Shilpa Gowda

Project title: Ambient Air Pollution and Lung Cancer Risk in the Women's Health Initiative Observational Study

Degree: MPH | Program: Occupational and Environmental Medicine (OEM) | Project type: Thesis/Dissertation
Completed in: 2016 | Faculty advisor: Joel D. Kaufman


Background: Outdoor air pollution was recently classified as a human carcinogen by the International Agency for Research on Cancer. However, epidemiologic studies supporting this classification have focused on lung cancer mortality rather than incidence, and most previous studies have included relatively few women. Women in older age groups with higher incidence of lung cancer have been particularly underrepresented. New studies that address these limitations are needed to further explore and quantify the strength of the association between ambient air pollution and lung cancer. 

Methods: We sought to evaluate this association, using data from the observational arm of the Women’s Health Initiative (WHI) study, a large, U.S.-based cohort of post-menopausal women (final sample size after applying exclusions = 86,146, with 1,332 lung cancer cases). Mean follow-up time of participants was 11 years, with a maximum follow-up time of 15 years. We used previously validated state-of-the-art geospatial models to estimate exposures to fine particulate matter (PM2.5) and nitrogen dioxide (NO2), two major constituents of ambient air pollution, based on participants’ residential addresses. NO2 is considered a proxy measure for traffic- related air pollutants. We also characterized participants’ exposures to traffic-related air pollution by the distances of their residential addresses to a primary limited-access highway. Using Cox proportional hazards regression models, we calculated hazard ratios (HRs) for the risk of lung cancer in association with these exposure metrics. We adjusted for several potential confounders, including age, race, body mass index, region of residence (including urbanicity), smoking habits, and socioeconomic status. Furthermore, we conducted exploratory analyses restricted to never smokers, adjusting for second-hand smoke exposures. Separate exploratory analyses stratified lung cancer outcomes by major histological subtypes. 

Results: In our primary analyses, no significant associations were observed. For example, when comparing the highest quartile of PM2.5 exposure (>14.59 µg/m3) to the lowest quartile (≤10.58 µg/m3), a HR of 0.91 [95% Confidence Interval (CI): 0.61-1.34] was observed. There were no consistent relationships seen between distance to roadway and lung cancer. There were also no significant associations observed in our exploratory analyses. 

Conclusions: Overall, we did not observe an association between ambient air pollution and lung cancer risk. However, despite the large sample size, our results do not exclude effect sizes seen in some previous studies. Additional years of follow-up and inclusion of participants from the WHI clinical trial arm would enhance study power and may make results more informative.