June 5, 2002
Researchers Update Findings of Air Pollution Study
Recently, researchers from the Johns Hopkins Bloomberg School of Public Health conducted new analyses of the National Morbidity, Mortality, and Air Pollution Study (NMMAPS) data set. NMMAPS comprised the first nationwide time series analyses of air pollution, weather, and mortality in the 90 largest cities in the United States, finding an association between particulate matter levels in the air and daily mortality counts in U.S. cities. The new analyses update some of the quantitative findings of NMMAPS; however, the researchers state that the qualitative findings remain unchanged and there is strong evidence for an association between particulate matter and mortality.
The primary investigator of the study is Jonathan Samet, MD, MS. The research team includes Scott Zeger, PhD, Francesca Dominici, PhD, and Aidan McDermott, PhD, Their work appeared in the New England Journal of Medicine and other peer-reviewed publications. While updating and expanding their work, the investigators recognized a limitation of the S-plus statistical program used to analyze data for the study. The S-plus program is standard statistical software used by many researchers around the world for time-series and other analyses. The Hopkins investigators determined that the default criteria in one regression program used for examining patterns and fitting the statistical model (referred to as convergence criteria) were too lenient for this application, resulting in an upward bias in the estimated effect of air pollution on mortality.
As a result of this discovery, the researchers have updated many of the original findings of the NMMAPS. For the 90 U.S. cities studied, the estimated effect of particulate matter on total mortality dropped from a 0.4 percent increase in deaths per 10 micrograms of particulate matter per cubic meter to about 0.2 percent when analyzed in a way that is insensitive to this bias. In addition to this quantitative change, the researchers reported that the key qualitative findings were unchanged. There remains strong evidence for a particulate-mortality association. They state that the association is strongest for cardiovascular and respiratory causes of death, and the association cannot be attributed to other gaseous pollutants such as carbon monoxide, ozone, nitrogen dioxide, and sulfur dioxide.
The NMMAPS team has reported these new findings to the Health Effects Institute (HEI), which is funding the NMMAPS. In turn, HEI has notified the Environment Protection Agency and other groups studying the NMMAPS data. The researchers are also in the process of notifying the scientific journals that published the original studies and are encouraging other scientists to review other time-series studies compiled using the S-plus software or other similar firstname.lastname@example.org.