Decomposing health inequality by dimensions and determinants in China using EQ-5D-5L: an integrated framework

The EQ-5D-5L is a widely used measure of health status that has been increasingly applied to study health inequalities. Despite growing recognition of health as a multidimensional construct, few decomposition studies have simultaneously considered both health dimension and their determinants as sources of inequality. This study fills this gap by providing the first integrated decomposition of socioeconomic-related health inequality in China using the EQ-5D-5L instrument. Using cross-sectional data from a nationally distributed sample of 29,882 Chinese adults, we applied the Erreygers-corrected concentration index (ECI) to measure socioeconomic inequality in EQ-5D-5L utility scores, with household income as the ranking variable. An integrated decomposition framework was developed to jointly quantify the contributions of the five EQ-5D-5L dimensions and their associated determinants to overall health inequality. This framework not only identifies the leading dimensions and determinants but also quantifies how determinants contribute to overall inequality through specific dimensions. The estimated ECI for overall EQ-5D-5L utility scores was 0.0275 (95% confidence interval: 0.0227, 0.0326), indicating modest but statistically significant pro-rich inequality in overall health. Pain/discomfort (PD) and mobility (MO) dimensions jointly accounted for more than 60% of overall inequality. Household income and chronic disease history were primary determinants, which contributed through all five dimensions. Sex–age composition contributed positively to overall inequality but showed an offsetting association with inequality in the anxiety/depression (AD) dimension. Negative life events contributed mainly through the AD and PD dimensions, whereas education was associated with reduced overall inequality via the AD dimension. High malnutrition risk contributed to overall inequality primarily through the MO, usual activities (UA), and SC dimensions, while poor sleep quality contributed predominantly through the AD and PD dimensions. These results were robust to using the Wagstaff-corrected concentration index and an alternative decomposition specification. The contributions of leading determinants to socioeconomic-related health inequality exhibited substantial heterogeneity across health dimensions. Our findings highlight the importance of moving beyond aggregate measures and suggest that interventions informed by the specific determinants and dimensions contributing most to overall inequality may help support more equitable health outcomes in China and other low- and middle-income countries.