Weight Status and Determinants of Health In Manitoba Children and Youth

ABSTRACT

Purpose: Because of the tremendous increase in overweight and obesity in Canadian children and youth in recent decades, we examined associations among health determinants, healthy living characteristics, and overweight and obesity in Manitoba children and youth.

Methods: Using descriptive statistics and logistic regression, we identified factors associated with measured overweight and obesity in a sample of 1651 Manitoba children and youth aged two to 17 years from the 2004 Canadian Community Health Survey 2.2-Nutrition.

Results: Thirty-one percent of the children and youth were overweight or obese. Males aged 12 to 17 or from foodinsecure homes were more likely to be overweight or obese than were younger males or males from food-secure households. Females from households with higher parental education were less likely to be overweight or obese than were those from households with lower parental education. Female youth who were sedentary for at least three hours daily were more likely to be overweight or obese than were less sedentary female youth. A trend toward significance with overweight or obesity in youth was noted with levels of daily fruit and vegetable consumption and regular physical activity.

Conclusions: Overweight and obesity in Manitoba children and youth are associated with socio-economic and demographic characteristics, and with food and activity behaviours. These findings can inform health and nutrition policy and practice by indicating health inequities that require particular attention.

(Can J Diet Prac Res. 2010;71:115-121)

(DOI: 10.3148/71.3.2010.115)

R�SUM�

Objectif. Dans les r�centes d�cennies, l'incidence d'exc�s de poids et d'ob�sit� a augment� de fa�on fulgurante chez les enfants et les jeunes gens du Canada. Pour cette raison, nous avons examin� des associations entre les d�terminants de la sant�, les caract�ristiques d'une vie saine, et l'exc�s de poids et l'ob�sit� chez les enfants et les jeunes gens du Manitoba.

M�thodes. � l'aide de la statistique descriptive et de la r�gression logistique, nous avons identifi� des facteurs associ�s � la pr�sence d'un exc�s de poids et de l'ob�sit� dans un �chantillon de 1651 enfants et jeunes gens du Manitoba �g�s de 2 � 17 ans ayant particip� � l'Enqu�te sur la sant� dans les collectivit�s canadiennes, cycle 2.2, Nutrition (2004).

R�sultats. Trente et un pour cent des enfants et des jeunes gens pr�sentaient un exc�s de poids ou �taient ob�ses. Les jeunes de 12 � 17 ans de sexe masculin ou provenant de m�nages caract�ris�s par une incertitude alimentaire �taient plus susceptibles de pr�senter un exc�s de poids ou d'�tre ob�ses que les gar�ons moins �g�s ou les gar�ons provenant de m�nages o� il n'y avait pas d'incertitude alimentaire. Les filles provenant de m�nages o� le niveau d'�ducation des parents �tait plus �lev� �taient moins susceptibles de pr�senter un surplus de poids ou d'�tre ob�ses que celles provenant de m�nages o� les parents avaient moins d'�ducation. Les jeunes filles s�dentaires pendant au moins trois heures par jour �taient plus susceptibles de pr�senter un surplus de poids ou d'�tre ob�ses que celles moins s�dentaires. Une tendance a �t� observ�e entre la pr�valence d'exc�s de poids et d'ob�sit� chez les jeunes gens, et la quantit� de fruits et l�gumes consomm�e quotidiennement et la pratique r�guli�re d'une activit� physique.

Conclusions. L'exc�s de poids et l'ob�sit� chez les enfants et les jeunes gens du Manitoba sont associ�s � des facteurs d�mographiques et socio�conomiques ainsi qu'aux comportements li�s � l'alimentation et � l'activit�. Ces conclusions peuvent servir de guide pour la pratique et l'�laboration de politiques relatives � la nutrition et � la sant� puisqu'elles indiquent les in�galit�s sur le plan de la sant� qui requi�rent une attention particuli�re.

(Rev can prat rech di�t�t. 2010;71:115-121)

(DOI: 10.3148/71.3.2010.115)

INTRODUCTION

Overweight and obesity among Canadian children and youth have increased dramatically at national (1) and provincial levels (2). As overweight and obese children and youth often become overweight adults (3), identifying which groups are at greatest risk for becoming overweight or obese is critical.

Few investigators have explored the effects of age, sex, ethnicity, and determinants of health on childhood overweight and obesity using measured heights and weights (4,5). Rather, many authors have relied on self-report or parental report (6,7), which may underestimate overweight and obesity (8-11). For example, a 2.6 body mass index (BMI) difference has been found between measured BMI and self-reported BMI; the difference was most pronounced in females and younger participants (10), and youth in some ethnic groups-specifically African Americans (11) and whites (10,11)-are more likely to underestimate their BMI. Determination of weight status based on self-report may lead to an incorrect identification of groups at risk for overweight and obesity and/or an incorrectly report prevalence of overweight and obesity.

PURPOSE

Associations were examined between determinants of health associated with overweight and obesity in Manitoba children and youth, through the use of measured height and weight.

METHODS

The study included a sample of Manitoba children and youth from the 2004 Canadian Community Health Survey 2.2 (CCHS 2.2)-Nutrition. The survey methodology and sample design are detailed elsewhere (12). The Government of Manitoba purchased an augmented children and youth sample size, which allowed enhanced analysis, including intra-provincial comparisons. Response rates were 76.5% nationally and 82.7% in Manitoba. Parent or guardian proxies were obtained for children aged five years or younger who did not have to be present for the interview, except for height and weight measurements. Survey data for six- to 11-year-olds were collected with parental or guardian assistance. Youth provided information independently. For children and youth aged two to 17, measured heights and weights were obtained for 8661 individuals (66%) nationally and for 1172 of 1651 (71%) in Manitoba. The participating Manitoba children and youth (53.0% male, 21.4% aged two to five, 34.1% aged six to 11, and 44.6% aged 12 to 17) were subsequently classified according to the BMI cut-offs of the International Obesity Task Force (13). As with other reports (1), underweight was not excluded from normal weight.

Measures and confounders

The CCHS 2.2 data were collected throughout 2004 and 2005. While 21% of the data was collected from January to March, 25% in April to May, 29% in June to August, and 25% in September to December, the effects of seasonality were controlled by the survey design. Participants' geographic locations (northern, southeastern central, southwestern, and Winnipeg) were based on Manitoba administrative health boundaries. Selfor proxy-reported ethnicity was dichotomized as Aboriginal or non-Aboriginal. Information on parental education and family income was collected as proxies of socio-economic status (SES). Highest parental education was classified as less than high school, high school, post-secondary training excluding university, or university. Family income adequacy was based on total gross family income in relation to the number of household individuals, as per Statistics Canada recommendations (12). Food insecurity, defined as the limited or uncertain availability of nutritionally adequate foods or the limited or uncertain ability to acquire foods in socially acceptable ways (12), was determined according to a knowledgeable household member's response to 18 questions. These questions were designed to determine the financial ability to purchase food over the previous 12 months (14). They captured food insecurity in four categories: food secure, food insecure without hunger, food insecure with hunger, and food insecure with severe hunger. The proportion of food-insecure households (those with hunger or severe hunger) was very small, and this variable was therefore further dichotomized into the categories of food secure or food insecure with or without hunger. Notably, members of the same household may experience different food-security status (15).

Daily fruit and vegetable consumption was dichotomized as five or more or fewer than five times daily, according to the thencurrent Canada's Food Guide to Healthy Eating (16). For six- to 11-year-olds, physical activity (PA) was determined according to a parental estimate of the hours a day the child engaged in PA at and outside school; estimates were dichotomized into the categories of fewer than two or two or more hours a day. Sedentary behaviour (SB) of six- to 11-year-olds was determined using proxy reports of the hours a day children were sedentary outside school. This variable included behaviours such as computer and video game use and television viewing, and was grouped as two or fewer or more than two hours a day. (Reading was excluded from sedentary behaviours.) Among youth, PA frequency was determined according to self-reported times a month in the previous three months that they were physically active for more than 15 minutes. The monthly average was then classified as regular (12 or more times a month), occasional (four to 11 times a month), or infrequent (zero to three times a month) (17). Sedentary behaviour outside school or work was based on self-reported hours a week in the previous three months, and was classified as low (10 or fewer hours a week), medium low (10 to 19 hours a week), medium high (20 to 29 hours a week), and high (30 or more hours a week). This grouping permitted an exploration of the classification of SB beyond the recommended 90 minutes a day maximum (18). This was important because questionnaire data on SB, particularly among highly sedentary youth, lead to an underestimation of actual time spent in SB (19).

Data analysis

Statistical analyses were conducted using SAS 9.1 for Windows (SAS Institute, Cary, NC, 2004). Two estimation procedures were employed. First, appropriate statistical weight was used to ensure that the data were representative of the general population. Second, a design-based variance estimation was conducted via bootstrap technique to reflect the complex design of the CCHS 2.2 (20). This generated standard errors and coefficients of variation (CV). Results were not reported if the CV for an analysis was greater than 33.3. Results with a CV of 16.6 to 33.3 should be interpreted with caution. Any reported confidence intervals (CIs) refer to the 95% CI, which is equivalent to p<0.05.

Descriptive statistics and logistic regression analyses were employed for male and female samples separately to control for gender effect. Cross-tabulation produced estimates for the prevalence of healthy weights and overweight and/or obesity in Manitoba children and youth. Multivariate logistic regression further tested statistical significance for the likelihood of being overweight and/or obese (dependent variable). Age, sex, parental education, household income, food insecurity, ethnicity, PA, SB, fruit and vegetable consumption, and geographic location were analyzed as both risk factors (independent variables) and confounding variables. Our statistical hypotheses were that age would be an independent predictor of weight status and that SES, demographic, PA, SB, and fruit and vegetable consumption variables would significantly predict the prevalence of overweight or obesity in Manitoba children and youth. The age effect regression model was adjusted for all social, demographic, and economic variables. All other models were adjusted for age because of overlapping effects of social, demographic, and economic status of households and small sample sizes.

Physical activity and SB were measured differently in six- to 11-year-olds and youth because of significant age-related developmental differences. This part of the analysis was conducted separately for two distinct age groups. No PA or SB data were collected for children aged two to five.

RESULTS

The prevalence of overweight was approximately 22% for both sexes, while the prevalence of obesity was somewhat higher in females (9.4%) than in males (8.5%) (Table 1). Overweight and obesity were highest in northern Manitoba, and lowest in the southwest (p<0.03, north versus southwest). No significant sex difference in overweight was found. The prevalence of obesity in females remained stable until age 11, after which it increased. A greater proportion of youth (36%) than children aged two to five (23%) were either overweight or obese (p<0.003). A linear increase in overweight and/or obesity by age was significant for males only (odds ratio [OR]=1.52, p<0.05) (Table 2).

Determinants of health had an impact on overweight and obesity (Tables 1 and 2). Overweight and/or obesity decreased as parental education increased. Nearly 50% of children and youth whose parents did not complete high school were overweight or obese. This finding was statistically significant in females only (OR=0.58, p<0.05). Over 41% of children and youth from food-insecure households were either overweight or obese. This percentage was significantly higher than those from food-secure homes (p<0.05), although the association was driven by males (OR=2.09, p<0.05).

Gender differences were also noted (Table 2). Males' risk of obesity increased with age (OR=1.5, p=0.03) and food insecurity (p=0.04), and varied by health region (OR=1.2, p=0.05). In females, the risk of obesity was inversely associated with parental education (OR=0.6, p=0.04).

Healthy living refers to individual behaviours and choices that promote health. Several of these behaviours were associated with a lower prevalence of overweight and obesity. Children who consumed fruits and vegetables more frequently were significantly less likely to be overweight or obese (p<0.04) (Table 1). Further, a linear trend was found between the increasing prevalence of overweight and obesity and decreasing frequency of PA among male youth (Table 3). Regularly active males tended to be less likely to be overweight or obese than were less active males (OR=0.23, p=0.06). Highly sedentary females were more likely to be overweight or obese (OR=1.45, p<0.05) than were those who were less sedentary (Figure 1 and Table 3). The prevalence of overweight or obesity in the lower three levels of SB was comparable (approximately 30%). Conversely, over 50% of youth who reported SB for 30 or more hours a week were overweight or obese (p<0.05 versus the lower three levels of SB groups) (Figure 1). High versus low fruit and vegetable consumption had little impact on the rates of overweight and obesity in male youths: 39% versus 36%, respectively. However, high fruit and vegetable consumption tended to be protective against overweight and obesity in females (OR=0.38, p=0.06). Similar associations between healthy living behaviours and overweight or obesity were identified using single or multivariable models.

DISCUSSION

While most Manitoba children and youth were within a healthy weight range, approximately one-third were overweight or obese, which is significantly higher than the Canadian average (26%) for the same year (1). As in other studies (21,22), overweight and obesity were more prevalent among children and youth in lower SES or northern-dwelling households, and they increased with age. Accordingly, the need is especially urgent for interventions that are likely to have an impact on these populations.

Almost 50% of children and youth whose parents had not completed high school were overweight or obese. Although higher than the 2004 Canadian average of 31% (21), this prevalence parallels those in other Canadian studies (4,6,23). In Manitoba, parental education level more profoundly influenced overweight and obesity rates of girls than boys. Reasons for this sex-specific association are unclear. Perhaps parents with higher education affect their daughters' body image and healthy choices more than their sons'. As role models, mothers with a higher level of education may also affect their daughters more than their sons. Consequently, female youth may adjust their lifestyle, including PA, SB, and eating habits. Notably, similar findings were reported in a meta-analysis in which the association between overweight and SES was examined (22).

Northern-dwelling children had a higher prevalence of overweight and obesity than did children in southern Manitoba, including Winnipeg. Social disparity factors compound the geographical differences in the prevalence of overweight and obesity because a general north-to-south gradient of increasing parental education and household income exists in Manitoba (24).

The inverse association between social disparities and overweight and obesity in children and youth is not unique to Canada. There is a trend, if not a significant negative correlation, between SES and overweight and obesity in children and young adults in other developed nations. Differences in the built environment, such as fear of crime in low-SES communities or a lack of green space, may have an impact on people's interactions with their environment, and thereby affect activity levels (25). Lower-income families may also consume low-cost, nutrient-poor, energy-dense foods (26) and have decreased access to recreational facilities (27).

Over 40% of children and youth from food-insecure households were overweight or obese. Low income has been directly related to food insecurity (26), a known risk factor for overweight and obesity (28). Males from food-insecure households were twice as likely as males from food-secure homes to be overweight or obese. Others have reported a sex difference in the incidence of overweight and food insecurity in adults (29). Our results suggest that this difference exists early in life, and provide further rationale for programs addressing determinants of health and individual behaviours for high-risk groups of children and youth.

Energy balance is important for the achievement and maintenance of a healthy body weight. Individuals with a higher SES generally consume more fruits, vegetables, and lower-fat dairy products, and less dietary fat (30). Our findings support the belief that consuming fruits and vegetables five or more times a day helps maintain a healthy weight in girls only. Like other Canadian researchers (1,21), we found that regular PA and decreased SB were associated with healthy weights. Infrequently active male youth were more likely to be overweight or obese; this was not found for female youth. The duration of PA (15 minutes) measured by the CCHS 2.2 may not be sufficient to affect overweight or obesity. Notably, Canadian guidelines recommend 90 minutes of PA a day for children and youth (31,32). While exploring these sex differences was beyond the scope of this study, others report that girls generally become less active through adolescence (33). Thus, the extent to which female youth are sedentary, rather than the time spent in PA, may influence the propensity toward overweight or obesity. The genderspecific associations of overweight and/or obesity with PA and SB also suggest that PA is a relevant, meaningful measure for male youth and SB for female youth.

In the current study, low PA, more SB, low fruit and vegetable consumption, and food insecurity were associated with a higher prevalence of overweight and obesity. This indicates a need for health promotion programs that target all these elements. Targeting a single component will not address individuals' and communities' needs for healthy living, and thus is unlikely to yield substantial weight change. Daycare centres, community centres, parks, and schools are important settings to ensure comprehensive programs. A recent review of the literature on schoolbased obesity and type 2 diabetes prevention in both general and high-risk populations in the school setting has shown that these programs have had modest success in promoting lifestyle change and increasing knowledge over the short term (34).

Study limitations and strengths

One study limitation is that the results from PA and SB data of six- to 11-year-olds did not reveal any significant association or potential trend in terms of overweight and/or obesity. The possibility that CCHS 2.2 questions failed to capture children's PA and SB accurately, and the small sample size could be potential reasons. Second, fruit and vegetable consumption was based on frequency, not quantity. Further, results are based on crosssectional data, which precludes an understanding of causality, with occasionally small sample sizes, including groups (e.g., Aboriginal children and youth) that previous studies indicate may be at high risk for overweight or obesity (1,35,36).

However, this study also has several strengths. Measured heights and weights accurately describe the prevalence of overweight and obesity among Manitoba children. Intra-provincial analyses contributed to the novelty of this study, as did the sexspecific associations identified among obesity and age, food insecurity, PA, SB, and fruit and vegetable consumption. The analysis of youth data revealed that both daily PA and SB are gender-specifically associated with a high risk of overweight and obesity. In-depth analysis on which level of SB affects the prevalence of overweight and obesity revealed that more than 50% of youth were overweight or obese if they engaged in SB for 30 or more hours a week (or 4.29 or more hours a day). This result supports Canada's guidelines on decreasing non-active time by 90 minutes a day for children and youth (31,32) to support a healthy weight. The other three levels of SB revealed similar trends in the prevalence of overweight and obesity (approximately 30%) for both male and female youth. This result suggests that 30 or more hours of SB a week are a significant indicator for a detrimental effect of SB on body weight.

We believe this is the first Canadian study in which overweight and obesity have been described at an intra-provincial level. This is also the first study in which gender-specific differences have been reported on the association of overweight (including obesity) with parental education and SB in female youth, and with food insecurity, geographic dwellings, and age in male youth. Furthermore, this study involved an exploration of the useful classification of SB of youth in CCHS 2.2 data (i.e., that more than 30 hours a week of SB are detrimental to maintenance of a healthy weight). Many studies support an association of overweight and obesity with social determinants of health (22,23) and lifestyle (36,37), including PA and SB (1,21).

Although our findings do not conclusively show differences between those of non-Aboriginal and Aboriginal ethnicity, this may be because of a small number of Aboriginal participants. The weighted Manitoba Aboriginal sample represented less than 12% of the total sample of Manitoba children and youth. In contrast, approximately 23% of Manitoba children and youth from birth to age 14 years are of Aboriginal ethnicity (38). Fortytwo percent of Manitoba's Aboriginal people live in northern Manitoba, while only 7% and 8% live in Winnipeg and southern Manitoba, respectively (39). Thus, the Aboriginal sample underrepresented the total number of Aboriginal children and youth. Caution therefore is warranted when interpreting our data on Aboriginal children and youth. Nonetheless, our findings identify vulnerable subpopulations: youth, children in the North, Aboriginal children and youth who live off reserve, males in food-insecure homes, and females whose parents did not complete high school. Recognition of those at greatest risk is critical to health and nutrition policy and practice.

Recently, broad policy directions have been developed to improve healthy eating behaviours and access to healthy food choices for all Canadians (40,41). These directions include supporting affordable, appropriate, and accessible PA and reducing the gap in PA levels across different age, sex, ability, education, and income levels (39). The Northern Healthy Foods Initiative is a multi-departmental program designed to increase access to affordable nutritious food in northern and remote communities (39), while the Manitoba Chronic Disease Prevention Initiative mobilizes communities to take action on healthy eating, physical activity, and smoking cessation (40).

RELEVANCE TO PRACTICE

As overweight and obesity continue to be a significant problem in Canadian youth, dietitians and other health care providers must advocate for population-wide strategies aimed at reducing the obesogenic environment, as well as develop effective interventions targeting high-risk groups. Engagement with families, communities, and other sectors is important because of their ability to influence daily conditions and settings. In addition, policies and programs should take into consideration socio-economic, gender, and cultural issues to meet the needs of Aboriginal children and other vulnerable populations (34). The scale and complexity of the issue suggest the need for more than educational approaches aimed at improving individual lifestyle behaviours. Broader consideration must be given to structural factors that shape food and activity behaviours, including access to affordable healthy foods and acceptable PA opportunities within environments that promote health.

Acknowledgements

We acknowledge Shuping Liu for her assistance with data analysis, and Joyce Slater for her editing, comments, and concrete suggestions on the manuscript. The Diabetes and Chronic Diseases Unit, Public Health Branch, and Healthy Populations Branch, Manitoba Health and Healthy Living, provided financial and administrative support.

[Reference]

References

1. Shields M. Measured obesity: overweight Canadian children and adolescents. Ottawa: Statistics Canada; 2005.

2. Willms J, Tremblay M, Katzmarzyk P. Geographic and demographic variation in the prevalence of overweight Canadian children. Obes Res. 2003;11:668-73.

3. Whitaker RC, Wright JA, Pepe MS, Seidel KD, Dietz WH. Predicting obesity in young adulthood from childhood and parental obesity. N Engl J Med. 1997;337(13):869-73.

4. Tremblay M, Willms J. Secular trends in the body mass index of Canadian children. CMAJ. 2000;163(11):1429- 33.

5. Tremblay M, Katzmarzyk P, Willms J. Temporal trends in overweight and obesity in Canada, 1981-1996. Int J Obes. 2002;26:538-43.

6. Janssen I, Katzmarzyk P, Boyce W, King M, Pickett W. Overweight and obesity in Canadian adolescents and their associations with dietary habits and PA patterns. J Adolesc Health. 2004;35:360-7.

7. Himes JH, Hannan P, Wall M, Neumark-Sztainer D. Factors associated with errors in self-reports of stature, weight, and body mass index in Minnesota adolescents. Ann Epidemiol. 2004;15:272-8.

8. Elgar F, Roberts C, Tudor-Smith C, Moore L. Validity of self-reported height and weight and predictors of bias in adolescents. J Adolesc Health. 2005;37:371-5.

9. Taylor AW, Dal Grande E, Gill TK, Chittleborough CR, Wilson DH, Adams RJ, et al. How valid are self-reported height and weight? A comparison between CATI self-report and clinic measurements using a large cohort study. Aust N Z J Public Health. 2006;30:238-46.

10. Brener ND, McManus T, Galuska DA, Lowry R, Wechsler H. Reliability and validity of self-reported height and weight among high school students. J Adolesc Health. 2003;32(4):281-7.

11. Morrissey SL, Whetstone LM, Cummings CM, Owen LJ. Comparison of selfreported and measured height and weight in eighth-grade students. J Sch Health. 2006;76(10):512-5.

12. Statistics Canada. Canadian Community Health Survey Cycle 2.2 (2004) Nutrition: general health component public use microdata file documentation. Ottawa: Statistics Canada; 2005 Oct.

13. Cole T, Bellizzi M, Flegal K, Dietz W. Establishing a standard definition for child overweight and obesity worldwide: international study. BMJ. 2000;320:1240-3.

14. Bickel G, Nord M, Price C, Hamilton WL, Cook JT. Guide to measuring household food security. Food and Nutrition Service. USDA; 2000 [cited 2007 Nov 13]. Available from: http: www.fns.usa

15. Radimer KL, Olson CM, Greene JC, Campbell CC, Habicht JP. Understanding hunger and developing indicators to assess it in women and children. J Nutr Educ. 1992;24:S36-44.

16. Health Canada. Canada's Food Guides from 1942-1992; [cited 2009 Mar 10]. Available from: http://www.hc-sc.gc.ca

17. Statistics Canada. Canadian Community Health Survey (CCHS) Cycle 2.2 (Nutrition). Public use microdata file (PUMF) derived and grouped variable specifications. Ottawa: Statistics Canada; 2005.

18. Canadian Paediatric Society. Healthy active living for children and youth. Paediatr Child Health. 2002;7(5):339-45.

19. Schmitz KH, Harnack L, Fulton JE, Jacobs DR Jr, Gao S, Lytle LA, et al. Reliability and validity of a brief questionnaire to assess television viewing and computer use by middle school children. J Sch Health. 2004;74(9):370-7.

20. Bailie L, Dufour J, Hamel M. Data quality assurance for the Canadian Community Health Survey (CCHS). Ottawa: Statistics Canada; 2002.

21. Shields M. Overweight and obesity among children and youth. Health Rep. 2006;17:27-42.

22. Wang Y, Beydoun MA. The obesity epidemic in the United States-gender, age, socioeconomic, racial/ethnic and geographic characteristics: a systematic review and meta-regression analysis. Epidemiol Rev. 2007;29:6-28.

23. Oliver L, Hayes M. Neighbourhood socio-economic status and the prevalence of overweight Canadian youth and children. Can J Public Health. 2005;96:415-20.

24. Statistics Canada. Census: analysis series; 2006 [cited 2009 Mar 7]. Available from: http://www12.statcan.ca/english/census06/analysis/education/urban.cfm

25. Papas MA, Alberg AJ, Ewing R, Helzlsouer KJ, Gary TL, Klassen, AC. The built environment and obesity. Epidemiol Rev. 2007;29;129-43.

26. Che J, Chen J. Food insecurity in Canadian households [1998/99 data]. Health Rep. 2001;12:11-22.

27. Moore LV, Diez Roux AV, Evenson KR, McGinn AP, Brines SJ. Availability of recreational resources in minority and low socioeconomic status areas. Am J Prev Med. 2008;34(1):16-22.

28. McIntyre L. Food insecurity. In: Raphael D, editor. Social determinants of health. Toronto: Canadian Scholars' Press; 2005. p. 187-200.

29. Vozoris N, Tarasuk V. Household food insufficiency is associated with poorer health. J Nutr. 2003;133:120-6.

30. Power EM. Determinants of healthy eating among low-income Canadians. Can J Public Health. 2005;96:S37-8.

31. Public Health Agency of Canada. Canada's physical activity guide for youth. Ottawa: Government of Canada; 2002.

32. Public Health Agency of Canada. Canada's physical activity guide for children. Ottawa: Government of Canada; 2002.

33. Pate RR, Dowda M, O'Neill JR, Ward DS. Change in PA participation among adolescent girls from 8th to 12th grade. J Phys Act Health. 2007;4:3-16.

34. Ng C, Anderson K, McQuillen K, Yu BN. School-based obesity and type 2 diabetes prevention programs: a public health perspective. Can J Diabetes. 2005; 29(3):211-9.

35. Katzmarzyk PT, Malina RM. Obesity and relative subcutaneous fat distribution among Canadians of First Nation and European ancestry. Int J Obes. 1998;22:1127-31.

36. Bernard L, Lavall�e C, Gray-Donald K, Delisle H. Overweight in Cree schoolchildren and adolescents associated with diet, low PA, and high television watching. J Am Diet Assoc. 1995;95:800-2.

37. Rose D, Bodor J. Household food insecurity and overweight status in young school children: results from the Early Childhood Longitudinal Study. Pediatrics. 2006;117:464-73.

38. Western Economic Diversification Canada. The Aboriginal context in Manitoba; 2001 [cited 2009 Jun 4]. Available from: http://www.wd.gc.ca/eng/6904.asp

39. Manitoba Aboriginal and Northern Affairs. Aboriginal people in Manitoba; 2000 [cited 2009 Mar 6]. Available from: http://www.gov.mb.ca/ana/apm2000/1/l.html

40. Manitoba healthy living, youth and seniors; 2009 [cited 2009 Mar 6]. Available from: http://www.gov.mb.ca/healthyliving/index.html

41. Public Health Agency of Canada. Integrated Pan-Canadian Healthy Living Strategy; 2005 [cited 2009 Mar 6]. Available from: http://www.phac-aspc.gc.ca/hl-vs-strat/pdf/hlse.pdf

[Author Affiliation]

BO NANCY YU, MD, MSc, PhD, Manitoba Health and Healthy Living, University of Manitoba, Department of Community Health Sciences, Winnipeg, MB; JENNIFER LISA PENNER PROTUDJER, MSc, Manitoba Health and Healthy Living, University of Manitoba, Department of Applied Health Sciences, Winnipeg, MB; KRISTIN ANDERSON, MSc, RD, Manitoba Health and Healthy Living, Winnipeg, MB; PAUL FIELDHOUSE, PhD, Manitoba Health and Healthy Living, University of Manitoba, Department of Human Nutritional Sciences, Winnipeg, MB