Assessing health effects of air pollution

Exposure to elevated concentrations of ambient air pollutants causes adverse human health effects. Two modes or methods of study are generally relied on to quantify the relationships between pollutants and specific effects. These are: human clinical experiments and epidemiological (or community exposure) studies. (Animal toxicological studies can be used to indicate the existence of an effect, but not the rate of the effect in humans.) Each method has limitations as a basis for quantifying the level of adverse effects anticipated in a given human population as a result of exposure. Consequently, care must be taken in deciding which studies are appropriate for assessment of health impacts in a population. Some limitations are inherent in the method. Epidemiological studies, for example, depend on adequate community monitoring and the ability to associate a cohort with ambient data. Clinical studies often do not represent the complex mix of pollutants in the atmosphere. Consequently, construction of dose/exposure-response functions is challenging. Another common complication in quantifying expected health impacts of a pollutant mix is lack of adequate ambient monitoring data coupled with little or no knowledge of a population's time and activity profiles. This paper summarizes the difficulties inherent in constructing estimates of health effects in populations living in densely populated and polluted areas and suggests approaches to making initial limited estimates. An example of the level of effort required to conduct a sophisticated study, using relatively rich databases, is given based on work done in the Southern California region.