A time-series study of the health effects of water-soluble and total-extractable metal content of airborne particulate matter

Abstract

Objectives: To assess whether adverse acute cardiopulmonary health outcomes are associated with concentration of trace metals in airborne particulate matter. Methods: Daily PM10 and PM2.5 was collected for one year in Edinburgh, UK, and the water-soluble and total-extractable content of 11 trace metals determined in each sample. Time series were analysed using generalised additive Poisson regression models, including adjustment for minimum temperature and loess smoothing of trends. Methods were explored of extending the time series of metal concentration in daily PM10 for the previous 7 years using multiple regression of the variation in metal content for the 1 year of measurements and the associated variation in air mass source region and other concurrently-measured potential predictor variables. Results: The 1 year of direct measurements showed no evidence of significant associations of particle-bound metal concentration with health outcomes beyond that expected by chance. Analysis of the extended time series showed significant positive associations with cardiovascular admissions both for total PM10 and for a number of the metals (e.g. Cu, Fe, Ni, V, Zn) but the metal effects were no longer significant after adjusting for PM10. Conclusions: Within the limitations of its power, the epidemiological results have not provided evidence for associations between particle-bound metal concentrations and adverse health outcomes that are substantially greater than for total PM. The generally strong correlations between metal and total PM suggest that quantifying independent effects of PM metal exposure on health will be difficult, even using more powerful time series of direct measurements.