ALTERNATIVE HYPOTHESES LINKING OUTDOOR PARTICULATE MATTER WITH DAILY MORBIDITY AND MORTALITY

Abstract

Numerous epidemiologic studies have reported associations between fluctuations in ambient particulate matter (PM) and small changes in daily mortality and morbidity. However, it has not been possible to causally link outdoor PM with changes in daily mortality/morbidity because of, among other things, problems relating to the exposuredose relationship. Alternative explanations are possible, but remain untested. Varying levels of ambient PM may be correlated to factors that modify mortality/morbidity by alternative pathways. We identify three such pathways: (1) weather conditions directly affecting health, (2) human behavior patterns directly affecting health, and (3) weather conditions and patterns of human behavior affecting exposure to indoor air contaminants, which directly affect health. These three pathways also are related to day-to-day fluctuations in ambient PM. Thus, PM levels may only be a surrogate for other, causal factors. For example, unpleasant weather or poor outdoor air quality causes people to spend more time indoors, and also promotes the use of climate-control systems, which can generate increased indoor levels of potentially toxic airborne particles. At the same time, weather extremes tend to elevate outdoor PM levels because of increased use of vehicles (instead of walking or cycling) and increased combustion emissions (for providing electric power for heating or air conditioning). Therefore, personal exposure to indoor contaminants may be linked to outdoor PM concentrations. The major significance of indoor air contaminants is that, unlike outdoor PM, toxicological and epidemiological studies of indoor air support biologic plausibility. Our analysis supports the importance of considering the factors that drive outdoor PM fluctuations in the first place to determine if the observed mortality/morbidity changes may be caused by pathways other than the hypothetical frank toxicity from inhalation of ambient PM.