Tobacco Smoking and Risk for Pulmonary Fibrosis: A Prospective Cohort Study From the UK Biobank

Chest. 2021 Sep;160(3):983-993. doi: 10.1016/j.chest.2021.04.035. Epub 2021 Apr 24.


Background: Idiopathic pulmonary fibrosis (IPF) is an interstitial lung disease of unknown origin. A limited number of small studies show an effect of tobacco smoking on risk of IPF, but second-hand smoking has not been examined.

Research question: Are smoking-related exposures associated with risk of IPF and does interaction between them exist?

Study design and methods: We designed a prospective cohort study using UK Biobank data, including 437,453 nonrelated men and women of White ethnic background (40-69 years of age at baseline). We assessed the effect of tobacco smoking-related exposures on risk for IPF using Cox regression adjusted for age, sex, Townsend deprivation index, and home area population density. We also examined potential additive and multiplicative interaction between these exposures. Multiple imputation with chained equations was used to address missing data.

Results: We identified 802 incident IPF cases. We showed an association between smoking status (hazard ratio [HR], 2.12; 95% CI, 1.81-2.47), and maternal smoking (HR, 1.38; 95% CI, 1.18-1.62) with risk of IPF. In ever smokers, a dose-response relationship was observed between pack-years of smoking and risk of IPF (HR per 1-pack-year increase, 1.013; 95% CI, 1.009-1.016). Furthermore, an additive and multiplicative interaction was observed between maternal smoking and smoking status, with a relative excess risk due to interaction of 1.00 (95% CI, 0.45-1.54) and a ratio of HRs of 1.50 (95% CI, 1.05-2.14).

Interpretation: Active and maternal tobacco smoking have an independent detrimental effect on risk of IPF and work synergistically. Also, intensity of smoking presents a dose-response association with IPF, strengthening the hypothesis for a potentially causal association.

Keywords: UK Biobank; cohort study; idiopathic pulmonary fibrosis; interaction; smoking.