Association between predicted skeletal muscle index and prevalence and incidence of chronic diseases: evidence from the national charls study

Skeletal muscle is essential for metabolic regulation and healthy aging. The predicted skeletal muscle index (pSMI) is an equation-derived composite proxy, but its associations with chronic diseases remain unclear in Chinese adults. This study included 6,676 adults aged ≥ 45 years using data from the China Health and Retirement Longitudinal Study (CHARLS), a nationally representative cohort. Cross-sectional associations between the pSMI and 14 chronic diseases were assessed using multivariable logistic regression, whereas longitudinal associations with incident diseases were evaluated through disease-specific Cox models. Restricted cubic spline and mediation analyses were used to explore nonlinear associations and a potential mediating role of the triglyceride–glucose (TyG) index. Sensitivity and subgroup analyses were conducted to assess robustness. At baseline, higher pSMI was associated with greater odds of prevalent hypertension, diabetes, dyslipidemia, and heart disease, and with lower odds of prevalent lung and memory diseases. Restricted cubic spline analyses indicated nonlinear associations for hypertension and diabetes, whereas other associations were approximately linear. Mediation analyses suggested that the TyG index statistically mediated associations between pSMI and several diseases, except memory disorders; these analyses do not establish a causal pathway. During longitudinal follow-up, higher baseline pSMI was associated with higher incidence of hypertension, diabetes, dyslipidemia, and liver disease, but lower incidence of lung disease, asthma, and memory disease. The pSMI exhibits disease-specific associations across chronic disease outcomes, highlighting its potential utility as an accessible proxy for epidemiologic risk stratification. Not applicable.