Ambient Particulate Matter (PM) and Stroke: a Review and Meta-analysis

摘要

Background: Short-term exposure to ambient particulate matter (PM) may be associated with increased risk of mortality and hospital admission for total cerebrovascular disease (CBVD), ischemic stroke (IS) or hemorrhagic stroke (HS), but the results have been inconsistent. Aims: To summarize the evidence of an association between short-term exposure to PM and mortality and hospital admission for CBVD, IS or HS. Methods: We conducted a systematic search of relevant epidemiological studies indexed by PubMed, Web of Knowledge, or Embase in English before February 2012.Two reviewers independently extracted key characteristics of each study and effect estimates for single-day lags up to 4 days and cumulative lags up to 4 days prior to the day of event. We used a random-effects model to estimate a summary incidence rate ratio (IRR) per a 10 μg/m3 increase in PM. We assessed publication bias with funnel plots and Egger's asymmetry test. Results: We identified 32 studies from 480 relevant articles. CBVD mortality was positively associated with both PM10 and PM2.5 with summary IRR (95% CI) of 1.005 (1.003,1.008) and 1.014 (1.008, 1.020), respectively. Rates of CBVD hospital admissions were also positively associated with PM10 and PM2.5, with summary IRR of 1.003 (0.997,1.008) and 1.014 (1.004,1.012), respectively. One studies suggested there was evidence of an association between IS mortality and PM2.5 and no study has studied the association between HS mortality and both PM10 and PM2.5. Results across studies were less consistent for the association between IS hospital admissions and both PM10 and PM2.5, with the summary IRR including the null hypothesis of no association. There was little evidence of an association between HS hospital admissions and either PM10 or PM2.5.There has no apparent publication bias in meta-analyses. Conclusions: There is evidence of an association between PM and both CBVD mortality and hospital admission, but less consistency among studies specifically of IS and HS hospital admission.