The Prevalence of Diabetes in the Cree of Western James Bay

Volume 21, No.3 - 2000

[Table of Contents]

The Prevalence of Diabetes in the Cree of Western James Bay

David AL Maberley, Will King and Alan F Cruess

Volume 21, No. 3 - 2000

Abstract

Diabetes prevalence and general demographic data for individuals with diabetes were evaluated in the Cree of Moose Factory, Ontario. Individuals with diabetes were identified through a retrospective review of the diabetes registry as well as of outpatient and inpatient records. The crude prevalence of diabetes was 62 (95% confidence interval: 54-72) per 1,000. The direct age-standardized prevalence of diabetes was 103 per 1,000 for the entire population (95% confidence interval: 89-118 per 1,000, standardized to the 1991 Canadian population). The estimated rate of diabetes in the Canadian population is approximately 5%. The average age of individuals with diabetes in the community was 53 years; the average duration of diabetes was 8.2 years. Most of the population with diabetes were female (64%) and were using anti-hypertensive medications (64%). This study presents diabetes prevalence data for the population of Moose Factory, Ontario, that indicate a higher prevalence than in both the Canadian population and other Cree populations in the region.

Key words: aboriginal; adult; diabetes mellitus; Indians, North American; non-insulin dependent diabetes; Ontario; prevalence

Introduction

The overall prevalence of self-reported diabetes among Canadian adults (aged 18-74) was approximately 5.1% in the Canada Health Survey.¹ Diabetes in North American Indians has become more prevalent over the past 50 years,² and certain native Canadian populations now suffer from a significantly higher prevalence of diabetes than the general Canadian population.^3-6

The prevalence of diabetes varies markedly from population to population, although it is difficult to compare prevalence data between studies because of major differences in methods. Young et al.'s 1990 review of the geographic distribution of diabetes in aboriginal Canadians provides an overview of regional differences in prevalence figures.⁷ This study demonstrated that the highest age-standardized diabetes rates were in the Atlantic region and the lowest rates were among the Inuit. The Algonkian speakers of the Northeast, who include the Cree of James Bay, have among the highest diabetes prevalence of all native groups in Canada.

Although few data exist for the Cree of western James Bay, Brassard et al.'s study of the Cree population of Quebec found the age-standardized prevalence of diabetes to be high in comparison with the Canadian population as a whole.⁸ This elevated prevalence in the Cree may indicate recent, accelerated changes in lifestyle and diet that have occurred in these communities over the past 50 years.^9,10

The objective of this paper is to evaluate the crude and age-standardized prevalence of diabetes in the community of Moose Factory, Ontario, and to compare these figures with those of other Canadian aboriginal populations.

Methods

The communities of western James Bay include Moose Factory, Moosonee, Attawapiskat, Kashechewan, Fort Albany, Peawanuk and New Post. The regional population is approximately 11,000 and is very stable, with little migration in or out. The inhabitants are predominantly Cree, one of several tribes that make up the Algonquin peoples.^11,12 Moose Factory is the largest community in the region and is located on an island in the mouth of the Moose River. The Cree inhabitants of Moose Factory over the age of 15 number 1,900.

Weeneebayko General Hospital, in Moose Factory, is the only hospital in the region. Health care for the population of Moose Factory is provided in this hospital or at an outpatient family medicine clinic that is based at the hospital.

Travel to the communities of western James Bay is primarily by air. A rail line does reach Moosonee, but there is no access road into the region from the south. All the communities of western James Bay can be considered "remote:" there is little contact from non-native populations, and traditional hunting and gathering practices are still maintained.

The data for this study were collected during a review of both the Weeneebayko Hospital's diabetes registry and all patient charts in the Moose Factory Outpatient Clinic. At the time the data were compiled, in June 1998, all known individuals with diabetes were cared for through this clinic. Thus there was no possibility of missing subjects who were receiving medical care elsewhere. Inpatient records were reviewed to confirm the diagnosis of diabetes in individuals whose diagnostic laboratory data were not available in the outpatient charts.

Data from the chart review were collected through the hand searching of patient files. A diagnosis of diabetes was confirmed after consideration of fasting blood glucose studies taken during the course of routine medical care at the Moose Factory medical clinic. The attending physicians at the clinic use standard World Health Organization (WHO) criteria to determine the diagnosis of diabetes. Specifically, a diagnosis of diabetes was assigned if patients' fasting blood glucose levels were above 7.8 mmol/L or oral glucose tolerance test levels were greater than 11.1 mmol/L.¹³ All laboratory studies were performed at the Weeneebayko Hospital Laboratory. For the glucose tolerance test, blood sugar levels were measured two hours after glucose load. Individuals with high random glucose tests were re-examined with fasting blood glucose tests at the outpatient clinic.

There is a very low prevalence of type I (insulin-dependent) diabetes in the Quebec Cree.⁹ During the course of this study, only one individual with type I diabetes was identified in Moose Factory. For this reason, all data presented are for individuals with type II (non-insulin-dependent) diabetes. Individuals with gestational diabetes or secondary diabetes were excluded from this study. Non-Cree people with diabetes were also excluded.

Basic demographic and medical history data for the cohort with diabetes were recorded during the chart review. Age and duration of diabetes were recorded as of January 1, 1998. Where physicians' notes did not document the onset of diabetes, this point was defined as the date of the oldest laboratory value meeting the WHO criteria for diabetes. Treatment status was recorded as the regimen an individual was receiving for the majority of the five-year period preceding January 1998. Hypertension was defined as present if a patient was taking medications to control his or her blood pressure. Macrovascular complications of diabetes were defined as a history of stroke with sequelae, or myocardial infarction as documented by electrocardiography.

Data were also collected for routine blood tests. Many individuals did not have extensive laboratory profiles over the course of their diabetic history; however, an attempt was made to standardize the laboratory results in a manner that would make these values somewhat meaningful. For all individuals, laboratory values were recorded from tests that were performed as close as possible to a point two years after each individual's diagnosis of diabetes. These values were chosen to provide some indication of the diabetic metabolic status that exists in this community during the early stages of diabetes.

Descriptive statistics were calculated using the SPSS statistical program. The direct age-standardized prevalence for Moose Factory was calculated using 1991 Canadian census data for individuals over the age of 15.¹⁴ Confidence intervals (CIs) for estimates of type II diabetes prevalence were calculated using the binomial distribution.¹⁵

This study was conducted with approval from the Weeneebayko General Hospital Board, the hospital's Chief Executive Officer and the Chief Medical Officer. The Hospital Board includes two representatives from each Cree community that receives medical care through the hospital. The study protocol was also approved by the Queen's University Health Sciences Human Research Ethics Board. All data were initially presented in an open forum in Moose Factory.

Results

Based on estimates supplied by the Health Planning Office of the Weeneebayko Hospital, the total Cree population of Moose Factory in 1997 was 2,819, with 1,900 individuals over 15 years of age.¹⁶

From the chart review, 174 individuals were identified as having diabetes. There were no individuals under age 15 with diabetes. The crude prevalence of diabetes for the entire population was 62 (95% CI: 53-72) per 1,000. The crude prevalence for individuals over age 15 was found to be 92 (95% CI: 79-105) per 1,000 [Table 1].

TABLE 1 Age-specific prevalence of diabetes by sex, Moose Factory population, 1998
Age (years)	Males		Females		Both sexes
Age (years)	Prevalence per 1,000	Crude numbers	Prevalence per 1,000	Crude numbers	Prevalence per 1,000
15-24	11.9	3/251	7.9	2/253	9.9
25-34	14.9	4/268	43.8	11/251	28.9
35-44	56.8	10/176	120.2	25/208	91.1
45-54	184.9	22/119	166.7	21/126	175.5
55-64	272.7	12/44	483.3	29/60	394.2
65-74	131.5	5/38	265.3	13/49	206.9
75-84	285.7	6/21	411.8	7/17	342.1
85+	0.0	0/6	307.7	4/13	210.5
All ages	67.2	62/923	114.6	112/977	91.6
Standardized rate^a (per 1,000) for ages 15 and up = 130.85 (95% CI: 115.7-146.0)
^a Age-standardized rates for Moose Factory subjects were calculated using the Canadian population distribution from the 1991 census as the standard.

To evaluate the effect of a person's sex on the prevalence of diabetes, sex-specific values were calculated for males and females over the age of 15 (Table 1). The male and female prevalences were 67 (95% CI: 51-83) per 1,000 and 115 (95% CI: 95-135) per 1,000 respectively for those over age 15. The highest prevalence of diabetes was among women between the ages of 55 and 64, of whom 48% were found to have diabetes.

Because the Moose Factory Cree have a larger proportion of their population under the age of 35 than the Canadian population, direct age-standardization was performed using 1991 Canadian census data with individuals over the age of 15 as the referent population.¹⁴ The direct age-standardized prevalence of diabetes among individuals over age 15 in Moose Factory was 131 (95% CI: 116-146) per 1,000. When all ages were considered, the age-standardized prevalence was 103 (95% CI: 92-114).

Demographic features of the full cohort are presented in Table 2. Data were not available for all subjects on each variable; for all variables, data were available for approximately 90% of subjects.

TABLE 2 Basic demographic data for Moose Factory population
Demographic variable	Subjects (with data)	Average number (continuous variables) or percentage (discrete variables)	Standard deviation
Total number of subjects	174
Average age of subjects	174	53 years	15 years
Average duration of diabetes	170	8.2 years	6.4 years
Number of males/females	174	36% / 64%
Treatment regimen (diet/oral/insulin)	171	28% / 53% / 19%
Hypertensives/Normotensives	169	64% / 36%
Stroke or myocardial infarction (ever/never)	155	14% / 86%
Average hemoglobin A_1C	166	10%	3%
Average body-mass index	155	32.4 kg/m²	4.9 kg/m²
Average serum cholesterol	155	5.25 mmol/L	1.10 mmol/L
Average blood urea nitrogen	164	5.24 mmol/L	1.74 mmol/L
Average serum creatinine	165	64.8 mmol/L	19.9 mmol/L

The average age of individuals with diabetes in Moose Factory was 53, and the average duration of diabetes was eight years. Sixty-four percent of all those with diabetes were women, and 64% were taking anti-hypertensive medication(s). Most subjects were being treated with oral hypoglycemics, and 14% had suffered a myocardial infarction or stroke (Table 2). With an average body-mass index (BMI) of 32.4 kg/m², this population with diabetes would be considered markedly obese (obesity defined as a BMI above 27).

The average hemoglobin A_lC for the Moose Factory population with diabetes was found to be 0.10-considered to indicate poor control for individuals with diabetes. The average serum cholesterol was 5.25 mmol/L, a level that corresponds to "borderline risk," while the average blood urea nitrogen and serum creatinine levels for the cohort were within the normal range for the Weeneebayko Hospital Laboratory.

Discussion

For the general Canadian population, the estimated prevalence of diabetes is approximately 50 per 1,000.¹ The present study demonstrates a significantly higher prevalence of diabetes in the Cree of James Bay, Ontario. In the community of Moose Factory, the crude prevalence of diabetes was found to be 62 per 1,000 overall, and 92 per 1,000 for individuals over age 15. Direct age-standardization provided a prevalence estimate of 131 per 1,000 in James Bay, Ontario. These figures are significantly higher than for the James Bay Cree of Quebec, among whom Brassard et al. found the crude prevalence of diabetes to be 27 (95% CI: 24-30) per 1,000 and the age-adjusted prevalence to be 66 (95% CI: 59-73) per 1,000 for those over age 20 (Table 3).⁸

TABLE 3 Direct age-standardized prevalence of diabetes for different Canadian populations
Region	Tribe/culture	Age range for standardization	Standard population	Direct age- standardized prevalence (95% CI)
Moose Factory, Ontario (present study)	Cree	15+	1991 Canadian	13.1% (11.5 -14.6)
James Bay, Quebec⁸	Cree	20+	1986 Canadian	6.6% (5.9-7.3)
Sandy Lake, Ontario²²	Ojibwa/Cree	10+	1991 Canadian	26.1% (22.9-29.3)
Southwestern Ontario³	Oneida/Chippewa/Muncey	5+	1985 Canadian	14.7% (12.7-16.7)
Southwestern Ontario³	Caucasian	5+	1985 Canadian	2.2% (1.6-2.8)
Moose Factory, Ontario (present study)	Cree	All ages	1991 Canadian	10.3% (9.2-11.4)
Atlantic region⁷	Algonkian	All ages	1985 Canadian	8.7% (8.1-9.3)
Ontario⁷	Algonkian/Iroquoian	All ages	1985 Canadian	7.6% (7.3-7.9)
Sioux Lookout, Ontario²²	Ojibwa/Cree	All ages	1986 Canadian	6.7% (4.6-8.7)

The differences in diabetes prevalence between these two related populations do not appear to be connected to sampling differences because both studies identified subjects through physician-diagnosed registries that used similar diagnostic criteria. The higher diabetes prevalence found in the present study may represent actual differences between the Ontario and Quebec study populations.

The increasing Westernization of North American aboriginal populations has resulted in striking dietary and lifestyle changes over the past 50 years.¹⁷ Brassard et al.'s study showed a marked geographic gradient in the prevalence of diabetes, such that more isolated communities were somewhat protected from the disease. The elevated diabetes prevalence in Moose Factory may indicate that this community is less remote than those examined by Brassard et al. In fact, Moose Factory is quite a developed community, with fast-food vendors, restaurants and department stores allowing increased access to Western dietary choices. The degree of Westernization in Moose Factory is similar to that of Chisasibi, Quebec; however, the inclusion of many more isolated communities in the Quebec estimate may have resulted in an overall lower prevalence. As well, the fact that sampling occurred a decade earlier suggests that the Quebec communities, as a whole, were less Westernized at the time of Brassard's study than Moose Factory is now. A very recent survey of the Quebec Cree population indicates that the crude prevalence of diabetes has increased over the past 10 years to more closely approximate the values presented in this study.¹⁸

Young et al.'s overview paper from 1990 presents age-standardized data for different native language groups based on the 1985 Canadian intercensal population estimate. The lowest rates of diabetes in the Canadian native population were found in the Northwest Territories, and the highest rates were noted in the Ontario and Atlantic regions.¹² As with the work of Brassard et al., a definite geographic gradient was observed. Lower diabetes prevalence figures were noted in more northerly latitudes and in more isolated regions.

To compare the diabetes prevalence figures found in this study with other North American Indian populations is complicated. Research methods vary greatly from study to study. Different population selection techniques, diagnostic criteria and standardized populations make for difficult comparisons of prevalence figures. Table 3 presents studies of native Canadian peoples that have incorporated age-standardized prevalence figures for diabetes using Canadian census data for standardization. To facilitate comparison, the direct age-standardized results of this study are presented for both the complete population and for individuals over age 15. When age-standardized prevalence figures were available for the entire population, the "all ages" figures were uniformly lower. This is due to the association between age and an increased prevalence of type II diabetes.

Studies of specific native populations of Ontario, Alaska and Arizona have demonstrated lower diabetes prevalence rates than in the Ontario Cree population of James Bay.^19-21 In contrast, higher recorded prevalence figures were found in the Ojibwa-Cree community of Sandy Lake, Ontario.²² Interestingly, for the tribes that demonstrated age-adjusted prevalence figures in the 100 per 1,000 range (Table 3), there was a marked homogeneity of heritage-the Chippewa, Cree, Ojibwa and Oneida are all part of the same language group (Algonkian).

The summary statistics for diabetes in Moose Factory (Table 2) demonstrate a significantly higher prevalence of diabetes among women than men. This finding has been documented in every other Canadian aboriginal diabetes prevalence study. Delisle et al. found a statistically different age-adjusted diabetes prevalence between men and women in the Lac Simon Algonquin of Quebec: 49% of women between 30 and 64 years of age were found to have diabetes in contrast to 24% of men.²³ Reasons for the recorded discrepancies between aboriginal men and women could be that women are often identified as having diabetes during routine pregnancy screening tests. Alternatively, the increased prevalence may relate to a higher rate of obesity, impaired glucose tolerance and type II diabetes at younger ages among native women.²¹ The impact of an increased prevalence of gestational diabetes must also be considered a contributing factor in the development of type II diabetes both maternally and in children.²⁴ Alternatively, women might be more active in seeking medical care than men.

The main limitation of the present study was the use of diabetes registry/chart review data for the diagnosis of cases. This may have resulted in a potentially incomplete picture of the state of diabetes in the Moose Factory Cree. Nonetheless, the data likely represent the minimum prevalence of diabetes in this community. The true rate could be as much as 25-30% higher.^25,26

Local Omuskegowuk Band Council initiatives have targeted diabetes as an area of community health focus in Moose Factory. The community is well educated about this condition and may have a higher capture rate for diabetes than other communities. If this is the case, the chronic disease registry used in this study may be more representative of the true diabetes prevalence in Moose Factory than in other communities with similar databases.²⁷

Another interesting statistic was the large number of individuals with diabetes who were taking anti-hypertensive medications (64%). This finding could be explained either by a significant association between diabetes and hypertension or by the more aggressive use of these medications by physicians for individuals with diabetes.

The average serum creatinine and blood urea nitrogen levels for the entire cohort were within the normal range for individuals without diabetes, suggesting that primary hypertension and not diabetic renal failure may be the likely mechanism for hypertension in these people. The cohort's average serum cholesterol was also within the normal non-diabetic range; however, as might be expected, hemoglobin A_1C levels were significantly higher than normal.

The importance of this study is that it documents a high prevalence of diabetes in the Moose Factory Cree when compared with the Quebec Cree population and other Canadian aboriginal populations. No data have been published about the prevalence of diabetes in the James Bay Cree of Ontario. This study serves as a baseline for future surveillance projects. The data also confirm local Band Council concerns about the magnitude of diabetes as a public health problem in Moose Factory and suggest that efforts are warranted toward furthering our understanding of diabetes, its related complications and sex/geographic variability in the James Bay Cree. It is known that the prevalence of diabetes in aboriginal Canadian communities is higher than that of the Canadian population as a whole. This study demonstrates that the prevalence of diabetes in the James Bay Cree is higher than in other Cree populations in the region and may be among the highest in the country. Moreover, the estimates presented here are likely an underestimation of the true prevalence of diabetes in the Moose Factory Cree.

References

1. Tan MH, MacLean DR. Epidemiology of diabetes mellitus in Canada. Clin Invest Med 1995;18(4):240-6.

2. West KM. Diabetes in North American Indians and other native populations of the New World. Diabetes 1974;23:841-55.

3. Evers S, McCracken E, Antone I, Deagle G. The prevalence of diabetes in Indians and Caucasians living in southwestern Ontario. Can J Public Health 1987;78:240-3.

4. Montour LT, Macaulay AC. High prevalence rates of diabetes mellitus and hypertension on a North American Indian reservation. Can Med Assoc J 1985;132:1110-2.

5. Worrall G, Fodor G, Butt M. A minimal prevalence study of diagnosed diabetes in Newfoundland and Labrador. Can J Public Health 1991;82:104-8.

6. Pioro M, Dyck R, Gillis D. Diabetes prevalence rates among First Nations adults on Saskatchewan reserves in 1990: comparison by tribal grouping, geography and with non-First Nations people. Can J Public Health 1996;87(5):325-8.

7. Young TK, Szathmary EJ, Evers S, Wheatley B. Geographical distribution of diabetes among the native population of Canada: a national survey. Soc Sci Med 1990;31(2):129-39.

8. Brassard P, Robinson E, Lavallée C. Prevalence of diabetes mellitus in the James Bay Cree of northern Quebec. Can Med Assoc J 1993;149(3):303-7.

9. Young TK, McIntyre LL, Dooley J, Rodriguez J. Epidemiologic features of diabetes mellitus among Indians in northwestern Ontario and northeastern Manitoba. Can Med Assoc J 1985;132:793-7.

10. Robinson E. The health of the James Bay Cree. Can Fam Physician 1988;34:1606-13.

11. Indian and Northern Affairs Canada. Health Planning Office, Weeneebayko Health Ahtuskaywin (Moose Factory, Ontario) [unpublished data].

12. Young TK, Emoke JES, Evers S, Wheatley B. Geographical distribution of diabetes among the native population of Canada: a national survey. Soc Sci Med 1990;31:129-39.

13. World Health Organization. Diabetes mellitus: report of a WHO Study Group. Geneva: WHO, 1985:11; WHO Technical Report Series No 727.

14. Statistics Canada. Profile of census metropolitan areas and census agglomerations (1991 census). Part A and Part B. Ottawa, 1992 and 1993; Cat 93-337 and 93-338.

15. Rosner B. Fundamentals of biostatistics. 4^th ed. Belmont (CA): Wadsworth Publishing, 1994:176.

16. Omuskegowuk Band Council (Moose Factory, Ontario). Population estimates, 1997 [unpublished data].

17. Szathmary EJE, Ritenbaugh C, Goodby CM. Dietary change and plasma glucose levels in an Amerindian population undergoing cultural transition. Soc Sci Med 1987;24:791-804.

18. Dannenbaum D, Verronneau M, Torrie J, Robinson E, Harris R, Dumont C. Prevalence of diabetes in Eeyou Istachee (Eastern James Bay Cree). Proceedings of the National Aboriginal Diabetes Association; 2000 June 1-4; Winnipeg, Manitoba.

19. Fox C, Harris SB, Whalen-Brough E. Diabetes among native Canadians in northwestern Ontario: 10 years later. Chronic Dis Can 1994;15(3):92-6.

20. Gohdes D, Kaufman S, Valway S. Diabetes in American Indians. Diabetes Care 1993;16 (Suppl 1):239-42.

21. Young TK, Schraer CD, Shubnikoff EV, Szathmary EJE, Nikitin YP. Prevalence of diagnosed diabetes in circumpolar indigenous populations. Int J Epidemiol 1992;21:730-6.

22. Harris SB, Gittelsohn J, Hanley A, Barnie A, et al. The prevalence of NIDDM and associated risk factors in native Canadians. Diabetes Care 1997;20(1):185-7.

23. Delisle H, Rivard J, Ékoé J. Prevalence estimates of diabetes and of other cardiovascular risk factors in the two largest Algonquin communities of Quebec. Diabetes Care 1995;18(9):1255-9.

24. Mohamed N, Dooley J. Gestational diabetes and subsequent development of NIDDM in aboriginal women of northwestern Ontario. Int J Circumpolar Health 1998;57 (Suppl 1):355-8.

25. Delisle HF, Ekoé JM. Prevalence of non-insulin dependent diabetes mellitus and impaired glucose tolerance in two Algonquin communities in Quebec. Can Med Assoc J 1993;148:41-7.

26. Young TK, Krahn J. Comparison of screening methods in a diabetes prevalence survey among northern Indians. Clin Invest Med 1988;11:180-5.

27. Worrall G. Diabetes among native people. Can Med Assoc J 1994;150(5):644-5.

Author References

David AL Maberley and Will King, Department of Ophthalmology, University of British Columbia, Vancouver, British Columbia

Alan F Cruess, Department of Ophthalmology, Queen's University, Kingston, Ontario

Correspondence: Dr David Maberley, Department of Ophthalmology, University of British Columbia, 2550 Willow Street, Vancouver, British Columbia V5Z 3N9

[Previous][Table of Contents [Next]


	Last Updated: 2002-10-04		Important Notices