taminants pose a risk to the exposed population cannot be determined in the absence of more detailed information about human exposures. Instead of focusing on the toxic chemicals that have been identified at a site itself, it is necessary to develop estimates of exposure to define and assess the population at risk, including estimation of the population size and exposure-related characteristics.
Although few studies directly assess airborne exposures to hazardous wastes, the committee finds persuasive evidence that health effects can occur from such exposures. Review of the relevant animal literature on compounds known to occur at hazardous-waste sites, along with those few epidemiologic studies of airborne exposures from sites, shows that a wide range of effects may occur, and they include serious diseases, such as cancer, neurobehavioral complaints, and constellations of self-reported symptoms. In addition, a few studies have explicitly linked airborne exposures from hazardous-waste sites to increased rates of birth defects, low birth weight, and chronic diseases in some small populations.
Lessons that have been learned in air pollution studies are relevant to epidemiologic studies of hazardous-waste sites. As with the assessment of other health effects, a crucial lesson from the recent history of epidemiologic studies of air pollution has been the critical role played by the general availability of monitoring data for criteria air pollutants. Without the extensive network of data on those pollutants for which monitoring data are routinely acquired, such as respirable particulates, epidemiologic studies of air pollution would not have been possible. Many of the pollutants of interest at hazardous-waste sites are not routinely evaluated. Exposure to hydrocarbons in urban air has not been monitored regularly since the 1970s, and there has been little work on their direct effects except in studies of the sick building syndrome and occupational studies. The committee understands that routine monitoring of ambient air around hazardous-waste sites is not generally feasible because of the number of sites, the low likelihood of detection in most cases, and the cost of monitoring. The small populations near most sites make the sites difficult to study with standard epidemiologic techniques. Nevertheless, more systematic determinations of where such specific monitoring and studies might be appropriate need to be made early in the process of identifying and describing sites for study. The methods of meta-analysis also could prove useful to the extent that examples of similar exposures at different sites could be combined.