Tobacco-related cancer in relation to prevalence of drinking and smoking in eastern Austria.

Eastern Austrian regional mortality patterns of oral cancer (oral cavity, pharynx and larynx) and oesophageal, lung and urinary bladder cancer were compared to smoker rates and to liver cirrhosis mortality by type of residence: Vienna (1.7 x 10(6) inhabitants), middle towns (50,000-100,000 and 10,000-50,000 inhabitants), small towns (2000-10,000 inhabitants) and rural areas categorized by agrarian quota less than or equal to 10%, 10%-20% and greater than 20%. The study area (Vienna, Lower Austria and Burgenland) covers 23,600 km2 with 3.23 x 10(6) inhabitants. In men, liver cirrhosis correlated negatively with smoker rates (r = 0.74, P = 0.1). Deaths from oral cancer and oesophageal cancer correlated significantly with deaths from liver cirrhosis (r = 0.81, P = 0.03; r = 0.78, P = 0.04, respectively) but not with smoker rates; lung cancer and bladder cancer correlated significantly with smoker rates (r = 0.91, P = 0.01; r0.83, P = 0.04, respectively), but not with liver cirrhosis. In women, similar urban-rural gradients for all parameters resulted in a positive correlation between liver cirrhosis and smoker rates (r = 0.59, P = 0.22) and a significant correlation of lung cancer with liver cirrhosis (r = 0.75, P = 0.05). Oral cancer correlated significantly with liver cirrhosis (r = 0.83, P = 0.02), but not with smoker rates; lung cancer correlated more significantly with smoker rates (r = 0.92, P = 0.01) than with liver cirrhosis; bladder cancer correlated positively with smoker rates (r = 0.70, P = 0.12). Geographical distribution of oral and oesophageal cancer in Eastern Austria seems thus to be highly subject to the prevalence of heavy drinking. Sociocultural influences upon the occurrence of these cancers seem to be mediated through drinking habits rather than through smoking habits alone.