UBC Faculty Research and Publications

Effectiveness of active school transport interventions: a systematic review and update Larouche, Richard; Mammen, George; Rowe, David A; Faulkner, Guy Feb 1, 2018

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

Item Metadata


52383-12889_2017_Article_5005.pdf [ 766.66kB ]
JSON: 52383-1.0363394.json
JSON-LD: 52383-1.0363394-ld.json
RDF/XML (Pretty): 52383-1.0363394-rdf.xml
RDF/JSON: 52383-1.0363394-rdf.json
Turtle: 52383-1.0363394-turtle.txt
N-Triples: 52383-1.0363394-rdf-ntriples.txt
Original Record: 52383-1.0363394-source.json
Full Text

Full Text

RESEARCH ARTICLE Open AccessEffectiveness of active school transportinterventions: a systematic reviewand updateRichard Larouche1,2, George Mammen3, David A. Rowe4 and Guy Faulkner5,6*AbstractBackground: Active school transport (AST) is a promising strategy to increase children’s physical activity. Asystematic review published in 2011 found large heterogeneity in the effectiveness of interventions in increasingAST and highlighted several limitations of previous research. We provide a comprehensive update of that review.Methods: Replicating the search of the previous review, we screened the PubMed, Web of Science, Cochrane,Sport Discus and National Transportation Library databases for articles published between February 1, 2010 andOctober 15, 2016. To be eligible, studies had to focus on school-aged children and adolescents, include anintervention related to school travel, and report a measure of travel behaviors. We assessed quality of individualstudies with the Effective Public Health Practice Project quality assessment tool, and overall quality of evidence withthe Grades of Recommendation, Assessment, Development, and Evaluation (GRADE) approach. We calculatedCohen’s d as a measure of effect size.Results: Out of 6318 potentially relevant articles, 27 articles reporting 30 interventions met our inclusion criteria.Thirteen interventions resulted in an increase in AST, 8 found no changes, 4 reported inconsistent results, and 5 didnot report inferential statistics. Cohen’s d ranged from −0.61 to 0.75, with most studies reporting “trivial-to-small”positive effect sizes. Three studies reported greater increases in AST over longer follow-up periods and two SafeRoutes to School studies noted that multi-level interventions were more effective. Study quality was rated as weakfor 27/30 interventions (due notably to lack of blinding of outcome assessors, unknown psychometric properties ofmeasurement tools, and limited control for confounders), and overall quality of evidence was rated as low.Evaluations of implementation suggested that interventions were limited by insufficient follow-up duration,incomplete implementation of planned interventions, and limited access to resources for low-income communities.Conclusions: Interventions may increase AST among children; however, there was substantial heterogeneity acrossstudies and quality of evidence remains low. Future studies should include longer follow-ups, use standardizedoutcome measures (to allow for meta-analyses), and examine potential moderators and mediators of travelbehavior change to help refine current interventions.Trial registration: Registered in PROSPERO: CRD42016033252Keywords: Active travel, Physical activity, Children, Safe routes to school, School travel plans, Walking school buses* Correspondence: guy.faulkner@ubc.ca5School of Kinesiology, University of British Columbia, D H Copp Building4606, 2146 Health Sciences Mall, Vancouver, BC V6T 1Z3, Canada6Center for Hip Health and Mobility, Robert H.N. Ho Research Centre, 5thFloor, 2635 Laurel St, Vancouver, BC V5Z 1M9, CanadaFull list of author information is available at the end of the article© The Author(s). 2018 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, andreproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link tothe Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver(http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.Larouche et al. BMC Public Health  (2018) 18:206 DOI 10.1186/s12889-017-5005-1BackgroundConsistent evidence shows that children and adolescentswho engage in active school transport (AST) are morephysically active than those who travel by motorized ve-hicles [1, 2]. Cycling to and from school can also in-crease cardiovascular fitness [1] and is associated with abetter cardiometabolic health profile [3]. At the popula-tion level, replacing motorized travel by AST could re-duce exhaust and greenhouse gas emissions [4, 5].Additional benefits of AST include positive emotionsduring the school trip [6], better way-finding skills [7]and superior school grades [8].Despite these benefits, the prevalence of AST has de-creased markedly during the last few decades in manycountries [9–13]. To address this issue, many interven-tions have been implemented. Perhaps the most well-known is the Safe Routes to School (SRTS) programwhich has received over one billion dollars in fundingfrom the US government [14]. Recent analysesconcluded that New York City’s SRTS program led to a33-44% reduction in injuries among school-aged chil-dren and the program was cost-effective even whendisregarding any potential benefits related to increasedphysical activity and decreased congestion and pollution[15, 16]. In other jurisdictions, school travel plans (STP)have been implemented to address key barriers to ASTat the local level, but often with limited funding [17–21].Moreover, walking school buses (WSB) wherein childrenwalk together on a set route with adult supervision havebeen implemented in many jurisdictions to address par-ental safety concerns [22, 23].To our knowledge, Chillón and colleagues [24] pub-lished the first systematic review of the effectiveness ofAST intervention among children and adolescents in2011. While the included interventions were quite het-erogeneous, most observed small increases in AST.However, quality of evidence for all interventions wererated as “weak” based on the Effective Public HealthPractice Project (EPHPP) quality assessment tool forquantitative studies [25]. Moreover, none of the interven-tions examined the moderators and mediators of travelbehavior change. A better understanding of moderatorsand mediators would enable researchers to understandwhat works for whom and why. We provide a comprehen-sive update on the effectiveness of AST interventions inchildren and youth that have been published over the last6 years. We also aimed to review the literature on themoderators and mediators of AST interventions.MethodsSearch strategyAs our goal was to update the previous review [24], wereplicated their search strategy. Databases searchedincluded PubMed, Web of Science (SCI and SSCI),SPORTDiscus, the Cochrane library, and the NationalTransportation Library. The search terms addressed fourmain categories: school-age children (adolescen* OR childOR children OR youth OR student* OR pupil OR pupils)AND active transportation (bike OR bikers OR biking ORbicycl* OR cycle OR cycling OR cyclist* OR commute*OR commuting OR transportation OR travel*) ANDintervention (intervention* OR implement* OR evaluat*OR change OR pilot OR project OR environment* ORengineer* OR encourage* OR planning OR impact OR“walk to school” OR “safe routes to school” OR “walkingschoolbus” OR “walking school bus” OR “walking schoolbuses”) AND school. Articles published between February1, 2010 (the cut-off date of the previous review) andOctober 15, 2016 were considered eligible. Our review isregistered in PROSPERO (CRD42016033252; seehttp://www.crd.york.ac.uk/PROSPERO/display_record.asp?ID=CRD42016033252).Inclusion and exclusion criteriaTo be included in the review, studies had to: 1) have beenconducted among children and adolescents (6-18 yearolds); 2) focus on AST; 3) include an intervention; and 4)examine the effect of the intervention on a measure of ac-tive transportation or physical activity. Studies that didnot meet all of these criteria were excluded. Language wasnot an exclusion criterion. Titles and abstract of all poten-tially relevant articles were screened by GM and GF. Fulltext copies of all articles that were not excluded at thisstage of the review were then screened by GM and GF.Any discrepancy was resolved by consensus.Data extractionThe following data were extracted from each includedstudy: lead author, country, a brief description of theintervention and its methodology, the effects on AST,the moderators and mediators examined, the effects onother outcomes and the types of strategies that wereused based on the Safe Routes to School 6E model [14].The 6 E’s are: 1) education (teaching students and com-munity members about the different transportationoptions and ensuring they have the skills and know-howto be safe in traffic); 2) encouragement (using events,activities and incentives to promote AST); 3) engineer-ing (making improvements to the built environment toincrease safety); 4) enforcement (partnering with law en-forcement to address traffic and crime concerns in theneighborhoods around schools and along school routes);5) evaluation (assessing the effectiveness of the interven-tions); and 6) equity (ensuring that initiatives are benefit-ing all demographic groups). By definition, all studiesthat met our inclusion criteria have used evaluations, sothis strategy was not extracted. Data extraction was doneby RL and GM for a subsample of studies, and only byLarouche et al. BMC Public Health  (2018) 18:206 Page 2 of 18RL for the remainder. When relevant information wasmissing from included papers, we attempted to contactthe lead author and/or the senior author.Quality assessmentTo assess the methodological quality of each study, weused an adapted version of the EPHPP. This tool includes6 components: 1) selection bias; 2) study design; 3) controlfor confounders; 4) blinding of participants and studystaff; 5) validity and reliability of the data collection tools;and 6) withdrawals and drop-outs. Each component wasrated as “weak”, “moderate” or “strong” based on stan-dardized criteria, and then the number of weak ratingswas tallied. Following the EPHPP approach, studies withzero weak ratings were rated as strong, studies with oneweak rating were rated as moderate, and studies with atleast two weak ratings were rated as weak. We retainedthe modifications proposed by Chillón and colleagues [24]to make the tool more suitable to studies in which theschool is the unit of allocation. We also added a numberof precisions to clarify the interpretation of the items. Ouradapted EPHPP tool is available in Additional file 1.Quality assessment was first performed by RL and DR fora subsample of five studies. After consensus was attainedfor these studies, the remaining articles were assessedeither by RL or DR. In case of doubt, the reviewer wasasked to indicate the issue in an Excel spreadsheet and allissues were resolved by consensus among the two re-viewers. Because blinding of participants was consideredunfeasible in the context of most AST interventions, wepresent results both with and without the blinding com-ponent of the EPHPP. In addition, we assessed overallquality of evidence using the “Grades of Recommendation,Assessment, Development, and Evaluation” (GRADE)approach [26, 27]. Following this approach, randomizedcontrolled trials begin as high quality evidence, but theymay be downgraded based on limitations in the designand implementation, indirectness of evidence, unex-plained heterogeneity of results, imprecision of estimates,and high probability of publication bias. Observationalstudies begin as low quality evidence, but may beupgraded if there are large effect sizes, a dose-responsegradient, or if all plausible confounding would reduce thetreatment effect [26, 27]. The overall quality of evidencewas rated by consensus among the authors.Statistical analysesFollowing the procedures of Chillón and colleagues [24],we computed Cohen’s d as a measure of effect size foreach intervention. For interventions that included a con-trol group, effect size was computed as the standardizedmean difference of the changes in AST between the ex-perimental and control groups. For those that includedonly an experimental group, it was computed betweenbaseline and follow-up data. Additional file 2 providescomprehensive details on how effect sizes were com-puted for each intervention. Authors were contacted toobtain information required to calculate d. FollowingCohen’s [28] guidelines, effect size was categorized astrivial (d < 0.2), small (d = 0.2), medium (d = 0.5), or large(d = 0.8). Due to the large methodological heterogeneityof the included studies (see Table 1), meta-analysis wasconsidered inappropriate.ResultsThe flow of papers in the review process is depicted inFig. 1. Overall, 6318 papers were identified by the searchincluding 2339 in PubMed, 1555 in Web of Science, 377in Cochrane, 882 in SPORTDiscus, and 1165 in theNational Transportation Library. One paper was identi-fied from the authors’ personal libraries. All abstractswere screened, and 54 papers were found to be poten-tially eligible for inclusion. After a thorough selectionprocess, 27 papers were excluded due to the followingexclusion criteria: no/ineligible intervention, n = 17; nomeasure of physical activity or AST, n = 8; review article,n = 2. A total of 27 papers, reporting on 30 different in-terventions, were included for analyses [17–20, 29–51].Results are presented at the intervention level becausethree papers reported the findings of two different inter-ventions. Specifically, Buckley et al. [29] included a fallevent without control group and a spring event withcontrol group, Crawford and Garrard [34] included a pilotstudy with control schools (pilot schools) and a main studywithout control group (program schools), and Johnson etal. [41] reported case-control analyses using data from twodifferent surveys conducted in distinct populations (Bike-ability and CensusAtSchool). Eleven interventions wereconducted in the US, five in the UK, three in Canada, twoin Australia, Belgium, Denmark and New Zealand, and onein Spain and China. Another intervention was conductedsimultaneously in Canada and the UK.Characteristics of interventionsOf these interventions, six evaluated Safe Routes to School(SRTS) interventions [38, 39, 42, 43, 46, 48], seven evalu-ated school travel plan (STP) projects [17–20, 30, 34], andtwo examined stand-alone walking school buses (WSB)schemes [45, 47]. Four interventions focused on theeffects of bicycle training programs [35, 36, 41], fiveexamined the effects of stand-alone events or contests[29, 31, 33, 40], and two were multi-component inter-ventions that examined, among other things, changes inAST following the intervention [32, 51]. Others includedtwo studies examining the effect of curriculum-based pro-grams on AST [44, 50], one intervention using a drop-offspot from which driven children could walk to school withadult supervision [49], and an investigation of the effect ofLarouche et al. BMC Public Health  (2018) 18:206 Page 3 of 18Table1Characteristicsoftheincludedinterventions(n=30)Author,CountryInterventionandstrategiesMethodsEffectonASTaEffectsonotheroutcomesModeratorsandmediatorsBuckleyetal.2013[29][fallevent]Idaho,USADesignatedASTday,aimedtoencouragestudentsandtheirfamiliestopracticeASTonaspecificday.Strategies:encouragementonly.Design:Observationalcasestudy|pre-post(during,1dayafter)Duration:1day.Sample:2primaryschools.ASTmeasure:observationcounts(tripstoschoolonly,walkingandcyclingcombined).RelativeincreaseinAST(101%)ontheday;remainedhighonedayfollowing.Noneexamined.Noneexamined.Buckleyetal.2013[29][springevent]Idaho,USADesignatedASTday,aimedtoencouragestudentsandtheirfamiliestopracticeASTonaspecificday.Strategies:encouragementonly.Design:Quasi-experimental|pre-post(during,1dayafter,2weeksafter)Duration:1day.Sample:3primaryschools(2experimental,1control).ASTmeasure:observationcounts(tripstoschoolonly,walkingandcyclingcombined).IncreaseinASTsustainedat2-weekfollow-uprelativetothecontrolschool(χ2=11.6;p=0.009).Parentescortincreasedby333%onASTday(p<0.001).Parentinterviewssuggestedthattheschooljourneyisanopportunistictimetospendwiththeirchild.Noneexamined.Buliungetal.2011[30]4Canadianprovinces:Ontario,Alberta,BritishColumbia,NovaScotiaSchoolTravelPlanning,aschool-specificinterventionaimedtoincreaseASTthroughacollaborativestakeholderapproach.Strategiesengineering,education,enforcementandencouragement(variedbetweenschools).Design:Observational|pre-post(1year).Sample:12primaryschools,1489parentquestionnaires.Duration:1-year.ASTmeasure:studentclassroomsurvey+parentquestionnaire(tripsto/fromschool,allactivemodescombined).StudentreporteddataindicatedamodestincreaseinASTatfollow-up(from43.8to45.9%).d=0.05.Pvalueisunavailable(G.Faulkner,personalcommunication).13%ofparentsreporteddrivinglessasaresultoftheintervention.Accordingtoparentsthe3mosteffectivestrategieswereeducation,specialevents,andinfrastructureimprovements.Noneexamined.Bungumetal.2014[31]LasVegas,USADesignatedASTday,aimedtoencouragestudentsandtheirfamiliestopracticeASTonaspecificday.Strategies:encouragementonly.Design:Quasi-experiment|pre-2post(during,1-week)assessmentsDuration:1-dayevent.Sample:2schools(1experimentaland1control)|1336studentsages5-11.ASTmeasures:observationcounts(tripstoschoolonly,allactivemodescombined).IncreaseinthemodeshareofASTby7.4percentagepointsonthedayoftheevent.ASTwasthensignificantlyhigherthaninthecontrolschool(χ2=27.2;p<.001;d=0.29).ASTdroppedtobaselineratesat1-weekassessment.Noeffectonthenumberofmotorvehiclesobservedaroundtheschools.TheincreaseinASTwaslargeringirls(χ2=13.5;p<.05)thanboys(χ2=1.79;p=.056),butformalmoderationanalysiswasnotreported.Christiansenetal.2014[32]DenmarkAcomprehensiveschool-basedinterventiontoimprovenon-curricularPAthroughchangesofthephysicalandschoolenvironmentsupportedbyeducationalactivities.InterventionschoolswereaskedtohaveapolicytargetingAST,tooffercyclingsafetyeducation.Strategies:engineering,education,enforcementandpolicy.Design:RCT|pre-post(2-year).Sample:14schools(7experimentaland7control)|1014students|ages11-13.Duration:2years.ASTmeasure:traveldiary(tripsto/fromschool,allactivemodescombined).TheprevalenceofASTincreasedfrom87.8%to88.8%intheexperimentalgroupandfrom84.3%to85.3%inthecontrolgroupwithnodifferencebetweengroups(p=0.30;d=0.13).Theinterventionhadnoeffectonperceivedsafetyoftheschoolroute,parentalencouragementofcyclingandattitudestowardcycling.Note:improvedcyclinginfrastructurewasnotimplementedasplannedduetolimitedfunding.StudentsreportinganunsaferoutetoschoolatbaselineweremorelikelytouseASTatfollow-upintheinterventiongroupcomparedtostudentswithanunsaferouteinthecontrolgroup(OR=2.69;95%CI=1.20–6.07).Nointeractionsforgender,parentencouragement,distance,walkabilityandbaselineAST.Larouche et al. BMC Public Health  (2018) 18:206 Page 4 of 18Table1Characteristicsoftheincludedinterventions(n=30)(Continued)Author,CountryInterventionandstrategiesMethodsEffectonASTaEffectsonotheroutcomesModeratorsandmediatorsCoombesetal.2016UnitedKingdomAtechnology-basedintervention(BeattheStreet)aimedtoincreaseASTviaincentive-motivatedapproaches.Strategies:encouragementonly.Design:quasi-experimental|pre-2post(7-weeks,20-weeks)assessments.Duration:9-weekintervention.Sample:2schools(1interventionand1control)|80students|ages8-10.ASTmeasure:traveldiary(tripsto/fromschool,allactivemodescombined).At7-weekfollow-up:nodifferenceinASTbetweengroups.At20-weekfollow-up:10%increaseinASTintheinterventiongroupand7%decreaseincontrolgroup(p=0.056).Nodifferenceinaccelerometercountsperminute,buttherewasasmallerdeclineinMVPAintheexperimentalgroup(−15.1vs.-23.3min/day;p=0.020).ChildrenwhotouchedaBeattheStreetboxmoreoftenweremoreactive(+3.5min/dayofMVPAforchildrenwhoengagedintheinterventiononthemeannumberofdays,thatis14.5days).Crawford&Garrard,2013[34]Victoria,Australia[pilotschools]TheRide2SchoolProgram,whichconsistedmostlyofpromotionalactivitieswithsomeinfrastructurechanges.Strategies:education,encouragement,andengineering.Design:Quasi-experimentalmixedmethodsstudy|pre-post(~1year).Sample:4primaryschools(2controland2intervention)|participantswereages10-13,butyoungerchildrenwerealsocountedintheobservations.Duration:9-12months.ASTmeasure:observationcounts+studenthands-upsurvey(tripstoschoolonly,allactivemodescombined).ObservationresultsshowthatASTincreasedsignificantlyintheinnersuburbanpilotschoolandtheoutersuburbancontrolschool.Hands-upsurveysshowthatASTincreasedintheinnersuburbanpilotschoolandnochangeswerefoundintheotherschools.Qualitativedatasuggestthattheprogramwaseasiertoimplementwithinaschoolthatwassmaller,moreestablished,withaculturethatwasacceptingandenthusiasticaboutAST,inanareaofhigherdensityandlowercaruse,withmoreinfrastructureimprovementsandamore“hands-on”approachfromtheCoordinator.Resultsdifferedbylevelofurbanization(see“effectsonAST”column).Crawford&Garrard,2013[34]Victoria,Australia[programschools]TheRide2SchoolProgram,whichconsistedmostlyofpromotionalactivities,withoutinfrastructurechanges.Strategies:educationandencouragement.Design:Observationalmixedmethodsstudy|pre-post(6months).Sample:13primaryschools.Duration:6months.ASTmeasure:studenthands-upsurveyandparentsurvey(tripstoschoolonly).Parentsurveys:increaseincycling(from13.9to15.9%)anddecreaseinscooterorskateboarduse(from6.2to5.0%).Theproportionofparentsreporting≥1activetripsincreased(adjustedOR=1.67;95%CI=1.04-2.68).Childsurveys:decreaseinscooterorskateboarduse(from7.2to4.9%)andadecreaseinAST(from51.1to48.7%).Thelatterwasnolongersignificantafteradjustment.d=0.04forparentreportand−0.06forchildreport.Qualitativedatasuggestthatprogramimplementationvariedbetweenschoolsandthattheprogramwasmoreeffectivewhenschoolcommunitieswerehighlymotivated,whensecurebikestoragefacilitieswereoffered,whenallactivemodeswerepromotedequallybydynamicschoolstaff.Noneexamined.Ducheyneetal.2014[35]BelgiumCyclingtraining,aimedtoincreasecyclingskillsandencourageuptakeofcycling.Strategies:educationonly.Design:RCT|pre-post(1-week,5-month)postassessments.Sample:3primaryschools(2distinctexperimentalgroupsand1controlgroup)|94students|age9.3±0.5.Duration:4sessions(45mineach).ASTmeasure:parentreported(to/fromschool,timespentcyclingonly).Changesinweeklytimespentcyclingdidnotdifferbetweeninterventionandcontrolgroup(F=1.9;p>0.05)Effectsizes:interventionvs.controlgroup:d=0.46;intervention+parentvs.control:d=0.03.Children’scyclingskillscoreincreasedsignificantlymoreintheinterventiongroupfrompretopost(F=16.9;p<0.001)andfrompreto5-monthsfollow-up(F=16.8;p<0.001)comparedtothecontrolgroup.Nointerventioneffectsforparentalattitudes.Noneexamined.Larouche et al. BMC Public Health  (2018) 18:206 Page 5 of 18Table1Characteristicsoftheincludedinterventions(n=30)(Continued)Author,CountryInterventionandstrategiesMethodsEffectonASTaEffectsonotheroutcomesModeratorsandmediatorsGoodmanetal.2016[36]UnitedKingdomBikeability,anationalcycletrainingprogramforchildrenandadults.Strategies:educationonly.Design:retrospectivenaturalexperiment.Sample:3336childrenwhoseschooleitherhadoffered(n=2563)orhadnotyetofferedBikeability(n=773)|age10-11.DurationN/A.ASTmeasure:parent-reportedfrequencyofcycling(atleastonceaweekvs.less),andwhethercyclingwaschildren’susualtravelmodetoschool.ChildrenattendingschoolsthathadofferedBikeabilitywerenotmorelikelytocycleatleastonceaweek(OR=0.99;95%CI=0.89-1.10)andtocycletoschool(OR=0.73;95%CI=0.41-1.29).ChildrenwhoreceivedBikeabilitytrainingweremorelikelytocycleatleastonceaweek(OR=1.26;95%CI=1.16-1.37).ChildrenattendingschoolsthathadofferedBikeabilityweremuchmorelikelytohavecompletedtheprogram(68%vs.28%;p<0.001).Children’sparticipationinBikeabilitywasidentifiedasapotentialmediatoroftherelationshipbetweenschoolexposuretoBikeabilityandcyclingfrequency;however,nomaineffectwasobserved.Gutierrezetal.2014[37]Miami,USAImplementationofcrossingguards&ASTawarenesscampaign.Strategies:education,equity.Design:Quasi-experimental|pre-post.Sample:58intersectionsnearelementaryschools(24whereanewcrossingguardwaspresentand34control).ASTmeasure:observationcounts(toschoolonly,walkingandcyclingexamineseparatelyandcombined).Thenumberofpedestrianandcyclistsdidnotchangefollowingtheadditionofcrossingguards(p>0.05;d=0.03).Safety:increaseinstudents’useofsupervisedrouteswithamoderateeffectsize(partialη2=0.008).NochangesinparentalattitudesregardingASTsafety.Noneexamined.Hendersonetal.2013Atlanta,USASafeRoutestoSchool,acomprehensive,federally-fundedprogramintheUSdesignedtoincreaseASTthroughnon-infrastructureandinfrastructurestrategies.Strategies:education,encouragement,andengineering.Design:Observational|pre-post.Sample:oneprimaryschool(658students).Duration:2years.ASTmeasure:parent-reported(to/fromschool);allactivemodescombined.TheprevalenceofASTincreasedfrom18%to42%inthemorningtrip(p<0.0001;d=0.66),andfrom18%to23%intheafternoon(NS;d=0.17).Parentalperceptionaboutthehealthbenefits,perceptionthattheschoolstronglyencouragedAST,andenjoymentofwalking/bikingtoschoolincreasedsignificantly(allp<0.01).Noneexamined.Hincksonetal.2011aNewZealandSchoolTravelPlanning,aschool-specificinterventionaimedtoincreaseASTthroughacollaborativestakeholderandmulti-strategicapproach.Strategies:engineering,education,enforcement,encouragement,andpolicyinterventions.Design:Observational|pre-post(1to2years).Sample:33primaryschools|13,631students.Duration:1-2years.ASTmeasure:studentreported(toschoolonly,allactivemodescombined).ASTincreasedby5.9±6.8%.d=0.45.Noneexamined.LargerincreaseinASTwithlongerfollow-upperiod.Longerfollow-upperiod,smallerschoolrollandhigherpre-interventionrateofASTpredictedhigherratesofASTatfollow-up.Hincksonetal.2011b[18]NewZealandSchoolTravelPlanning,aschool-specificinterventionaimedtoincreaseASTthroughacollaborativestakeholderandmulti-strategicapproach.Strategies:engineering,education,enforcement,encouragement,andpolicyinterventions.Design:Observational|pre-post(1-,2-and3-year).Sample:56primaryschools|57,096students|gradesK-5.Duration:3years.ASTmeasure:studentreported(toschoolonly,allactivemodescombined).TherewasanincreaseinASTbythe3rdyearofimplementation(from40.5to42.2%;OR=2.65;95%CI=1.75-4.02).dafter1,2,and3years=0.17,0.51and0.54respectively.Noneexamined.LargerincreaseinASTwithlongerfollow-upperiod.Theprogramwasmoreeffectiveinolderstudents,insmallerschoolsandinthecityofAuckland,butitwaslesseffectiveinlowSESschools(allp<0.05).Larouche et al. BMC Public Health  (2018) 18:206 Page 6 of 18Table1Characteristicsoftheincludedinterventions(n=30)(Continued)Author,CountryInterventionandstrategiesMethodsEffectonASTaEffectsonotheroutcomesModeratorsandmediatorsHoelscheretal.2016[39]Texas,USASafeRoutestoSchool,acomprehensive,federally-fundedprogramintheUSdesignedtoincreaseASTthroughinfrastructureandnon-infrastructurestrategies.Strategies:educationandencouragement;someschoolsreceivedinfrastructureimprovements(engineering).Design:Quasi-experimental|pre-post(3year).Sample:grade4studentsfrom78elementaryschools.Schoolswereallocatedtoeitherof3conditions:“infrastructure”(n=23),“non-infrastructure”(n=21)orcontrolgroup(n=34).Duration:3years.ASTmeasure:studentreported(to/fromschool,allactivemodescombined).Infrastructureandnon-infrastructureschoolshadsignificantlyhigherratesofASTinthemorning(p=0.024and0.013respectively)andnon-infrastructureschoolshadsignificantlyhigheroverallASTrelativetocontrolschools(p=0.036).However,differencesbetweengroupsattenuatedovertime.Studentsfrombothinfrastructureandnon-infrastructureschoolshadhigherAST-relatedself-efficacy,andasimilarfindingwasnotedininfrastructureschoolsforparents.Studentsinnon-infrastructureschoolsreportedengaginginPAonmoredaysthanstudentsfromcomparisonschools.Parentsfromalltypesofschoolsperceivedworsewalkabilityandbikeabilityintheirneighborhoodsandschoolsovertime.Noneexamined.Hunteretal.2015[40]London,EnglandReading,EnglandVancouver,CanadaInternationalschoolcompetition,aimedtoincreaseASTviaincentive-motivatedapproaches.Strategies:encouragementonly.Design:ObservationalMixed-Methods(4-week).Sample:12primaryandsecondaryschools|3817Students|9-13yearolds.Duration:4weeks.ASTmeasure:Objectiveswipecardtechnologyandchildreports(to/fromschool,walkingonly).Thepercentageofwalkingtripsmeasuredbytheswipecarddecreasedoverthe4-weekmeasurementperiodfrom29to12%.However,atbaseline77%ofchildrenstatedthattheywalkedtoschoolatleastonceinthepastweekandthisproportionwas86%atfollow-up.d=−0.61withswipecardmethodologyand0.34withself-report.Children’sattitudes:perceivedtheinterventiontohelpphysicalandmentalhealth.Adultattitudes:91%ofparentsand72%ofteacherssurveyedstatedthattheythoughtthecompetitionhadencouragedchildrentospendmoretimewalkingwiththeirfriends.Thiswascorroboratedwithfocusgroupsdata.Noneexamined.Johnsonetal.2016[41]England[Bikeabilityschooltravelsurvey]Bikeability,anationalcycletrainingprogramforchildrenandadults.Strategies:educationonly.Design:retrospectivecase-controlanalysis.Sample:1345year5and6students.Duration:N/A.ASTmeasure:frequencyofcyclingtoschoolaswellascyclingingeneral(childreport).StudentswhoreceivedBikeabilityweremorelikelytocycletoschool(OR=2.25;95%CI=1.83-3.52).d=0.45.StudentswhoreceivedBikeabilitydidnotreportmorecyclingingeneral(OR=1.01;95%CI=0.75-1.38).Year6studentswhoreceivedBikeabilityexpressedgreaterconfidence(OR=1.81;95%CI=1.26-2.59).NoneexaminedJohnsonetal.2016[41]England[CensusAtSchool]Bikeability,anationalcycletrainingprogramforchildrenandadults.Strategies:educationonly.Design:retrospectivecase-controlanalysis.1745year7-9students.Duration:N/A.ASTmeasure:frequencyofcyclingtoschoolaswellascyclingingeneral(childreport).StudentswhoreceivedBikeabilityweremorelikelytocycletoschool(OR=1.60;95%CI=1.17-2.21).d=0.26.StudentswhoreceivedBikeabilityweremorelikelytoreportcycling≥30mininthepastweek(OR=1.27;95%CI=1.07-1.51).Noneexamined.Mammenetal.2014a[19]CanadaSchoolTravelPlanning,aschool-specificinterventionaimedtoincreaseASTthroughacollaborativestakeholderandmulti-strategicapproach.Strategies(variedacrossschools):education,encouragement,engineering,enforcement.Design:retrospectiveanalysis(1-yearfollowingimplementation).Sample:53primaryschools,7827questionnaires.Duration:1year.ASTmeasure:parentquestionnaire(to/fromschool,allactivemodescombined).17%oftheparentsreporteddrivinglessasaresultoftheintervention.Ofthese,about83%reportedswitchingfromdrivingtoAST.Nobaselinedataavailableandnohypothesistestperformed.Noneexamined.Parentsofolderstudents,thoseliving<3kmawayfromschool,attendingurbanandsuburbanschools,andattendingmedium-SESschoolsweremorelikelytoreportlessdriving.Larouche et al. BMC Public Health  (2018) 18:206 Page 7 of 18Table1Characteristicsoftheincludedinterventions(n=30)(Continued)Author,CountryInterventionandstrategiesMethodsEffectonASTaEffectsonotheroutcomesModeratorsandmediatorsMammenetal.2014b[64]CanadaSchoolTravelPlanning,aschool-specificinterventionaimedtoincreaseASTthroughacollaborativestakeholderandmulti-strategicapproach.Strategies(variedacrossschools):education,encouragement,engineering,enforcement.Design:Observational|pre-post(1-year).Sample:53primaryschools.Duration:1year.ASTmeasure:Studentreportedhands-upsurvey(to/fromschool,allactivemodescombined).Baselineandfollow-updatashowedthat27%and31%ofchildrenengagedinASTtoandfromschool,withnointerventioneffect.d=−0.02formorningtripand0.01forafternoontrip.ChangesinASTrangedfroma26%decreaseanda23%increaseacrossschools.Noneexamined.SchoolsthatcollectedbaselinedataintheFall(i.e.,September)andfollow-updatainWinter(i.e.,February)observeda5%decreaseinAST(B=−5.36,p<.05).McDonaldetal.2013[42]Oregon,USASafeRoutestoSchool,acomprehensive,federally-fundedprogramintheUSdesignedtoincreaseASTthroughnon-infrastructureandinfrastructurestrategies.Strategies:engineering,education,encouragement,andenforcement(variedbetweenschools).Design:Quasi-experimental|pre-post.Sample:9primaryschoolsand5middleschools(includes9experimentaland5controlschools).Duration:upto4years.ASTmeasure:studentreported(to/fromschool,walkingandbikingcombined).Education+encouragementwereassociatedwithincreasesinwalkingandbikingby2and5percentagepointsrespectively.Augmentingeducationprogramswithengineeringimprovementswasassociatedwithincreasesinwalkingandbikingof5-20percentagepoints.Noneexamined.MorecomprehensiveprogramswereassociatedwithgreaterincreasesinAST(see“effectonAST”column).McDonaldetal.2014[43]Florida,Oregon,Texas,DistrictofColumbia,USASafeRoutestoSchool,acomprehensive,federally-fundedprogramintheUSdesignedtoincreaseASTthroughnon-infrastructureandinfrastructurestrategies.Strategies:engineering,education,encouragement,andenforcement(variedbetweenschools).Design:Quasi-experimental|pre-post(5-year).Sample:801schoolsofwhich83%wereelementaryschools(includes378experimentaland423controlschools)|65,289students.Duration:5years.ASTmeasure:studentreported(to/fromschool,walkingandbikingcombined).Relativetocontrolschools,eachyearofparticipationinSRTSwasassociatedwitha1.1%increaseinAST(p=0.002;d=0.019).Engineeringimprovementsledtoa3.3percentagepointincreaseinwalkingandbiking(p=0.031;d=0.12),whileeducation+encouragementinterventionsledtoa0.9percentagepointincreaseperyear(p=0.025;d=0.15).Noneexamined.MorecomprehensiveprogramswereassociatedwithgreaterincreasesinAST(see“effectonAST”column).McMinnetal.2012[44]Glasgow,ScotlandTravellingGreen,a6-weekschoolbasedinterventionaimedtoincreaseASTviateacherlessonplansandstudentpacks(e.g.,materialtosetwalkinggoals,recordbehavior).Strategies:educationandencouragement.Design:Quasi-experimental|pre-post.Sample:5primaryschools(2experimentaland1control)|166students|ages8-9.Duration:6weeks.ASTMeasure:StepcountsandMVPAmeasuredbyActigraphaccelerometersduringthetripstoandfromschool.Interventiongrouphadsmallerdecreasesinmeansteps(−47vs.-205)andsecondsofMVPA(−33vs.-85)duringthemorningtrip.Oppositeresultsontheafternoontripforsteps(−222vs.-120)andMVPA(−125vs.-59).d<0.1forchangesinstepsandMVPAduringtheschooltrip.Childrenwhoreceivedtheinterventionshowedasmallerdeclineindailystepcounts(−901vs.-2528;d=0.52)andtimespentengaginginMVPA(−429vs.-1171s;d=0.46).Noneexamined.Larouche et al. BMC Public Health  (2018) 18:206 Page 8 of 18Table1Characteristicsoftheincludedinterventions(n=30)(Continued)Author,CountryInterventionandstrategiesMethodsEffectonASTaEffectsonotheroutcomesModeratorsandmediatorsMendozaetal.2011[45]Texas,USAWalkingSchoolBus,aimedtoincreaseASTbyhavingchildrenwalkinginadult-supervisedgroups.Strategies:encouragementonly.Design:RCT|datacollectedbeforeandin4th&5thweekofintervention.Sample:8primaryschools(4experimentaland4control)|149students|averageage=10years.Duration:5weeks.ASTmeasure:Studentreported(to/fromschool,allactivemodescombined).Intheintention-to-treatanalyses,interventionchildrenincreasedtheirweeklypercentASTfrom23.8%±9.2%atbaselineto54.0%±9.2%atfollow-up,whereascontrolchildrendecreasedtheirweeklypercentASTfrom40.2%±8.9%to32.6%±8.9%(p<.0001;d=0.40).InterventionchildrenincreasedtheirMVPAfrom46.6±4.5to48.8±4.5min/daywhilecontrolsdecreasedtheirsfrom46.1±4.3to41.3±4.3min/day(p=.029;d=0.18).Acculturation(p=.014)andparentoutcomeexpectations(p=.025)werebothassociatedwithincreasedAST.Parentself-efficacywaspositivelyassociatedwithAST(r=0.182;p=.032).Østergaardetal.2015[46]DenmarkSafeandSecureCyclingtoschool,aimedtoincreasecyclingbehaviorsthroughamulticomponentcyclingpromotionprogram.Strategies:encouragement,enforcement,educationandengineering.Design:Quasi-experiment|pre-post(1-year)assessments.Sample:25primaryschools(13experimentaland12control)|2401students|4th&5thgradestudents.Duration:1year.ASTMeasure:Studentreportednumberofcyclingtripsto/fromschoolinthepastweek(range=0-10).Changeinthenumberofcyclingtripsto/fromschoolwerenotsignificant(B=0.15trips;95%CI=−0.25;0.54).d=0.02.Cardiorespiratoryfitnessdecreasedintheinterventiongrouprelativetothecontrolgroup(B=−1.45mlO2·kg−1 ·min−1 ;p<0.0001).Nochangeinrecreationalcycling,overallPA,BMIandobesity.Noneexamined.Sayersetal.2012[47]Columbia,USAWalkingSchoolBus,aimedtoincreaseASTbyhavingchildrenwalkinginadult-supervisedgroups.Strategies:encouragementonly.Design:Casecontrolanalysiswheretheresearcherscomparedaccelerometry-measuredPAbetweenWSBparticipantsandnon-participants.Sample:3primaryschools|77students|ages8-9.Duration:1week.None.PercentageoftimespentinMVPAdidnotdifferbetweenWSBparticipantsandcontrols(allp≥0.17).d=−0.32.Theage-relatedgradientinMVPAwasattenuatedinWSBparticipants.Noneexamined.Stewartetal.2014[48]Florida,Mississippi,Washington,Wisconsin,USASafeRoutestoSchool,acomprehensive,federally-fundedprogramintheUSdesignedtoincreaseASTthroughnon-infrastructureandinfrastructurestrategies.Strategies:engineering,education,encouragement,andenforcement(variedbetweenschools).Design:Observational|pre-post.Sample:53primaryschoolswithSRTSprojects.Duration:5years.ASTmeasure:studenthands-upsurveyorobservationcounts(traveltoschoolonly).Assessedwalkingandbikingseparatelyandcombined.Attheschoollevel,ASTincreasedfrom12.8%to19.8%(p<.001;d=0.50);walkingfrom8.8%to13.3%(p<.001;d=0.46);cyclingfrom2.0%to3.2%(p=0.085;d=0.32).Noneexamined.Smallerchangesincyclinginschoolsthathadhigherlevelsofcyclingatbaseline(r=−0.40;p=0.009).WerenotassociatedwithchangesinAST:funding,numberofstudents/schoolsperproject,projecttype,interventionstrategies,schoollevel,enrollment,%ofstudentseligibleforfree/reducedcostmeals,andcharacteristicsoftheschoolneighborhood.Larouche et al. BMC Public Health  (2018) 18:206 Page 9 of 18Table1Characteristicsoftheincludedinterventions(n=30)(Continued)Author,CountryInterventionandstrategiesMethodsEffectonASTaEffectsonotheroutcomesModeratorsandmediatorsVanwolleghemetal.2014[49]West-Flanders,BelgiumAdrop-offspot(500-800mdistancefromschool)wasorganizedthatincludedteachersupervisiononthewalkto/fromthedesignatedarea.Strategies:encouragementonly.Design:Observational|datacollectedbeforeandduringintervention.Sample:2primaryschools|58students|ages6-12(mean=9.7±1.6years).Length:1week.ASTmeasure:childreportedthenumberofactivetripsusingthedrop-offspotinadiary.Thenumberofreportedactivetripsperweekincreasedfrom1to3(χ2=52.9;p<0.001;d=1.00).Pedometer-determinedstepcountsbefore/afterschoolhoursincreasedsignificantly(+732stepcounts/day;χ2=12.2;p<0.001),butnotdailystepcounts(p=0.16).Positiveperceptionoftheinterventionbyprincipalsandparents,butteachersexpresseddoubtsaboutfutureimplementation.Noneexamined.Villa-Gonzálezetal.2016[50]SpainInterventionaimedinincreasingASTthroughchangingchildrensafetyperceptionsandattitudes.Strategies:educationandencouragement.Design:Quasi-experimental|pre-post(6months).Sample:5primaryschools(3experimentaland2controls)|206students|ages8-11.Length:6months.ASTmeasure:student-reportednumberofwalkingandcyclingtripsinthepreviousweek(range=0-10trips).Increaseinthefrequencyofactivetripsininterventionschools(0.6±0.2)relativetocontrolschools(−0.4±0.3)[p=0.019;d=0.40].Whenexaminingtravelmodesseparately,significantchangeswereonlyobservedforwalking.Noneexamined.Noneexamined.Xuetal.2015[51]ChinaClickObesityStudy,amulticomponentlifestylechildhoodobesitypreventionprogramaimedtoenhancelifestylebehaviors.Strategies(forlifestylebehaviorchangeingeneral):education,encouragement.Design:RCT|pre-post(1year).Sample:8primaryschools(4interventionand4control)|1182students|4thgrade.Length:1year.ASTmeasure:Studentreportedschooltravelmode(to/fromschool,walkingandcyclingcombined).Participantsininterventionschoolsweremorelikelytochangetheirtravelmodetowalkingorcyclingtoschool(OR=2.24,95%CI=1.47-3.40;d=0.45)relativetothoseinthecontrolschools.Interventionparticipantsweremorelikelytoshowa≥0.5kg/m2decreaseinBMI(OR=1.44,95%CI=1.10-1.87),toincreasethefrequencyofjoggingorrunning(OR=1.55,95%CI=1.18-2.02),andtodecreaseTV/computeruse(OR=1.41,95%CI=1.09-1.84)andredmeatconsumption(OR=1.50,95%CI=1.15-1.95).Noneexamined.CharacteristicsarereportedattheinterventionlevelbecausesomepapersreportedthefindingsoftwointerventionsASTactiveschooltransportation,MVPAmoderate-to-vigorousphysicalactivity,NSnon-significant,ORoddsratio,PAphysicalactivity,SESsocio-economicstatus,SRTSSafeRoutstoSchool,STPschooltravelplanning,WSBwalkingschoolbusa Detailsonthecalculationofstandardizedeffectsizes(Cohen’sd)areprovidedinAdditionalfile2Larouche et al. BMC Public Health  (2018) 18:206 Page 10 of 18deploying crossing guards on travel behaviors [37]. In-cluded studies assessed AST in a variety of ways includingclassroom hand-up surveys [17, 18, 20, 34, 42, 43, 48],child surveys and diaries [32, 33, 41, 45, 46, 49–51],parent surveys [19, 30, 34–36, 38], direct observation[29, 31, 32, 37, 48], using a swipe card technology [40]or by recording accelerometer steps taken during theschool journey [44]. One study compared accelerometry-measured PA among participants in a WSB and non-participants [47]. Moreover, there was substantialheterogeneity in how AST was operationalized (e.g., travelmode on the day of the survey, usual travel mode, fre-quency of AST, etc.) and whether different active modeswere assessed separately or pooled together (Table 1).The majority of interventions focused on the elementaryschool setting. Only three studies included some second-ary school students [40, 41, 43].The target sample size ofincluded interventions ranged from 80 to 65,289 students.Schools were randomized to an intervention or a controlgroup in four interventions [32, 35, 45, 51]. Of theremaining interventions, 11 used a pre-post design with-out a control group [17, 18, 20, 29, 30, 34, 38–40, 48, 49],10 were quasi-experimental studies with a control group[29, 31, 33, 34, 37, 42–44, 46, 50], four were retrospectivecase-control studies [36, 41, 47], and one was a retrospect-ive study [19]. A detailed description of the interventionsand their main results is provided in Table 1.Quality assessmentQuality ratings are shown in Table 2. For individualcomponents of the EPHPP, the proportion of weak rat-ings was 3.3% for study design, 30.0% for withdrawalsand dropouts, 56.7% for selection bias, 60.0% for controlfor confounders, 66.7% for data collection methods, and100% for blinding. Following Chillón and colleagues’[24] modifications of the EPHPP, four studies were rated“non-applicable” for withdrawals and dropouts becauseparticipants were recruited after the intervention oc-curred and could not have dropped out. No studyreported that outcome assessors or participants wereblinded, and only two studies discussed blinding andspecified that it was not feasible in their intervention[35, 45]. In analyses that included the blinding componentof the EPHPP tool, only three studies were rated as “mod-erate” [32, 39, 45], and the remainder were rated as“weak”. In a sensitivity analysis that excluded the blindingcomponent, study quality was rated as weak for 21 inter-ventions [19, 20, 29–31, 33, 34, 36, 38, 40–43, 46, 48–51],moderate for six interventions [17, 18, 35, 37, 44, 47], andstrong for three interventions [32, 39, 45]. While our re-view included some randomized controlled trials, most in-dividual studies were rated as “weak” and very seriouslimitations in the design and implementation of interven-tions were noted, as mentioned above. Therefore, weattributed a low grade for the overall quality of evidence.Fig. 1 Flow of articles in the review processLarouche et al. BMC Public Health  (2018) 18:206 Page 11 of 18Intervention effectivenessOverall, 13 interventions resulted in a statistically significantincrease in AST [18, 29, 31, 38, 41, 42, 45, 47–51] while eightreported no changes in AST [20, 32, 33, 35, 37, 43, 46, 47].Of the latter studies, McMinn et al. [43] reported asmaller seasonal decline in PA among children in theirintervention group, and this can be viewed as a positivefinding given that PA typically declines during the fall andwinter. Five interventions did not include an hypothesistest for changes in AST [17, 19, 29, 30, 40]. The remainingstudies reported inconsistent or conflicting results.Specifically, in their pilot study, Crawford & Garrard [34]reported a significant increase in AST in their inner subur-ban school, but no change in their outer suburban schoolrelative to the control group. In their “program” phase, theyreported a significant increase in AST in experimentalschools based on parent surveys after adjusting for con-founders, but their child surveys indicated no change inAST after statistical adjustment. Goodman and colleagues[36] reported that children attending a school that had of-fered the Bikeability program did not cycle more frequently;however, those who actually took part in Bikeability didcycle more frequently, suggesting that parents/children in-terested in cycling may have self-selected to participate.Table 2 Quality assessment of active school transportation interventionsLead author (year) Selection bias StudydesignControl forconfoundersBlinding DatacollectionWithdrawalsand dropoutsGlobalratingGlobal ratingwithout blindingBuckley (2013) [fall event] Weak Moderate Weak Weak Weak Strong Weak WeakBuckley (2013) [spring event] Weak Moderate Strong Weak Weak Strong Weak WeakBuliung (2011) Weak Moderate Weak Weak Weak Weak Weak WeakBungum (2014) Weak Moderate Weak Weak Weak Strong Weak WeakChristiansen (2014) Strong Strong Strong Weak Moderate Moderate Moderate StrongCoombes (2016) Weak Moderate Weak Weak Weak Strong Weak WeakCrawford (2013) [pilot] Weak Strong Strong Weak Weak Strong Weak WeakCrawford (2013) [program] Weak Moderate Weak Weak Weak Strong Weak WeakDucheyne (2014) Moderate Strong Strong Weak Weak Strong Weak ModerateGoodman (2016) Moderate Moderate Strong Weak Weak Weak Weak WeakGutierrez (2014) Moderate Strong Strong Weak Weak Strong Weak ModerateHenderson (2013) Moderate Moderate Weak Weak Weak Weak Weak WeakHinckson (2011a) Moderate Moderate Weak Weak Moderate Moderate Weak ModerateHinckson (2011b) Moderate Moderate Weak Weak Moderate Strong Weak ModerateHoelscher (2016) Moderate Moderate Strong Weak Strong Strong Moderate StrongHunter (2015) Weak Moderate Weak Weak Weak Weak Weak WeakJohnson (2016) [Bikeability] Weak Moderate Weak Weak Weak N/A Weak WeakJohnson (2016) [CensusAtSchool] Weak Moderate Weak Weak Weak N/A Weak WeakMammen (2014a) Weak Weak Weak Weak Weak N/A Weak WeakMammen (2014b) Moderate Moderate Weak Weak Strong Weak Weak WeakMcDonald (2013) Weak Moderate Weak Weak Moderate Weak Weak WeakMcDonald (2014) Weak Moderate Strong Weak Weak Weak Weak WeakMcMinn (2012) Moderate Moderate Weak Weak Strong Strong Weak ModerateMendoza (2011) Moderate Strong Strong Weak Strong Strong Moderate StrongØstergaard (2015) Weak Moderate Weak Weak Weak Moderate Weak WeakSayers (2012) Weak Moderate Strong Weak Strong N/A Weak ModerateStewart (2014) Moderate Moderate Weak Weak Weak Weak Weak WeakVanwolleghem (2014) Weak Moderate Weak Weak Strong Strong Weak WeakVilla-Gonzalez (2016) Moderate Moderate Strong Weak Weak Weak Weak WeakXu (2015) Weak Strong Strong Weak Weak Strong Weak WeakQuality assessment was conducted with a modified version of the Effective Public Health Practice Project quality assessment tool for quantitative studies (EPHPP, 2003),which is provided in Additional file 1. Following EPHPP guidelines, studies with no weak ratings are rated “strong”, studies with one weak rating are rated “moderate”and studies with more than one weak rating are rated “weak”. Considering that blinding of participants may not be feasible in the context of AST interventions, globalratings with and without the blinding component of the EPHPP are presentedLarouche et al. BMC Public Health  (2018) 18:206 Page 12 of 18Finally, Hoelscher et al. [39] observed that while interven-tion schools had higher rates of AST over the 4-year studyperiod, the differences between groups waned over time.Details on the computation of effect sizes (Cohen’s d)are provided in Additional file 2. Cohen’s d varied mark-edly across interventions with a range of −0.61 to 0.75.Effect size could not be calculated for five interventions,including two that provided only follow-up data [19, 42],and three that provided insufficient data to allow for com-putation of d [29, 39]. Effect size was rated as trivial for 10interventions [17, 20, 30, 32, 34, 36, 37, 43, 46, 47], smallfor eight interventions [31, 33, 39, 43, 48, 50, 51], andmedium for one intervention [49]. Data from Hinckson etal. [18] indicate a trivial effect size after 1 year of follow-up, but a medium effect size after 2 or 3 years. Hendersonand colleagues’ [38] SRTS intervention yielded a mediumeffect size for the morning trip and a trivial effect size forthe afternoon trip. Data from Hunter et al. [39] indicateda medium decrease in AST as estimated with the swipecard methodology, but a small increase for self-reportedAST. In Crawford and colleagues’ [34] pilot program,there was a small effect size for the inner suburban schooland a trivial one for the outer suburban school. In the 3-group intervention by Ducheyne et al. [35], there was asmall effect size when comparing the intervention andcontrol groups, but a trivial effect size when comparingthe intervention + parent (which targeted parents inaddition to children) vs. the control group. Finally, datafrom McMinn et al. [44] suggest a small effect size forchanges in minutes of moderate-to-vigorous PA per day,but a moderate effect size for changes in steps/dayalthough both effect sizes were similar (d = 0.46 and 0.52respectively); however, effect size was trivial for changes insteps and MVPA during the school trip. Table 3 summa-rizes effect sizes by type of intervention; however, no clearpattern is evident.Moderators and mediatorsThirteen studies examined potential moderators.Hinckson et al. [17, 18] noted that longer follow-up pe-riods, smaller school size, higher school SES, and higherpre-intervention rate of AST predicted higher rates ofAST at follow-up. Safe Routes to School interventionsusing multiple strategies (as defined by the 6P model)achieved larger increases in AST [42, 43], and a longerfollow-up period was also associated with more substan-tial increases in AST [43]. In contrast, a short follow-upperiod was discussed as a potential reason for the lack ofa significant mode shift in other interventions [20, 46].Mammen and colleagues [19] reported that parents ofolder students, those living closer to school and attend-ing urban or suburban schools (relative to rural) weremore likely to report “driving less” following the imple-mentation of an STP. Of the potential moderatorsexamined by Stewart et al. [48], only the percentage ofstudents cycling at baseline was negatively associatedwith changes in cycling. In addition, Mendoza and col-leagues’ [45] results suggest that greater acculturation,more positive parental self-efficacy and outcome expec-tations may facilitate children’s engagement in AST.Goodman and colleagues [36] intended to assess chil-dren’s participation in cycle training as a mediator of therelationship between exposure to the Bikeability pro-gram at the school level and children’s cycling behavior.However they found a similar frequency of cyclingamong children exposed and unexposed to the program.No other study described formal mediation analyses.DiscussionWe have provided a comprehensive update on the ef-fectiveness of AST interventions among children andadolescents. Our search strategy identified 27 papers, de-scribing the findings of 30 distinct interventions, whichhave been published since the previous review [24]. In-cluded interventions were quite diverse and changes intravel behaviors varied markedly across interventions. In-cluded studies suggest that interventions with longerfollow-up periods may achieve greater modal shifts. Theseobservations are of particular importance for policy-makers and practitioners implementing AST interventions.Two large SRTS interventions found that interventionsincluding both educational activities and infrastructurechanges resulted in greater increases in AST than inter-ventions using only one of these strategies [42, 43]. Theseresults are consistent with social-ecological models thatposit that behavior is determined by multiple levels of in-fluence including individual, interpersonal, community,policy and built environment factors [52, 53].We noted that few interventions targeted secondaryschool students. As the correlates of AST may differ byage [54], one should not assume that interventions thatare effective among children will work as well with ado-lescents. Adolescents generally have higher independentmobility [55] and, as such, the influence of parental per-ceptions on their school travel mode may be weaker.However, adolescents may have less favorable attitudestoward AST [56, 57], and this might be a key factor toaddress for interventions in secondary schools.In the previous systematic review [24], all studies wererated as “weak” based on the EPHPP tool. In our review,10% of the studies were rated “moderate” (even with astricter interpretation of the blinding component ofEPHPP) and, when the blinding component was dis-missed as unfeasible, 30% of the studies were rated as“moderate” or “strong”. This suggests a marginal im-provement in study quality over the last 6 years; howeverthe overall quality of evidence as assessed with theGRADE approach remains low. Our sensitivity analysisLarouche et al. BMC Public Health  (2018) 18:206 Page 13 of 18Table 3 Effect size of active school transportation interventions stratified by intervention typeMeasure of effect size Cohen’s dSafe Routes to schoolHenderson (2013) Change in prevalence of AST (morning trip/afternoon trip) 0.66/0.17McDonald (2014) Change in prevalence of AST 0.19Østergaard (2015) Change in number of weekly AST trips 0.02Stewart (2014) Change in prevalence of AST 0.28School travel planningBuliung (2011) Change in prevalence of AST 0.05Crawford (2013) Change in prevalence of AST – inner suburban pilot school(direct observation/hands-up survey)0.27/0.30Crawford (2013) Change in prevalence of AST – outer suburban pilot school(direct observation/hands-up survey)−0.12/0.04Crawford (2013) Change in prevalence of AST in the program schools(parent report/child report)0.04/-0.06Hinckson (2011a) Change in prevalence of AST 0.14Hinckson (2011b) Change in prevalence of AST according to length of follow-up(1 year/2 years/3 years)−0.17; 0.51; 0.54Mammen (2014b) Change in prevalence of AST (morning trip/afternoon trip) −0.02; 0.01Walking school busesMendoza (2011) Change in percentage of trips using AST 0.40Sayers (2012) Difference in % of time spent in MVPA −0.32Cycle trainingDucheyne (2014) Change in weekly time spent engaging in AST (intervention vs.control group/intervention + parent vs. control group)0.46/0.03Johnson (2016) Difference in odds of cycling to school between trained anduntrained children (Bikeability survey)0.45Johnson (2016) Difference in odds of cycling to school between trained anduntrained (CensusAtSchool survey)0.26Goodman (2016) Difference in odds of cycling to school between trained anduntrained (school level/individual level)−0.17; 0.18Special eventsBungum (2014) Change in number of students engaging in AST 0.29Coombes (2016) Change in proportion of trips using AST at 7-week and20-week follow ups respectively−0.32; 0.24Hunter (2015) Change in prevalence of AST (measured with swipecard/self-report)−0.61; 0.34Multi-component interventionsChristiansen (2014) Change in odds of engaging in AST 0.13Xu (2015) Change in odds of engaging in AST 0.45Curriculum-based interventionsMcMinn (2012) Difference in commuting steps and MVPA betweenintervention and control groups0.06/-0.03McMinn (2012) Difference in daily steps and MVPA betweenintervention and control groups0.52/0.46Villa-Gonzalez (2016) Changes in weekly number of active trips 0.40Drop-off spotsVanwolleghem (2014) Change in frequency of AST 0.75Crossing guardsGutierrez (2014) Change in number of students engaging in AST 0.03AST active school transportation, MVPA moderate-to-vigorous physical activity. Effect sizes were computed as detailed in Additional file 2. Some studies appearmore than once because they have multiple measures of effect size. Cohen’s d could not be computed for 5 interventions because insufficient information wasprovided by the authors. Following Cohen’s28 guidelines, effect size can be categorized as trivial (d < 0.2), small (d = 0.2), medium (d = 0.5), or large (d = 0.8)Larouche et al. BMC Public Health  (2018) 18:206 Page 14 of 18shows that the blinding component exerted a floor effecton quality scores. Because all interventions received a“weak” rating for blinding, they could not be ratedhigher than “moderate”. Future improvement in qualityratings could be made by controlling for confoundersand by using valid and reliable measures of AST, whichhave been reviewed elsewhere [58].The calculated effect sizes for most interventions weretrivial to small based on Cohen’s [28] thresholds. Al-though these widely-used thresholds are arbitrary, wehave used them in the absence of alternative options.Given the large reach of interventions such as SRTS andSTP, an effect size labeled as “trivial-to-small” may stillbe highly relevant from a population health perspective.Interestingly, a pooled intervention effect of d = 0.12 wasobtained in a meta-analysis of 30 controlled trials on PAinterventions among children and adolescents [59].Furthermore, while our review focused specifically onthe effect of interventions on travel behaviors, some in-cluded interventions have documented positive changes inother important outcomes such as children’s cycling skills[35], safe street crossing behaviors [37], attitudes towardAST [40], and higher daily PA [44, 45]. Substantial reduc-tions in road traffic injuries among children have alsobeen noted following implementation of SRTS [15]. Morebroadly, it has been proposed that interventions such asSRTS may benefit the larger communities in which theyare implemented, and not only children [60].Mediators and moderatorsA better understanding of the mediators and moderatorsof AST interventions could help identify what works forwhom and why [61, 62]. Of particular interest, manystudies emphasized the importance of having long termfollow-ups given that implementation of complex ASTinterventions may require a substantial amount of time[17–20, 43, 46]. Similarly, qualitative evaluations focus-ing on the implementation of AST interventions alsoidentify lack of time as a key challenge [63, 64]. To ad-dress the issue of follow-up length, some authors sug-gested that granting agencies should be encouraged toprovide more long term funding [63, 64].While there has been increased interest in studyingmoderators of AST interventions, none of the includedstudies conducted formal mediation analyses and most in-terventions did not include an explicit theoretical frame-work. Given the important role of parents in travel modedecision making [65], interventions that increase roadsafety may be more effective if they also target parents’self-efficacy in allowing their child to engage in AST [45].Implementation of interventionsUnderstanding the implementation of complex ASTinterventions may provide valuable information for thereader to contextualize the effectiveness of such inter-ventions. This may be particularly important for inter-ventions such as SRTS and STP that are essentiallyevaluated as “natural experiments” [66] because in mostcases, exposure to the intervention is not under the con-trol of the investigators. This is a threat to internal valid-ity because the fidelity of implementation varies, but atthe same time, it represents more closely how anintervention is implemented in the “real world”. Manyinterventions included in this review reported thatimplementation varied substantially between schools[19, 20, 32, 34, 46], and in some cases, planned changeswere not implemented as scheduled [32, 37, 46]. Crawfordand Garrard [34] also reported that the implementation ofthe Ride2School program was affected by the motivationof school communities. Such challenges and discrepanciesmay bias our results toward the null hypothesis.Lack of resources or unequal access to resources hasbeen noted by many authors as a limitation to AST inter-ventions [32, 63, 64]. In Canada, STPs and WSBs are im-plemented by non-governmental organizations and lack ofsupport from provincial and federal governments has beenidentified as a major barrier [64]. In Texas, stakeholdersexpressed difficulty in navigating the SRTS regulatoryprocess and emphasized that access to SRTS funding wasvery challenging for low income communities given thatno up-front funding was provided [63]. More generally,WSBs typically rely on volunteers which often makes longterm sustainability challenging [23, 67]. Providing paidWSB leaders may help overcome this issue.Strengths and limitationsAs in the previous review [24], we noted that many in-cluded studies did not include a control group. Anotherlimitation is that the original EPHPP tool seems bettersuited to assess studies where the unit of allocation isthe individual. To address this issue, we have modifiedthe tool so that the questions are more relevant toschool-based interventions (see Additional file 1). Never-theless, like other quality assessment tools, the scoringsystem of the EPHPP is rigid and may not always distin-guish more robust studies from weaker ones [68]. Forexample, in our review, no study reported that outcomeassessors were blinded, creating a floor effect wherebyno intervention can be rated higher than “moderate”.Notwithstanding the importance of blinding in prevent-ing observer bias and Hawthorne effects, a quality as-sessment tool should be able to discriminate strongerstudies from weaker ones. Our sensitivity analysis with-out the blinding component of the EPHPP intended toaddress this issue. We acknowledge that the use of adifferent quality assessment tool could have resulted indifferent ratings of study quality as observed previously[68]. Finally, the large heterogeneity in the measurementLarouche et al. BMC Public Health  (2018) 18:206 Page 15 of 18and operationalization of AST precluded meta-analysis.The development of a standard measurement protocolmay help address this issue.The rigorous systematic review process is an import-ant strength of the study. We followed the same searchstrategy as Chillón and colleagues [24] and computedstandardized effect sizes which should help readers in-terpret the effectiveness of interventions and performsample size calculations. Finally, the discussion of mod-erators, mediators and factors related to implementationshould help researchers refine current interventions.ConclusionsThe present systematic review highlights the diversity ofinterventions that have been implemented to promoteAST in the last few years, and shows that travel behaviorchange varied markedly between interventions. Many in-terventions have shown significant increases in AST, butcaution is required in interpretation given the lowquality of evidence. This underscores a need for inter-ventions using stronger study designs.Our findings have implications for researchers andpractitioners. First, it may take time for interventions tohave an effect on children’s travel behaviors. Therefore,follow-ups of at least 2 years should be conducted whenpossible to minimize the risk of type II error. Second,while many authors indicated that implementation of in-terventions varied markedly across schools, it is unclearhow this variation may influence effectiveness. Hence fu-ture research should examine the potential moderatingeffect of implementation. The fact that some interven-tions were not implemented as planned suggests thatsome of the effect sizes reported herein may be conser-vative. Third, there remains a clear need for investigationof the mediators of travel behavior change.Only three interventions included some high schools,highlighting a need for more research intervening in sec-ondary school settings. This is important given that thefactors associated with AST may differ markedly be-tween children and adolescents. Finally, because somechildren may live too far from their school, interventionsaiming to promote active transportation to/from otherdestinations such as parks, shops, sport venues, andfriends’ and relatives’ houses may also be warranted [69].Additional filesAdditional file 1: Appendix 1. Adjusted criteria for the Effective PublicHealth Practice Project quality assessment tool for quantitative studies.(DOCX 33 kb)Additional file 2: Appendix 2. Computation of effect sizes. (DOCX 34 kb)AbbreviationsAST: Active school transport; EPHPP: Effective public health practice project;GRADE: Grades of recommendation, assessment, development, andevaluation; PA: Physical activity; SRTS: Safe routes to school; STP: Schooltravel plans; UK: United Kingdom; US: United States; WSB: Walkingschool busesAcknowledgementsNot applicable.FundingGF holds a Canadian Institutes of Health Research-Public Health Agency ofCanada (CIHR-PHAC) Chair in Applied Public Health. RL was supported by apostdoctoral fellowship from the Canadian Institutes of Health Research.Funders had no role in the study.Availability of data and materialsThe data that was used to compute effect sizes is available in Additional file 2.Authors’ contributionsRL completed quality assessment, data extraction, statistical analyses, anddrafted the manuscript. GM conducted the search, screened papers forinclusion, assisted in data extraction and provided feedback on themanuscript. DAR conducted quality assessment, assisted with statisticalanalyses, and provided feedback on the manuscript. GF screened papers forinclusion and provided feedback on the manuscript. All authors read andapproved the final manuscript.Ethics approval and consent to participateNot applicable.Consent for publicationNot applicable.Competing interestsThe authors declare that they have no competing interests.Publisher’s NoteSpringer Nature remains neutral with regard to jurisdictional claims inpublished maps and institutional affiliations.Author details1Healthy Active Living and Obesity Research Group, Children’s Hospital ofEastern Ontario Research Institute, Ottawa ON K1H 8L1, Canada. 2Faculty ofHealth Sciences University of Lethbridge, 4401 University Drive, office M3049Lethbridge, Alberta T1K 3M4, Canada. 3Centre for Addiction and MentalHealth, Institute for Mental Health Policy Research, 1001 Queen St West,Toronto, ON M6J 1H4, Canada. 4School of Psychological Sciences and Health,University of Strathclyde, 16 Richmond St, Glasgow G1 1XQ, Glasgow, UK.5School of Kinesiology, University of British Columbia, D H Copp Building4606, 2146 Health Sciences Mall, Vancouver, BC V6T 1Z3, Canada. 6Center forHip Health and Mobility, Robert H.N. Ho Research Centre, 5th Floor, 2635Laurel St, Vancouver, BC V5Z 1M9, Canada.Received: 28 April 2017 Accepted: 20 December 2017References1. Larouche R, Saunders TJ, Faulkner GEJ, Colley RC, Tremblay MS. Associationsbetween active school transport and physical activity, body compositionand cardiovascular fitness: a systematic review of 68 studies. J Phys ActHealth. 2014;11(1):206–27.2. Schoeppe S, Duncan MJ, Badland H, Oliver M, Curtis C. Associations ofchildren’s independent mobility and active travel with physical activity,sedentary behaviour, and weight status: a systematic review. J Sci MedSport. 2013;16(4):312–9.3. Andersen LB, Wedderkopp N, Kristensen P, Moller NC, Froberg K, CooperAR. Cycling to school and cardiovascular risk factors: a longitudinal study. JPhys Act Health. 2011;8(8):1025–33.4. de Nazelle A, Nieuwenhuijsen MJ, Antó JM. Improving health throughpolicies that promote active travel: a review of evidence to supportintegrated health impact assessment. Environ Int. 2011;37(4):766–77.Larouche et al. BMC Public Health  (2018) 18:206 Page 16 of 185. Marshall JD, Wilson RD, Meyer KR, Rajangam SK, McDonald NC, Wilson EJ.Vehicle emissions during children’s school commuting: impacts ofeducation policy. Env Sci Technol. 2010;44(5):1537–43.6. Ramanathan S, O’Brien C, Faulkner G, Stone M. Happiness in motion: emotions,well-being and active school travel. J School Health. 2014;84(8):516–23.7. Rissotto A, Tonucci F. Freedom of movement and environmentalknowledge in elementary school children. J Environ Psychol. 2002;22:65–77.8. Martinez-Gomez D, Ruiz JR, Gomez-Martinez S, et al. Active commuting toschool and cognitive performance in adolescents. Arch Pediatr AdolescMed. 2011;165(4):300–5.9. Gray C, Larouche R, Barnes JD, et al. Are we driving our kids to unhealthyhabits? Results from the active healthy kids Canada 2013 report card onphysical activity for children and youth. Int J Environ Res Public Health.2014;11(6):6009–20.10. Grize L, Bringolf-Isler B, Martin E, Braun-Farhländer C. Trend in activetransportation to school among Swiss school children and its associatedfactors: three cross-sectional surveys 1994, 2000 and 2005. Int J Behav NutrPhys Act. 2010;7:28.11. McDonald NC. Active commuting to school: trends among USschoolchildren 1969-2001. Am J Prev Med. 2007;32(6):509–16.12. Trang NHHD, Hong TK, Dibley MJ. Active commuting to school amongadolescents in ho chi Minh City, Vietnam: changes and predictors in alongitudinal study, 2004 to 2009. Am J Prev Med. 2012;42(2):120–8.13. Tremblay MS, Gray CE, Akinroye KK, et al. Physical activity of children: aglobal matrix of grades comparing 15 countries. J Phys Act Health. 2014;11(Suppl 1):113–25.14. Safe Routes to School National Partnership. National Policy and Advocacy.2015. Available from: http://www.saferoutespartnership.org/blog/6-es-safe-routes-school-embracing-equity Accessed 11 Feb 2017.15. DiMaggio C, Li G. Effectiveness of a safe routes to school program inpreventing school-aged pedestrian injury. Pediatrics. 2013;131(2):290–6.16. Muennig PA, Epstein M, Li G, DiMaggio C. The cost-effectiveness of newYork City’s safe routes to school program. Am J Public Health. 2014;104(7):1294–9.17. Hinckson EZ, Badland HM. School travel plans: preliminary evidence forchanging school-related travel patterns in elementary school children. Am JHealth Promot. 2011;25(6):368–71.18. Hinckson EA, Garrett N, Duncan S. Active commuting to school in New Zealandchildren (2004-2008): a quantitative analysis. Prev Med. 2011;52(5):332–6.19. Mammen G, Stone MR, Faulkner G, et al. Active school travel: an evaluation ofthe Canadian school travel planning intervention. Prev Med. 2014;60:55–9.20. Mammen G, Stone MR, Buliung R, Faulkner G. “Putting school travel on themap”: facilitators and barriers to implementing school travel planning inCanada. J Transp Health. 2015;2(3):318–26.21. Rowland D, DiGuiseppi C, Gross M, Afolabi E, Roberts I. Randomisedcontrolled trial of site specific advice on school travel patterns. Arch DisChild. 2003;88:8–11.22. Kingham S, Ussher S. An assessment of the benefits of the walking school busin Christchurch, New Zealand. Transp Res Part A-Policy Pract. 2007;41(6):502–10.23. Smith L, Norgate SH, Cherrett T, Davies N, Winstanley C, Harding M. Walkingschool buses as a form of active transportation for children – a review ofthe evidence. J Sch Health. 2015;85:197–210.24. Chillón P, Evenson KR, Vaughn A, Ward DS. A systematic review ofinterventions for promoting active transportation to school. Int J Behav NutrPhys Act. 2011;8(10)25. Effective Public Health Practice Project. Quality Assessment Tool forQuantitative Studies. 2009. Available from http://www.ephpp.ca/tools.htmlAccessed 11 Feb 2017.26. Guyatt GH, Oxman AD, Vist GE, et al. GRADE: an emerging consensus on ratingquality of evidence and strength of recommendations. BMJ. 2008;336:924–6.27. Guyatt GH, Oxman AD, Vist GE, Kunz R, Falck-Ytter Y, Schünemann HJ.GRADE: what is “quality of evidence” and why is it important to clinicians?BMJ. 2008;336:995–8.28. Cohen J. Statistical power analysis for the behavioral science. 2nd ed.Hillsdale, NJ: Lawrence Erlbaum; 1988.29. Buckley A, Lowry MB, Brown H, Barton B. Evaluating safe routes to schoolevents that designate days for walking and bicycling. Transp Policy. 2013;30:294–300.30. Buliung R, Faulkner G, Beesley T, Kennedy J. School travel planning:mobilizing school and community resources to encourage active schooltransportation. J Sch Health. 2011;81:704–12.31. Bungum TJ, Clark S, Aguilar B. The effect of an active transport to schoolintervention at a suburban elementary school. Am J Health Educ. 2014;45(4):205–9.32. Christiansen LB, Toftager M, Ersbøll AK, Troelsen J. Effects of a Danishmulticomponent physical activity intervention on active school transport. JTransp Health. 2014;1:174–81.33. Coombes E, Jones A. Gamification of active travel to school: a pilotevaluation of the beat the street physical activity intervention. Health Place.2016;39:62–9.34. Crawford S, Garrard J. A combined impact-process evaluation of a programpromoting active transport to school: understanding the factors thatshaped program effectiveness. J Environ Public Health. 2013;81696135. Ducheyne F, de Bourdeaudhuij I, Lenoir M, Cardon G. Effects of a cycletraining course on children’s cycling skills and levels of cycling to school.Accid Anal Prev. 2014;67:49–60.36. Goodman A, van Sluijs EM, Ogilvie D. Impact of offering cycle training inschools upon cycling behaviour: a natural experimental study. Int J BehavNutr Phys Act. 2016;13:34.37. Gutierrez CM, Slagle D, Figueras K, Anon A, Huggins AC, Hotz G. Crossingguard presence: impact on active transportation and injury prevention. JTransp Health. 2014;1:116–23.38. Hendersen S, Tanner R, Klanderman S, Mattera A, Webb LM, Steward J. Saferoutes to school: a public health practice success story—Atlanta, 2008–2010. J Phys Act Health. 2013;10(2):141–2.39. Hoelscher D, Ory M, Dowdy D, et al. Effects of funding allocation for saferoutes to school programs on active commuting to school and relatedbehavioral, knowledge, and psychosocial outcomes: results from the Texaschildhood obesity prevention policy evaluation (T-COPPE) study. EnvironBehav. 2016;48(1):210–29.40. Hunter RF, de Silva D, Reynolds V, Bird W, Fox KR. International inter-schoolcompetition to encourage children to walk to school: a mixed methodsfeasibility study. Int J Behav Nutr Phys Act. 2015;8:19.41. Johnson R, Frearson M, Hewson P. Can bicycle training for children increaseactive travel? Eng Sustain. 2016;169(2):49–57.42. McDonald NC, Yang Y, Abbott SM, Bullock AN. Impact of the safe routes toschool program on walking and biking: Eugene. Oreg Study Transp Policy.2013;29:243–8.43. McDonald NC, Steiner RL, Lee C, Smith TR, Zhu X, Yang Y. Impact of thesafe routes to school program on walking and bicycling. J Am Plan Assoc.2014;80(2):153–67.44. McMinn D, Rowe DA, Murtagh S, Nelson NM. The effect of a school-basedactive commuting intervention on children's commuting physical activityand daily physical activity. Prev Med. 2012;54:316–8.45. Mendoza JA, Watson K, Baranowski T, Nicklas TA, Uscanga DK, Hanfling MJ.The walking school bus and children’s physical activity: a pilot clusterrandomized controlled trial. Pediatrics. 2011;128(3):e537–44.46. Østergaard L, Støckel JT, Andersen LB. Effectiveness and implementation ofinterventions to increase commuter cycling to school: a quasi-experimentalstudy. BMC Public Health. 2015;15:1199.47. Sayers SP, LeMaster JW, Thomas IM, Petroski GF, Ge B. A walking school busprogram: impact on physical activity in elementary school children inColumbia, Missouri. Am J Prev Med. 2012;43(5S4):S384–9.48. Stewart O, Moudon AV, Claybrooke C. Multistate evaluation of safe routes toschool programs. Am J Health Promot. 2014;28(3S):S89–96.49. Vanwolleghem G, D’Haese S, Van Dyck D, de Bourdeaudhuij I, Cardon G.Feasibility and effectiveness of drop-off spots to promote walking to school.Int J Behav Nutr Phys Act. 2014;11:136.50. Villa-González E, Ruiz JR, Ward DS, Chillón P. Effectiveness of an activecommuting school-based intervention at 6-month follow-up. Eur J PubHealth. 2016;26(2):272–6.51. Xu F, Ware RS, Leslie E, Tse LA, Wang Z, Li J, Wang Y. Effectiveness of arandomized controlled lifestyle intervention to prevent obesity amongChinese primary school students: CLICK-obesity study. PLoS One. 2015;10(10):e0141421.52. Giles-Corti B. People or places: what should be the target? J Sci Med Sport.2006;9(5):357–66.53. Sallis JF, Cervero RB, Ascher W, Henderson KA, Kraft MK, Kerr J. An ecologicalapproach to creating active living communities. Ann Rev Public Health.2006;27:297–322.54. Mitra R, Buliung RN. Exploring differences in school travel mode choicebehaviour between children and youth. Transp Policy. 2015;42:4–11.Larouche et al. BMC Public Health  (2018) 18:206 Page 17 of 1855. Shaw B, Bicket M, Elliot B, Fagan-Watson B, Mocca E, Hillman M. Children’sindependent mobility: an international comparison and recommendationsfor action. London, UK: Policy Studies Institute; 2015.56. Lorenc T, Brunton G, Oliver S, Oliver K, Oakley A. Attitudes to walking andcycling among children, young people and parents: a systematic review. JEpidemiol Community Health. 2008;62:852–7.57. Stuckless J. Project overview: high school pilot project 2010-2012. GreenCommunities Canada. 2012. Available from: http://activesafe.nonprofitwebsites.ca/sites/default/files/Overview%20of%20Ottawa%20High%20School%20STP%20Pilot%20Project%20-%20Jamie%20Stuckless.pdfAccessed 11 Feb 2017.58. Larouche R, Oyeyemi AL, Prista A, Onywera VO, Akinroye KK, Tremblay MS. Asystematic review of active transportation research in Africa and thepsychometric properties of measurement tools in children and youth. Int JBehav Nutr Phys Act. 2014;11:129.59. Metcalf B, Henley W, Wilkin T. Effectiveness of intervention on physicalactivity of children: systematic review and meta-analysis of controlled trialswith objectively measured outcomes (EarlyBird 54). BMJ. 2012;345:e5888.60. Watson M, Dannenberg AL. Investments in Safe Routes to schoolprojects: public health benefits for the larger community. Prev ChronicDis. 2008;5(3):A90.61. Baranowski T, Anderson C, Carmack C. Mediating variable framework inphysical activity interventions: how are we doing? How might we dobetter? Am J Prev Med. 1998;15(4):266–97.62. Bauman AE, Sallis JF, Dzewaltowski DA, Owen N. Toward a betterunderstanding of the influences on physical activity: the role ofdeterminants, correlates, causal variables, mediators, moderators, andconfounders. Am J Prev Med. 2002;23(Suppl 1):5–14.63. Atteberry H, Dowdy D, Oluyomi A, Nichols D, Ory MG, Hoelscher DM. Acontextual look at safe routes to school implementation in Texas. EnvBehav. 2016;48(1):192–209.64. Mammen G, Stone MR, Buliung N, Faulkner G. School travel planning inCanada: identifying child, family, and school-level characteristics associatedwith travel mode shift from driving to active school travel. J Transp Health.2014;1(4):288–94.65. Faulkner GEJ, Richichi V, Buliung RN, Fusco C, Moola F. What’s “quickest andeasiest?”: parental decision making about school trip mode. Int J Behav NutrPhys Act. 2010;7:62.66. Ramanathan S, Allison KR, Faulkner G, Dwyer JJM. Challenges in assessingthe implementation and effectiveness of physical activity and nutritionpolicy interventions as natural experiments. Health Promot Int. 2008;23(3):290–7.67. Kingham S, Ussher S. Ticket to a sustainable future: an evaluation of thelong-term durability of the walking school bus programme in Christchurch,New Zealand. Transp Policy. 2005;12(4):314–23.68. Armijo-Olivo S, Stiles CR, Hagen NA, Blondo PD, Cummings GG. Assessmentof study quality for systematic reviews: a comparison of the Cochranecollaboration risk of bias tool and the effective public health practiceproject quality assessment tool: methodological research. J Eval Clin Pract2012;18(1):12-8.69. Larouche R, Barnes J, Tremblay MS. Too far to walk or bike? Can J PublicHealth. 2013;104(7):e487–9.•  We accept pre-submission inquiries •  Our selector tool helps you to find the most relevant journal•  We provide round the clock customer support •  Convenient online submission•  Thorough peer review•  Inclusion in PubMed and all major indexing services •  Maximum visibility for your researchSubmit your manuscript atwww.biomedcentral.com/submitSubmit your next manuscript to BioMed Central and we will help you at every step:Larouche et al. BMC Public Health  (2018) 18:206 Page 18 of 18


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