Sex and racial/ethnic associations with post-stroke cognitive decline: longitudinal analyses from the health and retirement study

Background and Aims Evidence on sex and race/ethnic disparities in post-stroke cognitive decline have largely been inconclusive. We investigate these disparities using a series of quasi-experimental models. Methods Using the Health and Retirement Study (HRS), a biennial survey of a nationally representative sample of community dwellers in the US aged 51 years and older, we applied the two-way fixed effects model, event study design, and Bacon’s weighted difference-in-differences (DiD) estimation for survey panel data (1996 − 2016), to evaluate the differences in the modified Telephone Interview for Cognitive Status (TICS-m) score (lower scores representing lower cognitive functioning) between patients with and without self-reported incident stroke, while controlling for observed and unobserved time-invariant baseline participant characteristics such as age and education, as well as time-specific confounders. Results Of the 35,451 HRS participants that met the inclusion criteria (Figure.1), 4,620 (13.03%) had incident stroke. The mean (SD) TICS-m score post vs pre incident stroke was 12.8 (0.02) vs. 15.5 (0.07). The two-way fixed effects estimator for the relationship between incident stroke and cognitive decline was −1.21(95% CI:−1.32,−1.10), which was similar to the weighted DiD estimate of −1.25(95% CI:−1.38,−1.13). Event study revealed a continuous decline in TICS-m score, up to 10 years after stroke incidence in the overall sample, as well as among females and white non-Hispanics (Figure.1). However, the decline in TICS-m score plateaued from the third wave post-stroke for males and second wave post-stroke for black non-Hispanics. Conclusions Sex and race differences may illuminate variations in cognitive decline associated with stroke.