Background:
Lung cancer remains the leading cause of cancer-related mortality globally, with the number of cases projected to increase substantially in coming decades. The disease is often diagnosed at an advanced stage, resulting in poor survival and underscoring the importance of preventive strategies. While tobacco smoking is the dominant risk factor, ambient air pollution represents a pervasive exposure largely beyond individual control. In 2013, the International Agency for Research on Cancer classified outdoor air pollution and particulate matter (PM) as Group 1 carcinogens for lung cancer. However, uncertainty remains regarding which pollutants, sources and exposure lags contribute most to risk, limiting effective regulation. Health effects may vary by particle size, source and chemical composition. Improved exposure assessment capturing spatial and temporal variability is critical to reduce misclassification and clarify latency.
Objectives:
To examine associations between incident lung cancer and time-varying residential exposure to PM?.?, PM??, black carbon (BC) and nitrogen oxides (NO?) across multiple exposure lags in a large population-based cohort from southern Sweden.
Methods:
The study included 22 294 participants from the Malmö Diet and Cancer cohort, enrolled between 1991 and 1996 and followed until lung cancer diagnosis, death, emigration or end of follow-up in 2016. Incident lung cancer cases were identified through the Swedish Cancer Register. Annual residential exposures to PM?.?, PM??, BC and NO? from 1990 to 2016 were estimated using a high-resolution (50 × 50 m) Gaussian dispersion model based on detailed emission inventories and meteorological data. BC was estimated as a fraction of PM?.?. Time-dependent Cox proportional hazards models with age as the underlying time scale were used to estimate hazard ratios (HRs) and 95% confidence intervals (CIs) for baseline, 1–5 year lag and 6–10 year lag exposure windows. Models were progressively adjusted for sociodemographic factors, lifestyle variables and detailed smoking history.
Results:
During 325 966 person-years of follow-up, 499 incident lung cancer cases were identified. Mean baseline exposures were slightly higher among cases than non-cases for all pollutants. In minimally adjusted models, elevated HRs were observed for PM?.?, PM?? and BC, particularly for lagged exposure windows. These associations were attenuated after adjustment for smoking and socioeconomic factors, but remained suggestive. Fully adjusted HRs were 1.21 (95% CI: 0.51–2.89) per 10 µg·m?³ for PM?? (lag 1–5 years), 1.10 (0.67–1.80) per 0.5 µg·m?³ for BC (lag 1–5 years) and 1.26 (0.52–3.00) per 5 µg·m?³ for PM?.? (lag 6–10 years). Associations with NO? were consistently null across all exposure windows. Lag-specific patterns suggested stronger associations for PM?? and BC at shorter lags and for PM?.? at longer lags, although estimates were imprecise.
Conclusions:
Long-term exposure to PM?.?, PM?? and BC was weakly associated with lung cancer risk, with effect estimates sensitive to adjustment for smoking and other confounders. The null findings for NO? and attenuation of associations highlight the challenges of disentangling air pollution effects in low-to-moderate exposure settings. Further large-scale studies are needed.

Air pollution and lung cancer