Although outdoor air pollution and particulate matter in outdoor air have been consistently linked with increased lung cancer risk, the evidence for associations at other cancer sites is limited. Bladder cancer shares several risk factors with lung cancer and some positive associations of ambient air pollution and bladder cancer risk have been observed. This study examined associations of ambient air pollution and bladder cancer risk in the large‐scale Spanish Bladder Cancer Study. Estimates of ambient fine particulate matter (PM2.5) and nitrogen dioxide (NO2) concentrations were assigned to the geocoded participant residence of 938 incident bladder cancer cases and 973 hospital controls based on European multicity land‐use regression models. Adjusted odds ratios (ORs) and 95% confidence intervals (CI) for associations of ambient air pollution and bladder cancer risk were estimated using unconditional logistic regression models. Overall, there was no clear association between either ambient PM2.5 (OR per 5.9 μg/m3 = 1.06, 95% CI 0.71–1.60) or NO2 (OR per 14.2 μg/m3 = 0.97, 95% CI 0.84–1.13) concentrations and incident bladder cancer risk. There was no clear evidence for effect modification according to age group, sex, region, education, cigarette smoking status, or pack‐years. Results were also similar among more residentially stable participants and in two‐pollutant models. Overall, there was no clear evidence for associations of ambient PM2.5 and NO2 concentrations and incident bladder cancer risk. Further research in other large‐scale population studies is needed with detailed information on measured or modeled estimates of ambient air pollution concentrations and individual level risk factors.

Dear editor, We thank Kawada 1 for comments on our recent paper. 2 We agree that air pollution exposure assessment is one of the main limitations of our analysis, being based on modeled estimates of recent concentrations assigned to the participant residence at enrollment some 10 years prior. Although addressing limitations of previous work based on surrogate indicators of ambient air pollution, further work with more detailed individual-level estimates of historical exposure prior to cancer occurrence may be useful. Kawada 1 also points out that our analysis employed unconditional as opposed to conditional logistic regression models to estimate associations of ambient air pollution and incident bladder cancer risk. Unconditional logistic regression analysis of matched case-control studies adjusting for matching variables results in valid and possibly more precise estimates of association compared to conditional logistic regression when sparse data are not of concern. 3 Our analysis also adjusted for a range of bladder cancer risk factors in both the main analysis and in a range of sensitivity analyses with little change in findings observed. Other analyses of air pollution and bladder cancer based on either minimal or fully adjusted approaches have also noted little change in relative risk estimates observed. 4 Yours sincerely