Air pollution and lung cancer incidence in the Malmö diet and cancer cohort

Background Outdoor particulate air pollution is classified as causing lung cancer, but evidence on specific pollutants and exposure timing remains limited. Methods 22,294 participants in the population-based Malmö Diet and Cancer cohort, were enrolled between 1991 and 1996 and followed until 2016. Incident lung cancer cases were identified through national registers. Annual residential exposure to PM2.5, PM10, black carbon (BC), and nitrogen oxides (NOx) was estimated using high-resolution dispersion models and assigned based on residential history. Time-dependent Cox regression models with age as the time scale were used to estimate hazard ratios (HRs) and confidence intervals (CIs) for lung cancer incidence for exposure at baseline (1990–1994), the five years preceding diagnosis or censoring (lag 1–5), and the 6–10 years prior (lag 6–10). Models were progressively adjusted for smoking (status, intensity, duration), environmental tobacco smoke, employment, occupation, education, physical activity, cohabitation, and area-level socioeconomic status. Results During 325,966 person-years, 499 participants developed lung cancer. In models adjusted for age, sex, and calendar time, positive associations were observed for PM2.5, PM10, BC and NOₓ. Adjustment for smoking substantially attenuated the estimates leading to imprecise and not statistically significant associations. For example, the HR for PM2.5 (lag 1–5 years) was 1.03 (95% CI: 0.48–2.19) per 5 µg/m³, PM10 (lag 1–5 years) was 1.21 (95% CI: 0.51–2.89) per 10 µg/m³, and for BC (lag 1–5 years) was 1.10 (95% CI: 0.67–1.80) per 0.5 µg/m³, while associations with NOₓ were close to null. Discussion and Conclusion We observed suggestive but imprecise associations between long-term PM exposure and lung cancer incidence. The attenuation after adjustment for smoking highlights the importance of careful confounder control. Overall, findings do not provide strong evidence of an independent association in this low-exposure setting and should be interpreted cautiously.