Comprehensive Summary
This study examined the use of machine learning (ML) models to predict and map cardiometabolic multimorbidity (CMM), the coexistence of two or more cardiovascular and metabolic diseases, in Lingwu City, Northwest China. Researchers analyzed data from 11,353 residents aged 35–75 years collected through surveys, physical examinations, and laboratory tests. Four ML models (Logistic Regression (LR), Support Vector Machine (SVM), Random Forest (RF), and XGBoost) were developed to identify risk factors and spatial trends in CMM prevalence. The analysis revealed that CMM prevalence was highest in the northeastern region of Lingwu City, with spatial mapping showing how disease patterns shifted northward between 2017 and 2021. Among the four algorithms, the Random Forest model achieved the highest accuracy, performing best in both test and real-world datasets. The strongest predictors of CMM were age and waist circumference, while education level, alcohol use, and marital status had weaker associations. The study underscores the promise of ML and geospatial analysis in identifying high-risk populations and guiding targeted public health interventions for chronic disease prevention.
Outcomes and Implications
The findings from Lingwu City reveal how machine learning and spatial analysis can guide public health planning by pinpointing geographic and socioeconomic disparities in cardiometabolic multimorbidity (CMM). Identifying high-risk clusters, particularly in the city’s northeastern regions, enables policymakers to target prevention and health education programs more effectively. The association between higher socioeconomic status and CMM risk highlights the complex interplay between lifestyle, urbanization, and chronic disease. Applying such models across provinces could strengthen China’s chronic disease surveillance efforts, support equitable resource distribution, and inform early interventions aimed at reducing CMM prevalence before disease onset.