Abstract
OBJECTIVE: With advances in medical diagnosis, more people are diagnosed with more than one disease. The damage caused by different diseases varies, so relying solely on the number of diseases to represent multimorbidity is limited. The Charlson comorbidity index (CCI) is widely used to measure multimorbidity and has been validated in various studies. However, CCI's demographic and behavioral risk factors still need more exploration. METHODS: We conduct multivariate logistic regression analysis and restricted cubic splines to examine the influence factors of CCI and the relationship between covariates and risk of CCI, respectively. Our research employs the Multivariate Imputation by Chained Equations method to interpolate missing values. In addition, the CCI score for each participant is calculated based on the inpatient's condition using the International Classification of Diseases, edition 10 (ICD10). Considering the differences in the disease burden between males and females, the research was finally subgroup analyzed by sex. RESULTS: This study includes 5,02,411 participants (2,29,086 female) with CCI scores ranging from 0 to 98. All covariates differed between CCI groups. High waist-hip ratio (WHR) increases the risk of CCI in both males [OR = 19.439, 95% CI = (16.261, 23.241)] and females [OR = 12.575, 95% CI = (11.005, 14.370)], and the effect of WHR on CCI is more significant in males. Associations between age, Body Mass Index (BMI) and WHR, and CCI risk are J-shaped for all participants, males, and females. Concerning the association between Townsend deprivation index (TDI) and CCI risk, the U-shape was found in all participants and males and varied to a greater extent in males, but it is a J-shape in females. CONCLUSIONS: Increased WHR, BMI, and TDI are significant predictors of poor health, and WHR showed a greater role. The impact of deprivation indices on health showed differences by sex. Socio-economic factors, such as income and TDI, are associated with CCI. The association of social status differences caused by these socioeconomic factors with health conditions should be considered. Factors might interact with each other; therefore, a comprehensive, rational, and robust intervention will be necessary for health.