The effect of smoking on the risk of lung cancer mortality for asbestos workers in Great Britain (1971-2005).

OBJECTIVES Workers in the asbestos industry tend to have high smoking rates compared to the general population. Both asbestos exposure and cigarette smoking are recognized risk factors for lung cancer mortality, but the exact nature of the interaction between the two remains uncertain. The aim of this study was to examine the effect of smoking and smoking cessation among asbestos workers in Great Britain (GB) and investigate the interaction between asbestos exposure and smoking. METHODS The study population consisted of 98 912 asbestos workers recruited into the GB Asbestos Survey from 1971, followed-up to December 2005. Poisson regression was used to estimate relative risks of lung cancer mortality associated with smoking habits of the asbestos workers and to assess whether these effects differed within various categories of asbestos exposure. The interaction between asbestos exposure and smoking was examined using the Synergy (S) and Multiplicativity (V) indices, which test the hypotheses of additive and multiplicative interaction, respectively. The proportion of lung cancers among smokers attributable to the interaction of asbestos and smoking was also estimated. RESULTS During 1 780 233 person-years of follow-up, there were 1878 deaths from lung cancer (12% of all deaths). Risk of lung cancer mortality increased with packs smoked per day, smoking duration, and total smoke exposure (pack-years). Asbestos workers who stopped smoking remained at increased risk of lung cancer mortality up to 40 years after smoking cessation compared to asbestos workers who never smoked. The effects of smoking and stopping smoking did not differ by duration of asbestos exposure, main occupation, age at first asbestos exposure, year of first exposure, or latency period. The interaction between asbestos exposure and smoking for asbestos workers was greater than additive [S 1.4, 95% confidence interval (CI) 1.2-1.6], and the multiplicative hypothesis could not be rejected (V 0.9, 95% CI 0.3-2.4). For those asbestos workers who smoked, an estimated 26% (95% CI 14-38%) of lung cancer deaths were attributable to the interaction of asbestos and smoking. CONCLUSIONS This study emphasizes the importance of smoking prevention and cessation among those who work in the asbestos industry.