Estimating the causal effect of cardiometabolic conditions on socioeconomic and healthcare outcomes: a scoping review of Mendelian randomization studies.

backgroundCardiometabolic risk factors and conditions are the leading contributors to morbidity and mortality, yet quantifying their causal effects on socioeconomic outcomes using observational data is challenging due to endogeneity. Using genetic variants as instrumental variables, Mendelian randomization (MR) offers a unique approach to strengthen causal inference in this context and has also gained popularity in health economic literature.

aimsThis study aimed to: i) map the current landscape of MR studies evaluating the impact of cardiometabolic exposures on healthcare and socioeconomic outcomes; ii) describe how core MR assumptions were tested and reported; iii) summarize how additional assumptions underlying causal interpretation were discussed.

methodsWe searched MEDLINE and EMBASE for studies applying MR to examine the impact of cardiometabolic risk factors or conditions (e.g., obesity, blood pressure, cholesterol, coronary artery disease, type 2 diabetes) on socioeconomic and healthcare outcomes (e.g., education, income, occupational status, social deprivation, healthcare use and costs, health-related quality of life). Study characteristics, MR design choices, and reporting of assumption testing and causal interpretation were extracted and narratively summarized.

resultsSixteen studies were included, covering 79 exposure-outcome pairs. Most studies examined the effects of body mass index on employment or healthcare costs. Only one study assessed home ownership, social income transfers, resource utilization, and quality-adjusted life years as outcomes, respectively. Effects of childhood cardiometabolic exposures were rarely examined beyond educational outcomes. UK Biobank was the predominant data source. None of the core MR assumptions were mentioned across all studies. While weak instrument bias was frequently tested, less than 40% of studies assessed associations between instruments and observable confounders as falsification tests. Only few studies discussed monotonicity or homogeneity assumptions.

conclusionsAlthough MR is a promising identification strategy for assessing causal effects of cardiometabolic risk on healthcare and socioeconomic outcomes, reporting practices for assumption testing and causal interpretation vary widely. This review highlights opportunities to strengthen transparency and coherence in future MR applications. With increasing data availability and clearer methodological guidance, MR could complement conventional observational approaches in supporting policy decisions.