UBC Faculty Research and Publications

Systematic review of the relationships between sedentary behaviour and health indicators in the early… Poitras, Veronica J; Gray, Casey E; Janssen, Xanne; Aubert, Salome; Carson, Valerie; Faulkner, Guy; Goldfield, Gary S; Reilly, John J; Sampson, Margaret; Tremblay, Mark S Nov 20, 2017

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

Item Metadata


52383-12889_2017_Article_4849.pdf [ 800.62kB ]
JSON: 52383-1.0360655.json
JSON-LD: 52383-1.0360655-ld.json
RDF/XML (Pretty): 52383-1.0360655-rdf.xml
RDF/JSON: 52383-1.0360655-rdf.json
Turtle: 52383-1.0360655-turtle.txt
N-Triples: 52383-1.0360655-rdf-ntriples.txt
Original Record: 52383-1.0360655-source.json
Full Text

Full Text

RESEARCH Open AccessSystematic review of the relationshipsscreen-based sedentary behaviours such as reading and storytelling. Additional high-quality research using valid, Adiposity,RisksThe Author(s) BMC Public Health 2017, 17(Suppl 5):868DOI 10.1186/s12889-017-4849-8CanadaFull list of author information is available at the end of the article* Correspondence: veronicapoitras@gmail.com1Healthy Active Living and Obesity Research Group, RI #1, Children’s Hospitalof Eastern Ontario Research Institute, 401 Smyth Road, Ottawa, ON K1H 8L1,Keywords: Sedentary behaviour, Infants, Toddlers, Preschoolers, Early years, Screen time, Sitting, ReadingMotor development, Cognitive development, Bone and skeletal health, Cardiometabolic health, Fitness,and reliable measures is needed to more definitively establish the relationships between durations, patterns, andtypes of SB and health indicators, and to provide insight into the appropriate dose of SB for optimal health in theearly years.and health promotion in the early years, but also highlighindicators in the early years (0–4 years)Veronica J. Poitras1*, Casey E. Gray1, Xanne Janssen2, Salome Aubert1, Valerie Carson3, Guy Faulkner4,Gary S. Goldfield1, John J. Reilly2, Margaret Sampson1,5 and Mark S. Tremblay1AbstractBackground: The purpose of this systematic review was to examine the relationships between sedentary behaviour(SB) and health indicators in children aged 0 to 4 years, and to determine what doses of SB (i.e., duration, patterns[frequency, interruptions], and type) were associated with health indicators.Methods: Online databases were searched for peer-reviewed studies that met the a priori inclusion criteria:population (apparently healthy, 1 month to 4.99 years), intervention/exposure and comparator (durations, patterns,and types of SB), and outcome/health indicator (critical: adiposity, motor development, psychosocial health,cognitive development; important: bone and skeletal health, cardiometabolic health, fitness, risks/harm). The qualityof the evidence was assessed by study design and outcome using the Grading of Recommendations Assessment,Development and Evaluation (GRADE) framework.Results: Due to heterogeneity, meta-analyses were not possible; instead, narrative syntheses were conducted,structured around the health indicator and type of SB. A total of 96 studies were included (195,430 participantsfrom 33 countries). Study designs were: randomized controlled trial (n = 1), case-control (n = 3), longitudinal (n = 25),longitudinal with additional cross-sectional analyses (n = 5), and cross-sectional (n = 62). Evidence quality rangedfrom “very low” to “moderate”. Associations between objectively measured total sedentary time and indicators ofadiposity and motor development were predominantly null. Associations between screen time and indicators ofadiposity, motor or cognitive development, and psychosocial health were primarily unfavourable or null.Associations between reading/storytelling and indicators of cognitive development were favourable or null.Associations between time spent seated (e.g., in car seats or strollers) or in the supine position, and indicators ofadiposity and motor development, were primarily unfavourable or null. Data were scarce for other outcomes.Conclusions: These findings continue to support the importance of minimizing screen time for disease preventiont the potential cognitive benefits of interactive non-between sedentary behaviour and health© The Author(s). 2017 Open Access This articInternational License (http://creativecommonsreproduction in any medium, provided you gthe Creative Commons license, and indicate if(http://creativecommons.org/publicdomain/zele is distributed under the terms of the Creative Commons Attribution 4.0.org/licenses/by/4.0/), which permits unrestricted use, distribution, andive appropriate credit to the original author(s) and the source, provide a link tochanges were made. The Creative Commons Public Domain Dedication waiverro/1.0/) applies to the data made available in this article, unless otherwise stated.may have different health effects. Evidence from largebeing [21], and Carson et al. identified that differentThe Author(s) BMC Public Health 2017, 17(Suppl 5):868 Page 66 of 215BackgroundSedentary behaviour is defined as any waking behaviourwith an energy expenditure of ≤1.5 metabolic equiva-lents (METs) while in a sitting or reclining posture [1].It is increasingly recognized that too much sedentarybehaviour can have negative health effects across thelifespan [2–4], which are distinct from those that resultfrom low physical activity [5]. This may be of particularimportance in the early years of life, given that theseyears are critical for growth and development, and thatlifestyle behaviours established early in life tend to trackover time [6–8].In this regard, the Canadian Sedentary BehaviourGuidelines for the Early Years (ages 0–4 years) [9], andguidelines in other countries around the world (e.g.,Australia [10] and USA [11]), recommend that children<2 years of age have no exposure to screens, and thatthose aged 2 to 4 years have <1 h/day of screen time. Inaddition, guidelines (e.g., in Canada [9], Australia [10],and the United Kingdom [12]) recommend that parentsand caregivers minimize the time that children spendsitting or being restrained (e.g., in a stroller or highchair) while awake.In contrast to these recommendations, ≥ 80% of youngchildren are exposed to screens before the age of 2 years[13, 14], only 22% of Canadian children aged 3 to 4 yearsare meeting the screen time guidelines of <1 h/day, andon average parent-reported screen time for this agegroup is 2.0 h/day [15]. Moreover, young children arespending a substantial proportion of their time seden-tary, and no guidance regarding an “appropriate”amount of total sedentary time exists. This is a notablegap, given that a recent review including data from 10countries reported that children aged 2 to 5 years weresedentary for 34% to 94% of the day [16]. For instance,objectively measured data from a large, nationally repre-sentative sample of Canadian children showed that, onaverage, 3- to 4-year-olds were sedentary for 436 min/day (7 h, 16 min), which was roughly equivalent to 60%of their waking time [15].The Canadian Sedentary Behaviour Guidelines wereinformed by a systematic review of the evidence thatfound that high levels of television (TV) time wereassociated with increased adiposity and reduced psy-chosocial health and cognitive development [2]. How-ever, there was no evidence of benefits or harms forany other type of sedentary behaviour, for total seden-tary time, or for patterns (e.g., frequency, interrup-tions) of sedentary time. This may be in part becauseonly intervention and longitudinal studies wereincluded in this earlier review [2]. This is a criticallimitation because in recent years there has been adramatic shift in the media landscape (e.g., evolvingtechnologies including smartphones and tablets) [17],types of sedentary behaviour may have different effectson cognitive development in the early years of life (e.g.,screen time may be detrimental, and reading beneficial)[18]. These recent reviews present focused summaries;however, no previous review has provided a balancedconsideration of different types of sedentary behaviourand a range of holistic health indicators across studydesigns. Accordingly, a comprehensive review of the lit-erature was needed in order to: 1) understand the healtheffects of sedentary behaviour in the early years, 2)inform and update population-level recommendations,and 3) identify research gaps and guide the design offuture research and/or assist in the translation of currentresearch to practice.Therefore, the purpose of this study was to perform asystematic review that examined the relationshipsbetween sedentary behaviour and health indicators inchildren in their early years (0 to 4 years). An additionalaim was to determine what doses of sedentary behaviour(i.e., duration, patterns [frequency, interruptions], andtype) were associated with health indicators.MethodsProtocol and registrationThis systematic review was registered with the Inter-national Prospective Register of Systematic Reviews(PROSPERO; Registration no. CRD42016035270; availablefrom http://www.crd.york.ac.uk/PROSPERO/display_re-cord.asp?ID=CRD42016035270), and was conducted andreported following the Preferred Reporting Items forSystematic Reviews and Meta-Analyses (PRISMA)statement [22].Eligibility criteriaThe Population, Interventions, Comparisons, Outcomes,and Study design (PICOS) framework [23] was used tocross-sectional studies (with samples representative ofthe general population), together with new studiespublished since the original review, may provideadditional insight.In the intervening years, new systematic reviews havebeen conducted to investigate the relationships betweensedentary behaviour and particular health indicators. Forinstance, Hinkley et al. found that too little evidenceexisted to draw conclusions regarding associationsbetween sedentary behaviours and psychosocial well-and because different types of sedentary behaviour(e.g., reading, sitting, playing video games) [18, 19]and different patterns of sedentary behaviour [20]identify key study concepts in the research question, andto facilitate the search process.The Author(s) BMC Public Health 2017, 17(Suppl 5):868 Page 67 of 215PopulationThe population of interest was apparently healthy chil-dren (i.e., general populations, including those with over-weight and obesity; samples of clinical populations wereineligible) with a mean age of 1 month to 4.99 years (or,if no mean age was reported, samples described as:infants, toddlers, preschoolers, pre-elementary or pre-primary school age) for at least one sedentary behaviourmeasurement point. Subgroups were defined as follows:infants, 1 month to 1 year; toddlers, 1.1 to 3.0 years; andpreschoolers, 3.1 to 4.99 years.Intervention (exposure)The intervention/exposure was a specific measure of seden-tary behaviour (e.g., TV viewing, video gaming, iPad/tablet/touch-screen, smart phone, reading, puzzles, bouts, breaks,sedentary time, and “screen time” – defined as compositemeasures of screen use) obtained via objective (e.g., accel-erometry) or subjective (e.g., proxy-report) methods. Forinfants, sedentary behaviour was operationally defined asany waking behaviour characterized by low energy expend-iture (i.e., non-purposefully active) while restrained (e.g., ina stroller/pram, high chair, car seat/capsule), or when sedate(e.g., lying/sitting in a chair with little movement but not re-strained). Time spent in the prone position (“tummy time”)was not considered sedentary behaviour because this isdeemed “physical activity” in this age group. For toddlersand preschoolers, sedentary behaviour was defined as anywaking behaviour characterized by an energy expenditureof ≤1.5 METs while in a sitting or reclining posture [1].Studies defining sedentary behaviour as “physical inactivity”or “failing to meet physical activity guidelines” wereexcluded, because these definitions do not differentiatebetween sedentary behaviour and light-intensity physicalactivity. Studies of active video gaming exposures (e.g., Nin-tendo Wii™, Microsoft Kinect™, Sony’s Playstation Move™)were excluded because these games may elicit energyexpenditure > 1.5 METs [24], as were studies reportingbackground TV or screen access (e.g., TV is turned on, butnot necessarily being watched by the child) because thechild could be engaged in a non-sedentary behaviour. Forexperimental studies, interventions had to target sedentarybehaviour exclusively and not multiple health behaviours(e.g., both sedentary behaviour and diet).ComparisonVarious durations, patterns (frequencies, interruptions),and types of sedentary behavior were used for compari-son where available. A comparison or control group wasnot required.Outcomes (health indicators)Eight health indicators were chosen by expert consensusamong a 22-member group with expertise in movementbehaviours in children. The health indicators wereselected given consideration of the literature (previousreviews; e.g., [2]) and of the importance of including arange of holistic health indicators (i.e., physical, psycho-logical/social, and cognitive health). Four health indica-tors were identified as critical (primary) by expertconsensus: (1) adiposity (e.g., % body fat, weight status,waist circumference); (2) motor development (e.g., devel-opmental milestones, gross/fine motor skills, locomotor-object control); (3) psychosocial health (e.g., depressive/anxiety symptoms, prosocial behaviour, aggression, self-regulation); and (4) cognitive development (e.g., lan-guage development, attention, executive function). Fourhealth indicators were identified as important (second-ary) by expert consensus: (1) bone and skeletal health(e.g., bone mineral density, bone mineral content,skeletal area); (2) cardiometabolic health (e.g., bloodpressure, insulin resistance, blood lipids); (3) fitness(cardiovascular, musculoskeletal); and (4) risks (injury)/harm (e.g., plagiocephaly, torticollis).Study designsAll study designs were considered. For longitudinal stud-ies, any follow-up length was allowed as long as therewas at least one measure of sedentary behaviourbetween the ages of 1 month to 4.99 years. For logisticreasons, and to maximize generalizability, minimumsample size requirements were imposed [25]; random-ized controlled trials (RCTs) and non-randomized inter-vention studies were required to have at least 15participants in at least one intervention group, andobservational studies were required to have a minimumsample size of 100 participants. Published peer-reviewedoriginal manuscripts and in-press manuscripts, inEnglish or French, were eligible for inclusion. Greyliterature (except for registered clinical trials) and con-ference abstracts were excluded.Information sources and search strategyThe following databases were searched using the Ovidinterface: MEDLINE (1946 to April 13, 2016), EMBASE(1980 to 2016 week 15), PsycINFO (1806 to April Week1 2016), and CENTRAL (February 2016). PubMed wassearched for any additional studies not yet indexed inMEDLINE (April 11, 2016). SPORTdiscus (1949 to April14, 2016) and Communication Source (April 12, 2016)were searched using the EBSCOhost interface, and theCommunications and Mass Media Collection wassearched using Gale. The MEDLINE search strategy wascreated by a research librarian with expertise in system-atic review searching and peer-reviewed by a secondresearch librarian. The search was then adapted forother databases. No study design limits were applied,and searches were limited to English and FrenchThe Author(s) BMC Public Health 2017, 17(Suppl 5):868 Page 68 of 215publications. Updates to all search strategies, limited torandomized controlled trials for logistical reasons, wereperformed on November 1, 2016, to capture anyadditional studies that had been published in the interimbetween the initial searches and the data synthesis. Thesearch strategies are presented in Additional file 1. Trialregistries were also searched (https://clinicaltrials.gov/and http://www.who.int/ictrp/en/; October 11, 2016) forongoing clinical trials, using search terms for the seden-tary behaviour concept and age group of interest. TheInternational Journal of Child-Computer Interaction washand-searched, because this journal was not yet indexedin any of these databases.Bibliographic records were extracted as text files from theOvid, EBSCOHost, and Gale interfaces and imported intoReference Manager Software (Version 11; ThompsonReuters, San Francisco, CA, USA), where duplicate recordswere removed. Titles and abstracts of the remainingrecords were uploaded to DistillerSR (Evidence Partners,Ottawa, ON, Canada), a secure internet-based software,where they were screened against inclusion criteria inde-pendently by two reviewers. Exclusion by both reviewerswas required for a study to be excluded at the title andabstract stage; all other studies passed to full-text articlescreening. Two independent reviewers examined all full-text articles, and consensus was required for article inclu-sion in the review. Discrepancies between reviewers wereresolved by discussion between themselves, or with thelarger review team if needed. Relevant review articlesidentified during screening were also procured, and theirreference lists manually checked for studies potentiallymissed by the search.Data extractionData extraction forms were created by the study coordi-nators, and reviewed and piloted by the review team.Extraction was completed in Microsoft Excel by onereviewer and checked for accuracy by a second reviewer.Reviewers were not blinded to the authors or journalswhen extracting data. Information was extracted regard-ing important study characteristics (e.g., citation, studydesign, country, sample size, age, and sex of partici-pants); exposure (i.e., sedentary behaviour characteristics[e.g., type, volume, duration, frequency, pattern, andmeasurement and/or description of sedentary behaviourintervention]); outcome/health indicators (e.g., measure-ment type); results (e.g., odds ratio, difference in means);and covariates included in the analyses (if applicable;e.g., diet, physical activity). If data were unavailable forextraction (e.g., reported only in a graph, or described as“data not shown”), the authors were contacted. If datawere presented subdivided by sex, the data wereextracted independently for each sex only if data pooledacross sex were unavailable. If analyses were reported forany other subsets of data, results were extracted for onlythe analyses using the full sample. The results fromfinally adjusted models were extracted when studies pre-sented multiple models. Study findings were consideredstatistically significant at p < 0.05.Risk of bias and study quality assessmentThe risk of bias was systematically evaluated in each pri-mary research study using the methods described in theCochrane Handbook [26]. All individual studies wereassessed for the following potential sources of bias:selection bias, performance bias, detection bias, attritionbias, reporting bias, and other sources of bias (seePoitras et al. [25] for details).The quality of evidence for each health indicator byeach type of study design was assessed using theGrading of Recommendations, Assessment, Develop-ment and Evaluation (GRADE) framework [27]. The“quality of evidence” is the level of confidence in theestimate of effect. As such, the higher the quality of theevidence, the greater the confidence in the findings, andthe lower the quality, the more likely it is that futureresearch will change the level of confidence in the esti-mates and change the estimates themselves. Accordingto GRADE, there are four levels of quality (“high”, “mod-erate”, “low”, and “very low”); evidence quality ratingsstart at “high” for randomized studies and at “low” forall other studies. The quality of evidence is downgradedif there are limitations across studies due to serious riskof bias, inconsistency (e.g., unexplained heterogeneity inthe direction of the effect), indirectness (e.g., differencesbetween the population, intervention and/or outcomesin included studies and those of interest, such as a sur-rogate measure instead of a direct measure of an out-come), or imprecision (e.g., wide confidence intervalsthat lead to uncertainty about the true magnitude of theeffect) [28]. If there is no reason to downgrade, the qual-ity of evidence can be upgraded if there is a large effectsize, there is a dose-response gradient, or an effect isdetected in the presence of plausible confounders orother biases that would decrease an apparent treatmenteffect [29].In the present review, the overall quality of evidencefor each study design within each health indicator wasevaluated by two independent reviewers and verified bythe larger review team. The review team decided a priorinot to downgrade for risk of bias if the only potentialsources of bias identified were use of a conveniencesample or lack of exposure/outcome blinding, as in pre-vious movement behaviour systematic reviews [25, 30].Synthesis of resultsMeta-analyses were planned if data were sufficientlyhomogeneous in terms of statistical, clinical, andmethodological characteristics. If meta-analyses were notpossible, qualitative syntheses structured around thehealth indicator and type of sedentary behaviour wereconducted, with all studies weighted equally, and theresults presented narratively. Results were presented in“evidence profile” tables by outcome (health indicator)as per the GRADE framework (see Guyatt et al. [27] fordetails). For the purposes of this review, sedentarybehaviours were grouped into three categories: 1) object-ively measured sedentary time, 2) screen-based sedentarybehaviours, and 3) other sedentary behaviours (e.g.,reading, storytelling).ResultsDescription of studiesA total of 10,830 records were identified in the initialsearches, and an additional 11 were identified by check-ing the reference lists of review articles (Fig. 1). Afterde-duplication, 8915 records remained. In the searchupdate, an additional 106 records were identified (mak-ing a total of 10,936), and 101 of these remained afterde-duplication. No relevant records were identified inthe Trial Registry searches. After screening the 9016titles and abstracts (from the initial and updatedsearches), 334 full-text articles were obtained for furtherreview. Reasons for exclusion were: not in English orFrench language (n = 1), review paper (n = 2), sedentarybehaviour included only as a covariate or outcome andnot as the exposure (n = 2), sedentary behaviour definedas “failing to meet physical activity guidelines” (n = 2),sedentary behaviour exposure included backgroundscreens (n = 3), intervention did not target sedentarybehaviour specifically/exclusively (n = 9), not originalresearch (n = 9), no sedentary behaviour exposure (n =9), sample size (n = 15), did not assess the relationshipbetween sedentary behaviour and a relevant health indi-cator (n = 77), participants were not within appropriateage range (n = 92), and other (n = 17; e.g., comparatorThe Author(s) BMC Public Health 2017, 17(Suppl 5):868 Page 69 of 215Fig. 1 PRISMA flow diagram for the identification, screening, eligibility, anddo not sum to the total number of included studies because more than oninclusion of studies. *Note that the numbers for each health indicatore health indicator was reported in some studiesThe Author(s) BMC Public Health 2017, 17(Suppl 5):868 Page 70 of 215was the same “dose” of sedentary behaviour with differ-ent content, predatory publisher and problems with datasuch as incongruent values in text and tables). Somestudies were excluded for multiple reasons. A total of 96studies (from 73 unique samples) met the inclusion cri-teria (Fig. 1).Detailed findings for the individual 96 studies arepresented in Additional file 2: Tables S1-S7 and summa-rized in Tables 1, 2, 3, 4, 5, 6, 7 and 8. Data across studiesinvolved 195,430 participants (147,752 from 73 uniquesamples), ranging from 103 [31] to 50,589 [32] participants.Participants from one study were not included in this sam-ple size calculation because the sample size for the agegroup of interest was not reported [33]. Studies were con-ducted in 33 different countries, but were most commonlyconducted in the United States (n = 44), Belgium (n = 7),Canada (n = 7), Australia (n = 6), Germany (n = 5), and theNetherlands (n = 5), with four or fewer studies from allother countries (Additional file 2: Tables S1-S7). Theapproximate baseline age ranged from 0.3 to 4.95 years.One study used an experimental design (randomizedcontrolled trial); the remaining 95 studies used observa-tional designs, including case-control (n = 3), longitudinal(n = 25), longitudinal with additional cross-sectional ana-lyses (n = 5), and cross-sectional (n = 62).Quality of evidenceOverall, the quality of evidence ranged from “very low”to “moderate” across study designs and health indicators.The most common reason for downgrading the qualityof evidence was because of a serious risk of bias thatreduced the level of confidence in the observed effects.Common sources of bias included: not accounting forpotentially important confounders or mediating factors(e.g., diet); the use of potentially inappropriate measure-ment tools (e.g., exposure or outcome measures withunknown reliability and/or validity); and an unknownamount of, or reasons for, missing data. The quality ofevidence was not upgraded in any instance. For specificdetails regarding the quality of evidence by study designand health indicator, see Tables 1, 2, 3, 4, 5, 6 and 7.Data synthesisMeta-analyses could not be performed because ofheterogeneity in the sedentary behaviour exposure andhealth indicators (statistical, clinical, and methodo-logical). Instead, narrative syntheses are presented.Unless otherwise stated, results did not differ by sex,age, or specific sub-indicator within the eight healthindicator categories. Within each health indicator, resultsare presented first by study design, then by type of sed-entary behaviour exposure (objectively measured seden-tary time, screen-based sedentary behaviours, and othersedentary behaviours), and finally by sub-indicator (i.e.,specific measures of the eight health indicators). Thereader is referred to the Additional file 2: Tables S1-S7for statistic values and additional details.Critical (primary) health indicatorsAdiposityThe relationships between sedentary behaviour and adipos-ity were examined in 60 studies (see Table 1 andAdditional file 2: Table S1) [31–90]. Study designs were:randomized controlled trial (n = 1) [34], longitudinal (n =13) [33, 45, 54, 81–90], case-control (n = 2) [35, 36], andcross-sectional design or also reported cross-sectional find-ings (n = 47) [31–33, 37–80]. Indicators of adiposity (e.g.,body mass index [BMI]) were measured objectively (e.g.,measured by dual-energy X-ray absorptiometry) or assessedsubjectively (e.g., parent-reported height and weight; seeTable 1 for summary of measures). The quality of evidenceranged from “very low” to “moderate” across study designs(Table 1).In the randomized controlled trial of an interventionto reduce screen time, screen time was significantlylower for preschoolers in the intervention versus controlgroup at 2, 6, and 9 months post-intervention [34]. BMIz-scores were not different between the intervention andcontrol groups at baseline or 9-month follow-up, butBMI z-scores increased in both groups [34](Additional file 2: Table S1).Among the 13 longitudinal studies, sedentary behaviourwas assessed from age ~9 months to 4.95 years as screen-based (i.e., computer time, frequency of playing computergames, time watching DVDs, TV time, and total screentime) or other sedentary behaviours (i.e., time spent in thecar or in baby seats). Adiposity indicators were assessedbetween ~1.25 and 12 years follow-up.For screen-based sedentary behaviours, computer time[85], and frequency of playing computer games [82] at age4.8 years were not associated with total fat mass or leanmass, or weight status, at ~6 and 12 years of follow-uprespectively. Time watching DVDs at ages ~3–4 years wasunfavourably associated with weight status at kindergartenentry [83]. Total screen time in toddlers was unfavourablyassociated with weight status at preschool or school age in2/3 studies [33, 84]. In the third study, total screen timewas not associated with weight status [87].Ten longitudinal studies examined the relationshipsbetween TV time (at ages ranging from ~6 months to4.8 years) and adiposity indicators at ~1.5 to 12 years offollow-up. Of these, unfavourable associations werereported in 6/10 studies [33, 54, 81, 83, 88, 90], nullassociations in 1/10 studies [86], and mixed unfavour-able and null associations in 3/10 studies [82, 85, 89].Specifically, TV time was prospectively unfavourablyassociated with these adiposity indicators: BMI z-scorein 1/1 studies [88], BMI in 2/3 studies [54, 90], % changeTable1TherelationshipbetweensedentarybehaviourandadiposityNo.ofparticipants(No.ofstudies)DesignQualityassessmentAbsoluteeffectQualityRiskofbiasInconsistencyIndirectnessImprecisionOtherTherangeofmeanagesattimeofexposuremeasurementwas~0.75to4.95years;theoldestmeanageatfollow-upwas15.5years.Datawerecollectedbyrandomizedtrial,case-control,cross-sectionally,andupto12yearsoffollow-up.Adipositymeasureswere:BMI(absolute,z-score,SDscore,percentile);fatmassindex,leanmassindex,trunkfatmassindex;%bodyfat(measuredusingDXA);skinfoldratio(tricepsskinfoldthicknesstosubscapularskinfoldthickness);sumofskinfolds;waist-to-heightratio;waist-to-hipratio;weight-for-height(z-score);weight-for-age(z-score);waistcircumference(absolute,z-scoreforage);weightstatus(CDC,IOTF,orWHOcut-points;Flemishreferencedata;Frenchreferencestandards;RollandCacherareferencecurves;UnitedKingdomreferencestandardsin1999);totalfatmass(SDscore);leanmass(SDscore).412(1)RandomizedtrialaSeriousriskofbiasbNoseriousinconsistencyNoseriousindirectnessNoseriousimprecisionNoneScreentimecwassignificantlylowerintheinterventionvscontrolgroupat2,6,and9monthspost-interventiond.BMIz-scoreswerenotdifferentbetweentheinterventionandcontrolgroupsatbaselineor9-monthfollow-up,butBMIz-scoresincreasedinbothgroups[34].Moderatee32,699(13)LongitudinalfSeriousriskofbiasgNoseriousinconsistencyNoseriousindirectnessNoseriousimprecisionNoneScreen-basedsedentarybehaviours:Computer(duration):1/1studiesreportednullassociations[85]Computergames(frequency):1/1studiesreportednullassociations[82]Screentime(duration):2/3studiesreportedunfavourableassociations[33,84]1/3studiesreportednullassociations[87]TVtime(duration):6/10studiesreportedunfavourableassociations[33,54,81,83,88,90]1/10studiesreportednullassociations[86]3/10studiesreportedmixedunfavourableandnullassociations[82,85,89]WatchingDVDs(duration):1/1studiesreportedunfavourableassociations[83]Othersedentarybehaviours:Timeinbabyseats(duration):1/1studiesreportedmixedunfavourable,null,andfavourableassociations[45]Timeinthecar(duration):2/2studiesreportednullassociations[81,82]Verylowh1242(2)Case-controliSeriousriskofbiasjNoseriousinconsistencyNoseriousindirectnessNoseriousimprecisionNoneTVtime[35,36]andtotalsedentarytime[36]werenotdifferentbetweenchildrenwithoverweight/obese(casegroup)ornormalweight(controlgroup)status,butwatchingTVfor≥1h/daywasunfavourablyassociatedwithhavingoverweightstatus(OR=1.71,95%CI:1.07,2.75,p=0.02)[35].Verylowk94,191(47)Cross-sectionallSeriousriskofbiasmNoseriousinconsistencyNoseriousindirectnessNoseriousimprecisionNoneObjectivelymeasuredsedentarytime:Sedentarytime30-minbouts(accelerometerderived):1/1studiesreportednullassociations[40]Totalsedentarytime(accelerometer-derived):10/11studiesreportednullassociations[37–40,47,53,60,75,78,80]1/11studiesreportedmixedunfavourableandnullassociations[77]Screen-basedsedentarybehaviours:Computer(duration):3/4studiesreportednullassociations[63,67,79]1/4studiesreportedmixedunfavourableandnullassociations[71]Screentime(duration):6/18studiesreportedunfavourableassociations[32,33,46,50,59,73]10/18studiesreportednullassociations[44,52,57,58,62,64,65,71,72,79]2/18studiesreportedmixedunfavourableandnullassociations[41,61]TVtime(duration):5/23studiesreportedunfavourableassociations[33,55,66,67,71]11/23studiesreportednullassociations[31,42,43,49,50,56,60,63,69,75,76]5/23studiesreportedmixedunfavourableandnullassociations[48,51,54,68,79]1/23studiesreportedmixednullandfavourableassociations[74]1/23studiesreportedmixedunfavourable,null,andfavourableassociations[70]VerylownThe Author(s) BMC Public Health 2017, 17(Suppl 5):868 Page 71 of 215Table1Therelationshipbetweensedentarybehaviourandadiposity(Continued)No.ofparticipants(No.ofstudies)DesignQualityassessmentAbsoluteeffectQualityRiskofbiasInconsistencyIndirectnessImprecisionOtherUsingtheinternet(duration):1/1studiesreportednullassociations[69]Videogames(duration):1/1studiesreportedunfavourableassociations[69]WatchingDVDs/videos(duration):1/1studiesreportednullassociations[69]Othersedentarybehaviours:Sedentaryquietplay(duration):1/1studiesreportedmixedunfavourableandnullassociations[79]Timeinbabyseats(duration):1/1studiesreportednullassociations[45]Usingbooks(duration):1/1studiesreportednullassociations[69]BMIBodyMassIndex,CDCCentersforDiseaseControlandPrevention,DXAdual-energyX-rayabsorptiometry,IOTFInternationalObesityTaskForce,SDstandarddeviation,WHOWorldHealthOrganizationa Includes1randomizedcontrolledtrial[34]bSeriousriskofbias.Unclearifallocationwasadequatelyconcealedpriortogroupassignment;groupallocationwasadequatelyconcealedfromcontrol,butnotinterventiongroupduringthestudy;unclearifheightandweightweredirectlymeasuredorproxy-reported;baselinedatawerenotreported,makingitimpossibletodetermineifbaselineimbalancesexistedbetweengroups[34]c Screentimewassignificantlylowerintheinterventionvscontrolgroupat2mo,6mo,and9mofollow-uppost-intervention(mean±SD:2mo:39.48±16.36vs86.64±21.63min/day;6mo:24.72±4.45vs84.95±14.77min/day;9mo:21.15±6.12vs93.96±18.84min/day;allp<0.001)dIntervention:3printedmaterialsandinteractiveCDsandonecounsellingcallintendedtodecreasescreentime;8-weekduration.Control:Usualcare;unawareofcounsellinginterventionse Thequalityofevidencefromtherandomizedtrialwasdowngradedfrom“high”to“moderate”becauseofaseriousriskofbiasthatdiminishedthelevelofconfidenceintheobservedeffectsf Includes13longitudinalstudies[33,45,54,81–90]from9uniquesamples.Paganietal.[90]andFitzpatricketal.[89]reporteddatafromtheQuebecLongitudinalStudyofChildDevelopment;Reillyetal.[81]andLearyetal.[82]reporteddatafromtheAvonLongitudinalStudyofParentsandChildren(ALSPAC);Goozeetal.[84]andFloresandLin[83]reporteddatafromtheEarlyChildhoodLongitudinalStudy-BirthCohort(ECLS-B);andFuller-Tyszkiewiczetal.[54]andWheatonetal.[85]reporteddatafromtheLongitudinalStudyofAustralianChildren(LSAC).ResultsarepresentedseparatelyandparticipantsarecountedonlyoncegSeriousriskofbias.Questionablevalidityandreliabilityoftheexposuremeasure[33,45,54,81–90].Datawerereportedasmissing,butamountandreasonswerenotprovided[89].Heightandweightdatawereincompletewithoutexplanationfor23%oftheanalyzedsampleand60.7%oftheoriginalcohort[81].Possibleselectivereporting:differencesbetweenincludedandexcludedparticipantswerereportedforconfoundingvariablesbutnotexposurevariableswithoutexplanation[82].BMIatage3yrwasanalyzed,butwasnotreportedinthepurposeormethods[88].Didnotaccountforpotentiallyimportantconfoundingvariablesormediatingfactors:sugar-sweetenedbeverageconsumptionandsleepwereassessedbutnotaccountedfor[33];dietwasnotmeasuredorincludedintheanalysis[45];adjustedforphysicalactivity[89];ofthepotentialchildandfamilyconfoundersthatwereassessed,potentialconfounderswereincludedoromittedfromanalysesbasedontheauthors’determinationofwhatwas“likelytobelinkedtoourpredictororoutcomevariables,”withoutprovidingabasisforthatdetermination[89].Datawerepooledfromthecontrolandexperimentalgroupsofamessaging-basedobesitypreventioninterventionstudy[33]hThequalityofevidencefromthelongitudinalstudieswasdowngradedfrom“low”to“verylow”becauseofaseriousriskofbiasthatdiminishedthelevelofconfidenceintheobservedeffectsi Includes2case-controlstudies[35,36]j Seriousriskofbias.Questionablevalidityandreliabilityofthe1-dayphysicalactivityrecallquestionnaire[36].Potentiallyinappropriatestatisticalanalysis:investigatorsdichotomizedparticipantsbycategoryofTVviewingof≥1h/dayor<1h/daybasedonexploratorybivariateanalysesthatshowed1htobethedurationmostrelatedtochildren’sweightstatus[35]k Thequalityofevidencefromthecase-controlstudieswasdowngradedfrom“low”to“verylow”becauseofaseriousriskofbiasthatdiminishedthelevelofconfidenceintheobservedeffectsl Includes47cross-sectionalstudies[31–33,37–80]from40uniquesamples.Williamsetal.[37],Byunetal.[39],andByunetal.[38]reporteddatafromtheChildren’sActivityandMovementinPreschoolStudy(CHAMPS);Sijtsmaetal.[45]andSijtsmaetal.[46]reporteddatafromtheGroningenExpertCenterforKidswithObesity(GECKO)Drenthebirthcohort;Maniosetal.[48],Kourlabaetal.[49],andvanStralenetal.[50]reporteddatafromtheGrowth,ExerciseandNutritionEpidemiologicalStudyinpreSchoolers(GENESIS);Mendozaetal.[71]reporteddatafromtheNationalHealthandNutritionExaminationSurvey(NHANES)1999to2002,Fultonetal.[72]fromNHANES1999to2006,andTwarogetal.[73]fromNHANES2008to2012;TavernoRossetal.[76]andEspana-Romeroetal.[77]reporteddatafromtheStudyofHealthandActivityinPreschoolEnvironments(SHAPES);Brownetal.[55]andFuller-Tyszkiewiczetal.[54]reporteddatafromtheLongitudinalStudyofAustralianChildren(LSAC);Dolinskyetal.[53]andBolingTureretal.[45]reporteddatafromKidsandAdultsNow:DefeatObesity!(KAN-DO).ResultsarepresentedseparatelyandparticipantsarecountedonlyoncemSeriousriskofbias.Potentiallyinappropriatesamplingtechnique:participantswereanon-representativeconveniencesample[66];samplingdeviatedfromprotocolandspecificdeviationswerenotdocumented[57].Potentiallyinappropriatemeasurementtoolswereused:questionablevalidityandreliabilityoftheexposuremeasure[31–33,41,43–46,49–51,54–62,64–76]andoutcomemeasure[65];questionablevalidityofexposuremeasure[42,52,63,79];poorreliabilityofexposuremeasure[42];heightandweightwereobtainedbyparent-report[44,70];optionsfor2–3hand4–5hweremissingfromtheLikert-typescaleusedtoassessscreentime[74];appliedaccelerometrycut-pointswerenotvalidatedfortheagegroupofinterest[47].Potentialattritionbias:amountofunexplainedmissingexposureoroutcomedataisunknown[42,50]orrangedfrom14%to67%[39,40,42,43,59,60,69,71,73,74,76],andreasonmayberelatedtothetrueoutcomeofinterest[40,43,66,71].Potentialselectivereportingbias:statisticsfornon-significantrelationshipswerenotreported[48,64];authorsdecidedpost-hocnottoreportanalyseswithcontinuousexposurevariables[59];onlyfinalmodelwasreported[44];resultsforcorrelationsdescribedinthemethodssectionwerenotreported[62];compositeoutcomeswerepresentedwithoutindividualcomponents;resultsforcategoricalscreentimeandtotalscreentimedescribedinthemethodssectionwerenotreported[32];outcomesfrompooledhierarchicallinearregressionandvarianceinformationofincludedresultswerenotreported[70].Didnotaccountforpotentiallyimportantconfoundingvariablesormediatingfactors:diet[43,45,46,50,58,60,63,64,67,71,72,77,80];sugar-sweetenedbeverageconsumption;andsleep[33].Controlledforphysicalactivity[59,61,66,78].Sleepduringthedaywasconsideredsedentarytime[40]nThequalityofevidencefromthecross-sectionalstudieswasdowngradedfrom“low”to“verylow”becauseofaseriousriskofbiasthatdiminishedthelevelofconfidenceintheobservedeffectsThe Author(s) BMC Public Health 2017, 17(Suppl 5):868 Page 72 of 215Table2TherelationshipbetweensedentarybehaviourandmotordevelopmentNo.ofparticipants(No.ofstudies)DesignQualityassessmentAbsoluteeffectQualityRiskofbiasInconsistencyIndirectnessImprecisionOtherParticipantagesattimeofexposuremeasurementrangedfrom~4mo(0.3yr)to3–4years;theoldestmeanageatfollow-upwas5.4years.Datawerecollectedcross-sectionallyandupto3yearsoffollow-up.Motordevelopmentindicatorswereassessedbyparent-reportunlessotherwiseindicated;specificindicatorswere:ageatfirstsitting,ageatfirstcrawling,ageatfirstwalking,locomotion/locomotorskills(assessedbya“testofgrossmotordevelopment”orCHAMPSMotorSkillProtocol),motorskilldevelopment(assessedbythePDMS-2orCHAMPSMotorSkillProtocol),motorskills(assessedbya“neurologicaloptimalityscore”),objectcontrol(assessedbya“testofgrossmotordevelopment”,orCHAMPSMotorSkillProtocol),andvisual-motorabilities(assessedbytheWRAVMAtest).3413(3)LongitudinalaSeriousriskofbiasbNoseriousinconsistencyNoseriousindirectnessNoseriousimprecisionNoneScreen-basedsedentarybehaviours:TVtime(duration):2/3studiesreportednullassociations[88,91]1/3studiesreportedmixedunfavourableandnullassociations[92]Othersedentarybehaviours:Timeinababycarrier/sling(duration):1/1studiesreportednullassociations[91]Timeinacarseat(duration):1/1studiesreportedmixednullandfavourableassociations[91]Timeinahighchairorotherchair(duration):1/1studiesreportednullassociations[91]Timeinaplaypen(duration):1/1studiesreportednullassociations[91]Timeinastroller(duration):1/1studiesreportednullassociations[91]Verylowc681(4)Cross-sectionaldSeriousriskofbiaseNoseriousinconsistencyNoseriousindirectnessNoseriousimprecisionNoneObjectivelymeasuredsedentarytime:Sedentarytime30-minbouts(accelerometer-derived):1/1studiesreportednullassociations[40]Totalsedentarytime(accelerometer-derived):1/2studiesreportednullassociations[40]1/2studiesreportedmixedunfavourableandnullassociations[37]Screen-basedsedentarybehaviours:TVtime(duration):1/1studiesreportedunfavourableassociations[94]Othersedentarybehaviours:Timeinsupineposition(duration):1/1studiesreportedmixedunfavourableandnullassociations[93]VerylowfCHAMPSChildren’sActivityandMovementinPreschoolStudy,PDMS-2PeabodyDevelopmentalMotorScales–secondedition,WRAVMAWide-RangeAssessmentofVisualMotorAbilitya Includes3longitudinalstudies[88,91,92]from3uniquesamplesbSeriousriskofbias.Questionablevalidityandreliabilityofexposuremeasure[88,91,92]c Thequalityofevidencefromlongitudinalstudieswasdowngradedfrom“low”to“verylow”becauseofaseriousriskofbiasthatdiminishedthelevelofconfidenceintheobservedeffectsdIncludes4cross-sectionalstudies[37,40,93,94]from4uniquesamplese Seriousriskofbias.Questionablevalidityandreliabilityofexposuremeasure[93,94];largeamount(30.9%)ofunexplainedmissingdataandpatternofnonresponseindicatesreasonformissingdatamayhavebeenrelatedtotheoutcomeofinterest[40];sleepduringthedaywasincludedinsedentarytimeexposure[40]f Thequalityofevidencefromcross-sectionalstudieswasdowngradedfrom“low”to“verylow”becauseofaseriousriskofbiasthatdiminishedthelevelofconfidenceintheobservedeffectsThe Author(s) BMC Public Health 2017, 17(Suppl 5):868 Page 73 of 215Table3TherelationshipbetweensedentarybehaviourandpsychosocialhealthNo.ofparticipants(No.ofstudies)DesignQualityassessmentAbsoluteeffectQualityRiskofbiasInconsistencyIndirectnessImprecisionOtherTherangeofmeanagesattimeofexposuremeasurementwas~1to4.3years;theoldestmeanageatfollow-upwas~12years.Datawerecollectedbyrandomizedtrial,cross-sectionally,andupto9.5yearsoffollow-up.Psychosocialhealthmeasureswere:aggressiontowardasibling(assessedbytheAggressiveSiblingSocialBehaviorScale);aggressivebehaviours/aggression,delinquentbehaviours,totalbehaviourproblems,externalizingproblems,internalizingproblems,emotionalreactivity,anxiousordepressedsymptoms,andattentionproblems(assessedbytheCBCLorJapaneseCBCL);attentionalproblems(assessedbythehyperactivitysubscaleoftheBPI);attentionproblemsandhyperactivity(assessedbytheBASC-2);bullying(assessedbyunpublishedquestionnaire);co-operation,assertion,responsibility,self-control,andtotalsocialskills(assessedbytheSocialSkillsRatingSystem);emotionalsymptoms/problems,conductproblems,hyperactivity-inattention,peerproblems,andprosocialbehaviour(assessedusingtheSDQ);self-esteem,emotionalwell-being,familyfunctioning,andsocialnetworks(assessedusingtheKINDLR);social-emotionalcompetence(assessedbytheMIT-SEA);soothability,sociability,andemotionality(assessedbytheCTQ);victimization,anxiety,physicalaggression,andprosocialbehaviour(assessedbytheSBQ);andriskofbeingabully,victim,orbully-victim(assessedbyunpublishedquestionnaire).412(1)RandomizedtrialaSeriousriskofbiasbNoseriousinconsistencyNoseriousindirectnessNoseriousimprecisionNoneScreentimecwassignificantlylowerintheinterventionvscontrolgroupat2,6,and9monthspost-interventiond.Aggressiveanddelinquentbehaviourswerenotdifferentbetweentheinterventionandcontrolgroupsatbaseline,butweresignificantlylowerintheinterventionvscontrolgroupat9-monthspost-intervention[34].Moderatee13,301(9)LongitudinalfSeriousriskofbiasgNoseriousinconsistencyNoseriousindirectnessNoseriousimprecisionNoneScreen-basedsedentarybehaviours:Timee-gamingoronacomputer(duration):1/1studiesreportedmixedunfavourableandnullassociations[96]TVtime(duration):2/9studiesreportedunfavourableassociations[95,103]5/9studiesreportedmixedunfavourableandnullassociations[90,92,96,97,99]1/9studiesreportednullassociations[100]1/9studiesreportedmixednullandfavourableassociations[102]Verylowh9429(7)Cross-sectionaliSeriousriskofbiasjNoseriousinconsistencyNoseriousindirectnessNoseriousimprecisionNoneObjectivelymeasuredsedentarytime:Totalsedentarytime(accelerometer-derived):1/1studiesreportednullassociations[104]Screen-basedsedentarybehaviours:TVtime(duration):2/6studiesreportedunfavourableassociations[101,103]2/6studiesreportednullassociations[100,106]1/6studiesreportedmixedunfavourableandnullassociations[105]1/6studiesreportedmixednullandfavourableassociations[98]VerylowkBASC-2BehaviorAssessmentSystemforChildren,BPIBehaviorProblemsIndex,CBCLChildBehaviorChecklist,CTQChildTemperamentQuestionnaire,KINDLRQuestionnaireforMeasuringHealth-RelatedQualityofLifeinChildrenandAdolescents-RevisedVersion,MIT-SEAModifiedInfant-ToddlerSocialandEmotionalAssessment,SBQSocialBehaviorQuestionnaire,SDQStrengthsandDifficultiesQuestionnairea Includes1randomizedcontrolledtrial[34]bSeriousriskofbias.Unclearifallocationwasadequatelyconcealedpriortogroupassignment;groupallocationwasadequatelyconcealedfromcontrol,butnotinterventiongroupduringthestudy;knowledgeofoutcomeofinterestwasnotpreventedandoutcomemeasurementislikelytohavebeeninfluencedbylackofblinding;baselinedatawerenotreported,makingitimpossibletodetermineifbaselineimbalancesexistedbetweengroups[34]c Screentimewassignificantlylowerintheinterventionvscontrolgroupat2-,6-,and9-monthfollow-uppost-intervention(mean±SD:2month:39.48±16.36vs86.64±21.63min/day;6month:24.72±4.45vs84.95±14.77min/day;9month:21.15±6.12vs93.96±18.84min/day;allp<0.001)dIntervention:3printedmaterialsandinteractiveCDsandonecounsellingcall,intendingtodecreasescreentime;8-weekduration.Control:Usualcare;unawareofcounsellinginterventionseThequalityofevidencefromtherandomizedtrialwasdowngradedfrom“high”to“moderate”becauseofaseriousriskofbiasinthesinglerandomizedcontrolledtrialthatdiminishedthelevelofconfidenceintheobservedeffectsf Includes9longitudinalstudies[90,92,95–97,99,100,102,103]from6uniquesamples.Verlindenetal.[97,99]reporteddatafromtheGenerationRStudy;andPaganietal.[90,92]andWattetal.[95]reporteddatafromtheQuebecLongitudinalStudyofChildDevelopment(QLSCD).ResultsarepresentedseparatelyandparticipantsarecountedonlyoncegSeriousriskofbias.Questionablevalidityandreliabilityoftelevisiondurationexposuremeasure[90,92,97,99,100,102,103];questionablevalidityandreliabilityoftelevisiondurationexposuremeasureonweekdaysonly[96];poorreliabilityofoutcomemeasuresforresponsibility[102]andemotionalsymptoms,conductproblems,peerproblems,andprosocialbehaviour[100];largeamountofunexplainedmissingdataandpatternofnonresponseindicatesreasonformissingdatamayhavebeenrelatedtotheoutcomeofinterest[97];completeresultswerenotreportedforallrelationshipsexamined[99]hThequalityofevidencefromlongitudinalstudieswasdowngradedfrom“low”to“verylow”becauseofaseriousriskofbiasthatdiminishedthelevelofconfidenceintheobservedeffectsi Includes7cross-sectionalstudies[98,100,101,103–106]from7uniquesamplesj Seriousriskofbias.Questionablevalidityandreliabilityoftelevisiondurationexposuremeasure[98,100,101,103,105,106];poorreliabilityofoutcomemeasuresforemotionalsymptoms,conductproblems,peerproblems,andprosocialbehaviour[100];smallamount(218/4020)ofunexplainedmissingoutcomedataat3-yearfollow-up[92]k Thequalityofevidencefromcross-sectionalstudieswasdowngradedfrom“low”to“verylow”becauseofaseriousriskofbiasthatdiminishedthelevelofconfidenceintheobservedeffectsThe Author(s) BMC Public Health 2017, 17(Suppl 5):868 Page 74 of 215Table4TherelationshipbetweensedentarybehaviourandcognitivedevelopmentNo.ofparticipants(No.ofstudies)DesignQualityassessmentAbsoluteeffectQualityRiskofbiasInconsistencyIndirectnessImprecisionOtherTherangeofmeanagesattimeofexposuremeasurementwas~0.5to4.4years;theoldestagerangeatfollow-upwas9to10years.Datawerecollectedcross-sectionallyandupto8yearsoffollow-up.Cognitivedevelopmentindicatorswere:ADHDsymptoms(assessedbychecklistsbasedontheDSM-IV);attentionalproblems(assessedbytheBPI);attentionspan(assessedbytheCTQ);classroomengagement(assessedbyaClassroomEngagementScaleandanunpublishedquestionnaire);cognitiveability(assessedbytheImitationSortingTask);cognitivedevelopment(assessedbyBSID-IIandBSID-III);cognitiveinhibitorycontrol(assessedbytheAnimalStroopTask);executivefunction(assessedasacompositeofcognitiveinhibitorycontrolandworkingmemorycapacity;theBASC-2;fourtasks:grass/snow,whisper,backwarddigitspan,tower);languagedevelopment(total),auditorycomprehension,expressivecommunication(assessedbyASQ,PLS-4,CELF-P2,CELF-4,CDI,K-ASQ,ThaiCLAMS,medicaldiagnosis,anddevelopmentalassessmentwithDenver-IItest);mathematicalsuccess(assessedasrelativetotheclassdistribution);mathematics,readingrecognition,readingcomprehension(assessedbythePIAT);numberknowledge(assessedbyNKT);receptiveandtotalvocabulary(assessedbyPPVT);short-termmemory(assessedbytheMemoryforDigitSpanoftheWISC);speechdisorders(assessedbytheChuturiktestandChildBehaviorChecklistbyAchenbach,conversationwithparents,andclinicalexamination);andworkingmemorycapacity(assessedusingtheAnimalStroopTaskandK-ABCnumberrecalltest).8927(11)LongitudinalaSeriousriskofbiasbNoseriousinconsistencyNoseriousindirectnessNoseriousimprecisionNoneScreen-basedsedentarybehaviours:Electronicmediaexposure(duration):1/1studiesreportedunfavourableassociations[112]Othersedentarybehaviours:Parentsreading(frequency):1/1studiesreportedfavourableassociations[121]TVtime(duration):5/10studiesreportedunfavourableassociations[90,92,100,120,121]4/10studiesreportednullassociations[88,102,113,122]1/10studiesreportedmixedunfavourable,null,andfavourableassociations[119]Verylowc166(1)Case-controldSeriousriskofbiaseNoseriousinconsistencyNoseriousindirectnessNoseriousimprecisionNoneScreen-basedsedentarybehaviours:TVtime:1/1studiesreportedunfavourableassociations[116]Verylowf9330(16)Cross-sectionalgSeriousriskofbiashNoseriousinconsistencyNoseriousindirectnessNoseriousimprecisionNoneObjectivelymeasuredsedentarytime:Totalsedentarytime(accelerometer-derived):1/1studiesreportednullassociations[104]Screen-basedsedentarybehaviours:Computeruse(yes,no):1/1studiesreportednullassociations[109]Mobilephoneuse(yes,no):1/1studiesreportedunfavourableassociations[109]TVtime(duration):3/9studiesreportedunfavourableassociations[94,108,123]4/9studiesreportednullassociations[90,100,114,115,121]1/9studiesreportedmixedunfavourableandnullassociations[118]Totalmediaexposure(duration):1/1studiesreportedmixednullandunfavourableassociations[124]Videogames(duration):1/1studiesreportednullassociations[107]VerylowiThe Author(s) BMC Public Health 2017, 17(Suppl 5):868 Page 75 of 215Table4Therelationshipbetweensedentarybehaviourandcognitivedevelopment(Continued)No.ofparticipants(No.ofstudies)DesignQualityassessmentAbsoluteeffectQualityRiskofbiasInconsistencyIndirectnessImprecisionOtherOthersedentarybehaviours:Readingwithparents(duration,frequency):1/3studiesreportednullassociations[110]1/3studiesreportedfavourableassociations[117]1/3studiesreportedmixednullandfavourableassociations[124]Screentime(duration):1/1studiesreportedunfavourableassociations[111]Storytellingwithparents(frequency):2/2studiesreportedmixednullandfavourableassociations[117,124]ADHDAttention-Deficit/HyperactivityDisorder,ASQAgesandStagesQuestionnaire,BASC-2BehaviorAssessmentSystemforChildren,BSID-IIandBSID-IIIBayleyScalesofInfantDevelopment–secondandthirdeditions,BPIBehavioralProblemsIndex,CDICommunicativeDevelopmentInventory,CELF-P2ClinicalEvaluationofLanguageFundamentals–Preschool,CELF-4ClinicalEvaluationofLanguageFundamentalsFourthEdition,CLAMSClinicalLinguisticAuditoryMilestoneScale,CTQChildTemperamentQuestionnaire,DSM-IVDiagnosticandStatisticalManualofMentalDisorders–4,K-ABCKaufmanAssessmentBatteryforChildren,K-ASQKorean–AgesandStagesQuestionnaire,NKTNumberKnowledgeTest,PIATPeabodyIndividualAchievementTest,PLS-4PreschoolLanguageScale–4,PPVTPeabodyPictureVocabularyTest,WISCWechslerIntelligenceScaleforChildrena Includes11longitudinalstudies[88,90,92,100,102,112,113,119–122]from8uniquesamples.Tomopoulosetal.[112]reporteddatafromtheBellevueProjectforEarlyLanguage,Literacy,andEducationSuccess(BELLE);McKeanetal.[121]reporteddatafromtheEarlyLanguageinVictoriaStudy(ELVS);Paganietal.[90,92]reporteddatafromtheQuebecLongitudinalStudyofChildDevelopment(QLSCD);Schmidtetal.[88]reporteddatafromProjectViva;andFosterandWatkins[113],Christakisetal.[120]andZimmermanandChristakis[119]reporteddatafromtheNationalLongitudinalSurveyofYouth,Children,andYoungAdults(NLSY-Child).ResultsarepresentedseparatelyandparticipantsarecountedonlyoncebSeriousriskofbias.Questionablevalidityandreliabilityoftelevisiondurationexposuremeasureinallstudies[88,90,92,100,102,112,113,119–122];poorreliabilityofAttentionProblemssubscaleoftheChildBehaviorChecklist(ɑ=0.59)[102];possiblereportingbias,becausetherelationshipbetweenTVexposureandBMIatage3yrwasanalyzeddespitenotbeingdescribedinthemethodssection[88];twostudieshadunexplainedmissingdata(34%and40%missing)andthepatternofnonresponseindicatesthereasonformissingdatamayhavebeenrelatedtotheoutcomeofinterest[112,121];datawerereportedincompletelyfortherelationshipbetweenTVexposureandreadingachievement[90];themethodssectionofonestudyindicatedthatbivariateanalysiswouldbeperformed,butincludedvariablesandtheresultsoftheanalysiswerenotreported[121]c Thequalityofevidencefromlongitudinalstudieswasdowngradedfrom“low”to“verylow”becauseofaseriousriskofbiasthatdiminishedthelevelofconfidenceintheobservedeffectsdIncludes1case-controlstudy[116]e Seriousriskofbias.Exposuremeasurewasdescribedinpoordetail;questionablevalidityandreliabilityoftelevisiondurationexposuremeasure;theDenverIIScaleisusefulfordetectingseveredevelopmentalproblemsbuthasbeencriticizedasbeingunreliableforpredictinglesssevereorspecificproblems;theregressionmodelthatpredicteddevelopmentaldelayfromacompositeof“ageofonsetofTVviewing”and“TVviewing>2h/day”wasnotpre-specifiedinthemethods,andcompositevariableswerenotcombinedinanalyseswithotheroutcomes[116]f Thequalityofevidencefromthecase-controlstudywasdowngradedfrom“low”to“verylow”becauseofaseriousriskofbiasthatdiminishedthelevelofconfidenceintheobservedeffectsgIncludes16cross-sectionalstudies[90,94,100,104,107–111,114,115,117,118,121,123,124].Zimmermanetal.[117]andFergusonandDonnellan[124]reporteddatafromthesamesample.ResultsarepresentedseparatelyandparticipantsarecountedonlyoncehSeriousriskofbias.Potentiallyinappropriatesamplingtechniqueresultedinasamplewithhigherincomeandeducationthantheoverallpopulationfromwhichitwasrecruited[117,124];questionablevalidityandreliabilityoftheexposuremeasure[90,106–109,111,115,117,121–124];questionablevalidityofexposuremeasure[94];validationstudyshowedoverestimationofTVtimeexposuremeasure[110];questionablevalidityand/orreliabilityoftheoutcomemeasure[109,110];unknownamount[109,117]orbetween28%and60%[121,124]ofunexplainedmissingdataandpatternofnonresponseindicatesreasonformissingdatamayhavebeenrelatedtotheoutcomeofinterest;incompletereportingofexposure[109]andoutcome[90,110];longitudinalrelationshipswerereportedlycollectedbutnotreportedintheresults[115];themethodssectionofonestudyindicatedthatbivariateanalysiswouldbeperformed,butincludedvariablesandtheresultsoftheanalysiswerenotreported[121]i Thequalityofevidencefromlongitudinalstudieswasdowngradedfrom“low”to“verylow”becauseofaseriousriskofbiasthatdiminishedthelevelofconfidenceintheobservedeffectsThe Author(s) BMC Public Health 2017, 17(Suppl 5):868 Page 76 of 215Table 5 The relationship between sedentary behaviour and bone and skeletal healthNo. of participants(No. of studies)Design Quality assessment Absolute effect QualityRisk of bias Inconsistency Indirectness ImprecisionThe mean age was 4.4 years. Data were collected cross-sectionally. Bone and skeletal health were assessed objectively using quantitative ultrasound.1512 (1) Cross-sectionala Serious riskof biasbNo seriousinconsistencyNo seriousindirectnessSeriousimprecisioncObjectively measured sedentary time:After adjusting for MVPA, accelerometer-derived sedentarytime was no longer significantly associated with bonestiffness index (SI) in preschool children (β = -0.37;R2 = 19%; p = 0.28) [125].Very lowdprest5]s silowThe Author(s) BMC Public Health 2017, 17(Suppl 5):868 Page 77 of 215in BMI and % change in waist-to-height ratio in 1/1studies [33], fat mass in 1/1 studies [82], and weight sta-tus in 2/2 studies [81, 83] (Additional file 2: Table S1).TV time at age ~3 years was not associated with the rateof weight gain from ages 3 to 5 years [86]. TV time atage 2.4 years was not associated with waist circumfer-ence at age 10.15 years, but the change in TV time fromages 2.4 to 4.4 years was unfavourably associated withwaist circumference at age 10.15 years [89]. TV time atage 3.2 years was unfavourably associated with fat massat age 15 years.Regarding other sedentary behaviours, types of sittingwere examined in three longitudinal studies. Amongpreschoolers, time in the car was not prospectively asso-ciated with adiposity indicators in 2/2 studies [82, 85];however, among infants there were mixed unfavourable,null, and favourable associations between time in babyseats and adiposity indicators [45]. Specifically, time inMVPA moderate-to-vigorous physical activity, SI bone stiffness indexaIncludes 1 cross-sectional study that reported data from the Identification and(IDEFICS) sample [125]bSerious risk of bias. Study participants were selected by “judgment sample”; quand of quantitative ultrasound for measurement of bone stiffness in children [12cSerious imprecision. It was not possible to estimate the precision of the findingdThe quality of evidence from the cross-sectional study was downgraded from “confidence in the observed effects, and (2) serious imprecisionbaby seats at age ~9 months was unfavourably associ-ated with a change in weight-for-height and change inweight-for-age from ~9 months to 2 years, was not asso-ciated with weight-for-height or weight-for-age at age~2 years, and was favourably associated with waistcircumference-for-age at age ~2 years and change inwaist circumference-for-age from ~9 months to 2 years[45] (Additional file 2: Table S1).Table 6 The relationship between sedentary behaviour and cardiomNo. of participants(No. of studies)Design Quality assessmentRisk of bias Inconsistency IndirectnesThe mean age was 3.1 years. Data were collected cross-sectionally. Cardiometabolic hea276 (1) Cross-sectionala Serious risk of biasb No seriousinconsistencyNo seriousindirectnesaIncludes 1 cross-sectional study [126]bSerious risk of bias. Unknown reliability and validity of the exposure measure [126cThe quality of evidence from the cross-sectional study was downgraded from “lowconfidence in the observed effectsIn the two case-control studies, TV time [35, 36] andtotal sedentary time (assessed by one-day parent-recall)[36] were not significantly different between pre-schoolers with overweight/obese (case group) or normal-weight (control group) status, but watching TV for ≥1 h/day was unfavourably associated with having overweightstatus [35] (Additional file 2: Table S1).Among the 47 cross-sectional studies, sedentarybehaviour was assessed as accelerometer-derived seden-tary time, screen-based (i.e., computer time, time playinginactive video games, using the internet, watchingDVDs/videos, TV time, and total screen time), or othersedentary behaviours (i.e., sedentary quiet play, and timein the car or in baby seats).The relationships between accelerometer-derived seden-tary time and adiposity indicators in toddlers and pre-schoolers were examined in 11 cross-sectional studies; nullassociations were reported in 10/11 studies [37–40, 47, 53,Screen-based sedentary behaviours:There was no association between parent-reported screentime and SI (β = −0.04; R2 = 18.4%; p = 0.50) [125].evention of dietary- and lifestyle-induced health effects in children and infantsionable validity and reliability of subjective and objective exposure measures,nce the study did not provide a measure of variability in the results” to “very low” because of: (1) a serious risk of bias that diminished the level of60, 75, 78, 80] and mixed unfavourable and null associa-tions in 1/11 studies [77] (Additional file 2: Table S1).Specifically, total sedentary time was not associated with: %body fat, fat mass index, trunk fat mass index, or lean massindex in 1/1 studies [78]; BMI in 1/1 studies [75]; BMI z-score in 4/4 studies [37–39, 47]; and weight status in 4/4studies [40, 53, 60, 80] (Additional file 2: Table S1). Totalsedentary time was not associated with BMI z-scoreetabolic healthAbsolute effect Qualitys Imprecisionlth was assessed using an objective measure of blood pressure.sNo seriousimprecisionScreen-based sedentary behaviours:Watching TV for ≥ 2 h/day was not associated withhigh blood pressure (compared to <2 h/day,Prevalence Ratio = 0.9, 95% CI: 0.5, 1.4, p = 0.568) [126].Very lowc]” to “very low” because of a serious risk of bias that diminished the level ofspre.4 yareo sepreanicalseanw”asThe Author(s) BMC Public Health 2017, 17(Suppl 5):868 Page 78 of 215percentile or waist circumference, but was associated withwaist circumference percentile in girls (not boys) in 1/1studies [77]. Accelerometer-derived sedentary time in 30-min bouts was not associated with weight status [40].For screen-based sedentary behaviours, time playinginactive video games was unfavourably associated withpreschoolers’ BMI percentile, but using the internet andwatching DVDs/videos were not cross-sectionally associ-ated with BMI percentile [69] (Additional file 2: Table S1).Computer time was not associated with preschoolers’Table 7 The relationship between sedentary behaviour and fitnesNo. of participants(No. of studies)Design Quality assessmentRisk of bias Inconsistency Indirectness ImThe mean age at exposure measurement ranged from ~29 to 53 months (~2.4 to 4assessed as: lower body explosive strength (standing long jump) and fitness level (p1314 (2) Longitudinala Serious risk ofbiasbNo seriousinconsistencySeriousindirectnesscNimaIncludes 2 longitudinal studies [89, 90] from 1 unique sample (QLSCD)bSerious risk of bias. Questionable reliability and validity of the exposure [89, 90]included participants differed from missing participants [89]; controlled for physcSerious indirectness. Differences between outcomes of included studies and thofitness (lower-body strength assessed by standing long-jump performance) [89],reported direct measures of total body musculoskeletal or cardiovascular fitnessdThe quality of evidence from the longitudinal studies was downgraded from “loconfidence in the observed effects, and 2) indirectness of the comparisons beingweight status in 4/4 studies [63, 67, 71, 79], but wasunfavourably associated with sum of skinfold thicknessesin 1/1 studies [71].The relationships between total screen time and adi-posity indicators were examined in 18 cross-sectionalstudies; unfavourable associations were reported in 6/18studies [32, 33, 46, 50, 59, 73], null associations in 10/18studies [44, 52, 57, 58, 62, 64, 65, 71, 72, 79], and mixedunfavourable and null associations in 2/18 studies [41,61] (Additional file 2: Table S1). Of these, screen timewas unfavourably associated with: sum of skinfold thick-nesses in 0/1 studies, waist-to-height ratio in 1/1 studies[33], BMI in 2/2 studies [46, 50], and at least one meas-ure of weight status in 6/16 studies [32, 33, 41, 59, 61,73]. Only one of these studies was in infants (no associ-ation between screen time and weight status [58]); therest were in toddlers and preschoolers.The relationships between TV time and adiposity indi-cators in toddlers and preschoolers were examined in 23cross-sectional studies; unfavourable associations werereported in 5/23 studies [33, 55, 66, 67, 71], null associa-tions in 11/23 studies [31, 42, 43, 49, 50, 56, 60, 63, 69,75, 76], mixed unfavourable and null associations in 5/23 studies [48, 51, 54, 68, 79], mixed null and favourableassociations in 1/23 studies [74], and mixed unfavour-able, null, and favourable associations in 1/23 studies[70] (Additional file 2: Table S1). Of these, TV time wasunfavourably associated with: waist-to-hip ratio in 0/1studies, waist-to-height ratio in 1/1 studies [33], tricepsskinfold thickness in 0/1 studies, waist circumference in0/2 studies, sum of skinfolds in 1/3 studies [71], BMIpercentile in 0/1 studies, BMI in 2/11 studies [51, 54],and at least one measure of weight status in 9/13 studiesAbsolute effect Qualitycisionr). Data were collected longitudinally up to 8 years of follow-up. Fitness wasnt-report level relative to other children).riouscisionScreen-based sedentary behaviours:Higher TV time (hr/day) at age ~29 mowas unfavourably associated with standinglong-jump performance (cm) at age 97.8 mo(B = −0.361; 95% CI: −0.576, −0.145; p < 0.001) [89]and physical fitness level (scale from −2 to 2) in Grade 4(β = −0.09, SE = 0.0004; B = −0.01, 95% CI: −0.002, −0.02;p < 0.01) [90].A greater increase in TV time (hr/week) between age ~29and ~53 months was unfavourably associated with standinglong-jump performance (cm) at age 97.8 months (B = −0.285;95% CI: −0.436,-0.134; p < 0.01) [89] and physical fitness level(scale from −2 to 2, relative to other children) in Grade 4(β = −0.10, SE = 0.0003, p < 0.01) [90].Very lowdd outcome [90] measures; large unexplained loss to follow-up and unclear ifactivity [89, 90]of interest; only one study reported a measure of lower-body musculoskeletald one study reported an indirect measure of physical fitness [90]. No studiesto “very low” because of: 1) a serious risk of bias that diminished the level ofsessed[33, 48, 55, 66–68, 70, 71, 79]. Weekday (but not week-end) TV time was favourably associated with the ratio oftriceps to subscapular skinfold thickness (representinglimb-to-trunk adiposity ratio) in girls but not boys in 1/1studies [74]. TV time was favourably associated withBMI z-score in boys but not girls in 1/1 studies [70](Additional file 2: Table S1).Regarding other sedentary behaviours, infants’ time inbaby seats was not cross-sectionally associated withweight-for-height/age or waist circumference-for-age[45]. Among preschoolers, time using books [69] wasnot associated with BMI percentile [69]. Sedentary quietplay (defined as “e.g., looking into books, playing withblocks, playing with dolls, drawing, construction”) onweekdays or weekend days was not associated withweight status in boys [79]. In girls, sedentary quiet playon weekend days (but not weekdays) was unfavourablyassociated with weight status [79].Motor developmentThe relationships between sedentary behaviour and motordevelopment were examined in seven studies (see Table 2and Additional file 2: Table S2) [37, 40, 88, 91–94]. StudyTable 8 High-level summary of findings by health indicatorHealth indicator Number of studies Quality of evidence Summary of findings: Number of studies reporting unfavourable/null/favourableassociations with at least one health indicator measure by SB typeaCriticalAdiposity 60 Very low to moderate Objectively measured sedentary time:Sedentary time in 30-min bouts (accelerometer-derived): null (1)Total sedentary time (accelerometer-derived): unfavourable (1), null (12)Screen-based sedentary behaviours:Computer (duration, frequency): unfavourable (1), null (6)Internet (duration): null (1)Total screen time (duration): unfavourable (9), null (14)TV time (duration): unfavourable (20), null (24), favourable (2)Video games (duration): unfavourable (1)Other screens (DVDs/videos; duration): unfavourable (1), null (1)Other sedentary behaviours:Reading (duration): null (1)Sitting (baby seats, car, sedentary quiet play; duration): unfavourable (2),null (4), favourable (1)Motor development 7 Very low Objectively measured sedentary time:Sedentary time in 30-min bouts (accelerometer-derived): null (1)Total sedentary time (accelerometer-derived): unfavourable (1), null (2)Screen-based sedentary behaviours:TV time (duration): unfavourable (2), null (3)Other sedentary behaviours:Sitting (baby carrier/sling, car seat, high chair/other chair, playpen, stroller;duration): null (1), favourable (1)Supine position (duration): unfavourable (1), null (1)Psychosocial health 15 Very low to moderate Objectively measured sedentary time:Total sedentary time (accelerometer-derived): null (1)Screen-based sedentary behaviours:Computer (duration): unfavourable (1), null (1)Total screen time (duration): unfavourable (1)TV time (duration): unfavourable (9), null (11), favourable (2)Cognitive development 25 Very low Objectively measured sedentary time:Total sedentary time (accelerometer-derived): null (1)Screen-based sedentary behaviours:Computer (yes, no): null (1)Mobile phone use (yes, no): unfavourable (1)Total screen time (duration): unfavourable (1)TV time (duration): unfavourable (11), null (10), favourable (1)Video games (duration): null (1)Other screens (total or electronic media exposure; duration):unfavourable (2), null (1)The Author(s) BMC Public Health 2017, 17(Suppl 5):868 Page 79 of 215inutioThe Author(s) BMC Public Health 2017, 17(Suppl 5):868 Page 80 of 215designs were: longitudinal (n = 3) [88, 91, 92], and cross-sectional (n = 4) [37, 40, 93, 94]. Indicators of motordevelopment were measured objectively (e.g., visual-motorabilities measured using the Wide-Range Assessment ofVisual Motor Ability) or assessed subjectively by parent-report (e.g., age at first sitting; see Table 2 for summary ofmeasures). The quality of evidence was “very low” acrossstudy designs (Table 2).Among the three longitudinal studies, sedentarybehaviour was assessed from age 3.9 months to 2.4 years asTable 8 High-level summary of findings by health indicator (ContHealth indicator Number of studies Quality of evidenceImportantBone and skeletal health 1 Very lowCardiometabolic health 1 Very lowFitness 2 Very lowRisks / harms 0 N/AaNote that the number of studies reporting unfavourable/null/favourable associasome studies reported mixed associations. N/A: not applicablescreen-based (i.e., TV time) or other sedentary behaviours(i.e., time in a baby carrier/sling, car seat, high chair/otherchair, playpen, or stroller). Motor development indicatorswere assessed after 1.3 to 3 years of follow-up. For screen-based sedentary behaviours, TV time was not prospectivelyassociated with age at first sitting, crawling, or walking [91],visual-motor abilities [88], or object control [92], but wasunfavourably associated with locomotion skills [92].Regarding other sedentary behaviours, infants’ time ina baby carrier/sling, stroller, high chair or other chair, orplaypen was not associated with age at first sitting,crawling, or walking [91] (Additional file 2: Table S2).Greater time in a car seat at age ~9 months was associ-ated with earlier (i.e., favourable) age at first sitting andage at first crawling, but was not associated with age atfirst walking; time spent in a car seat at ages ~4 monthsand 1.7 years was not associated with age at first sitting,crawling, or walking [91].In the 4 cross-sectional studies, sedentary behaviourwas assessed as accelerometer-derived sedentary time,screen-based (i.e., TV time), or other sedentary behav-iours (i.e., time in the supine position). The relationshipsbetween accelerometer-derived sedentary time andmotor development were examined in two of the cross-sectional studies. Total sedentary time was not associ-ated with motor skills at age ~2 years [40] or ~3 to4 years [37], or with object control skills at age ~3 to4 years [37], but % sedentary time was unfavourablyassociated with locomotor skills at age ~3 to 4 years[37]. The number of 30-min bouts of sedentary behav-iour was not associated with motor skills [40].For screen-based sedentary behaviours, TV time wasunfavourably associated with motor skill development;ed)Summary of findings: Number of studies reporting unfavourable/null/favourableassociations with at least one health indicator measure by SB typeaOther sedentary behaviours:Reading (duration, frequency): null (2), favourable (3)Storytelling with parents (frequency): null (2), favourable (2)Screen-based sedentary behaviours:Screen time (duration): null (1)Objectively measured sedentary time:Total sedentary time (accelerometer-derived): null (1)Screen-based sedentary behaviours:TV time (duration): null (1)Screen-based sedentary behaviours:TV time (duration): unfavourable (2)N/Ans does not sum to the total number of studies for a given indicator sincechildren with delayed motor skill development spentmore time watching TV compared to children with typ-ical motor skill development, and children who were fre-quently exposed to TV (>0 h/day for children <2 yearsand >2 h/day for children ≥2 years) were more likely tohave delayed motor skill development than those whowere infrequently exposed [94].For other sedentary behaviours, time in the supineposition before 6 months of age was not associated withgross motor performance, but time in the supine pos-ition after age 6 months was unfavourably associatedwith gross motor performance [93].Psychosocial healthThe relationships between sedentary behaviour and psy-chosocial health in toddlers and preschoolers wereexamined in 15 studies (no studies in infants; see Table 3and Additional file 2: Table S3) [34, 90, 92, 95–106].Study designs were: randomized controlled trial (n = 1)[34], longitudinal (n = 9) [90, 92, 95–97, 99, 100, 102,103], and cross-sectional design or additionally reportedcross-sectional findings (n = 7) [98, 100, 101, 103–106].Indicators of psychosocial health (e.g., aggression,The Author(s) BMC Public Health 2017, 17(Suppl 5):868 Page 81 of 215symptoms of anxiety and depression) were assessed sub-jectively by parent-, teacher-, or self-report using ques-tionnaires (see Table 3 for summary of measures). Thequality of evidence ranged from “very low” to “moderate”across study designs (Table 3).In the randomized controlled trial of an interventionto reduce screen time, preschoolers’ screen time was sig-nificantly lower in the intervention versus control groupat 2, 6, and 9 months post-intervention [34]. Aggressiveand delinquent behaviours were not significantly differ-ent between the intervention and control groups at base-line, but were significantly lower in the interventionversus control group at 9-months post-intervention [34](Additional file 2: Table S3).Among the nine longitudinal studies, screen-basedsedentary behaviour (i.e., time e-gaming or on a com-puter, or TV time) was assessed from age ~1.5 to 5 years.Psychosocial health indicators were assessed after ~1 to9.5 years of follow-up.Time spent e-gaming or on a computer (on weekdaysor weekend days) at age 4.3 years was not associatedwith being at risk for the following at age 6.3 years: peerproblems, self-esteem problems, social well-being prob-lems, social functioning problems, or family functioningproblems [96]. Time spent e-gaming or on a computeron weekdays (but not weekend days) at age 4.3 yearswas unfavourably associated with being at risk for emo-tional problems at age 6.3 years in girls but not boys[96] (Additional file 2: Table S3).The relationships between TV time among toddlers/preschoolers and psychosocial health indicators atfollow-up were examined in nine longitudinal studies;unfavourable associations were reported in 2/9 studies[95, 103], null associations in 1/9 studies [100], mixedunfavourable and null associations in 5/9 studies [90, 92,96, 97, 99], and mixed null and favourable associationsin 1/9 studies [102] (Additional file 2: Table S3). Specif-ically, TV time was prospectively unfavourably associ-ated with the following psychosocial health indicators:victimization [90, 95], victimization by classmates [92],being a victim of bullying [97], being a bully [103], exter-nalizing problems [99], and being at risk for family func-tioning problems [96] (Additional file 2: Table S3). Nullassociations were reported between TV time and emo-tional symptoms [100]; conduct problems [100]; peer-problems [100]; prosocial behaviour [92, 100];externalizing problems [99, 102]; anxiety or depressivesymptoms [92, 102]; physical aggression [100] or aggres-sive behaviour [102]; being a bully, being a victim of bully-ing, or being a bully-victim [97]; being at risk foremotional problems, peer problems, self-esteem problems,emotional well-being problems, or social functioningproblems [96]; and co-operation, self-control, assertion,responsibility, or total social skills [102]. TV time at age~2.5 years was favourably associated with emotionalreactivity scores after ~3 years of follow-up [102].In the 7 cross-sectional studies, sedentary behaviourwas assessed as accelerometer-derived total sedentarytime or screen-based (i.e., TV time) sedentary behaviour.Total sedentary time (accelerometer-derived) was notcross-sectionally associated with preschoolers’ psycho-social health indicators (soothability, sociability, or emo-tionality) [104].The relationships between TV time and psychosocialhealth indicators in toddlers and preschoolers wereexamined in six cross-sectional studies; unfavourable as-sociations were reported in 2/6 studies [101, 103], nullassociations in 2/6 studies [100, 106], mixed unfavour-able and null associations in 1/6 studies [105], andmixed unfavourable and favourable associations in 1/6studies [98]. Specifically, TV time was unfavourablyassociated with aggression [101], bullying [103], totalexternalizing behaviour problems [105], and total be-haviour problems [105]. Null associations were re-ported between TV time and emotional symptoms,conduct problems, peer problems, and prosocial be-haviour [100], aggression toward a sibling [106], andinternalizing behaviour problems [105]. TV time wasfavourably associated with social-emotional compe-tence in one study [98].Cognitive developmentThe relationships between sedentary behaviour andcognitive development were examined in 25 studies (seeTable 4 and Additional file 2: Table S4) [88, 90, 92, 94,100, 102, 104, 107–124]. Study designs were: longitu-dinal (n = 11) [88, 90, 92, 100, 102, 112, 113, 119–122],case-control (n = 1) [116], and cross-sectional design oradditionally reported cross-sectional findings (n = 16)[90, 94, 100, 104, 107–111, 114, 115, 117, 118, 121, 123,124]. Indicators of cognitive development were mea-sured objectively (e.g., working memory capacitymeasured using the Memory for Digit Span test) orassessed subjectively by parent-report interview or ques-tionnaire (e.g., receptive vocabulary; see Table 4 forsummary of measures). The quality of evidence was“very low” across study designs (Table 4).Among the 11 longitudinal studies, sedentary behav-iour was assessed from age ~6 months to 5 years asscreen-based (i.e., electronic media exposure and TVtime) or other sedentary behaviours (i.e., frequency ofparents reading). Cognitive development indicators wereassessed after ~8 months to 8 years of follow-up.For screen-based sedentary behaviours, electronicmedia exposure at age ~6 months was unfavourablyassociated with the following at age 14 months: cognitivedevelopment, language development, and auditorycomprehension [112]. The relationships between TVThe Author(s) BMC Public Health 2017, 17(Suppl 5):868 Page 82 of 215time and cognitive development indicators in toddlersand preschoolers were examined in 10 longitudinalstudies; unfavourable associations were reported in 5/10studies [90, 92, 100, 120, 121], null associations in 4/10studies [88, 102, 113, 122], and mixed unfavourable, null,and favourable associations in 1/10 studies [119]. Specif-ically, TV time was prospectively unfavourablyassociated with the following cognitive developmentindicators: rate of change in language development[121], receptive vocabulary and number knowledge [92],classroom engagement [90, 92], mathematical achieve-ment [90], attentional problems [120], and hyperactivity-inattention [100] (Additional file 2: Table S4).Regarding other sedentary behaviours, the frequencyof parents reading to their child from ages ~8 months to4 years was favourably associated with both languagedevelopment at age 4 years and the rate of change inlanguage development between ages 5 to 7 years [121](Additional file 2: Table S4).In the case-control study, toddlers with language delay(cases) had significantly greater TV time than those withnormal language development (controls) [116]. Com-pared with toddlers who viewed ≤2 h/day TV time, thosewith >2 h/day TV time had increased odds of languagedelay [116].In the 16 cross-sectional studies, sedentary behaviourwas assessed as accelerometer-derived sedentary time,screen-based (i.e., computer use, mobile phone use, timeplaying inactive video games, TV time, total media expos-ure, and total screen time), or other sedentary behaviours(i.e., reading or storytelling with parents). Only one cross-sectional study examined the association betweenaccelerometer-derived total sedentary time and cognitivedevelopment indicators; total sedentary time was not asso-ciated with attention span in preschoolers [104].For screen-based sedentary behaviours, computer usewas not associated with the prevalence of speech disor-ders, but mobile phone use (any versus none) wasunfavourably associated with speech disorders in tod-dlers and preschoolers [109]. Time playing inactive videogames was not associated with hyperactivity or attentionproblems in preschoolers [107]. Total screen time wasunfavourably associated with communication develop-ment in toddlers [111], and total media exposure wasunfavourably associated with receptive language devel-opment and expressive language development in infantsand toddlers aged ~6 months to 1.3 years, but not withtotal language development in toddlers aged ~1.4 to2.3 years [124].The relationships between TV time and cognitivedevelopment in toddlers and preschoolers were exam-ined in nine cross-sectional studies; unfavourable associ-ations were reported in 3/9 studies [94, 108, 123], nullassociations in 5/9 studies [90, 100, 114, 115, 121], andmixed unfavourable and null associations in 1/9 studies[118] (see Additional file 2: Table S4 for statistics). Spe-cifically, TV time was unfavourably associated with lan-guage development or capacity in 2/5 studies [94, 108](Additional file 2: Table S4). TV time was unfavourablyassociated with delayed executive function [123] andcognitive development [94], but was not associated withcognitive ability [90] (Additional file 2: Table S4). TVtime was not associated with hyperactivity-inattention intoddlers [100], and was unfavourably associated withteacher-reported, but not parent-reported, attention-deficit/hyperactivity disorder (ADHD) symptoms in pre-schoolers [118] (Additional file 2: Table S4).Regarding other sedentary behaviours, the relation-ships between reading with parents and cognitive devel-opment indicators in infants, toddlers, and preschoolerswere examined in three cross-sectional studies [110, 117,124], two of which analyzed the same dataset in differentways [117, 124]; reading with parents was favourablyassociated with language development percentile in bothinfants and toddlers [117], but was not associated withabsolute language development in toddlers (not analyzedin infants) [124]. Reading with parents was favourablyassociated with absolute receptive language develop-ment, but not expressive language development, in in-fants [124]. In the third study, reading with parents wasnot associated with executive function in preschoolers[110]. Storytelling with parents was favourably associatedwith language development percentile in infants [117].In toddlers, storytelling was favourably associated withabsolute language development [124], but not languagedevelopment percentile [117]. Storytelling with parentswas favourably associated with absolute receptive lan-guage development, but not expressive language devel-opment, in infants [124] (Additional file 2: Table S4).Important (secondary) health indicatorsBone and skeletal healthThe relationship between sedentary behaviour and boneand skeletal health in preschoolers was examined in onecross-sectional study (see Table 5 and Additional file 2:Table S5) [125]. The quality of evidence was rated as“very low”. As summarized in Table 5, parent-reportedscreen time and accelerometer-derived total sedentarytime were not associated with bone stiffness index inpreschool children [125]. No other indices of bone andskeletal health were examined.Cardiometabolic healthThe relationship between sedentary behaviour and car-diometabolic health in preschoolers was examined inone cross-sectional study (see Table 6 and Add-itional file 2: Table S6) [126]. The quality of evidencewas rated as “very low”. Watching TV for ≥2 h/day wasThe Author(s) BMC Public Health 2017, 17(Suppl 5):868 Page 83 of 215not associated with high blood pressure in preschoolchildren [126]. No other cardiometabolic biomarkerswere examined.FitnessThe relationship between sedentary behaviour and fit-ness in toddlers and preschoolers was examined in twolongitudinal studies (no studies in infants; see Table 7and Additional file 2: Table S7) [89, 90]. The quality ofevidence was rated as “very low”.As summarized in Table 7, greater TV time at age~2.4 years was unfavourably associated with standinglong-jump performance at age ~8.2 years [89] and phys-ical fitness level (assessed as “relative to other children”via parent-report) in Grade 4 (age ~10 years) [90]. Agreater increase in TV time between age ~2.4 and~4.4 years was unfavourably associated with standinglong-jump performance at age 8.2 years [89] and phys-ical fitness level in Grade 4 [90].Risks/harmNo studies examined harms associated with sedentarybehaviour.DiscussionThe objective of this study was to perform a systematicreview that examined the relationships between seden-tary behaviours and health indicators in children 0 to4 years, and to determine what doses of sedentarybehaviours (i.e., duration, patterns [frequency, interrup-tions], and type) were associated with health indicators.The main findings are the following: 1) associationsbetween objectively measured total sedentary time andhealth indicators (adiposity and motor development)were predominantly null; 2) associations betweenscreen-based sedentary behaviours and health indicators(adiposity, motor or cognitive development, and psycho-social health) were largely unfavourable or null; 3) asso-ciations between reading or storytelling and cognitivedevelopment were favourable or null; and 4) associationsbetween time spent seated (e.g., in baby seats, car seats,high chairs or strollers) or in the supine position andhealth indicators (adiposity, motor development) wereprimarily unfavourable or null. Few studies examinedindicators of bone and skeletal health, cardiometabolichealth, or fitness, and no studies reported on risks orharms (e.g., torticollis, injuries) associated with sedentarybehaviours. These findings suggest that, in the earlyyears, total sedentary time may have a negligible impacton health, but the way that time is spent is important,with screen-based and seated/supine sedentary behav-iours likely to have unfavourable or null health effects(unlikely to have favourable effects), and interactive non-screen-based activities such as reading and storytellinglikely to have favourable health effects. A summary ofthe findings is presented in Table 8.The finding that there are no associations betweenobjectively measured total sedentary time and healthindicators in the early years (0 to 4 years) is in contrastto the relationships in older age groups, in particularadults [4, 127]. While this suggests that in the early yearsa certain amount of sedentary behaviour may be innocu-ous and perhaps even necessary for healthy growth anddevelopment, these findings should be interpreted withcaution. First, objectively measured total sedentary timewas examined only in cross-sectional studies, and aplausible explanation for the perceived lack of associ-ation between total sedentary time and health indicatorsis that there had simply been insufficient time for thoseeffects to manifest, rather than there being no effect.This hypothesis is supported by comparison of findingsfrom longitudinal and cross-sectional studies for subsetsof total sedentary behaviour. For instance, 9/10 (90%)longitudinal studies reported at least one unfavourableassociation between TV time and adiposity indicators,compared to only 11/22 (50%) cross-sectional studies.However, total sedentary time was examined only in re-lation to adiposity and motor development (and inone study each for indicators of psychosocial health,cognitive development, and bone and skeletal health);it remains possible that total sedentary time is associ-ated with other health indicators, particularly thoselikely to be acutely affected in the early years, such ascognitive development. More well-designed studieswith objective measures of sedentary behaviour areneeded.Second, in the present review, studies that utilizedaccelerometry measures applied a range of samplingintervals (epochs) and cut-points. Given that these meas-urement parameters influence the amount of sedentarybehaviour captured [128, 129], individual studies mayhave under- or overestimated the total amount of seden-tary time and may therefore have resulted in an under-estimation or overestimation of true effects. However,Byun et al. applied three different accelerometry cut-points in two cross-sectional datasets to test whetherthis would influence the findings, and found no associ-ation between total sedentary time and BMI z-score,regardless of the cut-points used [38]. Nonetheless, themost appropriate way to objectively measure sedentarybehaviour in the early years is still unknown and remainsan important area for future work.Lastly, total sedentary time was not objectivelyassessed in any studies in the infant age group; however,such measures may not be meaningful in non-ambulatory infants. Although the associations betweentotal sedentary time and health indicators were primarilynull, the present data do not allow for recommendationsThe Author(s) BMC Public Health 2017, 17(Suppl 5):868 Page 84 of 215regarding “appropriate” amounts or patterning (e.g.,breaks) of total sedentary time.Regarding screen-based sedentary behaviours, thepresent findings support and extend those of the earliersystematic review [2]; overall, screen time (namely TVtime) was unfavourably associated with a range of healthindicators. Notably, TV time was the predominant meas-ure of screen-based behaviour, followed by total screentime, with only eight studies reporting relationshipsbetween computer use and any health indicator; twostudies for each of DVDs/videos, electronic/total mediaexposure, and inactive video games; and one study formobile phone and internet use. Findings for these otherscreen exposures were mixed (unfavourable or null), andsuggest no benefits and some potential for harm.Although it seems intuitive that different types ofscreens may exert different effects (e.g., interacting onvideo-chat versus passive screen use), research on chil-dren’s use of such technologies lags behind their adop-tion [130]; this is a substantial research gap. Importantly,screen-based behaviours are used as a proxy for seden-tary behaviour; however, it is uncertain whether childrenin this age group are actually sedentary while usingscreens, and there may be screen-related health effectsthat are independent of the “lack of movement” [131,132]. Notwithstanding these limitations, the presentfindings indicate that less screen-based sedentary behav-iour is better for optimal health in the early years of life.Other sedentary behaviour exposures were less fre-quently examined, and findings were mixed. In general,reading [110, 117, 121, 124] and storytelling [117, 124]were favourably associated with cognitive development,while various types of time spent seated (e.g., in a carseat, high chair, or stroller) had mixed unfavourable andnull associations with indicators of adiposity and motordevelopment [45, 81, 82, 91]. An age-dependent effectwas observed in the only study that assessed time in thesupine position; time spent supine before 6 months ofage was not associated with gross motor performance,but greater time in the supine position after age6 months was associated with worse gross motor per-formance [93]. Overall, there was a paucity of dataregarding the relationships between other types of sed-entary behaviours and health indicators. Research showsthat children are spending ~7 h of the day in sedentarypursuits [15], and ~2 h of these are occupied by screentime [15]; this leaves an additional 5 h that areunaccounted for. Other types of sedentary behavioursare thus highly understudied, and this is an importantresearch gap.Most studies examined the duration of sedentarybehaviours in relation to health indicators, with onlythree studies specifically examining the impact of pat-terns of behaviour (i.e., breaks, frequency). Specifically,there was no association between accelerometer-derivedsedentary time in 30-min bouts and indicators of adipos-ity and motor development [40], or between the fre-quency of playing computer games and adiposityindicators [82], but there were favourable associationsbetween the frequency of parents reading or storytellingand child cognitive development [121]. These findingsare consistent with those of studies that examined sed-entary behaviour duration; however, it remains difficultto draw conclusions regarding patterns of sedentarybehaviour for optimal health in the early years.Strengths, limitations, and future directionsStrengths of this review include the use of a comprehen-sive search strategy that was developed and peer-reviewed by librarians with expertise in systematicreviews, as well as inclusion of all study designs and abroad range of health indicators that represent variousdimensions of health. Rigorous methodological stan-dards were used in this review, including application ofthe GRADE framework to guide the review process andassess the quality of the evidence [27]. To our know-ledge, this systematic review is the first to synthesize theevidence regarding the relationships between objectivelyand subjectively measured sedentary behaviour acrossthe most comprehensive range of health indicators inchildren in the early years of life.In terms of limitations, sample size restrictions wereimposed for feasibility reasons and to maximizegeneralizability, but it is possible that studies withsmaller sample sizes might have provided additionalinsight. Further, because of heterogeneity in the meas-urement of sedentary behaviour and health indicators,meta-analyses were not possible and all studies wereweighted equally in the narrative synthesis. The direc-tion of associations (i.e., unfavourable, null, favourable)was based on statistical significance; clinical significancewas not considered.Although an abundance of evidence was synthesizedin this review, several limitations of this area of researchwere identified that remain to be addressed. As men-tioned, data were limited regarding the relationshipsbetween sedentary behaviour and four relevant healthindicators (two or fewer studies for each of bone andskeletal health, cardiometabolic health, fitness, and risks/harms); TV time was the primary sedentary exposure,with few studies examining “other” types of screens (e.g.,tablets, mobile phones) or sedentary behaviours (e.g.,reading, puzzles); and objective measures of total seden-tary time were employed only in cross-sectional studies.Although adiposity was the most commonly measuredhealth indicator (60 studies), direct measures of adipos-ity were used in only two studies [78, 82] while theremainder used surrogate measures such as BMI. OnlyThe Author(s) BMC Public Health 2017, 17(Suppl 5):868 Page 85 of 215one randomized controlled study was included in thepresent review, and the quality of the evidence rangedfrom “very low” to “moderate” across the study designsand health indicators. There is a need for high-qualitystudies with strong designs to better establish the magni-tude of effects and the nature of dose-response gradients(if applicable), to assess cause-and-effect relationships,and to examine potential subgroup differences (e.g.,based on age, sex, or socio-economic status). WhenRCTs are not possible because of the inherent challengesof research in this age group, quasi-experimental or lon-gitudinal designs that use validated sedentary behaviourmeasures and outcome measures that are sensitiveenough to detect changes are recommended.Across the health indicators, the most common reasonfor downgrading the quality of evidence was the serious riskof bias associated with sedentary behaviour measures withno known psychometric properties. Consequently, develop-ment and use of reliable and valid subjective measures ofsedentary behaviour are needed. Defining and measuringsedentary behaviour in young children, particularly in non-ambulatory infants, remains a challenge. For instance,infants in the supine position may be vigorously movingarms and legs, and thus being “active”, but existingquestionnaire-based measures do not capture this. Futureresearch using inclinometers, which can more accuratelycapture postures [133], as well as limb-worn devices, willhelp to address the challenges associated with quantifyingsedentary behaviours in the early years. Finally, the questionof whether different types of sedentary behaviour “content”(e.g., educational versus recreational TV programming)exert different health effects was beyond the scope of thisreview, and remains an important area for future work.ConclusionsThis systematic review synthesized findings from 96 studieswith ~200,000 participants in 33 countries around theworld; the quality of the evidence ranged from “very low”to “moderate”. In summary, the findings demonstrate thatin the early years (0 to 4 years), total sedentary time mayhave a negligible impact on health, but the quality of thattime is important, with screen-based and seated/supinesedentary behaviours likely to have no benefit and a poten-tial for harm, and interactive non-screen-based activitiessuch as reading with caregivers having favourable healtheffects. These findings continue to support the importanceof minimizing screen time for disease prevention andhealth promotion in the early years [2, 9], and also highlightthe potential benefits of interactive non-screen-based sed-entary behaviours such as reading and storytelling. There isa need for additional research using valid and reliable mea-sures and high-quality study designs, to more definitivelyestablish the relationships between sedentary behavioursand health indicators, and to provide insight into theappropriate dose (durations, patterns, type) of sedentary be-haviour for optimal health in the early years.Additional filesAdditional file 1: Search strategies. (PDF 58 kb)Additional file 2: Supplementary Tables S1-S7. (PDF 1166 kb)AbbreviationsADHD: Attention-deficit/hyperactivity disorder; BMI: Body mass index;GRADE: Grading of recommendations, assessment, development andevaluation; METS: Metabolic equivalent; PICOS: Population, intervention,comparison, outcomes and study designs framework; PRISMA: Preferredreporting items for systematic reviews and meta-analyses; RCT: Randomizedcontrolled trial; SB: Sedentary behaviour; TV: TelevisionAcknowledgmentsThe authors wish to acknowledge the work of Linda Slater for peer review ofthe search strategies, Alejandra Jaramillo Garcia and Véronique Dorais formethodological consultation, and Nick Barrowman and Katie Gunnell forstatistical consultation.FundingThis study has been made possible through funding from the CanadianInstitutes of Health Research (CIHR; Grant KRS-144044), Canadian Society forExercise Physiology, Healthy Active Living and Obesity Research Group at theChildren’s Hospital of Eastern Ontario Research Institute, and the PublicHealth Agency of Canada. Publication charges for this article have beenfunded by CIHR Grant KRS-144044. Valerie Carson is supported by a CanadianInstitutes of Health Research New Investigator Salary Award. Guy Faulkner issupported by a Chair Award in Applied Public Health funded by the PublicHealth Agency of Canada in partnership with the Canadian Institutes ofHealth Research.Availability of data and materialsNot applicable.About this supplementThis article has been published as part of BMC Public Health Volume17 Supplement 5, 2017: 24-Hour Movement Guidelines for the EarlyYears: An Integration of Physical Activity, Sedentary Behaviour, andSleep. The full contents of the supplement are available online athttps://bmcpublichealth.biomedcentral.com/articles/supplements/vol-ume-17-supplement-5.Authors’ contributionsVJP, CEG, and MST were responsible for the initiation, conceptualization,and design of the systematic review. VJP, CEG, XJ, and SA wereresponsible for data collection and extraction and for risk of biasassessment. VJP and CEG were responsible for the GRADE analysis andinterpretation of data. VJP was the primary author of the manuscript.All authors (VJP, CEG, XJ, SA, VC, GF, GSG, JJR, MS, and MST) wereresponsible for revising the manuscript critically for important intellectualcontent. All authors read and approved 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.The Author(s) BMC Public Health 2017, 17(Suppl 5):868 Page 86 of 215Author details1Healthy Active Living and Obesity Research Group, RI #1, Children’s Hospitalof Eastern Ontario Research Institute, 401 Smyth Road, Ottawa, ON K1H 8L1,Canada. 2University of Strathclyde, School of Psychological Science andHealth, G1 1QE, Glasgow, Scotland, UK. 3Faculty of Physical Education andRecreation, University of Alberta, Edmonton, AB T6G 2H9, Canada. 4School ofKinesiology, University of British Columbia, Vancouver, BC V6T 1Z3, Canada.5Library and Media Services, Children’s Hospital of Eastern Ontario, Ottawa,ON K1H 8L1, Canada.Published: 20 November 2017References1. Sedentary Behaviour Research Network. Letter to the editor: standardizeduse of the terms “sedentary” and “sedentary behaviours”. Appl Physiol NutrMetab. 2012;37:540–2. http://dx.doi.org/10.1139/h2012-024. Accessed 25Nov 2016.2. LeBlanc AG, Spence JC, Carson V, Connor Gorber S, Dillman C, Janssen I,et al. Systematic review of sedentary behaviour and health indicators in theearly years (aged 0-4 years). Appl Physiol Nutr Metab. 2012;37:753–72. PM:22765839.3. Tremblay MS, LeBlanc AG, Janssen I, Kho ME, Hicks A, Murumets K, et al.Canadian sedentary behaviour guidelines for children and youth. ApplPhysiol Nutr Metab. 2011;36:59–64. PM:21326378.4. Thorp AA, Owen N, Neuhaus M, Dunstan DW. Sedentary behaviorsand subsequent health outcomes in adults: a systematic review oflongitudinal studies, 1996-2011. Am J Prev Med. 2011;41:207–15. PM:21767729.5. Biswas A, Oh PI, Faulkner GE, Bajaj RR, Silver MA, Mitchell MS, et al.Sedentary time and its association with risk for disease incidence, mortality,and hospitalization in adults: a systematic review and meta-analysis. AnnIntern Med. 2015;162:123–32. PM:25599350.6. Reilly JJ, Jackson DM, Montgomery C, Kelly LA, Slater C, Grant S, et al. Totalenergy expenditure and physical activity in young Scottish children: mixedlongitudinal study. Lancet. 2004;363:211–2. http://www.sciencedirect.com/science/article/pii/S0140673603153317.7. Janz KF, Burns TL, Levy SM. Tracking of activity and sedentary behaviors inchildhood: the Iowa bone development study. Am J Prev Med. 2005;29:171–8.http://www.sciencedirect.com/science/article/pii/S0749379705002023.8. Biddle SJ, Pearson N, Ross GM, Braithwaite R. Tracking of sedentarybehaviours of young people: a systematic review. Prev Med. 2010;51:345–51.PM:20682330.9. Tremblay MS, LeBlanc AG, Carson V, Choquette L, Connor Gorber S,Dillman C, et al. Canadian sedentary behaviour guidelines for the early years(aged 0-4 years). Appl Physiol Nutr Metab. 2012;37:370–91. PM:22448609.10. Australian Government Department of Health. Move and play every day: nationalphysical activity recommendations for children 0–5 years. 2014. http://www.health.gov.au/internet/main/publishing.nsf/Content/npra-0-5yrs-brochure.11. American Academy of Pediatrics, Council on Communications and Media.Children, adolescents, and the media. Pediatrics. 2013;132:958–61. 10.1542/peds.2013-2656. http://pediatrics.aappublications.org/content/132/5/958.abstract.12. Department of Health PAHIaP. Start active, stay active: a report on physicalactivity for health from the four home countries’ Chief Medical Officers.2011. https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/216370/dh_128210.pdf.13. Carson V, Tremblay MS, Spence JC, Timmons BW, Janssen I. The Canadiansedentary behaviour guidelines for the early years (zero to four years ofage) and screen time among children from Kingston, Ontario. Paediatr ChildHealth. 2013;18:25–8. PM:24381488.14. Zimmerman FJ, Christakis DA, Meltzoff AN. Television and DVD/videoviewing in children younger than 2 years. Arch Pediatr Adolesc Med.2007;161:473–9. PM:17485624.15. Garriguet D, Carson V, Colley RC, Janssen I, Timmons BW, Tremblay MS.Physical activity and sedentary behaviour of Canadian children aged 3 to 5.Health Rep. 2016;27:14–23. PM:27655168.16. Hnatiuk JA, Salmon J, Hinkley T, Okely AD, Trost S. A review of preschoolchildren’s physical activity and sedentary time using objective measures.Am J Prev Med. 2014;47:487–97. PM:25084681.17. Inchley J, Currie D, Young T, Samdal O, Torsheim T, Augustson L, et al.,editors. Growing up unequal: gender and socioeconomic differences inyoung people’s health and well-being. Health behaviour in school-agedchildren (HBSC) study: international report from the 2013/2014 survey.Copenhagen: World Health Organization; 2016.18. Carson V, Kuzik N, Hunter S, Wiebe SA, Spence JC, Friedman A, et al.Systematic review of sedentary behavior and cognitive development inearly childhood. Prev Med. 2015;78:115–22. PM:26212631.19. Carson V, Hunter S, Kuzik N, Gray CE, Poitras VJ, Chaput JP, et al. Systematicreview of sedentary behaviour and health indicators in school-agedchildren and youth: an update. Appl Physiol Nutr Metab. 2016;41:S240–65.PM:27306432.20. Carson V, Wong SL, Winkler E, Healy GN, Colley RC, Tremblay MS. Patterns ofsedentary time and cardiometabolic risk among Canadian adults. Prev Med.2014;65:23–7.21. Hinkley T, Teychenne M, Downing KL, Ball K, Salmon J, Hesketh KD. Earlychildhood physical activity, sedentary behaviors and psychosocial well-being: a systematic review. Prev Med. 2014;62:182–92. PM:24534461.22. Moher D, Liberati A, Tetzlaff J, Altman DG. The PRISMA group. Preferredreporting items for systematic reviews and meta-analyses: the PRISMAstatement. J Clin Epidemiol. 2009;62:1006–12. PM:19631508.23. Schardt C, Adams MB, Owens T, Keitz S, Fontelo P. Utilization of the PICOframework to improve searching PubMed for clinical questions. BMC MedInform Decis Mak. 2007;7:16. PM:17573961.24. LeBlanc AG, Chaput JP, McFarlane A, Colley RC, Thivel D, Biddle SJH, et al.Active video games and health indicators in children and youth: asystematic review. PLoS One. 2013;8:e65351. PM:23799008.25. Poitras VJ, Gray CE, Borghese MM, Carson V, Chaput JP, Katzmarzyk PT.Systematic review of the relationships between objectively measuredphysical activity and health indicators in school-aged children and youth.Appl Physiol Nutr Metab. 2016;41:S197–239. PM:27306431.26. Higgins JP, Altman DG, Gotzsche PC, Juni P, Moher D, Oxman AD, et al. TheCochrane Collaboration’s tool for assessing risk of bias in randomised trials.BMJ. 2011;343:d5928. PM:22008217.27. Guyatt G, Oxman AD, Akl EA, Kunz R, Vist G, Brozek J, et al. GRADEguidelines: 1. Introduction – GRADE evidence profiles and summary offindings tables. J Clin Epidemiol. 2011;64:383–94. PM:21195583.28. Balshem H, Helfand M, Schunemann HJ, Oxman AD, Kunz R, Brozek J, et al.GRADE guidelines: 3. Rating the quality of evidence. J Clin Epidemiol. 2011;64:401–6. PM:21208779.29. Guyatt GH, Oxman AD, Sultan S, Glasziou P, Akl EA, Alonso-Coello P, et al.GRADE guidelines: 9. Rating up the quality of evidence. J Clin Epidemiol.2011;64:1311–6. PM:21802902.30. Timmons BW, LeBlanc AG, Carson V, Connor Gorber S, Dillman C,Janssen I, et al. Systematic review of physical activity and health in theearly years (aged 0-4 years). Appl Physiol Nutr Metab. 2012;37:773–92.PM:22765840.31. Proctor MH, Moore LL, Gao D, Cupples LA, Bradlee ML, Hood MY, et al.Television viewing and change in body fat from preschool to earlyadolescence: the Framingham Children’s study. Int J Obes Relat MetabDisord. 2003;27:827–33. 12821969.32. Chiasson M, Scheinmann R, Hartel D, McLeod N, Sekhobo J, EdmundsLS, et al. Predictors of obesity in a cohort of children enrolled in WICas infants and retained to 3 years of age. J Community Health. 2016;41:127–33.33. Olafsdottir S, Berg C, Eiben G, Lanfer A, Reisch L, Ahrens W, et al. Youngchildren’s screen activities, sweet drink consumption and anthropometry:results from a prospective European study. Eur J Clin Nutr. 2014;68:223–8.PM:24253759.34. Yilmaz G, Demirli CN, Karacan CD. An intervention to preschool children forreducing screen time: a randomized controlled trial. Child Care Health Dev.2015;41:443–9. PM:24571538.35. Koleilat M, Harrison GG, Whaley S, McGregor S, Jenks E, Afifi A. Preschoolenrollment is associated with lower odds of childhood obesity among WICparticipants in LA County. Matern Child Health J. 2012;16:706–12. PM:21431308.36. Kain J, Andrade M. Characteristics of the diet and patterns of physicalactivity in obese Chilean preschoolers. Nutr Res. 1999;19:203–15.37. Williams HG, Pfeiffer KA, O’Neill JR, Dowda M, McIver KL, Brown WH, et al.Motor skill performance and physical activity in preschool children. Obesity(Silver Spring). 2008;16:1421–6. PM:18388895.38. Byun W, Liu J, Pate RR. Association between objectively measured sedentarybehavior and body mass index in preschool children. Int J Obes. 2013;37:961–5. PM:23318716.The Author(s) BMC Public Health 2017, 17(Suppl 5):868 Page 87 of 21539. Byun W, Dowda M, Pate RR. Correlates of objectively measured sedentarybehavior in US preschool children. Pediatrics. 2011;128:937–45. PM:22007010.40. Johansson E, Hagstromer M, Svensson V, Anna EK, Forssen M, Nero H, et al.Objectively measured physical activity in two-year-old children – levels,patterns and correlates. Int J Behav Nutr Phys Act. 2015;12:ArtID 3.41. Anderson SE, Whitaker RC. Household routines and obesity in US preschool-aged children. Pediatrics. 2010;125:420–8. PM:20142280.42. DuRant RH, Baranowski T, Johnson M, Thompson WO. The relationshipamong television watching, physical activity, and body composition ofyoung children. Pediatrics. 1994;94:449–55. PM:7936851.43. Burdette HL, Whitaker RC. A national study of neighborhood safety, outdoorplay, television viewing, and obesity in preschool children. Pediatrics. 2005;116:657–62. PM:16140705.44. Lioret S, Maire B, Volatier JL, Charles MA. Child overweight in France and itsrelationship with physical activity, sedentary behaviour and socioeconomicstatus. Eur J Clin Nutr. 2007;61:509–16. PM:16988644.45. Sijtsma A, Sauer PJ, Stolk RP, Corpeleijn E. Infant movement opportunitiesare related to early growth – GECKO Drenthe cohort. Early Hum Dev. 2013;89:457–61. PM:23631950.46. Sijtsma A, Koller M, Sauer PJ, Corpeleijn E. Television, sleep, outdoor playand BMI in young children: the GECKO Drenthe cohort. Eur J Pediatr. 2015;174:631–9. PM:25367053.47. Wijtzes AI, Kooijman MN, Kiefte-de Jong JC, de Vries SI, Henrichs J, JansenW, et al. Correlates of physical activity in 2-year-old toddlers: the generationR study. J Pediatr. 2013;163:791–9. PM:23523279.48. Manios Y, Kourlaba G, Kondaki K, Grammatikaki E, Anastasiadou A,Roma-Giannikou E. Obesity and television watching in preschoolers inGreece: the GENESIS study. Obesity (Silver Spring). 2009;17:2047–53. PM:19282823.49. Kourlaba G, Kondaki K, Liarigkovinos T, Manois Y. Factors associated withtelevision viewing time in toddlers and preschoolers in Greece: the GENESISstudy. J Public Health (Oxf). 2009;31:222–30. PM:19224946.50. van Stralen MM, te Velde SJ, van Nassau F, Brug J, Grammatikaki E, Maes L,et al. Weight status of European preschool children and associations withfamily demographics and energy balance-related behaviours: a pooledanalysis of six European studies. [review]. Obes Rev. 2012;13 Suppl 1:29–41.PM:22309063.51. Wen LM, Baur LA, Rissel C, Xu H, Simpson JM. Correlates of body massindex and overweight and obesity of children aged 2 years: findings fromthe healthy beginnings trial. Obesity (Silver Spring). 2014;22:1723–30.PM:24415528.52. Turer CB, Stroo M, Brouwer RJ, Krause KM, Lovelady CA, Bastian LA, et al. Dohigh-risk preschoolers or overweight mothers meet AAP-recommendedbehavioral goals for reducing obesity? Acad Pediatr. 2013;13:243–50.PM:23491583.53. Dolinsky DH, Brouwer RJ, Evenson KR, Ostbye T, Evenson KR, Siega-Riz AM.Correlates of sedentary time and physical activity among preschool-agedchildren. Prev Chronic Dis. 2011;8:A131. PM:22005624.54. Fuller-Tyszkiewicz M, Skouteris H, Hardy LL, Halse C. The associationsbetween TV viewing, food intake, and BMI. A prospective analysis of datafrom the longitudinal study of Australian children. Appetite. 2012;59:945–8.PM:23000277.55. Brown JE, Broom DH, Nicholson JM, Bittman M. Do working mothers raisecouch potato kids? Maternal employment and children’s lifestyle behavioursand weight in early childhood. Soc Sci Med. 2010;70:1816–24. PM:20299142.56. Dubois L, Farmer A, Girard M, Peterson K. Social factors and television useduring meals and snacks is associated with higher BMI among pre-schoolchildren. Public Health Nutr. 2008;11:1267–79. PM:18547454.57. Dennison BA, Erb TA, Jenkins PL. Television viewing and television inbedroom associated with overweight risk among low-income preschoolchildren. Pediatrics. 2002;109:1028–35. PM:12042539.58. Asplund KM, Kair LR, Arain YH, Cervantes M, Oreskovic NM, Zuckerman KE.Early childhood screen time and parental attitudes toward child televisionviewing in a low-income Latino population attending the specialsupplemental nutrition program for women, infants, and children. Child.2015;11:590–9. PM:26390321.59. Nelson JA, Carpenter K, Chiasson MA. Diet, activity, and overweight amongpreschool-age children enrolled in the special supplemental nutritionprogram for women, infants, and children (WIC). Prev Chronic Dis. 2006;3:A49. PM:16539790.60. LaRowe TL, Adams AK, Jobe JB, Cronin KA, Vannatter SM, Prince RJ.Dietary intakes and physical activity among preschool-aged childrenliving in rural American Indian communities before a family-basedhealthy lifestyle intervention. J Am Diet Assoc. 2010;110:1049–57.PM:20630162.61. Chen LP, Ziegenfuss JY, Jenkins SM, Beebe TJ, Ytterberg KL. Pediatric obesityand self-reported health behavior information. Clin Pediatr (Phila). 2011;50:872–5. PM:21357199.62. Minh Do LM, Tran TK, Eriksson B, Petzold M, Nguyen CTK, Ascher H.Preschool overweight and obesity in urban and rural Vietnam: differences inprevalence and associated factors. Glob Health Action. 2015;8:28615. PM:26452338.63. Hajian-Tilaki K, Heidari B. Childhood obesity, overweight, socio-demographicand life style determinants among preschool children in Babol, northernIran. Iran J Public Health. 2013;42:1283–91. PM:26171341.64. Koubaa AA, Younes K, Gabsi Z, Bouslah A, Maalel I, Maatouk EMW, et al. [riskfactors of children overweight and obesity]. [French]. Tunis Med. 2012;90:387–93. PM:22585646.65. Sasaki A, Yorifuji T, Iwase T, Komatsu H, Takao S, Doi H. Is there anyassociation between TV viewing and obesity in preschool children in Japan?Acta Med Okayama. 2010;64:137–42. PM:20424669.66. Jouret B, Ahluwalia N, Cristini C, Dupuy M, Negre-Pages L, Grandjean H,et al. Factors associated with overweight in preschool-age children insouthwestern France. Am J Clin Nutr. 2007;85:1643–9. PM:17556704.67. Jiang J, Rosenqvist U, Wang H, Greiner T, Ma Y, Toschke AM. Risk factors foroverweight in 2- to 6-year-old children in Beijing, China. Int J Pediatr Obes.2006;1:103–8. PM:17907322.68. Levin S, Martin MW, Riner WF. TV viewing habits and body mass indexamong South Carolina head start children. Ethn Dis. 2004;14:336–9. PM:15328934.69. Harrison K, Liechty JM. US preschoolers’ media exposure and dietary habits:the primacy of television and the limits of parental mediation. J Childrenand Media. 2012;6:18–36.70. Vandebosch H, Cleemput KV. Television viewing and obesity amongpre-school children: the role of parents. Communications. 2007;32:417–46.71. Mendoza JA, Zimmerman FJ, Christakis DA. Television viewing, computeruse, obesity, and adiposity in US preschool children. Int J Behav Nutr PhysAct. 2007;4:44. PM:17894878.72. Fulton JE, Wang X, Yore MM, Carlson SA, Galuska DA, Caspersen CJ.Television viewing, computer use, and BMI among U.S. children andadolescents. J Phys Act Health. 2009;6 Suppl 1:S28–35. PM:19998847.73. Twarog JP, Politis MD, Woods EL, Boles MK, Daniel LM. Daily televisionviewing time and associated risk of obesity among U.S. preschool agedchildren: an analysis of NHANES 2009-2012. Obes Res Clin Pract. 2015;9:636–8.74. Tremblay L, Rinaldi CM. The prediction of preschool children’s weight fromfamily environment factors: gender-linked differences. Eat Behav. 2010;11:266–75. PM:20850062.75. Jago R, Baranowski T, Baranowski JC, Thompson D, Greaves KA. BMI from 3-6 y of age is predicted by TV viewing and physical activity, not diet. Int JObes. 2005;29:557–64. PM:15889113.76. Taverno RS, Dowda M, Saunders R, Pate R. Double dose: thecumulative effect of TV viewing at home and in preschool onchildren’s activity patterns and weight status.[erratum appears inPediatr Exerc Sci. 2014 Feb;26(1):120]. Pediatr Exerc Sci. 2013;25:262–72. PM:23502043.77. Espana-Romero V, Mitchell JA, Dowda M, O'Neill JR, Pate RR. Objectivelymeasured sedentary time, physical activity and markers of body fat inpreschool children. Pediatr Exerc Sci. 2013;25:154–63. PM:23406703.78. Collings PJ, Brage S, Ridgway CL, Harvey NC, Godfrey KM, Inskip HM, et al.Physical activity intensity, sedentary time, and body composition inpreschoolers. Am J Clin Nutr. 2013;97:1020–8. PM:23553158.79. Cardon G, De Bourdeaudhuij I, Iotova V, Latomme J, Socha P, Koletzko B,et al. Health related behaviours in normal weight and overweightpreschoolers of a large pan-European sample: the ToyBox-study. PLoS One.2016;11(3):e0150580. 10.1371/journal.pone.0150580. PM:26950063.80. Bonvin A, Barral J, Kakebeeke TH, Kriemler S, Longchamp A, Marques-Vidal P,et al. Weight status and gender-related differences in motor skills and inchild care-based physical activity in young children. BMC Pediatr. 2012;12:23.PM:22405468.The Author(s) BMC Public Health 2017, 17(Suppl 5):868 Page 88 of 21581. Reilly JJ, Armstrong J, Dorosty AR, Emmett PM, Ness A, Rogers I, et al. Earlylife risk factors for obesity in childhood: cohort study. BMJ. 2005;330:1357.PM:15908441.82. Leary SD, Lawlor DA, Davey SG, Brion MJ, Ness AR. Behavioural early-lifeexposures and body composition at age 15 years. Nutr Diabetes. 2015;5:e150. PM:25664839.83. Flores G, Lin H. Factors predicting overweight in US kindergartners. Am JClin Nutr. 2013;97:1178–87. PM:23553169.84. Gooze RA, Anderson SE, Whitaker RC. Prolonged bottle use and obesity at 5.5 years of age in US children. J Pediatr. 2011;159:431–6. PM:21543085.85. Wheaton N, Millar L, Allender S, Nichols M. The stability of weight statusthrough the early to middle childhood years in Australia: a longitudinalstudy. BMJ Open. 2015;5:e006963. PM:25922101.86. Griffiths LJ, Hawkins SS, Cole TJ, Dezateux C. Millennium cohort studychild health group. Risk factors for rapid weight gain in preschoolchildren: findings from a UK-wide prospective study. Int J Obes. 2010;34:624–32.87. De Coen V, De Bourdeaudhuij I, Verbestel V, Maes L, Vereecken C. Riskfactors for childhood overweight: a 30-month longitudinal study of 3-to 6-year-old children. Public Health Nutr. 2014;17:1993–2000. PM:24172063.88. Schmidt ME, Rich M, Rifas-Shiman SL, Oken E, Taveras EM. Televisionviewing in infancy and child cognition at 3 years of age in a US cohort.Pediatrics. 2009;123:e370–5. PM:19254972.89. Fitzpatrick C, Pagani LS, Barnett TA. Early childhood television viewingpredicts explosive leg strength and waist circumference by middlechildhood. Int J Behav Nutr Phys Act. 2012;9:87. PM:22793018.90. Pagani LS, Fitzpatrick C, Barnett TA, Dubow E. Prospective associationsbetween early childhood television exposure and academic, psychosocial,and physical well-being by middle childhood. Arch Pediatr Adolesc Med.2010;164:425–31. PM:20439793.91. Hesketh KD, Crawford DA, Abbott G, Campbell KJ, Salmon J. Prevalence andstability of active play, restricted movement and television viewing ininfants. Early Child Dev Care. 2015;185:883–94.92. Pagani LS, Fitzpatrick C, Barnett TA. Early childhood television viewingand kindergarten entry readiness. Pediatr Res. 2013;74:350–5. PM:23788060.93. De Kegel A, Peersman W, Onderbeke K, Baetens T, Dhooge I, VanWaelvelde H. New reference values must be established for the Albertainfant motor scales for accurate identification of infants at risk formotor developmental delay in Flanders. Child Care Health Dev. 2013;39:260–7. PM:22676145.94. Lin LY, Cherng RJ, Chen YJ, Yang HM. Effects of television exposure ondevelopmental skills among young children. Infant Behav Dev. 2015;38:20–6. PM:25544743.95. Watt E, Fitzpatrick C, Derevensky JL, Pagani LS. Too much television?Prospective associations between early childhood televiewing and later self-reports of victimization by sixth grade classmates. J Dev Behav Pediatr.2015;36:426–33. PM:26075581.96. Hinkley T, Verbestel V, Ahrens W, Lissner L, Molnar D, Moreno LA, et al.Early childhood electronic media use as a predictor of poorer well-being: a prospective cohort study. JAMA Pediatr. 2014;168:485–92. PM:24639016.97. Verlinden M, Tiemeier H, Veenstra R, Mieloo CL, Jansen W, JaddoeVWV, et al. Television viewing through ages 2-5 years and bullyinginvolvement in early elementary school. BMC Public Health. 2014;14:157. PM:24520886.98. Intusoma U, Mo-Suwan L, Ruangdaraganon N, Panyayong B,Chongsuvivatwong V. Effect of television viewing on social-emotionalcompetence of young Thai children. Infant Behav Dev. 2013;36:679–85. PM:23948636.99. Verlinden M, Tiemeier H, Hudziak JJ, Jaddoe VW, Raat H, Guxens M, et al.Television viewing and externalizing problems in preschool children: thegeneration R study. Arch Pediatr Adolesc Med. 2012;166:919–25. PM:22869354.100. Cheng S, Maeda T, Yoichi S, Yamagata Z, Tomiwa K, Japan Children’s studygroup. Early television exposure and children’s behavioral and social outcomesat age 30 months. J Epidemiol. 2010;20 Suppl 2:S482–9. PM:20179364.101. Manganello JA, Taylor CA. Television exposure as a risk factor for aggressivebehavior among 3-year-old children. Arch Pediatr Adolesc Med. 2009;163:1037–45. PM:19884595.102. Mistry KB, Minkovitz CS, Strobino DM, DLG B. Children’s television exposureand behavioral and social outcomes at 5.5 years: does timing of exposurematter? Pediatrics. 2007;120:762–9. PM:17908763.103. Zimmerman FJ, Glew GM, Christakis DA, Katon W. Early cognitivestimulation, emotional support, and television watching as predictors ofsubsequent bullying among grade-school children. Arch Pediatr AdolescMed. 2005;159:384–8. PM:15809395.104. Irwin JD, Johnson AM, Vanderloo LM, Burke SM, Tucker P. Temperamentand objectively measured physical activity and sedentary time amongCanadian preschoolers. Prev Med Rep. 2015;2:598–601. PM:26844125.105. Teramoto S, Soeda A, Hayashi Y, Saito K, Urashima M. Problematicbehaviours of 3-year-old children in Japan: relationship withsocioeconomic and family backgrounds. Early Hum Dev. 2005;81:563–9.106. Miller LE, Grabell A, Thomas A, Bermann E, Graham-Bermann SA. Theassociations between community violence, television violence, intimatepartner violence, parent-child aggression, and aggression in siblingrelationships of a sample of preschoolers. Psychology of Violence. 2012;2:165–78.107. Linebarger DL. Contextualizing video game play: the moderating effectsof cumulative risk and parenting styles on the relations among videogame exposure and problem behaviors. Psychol Pop Media Cult. 2015;4:375–96.108. Byeon H, Hong S. Relationship between television viewing and languagedelay in toddlers: evidence from a Korea national cross-sectional survey.PLoS One. 2015;10:e0120663. PM:25785449.109. Rajchanovska D, Ivanovska BZ. The impact of demographic and socio-economic conditions on the prevalence of speech disorders in preschoolchildren in Bitola. Srp Arh Celok Lek. 2015;143:169–73. PM:26012126.110. Linebarger DL, Barr R, Lapierre MA, Piotrowski JT. Associations betweenparenting, media use, cumulative risk, and children’s executive functioning.J Dev Behav Pediatr. 2014;35:367–77. PM:25007059.111. Duch H, Fisher EM, Ensari I, Font M, Harrington A, Taromino C. Associationof screen time use and language development in Hispanic toddlers: a cross-sectional and longitudinal study. Clin Pediatr (Phila). 2013;52:857–65. PM:23820003.112. Tomopoulos S, Dreyer BP, Berkule S, Fierman AH, Brockmeyer C,Mendelsohn AL. Infant media exposure and toddler development. ArchPediatr Adolesc Med. 2010;164:1105–11. PM:21135338.113. Foster EM, Watkins S. The value of reanalysis: TV viewing and attentionproblems. Child Dev. 2010;81:368–75. PM:20331673.114. Zimmerman FJ, Gilkerson J, Richards JA, Christakis DA, Xu D, Gray S, et al.Teaching by listening: the importance of adult-child conversations tolanguage development. Pediatrics. 2009;124:342–9. PM:19564318.115. Ruangdaraganon N, Chuthapisith J, Mo-Suwan L, Kriweradechachai S,Udomsubpayakul U, Choprapawon C. Television viewing in Thai infants andtoddlers: impacts to language development and parental perceptions. BMCPediatr. 2009;9:34. PM:19460170.116. Chonchaiya W, Pruksananonda C. Television viewing associates with delayedlanguage development. Acta Paediatr. 2008;97:977–82. PM:18460044.117. Zimmerman FJ, Christakis DA, Meltzoff AN. Associations between mediaviewing and language development in children under age 2 years. JPediatr. 2007;151:364–8. PM:17889070.118. Miller CJ, Marks DJ, Miller SR, Berwid OG, Kera EC, Santra A, et al. Briefreport: television viewing and risk for attention problems in preschoolchildren. J Pediatr Psychol. 2007;32:448–52. PM:17012738.119. Zimmerman FJ, Christakis DA. Children’s television viewing and cognitiveoutcomes: a longitudinal analysis of national data. Arch Pediatr AdolescMed. 2005;159:619–25. PM:15996993.120. Christakis DA, Zimmerman FJ, DiGiuseppe DL, McCarty CA. Early televisionexposure and subsequent attentional problems in children. Pediatrics. 2004;113:708–13. PM:15060216.121. McKean C, Mensah FK, Eadie P, Bavin EL, Bretherton L, Cini E, et al. Leversfor language growth: characteristics and predictors of language trajectoriesbetween 4 and 7 years. PLoS One. 2015;10:e0134251.122. Blankson A. Do hours spent viewing television at ages 3 and 4 predictvocabulary and executive functioning at age 5? Merrill-Palmer Q. 2015;61:264–89.123. Nathanson AI, Fries PT. Television exposure, sleep time, andneuropsychological function among preschoolers. Media Psychology. 2014;17:237–61.124. Ferguson CJ, Donnellan MI. The association between children’s babyvideo viewing and poor language development robust? A reanalysisof Zimmerman, Christakis, and Meltzoff (2007). Dev Psychol. 2014;50:129–37.125. Herrmann D, Buck C, Sioen I, Kouride Y, Marild S, Molnar D, et al.Impact of physical activity, sedentary behaviour and muscle strengthon bone stiffness in 2-10-year-old children – cross-sectional resultsfrom the IDEFICS study. Int J Behav Nutr Phys Act. 2015;12:112. PM:26377674.126. Crispim PA, Peixoto MR, Jardim PC. Risk factors associated with high bloodpressure in two-to five-year-old children. Arq Bras Cardiol. 2014;102:39–46.PM:24263779.127. Wilmot EG, Edwardson CL, Achana FA, Davies MJ, Gorely T, Gray LJ, et al.Sedentary time in adults and the association with diabetes, cardiovasculardisease and death: systematic review and meta-analysis. Diabetologia. 2012;55:2895–905. PM:22890825.128. Cliff DP, Okely AD. Comparison of two sets of accelerometer cut-off pointsfor calculating moderate-to-vigorous physical activity in young children. JPhys Act Health. 2007;4:509–13. PM:18209240.129. Colley RC, Harvey A, Grattan KP, Adamo KB. Impact of accelerometer epochlength on physical activity and sedentary behaviour outcomes forpreschool-aged children. Health Rep. 2014;25:3–9. PM:24430918.130. Kabali HK, Irigoyen MM, Nunez-Davis R, Budacki JG, Mohanty SH, Leister KP,et al. Exposure and use of mobile media devices by young children.Pediatrics. 2015;136:1044–50. PM:26527548.131. Mazarello Paes V, Ong KK, Lakshman R. Factors influencing obesogenicdietary intake in young children (0-6 years): systematic review of qualitativeevidence. BMJ Open. 2015;5:e007396. PM:26377503.132. Christakis DA, Gilkerson J, Richards JA, Zimmerman FJ, Garrison MM, Xu D,et al. Audible television and decreased adult words, infant vocalizations, andconversational turns: a population-based study. Arch Pediatr Adolesc Med.2009;163:554–8. PM:19487612.•  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 Submit your next manuscript to BioMed Central and we will help you at every step:The Author(s) BMC Public Health 2017, 17(Suppl 5):868 Page 89 of 215133. Ridgers ND, Salmon J, Ridley K, O'Connell E, Arundell L, Timperio A.Agreement between activPAL and ActiGraph for assessing children’ssedentary time. Int J Behav Nutr Phys Act. 2012;9:15. PM:22340137.•  Maximum visibility for your researchSubmit your manuscript atwww.biomedcentral.com/submit


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