Ambient fine particulate matter exposure and lung cancer risk in the NIH-AARP Diet and Health Study

Abstract

Background and Aim: Fine particulate matter (PM2.5) is an established cause of lung cancer; however, few studies have evaluated risk by histologic subtype. Methods: We estimated outdoor PM2.5 concentrations at the residential enrollment (1995-1996) address for 494,728 participants of the NIH-AARP Diet and Health Study, a cohort located in 6 states (California, Florida, Louisiana, New Jersey, North Carolina, and Pennsylvania) and 2 metropolitan areas (Atlanta, Georgia, and Detroit, Michigan). We used annual estimates from a national spatiotemporal model to assess PM2.5 concentrations for a 5-year pre-enrollment period (1990-1994). We used Cox regression to estimate hazard ratios and 95% confidence intervals (HR[CI]) for risk of incident lung cancer overall and by histologic subtype per 10µg/m3 increase in PM2.5. Models were adjusted for demographic characteristics, smoking status and intensity, and census-tract median household income. We also examined models stratified by smoking status, sex, and study catchment area at baseline, and we tested for statistical interaction via Wald tests. Results: 16,766 lung cancers were diagnosed in follow-up through 2011. We observed a small, nonsignificant increased risk for lung cancer overall (HR=1.02[0.96-1.08] per 10µg/m3). Associations were stronger for adenocarcinoma (HR=1.10[1.00-1.21]) and squamous cell carcinoma (HR=1.08[0.94-1.24]), whereas we found no association for small cell carcinoma (HR=0.98[0.83-1.15]). No significant differences in risk were noted by smoking status or sex (p-interaction all >0.05). Associations between PM2.5 and lung cancer overall varied widely among the 8 cohort catchment areas (p-interaction <0.01). State-specific associations for lung cancer ranged from HR=0.78[0.46-1.30] in Louisiana to HR=1.60[1.27-2.01] in Pennsylvania. Conclusions: Our findings from this large U.S. cohort indicate that risk of lung cancer increases with PM2.5, particularly for adenocarcinoma and squamous cell carcinoma subtypes. Associations varied by region, suggesting that PM2.5 constituency may influence differences in lung cancer risk. Keywords: lung cancer, outdoor air pollution, particulate matter