Abstract
OBJECTIVE: To describe and compare the determinants of 1-year mortality after premature vs non-premature acute coronary syndrome (ACS). PATIENTS AND METHODS: Participants presenting with ACS were enrolled in a prospective registry of 29 hospitals in 4 countries, from January 22, 2012 to January 22, 2013, with 1-year of follow-up data. The primary outcome was all-cause 1-year mortality after premature ACS (men aged <55 years and women aged <65 years) and non-premature ACS (men aged ≥55 years and women aged ≥65 years). The associations between the baseline patient characteristics and 1-year mortality were analyzed in models adjusting for the Global Registry of Acute Coronary Events (GRACE) score and reported as adjusted odds ratio (aOR) (95% CI). RESULTS: Of the 3868 patients, 43.3% presented with premature ACS that was associated with lower 1-year mortality (5.7%) than those with non-premature ACS. In adjusted models, women experienced higher mortality than men after premature (aOR, 2.14 [1.37-3.41]) vs non-premature ACS (aOR, 1.28 [0.99-1.65]) (P (interaction)=.047). Patients lacking formal education vs any education had higher mortality after both premature (aOR, 2.92 [1.87-4.61]) and non-premature ACS (aOR, 1.78 [1.36-2.34]) (P (interaction)=.06). Lack of employment vs any employment was associated with approximately 3-fold higher mortality after premature and non-premature ACS (P (interaction)=.72). Using stepwise logistic regression to predict 1-year mortality, a model with GRACE risk score and 4 characteristics (education, employment, body mass index [kg/m(2)], and statin use within 24 hours after admission) had higher discrimination than the GRACE risk score alone (area under the curve, 0.800 vs 0.773; P (comparison)=.003). CONCLUSION: In this study, women, compared with men, had higher 1-year mortality after premature ACS. The social determinants of health (no formal education or employment) were strongly associated with higher 1-year mortality after premature and non-premature ACS, improved mortality prediction, and should be routinely considered in risk assessment after ACS.