Geographic areas with extremely high mortality correlations for cancers in the province of Alberta, Canada

Age-adjusted mortality rates, calculated by the direct method and based on the 1971 census-population of Alberta, for 24 male and 26 female cancer sites, from 29 census sub-divisions of the province of Alberta, collected over the period 1961 through 1981, were employed in this study. A total of 50,374 cases of cancer mortality were utilized. Correlations between pairs of cancer sites across census sub-divisions were calculated and cancer sites noted that varied little in mortality rates over Alberta's geographic areas. Significant correlations of mortalities (p less than 0.01) were observed between 5 pairs of male cancer and 18 pairs of female cancer sites. Risk factors common to correlating pairs are discussed. Similarities of our results with other studies are pointed out. The geographical distribution of extreme mortality correlations is analyzed and possible risk factors are discussed.