Predictors in Youth of Adult Cardiovascular Events.

Abstract

BACKGROUND AND OBJECTIVES Childhood risk factors are associated with cardiovascular events in adulthood. We compared the utility of a risk model based solely on nonlaboratory risk factors in adolescence versus a model that additionally included lipids to predict cardiovascular events in adulthood. METHODS The study comprised 11 550 participants from 7 longitudinal cohort studies in the United States, Australia, and Finland with risk factor measurements in adolescence and followed into adulthood. The adolescent risk factors were defined by using clinical standards including overweight or obesity, elevated blood pressure, smoking, and borderline high or high levels of total cholesterol and triglycerides. The main outcomes were medically adjudicated fatal or nonfatal cardiovascular disease events occurring after age 25. RESULTS Of 11 550 participants (55.1% female, mean age 50.0 ± 7.7 years), 513 (4.4%) had confirmed cardiovascular events. In a multivariable model (hazard ratio [95% confidence interval]), elevated blood pressure (1.25 [1.03-1.52]), overweight (1.76 [1.42-2.18]), obesity (2.19 [1.62-2.98]), smoking (1.63 [1.37-1.95]), and high total cholesterol (1.79 [1.39-2.31]) were predictors of cardiovascular events (P < .05). The addition of lipids (total cholesterol and triglycerides) into the nonlaboratory model (age, sex, blood pressure, BMI, and smoking) did not improve discrimination in predicting cardiovascular events (C-statistics for the lipid model 0.75 [SD 0.07] and nonlaboratory model 0.75 [0.07], P = .82). CONCLUSIONS Nonlaboratory-based risk factors and lipids measured in adolescence independently predicted adult cardiovascular events. The addition of lipid measurements to nonlaboratory risk factors did not improve the prediction of cardiovascular events.