Abstract
BACKGROUND: Dietary choices are inextricably linked to the incidence of cardiovascular disease (CVD), whereas an optimal dietary pattern to minimize CVD morbidity in high-risk subjects remains challenging. METHODS: We comprehensively assessed the relationship between food consumption frequencies and CVD in 28,979 high-risk subjects. The outcome was defined as the composite of the incidence of major CVD events, including coronary heart disease and stroke. Risk factors associated with CVD were screened through a shrinkage approach, specifically least absolute shrinkage and selection operator (LASSO) regression. Hazard ratios (HRs) for various dietary consumption frequencies were assessed using multivariable Cox frailty models with random intercepts. RESULTS: Increased egg and seafood consumption were associated with a lower risk of CVD (daily vs little, HR 1.70, 95% confidence interval, CI: 0.79-3.64, p (trend) = 0.0073 and HR 1.86, 95% CI: 1.24-2.81, p (trend) = 0.024, respectively). 6 non-food (age, sex, smoke, location, heart ratio, and systolic blood pressure) and 3 food (fruit, egg, and seafood) related risk factors were included in the nomogram to predict 3 and 5-year incidence of CVD. The concordance indexes of the training and validation cohorts were 0.733 (95% CI: 0.725-0.741) and 0.705 (95% CI: 0.693-0.717), respectively. The nomogram was validated using the calibration and time-dependent receiver operating characteristic curves, demonstrating respectable accuracy and discrimination. CONCLUSIONS: Guided by the concept of "food as medicine", this nomogram could provide dietary guidance and prognostic prediction for high cardiac risk subjects in CVD prevention.