SununrThis report presents information on n'sk factors for oesophageal cancer in Bulawayo. Zimbabwe. The data analysed were from the Cancer Registry of Bulawayo for the years . when all registered patients were interviewed using a standard questionnaire. The age-standardised incidence rates in the urban population of Bulawayo in the first 10 year period were 58.6 per 100 000 in men and 8.1 in women. The distribution of risk factors was assessed in 881 oesophageal cancer cases (826 male. 55 female) and a control group comprising other non-tobacco-and non-alcohol-related cancer (5238) Cancer of the oesophagus varies widely in incidence between different parts of the world. One area of high risk is in southern and eastern Africa, where it constitutes 11.5% of cancers in men, corresponding to an annual age-standardised incidence rate of 20.7 10-5 . Previous reviews (Oettle. 1964: Cook, 1971) have drawn attention to the male excess (ranging from 1.5 to 12 in different series). and the apparent increase in incidence over time, with hospital series from the 1930s and 1940s suggesting that it was at that time a relatively rare condition. Within east and southern Africa, areas of high and low incidence are intermingled, and the sharp geographic gradients in nrsk suggest a predominant role for environmental carcinogens in the aetiology (Day and Mufioz. 1982).In al., 1970, 1976. 1993 Odds ratios (ORs) for each risk factor, with 95% confidence limits, were estimated. while controlling for potential confounding variables, by multivariate regression. using the GLIM software package (Baker and Nelder, 1978). Subjects with missing values for the variables studied were included in the analyses as a separate category. Individuals non-resident in Zimbabwe were excluded from the analysis. P-values for trend were calculated by categorising and scoring the exposure variable and treating it as continuous.
ResultsNot all of the patients (or their relatives) could be interviewed, and in those who were, the questionnaire was not always completed. Overall, a complete interview was obtained in 72.6% of oesophageal cancer cases, and in 71.3% of the control group cancers.