UBC Faculty Research and Publications

Associations between the school food environment, student consumption and body mass index of Canadian… Mâsse, Louise C; de Niet-Fitzgerald, Judith E; Watts, Allison W; Naylor, Patti-Jean; Saewyc, Elizabeth M Mar 26, 2014

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


52383-12966_2013_Article_872.pdf [ 208.02kB ]
JSON: 52383-1.0132615.json
JSON-LD: 52383-1.0132615-ld.json
RDF/XML (Pretty): 52383-1.0132615-rdf.xml
RDF/JSON: 52383-1.0132615-rdf.json
Turtle: 52383-1.0132615-turtle.txt
N-Triples: 52383-1.0132615-rdf-ntriples.txt
Original Record: 52383-1.0132615-source.json
Full Text

Full Text

RESEARCH Open AccessAssociations between the school foodenvironment, student consumption and bodymass index of Canadian adolescentsLouise C Mâsse1,4*, Judith Evelyn de Niet-Fitzgerald1, Allison W Watts1, Patti-Jean Naylor2 and Elizabeth M Saewyc3AbstractBackground: Increasing attention has been paid to the school food environment as a strategy to reducechildhood obesity. The purpose of this study was to examine associations between the school food environment,students’ dietary intake, and obesity in British Columbia (BC), Canada.Methods: In 2007/08, principal responses about the school environment (N = 174) were linked to grades 7-12students (N = 11,385) from corresponding schools, who participated in the BC Adolescent Health Survey. Hierarchicalmixed-effect regression analyses examined the association between the school food environment and student’s intakeof sugar-sweetened beverages (SSBs), food consumption, and body mass index. Analyses controlled for school setting,neighborhood education level and student’s age and sex.Results: School availability of SSBs was positively associated with moderate (Odds Ratio (OR) = 1.15, 95% ConfidenceInterval (CI) = 1.02-1.30) and high (OR = 1.43, 95% CI = 1.13-1.80) SSB intake as were less healthful school nutritionguidelines for moderate SSB consumers only (OR = 0.65, 95% CI = 0.48-0.88). Availability of SSBs at school and itsconsumption were positively associated with student obesity (OR = 1.50, 95% CI = 1.12-2.01 and OR = 1.66, 95%CI = 1.19-2.34, respectively) but not with overweight. In contrast, consumption of less healthful food waspositively associated with overweight (OR = 1.03, 95% CI = 1.01-1.06).Conclusions: The results of this study provide further evidence to support the important role of schools in shapingadolescents’ dietary habits. Availability and consumption of SSBs, but not less healthful foods, at school were associatedwith higher adolescent obesity highlighting that other environments also contribute to adolescent obesity.Keywords: School food environment, Diet, Body mass index, School policy, Sugar-sweetened beverages, AdolescentsBackgroundSimilar to other developed countries, the prevalence ofmeasured overweight and obesity among Canadian youthaged 6 to 17 years in 2010 was 19.5% and 11.6%, respect-ively [1]. Over consumption of empty calories, defined ascalories originating from solid fat and added sugar, is seenas an important contributor to childhood obesity [2]. Theconsumption of empty calories accounts for about 40% ofthe total calories consumed by US children (2-18 year) ofwhich 22% are from sugar-sweetened beverages (SSBs) [3].Similarly, intake of all sugar (natural and added) accountsfor 25% of total calories of Canadian adolescents of which36%-44% are from added sugars (predominantly fromSSBs) [4]. Thus public health strategies that promote andenable healthy eating are seen as important investmentsto address childhood obesity [2,5].The school food environment is often targeted as chil-dren consume roughly 35-47% of their daily dietary intakewhile at school [6] and schools reach most children acrossvarious cultural and socio-demographic backgrounds [7].In the US, research has shown that students are exposedto a wide variety of less healthful food and beverages whileat school [8-10] and are consuming high amounts of lesshealthful food while at school; including SSBs and energydense food (pizza, french fries, chips and candies) [6,8,11].* Correspondence: lmasse@cfri.ubc.ca1School of Population and Public Health, University of British Columbia,F508-4480 Oak Street, Vancouver, British Columbia V6H 3V4, Canada4Department of Pediatrics/School of Population and Public Health, Universityof British Columbia, F508-4480 Oak Street, Vancouver, BC V6H 3V4, CanadaFull list of author information is available at the end of the article© 2014 Mâsse et al.; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the CreativeCommons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, andreproduction in any medium, provided the original work is properly credited. The Creative Commons Public DomainDedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article,unless otherwise stated.Mâsse et al. International Journal of Behavioral Nutrition and Physical Activity 2014, 11:29http://www.ijbnpa.org/content/11/1/29In Canada, there is limited data describing the school foodenvironment, but a recent study from British Columbia(BC) reported that “junk” foods (e.g., pop, cookies, chips,candies) were widely available in middle and high schoolsthrough vending machines, cafeterias, tuck shops, andschool fundraisers [12]. Furthermore, studies have foundthat the availability of particular food or beverages atschool is associated with consumption of those sameitems [9,13]. These findings suggest that improvementsto the school food environment may enable students tomake healthier food choices and lower their body massindex (BMI).Several studies have reported that school policies, prac-tices, and nutritional capacity and resources restrict theavailability of less healthful food and beverages at school[14-17] and improve student dietary intake (e.g., increasedfruit consumption and decreased intake of low-nutrientenergy dense food) [18,19]. In addition, a large-scale longi-tudinal evaluation of nutrition policies in US schoolsfound that middle and high school students in states withstronger school nutrition policies gained fewer BMI unitsand were less likely to remain overweight or obese overtime, based on measured BMI [20]. In contrast, otherstudies have found no relationship between school nutri-tion policies, practices, or resources with availability ordietary intake, including inconsistent findings for elemen-tary vs. middle/high schools [9,10,16]. Despite mixed re-sults, targeting the school food environment appearspromising for addressing obesity globally. As policies areincreasingly being used to improve the school food envir-onment, there is a need to gain a clearer understanding ofthe role of the school food environment on student eatingbehaviors and BMI.This study examined the extent to which the school foodenvironment of grades 7 to 12 students in BC, Canada wasassociated with consumption of SSBs, specific food itemsand BMI. It was hypothesized that schools with morehealthful nutritional environments (e.g. stronger policies,more restriction of unhealthy foods) would have studentswho consumed fewer SSBs and more healthful food itemsand have lower BMIs.MethodsData sourcesIn the 2007/08 school year, students in grades 7 to 12 com-pleted the BC Adolescent Health Survey (AHS) adminis-tered by the McCreary Centre Society every 5th year tomonitor the health of BC youth [21]. Of the 283,120 eligiblestudents (eliminating students in non-participating districts(9 out 59) and non-public schools), a random sample of44,104 students in 463 schools and 1760 classrooms strati-fied by grades and classes were recruited. In total, 29,315students completed the survey after incomplete and un-usable data were eliminated (66% response rate). Thesampling frame ensured a representative sample of BCpublic school students from grades 7 to 12 (for furtherdetails see [21]).During the same year, public school principals fromelementary, middle, and high schools in BC, Canadawere invited to complete a nutritional and physical activityschool environment survey. In total, 43 of the 59 schooldistricts (73% response rate) provided approval for thestudy; however, three districts were excluded as they partic-ipated in another study conducted by our team. Amongschools with students in grades 7 or higher, the school en-vironment survey was completed by 380 principals (48% re-sponse rate). For this paper, school and student level datawere linked resulting in an analytic sample of 174 schools(67 middle schools, 105 high schools, and two kindergartento grade 12 schools from 36 districts) and 11,385 students.ProceduresAll data collection procedures received ethics approvalfrom the University of British Columbia and University ofVictoria Research Ethics Board and from school districts.School data collectionIn January of 2008, school principals were invited tocomplete the school environment survey (with a pre-paid return envelope). To increase participation, a sec-ond mailing and a reminder email with an online link tothe consent form and survey were sent. Principals pri-marily filled out the 30-minute survey but were encour-aged to seek the expertise of their nutrition staff toaccurately answer sections of the survey. Principals re-ceived a nominal incentive ($10 CDN gift card).Student data collectionFrom February to June 2008, students completed the AHSsurvey. In about half of school districts, parental consentwas required, and students in selected classrooms receiveda consent form to bring home; only students who returnedthe signed consent form were allowed to complete the sur-vey. In other districts, parental notification with studentconsent was required, and students within selected class-rooms received a notification letter for parents and one forthemselves, inviting them to complete the AHS. To pre-serve anonymity, students were not asked to sign consentforms; instead, completion of the survey indicated theirconsent or assent. The 45-minute survey was completedduring classroom time, for further details see [21].MeasuresSchool nutrition environment surveyThe school environment survey integrated five constructsfrom the Theories of Organizational Changes and Stillman’sTobacco Policy Framework [22] (adapted for obesityprevention) to measure the school food environment.Mâsse et al. International Journal of Behavioral Nutrition and Physical Activity 2014, 11:29 Page 2 of 9http://www.ijbnpa.org/content/11/1/29The psychometric properties of the survey have previouslybeen published [16]. The survey included assessmentof: [1] District policy institutionalization – a 3-itemscale assessing perceived strength of district guidelineswith respect to the types of food and beverages madeavailable at school and requirements for nutritionalstaff and education (α = .79) [16]. Response optionswere above average, average, and below average; [2]School food guidelines – a 7-item scale assessingwhether the school had guidelines for; advertising offood/beverage, rewarding with food, subsidizinghealthier food/beverage items and requirements forstaff and student education (α = .64) [16]. All responseoptions were dichotomized as “yes”/“no” for the ana-lyses; [3] Nutritional resources – a 5-item scale mod-eled after Hoy’s school organization inventory [23]assessed perceived adequacy of nutritional resourcescompared to other schools in terms of food servicestaff, food service facilities, access to nutritional ex-pertise, catering options, and food/beverage offeringsat school (α = .72) [16]. Response options were aboveaverage, average, and below average; [4] Program par-ticipation – the answers to two “yes”/“no” items asses-sing participation in the BC Milk Program or the BCSchool Fruit and Vegetable Nutrition Program werecombined to create a 1-item index with participationdenoted as none, 1 program, and 2 programs; [16] and[5] Internal and external support - a 7-item scalemeasuring perceived support from the school com-munity (parents, staff, and students) for enacting stric-ter nutritional guidelines and whether they themselvesperceived schools to play a role in obesity prevention(α = .72) [16]. Response options were strongly agree, agree,disagree, and strongly disagree.Availability of SSBs and food at school was measuredwith the School Health Policies and Programs Study(SHPPS) questions that assessed food availability inschools (fruit; vegetables; cookies, crackers, cakes, pas-tries not low in fat; chocolate candy; pizza, ham-burgers, or hot dogs; French fried potatoes; saltysnacks not low in fat such as regular potato chips andcheese puffs) [24]. We computed a Food AvailabilityIndex using the Rideout scoring approach [12] as ittakes into account the proportion of healthful to lesshealthful food items offered at school. The index com-bined the seven items into 9 possible groups denotingthe extent to which more or less healthful food itemsare available at school. The index regroups various pat-terns of availability into specific groups, where 1 de-notes that only less healthful food items are offered atschool, (i.e., sweet/salty snacks and baked goods thatare not low in fat; pizza, hamburgers, or hot dogs; andFrench fried potatoes) and 9 that only healthful fooditems are offered (i.e., fruit and vegetables).Student consumption of SSBs and foodIn the AHS, student consumption of SSBs and specificfood items was measured with 5 items that asked aboutthe food or beverages consumed yesterday from the timethey got up until they went to bed. Student consumptionof SSBs was measured by the item that asked whetherthey drank “pop/soda” yesterday. Student consumptionof food was assessed by four items asking whether theyate marker foods fruit; green salad or vegetables; cookies,cake, donuts, and chocolate bars; and pizza, hot dogs, po-tato chips, and French fries. Response options to theseitems were “no”, “yes, once” and “yes, twice or more” andwere combined to compute a Food Consumption Indexutilizing the same scoring approach as the Food Availabi-lity Index (described above), with scores ranging from 1 to9 (where 1 indicates consumption of only less healthfulfood yesterday and 9 indicates consumption of onlyhealthful food yesterday).Student BMIAs part of the AHS, students self-reported their heightand weight, with BMI unit computed as kg/m2. StudentBMI was categorized based on age and gender as follows:“underweight”, “normal weight”, “overweight” and “obese”using Cole et al. criteria [25].Census dataThe setting and area-level educational attainment of eachschool was determined by linking school postal codes withthe 2006 Canadian Census. The school setting wascomputed by regrouping Statistics Canada census areasto identify schools located in: an urban setting (com-munities with an urban core ≥50,000 and a popula-tion ≥ 100,000), a suburban setting (communities withan urban core ≥10,000, and a rural setting (all othercommunities). Educational attainment was operational-ized as the percentage of the population with a highschool diploma.Statistical analysesWe used hierarchical mixed-effects linear or logistic re-gressions to account for the nesting of students withinschools and districts. To examine associations betweenthe school food environment and student consumptionof SSBs, we conducted two hierarchical logistic regres-sion analyses comparing no consumption to consumingone pop/soda yesterday (none vs. once) and comparingno consumption to consuming at least two pops/sodasyesterday (none vs. twice+). We used linear regression toexamine the association between the school nutritionenvironment and the student Food Consumption Index.We used logistic regression for the BMI analyses com-paring first the normal weight students versus the over-weight students and second the normal weight studentsMâsse et al. International Journal of Behavioral Nutrition and Physical Activity 2014, 11:29 Page 3 of 9http://www.ijbnpa.org/content/11/1/29versus the obese students. We combined the under-weight and normal weight students because we foundno difference in the results. For these analyses, we exam-ined associations with the school nutrition environment(Model 1), with student consumption of SSBs and theFood Consumption Index (Model 2) and with both theschool nutrition environment and student consumptionof SSBs and the Food Consumption Index entered as in-dependent variables (Model 3).Measures of neighbourhood-level postsecondary edu-cation and school setting were entered as school levelcovariates, while age and sex were entered as studentlevel covariates in the analyses.Missing data (6.9% SD = 4.6%) for the independentvariables and covariates were all imputed using the Ex-pectation Maximization multiple imputation techniqueswith five replicates [26]. All analyses were conducted inSTATA version 11.2 (StataCorp, Texas, US).ResultsDescriptive information is provided in Table 1. Studentswere on average 15 years old, equally split by sex (48.1%boys) and 12.7% were categorized as overweight and3.9% obese. In total, 42.3% of students reported consum-ing a SSB in the previous day and students scored 5.7 onthe Food Consumption Index indicating they consumedsome fruits or vegetables as well as some less healthfulfoods (sweets or fast foods items) in the previous day.Overall, 62.1% of schools were in an urban setting. Withrespect to the school environment, districts were in theprocess of institutionalizing nutrition policies. In total,42% of schools reported healthy nutrition practices withTable 1 Characteristics of schools (N = 174) and grade 7-12 students (N = 11,385), British Columbia, Canada% or mean (SD) [range]Student characteristicsAge (N = 11375) 14.9 (1.8) [12.0 – 19.0]Sex (N = 11368) Male 48.1%Female 51.9%Body Mass Index (BMI) (N = 9363) Underweight 4.9%Normal weight 78.4%Overweight 12.7%Obesity 3.9%Student behaviorSugar-sweetened beverage consumption (N = 10879) No 57.7%1 yesterday 31.9%2+ yesterday 10.4%Food consumption index (N = 10735) 5.7 (2.4) [1.0 – 9.0]School socio-demographic characteristicsPostsecondary education (N =171) 32.0% (8.0) [14.61 – 56.86]School setting (N = 174 ) Urban 62.1%Suburban 16.1%Rural 21.8%Median family income (N = 171) $69,006 ($24,216) [0 – $161,725]School environmentPolicy institutionalization – District guidelines (N = 152 ) 2.1 (0.4) [1.0 – 3.0]Policy institutionalization – school nutrition practices (N = 130 ) 0.4 (0.2) [0.0 – 1.0]Capacity & resources Nutritional resources (N = 151) 1.9 (0.5) [1.0 – 3.0]Program participation (N = 144) None 55.6%1 program 28.5%2 programs 16.0%Internal and external support (N = 148) 2.7 (0.4) [1.7 – 3.7]Sugar-sweetened beverages availability (N = 174) No 56.9%Yes 43.1%Food availability index (N = 174) 4.7 (2.1) [1.0 – 9.0]Mâsse et al. International Journal of Behavioral Nutrition and Physical Activity 2014, 11:29 Page 4 of 9http://www.ijbnpa.org/content/11/1/29average nutritional resources and 55.6% did not partici-pate in any nutritional program. The school communitysupport for stricter nutritional policies was slightly aboveaverage ((2.72/4)/100 = 68%). In total, 43.1% of schoolsindicated that students had access to SSBs. Finally ascore of about 5 for the Food Availability Index indicatedthat students had access to less healthier food but littleto no access to fruit and vegetables.Association with student consumptionFour variables were significantly associated with the con-sumption of SSBs (Table 2) – percent of postsecondaryeducation surrounding the school neighborhood, sex,school guidelines, and availability of SSBs at school. Over-all, student consumption of SSBs was lower in schools lo-cated in communities with higher rates of post-secondaryeducation (Odds Ratio (OR) = 0.89; p = .006 comparing noconsumption to one and OR = 0.84; p = .048 comparingno consumption to two+). In any school environment, theodds of being a moderate or high consumer of SSBs waslower for boys than for girls (OR = 0.49; p < .001 compar-ing no consumption to one and OR = 0.28; p < .001 com-paring no consumption to two+). In addition, the odds ofa student being a moderate consumer of SSBs was 0.65times lower (p = .006) in schools that had healthier nu-trition guidelines than those without; however therewere no effects for high consumers. Finally, the odds ofmoderate or high consumption of SSBs were higher inschools that reported having SSBs available than inthose that did not (OR = 1.15; p = .022 comparing noconsumption to one and OR = 1.43; p = .003 comparingno consumption to two+).Overall, postsecondary education and sex were theonly factors associated with the Food ConsumptionIndex (Table 2). Students attending a school locatedin a community with higher rates of post-secondaryeducation reported consuming more healthful food(b = 0.12, p = .04). In addition, girls had a healthierFood Consumption Index score than boys (b = 0.51,p < .001).Association with student BMIThe findings that examined associations with studentBMI are shown in Table 3 for comparisons between nor-mal versus overweight and in Table 4 for comparisonsbetween normal versus obese.Table 2 School factors associated with grade 7-12 students’ sugar-sweetened beverage consumption (N = 10879) andfood consumption index (N = 10735)Sugar-sweetened beverage Food consumptionindex (n = 10496)None versus one(n = 9518)None versus2+ (n = 7247)OR [95% CI], p-value OR [95% CI], p-value b [95% CI], p-valueConstant - - 4.48 [3.59; 5.37], p < .001CovariatesSchool postsecondary education 0.89 [0.81–0.96], p = .01 0.84 [0.71–0.99], p = .048 0.12 [0.01; 0.23], p = .04School setting Urban (reference) 1.00 1.00 1.00Suburban 1.15 [0.91–1.41], p = .19 1.09 [0.76–1.54], p = .64 0.03 [−0.23; 0.29], p = .83Rural 1.01 [0.85–1.21], p = .89 1.11 [0.80–1.52], p = .54 0.04 [−0.21; 0.28], p = .77Age 0.98 [0.95–1.01], p = .17 1.01 [0.96–1.05], p = .77 0.02 [−0.02; 0.05], p = .33Sex Male (reference) 1.00 1.00 1.00Female 0.49 [0.45–0.54], p < .001 0.28 [0.24–0.33], p < .001 0.51 [0.42; 0.60], p < .001School environmentPolicy institutionalization – district guidelines 1.08 [0.90–1.28], p = .44 0.98 [0.70–1.35], p = .88 −0.09 [−0.31; 0.14], p = .44Policy institutionalization – school nutrition practices 0.65 [0.48–0.88], p = .01 0.69 [0.41–1.15], p = .16 0.23 [−0.12; 0.58], p = .19Capacity & resources Nutritional resources 1.01 [0.88–1.16], p = .90 0.99 [0.76–1.27], p = .91 −0.06 [−0.24; 0.12], p = .53Program participation None (reference) 1.00 1.00 1.001 program 0.96 [0.84–1.08], p = .48 0.96 [0.73–1.26], p = .77 0.01 [−0.18; 0.20], p = .932 programs 0.97 [0.76–1.23], p = .81 0.88 [0.75–1.26], p = .49 0.07 [−0.15; 0.29], p = .54Internal and external support 0.91 [0.78–1.06], p = .25 1.08 [0.78–1.48], p = .65 −0.03 [−0.24; 0.19], p = .80Sugar-sweetened beverages availability No (reference) 1.00 1.00 NAYes 1.15 [1.02–1.30], p = .02 1.43 [1.13–1.80], p = .003 NAFood availability index NA NA 0.02 [−0.02; 0.05], p = .41OR = Odds ratio; CI = Confidence Interval; b = non-standardized parameter estimate.Mâsse et al. International Journal of Behavioral Nutrition and Physical Activity 2014, 11:29 Page 5 of 9http://www.ijbnpa.org/content/11/1/29Student and school demographic factors and studentconsumption, but not school environment, were signifi-cantly associated with the odds of a student being over-weight versus normal weight (model 3, Table 3). Studentsfrom schools in suburban and rural settings had higherodds of being overweight than those in an urban setting(OR = 1.37, p = .011 and OR = 1.41, p = .003, respectively).In addition, girls had lower odds of being overweight thanboys (OR = 0.46, p < .001) and students who reported con-suming less healthful foods had higher odds of being over-weight (OR = 1.03, p = .020).Results comparing obese versus normal weight students(Model 3, Table 4) indicate that the school setting, sex,availability of SSBs at school, and student consumption ofSSBs were significantly associated with the odds of a stu-dent being obese versus normal weight. Overall, studentsin suburban and rural schools had greater odds of beingobese than those who attended a school in an urban set-ting (OR = 1.52, p = .035 and OR = 1.89, p < .001, res-pectively). Girls had lower odds of being obese than boys(OR = 0.51, p = <.001). Notably, students had greater oddsof being obese than normal weight in schools where SSBswere readily available (OR = 1.50, p = .007) and if they re-ported consuming more than one SSB in the previous day(OR = 1.66, p = .003).DiscussionThis study comprehensively examined associations be-tween the school food environment and student con-sumption, and in turn, associations with BMI. We foundthat the availability of SSBs and nutritional practices atschool were associated with consumption of SSBs, butno associations with consumption of other less healthfulfoods. We also found that both the availability of SSBsand less healthful foods were associated with studentBMI; although these associations differed by BMI category(overweight versus obese). Specifically, the association withSSBs was only observed among the obese adolescents,whereas an association with less healthful foods was onlyobserved among the overweight adolescents. Our findingsTable 3 School factors and grade 7-12 students’ food/beverage consumption associated with Body Mass Index (normalversus overweight) (N = 8995)Model 1 (n = 8834) Model 2 (n = 8361) Model 3 (n = 8209)OR [95% CI], p-value OR [95% CI], p-value OR [95% CI], p-valueCovariatesSchool postsecondary education 0.90 [0.79–1.02], p = .09 0.92 [0.80–1.04], p = .19 0.90 [0.79–1.03], p = .11School setting Urban (reference) 1.00 1.00 1.00Suburban 1.41 [1.13–1.77], p = .003 1.45 [1.13–1.86], p = .003 1.37 [1.07–1.73], p = .01Rural 1.41 [1.14–1.73], p = .002 1.44 [1.14–1.82], p = .003 1.41 [1.13–1.77], p = .003Age 1.03 [0.99–1.07], p = .17 1.04 [0.99–1.07], p = .09 1.04 [1.00–1.08], p = .05Sex Male (reference) 1.00 1.00 1.00Female 0.47 [0.41–0.53], p < .001 0.47 [0.41–0.53], p < .001 0.46 [0.41–0.53], p < .001School environmentPolicy institutionalization – district guidelines 0.92 [0.73–1.15], p = .45 - 0.92 [0.73–1.16], p = .49Policy institutionalization – school nutrition practices 1.16 [0.79–1.67], p = .46 - 1.23 [0.84–1.80], p = .30Capacity and resources Nutritional resources 0.85 [0.72–1.02], p = .08 - 0.85 [0.70–1.02], p = .08Program participation None (reference) 1.00 - 1.001 program 1.03 [0.83–1.28], p = .78 - 1.02 [0.82–1.27], p = .872 programs 1.10 [0.83–1.46], p = .49 - 1.14 [0.84–1.54], p = .37Internal and external support 0.90 [0.73–1.12], p = .35 - 0.88 [0.71–1.11], p = .28Sugar-sweetened beverages availability No (reference) 1.00 - 1.00Yes 1.16 [0.97–1.39], p = .10 - 1.13 [0.94–1.36], p = .20Food availability index 1.00 [0.96–1.04], p = .99 - 1.00 [0.96–1.04], p = .86Student consumptionSugar-sweetened beverage consumption None (reference) - 1.00 1.001 yesterday - 1.13 [0.98–1.31], p = .10 1.13 [0.97–1.31], p = .122+ yesterday - 1.12 [0.90–1.39], p = .32 1.13 [0.90–1.40], p = .29Food consumption index - 1.03 [1.00–1.06], p = .02 1.03 [1.01–1.06], p = .02OR = Odds Ratio; CI = Confidence Interval.Mâsse et al. International Journal of Behavioral Nutrition and Physical Activity 2014, 11:29 Page 6 of 9http://www.ijbnpa.org/content/11/1/29add to the limited and inconsistent findings in this area[27] and provide further support for improving the schoolfood environment to reduce childhood obesity.Similar to previous studies [18,19,28], the availabilityof SSBs and less healthful nutritional practices at schoolwere both associated with greater SSB consumptionhighlighting the importance of schools in promotinghealthy dietary habits. In contrast, Taber et al. reportedthat reduced access to SSBs at school reduced purchas-ing but not overall consumption of SSB [29]. Unlike pre-vious studies [18,19,28,30], associations in this studywere observed for SSB consumption only and no asso-ciations were found between the other school food envir-onment variables (policies, programs, resources, support,availability) and the Food Consumption Index. Our fin-dings may indicate a need for attention to specificity infood environment measures or reflect limitations in howthe Food Consumption and Food Availability Indices weremeasured.Access to SSBs at school and their consumption wereboth associated with obesity providing further supportfor targeting schools to help address adolescent obesity[20,31]. While the association between SSB consumptionand BMI is supported by a recent review [32], the asso-ciation was present for obese but not overweight adoles-cents in our study. It is possible that access to SSBs inthe school setting may disproportionately affect studentswho come from a less healthy home environment asthey likely consume SSBs both at home and school.Interestingly, we found an association with our FoodConsumption Index and weight, but only for overweightadolescents compared to normal weight adolescents.This association was somewhat expected as a review byPerez-Escamilla [33] supports an association betweenenergy density and BMI in children and adolescents;however, it is somewhat inconsistent that we observedthis relationship only among those who were overweightand that consumption of SSBs appeared most related toTable 4 School factors and grade 7-12 students’ beverage/food consumption associated with Body Mass Index(normal versus obese) (N = 8172)Model 1 (n = 8018) Model 2 (n = 7604) Model 3 (n = 7458)OR [95% CI], p-value OR [95% CI], p-value OR [95% CI], p-valueCovariatesSchool postsecondary education 1.00 [0.83–1.20], p = .98 1.00 [0.81–1.23], p = .98 0.98 [0.80–1.20], p = .84School setting Urban (reference) 1.00 1.00 1.00Suburban 1.53 [1.06–2.23], p = .02 1.67 [1.15–2.44], p = .01 1.52 [1.35–2.23], p = .04Rural 1.94 [1.40–1.11], p < .001 2.01 [1.39–2.89], p < .001 1.89 [1.03–2.66], p < .001Age 1.06 [0.99–1.14], p = .09 1.07 [1.00–1.14], p = .06 1.06 [0.99–1.14], p = .10Sex Male (reference) 1.00 1.00 1.00Female 0.47[0.38–0.59], p < .001 0.51[0.40–0.64], p < .001 0.51[0.40–0.64], p < .001School environmentPolicy institutionalization – district guidelines 0.87 [0.62–1.23], p = .44 - 0.85 [0.59–1.21], p = .36Policy institutionalization – school nutrition practices 1.23 [0.66–2.29], p = .50 - 1.15 [0.59–2.23], p = .67Capacity and resources Nutritional resources 1.03 [0.77–1.36], p = .85 - 0.97 [0.73–1.31], p = .89Program participation None (reference) 1.00 - 1.001 program 1.22 [0.88–1.68], p = .23 - 1.20 [0.86–1.67], p = .272 programs 0.89 [0.52–1.52], p = .65 - 0.99 [0.57–1.72], p = .96Internal and external support 0.78 [0.56–1.07], p = .13 - 0.74 [0.53–1.04], p = .08Sugar-sweetened beverages availability No (reference) 1.00 - 1.00Yes 1.58 [1.20–2.10], p = .001 - 1.50 [1.12–2.01], p = .01Food availability index 1.03 [0.97–1.09], p = .32 - 1.03 [0.97–1.11], p = .29Student consumptionSugar-sweetened beverage consumption None (reference) - 1.00 1.001 yesterday - 1.26 [0.98–1.62], p = .07 1.19 [0.92–1.54], p = .182+ yesterday - 1.69 [1.20–2.36], p = .003 1.66 [1.19–2.34], p = .003Food consumption index - 1.04 [0.99–1.08], p = .15 1.03 [0.98–1.08], p = .27OR = Odds ratio; CI = Confidence Interval.Mâsse et al. International Journal of Behavioral Nutrition and Physical Activity 2014, 11:29 Page 7 of 9http://www.ijbnpa.org/content/11/1/29BMI among obese adolescents. This difference may re-late to the limitations of our methodology and is dis-cussed further below.With respect to the covariates we included in ourmodel, similar to other studies boys were more likely toconsume SSBs [34,35], consume energy dense foods [36],and to be overweight and obese than girls [1,35]. Boys arethus a vulnerable group that may benefit from changes tothe school food environment to a greater extent than girlsmight. Similar to other studies, we found that adolescentsfrom more disadvantaged neighborhoods also consumedmore SSBs [34,35] and energy dense food [6] and that stu-dents in suburban and rural schools were more likely tobe overweight and obese [37,38]. These findings highlightgroups of adolescents who may be differentially affectedby creating healthier eating environments at school.Our findings should be interpreted in light of the con-text in which the data were collected. Unlike the US,Canada does not have a national breakfast or school lunchprogram that is subsidized by the federal government [39].In BC, while subsidies for school lunch targeting studentsin need can be obtained from the provincial government,guidelines to regulate the school food environment werefirst written in 2005, but full implementation was only ex-pected in 2008, after the data were collected for this paper.Interestingly, even in such a varied context and with aguideline in place (albeit without an accountability mech-anism) many of our findings echo what others have foundin the US [28,30]. Given that the school food environmentof Canadian schools was quite different than that of USschools, it remains important to understand whether re-search from the US context translates into other jurisdic-tional contexts.Finally, the study limitations are important to considerwhen interpreting our results. First, the cross-sectional na-ture of the data limits our ability to make causal infer-ences. Second, we used self-report to measure studentconsumption and BMI, and although commonly used inlarge studies, are known to be associated with measure-ment errors that can mask or dampen existing associa-tions. Third, many principals did not complete the surveyand half the schools required written parental consentresulting in lower student participation; therefore, we donot know how non-response bias may have affected theresults. Fourth, although we evaluated the psychometricproperties of our school food environment measures, theywere developed or adapted from other measures to fit theBC context. Fifth, we utilized an established measure forthe availability of food and beverages at school; [24] how-ever, the measure did not identify if healthier versions ofspecific food were offered. An unpublished governmentreview suggests that little change in the school food envi-ronment had occurred before the full implementation ofthe first food guidelines were expected in schools whichwas after we collected the data. Sixth, we highlighted anyeffects that were significant at a p < .05 but some effectsmight be less stable as they were not significant at ap < .01 as well (i.e., the association between less healthfulfoods and overweight adolescents). Given the exploratorynature of our analyses, all of our findings should be repli-cated. Seventh, each province and territory in Canada hasdifferent policies/guidelines affecting the food environ-ment of public schools, and without a federally subsidizedschool breakfast/lunch program, the generalizability ofour findings to other jurisdictions is limited. Finally, meas-uring consumption over the entire day limits our ability todetermine associations with school-specific consumption.This is important as recent data among US childrenshowed a shift in the amount of energy intake obtainedfrom school sources to fast food places [40]. A better un-derstanding of where students’ food purchases occur mayshed further light on these findings.ConclusionsThe results of this study provide further evidence tosupport the important role of schools in shaping adoles-cents’ dietary habits. Availability of SSBs at school in-creased students’ odds of consuming SSBs and beingobese and availability of less healthful foods was associ-ated with higher consumption. Creating school environ-ments that are more conducive to healthy eating andimplementing a comprehensive approach that includesall of the environments in which adolescents spend theirtime will likely provide the greatest benefit in supportinghealthy food choices and healthy weights.Competing interestsThe authors declare that they have no competing interests.Authors’ contributionsLCM contributed to the conception and design of the study, interpretation offindings, and writing/reviewing the manuscript. JEDNF contributed to the dataanalysis, interpretation of the findings and writing/reviewing the manuscript.AWW contributed to the interpretation of findings and writing/reviewing themanuscript. PJN contributed to the conception and design of the study,interpretation of findings and reviewing the manuscript. EMS contributed tothe conception and design of the study, interpretation of findings andreviewing the manuscript. All authors read and approved the final manuscript.AcknowledgementsThe authors acknowledge McCreary Centre Society for permission to accessto the BC Adolescent Health Survey data. LCM was funded through salarysupport provided by the Child and Family Research Institute (CFRI) located atthe Children’s & Women’s Health Centre of British Columbia (C&W). Thisstudy was funded by the Canadian Institutes of Health Research (CIHR)Institute of Nutrition, Metabolism and the Institute of Human Development,Child and Youth Health 200905GIR-206392-GIR-CAAA-143786. Establishmentfunds from the Michael Smith Foundation for Health Research were used tocollect some of the data. JEDNF received post-doctoral salary support fromthe CFRI at C&W and from the Heart and Stroke Foundation of Canada.AWW is funded through a CIHR Doctoral Research Award in partnership withthe Danone Institute of Canada and through a CIHR fellowship in populationinterventions for chronic disease prevention in partnership with the Heartand Stroke Foundation of Canada. EMS was funded through an AppliedPublic Health Chair, CIHR and Public Health Agency of Canada #CPP86374.Mâsse et al. International Journal of Behavioral Nutrition and Physical Activity 2014, 11:29 Page 8 of 9http://www.ijbnpa.org/content/11/1/29Author details1School of Population and Public Health, University of British Columbia,F508-4480 Oak Street, Vancouver, British Columbia V6H 3V4, Canada. 2Schoolof Exercise Science, Physical and Health Education, University of Victoria, POBox 3015 STN CSC, Victoria, British Columbia V8W 3P1, Canada. 3School ofNursing, University of British Columbia, 2211 Wesbrook Mall, Vancouver,British Columbia V6T 2B5, Canada. 4Department of Pediatrics/School ofPopulation and Public Health, University of British Columbia, F508-4480 OakStreet, Vancouver, BC V6H 3V4, Canada.Received: 1 October 2013 Accepted: 13 February 2014Published: 26 March 2014References1. Roberts KC, Shields M, de Groh M, Aziz A, Gilbert JA: Overweight andobesity in children and adolescents: results from the 2009 to 2011Canadian Health Measures Survey. Health Rep 2012, 23:37–41.2. Institute of Medicine: Accelerating Progress in Obesity Prevention: Solving theWeight of the Nation. Washington, DC: The National Academies Press; 2012.3. Reedy J, Krebs-Smith SM: Dietary sources of energy, solid fats, and addedsugars among children and adolescents in the United States. J Am DietAssoc 2010, 110:1477–1484.4. Langlois K, Garriguet D: Sugar consumption among Canadians of all ages.Health Rep 2011, 22:23–27.5. World Health Organization: Obesity and Overweight. http://www.who.int/mediacentre/factsheets/fs311/en/.6. Briefel RR, Wilson A, Gleason PM: Consumption of low-nutrient, energy-densefoods and beverages at school, home, and other locations among schoollunch participants and nonparticipants. J Am Diet Assoc 2009, 109:S79–S90.7. Story M, Nanney MS, Schwartz MB: Schools and obesity prevention:creating school environments and policies to promote healthy eatingand physical activity. Milbank Q 2009, 87:71–100.8. Fox MK, Dodd AH, Wilson A, Gleason PM: Association between schoolfood environment and practices and body mass index of US publicschool children. J Am Diet Assoc 2009, 109:S108–S117.9. Park S, Sappenfield WM, Huang Y, Sherry B, Bensyl DM: The impact of theavailability of school vending machines on eating behavior duringlunch: the Youth Physical Activity and Nutrition Survey. J Am Diet Assoc2010, 110:1532–1536.10. Rovner AJ, Nansel TR, Wang J, Iannotti RJ: Food sold in school vendingmachines is associated with overall student dietary intake. J AdolescHealth 2011, 48:13–19.11. Briefel RR, Crepinsek MK, Cabili C, Wilson A, Gleason PM: School foodenvironments and practices affect dietary behaviors of US public schoolchildren. J Am Diet Assoc 2009, 109:S91–S107.12. Rideout K, Levy-Milne R, Martin C, Ostry AS: Food sales outlets, food avail-ability, and the extent of nutrition policy implementation in schools inBritish Columbia. Can J Public Health 2007, 98:246–250.13. Templeton SB, Marlette MA, Panemangalore M: Competitive foods increasethe intake of energy and decrease the intake of certain nutrients byadolescents consuming school lunch. J Am Diet Assoc 2005, 105:215–220.14. Kubik MY, Wall M, Shen L, Nanney MS, Nelson TF, Laska MN, Story M: Statebut not district nutrition policies are associated with less junk food invending machines and school stores in US public schools. J Am DietAssoc 2010, 110:1043–1048.15. Kubik MY, Lytle LA, Farbakhsh K: School and district wellness councils andavailability of low-nutrient, energy-dense vending fare in Minnesota middleand high schools. J Am Diet Assoc 2011, 111:150–155.16. Masse LC, de Niet JE: School nutritional capacity, resources and practicesare associated with availability of food/beverage items in schools. Int JBehav Nutr Phys Act 2013, 10:26.17. Whatley Blum JE, Beaudoin CM, O’Brien LM, Polacsek M, Harris DE, O’RourdeKA: Impact of Maine’s statewide nutrition policy on high school foodenvironments. Prev Chronic Dis 2011, 8:A19.18. Taber DR, Chriqui JF, Chaloupka FJ: Differences in nutrient intakeassociated with state laws regarding fat, sugar, and caloric content ofcompetitive foods. Arch Pediatr Adolesc Med 2012, 166:452–458.19. Mullally ML, Taylor JP, Kuhle S, Bryanton J, Hernandez KJ, MacLellan DL,McKenna ML, Gray RJ, Veugelers PJ: A province-wide school nutrition policyand food consumption in elementary school children in Prince Edward Is-land. Can J Public Health 2010, 101:40–43.20. Taber DR, Chriqui JF, Perna FM, Powell LM, Chaloupka FJ: Weight statusamong adolescents in States that govern competitive food nutritioncontent. Pediatrics 2012, 130:437–444.21. Saewyc E, Green R: Survey Methodology for the 2008 BC Adolescent HealthSurvey IV. http://www.mcs.bc.ca/pdf/AHS4_Methodology_Fact_Sheet.pdf.22. Stillman F, Hartman A, Graubard B, Gilpin E, Chavis D, Garcia J, Wun LM,Lynn W, Manley M: The American Stop Smoking Intervention Study.Conceptual framework and evaluation design. Eval Rev 1999, 23:259–280.23. Hoy WK, Tarter CJ, Kottkamp RB: The organizational health inventory forelementary schools. In Open Schools/Healthy Schools: MeasuringOrganizational Climate. Newbury Park: Sage Publications; 1991:85–106.24. Centers for Chronic Disease and Prevention: Healthy Youth: SHPPS 2006 DataFiles and Documentation. http://www.cdc.gov/healthyyouth/shpps/questionnaires.htm#2006.25. Cole TJ, Bellizzi MC, Flegal KM, Dietz WH: Establishing a standard definitionfor child overweight and obesity worldwide: international survey. BMJ2000, 320:1240–1243.26. Sterne JA, White IR, Carlin JB, Spratt M, Royston P, Kenward MG, Wood AM,Carpenter JR: Multiple imputation for missing data in epidemiologicaland clinical research: potential and pitfalls. BMJ 2009, 338:b2393.27. Jaime PC, Lock K: Do school based food and nutrition policies improvediet and reduce obesity? Prev Med 2009, 48:45–53.28. Schwartz MB, Novak SA, Fiore SS: The impact of removing snacks of lownutritional value from middle schools. Health Educ Behav 2009, 36:999–1011.29. Taber DR, Chriqui JF, Powell LM, Chaloupka FJ: Banning all sugar-sweetened beverages in middle schools: reduction of in-school accessand purchasing but not overall consumption. Arch Pediatr Adolesc Med2012, 166:256–262.30. Kubik MY, Lytle LA, Hannan PJ, Perry CL, Story M: The association of theschool food environment with dietary behaviors of young adolescents.Am J Public Health 2003, 93:1168–1173.31. Levy DT, Friend KB, Wang YC: A review of the literature on policiesdirected at the youth consumption of sugar sweetened beverages.Adv Nutr 2011, 2:182S–200S.32. Te ML, Mallard S, Mann J: Dietary sugars and body weight: systematicreview and meta-analyses of randomised controlled trials and cohortstudies. BMJ 2013, 346:e7492.33. Perez-Escamilla R, Obbagy JE, Altman JM, Essery EV, McGrane MM, Wong YP,Spahn JM, Williams CL: Dietary energy density and body weight in adultsand children: a systematic review. J Acad Nutr Diet 2012, 112:671–684.34. Wang YC, Bleich SN, Gortmaker SL: Increasing caloric contribution fromsugar-sweetened beverages and 100% fruit juices among US childrenand adolescents, 1988-2004. Pediatrics 2008, 121:e1604–e1614.35. Ogden CL, Kit BK, Carroll MD, Park S: Consumption of sugar drinks in theUnited States, 2005–2008, NCHS data brief, no 71. Hyattsville, MD: NationalCenter for Health Statistics; 2011.36. Piernas C, Popkin BM: Increased portion sizes from energy-dense foods affecttotal energy intake at eating occasions in US children and adolescents:patterns and trends by age group and sociodemographic characteristics,1977-2006. Am J Clin Nutr 2011, 94:1324–1332.37. Bruner MW, Lawson J, Pickett W, Boyce W, Janssen I: Rural Canadianadolescents are more likely to be obese compared with urbanadolescents. Int J Pediatr Obes 2008, 3:205–211.38. Davis AM, Bennett KJ, Befort C, Nollen N: Obesity and related healthbehaviors among urban and rural children in the United States: Datafrom the National Health and Nutrition Examination Survey 2003-2004and 2005-2006. J Pediatr Psychol 2011, 36:669–676.39. McKenna ML: Policy options to support healthy eating in schools. Can JPublic Health 2010, 101(Suppl 2):S14–S17.40. Poti JM, Popkin BM: Trends in energy intake among US children byeating location and food source, 1977-2006. J Am Diet Assoc 2011,111:1156–1164.doi:10.1186/1479-5868-11-29Cite this article as: Mâsse et al.: Associations between the school foodenvironment, student consumption and body mass index of Canadianadolescents. International Journal of Behavioral Nutrition and PhysicalActivity 2014 11:29.Mâsse et al. International Journal of Behavioral Nutrition and Physical Activity 2014, 11:29 Page 9 of 9http://www.ijbnpa.org/content/11/1/29


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items