Abstract
BACKGROUND: Young adults may have high long-term atherosclerotic cardiovascular disease (ASCVD) risk despite low short-term risk. OBJECTIVES: In this study, we sought to compare the performance of short-term and long-term ASCVD risk prediction tools in young adults and evaluate ASCVD incidence associated with predicted short-term and long-term risk. METHODS: We included adults aged 18 to 39 years, from 2008 to 2009 in a U.S. integrated health care system, and followed them through 2019. We calculated 10-year and 30-year ASCVD predicted risk and assessed ASCVD incidence. RESULTS: Among 414,260 young adults, 813 had an incident ASCVD event during a median of 4 years (maximum 11 years). Compared with 10-year predicted risk, 30-year predicted risk improved reclassification (net reclassification index: 16%) despite having similar discrimination (Harrell's C: 0.749 vs 0.726). Overall, 1.0% and 2.2% of young adults were categorized as having elevated 10-year (≥7.5%) and elevated 30-year (≥20%) predicted risk, respectively, and 1.6% as having low 10-year (<7.5%) but elevated 30-year predicted risk. The ASCVD incidence rate per 1,000 person-years was 2.60 (95% CI: 1.92-3.52) for those with elevated 10-year predicted risk, 1.87 (95% CI: 1.42-2.46) for those with low 10-year but elevated 30-year predicted risk, and 0.32 (95% CI: 0.30-0.35) for those with low 10-year and 30-year predicted risk. The age- and sex-adjusted incidence rate ratio was 3.04 (95% CI: 2.25-4.10) comparing those with low 10-year but elevated 30-year predicted risk and those with low 10-year and 30-year predicted risk. CONCLUSIONS: Long-term ASCVD risk prediction tools further discriminate a subgroup of young adults with elevated observed risk despite low estimated short-term risk.