Tobacco smoking and risk for idiopathic pulmonary fibrosis: a prospective cohort study in UK Biobank

Introduction: Idiopathic pulmonary fibrosis (IPF) is a disease with unknown etiology. Smoking is considered a potential risk factor, but only a limited number of studies is available. Aims and objectives: Our objective was to examine the association between tobacco smoking and IPF in a large prospective cohort study. Methods: We used data from UK Biobank, a large population-based study of more than 500,000 individuals. We defined IPF cases as patients with an ICD10 diagnosis of J84.1. We excluded cases with a diagnosis of connective tissue disease (ie rheumatoid arthritis, systematic lupus erythematosus, polyarteritis nodosa, necrotizing vasculopathies, dermatopolymyositis, scleroderma, Sjogren syndrome and sarcoidosis) and cases with an ICD10 diagnosis of J60 – J70 (ie lung diseases due to external agents). We ran logistic regression models for tobacco smoking (ever versus never smokers) and maternal smoking with risk for IPF. The estimates were adjusted for sex and age. Results: A total of 1,275 cases with an ICD10 code of J84.1 were identified. After exclusion of cases with a connective tissue disease and lung disease due to external agents, 685 cases of IPF and 501,362 controls were identified. The association of ever smoking with risk for IPF presented an odds ratio (OR) of 1.79 (95% confidence interval [CI], 1.50-2.14; P= 4 × 10⁻¹⁰). We also found a novel association between maternal smoking and risk for IPF with an OR of 1.41 (95% CI, 1.19-1.68; P= 1 × 10⁻⁴). In a multivariable model combining the two factors, the effect estimates and the level of significance remained unchanged. Conclusions: Tobacco smoking and maternal smoking are strong and independent risk factors for IPF.