Carbon monoxide and stroke: A time series study of ambient air pollution and emergency hospitalizations

Recent experimental and clinical studies suggested that exogenous carbon monoxide (CO) at low concentrations may have beneficial neuroprotective effects under certain circumstances. However, population-based epidemiological studies of environmentally relevant CO exposure generated mixed findings. The present study aimed to examine the short-term association of ambient CO with emergency stroke hospitalizations.

A time series study was conducted. Daily air pollution concentrations and emergency hospital admission data from January 2004 to December 2011 in Hong Kong were collected. Generalized additive Poisson models were used to estimate the associations between daily 24-hour mean concentrations of CO and emergency hospital admissions for stroke, while controlling for other traffic related co-pollutants: NO₂ and PM₂.₅. Sensitivity analyses were performed using daily 1-hour maximum concentration of CO as exposure indicator.

Negative associations were observed between ambient CO concentrations and emergency hospital admissions for stroke. The previous 1-3 day cumulative exposure to CO was associated with a -2.0% (95%CI, -3.3% to -0.7%) decrease in stroke admissions per interquartile range (IQR) increment in CO concentration (0.3 ppm). Similar results were obtained when using 1-hour maximum concentration of CO as exposure indicator. The negative association was robust to the co-pollutant adjustment for either NO₂ or PM₂.₅. Females and elders appeared to be more sensitive to ambient CO exposure. The negative association tended to be larger in cool season.

Short-term exposure to ambient CO was associated with decreased risk of emergency hospitalizations for stroke, suggesting some acute protective effects of CO exposure against stroke onsets.