Epidemiological evidence relating environmental smoke to COPD in lifelong non-smokers: a systematic review [version 3; peer review: 2 approved]

Abstract

Background: Some evidence suggests environmental tobacco smoke (ETS) might cause chronic obstructive pulmonary disease (COPD). We reviewed available epidemiological data in never smokers. Methods: We identified epidemiological studies providing estimates of relative risk (RR) with 95% confidence interval (CI) for various ETS exposure indices. Confounder-adjusted RRs for COPD were extracted, or derived using standard methods. Meta-analyses were conducted for each exposure index, with tests for heterogeneity and publication bias. For the main index (spouse ever smoked or nearest equivalent), analyses investigated variation in RR by location, publication period, study type, sex, diagnosis, study size, confounder adjustment, never smoker definition, and exposure index definition. Results: Twenty-eight relevant studies were identified; nine European or Middle Eastern, nine Asian, eight American and two from multiple countries. Five were prospective, seven case-control and 16 cross-sectional. The COPD definition involved death or hospitalisation in seven studies, GOLD stage 1+ criteria in twelve, and other definitions in nine. For the main index, random-effects meta-analysis of 33 heterogeneous (p<0.001) estimates gave a RR of 1.20 (95%CI 1.08-1.34). Higher estimates for females (1.59,1.16-2.19, n=11) than males (1.29,0.94-1.76, n=7) or sexes combined (1.10,0.99-1.22, n=15 where sex-specific not available), and lower estimates for studies of 150+ cases (1.08,0.97-1.20, n=13) partly explained the heterogeneity. Estimates were higher for Asian studies (1.34,1.08-1.67, n=10), case-control studies (1.55,1.04-2.32, n=8), and COPD mortality or hospitalisation (1.40,1.12-1.74, n=11). Some increase was seen for severer COPD (1.29,1.10-1.52, n=7). Dose-response evidence was heterogeneous. Evidence for childhood (0.88,0.72-1.07, n=2) and workplace (1.12,0.77-1.64, n=4) exposure was limited, but an increase was seen for overall adulthood exposure (1.20,1.03-1.39, n=17). We discuss study weaknesses that may bias estimation of the association of COPD with ETS. Conclusions: Although the evidence strongly suggests that ETS increases COPD, study weaknesses and absence of well-designed large studies preclude reliable effect estimation. More definitive evidence is required.