Relationship between gluten availability and celiac disease prevalence: A geo-epidemiologic systematic review

Celiac disease is a global disease requiring genetic susceptibility and gluten exposure to trigger immune-mediated enteropathy. The effect of the degree of gluten-containing grain availability on celiac disease prevalence is unknown. Our objective was to compare country-based gluten availability to celiac prevalence using a systematic literature review. We searched MEDLINE, Embase, Cochrane, and Scopus until May 2021. We included population-based serum screening with confirmatory testing (second serological study or small intestine biopsy) and excluded specific, high-risk, or referral populations. We determined country-specific gluten availability using the United Nations food balance for wheat, barley, and rye. Human leukocyte antigen (HLA) frequencies were obtained from allelefrequencies.net. The primary outcome was association between gluten-containing grain availability and celiac disease prevalence. Generalized linear mixed models method with Poisson's link was used for analysis. We identified 5641 articles and included 120 studies on 427 146 subjects from 41 countries. Celiac disease prevalence was 0-3.1%, median 0.75% (interquartile range 0.35, 1.22). Median wheat supply was 246 g/capita/day (interquartile range 214.8, 360.7). The risk ratio (RR) for wheat availability on celiac disease was 1.002 (95% confidence interval [CI]: 1.0001, 1.004, P = 0.036). A protective association was seen with barley, RR 0.973 (95% CI: 0.956, 0.99, P = 0.003), and rye, RR 0.989 (95% CI: 0.982, 0.997, P = 0.006). The RR for gross domestic product on celiac disease prevalence was 1.009 (95% CI: 1.005, 1.014, P < 0.001). The RR for HLA-DQ2 was 0.982 (95% CI: 0.979, 0.986, P < 0.001), and that for HLA-DQ8 was 0.957 (95% CI: 0.950, 0.964, P < 0.001). In this geo-epidemiologic study, gluten-containing grain availability showed mixed associations with celiac disease prevalence.