Introduction

There is widespread concern that the prevalences of overweight and obesity are reaching epidemic levels in many developed and developing countries.¹ In Canada, there is strong evidence that the prevalence of childhood overweight and obesity is rising rapidly.^2,3,4 For example, between 1981 and 1996, the prevalence of obesity among Canadian children aged 7–13 y tripled from 5 to 15%.² Since the rate of change is so great, it is generally believed that environmental factors, not genetic factors, explain the increase in childhood obesity. It is believed that automated labour-saving devices and inexpensive and ubiquitous access to calorie-dense food create an environment conducive to obesity.

Analyses of the 1988 Campbell's Survey on the Well-Being of Canadians suggest that there is a negative relation between self-reported physical activity and body mass index (BMI),⁵ and data from other countries support this relation.^1,6 Excessive television (TV) watching and video game use have been identified as a stimulus for excessive eating and sedentary behaviour.⁷ Gortmaker et al⁸ calculated that the odds of being overweight were 4.6 times greater for youth in the United States who watch >5 h of TV per day compared with those watching 0–2 h/day. This relation remained similar after controlling for baseline maternal overweight, socioeconomic status (SES), household structure, ethnicity, and maternal and child aptitude test scores. It appears that both physical activity and physical inactivity may contribute to a positive caloric balance resulting in an increase in the prevalence of obesity.

To our knowledge, there are no published findings examining the relations among physical activity, sedentary behaviour (video game use and TV/video watching), and BMI on a nationally representative sample of Canadian children. Furthermore, it is not known how physical activity and inactivity mediate the effect of socioeconomic status and family background on the risk of overweight and obesity. An assessment of these relations will assist in understanding the factors that contribute to the obesity epidemic, and may lend support to certain intervention strategies. Therefore, the purpose of this study was to examine, on a large representative sample of Canadian children, the relations among children's physical activity, sedentary behaviours, and BMI, while controlling for sex, family structure, and socioeconomic status.

Methods

Data from the first wave (1994) of the National Longitudinal Survey of Children and Youth (NLSCY)⁹ were used to assess the relations among children's physical activity (organized sport, unorganized sports and physical activities), sedentary behaviours (computer and video game use, and TV/video watching), and BMI (mass (kg)/height (m²)), while controlling for age, sex, and family background characteristics. The NLSCY is a nationally representative longitudinal study of Canadian children. The NLSCY began data collection in 1994 and has been repeated every 2 y. The 1994 NLSCY data were collected between the fall of 1994 and spring of 1995. The target population comprised all Canadian children from newborn to age 11 y (ages 7–11 y used in these analyses). The participating households were selected from the sampling frame of Statistics Canada's Labour Force Survey, which is representative of Canada's population. The Labour Force Survey employs a stratified, multistage probability sample design based on an area frame in which dwellings (residences) are the sampling units.⁹ Further details of the survey and sampling procedures are available elsewhere.⁹ The responding sample of 13 439 households was drawn from a household sample of 15 600 (86% response rate). In this survey, the height and body mass data were gathered through parental report, which in some cases (proportion not known) included direct measurement. BMI results were available for 7216 children aged 7–11 y from the NLSCY.

The outcome measures used in our analysis are dichotomous variables indicating whether or not a child was obese or overweight. The "person most knowledgeable" (PMK) reported the child's height and body mass, and in some cases (proportion unknown) these were assessed directly. The limitations of self- and parental-report data for the study are discussed elsewhere.^2,3 Overweight and obesity were defined using the recently developed international age- and sex-specific cutoffs presented in Table 1.¹⁰

Table 1 - Published cutoffs for BMI for overweight and obesity by sex between 7 and 11y of age. The cutoffs are defined to pass through BMI of 25 kg/m² (overweight) and 30 kg/m² (obese) at age 18 y.

Full table

The behavioural characteristics were assessed by parental response to the following questions:

• In the last 12 months, outside of school hours, how often has_______ taken part in any sports which involved coaching or instruction?
• Taken part in unorganized sports or physical activities?
• Taken lessons or instruction in music, dance, art or other non-sport activities?
• Taken part in any clubs, groups or community programs with leadership, such as Boys and Girls Clubs, Scouts, Guides or church groups?
• Played computer or video games?

Each of the above questions was answered on a five-point scale: almost never, about once a month, about once a week, a few times a week, most days. In preliminary analyses, we found that the critical distinction for the relation between the activity variables and BMI tended to be between participation of once a week or less, vs a few times a week or more. Thus, to simplify the analyses, we collapsed the categories to create dichotomous variables indicating whether or not a youth participated at least a few times a week in each activity. The variable describing video watching was also coded to indicate whether or not a youth watched videos at least a few times each week.

Parents were also asked About how many days a week on average does_______ watch TV or videos at home? and On those days, how many hours on average does he/she spend watching TV or videos? Children were then categorized as watching TV on average less than 2 h/day, between 2 and 3 h/day, and between 3 and 5 h/day.

Our analyses also included a measure of socioeconomic status (SES) constructed for the NLSCY. It is a statistical composite comprising parents' level of education, the prestige of the parents' occupations, and family income.¹¹ We identified youth who were in the lowest and highest quartiles of SES, which were also used in our analyses as dichotomous variables. Family structure indicates whether the child was living in a single- vs a two-parent family.

Our first analysis simply examined cross-tabulations of the relations between obesity, overweight, and activity/inactivity with socioeconomic status and family structure. A one-way analysis of variance F-test was used to assess statistical significance. Our second analysis fitted three logistic regression models to assess (a) the relation between obesity and overweight and activity and inactivity, (b) the relation between obesity and overweight and SES and family structure, and (c) the joint relation between obesity and overweight and activity/inactivity, SES, and family structure. In all three models, age and sex were included as control variables, and the sample design weights provided by Statistics Canada were used to take account of nonresponse bias and the sample design.

Results

Table 2 shows the relation between SES and family structure with obesity and overweight, participation in sport, physical activity, other organized leisure-time pursuits, and video game and TV usage. Obesity and overweight are both significantly related to SES and family structure. Participation in organized sports is more strongly related to SES and family structure than other forms of activity/inactivity. Watching TV less than 2 h/day, and more than 3 h/day is also strongly related to family background, while playing video games and watching TV 2–3 h/day are weakly related to family background. Overall, this analysis shows that SES, family structure, and activity and inactivity are 'overlapping' risk factors.¹²

Table 2 - Relation of obesity, overweight, and activity/inactivity with socioeconomic status and family structure (percentage and ANOVA F).

Full table

Table 3 presents the results from three different statistical models (I, II, III) which describe the relations of obesity and overweight with child's age and sex, activity/inactivity, and family background. Model I shows the relations between obesity and overweight with activity and inactivity, while controlling for age and sex. Participation in organized or unorganized sports was negatively associated with obesity, while watching TV more than 3 h/day was a risk factor. Playing video games and watching TV more than 2 h/day were risk factors for overweight, while participating in unorganized sports was negatively associated with overweight.

Table 3 - Relation of obesity and overweight with child's age and sex, activity/inactivity, and family background (presented as odds ratios (95% confidence intervals) from logistic regression models).

Full table

In preliminary analyses, we also examined the effects of a finer breakdown of TV watching (ie, 0–1, 1–2 h, etc), but because of relatively small sample sizes within these categories, and because the relative effects associated with obesity and overweight were small, we found the three categories (<2, 2–2.99, and 3–5 h/day) to be preferable. There appears to be an important threshold at 2 h/day, with respect to overweight, and 3 h/day with respect to obesity.

The results for Model II indicate that children living in single-parent families were more likely to be obese than those in two-parent families, while living in a low-SES family was not a strong risk factor for obesity. Conversely, children in low-SES families were more likely to be overweight than children in the middle two quartiles of SES, while living in a single-parent family was not a strong risk factor for overweight. However, one should note that many single parents have low household income and low levels of education. In this sample, 47% of single parents were in the lowest SES quartile. Living in a high-SES family was negatively associated with both overweight and obesity.

Model III includes both sets of factors. The child's family background and activity/inactivity are overlapping risk factors, but only partially. Children who participated in unorganized sports, and those who watched TV less than 2 h/day, were less likely to be overweight or obese. These associations cut across levels of SES. Similarly, differences in the prevalence of obesity and overweight associated with SES are only partially attributable to differences in the levels of children's activity and inactivity. For both obesity and overweight, the effects of living in a high-SES family are equivalent to those associated with regular participation in unorganized sports activities.

A test for interactions between age and the activity/inactivity variables indicated that the potential benefits of unorganized sports increase with age, while the reverse was found for organized sports.

Discussion

This study is the first to examine the relations among physical activity, sedentary behaviour, and overweight and obesity, after controlling for SES and family background, in a representative sample of Canadian children. The findings support the intuitive belief that physical activity provides protection from being overweight or obese, while TV watching and video game use are risk factors. The results also suggest that the benefits of participation in unorganized sport and physical activity increase with age, and are more important than involvement in organized sport.

The challenge to overcome the growing epidemic of obesity in Canada is immense. Automated labour-saving devices and inexpensive and ubiquitous access to calorie-dense food create an environment conducive to obesity. Excessive TV watching and video game use have been identified as a stimulus for excessive eating and sedentary behaviour.⁷ The present study suggests that this behaviour pattern increases the risk of being overweight or obese, regardless of SES, and is consistent with the work of Gortmaker et al.⁸ From their analysis,⁸ attributable risk estimates suggest that more than 60% of overweight incidence can be linked to excess TV viewing time. This relation was also reaffirmed for US children aged 8–16 y¹³ and Mexican children aged 9–16 y.¹⁴ Researchers in Belgium observed a significant positive relation between days per week of TV watching and BMI but in males only, suggesting that factors contributing to obesity development may be gender specific.¹⁵ Crespo et al¹⁶ found that the prevalence of obesity was lowest among children watching 1 h of TV, and highest among those watching 4 h of TV per day. In contrast to Guillaume et al,¹⁵ Crespo et al¹⁶ found that TV watching was positively associated with obesity among girls, even after controlling for age, race/ethnicity, family income, weekly physical activity, and energy intake.

Similar to excessive TV watching, the growing use of video and computer games has likely contributed to the progressive fattening of Canadian children. An inverse relation between video game playing and daily physical activity has recently been observed.¹⁷ Robinson reported that reducing children's TV, video-tape, and video game use may be a promising, population-based approach to prevent childhood obesity.¹⁸

An interesting observation from these analyses is that art and dance involvement is negatively associated with being overweight. This relation, which is evident even after controlling for SES and other activities included in the model (see Model III for overweight), suggests that different forms of leisure 'engagement' may be beneficial. Although just speculation, it may be that even low-activity forms of leisure (ie art) may be somewhat protective by preventing negative behaviours related to typical sedentary living (eating junk food while watching TV). Alternatively, this effect may be attributable to self-selection, or other physical activities not accounted for in the models used. This finding deserves further investigation.

Ecological data support the findings observed in this study. Canadian information clearly demonstrates that access to in-home entertainment that promotes sedentary behaviour has increased.¹⁹ For example, between 1986 and 1996 the number of households with home computers more than tripled (10.3–31.6%) and the households with two or more colour TVs more than doubled (23.8–51.5%).¹⁹ Clearly, the physical and social environment our children are exposed to differs significantly from previous generations, and is conducive to a reduction in energy expenditure. According to the Health Behaviours of School-Aged Children Survey²⁰ between 1990 and 1998 fewer Canadian students aged 11–15 y exercised two or more times a week outside of school hours, the percentage of Canadian students who watched TV four or more hours a day increased, and more Canadian children reported playing video games >4 h/week than in most other countries in the survey. In general, Canadians report about three times as much time spent on passive leisure compared to active leisure.¹⁹ Given the evolving environment in which Canadian children live, and given the tendency for obese children to remain obese as adults,^21,22 the obesity epidemic in Canada will not likely subside any time soon unless aggressive public health campaigns against obesity are implemented.

The relation between overweight and obesity, and morbidity and mortality is well established and has been reviewed recently.²³ Results from the Canadian Heart Health Survey have shown significant positive relations between cardiovascular disease risk factors (increased blood pressure, physical inactivity, diabetes mellitus, and dyslipidaemia) and BMI.²⁴ The economic cost of obesity has been estimated at $51.6 billion in direct medical costs in the US in 1995²⁵and $1.8 billion in Canada in 1997.²⁶ Although it is difficult to predict exactly how the increase in childhood obesity will translate into adult health problems,²⁷ it seems likely that the future economic ramifications will be substantial unless significant efforts to change physical activity behaviours and nutrition practices are made.

Despite using a fairly large, nationally representative sample, there are clear limitations to the findings from this study. Although large, multipurpose national studies generally require cursory, parental- or self-reported measurements of activity and inactivity, caution must be used when drawing conclusions from these data. Verification of these findings should be made, even if on a small sample, using more comprehensive and objective measures of physical activity/inactivity behaviour. Also, the accuracy of self- and parental reports of height and weight have been questioned and discussed.^2,3,28,29,30 In general, there appears to be a tendency for BMIs to be underestimated as a result of an underestimation of weight and an overestimation of height; so the absolute prevalences listed here are likely conservative estimates. Finally, the use of BMI as an indicator of obesity or overweight has been questioned;³¹ nevertheless, BMI is generally accepted as a valid indicator of body composition for assessments at the population level.^1,32

Caution must also be used when speculating on causation. It may be that children who are prone to overweight and obesity may also be prone to living a more sedentary lifestyle; that is, obesity and overweight can discourage children from participating. Other factors such as the opportunities for recreation and parenting styles may also be relevant and require further study.

Despite the limitations discussed above, this study provides evidence supporting the link between physical inactivity and obesity of Canadian children. The findings from this study provide further support for policy initiatives aimed at increasing physical activity opportunities while decreasing sedentary behaviour as a means of controlling the childhood obesity epidemic in Canada. These conclusions are consistent with the recently published guidelines in Canada's Physical Activity Guide for Children,³³ which calls for both an increase in physical activity time and a reduction in inactive time. Further work is required to understand how adherence to these recommendations can be maximized, particularly for children from single-parent families, and those with low SES.

References

1. World Health Organization. Obesity: preventing and managing the global epidemic. Report of a WHO Consultation on Obesity. World Health Organization: Geneva; 1998.
2. Tremblay MS, Willms JD. Secular trends in the body mass index of Canadian children. Can Med Assoc J 2000; 163: 1429–1433; 2001; 164: 970.
3. Tremblay MS, Katzmarzyk PT, Willms JD. Temporal trends in overweight and obesity in Canada, 1981–1996. Int J Obes Relat Metab Disord 2002; 26:538–543. | Article | PubMed | ChemPort |
4. Willms JD, Tremblay MS, Katzmarzyk PT. Geographic and demographic variation in the obesity of Canadian children. Obes Res 2003; 11: 668–673. | PubMed | ISI |
5. Stephens T, Craig CL. The well-being of canadians: highlights of the 1988 Caompbell's survey. Canadian Fitness and Lifestyle Research Institute: Ottawa, ON; 1990.
6. U.S. Department of Health and Human Services. Physical activity and health: a report of the surgeon general. U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion: Atlanta, GA; 1996.
7. Robinson TN, Killen JD. Ethnic and gender differences in the relationships between television viewing and obesity, physical activity, and dietary fat intake. J Health Educ 1995; 26(2 Suppl): S91–S98.
8. Gortmaker SL, Must A, Sobol AM, Peterson K, Colditz GA, Dietz WH. Television viewing as a cause of increasing obesity among children in the United States, 1986–1990. Arch Pediatr Adolesc Med 1996; 150: 356–362. | PubMed | ChemPort |
9. Statistics Canada and Human Resources Development Canada. National longitudinal survey of children: overview of survey instruments for 1994–1995, data collection 1. Statistics Canada Catalogue No. 95-02. Minister of Industry: Ottawa, ON; 1995.
10. Cole TJ, Bellizzi MC, Flegal KM, Dietz WH. Establishing a standard definition for child overweight and obesity worldwide: international survey. BMJ 2000; 320: 1–6. | Article | PubMed |
11. Willms JD, Shields M. A measure of socioeconomic status for the National Longitudinal Study of Children. Report prepared for Statistics Canada;1996.
12. Kraemer HC, Stice E, Kazdin A, Offord D, Kupfer D. How do risk factors work together? Mediators, moderators, and independent, overlapping, and proxy risk factors. Am J Psychiatr 2001; 158: 848–856.
13. Andersen RE, Crespo CJ, Bartlett SJ, Cheskin LJ, Pratt M. Relationship of physical activity and television watching with body weight and level of fatness among children. JAMA 1998; 279: 938–942. | Article | PubMed | ChemPort |
14. Hernandez B, Gortmaker SL, Colditz GA, Peterson KE, Laird NM, Parra-Caberra S. Association of obesity with physical activity, television programs and other forms of video viewing among children in Mexico City. Int J Obes Relat Metab Disord 1999; 23: 845–854. | Article | PubMed | ChemPort |
15. Guillaume M, Lapidus L, Bjorntorp P, Lambert A. Physical activity, obesity, and cardiovascular risk factors in children. The Belgian Luxembourg Child Study II. Obes Res 1997; 5: 549–556. | PubMed | ISI | ChemPort |
16. Crespo CJ, Smit E, Troiano RP, Bartlett SJ, Macera CA, Andersen RE. Television watching, energy intake, and obesity in US children. Arch Pediatr Adolesc Med 2001; 155: 360–365. | PubMed | ISI | ChemPort |
17. Janz KF, Mahoney LT. Maturation gender, and video game playing are related to physical activity intensity in adolescents: the Muscatine Study. Pediatr Exerc Sci 1997; 9: 353–363.
18. Robinson TN. Reducing television viewing to prevent obesity: a randomized controlled trial. JAMA 1999; 282: 1561–1567. | Article | PubMed | ISI | ChemPort |
19. Statistics Canada. Canada year book, 2001. Catalogue No. 11-402-XPE. Minister of Industry: Ottawa, ON; 2001.
20. World Health Organization. Health and health bahaviour among young people. WHO Policy Series: Health Policy for Children and Adolescents Issue 1: International Report. World Health Organization: Geneva; 2000.
21. Power C, Lake JK, Cole TJ. Body mass index and height from childhood to adulthood in the 1958 British birth cohort. Am J Clin Nutr 1997; 66: 1094–1101. | PubMed | ISI | ChemPort |
22. 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: 869–873. | Article | PubMed | ISI | ChemPort |
23. Bray GA. Overweight, mortality, and morbidity. In: Bouchard C (ed.). Physical activity and obesity. Human Kinetics: Champaign, IL; 2000. pp 31–53.
24. Rabkin SW, Chen Y, Leiter L, Liu L, Reeder BA. Canadian Heart Health Surveys Research Group. Risk factor correlates of body mass index. Can Med Assoc J 1997; 157(1 Suppl): S26–S31. | ISI |
25. Wolf AM, Colditz GA. Current estimates of the economic cost of obesity in the United States. Obes Res 1998; 6: 97–106. | PubMed | ISI | ChemPort |
26. Birmingham CL, Muller JL, Palepu A, Spinelli JJ, Anis AH. The cost of obesity in Canada. Can Med Assoc J 1999; 160: 483–488. | ISI | ChemPort |
27. Power C, Lake JK, Cole TJ. Measurement and long-term health risks of child and adolescent fatness. Int J Obes Relat Metab Disord 1997; 21: 507–526. | Article | PubMed | ChemPort |
28. Strauss RS. Comparison of measured and self-reported weight and height in a cross-sectional sample of young adolescents. Int J Obes Relat Metab Disord 1999; 23: 904–908. | Article | PubMed | ChemPort |
29. Paccaud F, Wietlisbach V, Rickenbach M. Body mass index: comparing mean values and prevalence rates from telephone and examination surveys. Rev Epidemiol Sante Publique 2001; 49: 33–40. | PubMed |
30. Mokdad AH, Serdula MK, Dietz WH, Bowman BA, Marks JS, Koplan JP. The spread of the obesity epidemic in the United States, 1991–1998. JAMA 1999; 282: 1519–1522. | Article | PubMed | ISI | ChemPort |
31. Eston RG. Use of the body mass index (BMI) for individual counselling: the new section editor for Kinanthropometry is 'Grade 1 Obese, Overweight' (BMI 27.3), but dense and 'distinctly muscular' (FFMI 23.1)! J Sport Sci 2002; 20: 515–518.
32. United States National Institutes of Health. Clinical guidelines for the identification, evaluation, and treatment of overweight and obesity in adults. National Institutes of Health: Bethesda, MD; 1998.
33. Health Canada. Canada's physical activity guide for children. Minister of Public Works and Government Services, Canada: Ottawa, ON; 2002.

Acknowledgements

This project was supported by the Canadian Population Health Initiative of the Canadian Institute for Health Information. We would like to thank Human Resources Development Canada for the use of data from the 1994 National Longitudinal Survey of Children and Youth. The results presented and the views expressed in this paper are those of the authors.