1197 Prevalence of Celiac Disease in Patients Presenting With Short Stature: A Systematic Review and Meta-Analysis

Abstract

INTRODUCTION: Short stature is a common extra-intestinal manifestation of celiac disease (CeD). The proportion of patients with short stature having CeD varies. We conducted a systematic review and meta-analysis to assess the prevalence of CeD in patients with presenting with short stature. METHODS: We searched Medline and EMBASE databases for the keywords 'celiac disease, coeliac disease, celiac, tissue transglutaminase antibody, anti-endomysial antibody, short stature and growth retardation'. All studies published from 1991 to February 2019 were included. Patients without any prior evaluation for short stature were classified as all-cause short stature, while previously evaluated patients, where no cause was found for short stature, were classified as having idiopathic short stature. Diagnosis of CeD was based on the European Society of Pediatric Gastroenterology, Hepatology and Nutrition guidelines. A random-effects model was used to pool the data. Risk of bias was assessed across each study using a validated tool developed for prevalence studies by Hoy et al (1). RESULTS: Seventeen studies including 3959 patients (1729 with all-cause short stature and 2230 with idiopathic short stature) were included. The pooled seroprevalence of CeD based on positive anti-tissue transglutaminase antibody, anti-endomysial antibody and antigliadin antibody was 11.5% (95% CI 2.34, 20.7; I2 = 47%) and 12.7% (95% CI 8.2, 17.3; I2 = 93.3%) for all-cause and idiopathic short stature (Figure 1), respectively. The pooled prevalence of biopsy-proven CeD was 7.2% (95% CI 4.44, 9.9; I2 = 77.7%) and 9.6 (95% CI 6.8,12.4) for all-cause (Figure 2) and idiopathic short stature (Figure 3), respectively. There was an overall moderate risk of selection bias and significant heterogeneity in the present studies. Statistical heterogeneity was partially due to differences in the diagnostic strategy (biopsy in seropositive individuals vs biopsy in all suspected patients), study design (retrospective vs prospective studies). Age, sample size did not influence the results on meta-regression analysis. CONCLUSION: Approximately 1 in 14 patients with all-cause short stature and 1 in 10 patients with idiopathic short stature have biopsy-confirmed CeD. Therefore, all patients with short stature should be evaluated for CeD.