What drives labor market losses among diabetics—physiological impairment or psychosocial burden? New evidence from triangulation

Diabetes imposes substantial labor market costs, yet the underlying drivers remain poorly understood. This study investigates the causal impact of diabetes on individuals’ labor market outcomes and explores the underlying mechanisms involving both physiological impairment and psychosocial burden. Given the strong endogeneity between diabetes and individual behaviors, we employ a triangulation framework with nationally representative microdata from the UK Biobank (2006–2010; 502,364 participants) and the China Health and Retirement Longitudinal Study (2015; 21,095 participants). By linking genetic and labor market information and applying Ordinary Least Squares (OLS), one-sample Mendelian Randomization (1SMR), and two-sample Mendelian Randomization (2SMR), we estimate the causal effects of diabetes on labor market outcomes. While we find no significant effect of diabetes on labor force participation and only inconclusive evidence regarding labor income, diabetes is shown to increase the risk of unemployment by 14.96 %. This suggests that labor market losses among diabetics primarily reflect constraints on work capacity and opportunity, rather than reduced willingness to work. Leveraging the dual diagnostic criteria in the data—self-reported versus medically diagnosed diabetes—we further decompose the underlying mechanisms. The analysis reveals that approximately 40 % of the unemployment loss can be attributed to physiological impairment, while 60 % is driven by psychosocial burden. These findings underscore the potential of targeted psychosocial interventions to improve labor market outcomes for individuals with diabetes—even in the absence of improvements in their physical health.