Does where you live matter? Leisure-time physical activity among Canadian youth: a multiple cross-sectional study
Charles Nadeau MSc, Laurence Letarte MSc, Ramona Fratu MSc, E. Owen D. Waygood PhD,Alexandre Lebel PhD
* This is the accepted version of the paper. The published version can be found here:
Abstract
Background: The aim of this study was to explore the population wide distribution in the practice of Leisure Time Physical Activity (LPA) among young Canadians and how it is influenced by contextual features of the living environment.
Methods: This study relied on the self-reported LPA behaviours of 54,832 Canadians aged 12 to 17. Observations were structured according to a four-level geographical hierarchy. The outcome studied was a dichotomous indicator which refers to achieving (or not) the recommended daily LPA level. To investigate the influence of the contextual features, multilevel logistic regressions were conducted.
Results: Among both sexes, significant variations were observed between health regions and between neighbourhoods within the provinces. Girls showed lower odds of achieving LPA standards when living in an urban setting and during winter. Although boys also showed lower odds of achieving standards during winter, living in an urban setting had no impact on them. Residual analysis highlighted provinces with a significantly different LPA level from the Canadian average.
Interpretation: The results show that LPA was associated with environmental factors at multiple scales among Canadian youth. The variation was more important at the neighbourhood level and suggests that youth had up to a 17% increased chance of achieving the recommended LPA level if they lived in a context with a higher LPA achievement level. Contextual influences may differ between girls and boys. The study provides rationale for further investigation on how LPA is promoted in different contexts.
Leisure-time physical activity (LPA) provides fundamental health benefits for young people(1). Benefits include increased physical fitness, reduced adiposity, favourable cardiovascular and metabolic disease risk profiles and reduced symptoms of depression (2-5). Among young people, the growth period is a critical time for the development of factors that have a great influence on health in adulthood, such as achieving an optimal bone health(6).
Understanding what influences youth to engage in LPA contributes to evidence-based planning of public health interventions, as effective programs will target factors known to contribute to physical (in)activity (7). Research into correlates or determinants of LPA has mostly focused on individual-level factors (8). Among them, socioeconomic status indicators such as education (9) and biological factors such as body mass index (10) have been associated with differences in LPAparticipation.However, contextual factors such as the quality of the built or social environment are less studied, but are thought to have widespread effects that may vary across geographical settings (8, 11, 12)[OW1].
A better understanding of the influence of geographic settings (local to national) on youth LPA can inform the development of multifaceted interventions which likely offer the best chance for success (13).
The general aim of this study was to explore the population wide distribution of LPA practice among young Canadians and the influence of area-based characteristics, while taking into account individual factors. The objectives were: (a) to describe the geographic variations of LPA among young Canadians and (b) to explore the contribution of contextual factors to explain these variations.
Methods
Data source, study population and sample size
The study used the Canadian Community Health Survey (CCHS) from 2003 to 2011. The CCHS provides self-reported information for a nationally representative sample of the
non-institutionalized civilian populationof 12 years and older in the 10 Canadian provinces (14). The sample for this study included individuals of 12-17 years old for whom complete information on ethnic origin, education level, body mass index and geographic location were available. Pregnant girls and respondents interviewed by proxies were excluded for consistency with the research design. The final sample was 54,832 observations.
Hierarchical structure
Observations were structured according to a four-level geographical hierarchy based on Statistics Canada 2006 census administrative units. When a place of residence was located in a census metropolitan area or a census agglomeration, the “neighbourhood” corresponds to the census tract, otherwise it was attributed to the municipality (census subdivision). The combination of census tractand census subdivisions enables the creation of comparable neighbourhood units for the entire Canadian population, whether they are located in an urban or a rural setting, and to consider the idiosyncrasies of the characteristics of the individuals’ place of residence. In order to take into account changes in the boundaries of the health regions during the study period, the geographic structure was harmonized using a digital boundary file reflecting the boundaries of the health regions as of October 2011 (15). A detailed methodology of the geographical structure is presented by Lalonde (16). ArcMap 10.1 software was used for the geospatial processing. The final hierarchical structure comprised 6004 neighbourhoods nested within 112 health regions, and the 10 Canadian provinces.
Outcome
The dependant variable studied was a dichotomous indicator of LPA which refers to achieving (or not) the recommended daily level for physical activity. Guidelines recommend that young people aged 12-17 years should accumulate an average of at least 60 minutes of moderate- to vigorous-intensity physical activity daily (17, 18). An “active” youth is one that achieves an index of energy expenditure of at least 30 kcal/kg/week (19) with a frequency of 5 times/week or more. This achievement corresponds to walking briskly (4.3 METs) for 1 hour a day, 7 days a week (20). Physical activity level was estimated by a validated index of energy expenditure that considered the frequency, duration and intensity of 18 types of self-reported LPA (21). To overcome the absence of a measure of intensity in the survey, thresholds were derived from a table of Metabolic Equivalent of Task (METs) attributed to various activities (21).
Individual variables
To account for the documented influences of individual characteristics on youth LPA (22-24), age, ethnic origin, highest education level of the household, and body mass index were used as control variables (Table 1).
Cycle and season
The survey contains data collected over a 12-month period (cycle 2003 and 2005) and a 24-month period (cycle 2007, 2009 and 2011). The exact date of sampling allowed for the inclusion of season in the explanatory model. Three seasons categories were created: Summer (July to October), Winter (February to March) and Transitional (combining November to January and April to June).
Area-based variables
Three independent area-based variables were considered. The first is the Census metropolitan influenced zone (MIZ) that indicates areas which are urban (inside census metropolitan areas and census agglomerations), urban outskirt (commuting flow to urban area higher than 30%) and rural (commuting flow to urban area lower than 30%) (25).
The two other area-based variables are the neighbourhoods’ social and material deprivation levels. The measures of deprivation are based on the factor score of two dimensions resulting from a principal component analysis that estimates the material and social deprivation of the neighbourhood units using Canada-wide census information (proportion of adults that are living alone, single parents, without high school diploma, unemployed, and the mean income of the area) (26). The distribution of the neighbourhoods’ deprivation levels was broken into quintiles by province for all census units.
Statistical analysis
Multilevel analysis was used to estimate the impact of contextual factors on LPA. The use of multilevel analysis is recommended to account for the clustering of persons within areas and to decompose the variance in the outcome being studied into between and within area components (27, 28). Thus, to investigate the influence of the contextual variables, a series of multilevel logistic regression analyses were conducted using the Bayesian estimation procedure as implemented via Markov chain Monte Carlo (MCMC) methods in MlwiN 2.28 software. This procedure is detailed in Browne and al. (29). The benefits of this type of analysis as compared to a standard logistic regression are to produce unbiased estimates of the random effects, and to provide a measure to compare the relative effectiveness of different models in accounting for the variations in LPA, the Deviance Information Criteria (DIC)(27).
All analyses were stratified by sex to control for the differences in physical maturation between girls and boys during the puberty(30).[OW2]
The modelling strategy was based on four incremental models in order to estimate the impact of the additional groups of variables in the fixed part of the models on the between area variance (random part of the models). The models’ equations are provided in the supplementary file (S1).
The distribution of the LPA between the geographic levels (random part) without covariates is seen in the first (null) model. The second model introduced the control variables of cycle and season. The third introduced the census metropolitan influenced zone. The last model introduced the neighbourhood’s social and material deprivation indices.
The Deviance Information Criteria (DIC) was used to compare each model’s goodness-of-fit (31) with the previous model’s (ΔDIC). The median odds ratio (OR) was used to translate the area level variance to an OR scale, which has a consistent and intuitive interpretation (32). The median OR can be directly compared to a regular OR and allows one to estimate the relative importance of the unexplained area-level variation (here Province, Health region, and Neighbourhood) with other variables included in the model.
Province-level residuals analysis
The province-level residuals and associated standard error (S.E.) were used to plot and rank the OR of achieving the LPA threshold along with the 95% confidence interval (CI) for each province. As such, those provinces that differ from the national mean (above individual, neighbourhood and regional effects) can easily be identified.
Results
The distribution of all variables used in the analyses is found in Table 1. The distributions are comparable by sex and the largest difference is found in the outcome where 36.9% of the girls and 51.9% of the boys achieved the recommended level of LPA.
[Table 1: Outcome and covariates distribution for girls and boys]
All models present the OR of achieving the recommended LPA within 95% CI for each characteristic considered (fixed part). The between-area variance structure (random part) was analysed using the median OR (i.e. Province, Health Region, and Neighbourhood). Variation in the median OR was considered significant when 1.96 times its S.E. remained lower than the between area variance.
Apart from the null model, the reference category is aged 12-15, Caucasian, living in a highly educated household, reporting a normal weight, surveyed during summer, living in an urban setting (for Models 3 and 4), and living in the least deprived neighbourhoods (Model 4).
Results for the models on girls’ LPA
The null model shows a significant difference in achieving the recommended LPA between health regions and neighbourhoods, but not between provinces (Table 2, random part). Introducing individual characteristics (Model 2), and thus controlling for age, ethnic origin and body mass index category, shows lower odds of achieving the recommended LPA level among girls surveyed during winter (OR=0.58, 95% CI[0.54,0.62]) or the transitional season (OR=0.66, 95% CI[0.62,0.70]) as compared with summer. No differences in LPA was observed in the studied period from 2003 to 2012, but a significant variation between health region remains.
Introducing the MIZ of the neighbourhoods (e.g. urban, urban outskirt, or rural; Model 3) suggests that a lower likelihood of achieving the recommended LPA level exists for girls in urban areas as compared with those in urban outskirt areas (OR=1.12, 95% CI[1.04,1.21]) or rural areas (OR=1.13, 95% CI[1.04,1.23]). Further, the location of the neighbourhoods does not explain much of the significant variation observed between health regions (Var=0.014, S.E.=0.005), but shows an important increase in the variation explained at the neighbourhood level (Var=0.026, S.E.=0.009). The median OR indicates that girls living in a urban neighbourhood have a 12% increased chance (median OR=1.12) of achieving the recommended LPA level if they would move to a Health Region with a higher achievement of recommended LPA level, and a 17% increased chance if they lived in a province with a higher achievement level (median OR=1.17).
Model 4 introduced the deprivation level of the neighbourhood and shows no significant association in achieving the recommended LPA and reduces the goodness-of-fit as compared with Model 3.
[Table 2. Individual, cycle, season and contextual factors on girls’ leisure-time physical activity[AL3]]
Results for the models on boys’ LPA
As above, the null model shows a significant difference in achieving the recommended LPA only between health regions (Table 3). Introducing individual characteristics and season (Model 2) shows lower odds of achieving the recommended LPA level among boys surveyed during winter (OR=0.55, 95% CI[0.51,0.59]) or the transitional season (OR=0.70, 95% CI[0.67,0.74]) as compared with summer. A slight, but statistically significant, decrease in LPA was observed from 2003 to 2012 (OR=0.98, 95% CI[0.96,0.99]). Importantly, the individual characteristics improve the goodness-of-fit of the model, but do not explain the between health region variation suggesting an homogeneous distribution of these characteristics between areas.
Unlike for girls, the type of neighbourhood is not associated with the odds of achieving the recommended LPA level (Model 3). Nevertheless, the model now shows a significant variation between health regions (Var=0.017, S.E.=0.005) and between neighbourhoods (Var=0.023, S.E.=0.009) within the provinces. The median OR indicates that boys had a 13% increased chance (median OR=1.13) of achieving the recommended LPA level if they lived Health Region with a higher achievement of recommended LPA level and a 16% increased chance if they lived in a neighbourhood with a higher achievement level (median OR=1.16).
The deprivation level of the neighbourhood was introduced in Model 4, but shows no significant association with achieving the recommended level of LPA and reduces the goodness-of-fit as compared with Model 3.
[Table 3. Individual, cycle, season and contextual factors on boys’ leisure-time physical activity]
Variation between provinces
When individual characteristics and contextual factors are taken into account, the distribution of the odds of achieving the recommended LPA level showed no significant variation between the provinces in girls and boys (Model 3, Table 2 and Table 3). However, province-specific residuals are used here to identify provinces presenting significantly higher or lower odds of achieving the recommended LPA than the countrywide estimates. Girls living in Quebec were less likely to achieve the recommended level as compared with the national mean (Figure 1A), whereas girls living in Ontario and British-Columbia were more likely to achieve that threshold. Among boys (Figure 1B), only those living in Ontario were more likely to achieve the level, whereas boys living in other provinces remained within the CI of the national average.
[Figure 1. Province-level residuals of the logarithm of the odds ratio among (1A) Girls and (1B) Boys[AL4]]
Interpretation
Main findings
This study explored differences in reported LPA time among Canadian youth according to their geographic context, and mainly focused on the contextual factors. A four-level model was applied that controlled for age, ethnic origin, household education level and BMI. The results showed that beyond individual characteristics, the contextual area characteristics were associated with the odds of an individual accumulating an energy equivalent of at least 60 minutes of moderate- to vigorous-intensity LPA. Moreover, these contextual influences were sometimes different between girls and boys.
Explanation and comparison with other studies
Girls’ LPA was observed to be significantly lower than boys’(36.9% vs 51.9%) as reported in other Canadian investigations (33, 34), reinforcing the need to apply consideration to sex-specific needs when planning public health interventions that relate to LPA. A proposed area of improvement involves strategies to ensure equitable access to resources, including availability and access to suitable physical education classes and/or organized sports which may be subject to sex-related inequities (35, 36).
Season had a highly significant influence. Regardless of sex, winter was observed as a major barrier to performing LPA (girls: OR=0.58, 95% CI[0.54,0.62]; boys: OR=0.55, 95% CI[0.51,0.59]), even taking into account the potential addition of the compulsory physical education class in the results. This finding supports previous results that found an influence of seasonality on LPAamong various populations, including young Canadians (37, 38). For that reason, seasonality needs to be taken into account when developing interventions and programs targeting physical activity.
Living in an urban outskirt or a rural area [AL5]was associated with higher odds of meeting physical activity guidelines among girls as compared to urban neighbourhoods. This result conflicts with previous findings showing the opposite pattern (39). It suggests that facilities available in urban areas might be more suitable for boys (where this measure was not significant).[OW6]The analysis of the between neighbourhood variance further shows that most of the variation in the practice of LPA among Canadian youth occurs between urban neighbourhoods. This suggests that greater inequalities exist in LPA opportunities among urban neighbourhoods.
Whereas the between-province variance distribution showed no significant differences globally (girls: Var=0.028, S.E.=0.019; boys: Var=0.016, S.E.=0.013), residual analysis of provincial units highlighted some provinces that are different from the Canadian mean. The relatively poor odds of achieving the recommended LPA level among girls living in Quebec (log OR=-[OW7]0.240, 95% CI[-0.376,-0.103]) raises the question as to why Quebec girls tend to be less active than other Canadians, and provide rationale for further investigation on how physical activity is promoted in that province as compare to Ontario or British-Columbia where girls had higher LPA level than the national.