Long-Term Exposure to Air Pollution and Survival After Ischemic Stroke

Abstract

Background and Purpose- China bears a heavy burden of stroke because of its large population of elderly people and the propensity for stroke. Previous studies have examined the association between air pollution and stroke mortality or hospital admission. However, the global evidence for adverse effects of air pollution on survival after stroke is scarce. Methods- We used the first national hospital-based prospective registry cohort of stroke in China, which included 12 291 ischemic stroke patients who visited hospitals during 2007 to 2008. All patients were followed for 1-year poststroke. Deaths during the follow-up period were recorded. Participants' 3-year prestroke exposures to ambient PM₁, PM_2.5, PM₁₀ (particulate matter with aerodynamic diameters ≤1, ≤2.5, and ≤10 μm, respectively) and NO₂ (nitrogen dioxide) were estimated by machine learning algorithms with satellite remote sensing, land use information, and meteorological data. Cox proportional hazards models were used to examine the association between air pollution and survival after ischemic stroke. Results- In total, 1649 deaths were identified during the 1-year follow-up period. After controlling for potential confounders, significant associations were observed between exposure to PM₁ and PM_2.5 and incident fatal ischemic stroke. The corresponding hazard ratios and 95% CIs associated with 10 µg/m³ increase in PM₁ and PM_2.5 were 1.05 (1.02-1.09) and 1.03 (1.00-1.06), respectively. No significant association was observed for PM₁₀ or NO₂ (hazard ratios and 95% CIs, 1.01 [1.00-1.03] and 1.03 [0.99-1.06], respectively). Higher hazard ratios (and 95% CIs) were observed for male, elderly and obese individuals. Conclusions- Prestroke exposure to PM₁ and PM_2.5 was associated with increased incident fatal ischemic stroke in the year following an ischemic stroke in China. Improved air quality may be beneficial for people to recover from stroke.

Keywords: air pollution; machine learning; nitrogen dioxide; particulate matter; proportional hazards models.

Figure 1.

Locations of 12,291 ischaemic stroke patients included in this study.

Figure 2.

Hazard ratios (and 95% confidence intervals) of death from ischaemic stroke associated with per 10 μg/m³ increase in mean exposures to air pollutants during three years before stroke. Note: Model 1 included mean level of one pollutant and a smooth term of the interaction of longitude and latitude. Model 2 included all variables in the Model 1 and added age, sex and BMI. Model 3 included all variables shown in Table 1.