Measuring the Effects of Air Pollution among Persons with Severe Emphysema: The National Emphysema Treatment Trial (NETT)

Abstract

Background: Emphysema is a chronic obstructive pulmonary disease that affects approximately 5 million people in the US. For subjects with severe emphysema, few effective treatment options exist; however, the recent adoption of lung volume reduction surgery (LVRS) is considered as a promising alternative to traditional therapy. The National Emphysema Treatment Trial (NETT) was a randomized controlled trial designed to assess the efficacy of LVRS and medical therapy versus standard medical management among participants with severe emphysema. Results from the trial demonstrated less morbidity and mortality among participants receiving LVRS. We evaluated the effects of air pollution on the health of NETT participants. Methods: Data from the NETT study (1998-2003) included 1218 subjects, men and women ages 39-84. We also obtained data from the US Environmental Protection Agency Air Quality Systems database, which included daily values of fine particulate matter (PM2.5) and ozone. ZIP code specific exposures were spatially interpolated with the use of log-normal kriged models. This methodology was employed in order to assign exposures in areas with few nearby air pollution monitors. We investigated whether there was evidence of differential air pollution exposure by individual and area-level measures of socioeconomic status (SES). We assessed both the short- and long-term health effects of air pollution on the respiratory morbidity of NETT participants using mixed linear and Poisson models, which accounted for the daily (repeated) measures of air pollution data. In addition, we constructed survival models to investigate the impact of air pollutant exposures on post-LVRS mortality. Results: NETT participants experienced varying levels of exposure to ozone and PM2.5 and increasing cumulative pollutant exposures were associated with decreasing values of area-level SES. We found that PM2.5 and ozone significantly worsened respiratory function in these subjects. There also appeared to be evidence of a differential effect of ambient air pollutants on pulmonary function and respiratory symptoms according to their randomization arm. Mean ozone concentrations were significantly associated with increased post-LVRS mortality risk. The benefits reported for participants who received LVRS surgery did not persist in the presence of ambient air pollutants, suggesting that the surgery may not have conferred a protective survival effect for those participants who underwent the procedure. Conclusion: There were significant adverse impacts of ambient ozone and PM2.5 on the respiratory health and survival of NETT participants. LVRS subjects were more likely to experience worsened air pollution-related lung function outcomes. Implications from our analyses could lead to the recommendation of lowered acceptable PM2.5 and ozone limits for individuals with existing respiratory disease.