Trends in Stroke Recurrence in Mexican Americans and Non-Hispanic Whites

Abstract

Background and Purpose: Stroke incidence and mortality are declining rapidly in developed countries. Little data on ethnic-specific stroke recurrence trends exist. Fourteen-year stroke recurrence trend estimates were evaluated in Mexican Americans and non-Hispanic whites in a population-based study. Methods: Recurrent stroke was ascertained prospectively in the population-based BASIC (Brain Attack Surveillance in Corpus Christi) project in Texas, between 2000 and 2013. Incident cases were followed forward to determine 1- and 2-year recurrence. Fine & Gray subdistribution hazard models were used to estimate adjusted trends in the absolute recurrence risk and ethnic differences in the secular trends. The ethnic difference in the secular trend was examined using an interaction term between index year and ethnicity in the models adjusted for age, sex, hypertension, diabetes mellitus, smoking, atrial fibrillation, insurance, and cholesterol and relevant interaction terms. Results: From January 1, 2000 to December 31, 2013 (N=3571), the cumulative incidence of 1-year recurrence in Mexican Americans decreased from 9.26% (95% CI, 6.9%–12.43%) in 2000 to 3.42% (95% CI, 2.25%–5.21%) in 2013. Among non-Hispanic whites, the cumulative incidence of 1-year recurrence in non-Hispanic whites decreased from 5.67% (95% CI, 3.74%–8.62%) in 2000 to 3.59% (95% CI, 2.27%–5.68%) in 2013. The significant ethnic disparity in stroke recurrence existed in 2000 (risk difference, 3.59% [95% CI, 0.94%–6.22%]) but was no longer seen by 2013 (risk difference, −0.17% [95% CI, −1.96% to 1.5%]). The competing 1-year mortality risk was stable over time among Mexican Americans, while for non-Hispanic whites it was decreasing over time (difference between 2000 and 2013: −4.67% [95% CI, −8.72% to −0.75%]). Conclusions: Mexican Americans had significant reductions in stroke recurrence despite a stable death rate, a promising indicator. The ethnic disparity in stroke recurrence present early in the study was gone by 2013.