Causal relationship between diet and celiac disease in leukemia: A two-sample Mendelian randomization study

Abstract

Objective To explore the causal relationship between celiac disease and a wheat-free diet with leukemia. Methods We obtained genome-wide association study (GWAS) data on celiac disease, a wheat-free diet, and leukemia from published literature. Single nucleotide polymorphisms (SNPs) representing celiac disease or a wheat-free diet were selected as instrumental variables. Two-sample Mendelian randomization (MR) analyses were conducted using inverse-variance weighted (IVW), weighted median, MR-Egger regression, simple mode, and weighted mode methods. The reliability of the results was evaluated using MR-Egger test, MR-PRESSO test, Cochran Q test, and leave-one-out test. Results We identified 16 SNPs associated with celiac disease as instrumental variables and 5 SNPs associated with a wheat-free diet as instrumental variables. MR analysis showed that an increase in celiac disease is associated with an increased risk of leukemia (OR = 1.03, 95% CI: 1.01-1.07, P = 0.03) according to the IVW method, and similar results were obtained using the MR-Egger regression (OR = 1.04, 95% CI: 1.00-1.09, P = 0.04). Consistent causal effects were observed using the five methods. The IVW method revealed that a wheat-free diet in healthy individuals leads to an increased risk of leukemia (OR = 1.06, 95% CI: 1.03-1.08, P = 2.36E-05), and similar results were obtained using the MR-Egger regression (OR = 1.06, 95% CI: 0.98-1.15, P = 0.27). The five methods consistently supported these causal directions. Reliability analysis indicated no heterogeneity, no pleiotropy, and no significant bias in both sets of MR analyses, suggesting the reliability of the results. Conclusion Celiac disease is a risk factor for leukemia, and a wheat-free diet in healthy individuals also increases the risk of developing leukemia. Adjusting one's own immune system may be the best choice for individuals with celiac disease.