Waist-to-hip ratio and nonalcoholic fatty liver disease: a clinical observational and Mendelian randomization analysis

Abstract

Background: Obesity often coincides with non-alcoholic fatty liver disease (NAFLD), yet a significant portion of NAFLD patients exhibit normal body mass index (BMI) but have abdominal obesity. Recognizing this discrepancy, we aimed to delve deeper into this phenomenon through observational studies coupled with two-sample Mendelian randomization (MR) analysis, with waist-to-hip ratio (WHR) serving as the indicator for abdominal obesity. Our objective was to ascertain whether WHR correlates with an increased risk of NAFLD development.

Methods: This study utilized data from the National Health and Nutrition Examination Survey (NHANES) 2017-2018 to examine the association between WHR and NAFLD through weighted multivariate logistic regression models. On this basis, subgroup analyses were performed to further explore the correlation between WHR and NAFLD. Subsequently, a two-sample MR analysis was conducted using genome-wide association studies (GWAS) data to investigate the potential causal relationship between WHR and NAFLD. Sensitivity analyses were also employed to ensure the robustness of our findings.

Results: A total of 3,732 eligible participants were included in the analysis. Weighted multivariable-adjusted logistic regression models revealed a positive association between WHR and the risk of NAFLD (Q2vsQ1: OR = 1.94 [95% CI: 1.55-2.44]; Q3vsQ1: OR = 2.08 [95% CI: 1.51-2.85]; Q4vsQ1: OR = 3.70 [95% CI: 2.13-6.43], p < 0.05). The results of the subgroup analysis suggested that there was an interaction in the correlation between WHR and NAFLD in normal weight, overweight, and obese populations (p < 0.05). The RCS curves indicated that there was a nonlinear relationship between WHR and NAFLD in populations with BMI in the normal versus obese categories. Furthermore, MR analysis provided additional support for the causal relationship between WHR and NAFLD. Using inverse variance weighting (IVW), the MR analysis yielded an OR of 2.062 (95% CI: 1.680-2.531, p<0.05). Consistent results were obtained with the other four MR methods, all supporting the same direction of causality. Sensitivity analyses were performed to assess the robustness of the findings (p > 0.5), further reinforcing the reliability of the observed associations.

Conclusion: WHR elevation heightens the susceptibility to NAFLD.

Keywords: Mendelian randomization analysis; NHANES; clinical observational study; non-alcoholic fatty liver disease; waist-to-hip ratio.