The association between Life’s Crucial 9 and cardiometabolic multimorbidity among US older adults: Research based on machine learning

To explore the association between Life’s Crucial 9 (LC9) and cardiometabolic multimorbidity (CMM) in American adults aged ≥60 years, as well as key LC9 components. Using data from the National Health and National Health and Nutrition Examination Survey 2005 to 2018, 4913 participants were included. Weighted logistic regression, restricted cubic splines, subgroup analyses, and 4 machine learning models (e.g., XGBoost) were used. Model performance was evaluated using the area under the curve, and SHAP analysis was used to interpret the optimal model. Each 10-point increase in LC9 was associated with a 48% lower CMM risk (OR = 0.52, 95% CI: 0.47–0.58, P < .001), with a significant nonlinear relationship. XGBoost exhibited the best predictive performance (area under the curve = 0.829). Fasting blood glucose, serum creatinine (Scr), and the poverty income ratio were the top predictive factors. Higher LC9 scores are strongly and nonlinearly linked to reduced CMM risk in older adults. XGBoost combined with SHAP is a useful tool for CMM risk stratification, highlighting blood glucose level as a key factor for CMM based on LC9.