Jason Wong

Sc.D.

National Cancer Institute

Postdoctoral Fellow

PLCO (Learn more about this study)

PLCO-264

Mar 24, 2017

A case-crossover study of air pollution exposure and lung cancer related mortality in PLCO

I plan to conduct a case-crossover study to examine whether high exposure to air pollutants (PM10, SO2, NO2, CO, O3) above certain levels in certain time-windows is related to lung cancer risk in the overall PLCO study population, and lung cancer specific mortality among those diagnosed with lung cancer. In 2012, PLCO collected information on recent addresses for ~83,000 participants. The addresses were geocoded and linked to the 2000 and 2010 census data. The PLCO study also planned to reconstruct residential history for these participants. I propose to link the PLCO lung cancer dataset to the US Environmental Protection Agency's AIRS database which contains daily average and maximum estimates of PM10, SO2, NO2, CO, O3, based on the residential location of each participant. I will use multivariable-adjusted conditional logistic regression models to evaluate air pollution levels in specific time-windows before occurrence of lung cancer diagnosis and lung cancer related death, in relation to air pollution levels in a specified "control" time window. The case-crossover is a powerful study design in which each participant acts as their own control (each person is their own strata), which circumvents the need to consider control subjects. Further, it is the most effective study design to assess relationships with transient/temporary exposures, such as spikes in air pollution levels. The case-crossover inherently accounts for between-subject confounding bias, although within person confounding and secular trends still need to be accounted for.

1) To determine if high levels of PM10, SO2, NO2, CO, O3 in specific time-periods are associated with subsequent risk of incident lung cancer diagnosis among all PLCO participants.

2) To determine if high levels of PM10, SO2, NO2, CO, O3 in specific time-periods are associated with risk of subsequent lung cancer-related death among participants diagnosed with lung cancer.