The association between intra-oral cancer and surrogate markers of smokingand alcohol consumption

Objectives 1) To describe the relationships between intra-oral cancer incid ence, in England and Wales and the corresponding incidence of lung cancer ( a surrogate for cigarette smoking), and liver cirrhosis mortality (a surrog ate for alcohol consumption). 2) To elucidate the role and relative importa nce of cigarette smoking and alcohol consumption in the aetiology of intra- oral cancer. Method Registration data for intra-oral cancer in males and fe males aged 35 years and over for the 15 regional health authorities (RHAs) in England and Wales in 1979-1983 were obtained from the Office of Populati on Censuses and Surveys and aggregated for the five-year period. Census-bas ed population data enabled standardised incidence ratios (SIR) to be calcul ated for each RHA. Using the identical method, SIRs were also computed for lung cancer and standardised mortality ratios (SMR) for liver cirrhosis, th e latter for the period 1974-1978 to reflect the shorter induction time for liver disease than for cancer. Correlation coefficients (Spearman's rho) w ere calculated between the surrogate measures and intra-oral cancer inciden ce. Results For all three diseases a north-south gradient, favouring the mo re southerly RHAs was generally apparent. For males, the correlation betwee n liver cirrhosis mortality and intra-oral cancer incidence was greater and statistically more significant (rho=0.75, P<0.01)than that between lung an d intra-oral cancer incidence (rho = 0.63, P<0.05). For females, the corres ponding correlation coefficients were positive but not significant. Conclus ion For males at least, the association between a surrogate marker for alco hol consumption and intra-oral cancer was greater than that observed for ci garette smoking. This adds further evidence to the current view that alcoho l consumption may be more important in the aetiology of intra-oral cancer t han cigarette smoking.