Mediation and Interaction Analysis of Ischemic Heart Disease and Depression in the Causal Effect of Central Obesity on Health Status Based on Inverse Probability Weighting

Objective To investigate the direct causal effect between central obesity and health status, as well as the mediating and interactive effects of ischemic heart disease and depression in this causal relationship. Methods A total of 58,234 participants aged 45–65 years from the UK Biobank cohort between 2006 and 2010 were included and classified into a central obesity group and a control group according to central obesity status. And the health status was expressed as health status score. The main direct causal effect between central obesity and health status was explored through a causal effect analysis, while multi-mediation analysis approach based on inverse probability weighting was used to adjust for sleep disturbance influenced by central obesity and to examine the mediating and interactive effects between them. The mediating and interactive effects comes mainly from two binary mediate variables, ischemic heart disease and depression. The E-value method was used to assess the potential influence of unmeasured confounders. And subgroup analyses were further conducted by gender. Results The inverse probability weights for exposure balanced the baseline covariates between central obesity groups, while within each exposure group, the inverse probability weights for mediators further balanced the covariate distributions between ischemic heart disease and depression groups, including sleep disturbance influenced by central obesity, thereby effectively controlling for confounding factors. In the causal effect estimation, the direct effect between central obesity and health status was statistically significant ( P  < 0.001), with a pure direct effect of 0.53 (95% CI: 0.50–0.57). Ischemic heart disease and depression partially mediated and interacted in the relationship between central obesity and health status statistically significant, with a total indirect effect of 0.04 (95% CI: 0.03–0.05), a reference interaction effect of 0.01 (95% CI: 0.01–0.02), and a mediated interaction effect of 0.01 (95% CI: 0–0.02). Subgroup analysis results indicate that direct and indirect effects differ across genders. There is no interaction between central obesity in men and ischemic heart disease or depression. Sensitivity analysis indicates that the pure direct effect of central obesity on health status is relatively robust to unmeasured confounding factors, with greater robustness observed in women. The total indirect effect, however, exhibits higher sensitivity and is susceptible to unmeasured confounding factors in both men and women. Conclusion The adverse impact of central obesity on health status is partly mediated through ischemic heart disease and depression, but its direct causal effect remains predominant. When formulating intervention strategies, priority should be given to controlling central obesity itself, while concurrently managing its mediating and interacting pathways, and fully considering gender differences.