- Email: [email protected]
Prevalence of Overweight and Obesity in a Woodland Cree Community: 14 Year Trends
Prevalence of overweight and obesity in a woodland cree community: 14 year trends
ORIgINAL RESEARCh
Prevalence of Overweight and Obesity in a Woodland Cree Community: 14 Year Trends brenda g. bruner1 Phd, Karen e. chad2 Phd, roland f. dyck2 md Queen’s university, Kingston, ontario, canada university of saskatchewan, saskatoon, saskatchewan, canada
1 2
ABSTRACT
RÉSUMÉ
O B J E C T I v E : To estimate the current prevalence of over-
O B J E C T I F : Estimer la prévalence actuelle de l’embonpoint/
weight/obesity in a Woodland Cree First Nations community and examine trends over time.
obésité dans une communauté de la Première nation crie de Woodland et examiner son évolution dans le temps.
M E T h O D S : Anthropometric measures (waist circumfer-
M É T h O D E S : Des mesures anthropométriques (tour de
ence [WC], height and weight) were used to determine the prevalence of overweight/obesity. Results were compared to those obtained in 1991.
taille [TT], taille et poids) ont servi à déterminer la prévalence de l’embonpoint/obésité. Les résultats ont été comparés à ceux obtenus en 1991.
R E S U L T S : Data from 289 males and females aged 7 to 82
R É S U L T A T S : Les données obtenues auprès de 289 per-
years (mean 31.0±16.5 years) were compared to 316 participants aged 7 to 78 years (mean 24.6±15.6 years) from 1991. In 1991, 24% of youth were overweight/obese, compared to 26% in 2005. Among adults, 74% were overweight/obese in 1991, compared to 68% in 2005. Differences were not significant for youth or adults. WC increased with age in males and females. Mean WC for all female age groups exceeded the Canadian guidelines cutoff; WC for all male age groups was below the guidelines cutoff, with the exception of those aged over 65.
sonnes des deux sexes de 7 à 82 ans (âge moyen de 31,0 ± 16,5 ans) ont été comparées à celles obtenues en 1991 auprès de 316 participants de 7 à 78 ans (âge moyen de 24,6 ± 15,6 ans). En 1991, 24 % des jeunes avaient de l’embonpoint ou étaient obèses, par rapport à 26 % en 2005. Chez les adultes, la prévalence de l’embonpoint/ obésité était de 74 % en 1991, par rapport à 68 % en 2005. Les différences n’étaient significatives ni chez les jeunes ni chez les adultes. Le TT augmentait avec l’âge dans les deux sexes. Chez les femmes, le TT moyen dans tous les groupes d’âge était supérieur au seuil recommandé dans les lignes directrices canadiennes; chez les hommes, le TT était inférieur au seuil recommandé dans tous les groupes d’âge, sauf chez les hommes de 65 ans et plus.
C O N C L U S I O N S : High prevalence rates of overweight/
obesity among adults remained over time, particularly among females. The concern for unhealthy body weights among females suggests intervention programs targeting women are needed. Future research exploring communityspecific causes of overweight/obesity to identify local belief systems and lifestyle practices is required. K E y w O R D S : Aboriginal, adolescents, adults, body mass
index, obesity, overweight
C O N C L U S I O N S : La prévalence de l’embonpoint/obésité
est demeurée élevée au fil des ans, surtout chez les femmes. Les inquiétudes liées à la prévalence de l’embonpoint/ obésité chez les femmes font ressortir la nécessité de programmes d’intervention s’adressant aux femmes. Des recherches devront être menées pour déterminer les causes
address for correspondence: Brenda Bruner, School of Kinesiology and Health Studies Queen’s University, 69 Union Street, PEC 207, Kingston, Ontario, Canada, K7L 3N6 Telephone: (613) 533-6000 ext. 75210, Fax: (613) 533-2009, E-mail: [email protected] CANADIAN JOURNAL OF DIABETES. 2009;33(2):105-113.
| 105
106 |
CANADIAN JOURNAL OF DIABETES
de l’embonpoint/obésité à l’échelle communautaire, soit les croyances et les habitudes de vie qui y contribuent. M O T S C L É S : embonpoint, obésité, indice de masse cor-
porelle, autochtone, adolescents, adultes
INTRODUCTION It is well recognized that the global increases of overweight and obesity are public health concerns, and given their substantial collective increase in prevalence are regarded as a pandemic affecting millions worldwide (1). The most recent Canadian data indicate that 36.1% of adults are overweight (i.e. body mass index [BMI] 25 to 29.9 kg/m2) and an additional 23.1% are obese (i.e. BMI >30 kg/m2), a substantial increase from the 13.8% reported 25 years ago (2). One of the most notable increases is among those aged 25 to 34 years, where obesity rates more than doubled, from 8.5% in 1978/79 to 20.5% in 2004 (2). While these data revealed similar rates of obesity for males (22.9%) and females (23.2%), females were more likely to have class III obesity (BMI >40 kg/m2; females 4% vs. males 2%) (2). Overweight and obesity have been shown to be strongly associated with chronic diseases such as type 2 diabetes mellitus, cardiovascular disease and some cancers (3). However, type 2 diabetes appears to be the most directly related to increasing obesity (4), with the prevalence of type 2 diabetes also increasing worldwide (5). While the Canadian data for overweight and obesity prevalence trends were derived from a target population that included long-term residents in health institutions and household residents in all provinces and territories, 1 notable group omitted from these population studies is those living on First Nations reserves. Although the Canadian data suggest higher rates of obesity in the Aboriginal population (based on the notion that Aboriginal peoples comprise a large percentage of the northern population), data on the prevalence of overweight and obesity in this population are limited (3). (Note: the terms Aboriginal and Aboriginal peoples do not describe a homogeneous group. However, for the purpose of this paper, the terms will be used to describe First Nations Status, Non-Status, Métis and Inuit peoples, with the exception of the use of specific terms when discussing the results of research studies that have employed these terms.) To date, Canadian research has focused primarily on the pediatric population (6–8) or has been community-specific to Manitoba and Northern Ontario (9,10). Nevertheless, results from these studies have reported an increased prevalence of obesity in the Aboriginal population. These findings have also highlighted a predominantly central pattern of obesity (9), associated with an increased risk for type 2 diabetes (11), of particular concern for Aboriginal peoples, who are already at higher risk (12).
One of the few studies to determine rates of obesity among the adult Aboriginal population was conducted in 3 northern Saskatchewan communities with varying accessibility to urban centres (13). The results of this study showed that those with increased accessibility to urban centres also showed an increased BMI (13). For this particular study, obesity was defined as >10% above the Canadian mean; according to this criterion, approximately 45% of males and 65% of females in the “most accessible” community were classified as obese — proportions that were greater than for the 2 communities with less urban accessibility (13). Given the reported high prevalence of obesity among Aboriginal peoples, the purpose of this study was to determine current overweight and obesity rates in the “most accessible” Aboriginal community in the 1991 study (13). This study also afforded the unique opportunity to compare the current rates of overweight and obesity with those from 1991 (14). Therefore, the second purpose of this study was to determine if there had been an increase in rates of overweight and obesity over time.
METhODS research project approach Historically, there has often been a “helicopter approach” (15) to research in Aboriginal communities, whereby investigators arrive in a community, collect the information they require with minimal interaction with members of the community and publish their findings in scientific journals without input from the community where the research was based. This “style” of research has resulted in distrust and resistance on the part of Aboriginal communities in relation to university-based researchers, given the lack of community control, lack of local benefits and interpretation of the data in isolation of the social context (16). In keeping with more recently developed protocols and ethical guidelines for working with Aboriginal communities (17,18), this project began by establishing a partnership with the health services organization representing the Band and health director in the community. In addition, initial exploratory meetings were held with community members who expressed an interest in participating in a follow-up project on healthy body weights. The involvement of community members is essential for community-based research to be successful, as the community is seen as an integral component of the planning, development, implementation and evaluation of health-promotion strategies. As a result, a participatory approach based on the principles of community development was applied to this project. This approach has been increasingly accepted in research projects involving Aboriginal communities (15). Subsequently, a community project worker was hired through the health clinic to assist with data collection and project organization. Regularly
Prevalence of overweight and obesity in a woodland cree community: 14 year trends
scheduled visits were made by the university researchers to establish rapport and trust with employees at the health clinic and members of the community, as well as to help develop and plan the research project.
consent Informed consent for this project was developed in stages. The first stage involved obtaining consent from local governance, including board members, health portfolio councillors, the health committee and appropriate community leaders and elders. The details of the proposed project were outlined, including the purpose and scope of the study, proposed methods to obtain individual informed consent and how the information gathered would be shared with the community. Approval for the community’s participation in the project was established prior to obtaining individual consent. Once all details of the project were mutually agreed upon and community consent was granted, informed consent was obtained from the University Advisory Committee on Ethics in Behavioral Science Research, following which informed consent was obtained from each participant. The process of obtaining individual consent was approved by the health director, and all details of the consent were done in collaboration and consultation with the health director and the community project worker to ensure appropriate language and content. the community The setting for this research project was a reserve community in northern Saskatchewan, located approximately 500 km northeast of a major Saskatchewan city and 140 km west of the Manitoba border. The community is accessible throughout the year by a gravel road approximately 30 km off a main highway. At the time of data collection (September 2004 to May 2005), the population was estimated to be 972 (48% female, 52% male), with 60% under the age of 25 years. Healthcare in the community is delivered through a local health centre, which offers a variety of health programs, treatment services and 24-hour emergency care, provided by nurses who live in the community. A family physician travels to the community 2 days per week, and specialist referrals are made to the most accessible urban centre. Participants All individuals aged 7 years and older were eligible to participate. A number of strategies were used to obtain a representative sample of individuals and to maximize the number of participants in the project. Recruitment strategies were facilitated through the health clinic and included 2 health fairs, weekly health clinic programs and personal invitations to those who may have been overlooked in previous recruit-
ment methods. These individuals were invited to attend the clinic at their convenience to participate in the project.
Procedures Anthropometric measurements included waist circumference (WC), and height and weight to calculate body mass index (BMI). All measures were obtained using standard procedures (19), with the participants dressed in light clothing and without shoes. Height was determined using a wall-mounted tape measure and set square, and weight was measured using a standing floor scale (Detecto, Webb City, MO). WC was measured by positioning a tape measure horizontally at the level of noticeable waist narrowing, and the measurement was read at the end of a normal expiration. When the point of narrowing could not be found, an indeterminate waist was approximated by finding the lateral level of the 12th or lower floating rib and the girth was recorded at that site (19). Waist measurements were taken twice and recorded to the nearest 0.5 cm. When the difference between the first and second measure was greater than 0.5 cm, a third measurement was taken, and the average was determined from all 3 measurements. WC was measured only in 2005. The data were collected between September 2004 and May 2005. data analyses The sample was divided into two age cohorts: youth aged 7 to 17 years and adults aged ≥18 years. The BMI classifications used for overweight and obesity for adults were those outlined by Health Canada and the Obesity Canada Clinical Practice Guidelines Expert Panel, who define overweight as BMI ≥25.0 kg/m2 and obesity as ≥30.0 kg/m2 (20,21). Given that these guidelines are not intended for use with individuals under the age of 18 years, the classification for overweight and obesity for youth aged 7 to 17 years was based on the age-specific BMI cutoffs for children from the International Obesity Task Force (22). These cutoffs were derived from a large international sample by constructing centile curves for BMI that correspond to the adult cutoffs for overweight (25 kg/m2) and obesity (30 kg/m2) (22). Health risk associated with increased WC for adults was defined according to the Canadian Guidelines for Body Weight Classification in Adults (20). Analyses were performed using SPSS version 14.0 (SPSS Inc. Chicago, IL) and SAS version 9.0 (SAS Institute Inc., Cary, NC). Rates of overweight and obesity and mean BMI for each time point were determined by descriptive statistics. As the sample was partially dependent and partially independent (i.e. some participants were measured at both time points), generalized estimating equations (GEE) were used to test for differences in prevalence rates of overweight and obesity between the 2 time points (1991 and 2005). GEE is a CANADIAN JOURNAL OF DIABETES. 2009;33(2):105-113.
| 107
CANADIAN JOURNAL OF DIABETES
RESULTS A total of 289 male and female participants between the ages of 7 and 82 years (mean age 31.0±16.5 years) participated in this study. This included 81 youth (mean age 12.7±2.9 years) and 208 adults (mean age 38.1±14.0 years) (Table 1). The data for this study were compared to those from the 316 male and female participants between the ages of 7 and 78 years (mean age 24.6±15.6) who participated in the 1991 study; 138 youth (mean age 11.3±3.2 years) and 178 adults (mean age 34.8±13.4 years) (Table 1). As shown in Table 1, the total sample size in 2005 was slightly smaller and participants were slightly older than in 1991. In addition, a greater proportion of females participated in the follow-up study in 2005; the proportion of female and male participants in 1991 was roughly equal (Table 1). Table 1. Comparison of male and female adult and youth participants, 1991 and 2005 1991
2005
N=316
n=289
24.6±15.6
31.0±16.5
Males, n (%)
158 (50)
108 (37)
Females, n (%)
158 (50)
181 (63)
Youth (7–17 y)
N=138
n=81
Mean age, y (±SD)
11.3±3.2
12.7±16.5
Males, n (%)
68 (49)
30 (37)
Females, n (%)
70 (51)
51 (63)
Adults (18+ y)
N=178
n=208
34.8±13.4
38.1±14.0
Males, n (%)
90 (51)
78 (38)
Females, n (%)
88 (49)
130 (62)
Overall Mean age, y±SD
Mean age, y (±SD)
youth (7 to 17 years) Figure 1a shows that in 1991 nearly one-quarter (24%) of the youth population was overweight/obese (23% females and 25% males); however, there was no significant difference in rates of overweight/obesity over time (p=0.74) as
26% of the sample was overweight/obese in 2005 (32% females and 17% males). When stratified by gender, there was also no significant difference in rates of overweight/ obesity over time for males (p=0.37) or females (p=0.30). Although the overall differences were not significant, we felt it was important to indicate whether a specific age group influenced the change in overweight/obesity rates over time, particularly among the female youth. Given the small sample sizes for individual ages, the data were stratified by age group (7 to 11 years and 12 to 17 years) and are reported as the prevalence of normal weight, overweight and obesity for each age group (Table 2). Based on the criteria outlined by Cole and colleagues (2000), Table 2 shows that Figure 1. (A) youth (7–17 years)* and (B) adults (≥18 years) classified as overweight/ obese, † 1991 and 2005
A 35
1991 2005
30
Percent overweight/obese
statistical technique used to analyze dichotomous outcomes of longitudinal correlated data (23). Therefore, the raw BMI data for the GEE analysis was dichotomized into 2 categories: overweight or obese and not overweight or obese. A total of 77 individuals who had anthropometry data for both time points were included in this analysis. A subanalysis using the Kappa statistic was used to assess change in BMI classification (i.e. normal weight, overweight, obese) over time in the individuals with repeated data.
25
20
15
10
5
0
All
Males
Females
B 100
Percent overweight/obese
108 |
1991 2005
80
‡
60
40
20
0
All
Males
*Base on age-specific cutoffs † BMI ≥25.0 kg/m2 ‡ Significantly different from 1991; p<0.05
Females
Prevalence of overweight and obesity in a woodland cree community: 14 year trends
Table 2. Body weight classification among youth, 1991 and 2005 Males
Females
Age group, y
1991
2005
1991
2005
7–11 y
n=33
n=8
n=43
n=20
Normal weight, %
72.7
87.5
81.4
55.0
Overweight, %
24.2
12.5
11.6
15.0
3.0
0
7.0
30.0
n=35
n=22
n=27
n=31
Normal weight, %
77.1
81.8
70.4
74.2
Overweight, %
20.0
9.1
18.5
12.9
2.9
9.1
11.1
12.9
Obese, % 12–17 y
Obese, %
among males aged 7 to 11 years, there was a decrease in the proportion of overweight and obesity from 1991 to 2005, whereas a greater proportion of females were overweight and obese in 2005 than in 1991. Among youth aged 12 to 17 years, there was a decrease in the proportion of both males and females who were overweight, but an increase in the proportion of males who were obese.
adults (≥18 years) The mean age of adult participants in 1991 was 34.8±13.4 years, slightly lower than the mean age in 2005 (38.1±14.0 years); however, this difference was not significant. Similar to the findings reported among youth in the community, there was no significant difference in overall prevalence of overweight/obesity over time (p=0.59). However, Figure 1b shows that three-quarters (74%) of the population in 1991 were overweight/obese (10% class II and 3% class III obese) and over two-thirds (68%) were overweight/obese in 2005 (14% class II and 1% class III obese). When stratified by
gender, there was a significant decrease (p=0.04) in rates of overweight/obesity over time for males (74 to 57%) but not for females (p=0.20), whose prevalence rates remained essentially unchanged (74 to 75%) (Figure 1b). As shown in Table 3, BMI tended to increase with age in both 1991 and 2005. The mean BMIs for all age groups at both time points are consistent with the criteria for either overweight or obesity (20,24). When stratified by gender, the mean BMIs for all age groups were classified as overweight or obese with the exception of males aged 18 to 34 years in 2005. Mean BMI values for females aged 18 to 34 years in both 1991 and 2005 were above the Canadian guidelines for overweight; for all other female age groups, mean BMI values exceeded the cutoffs for obesity. Overall, the mean WC for females was 102.1±12.5 cm (46% >88 cm) and for males was 96.2±13.0 cm (9% >102 cm). As shown in Figure 2, WC increased with age for both males and females. The mean WC for all female age groups were above the recommended Canadian guidelines of 88 cm (20,24), while the mean WC for all male age groups, with the exception of those aged ≥65, were below the guidelines of 102 cm (20,24).
Paired data A separate subanalysis was performed on paired data from 77 individuals (39% males and 61% females) measured in both 1991 and 2005. The data revealed that of the 35 participants classified as normal weight in 1991, over half were either overweight (42.9%) or obese (8.6%) in 2005. Of the 23 participants classified as overweight in 1991, 52.2% remained overweight, 39.1% were obese and 8.7% were normal weight in 2005. Finally, of the 19 participants who were obese in 1991, all remained obese in 2005, with the exception of 2 who were reclassified as overweight. Overall, 35% of the sample had an increase in BMI classification; of these, 67% were female. A larger proportion (45.5%) of
Table 3. Body mass index among adults, 1991 and 2005 Overall Age group, y
Males
Females
1991
2005
1991
2005
1991
2005
18–34
n=105 27.2±4.3
n=91 26.2±5.0
n=52 26.6±3.3
n=34 24.6±5.2
n=53 27.9±5.1
n=57 27.2±4.7
35–49
n=48 30.2±5.9
n=74 29.9±5.2
n=23 30.1±6.4
n=25 28.7±4.3
n=25 30.3±5.6
n=49 30.6±5.5
50–64
n=16 31.0±4.5
n=32 29.5±5.4
n=9 30.7±3.9
n=16 28.4±5.6
n=7 31.2±5.6
n=16 30.6±5.2
≥65
n=9 31.8±4.8
n=11 31.1±6.5
n=6 30.3±3.8
n=3 30.3±6.8
n=3 34.9±6.1
n=8 31.5±6.8
Values are mean±SD
CANADIAN JOURNAL OF DIABETES. 2009;33(2):105-113.
| 109
CANADIAN JOURNAL OF DIABETES
Figure 2. Mean waist circumference by age group, 2005 120
100
Waist circumference (cm)
110 |
80
60
40
20
0
Males 18–34 y
Females 35–49 y
50–64 y
≥65 y
*BMI ≥25.0 kg/m2 Significantly different from 1991; p<0.05
†
youth participants (7 to 17 years) in 1991 had an increase in BMI classification over time compared to those who were adults (31%). A kappa test was estimated as 0.410 and indicated that there was a significant change in the classification of BMI between the 2 time points (p<0.001).
DISCUSSION The results of this study revealed there was a significant decrease in rates of overweight/obesity among adult males in the community, but no significant difference for youth or adult females over time. Although these findings were not significant, there is some indication that overweight/obesity has increased among the female youth, specifically those aged 7 to 11 years, as reflected in the increased prevalence of overweight/obesity from 1991 to 2005. Although the prevalence of overweight/obesity among adult males decreased, there was relatively little change among females, which is not surprising given the higher prevalence of overweight/ obesity when measured initially. The findings also show that in the subsample of participants with paired data, there was a significant increase in BMI classification over time. More specifically related to the younger population, it appears that the risk for unhealthy body weight is greater for females and may begin at an earlier age compared to their male counterparts. This is of concern because Aboriginal ethnicity has been shown to be an independent predictor of gestational diabetes when it interacts with pre-pregnancy obesity (25). Obesity increases the risk of developing gestational diabetes (26), and a woman who has had gestational
diabetes is at a high risk of perpetuating the cycle by giving birth to an infant with a high likelihood of becoming obese and developing type 2 diabetes at an early age (26,27). In addition, women with gestational diabetes have up to a 70% chance of developing type 2 diabetes later in life (28) and a 50% risk of developing type 2 diabetes within 5 years postpartum (26). The prevalence rates among youth at both time points were lower than those reported in other community-specific surveys involving Aboriginal youth (7,10,29,30). However, one similarity between these studies and the current study is that a greater proportion of females were overweight and/or obese compared to their male counterparts. One exception is the data reported by Katzmarzyk and Malina, where rates of obesity were similar for males and females (28.6 and 29.4%, respectively) (10). It should be noted that the reference data used to define overweight and obesity in these studies were based on the 85th and 95th percentiles from the National Health and Nutrition Examination Survey (NHANES) II and III, but the classifications were not consistent across all studies. For example, some studies defined overweight as ≥85th percentile (7,29,30), while Katzmarzyk and Malina used the same criteria (i.e. >85th percentile) to define obesity (10) . Given that the reference standards for defining overweight and obesity have varied between studies, direct comparisons should be made with caution. What is important to note, however, is that regardless of the reference standards used and communities or First Nations groups assessed (e.g. Mohawk, Oji-Cree, Woodland Cree), the findings of this study and others have reported a high proportion of children and youth classified as overweight or obese, particularly females. These findings are important given that obesity is believed to track from childhood into adulthood (31,32). Furthermore, the duration of obesity is thought to increase the risk of developing type 2 diabetes (33), thus increasing health risks as the Aboriginal population ages. In this regard, the serial data, which showed a significant increase in BMI classification over time, highlight the importance of continued surveillance and early intervention strategies. Decreasing rates of overweight and obesity could reduce the associated long-term health consequences within our First Nations communities. Within the adult population, overweight and obesity continue to remain a special concern for females, for whom mean BMIs in all age groups were classified as obese. In Canada, there have been few studies assessing prevalence rates of overweight and obesity in the adult Aboriginal population. However, of those that have been published, Katzmarzyk and Malina reported the prevalence of obesity among Anishabai Temagami First Nation males and females to be 51% and 60%, respectively (10). Although these results are higher than our findings for both males and females
Prevalence of overweight and obesity in a woodland cree community: 14 year trends
at both time points, it should be noted that Katzmarzyk and Malina defined obesity as a BMI ≥27.8 kg/m2 for males and ≥27.3 kg/m2 for females (10), the criteria outlined in the 1988 Canadian Guidelines for Healthy Weights (34). Since the present study used the current guidelines defining overweight as a BMI ≥25.0 kg/m2 and obesity as ≥30.0 kg/m2 for both males and females (20,21), this again limits direct comparison between studies. However, similar to the data reported among the youth, higher rates of obesity are demonstrated in both studies of Aboriginal females compared to males. Compared to our study, earlier data from 6 First Nations communities in northwestern Ontario and northeastern Manitoba also reported higher combined rates of overweight/obesity among both males (40 to 80%) and females (40 to 90%) ranging in age from 20 to 64 years (9). The criteria used to classify overweight and obesity were similar to the present study (i.e. overweight ≥26.0 kg/m2 and obesity ≥30.0 kg/m2). While BMI is a practical method for measuring and classifying overweight and obesity, it does not measure the pattern of body adiposity that is a risk factor for disease, independent of level of obesity (32). More specifically, abdominal fat distribution is associated with an increased risk of cardiovascular disease and type 2 diabetes (35). In adults, WC is commonly used to measure abdominal obesity (32,35) and, thus, health risk. Few studies have assessed WC among Canadian Aboriginal peoples that can be used for comparisons to the current data. In Sandy Lake, similar results to our data for males (i.e. mean WC was <102 cm) have been reported (36). However, the data for females was somewhat different. Although both studies report mean WC >88 cm, the mean WC in this study was over 10 cm higher than that reported by Hanley and colleagues (102.1±12.5 cm vs. 90.6±13.8 cm, respectively) (36). While WC values were not reported by Young and Sevenhuysen, the authors did report that the pattern of obesity was centrally located (9). Contrary to the current findings, however, this was higher among males compared to females. The results from this study suggest that females in all adult age ranges (i.e. 18 years and older) are at an increased risk of developing health problems, as the mean WC values were greater than the Canadian guidelines for women of 88 cm (20). As WC increases with age, these results further highlight the health risks for women in this community and the need for intervention programs targeting unhealthy body weights. Although there are limited Canadian data regarding the prevalence of overweight and obesity among First Nations groups, the results of this study support findings from previous studies, which suggest that high levels of overweight and obesity exist in First Nations communities. While there was no significant increase in overweight/obesity over time using a cross-sectional sample in the community, the results
from the paired data showed a significant difference in the number of individuals with an increase in BMI classification. The findings of this study and other Canadian studies reporting high rates of overweight and obesity, particularly among Aboriginal females, highlights the need for intervention strategies targeting unhealthy body weights. The results of this study and previous literature also emphasize the importance of developing surveillance measures to document changes in age-specific prevalence rates of overweight and obesity among various First Nations groups to determine the effectiveness of intervention programs. This study contributes new knowledge regarding overweight and obesity among Aboriginal people in Canada. One of the strengths of this study is the use of measured height, weight and WC. Self-reported values tend to underestimate weight, particularly among heavier participants, while height tends to be over-estimated (32). An additional strength is the follow-up nature of the study. As limited surveillance data exist, this study is a first step in tracking overweight and obesity at the community level, with future research attempting to obtain measurement data from individuals participating in the previous studies. The limitations of this study must also be acknowledged. The first is the cross-sectional nature of this research. As in many studies conducted within First Nations communities (7–10), all eligible individuals in the community were invited to participate, but many did not. This may have resulted in undersampling of certain age groups and may limit the sample representativeness. However, it should be noted that the proportion of individuals in each age group in this study was reflective of the demographics of eligible participants in the community (i.e. a greater proportion ≥18 years of age). Second, although we did not assess changes in socioeconomic status over time, a crude comparison of the 1991 and 2001 census data showed an increase in median household income ($16 928 to $23 915), still well below the provincial median ($40 251). This may influence health-related determinants that impact overweight and obesity, such as diet and physical activity, as socioeconomic status has been shown to be inversely related to healthy lifestyles (37). Third, we encountered a limited ability to compare results to other studies. In this study, we elected to use the cutoffs for overweight and obesity derived by Cole and colleagues (22) for the youth sample, given the lack of an accepted definition of obesity in the pediatric population (22), and because this approach is said to avoid arbitrarily choosing reference data (i.e. National Health and Nutrition Examination [NHANES] surveys) and cutoff points for overweight and obesity (e.g. 85th and 95th percentiles, respectively). Additionally, the cutoffs derived by Cole and colleagues are based on a worldwide population and correspond to the adult cutoff points of overweight CANADIAN JOURNAL OF DIABETES. 2009;33(2):105-113.
| 111
112 |
CANADIAN JOURNAL OF DIABETES
and obesity, the latter allowing us to examine differences in BMI in the subsample over time. Of note, we did analyze the data using the BMI percentiles, and the proportion of individuals classified as overweight and obese were similar with both methods. While the Canadian Guidelines for Body Weight Classification in Adults (20) are said to be appropriate for all racial/ethnic groups in Canada, further research is still required to determine if the cutoffs used for the general population are appropriate for Aboriginal populations (38). Finally, the procedure used to measure WC in this study (i.e. minimal waist) is commonly used in clinical and research settings; however, it should be noted that to date there is no consensus regarding the optimal protocol to measure WC. A recent review by Ross and colleagues identified 8 protocols, of which 3 (umbilicus, midpoint between the superior border of the iliac crest and lowest rib, and minimal waist) predominate in the literature (39). While the 2008 Canadian clinical practice guidelines recommends obtaining WC measures at the top of the iliac crest (21), the data for this study was collected prior to the release of the guidelines. Of note, Wang and colleagues compared WC measures taken at 4 sites, including the minimal waist (40). The results showed that the WC measure was significantly smaller at the minimal waist compared to the other 4 sites (40), suggesting that our results may underestimate the WC measures in our sample.
CONCLUSIONS The results of this study support previous literature showing high rates of overweight and obesity among Aboriginal peoples. The findings also show that females, whether youth or adult, have higher rates of overweight/obesity compared to their male counterparts, which increases their risk of developing associated health problems. Gestational diabetes and type 2 diabetes are of particular concern given that obese women are at increased risk of developing gestational diabetes, thus, increasing the risk of both the mother and offspring developing type 2 diabetes later in life. The high proportion of females classified as overweight/obese highlights the need for intervention programs targeting this subgroup, particularly young women in the community. While the cross-sectional data showed no significant increase in overweight/obesity over time, we did demonstrate a significant increase in BMI classification in the individuals measured in both 1991 and 2005. Since obesity tracks from childhood to adulthood, prevention programs, targeting unhealthy body weights in Aboriginal communities should begin at a young age. Prior to developing such programs, however, future research should be undertaken to gain an understanding of the local belief systems (41) and practices regarding modifiable risk factors for overweight and obesity, such as physical activity and dietary practices
ACKNOwEDgEMENTS We wish to acknowledge the Peter Ballantyne Cree Nation Health Services Inc, the chief and council of the community, health clinic staff, as well as Ms. Jane Ballantyne and Ms. Janet Ballantyne, for their cooperation with, assistance with and support of this project. This project was supported through a research grant from the Canadian Institutes of Health Research and the Heart and Stroke Foundation of Canada. We would also like to acknowledge Josh Lawson who contributed to the statistical analysis.
DUALITIES OF INTEREST No dualities declared.
AUThOR CONTRIBUTIONS KC and RD obtained grant funding. BB and KC created the study design. BB wrote the first draft of the manuscript. BB and JL conducted the statistical analysis. BB conducted the data collection and management in conjunction with the community. All authors contributed to the interpretation of results and review of the final manuscript.
REFERENCES 1. World Health Organization. Obesity: Preventing and managing the global epidemic. Geneva, Switzerland; WHO: 2000. 2. Tjepkema M. Measured adult obesity in Canada: Measured height and weight. In Nutrition: Findings from the Canadian Community Health Survey: Statistics Canada. Ottawa, ON; 2005:1-32. 3. Raine K. Overweight and obesity in Canada: A population health perspective. Ottawa, ON: Canadian Institute for Health Information; 2004. 4. Jung R. Obesity as a disease. Br Med Bull. 1997;53:307-321. 5. Wild S, Roglic G, Green A, et al. Global prevalence of diabetes. Diabetes Care. 2004;27:1047-1053. 6. Bernard L, Lavallee C, Gray-Donald K, et al. Overweight in Cree schoolchildren and adolescents associated with diet, low physical activity, and high television viewing. J Am Diet Assoc. 1995; 95:800-802. 7. Hanley A, Harris S, Gittelsohn J, et al. Overweight among children and adolescents in a native Canadian community: Prevalence and associated risk factors. Am J Clin Nutr. 2000;71: 693-700. 8. Potvin L, Desrosiers S, Trifonopoulos M, et al. Anthropometric characteristics of Mohawk children aged 6 to 11 years: a population perspective. J Am Diet Assoc. 1999;99:955-961. 9. Young T, Sevenhuysen G. Obesity in northern Canadian Indians: patterns, determinants, and consequences. Am J Clin Nutr. 1989;49:786-793. 10. Katzmarzyk P, Malina R. Obesity and relative subcutaneous fat distribution among Canadians of first nation and European ancestry. Int J Obes Relat Metab Disord. 1998;22:1127-1131.
Prevalence of overweight and obesity in a woodland cree community: 14 year trends
11. Young T, Reading J, Elias B, et al. Type 2 diabetes mellitus in Canada’s first nations: Status of an epidemic in progress. CMAJ. 2000;163:561-566. 12. Health Canada. Center for Chronic Disease Prevention and Control, Population and Public Health Branch. Diabetes in Canada. 2nd ed. Ottawa, ON: 2002. 13. Dyck RF, Tan L, Hoeppner VL. Body mass index, gestational diabetes and diabetes mellitus in three northern Saskatchewan aboriginal communities. Chronic Dis Can. 1995;16:24-26. 14. Dyck RF, Tan L, Hoeppner VL. Brief communication: relationship of body mass index and self-reported rates of gestational diabetes and diabetes mellitus with geographic accessibility: a comparison of three northern Saskatchewan aboriginal communities. Chronic Dis Can. 1995;16:24-26. 15. Smylie J, Kaplan-Myrth N, Tait C, et al. Health sciences research and aboriginal communities: pathway or pitfall? J Obstet Gynaecol Can. 2004;26:211-216. 16. Henderson R, Simmons D, Bourke L, et al. Development of guidelines for non-indigenous people undertaking research among the indigenous population of north-east Victoria. MJA. 2002;176:482-485. 17. American Academy of Pediatrics, Ethics considerations in research with socially identifiable populations. Pediatrics. 2004;113:148-151. 18. Tricouncil Policy Statement. Research involving aboriginal peoples. In: Ethical Conduct for Research Involving Humans. Ottawa, ON; 1998:6.1-6.4. 19. Canadian Society for Exercise Physiology. The Canadian physical activity, fitness and lifestyle appraisal resource manual. Ottawa, ON; 1996:7.12-7.18. 20. Health Canada. Office of Nutrition, Policy and Promotion. Canadian Guidelines for Body Weight Classification in Adults. Ottawa, ON; 2003. 21. Depres J, Tchernoff A. Classification of overweight and obesity in adults. CMAJ. 2007;176(suppl 8):S21-26. 22. Cole T, Bellizzi M, Flegal K, et al. Establishing a standard definition for child overweight and obesity worldwide: International survey. BMJ. 2000;320:1240-1243. 23. Hanley J, Negassa A, Edwardes M, et al. Statistical analysis of correlated data using generalized estimating equations: an orientation. Am J Epidemiol. 2003;157:364-375. 24. Lau D, Douketis J, Morrison K, et al. Executive summary: 2006 Canadian clinical practice guidelines on the management and prevention of obesity in adults and children. CMAJ. 2007;176(suppl 8):S1-13. 25. Dyck R, Klomp H, Tan L, et al. A comparison of rates, risk factors, and outcomes of gestational diabetes between aboriginal and non-aboriginal women in Saskatoon district health. Diabetes Care. 2002;25:487-493. 26. Inzucchi S. Diabetes in pregnancy. In: Burrow G., ed. Medical Complications in Pregnancy. Philadelphia, PA: W.B. Saunders Co; 1999:25-52.
27. Dabelea D, Pettitt DJ. Intrauterine diabetic environment confers risks for type 2 diabetes and obesity in the offspring, in addition to genetic susceptibility. J Pediatr Endocrinol Metab. 2001;14:1085-1091. 28. Kim C, Newton K, Knopp R. Gestational diabetes and the incidence of type 2 diabetes. Diabetes Care. 2002;25:1862-1868. 29. Young TK, Dean HJ, Flett B, et al. Childhood obesity in a population at high risk for type 2 diabetes. J Pediatr. 2000;136: 365-369. 30. Trifonopoulos M. Anthropometry and diet of Mohawk schoolchildren in Kahnawake [dissertation]. Montreal, QC; 1995. 31. Serdula M, Ivery D, Coates R, et al. Do obese children become obese adults. A review of the literature. Prev Med. 1993;22: 167-177. 32. Power C, Lake J, Cole T. Measurement and long-term health risks of child and adolescent fatness. Int J Obes Relat Metab Disord. 1997;21:507-526. 33. Everhart J, Pettitt D, Bennett P, et al. Duration of obesity increases the incidence of NIDDM. Diabetes. 1992;41:235-240. 34. Health and Welfare Canada. Health Services and Health Promotion Branch. Canadian guidelines for healthy weights. In: Report of an Expert Group Convened by Health Promotion Directorate. Ottawa, ON; 1988. 35. Snijder M, Van Dam R, Visser, et al. What aspects of body fat are particularly hazardous and how do we measure them? Int J Epidemiol. 2006;35:83-92. 36. Hanley A, Connelly P, Harris S, et al. Adiponectin in a native Canadian population experiencing rapid epidemiological transition. Diabetes Care. 2003;26:3219-3225. 37. Canadian Institute for Health Information. Improving the Health of Canadians: Income. Ottawa, ON; 2004. 38. Douketis J, Paradis G, Keller H, Martineau C. Canadian guidelines for body weight classification in adults: Application in clinical practice to screen for overweight and obesity and to assess disease risk. CMAJ. 2005;172:995-998. 39. Ross R, Berentzen T, Bradshaw A, et al. Does the relationship between waist circumference, morbidity and mortality depend on measurement protocol for waist circumference? Obes Rev. 2007;9:312-325. 40. Wang J, Thornton J, Bari S, et al. Comparison of waist circumferences measured at 4 sites. Am J Clin Nutr. 2002;77: 378-384. 41. Willows ND. Overweight in first nations children: Prevalence, implications, and solutions. Journal of Aboriginal Health. 2005; 2:76-85.
CANADIAN JOURNAL OF DIABETES. 2009;33(2):105-113.
| 113
ORIgINAL RESEARCh
Prevalence of Overweight and Obesity in a Woodland Cree Community: 14 Year Trends brenda g. bruner1 Phd, Karen e. chad2 Phd, roland f. dyck2 md Queen’s university, Kingston, ontario, canada university of saskatchewan, saskatoon, saskatchewan, canada
1 2
ABSTRACT
RÉSUMÉ
O B J E C T I v E : To estimate the current prevalence of over-
O B J E C T I F : Estimer la prévalence actuelle de l’embonpoint/
weight/obesity in a Woodland Cree First Nations community and examine trends over time.
obésité dans une communauté de la Première nation crie de Woodland et examiner son évolution dans le temps.
M E T h O D S : Anthropometric measures (waist circumfer-
M É T h O D E S : Des mesures anthropométriques (tour de
ence [WC], height and weight) were used to determine the prevalence of overweight/obesity. Results were compared to those obtained in 1991.
taille [TT], taille et poids) ont servi à déterminer la prévalence de l’embonpoint/obésité. Les résultats ont été comparés à ceux obtenus en 1991.
R E S U L T S : Data from 289 males and females aged 7 to 82
R É S U L T A T S : Les données obtenues auprès de 289 per-
years (mean 31.0±16.5 years) were compared to 316 participants aged 7 to 78 years (mean 24.6±15.6 years) from 1991. In 1991, 24% of youth were overweight/obese, compared to 26% in 2005. Among adults, 74% were overweight/obese in 1991, compared to 68% in 2005. Differences were not significant for youth or adults. WC increased with age in males and females. Mean WC for all female age groups exceeded the Canadian guidelines cutoff; WC for all male age groups was below the guidelines cutoff, with the exception of those aged over 65.
sonnes des deux sexes de 7 à 82 ans (âge moyen de 31,0 ± 16,5 ans) ont été comparées à celles obtenues en 1991 auprès de 316 participants de 7 à 78 ans (âge moyen de 24,6 ± 15,6 ans). En 1991, 24 % des jeunes avaient de l’embonpoint ou étaient obèses, par rapport à 26 % en 2005. Chez les adultes, la prévalence de l’embonpoint/ obésité était de 74 % en 1991, par rapport à 68 % en 2005. Les différences n’étaient significatives ni chez les jeunes ni chez les adultes. Le TT augmentait avec l’âge dans les deux sexes. Chez les femmes, le TT moyen dans tous les groupes d’âge était supérieur au seuil recommandé dans les lignes directrices canadiennes; chez les hommes, le TT était inférieur au seuil recommandé dans tous les groupes d’âge, sauf chez les hommes de 65 ans et plus.
C O N C L U S I O N S : High prevalence rates of overweight/
obesity among adults remained over time, particularly among females. The concern for unhealthy body weights among females suggests intervention programs targeting women are needed. Future research exploring communityspecific causes of overweight/obesity to identify local belief systems and lifestyle practices is required. K E y w O R D S : Aboriginal, adolescents, adults, body mass
index, obesity, overweight
C O N C L U S I O N S : La prévalence de l’embonpoint/obésité
est demeurée élevée au fil des ans, surtout chez les femmes. Les inquiétudes liées à la prévalence de l’embonpoint/ obésité chez les femmes font ressortir la nécessité de programmes d’intervention s’adressant aux femmes. Des recherches devront être menées pour déterminer les causes
address for correspondence: Brenda Bruner, School of Kinesiology and Health Studies Queen’s University, 69 Union Street, PEC 207, Kingston, Ontario, Canada, K7L 3N6 Telephone: (613) 533-6000 ext. 75210, Fax: (613) 533-2009, E-mail: [email protected] CANADIAN JOURNAL OF DIABETES. 2009;33(2):105-113.
| 105
106 |
CANADIAN JOURNAL OF DIABETES
de l’embonpoint/obésité à l’échelle communautaire, soit les croyances et les habitudes de vie qui y contribuent. M O T S C L É S : embonpoint, obésité, indice de masse cor-
porelle, autochtone, adolescents, adultes
INTRODUCTION It is well recognized that the global increases of overweight and obesity are public health concerns, and given their substantial collective increase in prevalence are regarded as a pandemic affecting millions worldwide (1). The most recent Canadian data indicate that 36.1% of adults are overweight (i.e. body mass index [BMI] 25 to 29.9 kg/m2) and an additional 23.1% are obese (i.e. BMI >30 kg/m2), a substantial increase from the 13.8% reported 25 years ago (2). One of the most notable increases is among those aged 25 to 34 years, where obesity rates more than doubled, from 8.5% in 1978/79 to 20.5% in 2004 (2). While these data revealed similar rates of obesity for males (22.9%) and females (23.2%), females were more likely to have class III obesity (BMI >40 kg/m2; females 4% vs. males 2%) (2). Overweight and obesity have been shown to be strongly associated with chronic diseases such as type 2 diabetes mellitus, cardiovascular disease and some cancers (3). However, type 2 diabetes appears to be the most directly related to increasing obesity (4), with the prevalence of type 2 diabetes also increasing worldwide (5). While the Canadian data for overweight and obesity prevalence trends were derived from a target population that included long-term residents in health institutions and household residents in all provinces and territories, 1 notable group omitted from these population studies is those living on First Nations reserves. Although the Canadian data suggest higher rates of obesity in the Aboriginal population (based on the notion that Aboriginal peoples comprise a large percentage of the northern population), data on the prevalence of overweight and obesity in this population are limited (3). (Note: the terms Aboriginal and Aboriginal peoples do not describe a homogeneous group. However, for the purpose of this paper, the terms will be used to describe First Nations Status, Non-Status, Métis and Inuit peoples, with the exception of the use of specific terms when discussing the results of research studies that have employed these terms.) To date, Canadian research has focused primarily on the pediatric population (6–8) or has been community-specific to Manitoba and Northern Ontario (9,10). Nevertheless, results from these studies have reported an increased prevalence of obesity in the Aboriginal population. These findings have also highlighted a predominantly central pattern of obesity (9), associated with an increased risk for type 2 diabetes (11), of particular concern for Aboriginal peoples, who are already at higher risk (12).
One of the few studies to determine rates of obesity among the adult Aboriginal population was conducted in 3 northern Saskatchewan communities with varying accessibility to urban centres (13). The results of this study showed that those with increased accessibility to urban centres also showed an increased BMI (13). For this particular study, obesity was defined as >10% above the Canadian mean; according to this criterion, approximately 45% of males and 65% of females in the “most accessible” community were classified as obese — proportions that were greater than for the 2 communities with less urban accessibility (13). Given the reported high prevalence of obesity among Aboriginal peoples, the purpose of this study was to determine current overweight and obesity rates in the “most accessible” Aboriginal community in the 1991 study (13). This study also afforded the unique opportunity to compare the current rates of overweight and obesity with those from 1991 (14). Therefore, the second purpose of this study was to determine if there had been an increase in rates of overweight and obesity over time.
METhODS research project approach Historically, there has often been a “helicopter approach” (15) to research in Aboriginal communities, whereby investigators arrive in a community, collect the information they require with minimal interaction with members of the community and publish their findings in scientific journals without input from the community where the research was based. This “style” of research has resulted in distrust and resistance on the part of Aboriginal communities in relation to university-based researchers, given the lack of community control, lack of local benefits and interpretation of the data in isolation of the social context (16). In keeping with more recently developed protocols and ethical guidelines for working with Aboriginal communities (17,18), this project began by establishing a partnership with the health services organization representing the Band and health director in the community. In addition, initial exploratory meetings were held with community members who expressed an interest in participating in a follow-up project on healthy body weights. The involvement of community members is essential for community-based research to be successful, as the community is seen as an integral component of the planning, development, implementation and evaluation of health-promotion strategies. As a result, a participatory approach based on the principles of community development was applied to this project. This approach has been increasingly accepted in research projects involving Aboriginal communities (15). Subsequently, a community project worker was hired through the health clinic to assist with data collection and project organization. Regularly
Prevalence of overweight and obesity in a woodland cree community: 14 year trends
scheduled visits were made by the university researchers to establish rapport and trust with employees at the health clinic and members of the community, as well as to help develop and plan the research project.
consent Informed consent for this project was developed in stages. The first stage involved obtaining consent from local governance, including board members, health portfolio councillors, the health committee and appropriate community leaders and elders. The details of the proposed project were outlined, including the purpose and scope of the study, proposed methods to obtain individual informed consent and how the information gathered would be shared with the community. Approval for the community’s participation in the project was established prior to obtaining individual consent. Once all details of the project were mutually agreed upon and community consent was granted, informed consent was obtained from the University Advisory Committee on Ethics in Behavioral Science Research, following which informed consent was obtained from each participant. The process of obtaining individual consent was approved by the health director, and all details of the consent were done in collaboration and consultation with the health director and the community project worker to ensure appropriate language and content. the community The setting for this research project was a reserve community in northern Saskatchewan, located approximately 500 km northeast of a major Saskatchewan city and 140 km west of the Manitoba border. The community is accessible throughout the year by a gravel road approximately 30 km off a main highway. At the time of data collection (September 2004 to May 2005), the population was estimated to be 972 (48% female, 52% male), with 60% under the age of 25 years. Healthcare in the community is delivered through a local health centre, which offers a variety of health programs, treatment services and 24-hour emergency care, provided by nurses who live in the community. A family physician travels to the community 2 days per week, and specialist referrals are made to the most accessible urban centre. Participants All individuals aged 7 years and older were eligible to participate. A number of strategies were used to obtain a representative sample of individuals and to maximize the number of participants in the project. Recruitment strategies were facilitated through the health clinic and included 2 health fairs, weekly health clinic programs and personal invitations to those who may have been overlooked in previous recruit-
ment methods. These individuals were invited to attend the clinic at their convenience to participate in the project.
Procedures Anthropometric measurements included waist circumference (WC), and height and weight to calculate body mass index (BMI). All measures were obtained using standard procedures (19), with the participants dressed in light clothing and without shoes. Height was determined using a wall-mounted tape measure and set square, and weight was measured using a standing floor scale (Detecto, Webb City, MO). WC was measured by positioning a tape measure horizontally at the level of noticeable waist narrowing, and the measurement was read at the end of a normal expiration. When the point of narrowing could not be found, an indeterminate waist was approximated by finding the lateral level of the 12th or lower floating rib and the girth was recorded at that site (19). Waist measurements were taken twice and recorded to the nearest 0.5 cm. When the difference between the first and second measure was greater than 0.5 cm, a third measurement was taken, and the average was determined from all 3 measurements. WC was measured only in 2005. The data were collected between September 2004 and May 2005. data analyses The sample was divided into two age cohorts: youth aged 7 to 17 years and adults aged ≥18 years. The BMI classifications used for overweight and obesity for adults were those outlined by Health Canada and the Obesity Canada Clinical Practice Guidelines Expert Panel, who define overweight as BMI ≥25.0 kg/m2 and obesity as ≥30.0 kg/m2 (20,21). Given that these guidelines are not intended for use with individuals under the age of 18 years, the classification for overweight and obesity for youth aged 7 to 17 years was based on the age-specific BMI cutoffs for children from the International Obesity Task Force (22). These cutoffs were derived from a large international sample by constructing centile curves for BMI that correspond to the adult cutoffs for overweight (25 kg/m2) and obesity (30 kg/m2) (22). Health risk associated with increased WC for adults was defined according to the Canadian Guidelines for Body Weight Classification in Adults (20). Analyses were performed using SPSS version 14.0 (SPSS Inc. Chicago, IL) and SAS version 9.0 (SAS Institute Inc., Cary, NC). Rates of overweight and obesity and mean BMI for each time point were determined by descriptive statistics. As the sample was partially dependent and partially independent (i.e. some participants were measured at both time points), generalized estimating equations (GEE) were used to test for differences in prevalence rates of overweight and obesity between the 2 time points (1991 and 2005). GEE is a CANADIAN JOURNAL OF DIABETES. 2009;33(2):105-113.
| 107
CANADIAN JOURNAL OF DIABETES
RESULTS A total of 289 male and female participants between the ages of 7 and 82 years (mean age 31.0±16.5 years) participated in this study. This included 81 youth (mean age 12.7±2.9 years) and 208 adults (mean age 38.1±14.0 years) (Table 1). The data for this study were compared to those from the 316 male and female participants between the ages of 7 and 78 years (mean age 24.6±15.6) who participated in the 1991 study; 138 youth (mean age 11.3±3.2 years) and 178 adults (mean age 34.8±13.4 years) (Table 1). As shown in Table 1, the total sample size in 2005 was slightly smaller and participants were slightly older than in 1991. In addition, a greater proportion of females participated in the follow-up study in 2005; the proportion of female and male participants in 1991 was roughly equal (Table 1). Table 1. Comparison of male and female adult and youth participants, 1991 and 2005 1991
2005
N=316
n=289
24.6±15.6
31.0±16.5
Males, n (%)
158 (50)
108 (37)
Females, n (%)
158 (50)
181 (63)
Youth (7–17 y)
N=138
n=81
Mean age, y (±SD)
11.3±3.2
12.7±16.5
Males, n (%)
68 (49)
30 (37)
Females, n (%)
70 (51)
51 (63)
Adults (18+ y)
N=178
n=208
34.8±13.4
38.1±14.0
Males, n (%)
90 (51)
78 (38)
Females, n (%)
88 (49)
130 (62)
Overall Mean age, y±SD
Mean age, y (±SD)
youth (7 to 17 years) Figure 1a shows that in 1991 nearly one-quarter (24%) of the youth population was overweight/obese (23% females and 25% males); however, there was no significant difference in rates of overweight/obesity over time (p=0.74) as
26% of the sample was overweight/obese in 2005 (32% females and 17% males). When stratified by gender, there was also no significant difference in rates of overweight/ obesity over time for males (p=0.37) or females (p=0.30). Although the overall differences were not significant, we felt it was important to indicate whether a specific age group influenced the change in overweight/obesity rates over time, particularly among the female youth. Given the small sample sizes for individual ages, the data were stratified by age group (7 to 11 years and 12 to 17 years) and are reported as the prevalence of normal weight, overweight and obesity for each age group (Table 2). Based on the criteria outlined by Cole and colleagues (2000), Table 2 shows that Figure 1. (A) youth (7–17 years)* and (B) adults (≥18 years) classified as overweight/ obese, † 1991 and 2005
A 35
1991 2005
30
Percent overweight/obese
statistical technique used to analyze dichotomous outcomes of longitudinal correlated data (23). Therefore, the raw BMI data for the GEE analysis was dichotomized into 2 categories: overweight or obese and not overweight or obese. A total of 77 individuals who had anthropometry data for both time points were included in this analysis. A subanalysis using the Kappa statistic was used to assess change in BMI classification (i.e. normal weight, overweight, obese) over time in the individuals with repeated data.
25
20
15
10
5
0
All
Males
Females
B 100
Percent overweight/obese
108 |
1991 2005
80
‡
60
40
20
0
All
Males
*Base on age-specific cutoffs † BMI ≥25.0 kg/m2 ‡ Significantly different from 1991; p<0.05
Females
Prevalence of overweight and obesity in a woodland cree community: 14 year trends
Table 2. Body weight classification among youth, 1991 and 2005 Males
Females
Age group, y
1991
2005
1991
2005
7–11 y
n=33
n=8
n=43
n=20
Normal weight, %
72.7
87.5
81.4
55.0
Overweight, %
24.2
12.5
11.6
15.0
3.0
0
7.0
30.0
n=35
n=22
n=27
n=31
Normal weight, %
77.1
81.8
70.4
74.2
Overweight, %
20.0
9.1
18.5
12.9
2.9
9.1
11.1
12.9
Obese, % 12–17 y
Obese, %
among males aged 7 to 11 years, there was a decrease in the proportion of overweight and obesity from 1991 to 2005, whereas a greater proportion of females were overweight and obese in 2005 than in 1991. Among youth aged 12 to 17 years, there was a decrease in the proportion of both males and females who were overweight, but an increase in the proportion of males who were obese.
adults (≥18 years) The mean age of adult participants in 1991 was 34.8±13.4 years, slightly lower than the mean age in 2005 (38.1±14.0 years); however, this difference was not significant. Similar to the findings reported among youth in the community, there was no significant difference in overall prevalence of overweight/obesity over time (p=0.59). However, Figure 1b shows that three-quarters (74%) of the population in 1991 were overweight/obese (10% class II and 3% class III obese) and over two-thirds (68%) were overweight/obese in 2005 (14% class II and 1% class III obese). When stratified by
gender, there was a significant decrease (p=0.04) in rates of overweight/obesity over time for males (74 to 57%) but not for females (p=0.20), whose prevalence rates remained essentially unchanged (74 to 75%) (Figure 1b). As shown in Table 3, BMI tended to increase with age in both 1991 and 2005. The mean BMIs for all age groups at both time points are consistent with the criteria for either overweight or obesity (20,24). When stratified by gender, the mean BMIs for all age groups were classified as overweight or obese with the exception of males aged 18 to 34 years in 2005. Mean BMI values for females aged 18 to 34 years in both 1991 and 2005 were above the Canadian guidelines for overweight; for all other female age groups, mean BMI values exceeded the cutoffs for obesity. Overall, the mean WC for females was 102.1±12.5 cm (46% >88 cm) and for males was 96.2±13.0 cm (9% >102 cm). As shown in Figure 2, WC increased with age for both males and females. The mean WC for all female age groups were above the recommended Canadian guidelines of 88 cm (20,24), while the mean WC for all male age groups, with the exception of those aged ≥65, were below the guidelines of 102 cm (20,24).
Paired data A separate subanalysis was performed on paired data from 77 individuals (39% males and 61% females) measured in both 1991 and 2005. The data revealed that of the 35 participants classified as normal weight in 1991, over half were either overweight (42.9%) or obese (8.6%) in 2005. Of the 23 participants classified as overweight in 1991, 52.2% remained overweight, 39.1% were obese and 8.7% were normal weight in 2005. Finally, of the 19 participants who were obese in 1991, all remained obese in 2005, with the exception of 2 who were reclassified as overweight. Overall, 35% of the sample had an increase in BMI classification; of these, 67% were female. A larger proportion (45.5%) of
Table 3. Body mass index among adults, 1991 and 2005 Overall Age group, y
Males
Females
1991
2005
1991
2005
1991
2005
18–34
n=105 27.2±4.3
n=91 26.2±5.0
n=52 26.6±3.3
n=34 24.6±5.2
n=53 27.9±5.1
n=57 27.2±4.7
35–49
n=48 30.2±5.9
n=74 29.9±5.2
n=23 30.1±6.4
n=25 28.7±4.3
n=25 30.3±5.6
n=49 30.6±5.5
50–64
n=16 31.0±4.5
n=32 29.5±5.4
n=9 30.7±3.9
n=16 28.4±5.6
n=7 31.2±5.6
n=16 30.6±5.2
≥65
n=9 31.8±4.8
n=11 31.1±6.5
n=6 30.3±3.8
n=3 30.3±6.8
n=3 34.9±6.1
n=8 31.5±6.8
Values are mean±SD
CANADIAN JOURNAL OF DIABETES. 2009;33(2):105-113.
| 109
CANADIAN JOURNAL OF DIABETES
Figure 2. Mean waist circumference by age group, 2005 120
100
Waist circumference (cm)
110 |
80
60
40
20
0
Males 18–34 y
Females 35–49 y
50–64 y
≥65 y
*BMI ≥25.0 kg/m2 Significantly different from 1991; p<0.05
†
youth participants (7 to 17 years) in 1991 had an increase in BMI classification over time compared to those who were adults (31%). A kappa test was estimated as 0.410 and indicated that there was a significant change in the classification of BMI between the 2 time points (p<0.001).
DISCUSSION The results of this study revealed there was a significant decrease in rates of overweight/obesity among adult males in the community, but no significant difference for youth or adult females over time. Although these findings were not significant, there is some indication that overweight/obesity has increased among the female youth, specifically those aged 7 to 11 years, as reflected in the increased prevalence of overweight/obesity from 1991 to 2005. Although the prevalence of overweight/obesity among adult males decreased, there was relatively little change among females, which is not surprising given the higher prevalence of overweight/ obesity when measured initially. The findings also show that in the subsample of participants with paired data, there was a significant increase in BMI classification over time. More specifically related to the younger population, it appears that the risk for unhealthy body weight is greater for females and may begin at an earlier age compared to their male counterparts. This is of concern because Aboriginal ethnicity has been shown to be an independent predictor of gestational diabetes when it interacts with pre-pregnancy obesity (25). Obesity increases the risk of developing gestational diabetes (26), and a woman who has had gestational
diabetes is at a high risk of perpetuating the cycle by giving birth to an infant with a high likelihood of becoming obese and developing type 2 diabetes at an early age (26,27). In addition, women with gestational diabetes have up to a 70% chance of developing type 2 diabetes later in life (28) and a 50% risk of developing type 2 diabetes within 5 years postpartum (26). The prevalence rates among youth at both time points were lower than those reported in other community-specific surveys involving Aboriginal youth (7,10,29,30). However, one similarity between these studies and the current study is that a greater proportion of females were overweight and/or obese compared to their male counterparts. One exception is the data reported by Katzmarzyk and Malina, where rates of obesity were similar for males and females (28.6 and 29.4%, respectively) (10). It should be noted that the reference data used to define overweight and obesity in these studies were based on the 85th and 95th percentiles from the National Health and Nutrition Examination Survey (NHANES) II and III, but the classifications were not consistent across all studies. For example, some studies defined overweight as ≥85th percentile (7,29,30), while Katzmarzyk and Malina used the same criteria (i.e. >85th percentile) to define obesity (10) . Given that the reference standards for defining overweight and obesity have varied between studies, direct comparisons should be made with caution. What is important to note, however, is that regardless of the reference standards used and communities or First Nations groups assessed (e.g. Mohawk, Oji-Cree, Woodland Cree), the findings of this study and others have reported a high proportion of children and youth classified as overweight or obese, particularly females. These findings are important given that obesity is believed to track from childhood into adulthood (31,32). Furthermore, the duration of obesity is thought to increase the risk of developing type 2 diabetes (33), thus increasing health risks as the Aboriginal population ages. In this regard, the serial data, which showed a significant increase in BMI classification over time, highlight the importance of continued surveillance and early intervention strategies. Decreasing rates of overweight and obesity could reduce the associated long-term health consequences within our First Nations communities. Within the adult population, overweight and obesity continue to remain a special concern for females, for whom mean BMIs in all age groups were classified as obese. In Canada, there have been few studies assessing prevalence rates of overweight and obesity in the adult Aboriginal population. However, of those that have been published, Katzmarzyk and Malina reported the prevalence of obesity among Anishabai Temagami First Nation males and females to be 51% and 60%, respectively (10). Although these results are higher than our findings for both males and females
Prevalence of overweight and obesity in a woodland cree community: 14 year trends
at both time points, it should be noted that Katzmarzyk and Malina defined obesity as a BMI ≥27.8 kg/m2 for males and ≥27.3 kg/m2 for females (10), the criteria outlined in the 1988 Canadian Guidelines for Healthy Weights (34). Since the present study used the current guidelines defining overweight as a BMI ≥25.0 kg/m2 and obesity as ≥30.0 kg/m2 for both males and females (20,21), this again limits direct comparison between studies. However, similar to the data reported among the youth, higher rates of obesity are demonstrated in both studies of Aboriginal females compared to males. Compared to our study, earlier data from 6 First Nations communities in northwestern Ontario and northeastern Manitoba also reported higher combined rates of overweight/obesity among both males (40 to 80%) and females (40 to 90%) ranging in age from 20 to 64 years (9). The criteria used to classify overweight and obesity were similar to the present study (i.e. overweight ≥26.0 kg/m2 and obesity ≥30.0 kg/m2). While BMI is a practical method for measuring and classifying overweight and obesity, it does not measure the pattern of body adiposity that is a risk factor for disease, independent of level of obesity (32). More specifically, abdominal fat distribution is associated with an increased risk of cardiovascular disease and type 2 diabetes (35). In adults, WC is commonly used to measure abdominal obesity (32,35) and, thus, health risk. Few studies have assessed WC among Canadian Aboriginal peoples that can be used for comparisons to the current data. In Sandy Lake, similar results to our data for males (i.e. mean WC was <102 cm) have been reported (36). However, the data for females was somewhat different. Although both studies report mean WC >88 cm, the mean WC in this study was over 10 cm higher than that reported by Hanley and colleagues (102.1±12.5 cm vs. 90.6±13.8 cm, respectively) (36). While WC values were not reported by Young and Sevenhuysen, the authors did report that the pattern of obesity was centrally located (9). Contrary to the current findings, however, this was higher among males compared to females. The results from this study suggest that females in all adult age ranges (i.e. 18 years and older) are at an increased risk of developing health problems, as the mean WC values were greater than the Canadian guidelines for women of 88 cm (20). As WC increases with age, these results further highlight the health risks for women in this community and the need for intervention programs targeting unhealthy body weights. Although there are limited Canadian data regarding the prevalence of overweight and obesity among First Nations groups, the results of this study support findings from previous studies, which suggest that high levels of overweight and obesity exist in First Nations communities. While there was no significant increase in overweight/obesity over time using a cross-sectional sample in the community, the results
from the paired data showed a significant difference in the number of individuals with an increase in BMI classification. The findings of this study and other Canadian studies reporting high rates of overweight and obesity, particularly among Aboriginal females, highlights the need for intervention strategies targeting unhealthy body weights. The results of this study and previous literature also emphasize the importance of developing surveillance measures to document changes in age-specific prevalence rates of overweight and obesity among various First Nations groups to determine the effectiveness of intervention programs. This study contributes new knowledge regarding overweight and obesity among Aboriginal people in Canada. One of the strengths of this study is the use of measured height, weight and WC. Self-reported values tend to underestimate weight, particularly among heavier participants, while height tends to be over-estimated (32). An additional strength is the follow-up nature of the study. As limited surveillance data exist, this study is a first step in tracking overweight and obesity at the community level, with future research attempting to obtain measurement data from individuals participating in the previous studies. The limitations of this study must also be acknowledged. The first is the cross-sectional nature of this research. As in many studies conducted within First Nations communities (7–10), all eligible individuals in the community were invited to participate, but many did not. This may have resulted in undersampling of certain age groups and may limit the sample representativeness. However, it should be noted that the proportion of individuals in each age group in this study was reflective of the demographics of eligible participants in the community (i.e. a greater proportion ≥18 years of age). Second, although we did not assess changes in socioeconomic status over time, a crude comparison of the 1991 and 2001 census data showed an increase in median household income ($16 928 to $23 915), still well below the provincial median ($40 251). This may influence health-related determinants that impact overweight and obesity, such as diet and physical activity, as socioeconomic status has been shown to be inversely related to healthy lifestyles (37). Third, we encountered a limited ability to compare results to other studies. In this study, we elected to use the cutoffs for overweight and obesity derived by Cole and colleagues (22) for the youth sample, given the lack of an accepted definition of obesity in the pediatric population (22), and because this approach is said to avoid arbitrarily choosing reference data (i.e. National Health and Nutrition Examination [NHANES] surveys) and cutoff points for overweight and obesity (e.g. 85th and 95th percentiles, respectively). Additionally, the cutoffs derived by Cole and colleagues are based on a worldwide population and correspond to the adult cutoff points of overweight CANADIAN JOURNAL OF DIABETES. 2009;33(2):105-113.
| 111
112 |
CANADIAN JOURNAL OF DIABETES
and obesity, the latter allowing us to examine differences in BMI in the subsample over time. Of note, we did analyze the data using the BMI percentiles, and the proportion of individuals classified as overweight and obese were similar with both methods. While the Canadian Guidelines for Body Weight Classification in Adults (20) are said to be appropriate for all racial/ethnic groups in Canada, further research is still required to determine if the cutoffs used for the general population are appropriate for Aboriginal populations (38). Finally, the procedure used to measure WC in this study (i.e. minimal waist) is commonly used in clinical and research settings; however, it should be noted that to date there is no consensus regarding the optimal protocol to measure WC. A recent review by Ross and colleagues identified 8 protocols, of which 3 (umbilicus, midpoint between the superior border of the iliac crest and lowest rib, and minimal waist) predominate in the literature (39). While the 2008 Canadian clinical practice guidelines recommends obtaining WC measures at the top of the iliac crest (21), the data for this study was collected prior to the release of the guidelines. Of note, Wang and colleagues compared WC measures taken at 4 sites, including the minimal waist (40). The results showed that the WC measure was significantly smaller at the minimal waist compared to the other 4 sites (40), suggesting that our results may underestimate the WC measures in our sample.
CONCLUSIONS The results of this study support previous literature showing high rates of overweight and obesity among Aboriginal peoples. The findings also show that females, whether youth or adult, have higher rates of overweight/obesity compared to their male counterparts, which increases their risk of developing associated health problems. Gestational diabetes and type 2 diabetes are of particular concern given that obese women are at increased risk of developing gestational diabetes, thus, increasing the risk of both the mother and offspring developing type 2 diabetes later in life. The high proportion of females classified as overweight/obese highlights the need for intervention programs targeting this subgroup, particularly young women in the community. While the cross-sectional data showed no significant increase in overweight/obesity over time, we did demonstrate a significant increase in BMI classification in the individuals measured in both 1991 and 2005. Since obesity tracks from childhood to adulthood, prevention programs, targeting unhealthy body weights in Aboriginal communities should begin at a young age. Prior to developing such programs, however, future research should be undertaken to gain an understanding of the local belief systems (41) and practices regarding modifiable risk factors for overweight and obesity, such as physical activity and dietary practices
ACKNOwEDgEMENTS We wish to acknowledge the Peter Ballantyne Cree Nation Health Services Inc, the chief and council of the community, health clinic staff, as well as Ms. Jane Ballantyne and Ms. Janet Ballantyne, for their cooperation with, assistance with and support of this project. This project was supported through a research grant from the Canadian Institutes of Health Research and the Heart and Stroke Foundation of Canada. We would also like to acknowledge Josh Lawson who contributed to the statistical analysis.
DUALITIES OF INTEREST No dualities declared.
AUThOR CONTRIBUTIONS KC and RD obtained grant funding. BB and KC created the study design. BB wrote the first draft of the manuscript. BB and JL conducted the statistical analysis. BB conducted the data collection and management in conjunction with the community. All authors contributed to the interpretation of results and review of the final manuscript.
REFERENCES 1. World Health Organization. Obesity: Preventing and managing the global epidemic. Geneva, Switzerland; WHO: 2000. 2. Tjepkema M. Measured adult obesity in Canada: Measured height and weight. In Nutrition: Findings from the Canadian Community Health Survey: Statistics Canada. Ottawa, ON; 2005:1-32. 3. Raine K. Overweight and obesity in Canada: A population health perspective. Ottawa, ON: Canadian Institute for Health Information; 2004. 4. Jung R. Obesity as a disease. Br Med Bull. 1997;53:307-321. 5. Wild S, Roglic G, Green A, et al. Global prevalence of diabetes. Diabetes Care. 2004;27:1047-1053. 6. Bernard L, Lavallee C, Gray-Donald K, et al. Overweight in Cree schoolchildren and adolescents associated with diet, low physical activity, and high television viewing. J Am Diet Assoc. 1995; 95:800-802. 7. Hanley A, Harris S, Gittelsohn J, et al. Overweight among children and adolescents in a native Canadian community: Prevalence and associated risk factors. Am J Clin Nutr. 2000;71: 693-700. 8. Potvin L, Desrosiers S, Trifonopoulos M, et al. Anthropometric characteristics of Mohawk children aged 6 to 11 years: a population perspective. J Am Diet Assoc. 1999;99:955-961. 9. Young T, Sevenhuysen G. Obesity in northern Canadian Indians: patterns, determinants, and consequences. Am J Clin Nutr. 1989;49:786-793. 10. Katzmarzyk P, Malina R. Obesity and relative subcutaneous fat distribution among Canadians of first nation and European ancestry. Int J Obes Relat Metab Disord. 1998;22:1127-1131.
Prevalence of overweight and obesity in a woodland cree community: 14 year trends
11. Young T, Reading J, Elias B, et al. Type 2 diabetes mellitus in Canada’s first nations: Status of an epidemic in progress. CMAJ. 2000;163:561-566. 12. Health Canada. Center for Chronic Disease Prevention and Control, Population and Public Health Branch. Diabetes in Canada. 2nd ed. Ottawa, ON: 2002. 13. Dyck RF, Tan L, Hoeppner VL. Body mass index, gestational diabetes and diabetes mellitus in three northern Saskatchewan aboriginal communities. Chronic Dis Can. 1995;16:24-26. 14. Dyck RF, Tan L, Hoeppner VL. Brief communication: relationship of body mass index and self-reported rates of gestational diabetes and diabetes mellitus with geographic accessibility: a comparison of three northern Saskatchewan aboriginal communities. Chronic Dis Can. 1995;16:24-26. 15. Smylie J, Kaplan-Myrth N, Tait C, et al. Health sciences research and aboriginal communities: pathway or pitfall? J Obstet Gynaecol Can. 2004;26:211-216. 16. Henderson R, Simmons D, Bourke L, et al. Development of guidelines for non-indigenous people undertaking research among the indigenous population of north-east Victoria. MJA. 2002;176:482-485. 17. American Academy of Pediatrics, Ethics considerations in research with socially identifiable populations. Pediatrics. 2004;113:148-151. 18. Tricouncil Policy Statement. Research involving aboriginal peoples. In: Ethical Conduct for Research Involving Humans. Ottawa, ON; 1998:6.1-6.4. 19. Canadian Society for Exercise Physiology. The Canadian physical activity, fitness and lifestyle appraisal resource manual. Ottawa, ON; 1996:7.12-7.18. 20. Health Canada. Office of Nutrition, Policy and Promotion. Canadian Guidelines for Body Weight Classification in Adults. Ottawa, ON; 2003. 21. Depres J, Tchernoff A. Classification of overweight and obesity in adults. CMAJ. 2007;176(suppl 8):S21-26. 22. Cole T, Bellizzi M, Flegal K, et al. Establishing a standard definition for child overweight and obesity worldwide: International survey. BMJ. 2000;320:1240-1243. 23. Hanley J, Negassa A, Edwardes M, et al. Statistical analysis of correlated data using generalized estimating equations: an orientation. Am J Epidemiol. 2003;157:364-375. 24. Lau D, Douketis J, Morrison K, et al. Executive summary: 2006 Canadian clinical practice guidelines on the management and prevention of obesity in adults and children. CMAJ. 2007;176(suppl 8):S1-13. 25. Dyck R, Klomp H, Tan L, et al. A comparison of rates, risk factors, and outcomes of gestational diabetes between aboriginal and non-aboriginal women in Saskatoon district health. Diabetes Care. 2002;25:487-493. 26. Inzucchi S. Diabetes in pregnancy. In: Burrow G., ed. Medical Complications in Pregnancy. Philadelphia, PA: W.B. Saunders Co; 1999:25-52.
27. Dabelea D, Pettitt DJ. Intrauterine diabetic environment confers risks for type 2 diabetes and obesity in the offspring, in addition to genetic susceptibility. J Pediatr Endocrinol Metab. 2001;14:1085-1091. 28. Kim C, Newton K, Knopp R. Gestational diabetes and the incidence of type 2 diabetes. Diabetes Care. 2002;25:1862-1868. 29. Young TK, Dean HJ, Flett B, et al. Childhood obesity in a population at high risk for type 2 diabetes. J Pediatr. 2000;136: 365-369. 30. Trifonopoulos M. Anthropometry and diet of Mohawk schoolchildren in Kahnawake [dissertation]. Montreal, QC; 1995. 31. Serdula M, Ivery D, Coates R, et al. Do obese children become obese adults. A review of the literature. Prev Med. 1993;22: 167-177. 32. Power C, Lake J, Cole T. Measurement and long-term health risks of child and adolescent fatness. Int J Obes Relat Metab Disord. 1997;21:507-526. 33. Everhart J, Pettitt D, Bennett P, et al. Duration of obesity increases the incidence of NIDDM. Diabetes. 1992;41:235-240. 34. Health and Welfare Canada. Health Services and Health Promotion Branch. Canadian guidelines for healthy weights. In: Report of an Expert Group Convened by Health Promotion Directorate. Ottawa, ON; 1988. 35. Snijder M, Van Dam R, Visser, et al. What aspects of body fat are particularly hazardous and how do we measure them? Int J Epidemiol. 2006;35:83-92. 36. Hanley A, Connelly P, Harris S, et al. Adiponectin in a native Canadian population experiencing rapid epidemiological transition. Diabetes Care. 2003;26:3219-3225. 37. Canadian Institute for Health Information. Improving the Health of Canadians: Income. Ottawa, ON; 2004. 38. Douketis J, Paradis G, Keller H, Martineau C. Canadian guidelines for body weight classification in adults: Application in clinical practice to screen for overweight and obesity and to assess disease risk. CMAJ. 2005;172:995-998. 39. Ross R, Berentzen T, Bradshaw A, et al. Does the relationship between waist circumference, morbidity and mortality depend on measurement protocol for waist circumference? Obes Rev. 2007;9:312-325. 40. Wang J, Thornton J, Bari S, et al. Comparison of waist circumferences measured at 4 sites. Am J Clin Nutr. 2002;77: 378-384. 41. Willows ND. Overweight in first nations children: Prevalence, implications, and solutions. Journal of Aboriginal Health. 2005; 2:76-85.
CANADIAN JOURNAL OF DIABETES. 2009;33(2):105-113.
| 113