Abstract
Nonalcoholic fatty liver disease (NAFLD) is a chronic liver condition with a steadily increasing prevalence. Evidence indicates that both smoking and obesity are significant risk factors for NAFLD, yet the extent to which smoking influences NAFLD through weight gain remains unclear. This study aimed to dissect the intricate relationship between smoking, body mass index (BMI), and NAFLD using Mendelian randomization (MR) analysis. We leveraged data from 30 genome-wide association studies involving over 1.2 million individuals, from which 123 single nucleotide polymorphisms were selected as instrumental variables for smoking. BMI data were sourced from the Genetic Investigation of Anthropometric Traits (GIANT) consortium, encompassing more than 700,000 individuals, with 521 single nucleotide polymorphisms serving as instrumental variables. NAFLD data were obtained from multiple databases, including the eMERGE Network, UK Biobank, Estonian Biobank, and FinnGen, comprising 8434 cases and 770,180 controls. All participants in this study were of European ancestry. We first applied univariate MR analysis to assess the causal relationship between smoking, NAFLD, and BMI. Subsequently, multivariate MR was used to assess the effect of smoking on NAFLD after adjusting for BMI. The coefficient product method was used to calculate the mediating effect of BMI. Results found that both smoking and high BMI were able to increase the risk of NAFLD, with odds ratios of 1.83 (95% confidence interval [CI]: 1.31-2.55) and 1.58 (95% CI: 1.42-1.77), respectively. BMI mediated 73.3% (95% CI: 62.3%-80.5%) of the effect of smoking on NAFLD. The findings support weight control and the encouragement of smoking cessation, especially in obese populations, as strategies to reduce the risk of NAFLD.