Abstract

The objective is to examine school-level program and policy characteristics and student-level behavioural characteristics associated with being overweight. Multilevel logistic regression analysis were used to examine the school- and student-level characteristics associated with the odds of a student being overweight among 1264 Grade 5–8 students attending 30 elementary schools in Ontario, Canada. Data were derived from the Physical Activity of Youth in Ontario Schools host study. Significant between-school random variation in overweight was identified [σμ02=0:187 (0:084), P < 0.001]; school-level differences accounted for 5.4% of the variability in the odds of a student being overweight versus a normal weight. A student attending a school that was in the action phase for the school-level construct ‘Availability and use of interschool programs’ was significantly less likely to be overweight than a similar student attending a school that was in the initiation phase for this construct. Important student-level characteristics included physical activity (PA) and gender. Developing a better understanding of the school- and student-level characteristics associated with overweight among youth is critical for informing school-based prevention policies. Future research should evaluate if implementing and promoting interschool PA programming, and which types of interschool activities, are effective in preventing or reducing overweight.

Introduction

The prevalence of overweight youth in North America has increased dramatically [1, 2] and it appears that the increasing trend is likely to continue in the future [3]. This is cause for concern as overweight among youth is not only associated with hypertension, high cholesterol and abnormal glucose tolerance [4, 5], but it is also associated with increased risk of some cancers [4–7]. Given the rapid increase in the prevalence of overweight among youth, it appears that modifiable factors [e.g. physical activity (PA), diet] are likely more important determinants of the current crisis than non-modifiable factors (e.g. genetics) [8, 9]. As such, it is suspected that the increased prevalence of obesity is primarily due to reduced energy expenditure and/or increased energy intake [8]. There is an immediate need to identify the modifiable characteristics associated with youth overweight in order to develop effective prevention initiatives.

Research has previously identified a number of individual characteristics that are associated with overweight among youth [4]. For instance, overweight youth are less likely to be physically active [10–12], less likely to participate in teams or organized sports [10, 13], more likely to spend time in sedentary screen-time behaviours [11, 13, 14], less likely to have physically active friends [12] and more likely to be male [10, 12, 13]. Although research has identified that some dietary factors are associated with overweight among older adolescents, data from the Continuing Survey of Food Intake for Individuals and the National Health and Nutrition Examination Survey III suggest that many dietary factors are not strongly associated with overweight among children [10]. As such, the present study focuses on examining the link between overweight and characteristics associated with PA, not diet.

The school environment can play an important role in providing youth with opportunities to engage in activities that help to prevent or reduce overweight. For instance, school-based sports programs represent key opportunities where youth can engage in PA of sufficient quantity and quality to maintain a healthy weight status [15]. However, few studies have examined the school characteristics associated with that variability [16]. Existing evidence [16–21] suggests that there is modest variability in overweight between schools (∼2–5%), and demographic characteristics (e.g. school size and school socioeconomic status) appear to be associated with student overweight [16, 19, 20]. However, the lack of evidence exploring how policies within the school environment are associated with an individual student's risk of being overweight continues to represent an important gap in the literature. If we really want to explain a large amount of the variance in child overweight, we must simultaneously examine both school policies and student characteristics and determine if there are any interactive effects [21].

As such, the current study seeks to: (i) characterize the prevalence of weight status, PA, screen time and sports participation in a sample of elementary school students, (ii) characterize the prevalence of different programs and policies within the school environment, (iii) identify if there is significant variability in the odds of being overweight across schools and (iv) explore if student- and school-level characteristics are associated with the odds of a student being overweight.

Materials and methods

Design

This cross-sectional study used self-reported data collected in 2007–2008 from a convenience sample of Grade 5–8 students attending 30 elementary schools in Ontario, Canada, as part of the Physical Activity of Youth in Ontario Schools (PLAY-On) study. The purpose of the PLAY-On study was to better understand PA and correlates of PA among a sample of Ontario elementary school students. With the Play-On host study, student-level data were collected from consenting students using the School Health Action, Planning and Evaluation System (SHAPES) Physical Activity Module (PAM) [22]. The PAM asks students about PA, height and weight, sedentary behaviours, correlates for PA and participation in teams and sporting activities at school. Additional details about the PLAY-On study are available online (http://www.shapes.uwaterloo.ca/projects/PLAYON) or in print [23], and additional details about SHAPES and the PAM measures are available [22, 24].

School-level data were collected using the PA categories of the elementary school version of the Healthy School Planner (HSP) [25]. The HSP was designed to assist schools in determining the health status of their school and support them in developing an action plan for making improvements in their school. The planner is based on comprehensive school health (CSH), an internationally recognized framework for supporting improvements in students’ educational outcomes while addressing school health in a planned, integrated and holistic way. The HSP was developed by staff at the Propel Centre for Population Health Impact at the University of Waterloo, under the guidance of an Advisory Committee formed by the JCSH funded by the Public Health Agency of Canada. The PA assessment portion of the HSP was derived from the SHAPES Healthy School Environment Survey and the Michigan Healthy Schools Action Tool. The HSP PA assessment tool used in PLAY-On is designed to assess programs, activities, committees, facilities and guidelines surrounding PA in the school environment. The four PA categories included in the HSP tool are aligned with the Government of Ontario's Foundations for a Healthy School (FHS) [26].

Data collection

In September 2007 eight Ontario schools boards were approached asking permission to recruit schools to participate in the PLAY-On study. Seven school boards agreed to participate. Since there were time constraints (PLAY-On data had to be collected during the 2007–2008 academic year) and budget limitations, random sampling could not be used. As such, the PLAY-On team worked closely with participating school boards to identify and recruit a convenience sample of 30 elementary schools. Within the 30 schools that agreed to participate, data collection occurred between November 2007 and April 2008. All students at the participating schools were eligible to participate. Prior to participating in the study, active consent from parents was required and at any time students were able to decline participation. The PAM was completed by eligible students during class time and there was no compensation for participation. At each participating school, the administrator most knowledgeable about the school's programs, policies and resources was asked to complete the HSP survey. The University of Waterloo Office of Research Ethics and appropriate School Board Ethics committees approved the study procedures.

Participants

Of the 4838 students enrolled in grades 5–8 at the 30 participating elementary schools, 50.6% (n = 2449) completed the survey. Missing respondents resulted from parent refusal (46.2%; n = 2237) and absenteeism on the day of the survey (3.2%; n = 152). This distribution is consistent with a previous active consent study examining overweight among Canadian elementary students [19]. Among participating students, body mass index (BMI) could not be calculated for 48.4% (n = 1185) of them due to missing data required to calculate BMI; 16.2% (n = 398) did not report their height and weight, 22.8% (n = 558) reported their weight but not height, 8.0% (n = 196) reported their height but not weight and 1.3% (n = 33) did not report their age. Overall, data from 1264 students were available for the analyses. Previous studies with youth populations have reported similar rates of missing height and weight data [27, 28]. All 30 elementary schools completed the HSP survey.

Measures

Outcome—weight status

BMI is a number calculated from weight and height which is used as a reliable indicator of body fatness for children [29]. In the present study, BMI was calculated using previously validated self-report measures of weight (kg) and height (m) (BMI = kg m−2) [22]. Validity testing has previously demonstrated significant criterion validity based on Spearman correlations for our self-reported measures of height (r = 0.97, P < 0.001) and weight (r = 0.98, P < 0.001), and reliability testing has demonstrated 1-week test–retest reliability based on weighted kappa coefficients for our self-reported BMI (K = 0.75, P < 0.01) [22]. Consistent with CDC guidelines and growth charts [30], students within the lowest 5th percentile for BMI adjusted for age and sex were classified as underweight, students within the 6th–84th percentile for BMI adjusted for age and sex were classified as normal weight, students within the 85th–94th percentile for BMI adjusted for age and sex were classified as overweight, and students within the highest 5th percentile for BMI adjusted for age and sex were classified as obese. For the multivariate analyses, students classified as overweight or obese were collapsed into one category (overweight) to represent all youth who may be at-risk for morbidity associated with being overweight and to ensure there was sufficient power for the multilevel analyses.

Student-level correlates

Physical activity level was measured by asking respondents how many minutes of vigorous physical activity (VPA) and moderate physical activity they engaged in on each of the last 7 days. Since youth tend to substantially over-report time spent doing PA in self-report [22, 31], the measures are more valid for differentiating students who report less time doing PA from those who report more time doing PA [22]. Hence, consistent with the existing literature [14, 23, 32], students more than 1 SD (≤16th percentile) below the sample mean were classified as low active, students more than 1 SD (≥84th percentile) above the sample mean were classified as high active; all others were classified as moderately active.

The measures for sedentary behaviour, sports participation and social influences were consistent with previous research [14, 23, 32]. Respondents reported the number of hours for each day of the week that they spent watching TV/movies, playing video/computer games, surfing the Internet, instant messaging or talking on the phone. The average screen time per day was calculated based on the average time reported over the previous week and grouped responses into three categories (<1, 1–3, >3 hours per day). Respondents also reported whether or not they participate in varsity or team sports at school (yes/no), whether or not they participate in intramural or house league sports at school (yes/no), whether or not they participate in league or team sports outside of school (yes/no), and how many of their close friends are physically active (0–5).

School-level characteristics

Consistent with the four components that form the basis of the FHS [26], the HSP PA tool measured indicators associated with: Healthy Physical Environment (availability of, access to, and adequacy in meeting student needs for, indoor and outdoor facilities, equipment and resources for safe, quality PA on or near school grounds, both during and outside of school hours); Instruction and Programs (availability, delivery and characteristics of curricular physical education, extracurricular PA programs and active transportation to school, including barriers to implementing such programs); Supportive Social Environment (characteristics of the school's social environment that predispose, reinforce and enable enjoyable, lifelong participation in PA or that hinder such activities) and Community Partnerships (the accessibility and availability of support services for PA which may include partnerships with public health units and community-based services and resources). The HSP does not measure the presence or absence of specific obesity prevention or PA promotion interventions within the school. Each indicator was assigned a classification based on the corresponding phase of implementation in the Healthy School Continuum as outlined by the Joint Consortium on School Health [33]: ‘Initiation’ (falls short or exhibits extensive room for improvement in meeting the recommendations related to school capacity for PA); ‘Action’ (meets the recommendations in several, but not all areas related to school capacity for PA, exhibits some room for improvement) and ‘Maintenance’ (consistently meets or exceeds the recommendations related to school capacity for PA, encouraged to maintain the current level of commitment to supporting PA at school). Each of the four FHS components was also assigned an ‘overall’ phase classification based on the combined responses to component indicators. The assessment schemes for the HSP measures were developed and designed to ensure concrete validity based on recommendations from current research literature, Government of Ontario guidelines and input from experts on PA in schools [25].

Analyses

Using student-level data, the prevalence of weight status, PA, screen time, social influences and team sports were examined by sex. Using school-level data, the prevalence of the FHS indicators for Healthy School Environment, Instruction and Programs, Supportive Social Environment and Community Partnerships were examined by the phase of implementation. To understand the student- and school-level factors associated with being overweight, a series of multilevel logistic regression analyses were performed. Consistent with other multilevel studies [34, 35], a step modelling procedure was used. Step 1 examined if differences in being overweight were random or fixed across schools. The school-level variance term from Step 1 (σμ02) was used to calculate the intraclass correlation (ICC) for binary outcomes (The formula for calculating the ICC for binary outcomes, where ICC = forumla), where the ICC represents the proportion of the total variance in overweight that is due to differences across schools [36]. In Step 2, a series of univariate analyses examined if each of the FHS indicators were associated with being overweight (Model 1) and another series of univariate analyses examined if the overall score for each FHS indicator was associated with being overweight (Model 2). Only significant school-level variables (P < 0.05) were retained for further analyses. In Step 3, a multivariate model was developed to examine how the student characteristics and the significant school characteristics identified in Step 2 were associated with being overweight (final model). After the final model, contextual interactions between all the significant (P < 0.05) school and student characteristics were explored. Since there was a large amount of missing BMI data, a sensitivity analysis was performed by imputing some of the missing BMI data. Research with the PLAY-On previously identified that students were more likely to have missing BMI data if they perceived themselves to be overweight [37]. As such, among the respondents with missing BMI data, we identified those who reported that they consider themselves overweight (n = 244) and grouped them with the respondents already classified as overweight and re-ran the final model. Statistical analyses were conducted on MLwiN Version 2.02 [38].

Results

Student characteristics

Demographic characteristics of students who provided BMI data are presented in Table I. The sample was 52.4% (n = 663) male and 47.5% (n = 601) female. The average age was 11.8 (±1.1) years; the age distribution was not significantly different between males and females [χ2 = 4.75, degrees of freedom (df) = 5, P = 0.447]. The mean BMI among males was 19.5 (±3.8) kg m−2 and 19.1 (±4.1) kg m−2 among females. Overall, 4.5% of the sample was considered underweight, 17.4% of the sample was considered overweight, 9.7% of the sample was considered obese and 68.4% were considered normal weight for their age and sex. Males were more likely to be overweight or obese compared with females (χ2 = 12.46, df = 3, P < 0.01). A total of 166 (13.1%) students were classified as low active, 845 (66.9%) were classified as moderately active and 253 (20.0%) were classified as highly active. Males were more likely to be high active compared with females (χ2 = 17.53, df = 2, P < 0.001). The majority of students (63.9%) reported 1–3 h of screen time per day. Males were more likely than females to report spending 3 or more hours pay day on screen time activities (χ2 = 12.17, df = 2, P < 0.01). Very few students reported that they have less than three close friends who are physically active (8.6%). Males were more likely than females to report having five close friends who are physically active (χ2 = 38.41, df = 5, P < 0.001). The majority of students also reported that they do participate in intramural sports at school (68.3%), varsity sports at school (70.5%) or league sports outside of school (77.2%). There were no gender differences in the prevalence of students participating in different team sports.

Table I.

Descriptive statistics for youth in grades 5–8 who provided data to calculate BMI by gender (Ontario, Canada)

Student-level characteristics  Male (n = 663), % (n)a Female (n = 601), % (n)a Chi square 
Weight statusb Underweight 4.4 (29) 4.7 (28) χ2 = 43.24, df = 3, P < 0.001 
Normal weight 64.5 (428) 72.7 (437) 
Overweight 19.3 (128) 15.3 (92) 
Obese 11.8 (78) 7.3 (44) 
Physical activity level High active 24.4 (162) 15.1 (91) χ2 = 17.53, df = 2, P < 0.001 
Moderately active 63.8 (423) 70.2 (422) 
Low active 11.8 (78) 14.6 (88) 
Screen time per day <1 hour per day 18.8 (124) 24.3 (145) χ2 = 12.17, df = 2, P < 0.01 
1–3 hours per day 63.6 (419) 64.2 (384) 
>3 hours per day 17.6 (116) 11.5 (69) 
Number of close friends who are physically active None 0.5 (3) 0.0 (0) χ2 = 38.41, df = 5, P < 0.001 
1.8 (12) 2.5 (15) 
6.2 (40) 6.2 (37) 
13.9 (90) 26.7 (159) 
26.8 (174) 25.4 (151) 
50.8 (329) 39.2 (233) 
Intramural sports at school Does not participate 32.1 (187) 31.3 (198) χ2 = 0.09, df = 1, P = 0.764 
Does participate 67.9 (395) 68.7 (434) 
Varsity sports at school Does not participate 27.5 (174) 31.7 (185) χ2 = 2.63, df = 1, P = 0.105 
Does participate 72.5 (459) 68.3 (398) 
League sports outside of school Does not participate 39.2 (233) 35.1 (232) χ2 = 2.16, df = 1, P = 0.142 
Does participate 60.8 (362) 64.9 (428) 
Grade 16.6 (110) 14.3 (86) χ2 = 1.76, df = 3, P = 0.624 
23.4 (155) 24.6 (148) 
28.6 (190) 30.6 (184) 
31.4 (208) 30.5 (183) 
Student-level characteristics  Male (n = 663), % (n)a Female (n = 601), % (n)a Chi square 
Weight statusb Underweight 4.4 (29) 4.7 (28) χ2 = 43.24, df = 3, P < 0.001 
Normal weight 64.5 (428) 72.7 (437) 
Overweight 19.3 (128) 15.3 (92) 
Obese 11.8 (78) 7.3 (44) 
Physical activity level High active 24.4 (162) 15.1 (91) χ2 = 17.53, df = 2, P < 0.001 
Moderately active 63.8 (423) 70.2 (422) 
Low active 11.8 (78) 14.6 (88) 
Screen time per day <1 hour per day 18.8 (124) 24.3 (145) χ2 = 12.17, df = 2, P < 0.01 
1–3 hours per day 63.6 (419) 64.2 (384) 
>3 hours per day 17.6 (116) 11.5 (69) 
Number of close friends who are physically active None 0.5 (3) 0.0 (0) χ2 = 38.41, df = 5, P < 0.001 
1.8 (12) 2.5 (15) 
6.2 (40) 6.2 (37) 
13.9 (90) 26.7 (159) 
26.8 (174) 25.4 (151) 
50.8 (329) 39.2 (233) 
Intramural sports at school Does not participate 32.1 (187) 31.3 (198) χ2 = 0.09, df = 1, P = 0.764 
Does participate 67.9 (395) 68.7 (434) 
Varsity sports at school Does not participate 27.5 (174) 31.7 (185) χ2 = 2.63, df = 1, P = 0.105 
Does participate 72.5 (459) 68.3 (398) 
League sports outside of school Does not participate 39.2 (233) 35.1 (232) χ2 = 2.16, df = 1, P = 0.142 
Does participate 60.8 (362) 64.9 (428) 
Grade 16.6 (110) 14.3 (86) χ2 = 1.76, df = 3, P = 0.624 
23.4 (155) 24.6 (148) 
28.6 (190) 30.6 (184) 
31.4 (208) 30.5 (183) 
a

Numbers may not add to total because of missing values.

b

BMI values used to determine weight status have been adjusted for age and gender.

Demographic characteristics comparing students who provided BMI data to those who did not provide sufficient data to calculate BMI are presented in Table II. Females were more likely to have BMI data missing compared with males (χ2 = 28.2, df = 1, P < 0.01). BMI data were more likely to be missing among low-active students compared with high-active students (χ2 = 33.6, df = 2, P < 0.001). BMI data were also more likely to be missing among students who report that they do not participate in intramural sports at school (χ2 = 19.8, df = 1, P < 0.001), varsity sports at school (χ2 = 27.6, df = 1, P < 0.001) or league sports outside school (χ2 = 30.1, df = 1, P < 0.001).

Table II.

Descriptive statistics for youth in grades 5–8 by whether or data were available to calculate BMI (Ontario, Canada)

  BMI data available (n = 1264), % (n)a BMI data missing (n = 1185), % (n)a Chi square 
Sex Male 52.6 (626) 41.5 (462) χ2 = 28.2, df = 1, P < 0.001 
Female 47.4 (565) 58.5 (651) 
Physical activity level High active 19.3 (230) 12.5 (139) χ2 = 33.6, df = 2, P < 0.001 
Moderately active 67.7 (806) 67.6 (422) 
Low active 13.0 (78) 19.9 (222) 
Screen time per day <1 hour per day 21.5 (275) 22.6 (254) χ2 = 0.87, df = 2, P = 0.646 
1 to 3 hours per day 63.8 (817) 61.9 (698) 
>3 hours per day 14.7 (189) 15.5 (175) 
Number of close friends who are physically active None to two friends 8.4 (99) 10.9 (119) χ2 = 3.8, df = 1, P = 0.052 
Three or more 91.6 (1,073) 89.1 (977) 
Intramural sports at school Does not participate 31.7 (363) 40.8 (430) χ2 = 19.8, df = 1, P < 0.001 
Does participate 68.3 (783) 59.2 (624) 
Varsity sports at school Does not participate 29.7 (340) 40.3 (432) χ2 = 27.6, df = 1, P < 0.001 
Does participate 70.3 (806) 59.7 (640) 
League sports outside of school Does not participate 22.8 (264) 33.2 (359) χ2 = 30.1, df = 1, P < 0.001 
Does participate 77.2 (892) 66.8 (721) 
  BMI data available (n = 1264), % (n)a BMI data missing (n = 1185), % (n)a Chi square 
Sex Male 52.6 (626) 41.5 (462) χ2 = 28.2, df = 1, P < 0.001 
Female 47.4 (565) 58.5 (651) 
Physical activity level High active 19.3 (230) 12.5 (139) χ2 = 33.6, df = 2, P < 0.001 
Moderately active 67.7 (806) 67.6 (422) 
Low active 13.0 (78) 19.9 (222) 
Screen time per day <1 hour per day 21.5 (275) 22.6 (254) χ2 = 0.87, df = 2, P = 0.646 
1 to 3 hours per day 63.8 (817) 61.9 (698) 
>3 hours per day 14.7 (189) 15.5 (175) 
Number of close friends who are physically active None to two friends 8.4 (99) 10.9 (119) χ2 = 3.8, df = 1, P = 0.052 
Three or more 91.6 (1,073) 89.1 (977) 
Intramural sports at school Does not participate 31.7 (363) 40.8 (430) χ2 = 19.8, df = 1, P < 0.001 
Does participate 68.3 (783) 59.2 (624) 
Varsity sports at school Does not participate 29.7 (340) 40.3 (432) χ2 = 27.6, df = 1, P < 0.001 
Does participate 70.3 (806) 59.7 (640) 
League sports outside of school Does not participate 22.8 (264) 33.2 (359) χ2 = 30.1, df = 1, P < 0.001 
Does participate 77.2 (892) 66.8 (721) 
a

Numbers may not add to total because of missing values.

School characteristics

The mean prevalence of overweight students at a school was 28.3% (range: 7.1–53.9%). There were also differences among the 30 elementary schools for the school-level environmental characteristics examined. The majority of schools were in the action phase for the overall indicator scores for Healthy Physical Environment (66.7%) and Supportive Social Environment (66.7%) and the maintenance phase for the overall score for Community Partnerships (56.6%). Conversely, the majority of schools were in the initiation phase for the overall score for Instruction and Programs (73.3%). Overall, no schools were in the maintenance phase for the overall scores for Healthy Physical Environment, Instruction and Programs and Supportive Social Environment. Within each of the four FHS components, there was substantial variability across schools in relation to the individual indicators measured. The descriptive statistics for these school-level indicators have been presented elsewhere [23].

School characteristics associated with overweight

Significant between-school random variation in the odds of being overweight was identified [σ2μ0 = 0.187 (0.084), P < 0.001]; school-level differences accounted for 5.4% of the variability in the odds of a student being overweight versus a normal weight. Fig. 1 illustrates the hypothetical population-level impact associated with skewing or normalizing the BMI distribution by the 5.4% between-school variance identified. As shown in Table III, univariate analyses identified that ‘Instructions and Programs’ was the only FHS category with a school-level indicator significantly associated with the likelihood of a student being overweight. A student was less likely to be overweight if he/she attended a school that was in the action phase for the indicator ‘Availability and use of interschool programs’ compared with a student attending a school that was in the initiation phase for that indicator [β-0.60 (0.15), P < 0.001]. A similar protective effect was identified for the maintenance phase for this indicator [β-0.41 (0.59)]; however, the association was not significant likely as a result of there only being two schools in the maintenance phase. Interestingly, none of the overall scores for the four FHS components were significantly associated with being overweight.

Fig. 1.

Illustration of the potential population-level impact resulting from school-based interventions that have an effect of either skewing or normalizing the population distribution for BMI.

Fig. 1.

Illustration of the potential population-level impact resulting from school-based interventions that have an effect of either skewing or normalizing the population distribution for BMI.

School- and student-level characteristics associated with overweight

The adjusted odds ratios for the significant school and student characteristics in the Final Model are presented in Table IV. Students who were moderately active [odds ratio (OR) 0.38, 95% confidence interval (CI) 0.26–0.55] or high active (OR 0.29, 95% CI 0.18–0.47) were substantially less likely to be overweight compared with low-active students. Male students were more likely to be overweight than female students (OR 1.52, 95% CI 1.16–2.00). Screen time, participation in intramural, varsity or league sports and having close friends who were physically active were not significantly associated with being overweight. There was one significant school characteristic associated with being overweight. If a student attended a school that was in the action phase for the indicator ‘Availability and use of interschool programs’, he/she was less likely to be overweight than a similar student attending a school that was in the initiation phase for that indicator. No significant contextual interactions between school- and student-level characteristics were identified. Moreover, the results of the sensitivity analysis using imputed BMI values (data not shown) were consistent with the results presented in Table IV.

Table III.

Multilevel logistic regression analyses examining school-level factors associated with being overweight among youth in grades 5–8 (Ontario, Canada)

Model estimates (standard error)
 
Phase  Model 1—overweight versus normal weight Model 2—overweight versus normal weight 
Healthy Physical Environment 
    Student access to a variety of facilities on and off school grounds during school hoursa Action 0.18 (0.68)  
Maintenance 0.63 (0.68)  
    Availability of physical activities during inclement weathera Action −0.29 (0.25)  
Maintenance −0.65 (0.55)  
    Student access to facilities and equipment outside of school hoursa Action −0.08 (0.32)  
Maintenance 0.08 (0.54)  
    Support for active transportation to and from schoola Action −0.36 (0.33)  
Maintenance −0.04 (0.31)  
Overall score for this indicatora Action  −0.02 (0.21) 
Instruction and Programs 
    Implementation of daily PAb Maintenance 0.22 (0.30)  
    Time spent per week engaged in PA during physical education classesa Action −0.61 (0.72)  
Maintenance 0.01 (0.61)  
    Classes taught by a qualified physical education specialista Action 0.29 (0.30)  
    Availability and use of intramural/club activitiesa Action −0.22 (0.25)  
Maintenance −0.01 (0.41)  
    Consistency of intramural programming across grade divisions and seasonsa Action −0.07 (0.20)  
Maintenance 0.12 (0.26)  
    Availability and use of interschool programsa Action −0.60 (0.15)*  
Maintenance −0.41 (0.59)  
    Consistency of interschool programming across seasonsa Maintenance 0.11 (0.13)  
Overall score for this indicatora Action  0.07 (0.22) 
Supportive Social Environment 
    Emphasis placed on maximizing participation in PA through school programsa Action −0.92 (0.63)  
Maintenance −0.45 (0.60)  
    Incorporation of PA into other school subjectsa Action −0.09 (0.52)  
Maintenance −0.81 (0.72)  
    Special recognition of students who participate in school physical activitiesa Action 0.59 (0.70)  
Maintenance 0.42 (0.64)  
    Formal collection of suggestions from the school community about PA at schoola Action −0.06 (0.39)  
Maintenance 0.47 (0.84)  
    Promotion of PA programs and events for students, families and school staffa Action −0.52 (0.46)  
Maintenance −0.29 (0.54)  
    Use of PA as a reward, not as disciplinea Action 0.17 (0.46)  
Maintenance 0.11 (0.54)  
    Presence of written policies or practices that support PAa Action 0.45 (0.50)  
Maintenance 0.61 (0.59)  
Overall score for this indicatora Action  0.11 (0.22) 
Community Partnerships 
    Support available for school staff involved with PAb Maintenance −0.08 (0.27)  
    Connection to community resourcesa Action 0.45 (0.40)  
Maintenance 0.32 (0.31)  
Overall score for this indicatora Action  0.38 (0.31) 
Maintenance  0.34 (0.28) 
Model estimates (standard error)
 
Phase  Model 1—overweight versus normal weight Model 2—overweight versus normal weight 
Healthy Physical Environment 
    Student access to a variety of facilities on and off school grounds during school hoursa Action 0.18 (0.68)  
Maintenance 0.63 (0.68)  
    Availability of physical activities during inclement weathera Action −0.29 (0.25)  
Maintenance −0.65 (0.55)  
    Student access to facilities and equipment outside of school hoursa Action −0.08 (0.32)  
Maintenance 0.08 (0.54)  
    Support for active transportation to and from schoola Action −0.36 (0.33)  
Maintenance −0.04 (0.31)  
Overall score for this indicatora Action  −0.02 (0.21) 
Instruction and Programs 
    Implementation of daily PAb Maintenance 0.22 (0.30)  
    Time spent per week engaged in PA during physical education classesa Action −0.61 (0.72)  
Maintenance 0.01 (0.61)  
    Classes taught by a qualified physical education specialista Action 0.29 (0.30)  
    Availability and use of intramural/club activitiesa Action −0.22 (0.25)  
Maintenance −0.01 (0.41)  
    Consistency of intramural programming across grade divisions and seasonsa Action −0.07 (0.20)  
Maintenance 0.12 (0.26)  
    Availability and use of interschool programsa Action −0.60 (0.15)*  
Maintenance −0.41 (0.59)  
    Consistency of interschool programming across seasonsa Maintenance 0.11 (0.13)  
Overall score for this indicatora Action  0.07 (0.22) 
Supportive Social Environment 
    Emphasis placed on maximizing participation in PA through school programsa Action −0.92 (0.63)  
Maintenance −0.45 (0.60)  
    Incorporation of PA into other school subjectsa Action −0.09 (0.52)  
Maintenance −0.81 (0.72)  
    Special recognition of students who participate in school physical activitiesa Action 0.59 (0.70)  
Maintenance 0.42 (0.64)  
    Formal collection of suggestions from the school community about PA at schoola Action −0.06 (0.39)  
Maintenance 0.47 (0.84)  
    Promotion of PA programs and events for students, families and school staffa Action −0.52 (0.46)  
Maintenance −0.29 (0.54)  
    Use of PA as a reward, not as disciplinea Action 0.17 (0.46)  
Maintenance 0.11 (0.54)  
    Presence of written policies or practices that support PAa Action 0.45 (0.50)  
Maintenance 0.61 (0.59)  
Overall score for this indicatora Action  0.11 (0.22) 
Community Partnerships 
    Support available for school staff involved with PAb Maintenance −0.08 (0.27)  
    Connection to community resourcesa Action 0.45 (0.40)  
Maintenance 0.32 (0.31)  
Overall score for this indicatora Action  0.38 (0.31) 
Maintenance  0.34 (0.28) 

1 = overweight (n = 342), 0 = normal weight (n = 865).

a

Reference group is Initiation.

b

Reference group is Action. *P < 0.001.

Table IV.

Odds ratios for school- and student-level factors associated with being overweight among youth in grades 5–8 (Ontario, Canada)

  Adjusted odds ratioa (95% CI)—Final Model—overweight versus normal weight 
Student-level characteristics 
    Physical activity level Low active 1.00 
Moderately active 0.38 (0.26, 0.55)** 
High active 0.29 (0.18, 0.47)** 
    Screen time per day <1 hour per day 1.00 
1–3 hours per day 0.87 (0.62, 1.21) 
>3 hours per day 1.05 (0.68, 1.65) 
    Number of close friends who are physically active None to two friends 1.00 
Three or more 0.86 (0.56, 1.31) 
    Intramural sports at school Does not participate 1.00 
Does participate 1.16 (0.82, 1.64) 
    Varsity sports at school Does not participate 1.00 
Does participate 0.87 (0.61, 1.23) 
    League sports outside of school Does not participate 1.00 
Does participate 0.94 (0.67, 1.32) 
    Gender Female 1.00 
Male 1.52 (1.16, 2.00)* 
School-level characteristics 
    Availability and use of interschool programs Initiation 1.00 
Action 0.55 (0.41, 0.74)** 
Maintenance 0.52 (0.13, 2.09) 
  Adjusted odds ratioa (95% CI)—Final Model—overweight versus normal weight 
Student-level characteristics 
    Physical activity level Low active 1.00 
Moderately active 0.38 (0.26, 0.55)** 
High active 0.29 (0.18, 0.47)** 
    Screen time per day <1 hour per day 1.00 
1–3 hours per day 0.87 (0.62, 1.21) 
>3 hours per day 1.05 (0.68, 1.65) 
    Number of close friends who are physically active None to two friends 1.00 
Three or more 0.86 (0.56, 1.31) 
    Intramural sports at school Does not participate 1.00 
Does participate 1.16 (0.82, 1.64) 
    Varsity sports at school Does not participate 1.00 
Does participate 0.87 (0.61, 1.23) 
    League sports outside of school Does not participate 1.00 
Does participate 0.94 (0.67, 1.32) 
    Gender Female 1.00 
Male 1.52 (1.16, 2.00)* 
School-level characteristics 
    Availability and use of interschool programs Initiation 1.00 
Action 0.55 (0.41, 0.74)** 
Maintenance 0.52 (0.13, 2.09) 

Final model: overweight (n = 342), 0 = normal weight (n = 865).

a

Odds ratios adjusted for all other variables in the table and controlling for grade. *P < 0.01, **P < 0.001.

Discussion

Progress in reducing or limiting the increase in the prevalence of youth overweight will require efforts from many different stakeholders in many different contexts. While school-based interventions alone will not be sufficient to solve the problem, it is unlikely that the current trends in youth overweight can be reversed without more effective school-based programming [39]. Considering that the school environment represents one of the key contexts for intervening, developing a better understanding of the modifiable school- and student-level factors associated with overweight among youth is critical for informing future programs and policies. In this study, a substantial number of youth were considered overweight, despite using self-report height and weight measures and the large amount of missing data. Moreover, the likelihood of youth in this sample being overweight was significantly associated with both school and individual characteristics. Given the sample demographics are consistent with other North American youth populations [1, 40], these findings should be fairly representative within that context.

Consistent with existing research [16–21], among students who provided self-reported height and weight data, there were significant differences in the risk of being overweight across schools within the PLAY-On sample. Specifically, it was identified that school-level differences accounted for 5.4% of the variability in the odds of a student being overweight. Although this may appear relatively modest relative to the variability which would be accounted for at the student level, this represents a substantial amount of variation which could be amenable to school-level intervention. Consistent with the concept of the ‘prevention paradox’ [41], if a school-level program or policy has a small effect on either shifting or normalizing the distribution of overweight across schools, the impact at the student-level would be substantial (Fig. 1). Even a modest school-level effect could impact a substantial portion of the student population. Moreover, these data can provide practitioners with realistic insight about the potential impact that even the most successful school-level interventions could be expected to have on the student population. Future studies should report the between-school variability in BMI identified in their samples, and in the case of school-level intervention studies, the pre- and post-intervention frequency distributions at the student-level should also be reported in order to better understand the impact that the intervention had from a population-level perspective.

Interschool programs consist of team or individual sport competitions between schools that are available to students. Although interschool programs typically emphasize competition and winning and do not serve the entire student body, these data suggest that these programs may be an important factor in combating youth overweight. Even when controlling for individual characteristics (including PA, screen time and sport participation at school), it was identified that a student attending a school in the action phase for ‘Availability and use of interschool programs’ was significantly less likely to be overweight than a similar student attending a school in the initiation phase for this construct. Although the availability of interschool programs is not directly associated with PA levels among youth [23], their impact on overweight may function indirectly through influencing PA patterns as research has previously suggested that interschool programs can provide important opportunities for youth to be active [42]. According to Rye et al. [42], such opportunities may be particularly important for youth who lack the time, skills or confidence to play in other organized sporting activities or youth living in rural or lower SES areas where access to community recreational facilities may be limited. Schools and school boards could look for creative options to increase funding and supervision to overcome barriers of providing these programs and seek to improve student access to activities by offering a variety of sports and available times from which the students can choose [43, 44]. Research is required to evaluate if implementing interschool programs, and which types of programs, are effective in preventing or reducing overweight. Moreover, considering that public health departments are currently underutilized in obesity prevention programming [45], research could also evaluate the benefit of linking public health capacity to compliment school-based obesity prevention programs outside school hours (i.e. interschool programs).

Numerous studies have examined individual-level factors associated with youth overweight. Consistent with this research [10–15], it was identified that physically active youth are less likely to be overweight than inactive youth. However, many of the associations between other student characteristics and overweight previously identified in the literature were not significant in this study. For instance, overweight was not associated with participating in teams or organized sports, time spent in sedentary behaviour or having physically active friends. Considering that this was an elementary school sample (rather than an older secondary student sample where such associations are typically reported) and that the analytical focus was geared towards identifying school-level correlates associated with overweight, this may not be surprising. Moreover, it may be that controlling for the presence or absence of interschool programs at the school-level may have resulted in some associations at the student level not being significant. For example, a school would have to offer interschool programs for a student to be able to participate in such programs at school.

Previous research on older adolescent populations has also reported problems with large amounts of missing self-reported BMI data [27, 28]. Both Hines and Faricy [27] and Tiggemann [28] report that missing data was associated with age, where BMI data were more likely to be missing among younger respondents. Consistent with those studies, additional analyses performed with the PLAY-On data identified an age-related trend where the prevalence of missing BMI data was higher for younger respondents [37]. Moreover, considering that the PLAY-On sample was composed of even younger respondents than the samples examined in previous studies [27, 28], a lack of awareness of one's height or weight among the younger aged respondents in the PLAY-On sample may be underlying this non-response. Exploratory analyses also revealed that students with missing BMI data in PLAY-On were more likely to perceive themselves to be overweight or to be low active and highly sedentary [37]. When the sensitivity analyses were performed, these assumptions were used to impute missing values among the sample. The results of the sensitivity analyses (data not shown) were almost identical to the results from our primary analyses. Despite such attempts to demonstrate these findings are robust, it is possible that some of the results identified in the present study may be biased (e.g. the lack of associations between school characteristics or sedentary behaviour and overweight); consequently, the results presented should be interpreted with caution. Strategies to reduce missing data in the self-reporting of height and weight in ways which are sensitive to children need to be developed and evaluated if self-report is used exclusively to monitor and evaluate obesity-related interventions and initiatives.

Limitations

This study is subject to some limitations. A large amount of BMI data were missing in the host study and analyses revealed that there were some differences in the sample characteristics of students with and without BMI data which may bias the results. The missing BMI values may be a result of motivated non-responding, which would have important implications for the feasibility of using self-report weight and BMI as a target for intervention and surveillance efforts. Data were not available to examine energy intake in the present study so the results presented only address part of the issue of weight status. Future research needs to include measures of energy consumption in order to better understand the issue of weight status [4]. Moreover, given that only one of the school-level PA indicators examined in this study was associated with overweight, a better understanding of the school characteristics that impact overweight might be garnered by examining the presence or absence of specific obesity prevention or PA promotion interventions within the school and/or school-level indicators associated with diet. Causal relationships cannot be inferred from these cross-sectional data. Although data were based on self-reports, the measures in the PAM have been previously demonstrated to be reliable and valid [22], and honest reporting was encouraged by ensuring confidentiality during data collection.

Conclusion

Developing a better understanding of the modifiable school- and student-level factors associated with overweight among youth is critical for informing intervention programs and policies designed to prevent and reduce overweight among youth populations. It was identified that even though most of the students in this sample are considered a healthy body weight, a substantial number of youth were overweight or obese. Moreover, youth in our sample were less likely to be overweight if they attended a school where interschool programs were available for students. Future research should evaluate if the optimal population-level impact for school-based obesity prevention programming might be achieved most economically if interventions selectively targeted the schools that are putting students at the greatest risk.

Funding

Heart and Stroke Foundation of Ontario (to S. L.).

Conflict of interest statement

There are no conflicts of interest or competing financial interests in relation to this work.

The project was conducted by the Population Health Research Group at the University of Waterloo under the management of Chad Bredin. Dr Leatherdale is a Cancer Care Ontario Research Chair in Population Studies. The Canadian Cancer Society provided funding to develop SHAPES, the system used to collect the PLAY-On data.

References

1.
Shields
M
Overweight and obesity among children and youth
Health Rep
 , 
2006
, vol. 
17
 (pg. 
27
-
42
)
2.
Ogden
CL
Carroll
MD
Curtin
LR
, et al.  . 
Prevalence of overweight and obesity in the United States, 1999–2004
JAMA
 , 
2006
, vol. 
295
 (pg. 
1549
-
55
)
3.
U.S. Department of Health and Human Services
The Surgeon General's Call to Action to Prevent and Decrease Overweight and Obesity
 , 
2001
Rockville, MD
USDHHS, Public Health Services; Office of the Surgeon General
4.
American Dietetic Association
Position of the American Dietetic Association: individual-, family-, school-, and community-based interventions for pediatric overweight
J Am Diet Assoc
 , 
2006
, vol. 
106
 (pg. 
925
-
45
)
5.
Dietz
WH
Health consequences of obesity in youth: childhood predictors of adult disease
Pediatrics
 , 
1998
, vol. 
101
 (pg. 
518
-
25
)
6.
Colditz
G
DeJong
W
Hunter
D
, et al.  . 
Harvard report on cancer prevention. Volume 1: causes of human cancer
Cancer Causes Control
 , 
1996
, vol. 
7
 (pg. 
S3
-
58
)
7.
Serdula
MK
Ivery
D
Coates
RJ
, et al.  . 
Do obese children become obese adults? A review of the literature
Prev Med
 , 
1993
, vol. 
22
 (pg. 
167
-
77
)
8.
Anderson
PM
Butcher
KE
Childhood obesity: trends and potential causes
Future Child
 , 
2006
, vol. 
16
 (pg. 
19
-
45
)
9.
Marti
A
Moreno-Aliaga
MJ
Hebebrand
J
, et al.  . 
Genes, lifestyles and obesity
Int J Obes Relat Metab Disord
 , 
2004
, vol. 
28
 (pg. 
S29
-
36
)
10.
Storey
ML
Forshee
RA
Weaver
AR
, et al.  . 
Demographic and lifestyle factors associated with body mass index among children and adolescents
Int J Food Sci Nutr
 , 
2003
, vol. 
54
 (pg. 
491
-
503
)
11.
Eisenmann
JC
Bartee
RT
Wang
MQ
Physical activity, TV viewing, and weight in U.S. youth: 1999 Youth Risk Behavior Survey
Obes Res
 , 
2002
, vol. 
10
 (pg. 
379
-
85
)
12.
Trost
SG
Kerr
LM
Ward
DS
, et al.  . 
Physical activity and determinants of physical activity in obese and non-obese children
Int J Obes
 , 
2001
, vol. 
25
 (pg. 
822
-
9
)
13.
Forshee
RA
Anderson
PA
Storey
ML
The role of beverage consumption, physical activity, sedentary behavior, and demographics on body mass index of adolescents
Int J Food Sci Nutr
 , 
2004
, vol. 
55
 (pg. 
463
-
78
)
14.
Wong
SL
Leatherdale
ST
Association between sedentary behaviour, physical activity, and obesity: inactivity among active kids
Prev Chronic Dis
 , 
2009
, vol. 
6
 (pg. 
1
-
13
)
15.
Beets
MW
Pitetti
KH
Contribution of physical education and sport to health-related fitness in high school students
J School Health
 , 
2005
, vol. 
75
 (pg. 
25
-
30
)
16.
O'Malley
PM
Johnston
LD
Delva
J
, et al.  . 
Variation in obesity among American secondary school students by school and school characteristics
Am J Prev Med
 , 
2007
, vol. 
33
 (pg. 
S187
-
94
)
17.
Singh
AS
Paw
MJMCA
Brug
J
, et al.  . 
Short-term effects of school-based weight gain prevention among adolescents
Arch Pediatr Adolesc Med
 , 
2007
, vol. 
161
 (pg. 
565
-
71
)
18.
Coleman
KJ
Tiller
CL
Sanchez
J
, et al.  . 
Prevention of the epidemic increase in child risk of overweight in low-income schools
Arch Pediatr Adolesc Med
 , 
2005
, vol. 
159
 (pg. 
217
-
24
)
19.
Veuglers
PJ
Fitzgerald
AL
Prevalence of and risk factors for childhood overweight and obesity
CMAJ
 , 
2005
, vol. 
13
 (pg. 
607
-
13
)
20.
Veuglers
PJ
Fitzgerald
AL
Effectiveness of school programs in preventing childhood obesity: a multilevel comparison
Am J Pub Health
 , 
2005
, vol. 
95
 (pg. 
432
-
5
)
21.
Leatherdale
ST
Papadakis
S
A multi-level examination of the association between older social models in the school environment and overweight and obesity among younger students
J Youth Adolescence
 , 
2010
 
doi: 10.1007/s10964-009-9491-z, epub ahead of print, 16 December 2009
22.
Wong
SL
Leatherdale
ST
Manske
S
Reliability and validity of a school-based physical activity questionnaire
Med Sci Sports Exerc
 , 
2006
, vol. 
38
 (pg. 
1593
-
600
)
23.
Leatherdale
ST
Manske
S
Faulkner
G
, et al.  . 
A multi-level examination of school programs, policies and resources associated with physical activity among elementary school youth in the PLAY-ON study
Int J Behav Nutr Phys Act
 , 
2010
, vol. 
7
 pg. 
6
 
24.
Leatherdale
ST
Manske
S
Wong
S
, et al.  . 
Integrating research, policy and practice in school-based physical activity prevention programming: the School Health Action, Planning and Evaluation System (SHAPES) Physical Activity Module
Health Promot Pract
 , 
2009
, vol. 
10
 (pg. 
254
-
61
)
25.
Joint Consortium of School Health
Healthy School Planner
  
26.
Ontario Ministry of Education
Foundations for a Healthy School
 , 
2009
 
27.
Himes
JH
Faricy
A
Validity and reliability of self-reported stature and weight of US adolescents
Am J Hum Biol
 , 
2001
, vol. 
13
 (pg. 
255
-
60
)
28.
Tiggemann
M
Nonreporting of body mass index: a research note on the interpretation of missing data
Int J Eat Disord
 , 
2006
, vol. 
39
 (pg. 
346
-
9
)
29.
Cole
TJ
Bellizzi
MC
Flegal
KM
, et al.  . 
Establishing a standard definition for child overweight and obesity worldwide: international survey
BMJ
 , 
2000
, vol. 
320
 (pg. 
1240
-
3
)
30.
Centers for Disease Control and Prevention
BMI—Body Mass Index: about BMI for Children and Teens
 , 
2006
 
31.
McMurray
RG
Ring
KB
Treuth
MS
, et al.  . 
Comparison of two approaches to structured physical activity surveys for adolescents
Med Sci Sports Exerc
 , 
2004
, vol. 
36
 (pg. 
2135
-
43
)
32.
Leatherdale
ST
Wong
SL
Modifiable characteristics associated with sedentary behaviours among youth
Int J Pediatr Obes
 , 
2008
, vol. 
3
 (pg. 
93
-
101
)
33.
Joint Consortium on School Health
JCSH Healthy School Planner
  
34.
Robertson-Wilson
JE
Leatherdale
ST
Wong
SL
Social-ecological correlates of active commuting to school among high school students
J Adolesc Health
 , 
2008
, vol. 
42
 (pg. 
486
-
95
)
35.
Murnaghan
DA
Sihvonen
M
Leatherdale
ST
, et al.  . 
The relationship between school-based smoking policies and prevention programs on smoking behavior among grade 12 students in Prince Edward Island: a multilevel analysis
Prev Med
 , 
2007
, vol. 
44
 (pg. 
317
-
22
)
36.
Snijders
T
Bosker
B
Multilevel Analysis: An Introduction to Basic and Advanced Multilevel Modeling
 , 
1999
London
Sage
pg. 
p. 224
 
37.
Arbour-Nicitopoulous
KP
Faulkner
GE
Leatherdale
ST
Learning from missing data: interpreting missing body mass index (BMI) values in young children
 
3rd International Congress on Physical Activity and Public Health, Toronto, ON, Canada, May 06, 2010
38.
Rasbash
J
Browne
W
Healy
M
, et al.  . 
MLwiN Version 2.02
 , 
2005
Bristol, UK
Institute of Education, Multilevel Models Project
39.
Story
M
Nanney
MS
Schwartz
MB
Schools and obesity prevention: creating school environments and policies to promote healthy eating and physical activity
Millbank Quart
 , 
2009
, vol. 
87
 (pg. 
71
-
100
)
40.
Janssen
I
Katzmarzyk
PT
Boyce
WF
, et al.  . 
Comparison of overweight and obesity prevalence in school-aged youth from 34 countries and their relationships with physical activity and dietary patterns
Obes Rev
 , 
2005
, vol. 
6
 (pg. 
123
-
32
)
41.
Rose
G
The Strategy of Preventive Medicine
 , 
1992
Oxford
Oxford University Press
42.
Rye
JA
O'Hara Tompkins
N
Eck
R
, et al.  . 
Promoting youth physical activity and healthy weight through schools
W V Med J
 , 
2008
, vol. 
104
 (pg. 
12
-
5
)
43.
Dwyer
JJ
Allison
KR
LeMoine
KN
, et al.  . 
A provincial study of opportunities for school-based physical activity in secondary schools
J Adolesc Health
 , 
2006
, vol. 
39
 (pg. 
80
-
6
)
44.
Powers
HS
Conway
TL
McKenzie
TL
, et al.  . 
Participation in extracurricular physical activity programs at middle schools
Res Q Exerc Sport
 , 
2002
, vol. 
73
 (pg. 
187
-
92
)
45.
Slater
SJ
Powell
LM
Chaloupka
FJ
Missed opportunities: local health departments as providers of obesity prevention programs for adolescents
Am J Prev Med
 , 
2007
, vol. 
33
 (pg. 
s246
-
50
)

Comments

0 Comments