Estimated Personal Soot Exposure is Associated With MI Onset in a Case-Crossover-Study

Abstract

TM6-O-02 Background: An association between vehicular traffic exposure and the exacerbation of cardiovascular disease has been suggested in previous studies. We had shown earlier that exposure to traffic might trigger myocardial infarction (MI) within 1 hour afterward. The current case-crossover study combines personal information and estimated ambient soot concentration to better estimate individual exposure to traffic-related air pollution. Methods: Cases of MI were routinely identified and interviewed in a population-based MI-Registry, from February 1999 through December 2003. A total of 960 MI survivors for whom the date and time of the MI were known, who had completed the Registry's interview, provided complete information on activities that may have triggered the MI in a patient diary, and spent case and control periods within the study area were included. Ambient soot was not continuously measured during the study period. Therefore, a prediction model for daily soot concentrations was built, using reflectometer measurements from approximately 100 PM2.5 Harvard Impactor filters of 1999 to 2000 as outcome and 24-hour PM2.5, NO2, NO, and relative humidity as explanatory variables in a linear regression. The estimates obtained were validated applying them to hourly soot aethalometer measurements of 2004 to 2005. Indicators for hour of the day were added (R2 = 0.72). Hourly soot for the study period was then estimated applying this formula. Running 24-hour-means were calculated for each hour. Individual soot exposure was estimated using a linear combination of estimated 24-hour-mean ambient soot concentration, number of hours spent outdoors, and number of hours spent in traffic over the 24 preceding hours. Its association with MI onset was then estimated in a case-crossover analysis controlling for temperature using conditional logistic regression analyses. The 24 hour-period directly before onset was considered as case period and 24 hour-periods 24 and 48 before onset were defined as control periods. Results: Daily estimated soot concentrations ranged between 1.1 μg/m3 and 8 μg/m3, with a mean of 2 μg/m3. Case-crossover analyses of the association of personal soot exposure with MI resulted in an odds ratio (OR) of 1.22 per interquartile range (IQR) of 1 μg/m3 (95% CI, 1.05–1.43). In comparison the ambient estimated soot (OR, 1.01; CI, 0.87–1.17 per 1 μg/m3) or PM2.5 (OR, 1.00; CI, 0.85–1.17 per 8.5 μg/m3) was not significantly associated with MI onset. Conclusions: The results suggest that health effects of traffic-related air pollution may be better estimated taking personal activities into account.