I would like to thank you for inviting me here to speak to you about this important topic. I'd also like to thank the committee for choosing to address the issue of childhood obesity this fall.
I'm a health geographer, which means I investigate how the local places where individuals live shape their health status. My research focuses in particular on how the neighbourhood environment influences childhood obesity by structuring opportunities to engage in activities that promote or inhibit weight gain.
I've been invited to come here today to present findings of my research on neighbourhood socio-economic influences on childhood obesity published in the Canadian Journal of Public Health. I would also like to acknowledge the contribution of Dr. Michael Hayes, who was co-author of this work. This research uses data from the Statistics Canada national longitudinal survey of children and youth.
Neighbourhood socio-economic status is measured from unemployment rates, median family income, and percent without post-secondary education. It is divided into four equal categories or quartiles, from low SES to high SES. Measures of overweight and obesity are based on children's heights and weights and use of body mass index cut-points from the International Obesity Task Force. Heights and weights are based on parent and self-reports.
The graph on page 4 shows the prevalence of overweight and obesity by neighbourhood SES among children and youth. First, a social gradient in overweight is apparent, in which the prevalence decreases from 41% in low SES neighbourhoods to 27% in high SES neighbourhoods. Similarly, the prevalence of childhood obesity systematically decreases from 19% in low SES neighbourhoods to 10% in high SES neighbourhoods. This pattern is also evident for obesity among youth.
I want to emphasize that the low category represents the bottom 25% of neighbourhoods in Canada. It does not represent only extreme poverty. Neighbourhood effects are real and persist after controlling for parental education, family income, and the child's age and gender.
The graph on page 6 indicates the likelihood of a child being overweight, adjusted and unadjusted for family factors. The solid bars show the likelihood of being overweight, adjusted for family factors. They show that relative to living in a high SES neighbourhood, a child is almost 1.3 times more likely to be overweight if they live in a low SES neighbourhood. This shows that neighbourhood effects are real, and a child is more likely to be overweight if they live in a low SES neighbourhood, independent of their family circumstances. Thus neighbourhood effects are not solely due to the clustering of low socio-economic status families in disadvantaged neighbourhoods.
What can be attributed to neighbourhood factors? The next few graphs provide some evidence to explain this pattern. The graph on page 7 shows the percent of parents reporting a lack of safe parks and playgrounds in the neighbourhoods. Again we see a similar pattern where a lack of safe parks and playgrounds is highest at 27% in low SES neighbourhoods and systematically decreases to 9% in high SES neighbourhoods. This is a three-fold difference in safe parks and playgrounds between the highest and lowest quarters of neighbourhoods in Canada.
The graph on page 8 shows the percentage of children and youth not participating in organized and unorganized sports. Organized sports include activities such as hockey or gymnastics. They have coaches or instructors and typically involve fees. Unorganized sports do not have coaches or instructors and include activities such as street hockey or simply playing in parks and playgrounds. I want to point out two things about this graph.
First, among children aged five to eleven, there was a clear graded relationship between participation in organized sports and neighbourhood SES, in which the percentage of children not participating is highest, at 52%, in low SES neighbourhoods and decreases to 30% in the high category. This is a substantial difference in non-participation rates between the highest and lowest quarters of neighbourhoods in Canada. It could be due to factors such as lack of programs in these neighbourhoods or to parental barriers, such as the ability to pay or to provide transportation for children to attend such programs.
Second, while non-participation rates are similar across all neighbourhoods for unorganized activities, the previous graph suggests that children in low SES neighbourhoods may undertake such activities in less safe environments.
Over the last few months, I have expanded upon this research by examining the emergence of neighbourhood disparities in obesity and overweight as children age. This research is important because if we can identify when these neighbourhood disparities emerge, then we can develop policies to address them.
Using the same data, from the national longitudinal survey of children and youth, but following children over time, the graph on page 9 shows the development of overweight and obesity by neighbourhood income among a cohort of children aged two and three.
First, among children aged two and three in 1994 there was no clear relationship between overweight and neighbourhood income. Second, when these children are assessed four years later, aged six and seven, a strongly graded relationship is apparent between neighbourhood SES and overweight and obesity, and this persists when children are aged ten and eleven.
The key findings of this research are that neighbourhoods with lower SES have higher levels of obesity, less participation in organized sports, and a lack of safe parks and playgrounds. Neighbourhood disparities in overweight and obesity seem to emerge between the ages of two and ten.
I have been asked to discuss what the federal government might do to address neighbourhood inequality and childhood obesity. Effective policy, I think, will require the federal government to take an active leadership role, involving both provinces and municipalities and other stakeholders as well.
Federal government support for the establishment of universal programs for physical activity through targeted support to municipalities may be needed. Universal programs run through community centres may begin to address the graded relationship between neighbourhood SES and non-participation in sports programs, and additional targeted programs may be required to improve participation rates in low SES neighbourhoods, which are very low right now.
I encourage the federal government to take an active role in reducing barriers to participation in physical activity for children. Targeted support could be directed towards municipalities or provinces to reduce such barriers. For example, providing transportation for children to and from home to recreational programs may increase participation, especially among children living in low SES neighbourhoods or families, and the elimination of user fees for programs, especially among low SES children or low SES neighbourhoods, may improve participation.
Increasing children's participation in unstructured activity is considered important to reduce childhood obesity, and improving the safety of parks and playgrounds would likely improve participation rates in unorganized physical activity. Targeted support for municipalities to address safety concerns relevant to their neighbourhood is needed.
For example, supervision of parks and playgrounds during after-school hours may improve safety and encourage use. Implementation of traffic calming in neighbourhoods, and especially around parks and playgrounds, may increase safety.
Without such initiatives, efforts to increase unorganized physical activity may meet with little success or could even have unintended consequences, such as exposing children to hazards.
Also, policies to address neighbourhood inequalities in overweight and obesity should focus on young children.
In conclusion, all Canadian children should be able to grow up in neighbourhoods with safe parks and playgrounds and opportunities for physical activity. Addressing childhood obesity will require policies that focus on the neighbourhoods in which Canadian children live. The federal government should take an active role in such initiatives.
When formulating anti-obesity policy, I'd encourage the federal government to ask what this policy will do for children living in low-income neighbourhoods. Effective anti-obesity policy must be relevant to children living in low-income neighbourhoods.
Again, thank you for inviting me here to speak to you about this important issue.
I would also like to thank the Statistics Canada research data centre program, which provided access to the micro-data files of the Statistics Canada NLSCY, which this analysis was based on.