Spatial variations and risk factors of multimorbidity in China: A population-based spatial modelling study

Multimorbidity burden is likely to vary across China, but relevant evidence is insufficient. The extent to which individual and provincial factors may affect spatial variations of multimorbidity has not been fully examined. This study aims to estimate the provincial multimorbidity burden among Chinese aged 45 and older, identifying which risk factors remain constant and which vary across China.

This study included 18,561 adults aged 45 and older from the China Health and Retirement Longitudinal Study in 2020. A Bayesian spatial varying coefficients model was adopted to estimate the multimorbidity burden and 95% Bayesian credible intervals, using the Chinese Multimorbidity-Weighted Index (CMWI) as a measurement. Partial correlation coefficients between covariates and CMWI in each region were calculated to investigate the need for varying coefficients. Spatial autocorrelation analyses were used to identify clusters of high and low multimorbidity burden.

The estimated CMWI across the 27 provinces in China ranged from 1.76 (95% BCI: 1.64, 1.89) to 4.42 (95% BCI: 4.16, 4.70). High multimorbidity burden areas were clustered in North and Northeast China, while areas with relatively low burden were in southern China. The top three provinces by median CMWI estimates were Neimenggu, Heilongjiang, and Jilin, whereas Guangdong, Zhejiang, and Beijing were among the lowest CMWI estimates. The effect of age and sex showed spatial variation across China, while other risk factors showed fixed effects.

The burden of multimorbidity varies across China and not all risk factors associated with multimorbidity are consistent across regions, providing valuable insights for chronic disease management.