Women's Health Reports (Sep 2020)

Rural?Urban Residence and Stroke Risk and Severity in Postmenopausal Women: The Women's Health Initiative

  • Shawnita Sealy-Jefferson,
  • Molly Roseland,
  • Michele L. Cote,
  • Amy Lehman,
  • Eric A. Whitsel,
  • Jason Booza,
  • Michael S. Simon

DOI
https://doi.org/10.1089/WHR.2020.0034
Journal volume & issue
Vol. 1, no. 1
pp. 326 – 333

Abstract

Read online

Background: The impact of rural?urban residence on stroke risk and poor stroke outcomes among postmenopausal women is unknown. Methods: We used data from the Women's Health Initiative (WHI) (1993?2014; n?=?155,186) to test the hypothesis that women who live in rural compared with urban areas have higher stroke risk and worse stroke outcomes than urban women. We used rural?urban commuting area codes to categorize geocoded participant addresses into urban, large rural, or small rural areas. Incident strokes during follow-up were adjudicated by neurologists who used standardized criteria for reviewing brain imaging reports and other medical records and determining stroke subtype. Stroke functional recovery was measured with the Glasgow Stroke Outcomes Scale ascertained from the hospital record. We used univariable and multivariable-adjusted Cox proportional hazards models as well as logistic regression models to test whether rural?urban residence predicted stroke risk and odds of poor stroke outcome. Results: Among the 155,186 women in our cohort, 2.3% (n?=?3514) had an incident stroke. We observed a modest reduction in risk of incident stroke among women who lived in urban (adjusted hazard ratio [aHR]: 0.86, confidence interval [95% CI]: 0.71?1.05) and large rural areas (aHR: 0.79, 95% CI: 0.60?1.04) compared with women who lived in small rural areas. In contrast, women who lived in urban compared with large rural areas had a similarly modest increased risk of stroke (aHR: 1.09, 95% CI: 0.89?1.32). Women who lived in urban compared with large rural areas were more likely to have poor stroke outcome (odds ratio [OR]: 1.41, 95% CI: 1.06?1.88), but the association was attenuated after adjustment for covariates (adjusted OR [aOR]: 1.27, 0.93?1.74). Conclusions: Future studies should confirm and examine the potential pathways of the reported associations among postmenopausal women.

Keywords