| pubmed-article:2389148 | pubmed:abstractText | The prevalence of diagnosed diabetes was determined for 76% of the registered Indian and Inuit (Eskimo) population of Canada from case registers maintained by the federal agency responsible for Indian health services. A total of 5324 cases were ascertained. The age-sex adjusted rate varied among the Indians from a low of 0.8% in the Northwest Territories to a high of 8.7% in the Atlantic region. Among Inuit, the prevalence was 0.4%. Most cases occur in middle-aged or older individuals, with a higher prevalence among Indian (but not Inuit) females. An ecologic analysis was performed with the crude prevalence of individual communities regressed upon independent variables that included longitude, latitude, geographic isolation, culture area, and language family. Stepwise regressions were also carried out within the Algonkian, Athapaskan, and Eskimoan language families. For the national sample, composite language phylum-culture area predictors were used. The results in the national sample confirmed most findings in the individual language family analysis. Six predictors: latitude, Northeast-Algonkian, Northeast-Iroquoian, Subarctic-Algonkian, Plains-Siouan, and Plains-Algonkian, ranked here in decreasing order of importance, explained 48.4% of the variation in diabetes prevalence. All the named groups had rates significantly greater than found in the reference group of Arctic-Eskimoans. We conclude that the distribution of diabetes among Canadian natives is determined by both genetic and environmental factors. | lld:pubmed |
