Epidemiology has become increasingly suspect as the results of this discipline are used more and more as tools for social engineering. Smoking, drinking and diet are but a few of many aspects of daily life which are subject to government intervention and private castigation as a consequence of the results of trials to determine the effects of lifestyle on health and mortality. I have shown elsewhere that many such trials have had their results misinterpreted both by their authors and by others (1).Here is another example
R.C. Brownson et al. in the abstract to their paper "Passive smoking and lung cancer in non smoking women"(2) state "Ours and other recent studies suggest a small but consistent increased risk of lung cancer from passive smoking". That is not the case. Whatever other studies may have suggested, theirs shows nothing of the sort. On the contrary, their study suggests that passive smoking prevents lung cancer. Their results are summarised in their Tables 1 to 3 which list a total of 76 odds ratios. These compare different exposure situations and also different types of lung cancer in different situations. Of the 76, 67 are non-significant. Their 9 significant odds ratios range from 0.4 to 0.7, all showing a reduction in lung cancer with passive smoke exposure. These, the only significant results in their tables, are not mentioned in their abstract or any other part of their text . Instead they say, for example, "There was little evidence of increased lung cancer risk associated with passive smoke exposure in childhood (Table 1)". Table 1 lists 16 odds ratios, all of them less than one and 7 of them significantly so. This is not "little evidence of increased lung cancer risk" but good evidence of decreased lung cancer risk.
This odd result appears to have gone unnoticed.
1. Johnstone, J.R. Health Scare : The Misuse of Science in Public Health Policy (Australian Institute for Public Policy , [now the Institute of Public Affairs] Perth ,1991)
2. Brownson, R.C., Alavanja, M.C.R., Hock, E.T. & Loy,T. S. American Journal of Public Health 82, 1525-1530 (1992)