Impaired lung function and lung cancer risk in 461 183 healthy individuals: a cohort study

Background It has been known that smoking and various lung diseases including lung cancer can cause lung function impairment. However, the impact of different types of lung function impairments, such as preserved ratio impaired spirometry (PRISm) and airflow obstruction (AO), on the incidence and mortality of lung cancer in both general and never-smoker populations remains unclear. We wished to examine the effect of lung function impairments on lung cancer risks. Methods This was a retrospective cohort study (1 January 1994 to 31 December 2017) of individuals from a health surveillance programme in Taiwan who underwent baseline spirometry tests at the entry point. PRISm was defined as an FEV1/FVC (forced expiratory volume in 1 s/forced vital capacity) ratio >0.7 and FEV1 <0.8, while AO was defined as an FEV1/FVC ratio <0.7. Cox proportional hazards models and cubic spline curves were used to examine the associations between lung function impairments and lung cancer risks. Results The study included 461,183 individuals, of whom 14.3% had PRISm and 7.9% had AO. A total of 4038 cases of lung cancer and 3314 lung cancer-related deaths were identified during the 23 years of follow-up. Individuals with PRISm and AO exhibited a higher risk of lung cancer incidence and mortality compared with those with normal lung function. The adjusted HRs and 95% CIs were 1.14 (1.03 to 1.26) and 1.23 (1.10 to 1.37) in the overall cohort, and 1.08 (0.93 to 1.24), and 1.23 (1.05 to 1.45) in the never-smoker cohort. The risks of both developing and dying of lung cancer increased with the severity levels of lung function impairments and lower FEV1 values. Conclusion Impaired lung function is associated with increased risks of developing lung cancer and subsequent mortality. The study highlights the importance of considering lung function in lung cancer screening for better candidate selection. Data are available upon reasonable request. The data used in this study was authorised by MJ Health Research Foundation (Authorisation Code: MJHRFB2014001C). The MJ Health Research Foundation administered MJ Health Survey Database and MJ BioData, and the data were available at the website: . The study design, data collection, data analysis, data interpretation, writing of the report and submission for publication were independently decided by the authors and had no relation to the funding source. We are grateful to the Health and Welfare Data Science Center and National Health Research Institutes for providing administrative and technical support.