Search

The 1991 Canadian Census Cohort is the largest population-based cohort in Canada (N=2,734,835). Prior to the creation of this Cohort, no national population-based Canadian cohort was available to examine mortality by socioeconomic indicators. The 1991 Canadian Census Cohort was created via the linkage of a sub-sample of respondents from the mandatory 1991 Canadian Census long-form to historical tax summary files, Canadian Mortality Database, Canadian Cancer Database, 1991 Health and Activity Limitation Survey and a sub-sample of the Longitudinal Worker File. Overall ascertainment of mortality and cancer is anticipated to be nearly complete and the Cohort is broadly representative of most groups in the Canadian population. The Cohort has been used to examine mortality outcomes by different indicators of socioeconomic status, occupational categories, ethnic groups, educational attainment, and for exposure to ambient air pollution. Results have shown that the estimated remaining years of life at age 25 differed substantially by income adequacy quintile, educational attainment, housing type and Aboriginal ancestry.

Background Increasingly more First Nations (FN) people have moved from rural to urban areas. It is unknown how disparities in infant mortality among FN versus non-FN women have changed over time in urban versus rural areas. Methods We conducted a birth cohort-based study of all 877 925 live births (56 771 FN and 821 154 non-FN) registered in British Columbia, 1981-2000. Main outcomes included rates, risk differences, and relative risks of neonatal, postneonatal, and overall infant death. Results Both neonatal and postneonatal mortality rates for FN infants showed a steady decline in rural areas but a rise-and-fall pattern in urban areas. Relative risks for overall infant death among FN versus non-FN infants declined steadily from 2.75 (95% CI: 2.04, 3.72) to 1.87 (95% CI: 1.24, 2.81) in rural areas from 1981-1984 to 1997-2000, but rose from 1.59 (95% CI: 1.27, 1.99) (1981-1984) to 2.80 (2.33-3.37) (1989-92) and then fell to 1.89 (1.44-2.49) (1997-2000) in urban areas. Risk differences for neonatal death among FN versus non-FN infants declined substantially over time in rural but not urban areas. The disparities in neonatal death among FN versus non-FN were largely explained by differences in preterm birth, while the disparities in postneonatal death were not explained by observed maternal and pregnancy characteristics. Conclusions Reductions in disparities in infant mortality among FN versus non-FN women have been less substantial and consistent over time in urban versus rural areas of British Columbia, suggesting the need for greater attention to FN maternal and infant health in urban areas.