Genetic liability to obesity and peptic ulcer disease: a Mendelian randomization study

Abstract

Background: Epidemiological evidence relating obesity to peptic ulcer disease (PUD) has been mixed. Here we sought to determine the causality in the association of obesity with PUD risk using the Mendelian randomization (MR) approach. Methods: This study was based on summary-level data for body mass index (BMI), waist-to-hip ratio (WHR), and PUD derived from large genome-wide association studies (GWASs). Single nucleotide polymorphisms significantly associated with BMI and WHR (P < 5 × 10–8) were leveraged as instrumental variables. Causal estimates were pooled using several meta-analysis methods. In addition, multivariable MR was employed to account for covariation between BMI and WHR, as well as to explore potential mediators. Results: Genetically predicted higher BMI has a causal effect on PUD, with an OR of 1.34 per SD increase in BMI (~ 4.8 kg/m2) (P = 9.72 × 10–16). Likewise, there was a 35% higher risk of PUD (P = 2.35 × 10–10) for each SD increase in WHR (0.09 ratio). Complementary analyses returned consistent results. Multivariable MR demonstrated that adjustment for WHR largely attenuated the BMI-PUD association. However, the causal association of WHR with PUD risk survived adjustment for BMI. Both the associations remained robust upon adjustment for several traditional risk factors. Replication analyses using different instrumental variants further strengthened the causal inference. Besides, we found no evidence for the causal association in the reverse analyses from PUD to BMI/WHR. Conclusions: This MR study revealed that obesity (notably abdominal obesity) is causally associated with higher PUD risk. Programs aimed at weight loss may represent therapeutic opportunities for PUD.