the spring birth excess was present in boys (33%; P = .0005), but not in girls (26%; P = .43). The sex difference in season of birth was less striking in patients with CD diagnosed at age ≥15 years. CONCLUSION: Season of birth is an environmental risk factor for CD, particularly in boys diagnosed before age 15 years. The results are consistent with a new theoretical model that integrates potential environmental factors (eg, gluten introduction, ultraviolet-B exposure, vitamin D status) and acute viral gastrointestinal infections in early childhood.

Predictors of gluten avoidance and implementation of a gluten-free diet in children and adolescents without confirmed celiac disease.

OBJECTIVES: To determine independent predictors of gluten avoidance and of a physician's decision to initiate a gluten-free diet (GFD) in children and adolescents without confirmed celiac disease. STUDY DESIGN: We performed a structured medical record review of 579 patients aged 1-19 years presenting for evaluation of celiac disease between January 2000 and December 2010 at a large Boston teaching hospital. We collected data including demographic information, medical history, serology, small intestinal biopsy, history of gluten avoidance, and the postworkup recommendation of implementation of a GFD. Predictors of gluten-related issues were identified by multivariate logistic regression. RESULTS: Among 579 children without a previous diagnosis of celiac disease (mean age, 8.7 years), 43 (7.4%) had ever avoided gluten. Independent predictors of gluten avoidance were irritability or poor temper (OR, 3.2), diarrhea (OR, 2.5), weight issues (OR, 0.4), pervasive developmental disorder (OR, 5.3), and family history of celiac disease (OR, 2.2). Among 143 children without confirmed celiac disease who underwent diagnostic evaluation, several predictive factors were associated with a physician- recommended/parent-initiated GFD: irritability (OR, 6.4), diarrhea (OR, 3.4), pervasive developmental disorder (OR, 7.9), and positive serology before the referral (OR, 4.3). CONCLUSION: Gluten avoidance among children and adolescents without a previous diagnosis of celiac disease is relatively common. The identified predictors suggest that gluten avoidance is associated with nonspecific behavioral and gastrointestinal complaints and perhaps with the perceived dietary responses in another family member thought to have celiac disease.