Long-term association of air pollution and incidence of lung cancer among older Americans: A national study in the Medicare cohort

BACKGROUND: Despite strong evidence of the association of fine particulate matter (PM(2.5)) exposure with an increased risk of lung cancer mortality, few studies had investigated associations of multiple pollutants simultaneously, or with incidence, or using causal methods. Disparities were also understudied. OBJECTIVES: We investigated long-term effects of PM(2.5), nitrogen dioxide (NO(2)), warm-season ozone, and particle radioactivity (PR) exposures on lung cancer incidence in a nationwide cohort. METHODS: We conducted a cohort study with Medicare beneficiaries (aged ≥ 65 years) continuously enrolled in the fee-for-service program in the contiguous US from 2001 to 2016. Air pollution exposure was averaged across three years and assigned based on ZIP code of residence. We fitted Cox proportional hazards models to estimate the hazard ratio (HR) for lung cancer incidence, adjusted for individual- and neighborhood-level confounders. As a sensitivity analysis, we evaluated the causal relationships using inverse probability weights. We further assessed effect modifications by individual- and neighborhood-level covariates. RESULTS: We identified 166,860 lung cancer cases of 12,429,951 studied beneficiaries. In the multi-pollutant model, PM(2.5) and NO(2) exposures were statistically significantly associated with increased lung cancer incidence, while PR was marginally significantly associated. Specifically, the HR was 1.008 (95% confidence interval [CI]: 1.005, 1.011) per 1-μg/m(3) increase in PM(2.5), 1.013 (95% CI: 1.012, 1.013) per 1-ppb increase in NO(2), and 1.005 (0.999, 1.012) per 1-mBq/m(3) increase in PR. At low exposure levels, all pollutants were associated with increased lung cancer incidence. Men, older individuals, Blacks, and residents of low-income neighborhoods experienced larger effects of PM(2.5) and PR. DISCUSSION: Long-term PM(2.5), NO(2), and PR exposures were independently associated with increased lung cancer incidence among the national elderly population. Low-exposure analysis indicated that current national standards for PM(2.5) and NO(2) were not restrictive enough to protect public health, underscoring the need for more stringent air quality regulations.