Validation and Recalibration of PCE, China-PAR and PREVENT Models for Estimating ASCVD Risk in China (JACC Asia May 2026)-Article: Validation and Recalibration of PCE, China-PAR and PREVENT Models for Estimating ASCVD Risk in China

Pdf Summary

This study evaluated three 10-year ASCVD risk prediction models—the U.S.-derived PCE, the China-specific China-PAR, and the newer PREVENT equations—in a large contemporary Chinese cohort from the Kailuan study. Among 79,497 adults aged 40 to 79 years without ASCVD at baseline, 4,425 ASCVD events occurred over a median 10-year follow-up.<br /><br />All three original models showed similar discrimination in women (good performance, C-index about 0.74) and moderate discrimination in men (C-index about 0.68). However, their calibration differed substantially. The original PCE and China-PAR generally overestimated risk, especially in men and in higher-risk groups. PCE overestimated 10-year risk by 34.9% in men and 15.1% in women. China-PAR overestimated by 17.7% in men and 48.2% in women. In contrast, PREVENT underestimated risk in men by 29.4% but was close to accurate in women.<br /><br />The authors then recalibrated the models using local cohort data, splitting the sample into derivation and validation sets to avoid overfitting. Recalibration improved agreement between predicted and observed risks for all three models. Decision-curve analysis suggested that recalibrated PCE improved clinical net benefit in men, and recalibrated China-PAR improved net benefit in women. Recalibration of PREVENT improved calibration, though clinical benefit changes were less pronounced.<br /><br />They also tested whether adding triglyceride-glucose index (TyG) and high-sensitivity C-reactive protein (hsCRP) would improve prediction. These biomarkers were associated with ASCVD, but they did not meaningfully improve discrimination, although they improved calibration and decision-curve performance.<br /><br />Overall, the study concludes that original ASCVD risk scores may misestimate risk in Chinese adults, but simple recalibration to local population data can substantially improve their usefulness for primary prevention.

Meta Tag

Concept Cardiovascular Risk Prediction Model

Concept Pooled Cohort Equation

Concept China-PAR

Concept PREVENT

Concept External Validation

Concept Recalibration

Keywords

ASCVD risk prediction

PCE

China-PAR

PREVENT equations

Chinese cohort

risk calibration

risk overestimation

decision-curve analysis

triglyceride-glucose index

hsCRP

Cardiovascular Risk Prediction Model

Pooled Cohort Equation

China-PAR

PREVENT

External Validation

Recalibration