Replication Data for: Competing-Risk-Aware Cardiovascular Risk Estimation in Older Chinese Adults: A 9-Year Longitudinal Cohort

This repository provides code, derived data, and tabulated outputs for competing-risk analysis and cardiovascular risk prediction in older Chinese adults using the China Health and Retirement Longitudinal Study (CHARLS, 2011-2020). The package is intended for reproducibility, secondary methodological comparison, and data-sharing purposes.

The materials include:

derived participant-level analytical data for 9,551 CVD-free adults;
stepwise Python scripts for data preparation, survival modeling, hazard-ratio estimation, calibration assessment, restricted cubic spline analysis, competing-risk evaluation, and sensitivity analyses; and
result tables summarizing baseline characteristics, hazard ratios, feature importance, calibration, subgroup performance, and competing-risk estimates.
The analytical framework compares conventional Kaplan-Meier estimation with Aalen-Johansen cumulative incidence estimation under competing mortality, and includes machine-learning and regression-based survival modeling workflows that can be adapted for related cohort studies.

Raw CHARLS data are not redistributed in this repository. Researchers can obtain access to the original source data from the CHARLS Data Portal: https://charls.charlsdata.com/