Setting ambient air quality standards for particulate matter.

Ambient air particulate matter (PM), unspecified as to chemical composition, is of concern because of its health effects. Air quality standards for PM have been established in many countries. The earliest standards were based on threshold models and use of a margin of safety. Initially, standards were based on the mass of total suspended material. In the 1980s a shift to a size-specific standard, PM(10), began. PM(10) is the fraction of PM captured with 50% efficiency at 10 microm and greater efficiency at smaller sizes. In the late 1990s, standards were proposed for PM(2.5), which is captured with 50% efficiency at 2.5 microm. The standards for PM are based almost exclusively on human epidemiological data, with laboratory animal and in vitro data used in a supporting role. During the 1990s, new statistical tools began to be used and demonstrated an association between increased PM and an increase in cardiorespiratory morbidity and mortality. The analyses are complicated by the effects of other pollutants such as ozone. Effects have been observed down to 10-20 microg of PM(10) per cubic meter, levels equal to or below background in many parts of the world. In many studies there has been no evidence of a threshold. In the absence of a threshold, a critical issue becomes how to determine how low is low enough? This paper reviews the current literature on PM health effects and suggests research avenues that may yield data which, combined with public policy considerations, may be able to address the issue of 'how low is low enough?'