On Oct. 26, 1948, a temperature inversion laid a blanket of cold, stagnant air over Donora, Pa., a tiny mill town on the Monongahela River. Over the next 5 days, the buildup of pollution cloaked the sun, sometimes restricting vision to just a few feet. Twenty people died outright and 50 more perished within a month from lingering health damage, says consulting epidemiologist Devra Davis, a former Donora resident whose own family survived the tragedy. As bad as her hometown's pollution had been, its impact would pale against a 5-day killer smog that settled on London in December 1952. It killed some 12,000 people within 3 months, according to calculations in a June 2001 report by Davis and Michelle L. Bell of Johns Hopkins University in Baltimore. "With a death rate more than three times the norm for this period, the London fog of 1952 is widely regarded as a catalyst for the study of air pollution epidemiology," the pair noted. That science would eventually show that even the diffuse dust wafting in seemingly clear air could kill. Its victims are just harder to identify than those in the London and Donora catastrophes because most who succumb are elderly or already in ill health. Indeed, a trailblazing 1991 analysis by Joel Schwartz, then at the Environmental Protection Agency, concluded that some 60,000 U.S. residents die from heart attacks and respiratory problems each year because of the effects of airborne dust at concentrations within federal pollution limits (SN: 4/6/91, p. 212). Stunning as those numbers were at first, they're now accepted by most researchers. In that 1991 study and subsequent ones, Schwartz, now at the Harvard School of Public Health in Boston, has shown that community death rates rise and fall nearly in lock-step with local changes in concentrations of tiny dust particles-even when concentrations of those particulates are just one-quarter of the federal limit for outdoor air. Yet more than a decade later, nagging questions remain: What makes dust and smoke particles, especially small ones, toxic? Is particulate matter, as scientists call it, inherently poisonous, regardless of its composition? Or does a large surface area per unit mass make those particles robust vehicles for ferrying toxicants such as metal atoms deep into the lungs? In the past 2 years, a flurry of new data has finally begun answering these questions. The research links the greatest harm to the tiniest dust: particulate matter no more than 2.5 micrometers in diameter, called the PM-2.5 fraction. Some studies suggest that the most
