Findings

View the most recent version.

Methods
Results
Discussion
Limitations
Summary and conclusion

According to the 2001 census, 976,000 Canadians (3% of the total population) self-identified as Aboriginal: First Nations (North American Indian), Métis or Inuit. About 5% of the Aboriginal-identity population, numbering more than 45,000, were Inuit. Inuit are descended from Aboriginal people who historically inhabited the Arctic regions of Canada, Alaska, Greenland and Siberia. Most Inuit in Canada now reside in one of four regions: the Inuvialuit region (along the Arctic coast of the Northwest Territories), Nunavut (eastern Arctic territory), Nunavik (northern Quebec), and Nunatsiavut (northern coast of Labrador).

Data on Inuit identity are collected for deaths that occur in Canada's two northern territories - the Northwest Territories and Nunavut - but not for deaths that take place in the provinces. As a result, since one-fifth of deaths to residents of the Northwest Territories and Nunavut occur in the provinces,¹ and 20% of the total Inuit population of Canada (according to self-identification questions on the census) reside in the provinces, basic health indicators such as life expectancy at birth cannot be estimated for the Inuit using standard data sources and methods.

Previously, life expectancies had been calculated from nominal list data (no longer available) for Inuit in the former Northwest Territories (including what is now Nunavut) and in Nunavik (northern Quebec) for the years 1941 to 1950 through 1978-1982.^{2, 3} Life expectancy at birth for Inuit of the former Northwest Territories rose from 29 years in 1941 to 1950 (38 years less than for Canada overall), to 37 years in 1951 to 1960 (33 years less), to 51 years in 1963 to 1966 (21 years less), and to 66 years in 1978 to 1982 (19 years less).⁴ For Inuit in Nunavik, life expectancy in 1984 to 1988 was 14 years less than for the total population of Quebec.⁵

Life expectancy figures are regularly published for Quebec's Nunavik health region, and since 2000, for the territory of Nunavut, covering the entire population of those areas, including non-Inuit. Results for 2000 to 2002⁶ showed that life expectancy at birth was about 67 years in Nunavik and about 69 years in Nunavut, or approximately 13 and 11 years less than for Canada as a whole at the time. However, the estimates are based on only three years of deaths; earlier results for what is now Nunavut are not available; and no figures have been published for the Inuit-inhabited areas of the Northwest Territories and Labrador.

Consequently, there are currently no national life expectancy estimates for the Inuit component of the Canadian population. To partially fill this data gap, a geographic-based approach was used to obtain life expectancy for all of the Inuit-inhabited areas of Canada over a 15-year period.

Methods

Areas with a relatively high proportion of Inuit residents were identified. Vital statistics death records and census population counts were used to compute life expectancy measures for these areas.

From census questions, Aboriginal groupings can be determined on the basis of ancestry, legal status (in the case of First Nations), or self-identification. The self-identification question was used to select communities for this study. The choice is important in the case of First Nations, but of less consequence for Inuit, since most who report Inuit ancestry also self-identify as Inuit.^7,8 

According to the Aboriginal identity question on the census, most Inuit live in Nunavut, followed by Nunavik, the Inuvialuit region and Nunatsiavut, and in each of these regions, the majority of the population self-identified as Inuit. The corresponding communities of residence can be readily determined on the basis of place-name-based locality codes, which are always included on vital statistics death records. Postal codes, which can also be used to determine these communities, are often missing on death records from Nunavut and the Northwest Territories.

For a given census subdivision, if the observed proportion of residents who self-identified as Inuit identity was equal to or greater than a chosen cut-off, it was included in the list of Inuit-inhabited areas. When two communities shared the same rural postal code, as was the case with Kuujjuarapik (predominately Inuit) and Whapmagoostui (predominately Cree) in northern Quebec, the population of the two communities was combined before calculation of the proportion Inuit, since assignment to the correct census subdivision based on postal code (or postal community name) would be uncertain.

The total Inuit-identity population of Canada (all provinces and territories) was 45,070. The proportion who would be included in this analysis depended on the cut-off chosen (Table 1). Choosing communities that were "at least 33% Inuit" rather than "at least 20% Inuit" did not change the number of Inuit-inhabited communities included (54). "At least 50% Inuit" would exclude North West River, Labrador (35% Inuit), Inuvik, Northwest Territories (36% Inuit), and Kuujjuuarapik/Whapmagoostui, Quebec (37% Inuit), leaving 51 Inuit-inhabited communities. "At least 67% Inuit" would also exclude Iqaluit, Nunavut (59% Inuit) and Aklavik, Northwest Territories (59% Inuit), leaving 49 Inuit-inhabited communities. (The list of communities defined by each of these potential cut-offs was virtually identical in 1991, 1996 and 2001.) In order to include all 54 of the largely Inuit communities (and all communities in the four Inuit land claims settlement areas), the 33% cut-off was selected for this analysis, although the area also includes a larger proportion of non-Inuit (20%: 5% other Aboriginal identity and 15% non-Aboriginal) than would have been the case with more restrictive cut-offs.

The census subdivisions selected were grouped into four regions: the Inuvialuit region (Northwest Territories, 6 communities), Nunavut (the entire territory, 28 communities), Nunavik (Quebec, 14 communities), and Nunatsiavut (Newfoundland and Labrador, 6 communities) (Map 1, Appendix table A).

Calendar-year deaths were compiled for three 5-year periods: 1989 through 1993, 1994 through 1998, and 1999 through 2003. Person-years at risk were estimated by multiplying by 5 the unadjusted census population counts (100% data, including the institutional population if any) for each mid-period census (1991, 1996 and 2001, respectively). Deaths were compiled based on usual place of residence, regardless of where the death occurred. For example, deaths to Nunavut residents in Ontario or Quebec were assigned to their respective home communities in Nunavut.

Computations were done for males and females, separately and combined. Age was grouped into 19 strata (less than 1, 1 to 4, 5 to 9, 10 to 14, 15 to 19, 20 to 24, 25 to 29, 30 to 34, 35 to 39, 40 to 44, 45 to 49, 50 to 54, 55 to 59, 60 to 64, 65 to 69, 70 to 74, 75 to 79, 80 to 84, 85 years or more). For each sex and age group, the death rate was calculated as the number of deaths divided by the estimated number of person-years at risk.

Abridged life tables and associated variances, standard errors, and 95% confidence intervals were calculated according to the method of Chiang⁹. The values of Chiang's a (the fraction of the last interval of life lived by those dying in the interval) was set at 0.1 for stratum 1 (to reflect the relatively high mortality in the first year of life) and to 0.5 for all other strata. Ninety-five percent confidence intervals for life expectancy were calculated as the estimate plus or minus 1.96 times its standard error. Ninety-five percent confidence intervals for differences in life expectancy (temporal increases or decreases) were calculated as the difference in life expectancy plus or minus 1.96 times the square root of the sum of the variances for each of the two life expectancies.

Special tabulations of 2001 census data were used to describe the socio-economic characteristics of the population of the Inuit-inhabited areas.

Trends in life expectancy in the Inuit-inhabited areas were compared with life expectancies reported for all Canada from 1951 to 2001.^10-14 Results for 1999 to 2003 (2001) were compared with life expectancies reported for other circumpolar regions (Greenland and Alaska), for Canadian First Nations and for other developed and developing countries.^15-18

Results

Top of Page

Characteristics of the population

The socio-demographic characteristics of the population in the Inuit-inhabited areas differed from those of the total population of Canada (Table 2). In 2001, adults in the Inuit-inhabited areas tended to have less formal education. As well, their employment-to-population ratio was somewhat lower, and while households were larger, household incomes were lower, resulting in much lower average income per person. Finally, the percentage of homes in need of major repairs was three times as high as in Canada overall.

To a large extent, these differences reflected the characteristics of Aboriginal people, particularly the Inuit, in the Inuit-inhabited areas. Among adults, 32% of Inuit and 26% of other Aboriginal peoples in these areas had no more than elementary school, compared with just 3% of the non-Aboriginal population. By contrast, only 1% of Inuit and 4% of other Aboriginal peoples had a university degree, compared with 34% of non-Aboriginal people. In the Inuit-inhabited areas, about half of Inuit and other Aboriginal people aged 15 to 64 had a job, compared with nearly 90% of non-Aboriginal people. And among those who were employed, around a quarter of Inuit and other Aboriginal people performed unskilled labour (27% and 24%, respectively), compared with 6% of the non-Aboriginal population. Fewer than one-fifth of employed Inuit and other Aboriginal people, versus almost half the non-Aboriginal group, held professional or managerial positions. And while about a quarter of Inuit and other Aboriginal people lived in homes needing major repairs, this was the case for 13% of non-Aboriginal people in these areas. (For information on progress over time with respect to such socio-economic indicators, see the Inuit social trends series recently published by Indian and Northern Affairs Canada.^19,20 Related information about each community, based on the 2001 Census of Canada, is available as a published document.²¹)

The percentage of the population who were Inuit ranged from 54% in the Inuvialuit region, to 68% in Nunatsiavut, and up to 84% in both Nunavut and Nunavik (Table 3).

Population and death data

Top of Page

From 1991 to 2001, the population of the Inuit-inhabited areas increased considerably (Table 4 ), mainly because of high birth rates among the Inuit and other Aboriginal inhabitants.²² In 2001, most of the population of the Inuit-inhabited areas resided in Nunavut (58%), followed by Nunavik (23%), the Inuvialuit region (11%), and Nunatsiavut (8%).

Over the 1991 to 2001 period, there were 3,474 deaths to residents of these areas out of a total of 643,275 person-years at risk. During this time, 18% of the deaths to residents of Nunavut and of the Inuvialuit region occurred outside of those territories, mostly in the adjacent provinces to the south (data not shown). As previously explained, such deaths were included in this analysis, according to the decedents' usual place of residence. Almost all deaths to residents of Nunavik and Nunatsiavut occurred in their respective provinces (Quebec, and Newfoundland and Labrador, respectively).

Life expectancy

Top of Page

In 1991 (1989-1993), life expectancy at birth (both sexes combined) in the Inuit-inhabited areas was about 68 years (95% CI 66.8 to 68.8) (Chart 1, Table 5). By 2001 (1999 to 2003), life expectancy in these areas had not increased, and may even have declined by about a year (95% CI -2.2 to +0.4). Life expectancy for males may have fallen by more than a year (95% CI -3.5 to +0.3) and was virtually unchanged for females (95% CI -1.5 to +1.9).

However, levels and trends in life expectancy varied by region (Chart 2). In Nunavut, life expectancy may have increased by about a year (95% CI -0.4 to +3.0), although the increase was limited to females, as that of males hardly changed. Life expectancy fell by nearly 4 years (95% CI -6.4 to -1.0) in Nunavik, and possibly, by about 3 years (95% CI -6.3 to +0.5) in the Inuvialuit region.

In 1991, life expectancy in the Inuit-inhabited areas had been 10 years less than in Canada overall, with a wider gap for females (11 years) than for males (9 years) (Table 6). By 2001, the difference was more than 12 years, and the gap was similar for males and females. At 67 years in 2001, life expectancy in the Inuit-inhabited areas was about what life expectancy had been for all Canada in 1946.

In 2001, life expectancy in Canada's Inuit-inhabited areas was the same as in Greenland, which is largely Inuit-populated¹⁶; slightly lower than for all Alaskan natives¹⁷ (only 47% of whom are Inuit²³); and about 6 years less than for Canada's First Nations (Table 7). Life expectancy in other developed countries tended to be considerably higher.¹⁸ Only developing countries had levels of life expectancy similar to those in Canada's Inuit-inhabited areas - for instance, the Dominican Republic, Egypt, and Guatemala,¹⁸ which had much lower Gross Domestic Product (GDP) per capita (about $4,000 to $5,000 in international dollars in 2004) than did Canada (about $31,000).²⁴

Infant mortality

The infant mortality rate for the Inuit-inhabited areas fell from 25.6 deaths per 1,000 population younger than age 1 (95% CI 21.6 to 30.3) in 1989 to 1993, to 21.9 (95% CI 18.2 to 26.4) in 1994 to 1998, and to 18.5 (95% CI 15.0 to 22.9) in 1999 to 2003 (data not shown elsewhere). These rates were about four times higher than those for Canada overall: 6.0, 5.2 and 4.8 deaths per 1,000 live births, respectively.²⁵ Nonetheless, the rate difference between the Inuit-inhabited areas and all Canada fell from 19.6 deaths (95% CI 15.2 to 23.9) to 16.7 deaths (95% CI 12.7 to 20.7) to 13.8 deaths (95% CI 9.9 to 17.6) per 1,000 over those years, a decrease of 5.8 deaths per 1,000, or 30%.

Discussion

Top of Page

Life expectancy in the Inuit-inhabited areas was far below that for the country overall, and considerably below that for other Aboriginal peoples in Canada. As well, while life expectancy in Canada overall continued to rise, it appears to have stagnated in the Inuit-inhabited areas, so the gap widened by more than two years during this period.

The substantial decline in life expectancy in Nunavik during this period is particularly striking. The former extraordinary gains — from 35 years in 1946 (1941 to 1951) to 61 years in 1976 (1971 to 1981)² — now appear to have stalled, with little if any lasting progress since the mid-1970s.

However, these findings for the Inuit-inhabited areas do not distinguish life expectancy for Inuit from that of non-Inuit. If the life expectancy of the 15% of the population who were non-Aboriginal is assumed to be the same as that for all Canada (79.5 years in 2001),¹⁴ and that of the 5% of the population who were other Aboriginal to be the same as that of all Registered Indians in Canada (72.8 years in 2000),^15,26,27 then, taking into account the relative population sizes of each group, the life expectancy of Inuit-identity residents would have been 64.2 years (95% CI 63.4 to 65.0) - or 2.7 years less (95% CI -3.0 to -1.6) than that of all residents of the Inuit-inhabited areas, and 15 years less than that for Canada as a whole. Also, because the non-Inuit proportion of the population varied considerably by region, the ranking of the regions according to these rough calculations of life expectancy for the Inuit-identity residents would change, putting the two more southern regions on the bottom and the two Arctic regions on top. Under these assumptions, Inuit life expectancy would have been 60.2 years (95% CI 58.6 to 61.8) in Nunavik, 60.6 years (95% CI 58.1 to 63.1) in Nunatsiavut, 64.4 years (95% CI 62.1 to 66.7) in the Inuvialuit region, and 66.2 years (95% CI 65.0 to 67.4) in Nunavut.

Limitations

Top of Page

This study used a geographic-based approach, and hence, the estimates (except those made hypothetically in the preceding Discussion) are for regions rather than for ethnic groups. That seriously limits how the findings may be interpreted, but it also has two useful implications.First, all residents of the Inuit-inhabited areas, regardless of ethnicity, may experience similar isolation and difficulty accessing health care and other services. Second, health care and other services are provided mainly on a geographic as opposed to an ethnic basis, so the target population of such services is all residents of a particular area.

Annual population estimates corrected for net census undercount are not available for census subdivisions, so uncorrected population counts were used to determine person-years at risk. This would have led to a slight overestimate of mortality rates and a corresponding slight underestimate of life expectancy—by about -0.1 year, to judge by a comparison of life tables for Nunavut calculated for 1999 to 2003 (data not shown) with those based on corrected population estimates.⁶

In the life table calculations, arbitrary values for Chiang's a were used, rather than values published for other populations or values calculated specifically for this population. However, use of a wide range of plausible alternative values had only a slight impact on life expectancy at birth: less than one-tenth of a year (data not shown), which is negligible compared with the typical 95% confidence intervals of roughly plus or minus one year in this study.

Deaths to residents of the Inuit-inhabited areas that occurred outside Canada or the United States are not included in Canadian vital statistics, so a few deaths may have been missed. This limitation also applies to deaths to residents of the rest of Canada.

Few long-term care facilities are located in the Inuit-inhabited areas. Former residents of those areas who moved south for long-term care could have been counted as residents of the south at the time of their death. This could result in an undercount of deaths at advanced ages for the usual residents of the regions considered in this study.

With a cut-off of at least 33% Inuit, 5% of the population of the Inuit-inhabited areas were other Aboriginal, and another 15%, non-Aboriginal. While the other Aboriginal group had socio-economic characteristics similar to those of the Inuit and probably somewhat higher life expectancy, the non-Aboriginal group had much more favourable socio-economic characteristics and probably considerably higher life expectancy. This mixing of high- and low-mortality populations would be expected to reduce overall mortality rates, an effect that might have been apparent had it been possible to confine the study to Inuit rather than to Inuit-inhabited areas.

Summary and conclusion

Top of Page

Areas where at least 33 % of residents were Inuit were identified, and census population counts and vital statistics death data were used to calculate life tables for those areas during three five-year periods: 1989 through 1993 (centered around 1991), 1994 through 1998 (1996), and 1999 through 2003 (2001). The population of these areas was 80% Inuit, and included 81% of all Inuit in Canada. In 1991, life expectancy in the Inuit-inhabited areas was approximately 10 years less than for Canada as a whole. And from 1991 to 2001, while life expectancy for Canada overall rose by about two years, it did not increase in the Inuit-inhabited areas (and may have fallen by about a year), further widening the gap.

Analysis of 2001 census data revealed lower levels of education and income, and poorer employment and housing conditions in the Inuit-inhabited areas compared with Canada as a whole, and within the Inuit-inhabited areas, for Inuit compared with the non-Aboriginal population. Any or all of these factors, in addition to others such as lifestyle risk factors and environmental conditions,²⁸ about which information was not available from death registrations, could be at least partly responsible for the lower life expectancy in the Inuit-inhabited areas. Moreover, although the calculations for these areas are dominated by the life expectancy of Inuit residents, they likely also reflect the presumably higher life expectancies of non-Aboriginal and other Aboriginal residents.

Over the entire study period, the infant mortality rate was approximately four times higher in the Inuit-inhabited areas, compared with all Canada. However, the absolute difference in the rates fell by 30% from 1989 to 1993 to 1999 to 2003, indicating considerable progress with respect to this key health indicator, although much remains to be accomplished.

This geographic-based method of identifying areas with a high proportion of Inuit residents could be useful for compiling a broad range of administrative data, including birth registrations, hospital morbidity statistics, and disease registry data. Future extensions of this research will examine causes of death in the Inuit-inhabited areas, and conceptually similar analyses will be undertaken for areas with a high proportion of First Nations and of Métis people. However, other methods of compiling data relevant to Aboriginal health should be considered, such as data linkages to Aboriginal population registries^29,30 and self-reporting (or reporting by next-of-kin) in vital statistics³¹ and other health records.