Associations between community income and cancer incidence in Canada and the United States

BACKGROUND. Associations between socioeconomic status (SES) and the inciden ce of cancer have been reported previously in the U.S. Canada has more comp rehensive health care and social programs than the U.S. The purpose of this study was to compare the strength of associations between SES and cancer i ncidence in Canada and the U.S. METHODS. The regions studied were the Canadian province of Ontario and the areas of the U.S. covered by the Surveillance, Epidemiology, and End Result s (SEER) program. The populations at risk were defined using the 1991 Canad ian Census and the 1990 U.S. Census. The populations of Ontario and of the SEER areas of the U.S. were each divided into deciles on the basis of media n household income. Population-based cancer registries were used to identif y incident cases. Age-standardized incidence rates for all major groups of malignant diseases were calculated for each SES decile in Ontario and in th e U.S. Income-associated incidence gradients observed in Ontario and the U. S. were compared. RESULTS. The incidence of most types of cancer was similar in Ontario and t he U.S. In both countries, there were moderately strong, inverse associatio ns between income level and the incidence of carcinomas of the cenix, the h ead and neck region, the lung, and the gastrointestinal tract. In both Onta rio and the U.S., several of these diseases were twice as common in the bot tom income decile than they were in the top decile. In contrast, carcinoma of the female breast and carcinoma of the prostate were more common among h igher income communities in both countries, but the observed associations w ere weaker in Ontario. CONCLUSIONS. Despite Canada's universal health insurance and more comprehen sive social security system, the association between lower socioeconomic st atus and the incidence of many common cancers is just as strong in Ontario as it is in the U.S. The mechanisms responsible for these associations requ ire further investigation. (C) 2000 American Cancer Society.