Passive smoking and lung cancer association: a result of bias?

1 Epidemiological studies have reported that non-smokers married to smokers have a lung cancer risk 20-50% higher than that of non-smokers married to non-smokers. 2 In contrast, extrapolation based on relative smoke exposure of passive and active smokers would predict a much smaller effect. 3 This paper examines the possibility that bias due to misclassification of smoking habits, coupled with between spouse smoking habit concordance, could account for this discrepancy. 4 One thousand seven hundred and seventy-five subjects were asked about their smoking habits and use of other nicotine products in a non-health context likely to minimize underreporting of smoking. One thousand five hundred and thirty-seven provided saliva for cotinine analysis. Of 808 who claimed not to be users of such products, 2.5% had cotinine values above 30 ng/ml, suggesting their self reports were false. In another study 540 subjects were interviewed in 1980 and in 1985. Ten per cent claiming on one occasion never to have smoked, made inconsistent statements on the other occasion. A third study showed a strong tendency for smokers to marry smokers. 5 Estimates of bias based on these data indicate that misclassification can explain the unexpectedly high lung cancer risk associated with spouse smoking in epidemiological studies of self-reported never smokers.