The iCARE-DM Model for Five-Year T2DM Risk Prediction in the Elderly Population from Chinese Routine Public Health...

Area of Science:

• Gerontology
• Public Health
• Epidemiology

Background:

• Type 2 Diabetes Mellitus (T2DM) risk assessment in high-risk elderly populations is crucial for prevention.
• Existing tools lack universal applicability, robustness, and generalizability for diverse Chinese elderly demographics.
• There is a need for culturally adapted, accurate risk assessment tools for Type 2 Diabetes Mellitus (T2DM) in China.

Purpose of the Study:

• To develop and validate the integrated Chinese Adapted Risk Evaluation for Diabetes Mellitus (iCARE-DM) model.
• To assess the iCARE-DM model's performance against existing risk scores and machine learning models.
• To ensure the iCARE-DM model's robustness and generalizability across different subgroups of elderly Chinese individuals.

Main Methods:

• Developed the iCARE-DM model using meta-analysis of East Asian cohort studies on T2DM risk factors.
• Validated the model in three independent multicenter Chinese populations.
• Evaluated predictive performance using AUC, sensitivity, specificity, accuracy, and log-rank tests, comparing with NCDRS and ML models.

Main Results:

• The iCARE-DM model achieved AUC values of 0.741-0.783, significantly outperforming the NCDRS model by over 12%.
• While machine learning models showed comparable AUC, their performance varied significantly across populations.
• Subgroup analyses confirmed consistent performance of iCARE-DM across age, gender, and rural-urban groups.

Conclusions:

• The iCARE-DM model demonstrates superior accuracy, robustness, and generalizability compared to the NCDRS and machine learning models.
• iCARE-DM provides a reliable, culturally adapted tool for T2DM risk assessment in elderly Chinese populations.
• This validated model supports targeted public health interventions for Type 2 Diabetes Mellitus (T2DM) prevention in China.