Abstract
Recent studies suggest a link between air pollution and lung cancer, but causality remains uncertain due to confounding and reverse causation. Mendelian randomization (MR) reduces such bias and offers a new way to explore this relationship. MR is a method that uses genetic variants as instrumental variables to assess the causal relationship between an exposure and an outcome, effectively controlling for confounding and reverse causation. The inverse-variance weighted method is a commonly used approach in MR analysis, which estimates the overall causal effect by weighting the effect ratios of multiple single nucleotide polymorphisms, assuming all instruments are valid. Based on 2-sample MR, this study incorporated 5 air pollution indices and conducted MR analyses with lung cancer outcome data from 2 different sources. Subsequently, a meta-analysis was performed on the primary inverse-variance weighted results, followed by multiple corrections of the thresholds after the meta-analysis to ensure accuracy. Finally, reverse causality was tested through MR analysis for air pollution indices significantly associated with lung cancer. And the selection criteria for instrumental variables were: P < 5 × 10⁻⁶, F > 10, minor allele frequency > 0.01, clump_kb = 10,000, and clump_r2 = 0.001. Five air pollution indices were analyzed using MR analysis and meta-analysis with lung cancer data from the FinnGen R12 and OpenGWAS databases. Multiple corrections were applied to the significance threshold results after the meta-analysis. The final results showed that only nitrogen dioxide (NO₂) exhibited a significant association, with an OR of 3.426 (95% CI: 1.897-6.186, P = 2.21 × 10⁻⁴). Additionally, the positive air pollution index NO₂ showed no evidence of reverse causality with lung cancer from either data source. This study demonstrates a significant causal association between NO₂ and lung cancer, indicating that NO₂ may be a potential risk factor for lung cancer.