The iCARE-DM Model for Five-Year T2DM Risk Prediction in the Elderly Population from Chinese Routine Public Health Services — China, 2017–2024

• DOI: 10.46234/ccdcw2025.111
• Publication Date: 2025-05-09T10:32:02Z
• Authors: Xinyue Han; ; Xiaotao Zhou; Huifang Yang; Qiao Deng; Wanting Feng; Yilin Teng; Yanan Wang; Jialu Yang; Yan Liu; Min Xia; Ben Zhang; Shouling Wu; Tao Zhang; Jiayuan Li
• Keywords: Chinese population

Risk assessment for high-risk populations is critical preventing Type 2 Diabetes Mellitus (T2DM). Although China's public health services have continuously contributed to early grass-roots diagnosis of diabetes years, universally applicable tools identifying latent elderly urgently need account heterogeneity, robustness, and generalizability. Therefore, this study developed validated the integrated Chinese Adapted Evaluation (iCARE-DM) model individuals. The iCARE-DM was based on pooled effect estimates from a meta-analysis cohort studies that identified T2DM risk factors in East Asian three multicenter populations. Predictive performance evaluated using area under curve (AUC), sensitivity, specificity, accuracy, log-rank tests, compared with guideline-recommended (i.e., New Score, NCDRS) as well four machine learning (ML) models. achieved AUC values 0.741, 0.783, 0.766, outperforming NCDRS by at least 12%. best-performing ML comparable model, its varied significantly across (with range high 9%). Subgroup analyses confirmed consistent age, gender rural-urban groups. demonstrated higher accuracy than exhibited superior robustness generalizability provides robust, culturally adapted tool