Effectiveness of intervention strategies exclusively targeting reductions in children’s sedentary time: a systematic review of the literature

An increasing number of interventions targeting sedentary behaviour in children have emerged in recent years. Recently published reviews included sedentary behaviour and physical activity interventions. This review critically summarizes evidence on the effectiveness of intervention strategies that exclusively targeted reducing sedentary time in children and adolescents. We performed a systematic literature search in Pubmed, Embase and the Cochrane Library through November 2015. Two independent reviewers selected eligible studies, extracted relevant data and rated the methodological quality using the assessment tool for quantitative studies. We included 21 intervention studies, of which 8 studies scored moderate on methodological quality and 13 studies scored weak. Four out of eight moderate quality studies reported significant beneficial intervention effects. Although descriptions of intervention strategies were not always clearly reported, we identified encouragement of a TV turnoff week and implementing standing desks in classrooms as promising strategies. Due to a lack of high quality studies and inconsistent findings, we found no convincing evidence for the effectiveness of existing interventions targeting solely sedentary behaviour. We recommend that future studies apply mediation analyses to explore which strategies are most effective. Furthermore, to increase the effectiveness of interventions, knowledge of children’s motives to engage in sedentary behavior is required, as well as their opinion on potentially effective intervention strategies. Electronic supplementary material The online version of this article (doi:10.1186/s12966-016-0387-5) contains supplementary material, which is available to authorized users.


Background
Children nowadays spend a large amount of their time in sedentary behaviours with average values of up to nine hours per day for total sedentary time and up to four hours per day for screen time [1][2][3][4][5]. Importantly, many children exceed the World Health Organization (WHO) recommendation of limiting electronic media use (i.e. <1 h/day for 2-4 year olds and <2 h/day for 5-17 year olds). For example, up to 80 % of 12-17 year olds spend more than 2 h/day on screen behaviours [4,6].
To date, evidence regarding the relationship between sedentary behaviour and health indicators in children is not convincing [7][8][9], partly due to a lack of methodologically sound studies [10]. Nevertheless, the growing public health concern regarding the health effects of excessive sedentary behaviour in children has led to an emerging number of interventions targeting sedentary behaviour in children and adolescents in recent years [11][12][13][14].
A number of reviews have been published on the effectiveness of interventions targeting sedentary behaviour [15][16][17][18][19][20][21][22][23][24][25], as well as a review of reviews [26]. However, a number of studies included in these reviews included strategies targeting promotion of moderate-tovigorous physical activity. It is therefore not known which intervention strategies resulted in sedentary behaviour change, and which intervention strategies that exclusively target reductions in sedentary behaviour are effective.
Therefore, the aim of the present review was to critically summarize the evidence regarding the effectiveness of interventions that exclusively targeted reductions in sedentary time in children and adolescents (0 to 18 years old), without strategies promoting physical activity simultaneously. Additionally, we aimed to identify effective intervention strategies to reduce sedentary time. Though current guidelines focus on screen-based sedentary behaviour, recent evidence highlights the importance of limiting prolonged sitting [27][28][29]. Therefore, we did not limit our review to interventions targeting a specific type of sedentary behaviour.

Literature search
We performed a systematic literature search in Pubmed, Embase and the Cochrane Library from inception through November 2015. The search strategy included terms related to 'interventions' (e.g. randomized controlled trial, controlled trial, control group) in ANDcombination with terms related to sedentary behaviour (e.g. screen time, television, computer use, sitting). The search was limited to studies in children and adolescents (i.e. participants aged 0-18 years). Additional file 1 provides the full search. In addition, reference lists were screened for potential additional studies.

Inclusion and exclusion criteria
Intervention studies were included if they (i) evaluated interventions (or intervention arms) targeting sedentary behaviour (e.g. TV viewing, computer use, reading, playing board games etc.) in children and adolescents (aged 0 to 18 years old), and (ii) included time spent in sedentary behaviour as an outcome measure. Studies were excluded if they also included strategies that promoted increases in physical activity. We included only full-text studies that were published in the English language in peer-reviewed journals.

Selection process and data extraction
Two reviewers (TA and JK) independently checked all identified titles and abstracts to establish potential relevant studies. Next, two reviewers (TA and JK) independently screened the full-text papers to check if they met the inclusion criteria. Any discrepancies were resolved through discussion and, if consensus was not reached, with a third reviewer (MC).
The following data were extracted from all included studies, using a structured form: (i) participant characteristics; (ii) intervention strategies; (iii) intervention setting; (iv) intervention duration and follow-up; (v) description of control group (vi) sedentary behaviour outcome; and (vii) results of the study.

Methodological quality assessment and data synthesis
Two reviewers (TA and MC) independently rated the methodological quality of all included studies using the quality assessment tool for quantitative studies of the Effective Public Health Practice Project (EPHPP) [30]. This tool uses three response options (strong, moderate, weak) on the following eight quality criteria: selection bias, study design, confounders, blinding, data collection methods, withdrawals and dropouts, intervention integrity, and analysis (see Additional file 2). Any discrepancies were resolved through discussion. The methodological quality of a study was rated as strong when at most one of the quality criteria was scored as weak and two as moderate. A study was rated as moderate when at most two weak quality criteria were scored as weak. When more than two quality criteria were scored as weak, a study was rated as weak.
To synthesize the study results, we applied a bestevidence synthesis [31], in accordance with the Cochrane Collaboration. Using this method, the number of studies, their methodological quality and the consistency of the results are taken into account: Strong evidenceconsistent findings in ≥ 2 studies of strong quality; Moderate evidenceconsistent findings in ≥2 studies of moderate quality; Conflicting/insufficient evidenceconflicting findings or lack of moderate/high quality studies.

Methodological quality and intervention effects
Additional file 3 shows the quality assessment scores and Table 2 shows the results of all included studies (sorted by age range and methodological quality). None of the studies were rated as being of strong methodological quality, six studies were moderate and 12 studies were of weak quality. Except for one study [44], all studies were randomized controlled trials, indicating adequate study designs. In contrast, study samples were often not representative, assessors not blinded to group assignment, participants not blinded to the research question(s) and the outcome measures were of unknown/ inadequate validity and reliability. Moreover, participation rate, intended intervention dose, analyses according to intention-to-treat and method of randomization were poorly described.
Of the eight moderate quality studies, four studies reported significant intervention effects on children's sedentary time [32,33,43,46] while four studies reported no significant effects [11,40,47,48]. Among children aged 2.5 to 5.5 years old, Dennison et al. [32] applied a 7-week intervention targeting children's TV viewing. Intervention strategies included knowledge transfer (take-home messages of sessions and brochures), a TV turnoff period of one week and targeting parental skills (monitoring and rewarding their child's behaviour). This study found that significantly fewer children in the intervention group watched >2 h per day TV than control children (parent-report). Moreover, they found that children in the intervention group watched significantly less TV/video on weekdays and Sundays (i.e. -36 and -60 min/day, respectively) than control children, but not            on Saturdays. No significant intervention effects were found on computer/video game use in this study [32]. Epstein et al. [33] found that, after a 2-year intervention period targeting both TV and computer time, 4-to-7 year old children (BMI >75 th percentile for age and sex) in the intervention group spent less time on objectively assessed (i.e. TV/computer allowance) TV viewing and computer games (i.e. -3.2 h/d) than children in the control group. This study included knowledge transfer (monthly newsletters to parents), the use of a TV control device, parental skills (rewarding their child's behaviour) and goal setting (goals set by research staff), as intervention strategies. Among 9-to-10 year olds, Hinckson et al. [46] implemented standing desks in classrooms of the intervention school for four weeks. Compared to children in the control classroom, children in the intervention classroom reduced their overall time spent sitting though the difference was small [46]. Among 8-to-9 year olds, Robinson et al. [43] applied a 6-month intervention targeting decreasing media use. Intervention strategies included a 10-day TV turnoff period, the use of a TV control device, knowledge transfer (children engaged in lessons, parents received newsletters), parental skills (rewarding their child's behaviour) and goal setting (goals set by research staff). This study found that children in the intervention group reported watching less TV (i.e. 47 min/d) and playing fewer videogames (i.e. 22 min/d) than control children, however, no effects were found on watching videotapes and other sedentary behaviours (i.e. using a computer, reading, listening to music). Based on parent-report, only children's TV viewing was significantly lower in the intervention group than the control group (i.e. -37 min/d) [43]. According to the best-evidence synthesis, based on the eight moderate quality studies, we found conflicting/inconsistent evidence for the effectiveness of interventions targeting sedentary behaviour. Nevertheless, from these moderate quality studies we identified two promising intervention strategies (i.e. moderate evidence). Second, one study in which standing desks were implemented in classrooms demonstrated promising intervention effects [46].

Discussion
This review aimed to summarize the evidence on the effectiveness of interventions targeting exclusively sedentary behaviour in children aged 0-18 years. We conclude that there is conflicting/insufficient evidence for the effectiveness of such interventions. Based on the three moderate quality studies reporting significant intervention effects, two intervention strategies seem promising: 1) encouragement of a TV turnoff week; and 2) implementation of standing desks in classrooms. Since there were only eight moderate quality studies, including children with a wide age range (i.e. varying from 3.1 to 11.2 years), we could not draw age-specific conclusions.
Two out of three studies that included the encouragement of a TV turnoff week as a strategy to reduce sedentary time found reductions on TV/video time [32,43]. Encouraging children not to watch TV for a certain time period may help them to experience the enjoyment of behaviours other than TV viewing. A premise for this is that the alternative behaviour should be as fun as the specific sedentary behaviour [49]. Robinson et al. [43] found no effects on other sedentary behaviours such as watching videotapes, reading and using the computer, indicating that TV viewing was not replaced by other sedentary activities. Future studies should confirm the potential of implementing a TV turnoff week to reduce TV viewing time and explore longer-term effects.
Recently, the implementation of furniture nudging interruptions in sedentary time, for example standing breaks, within the school setting has gained more attention. Although the effect was small, Hinckson et al. [46] found a reduction in overall time spent sitting when implementing standing desks. To date, it is unclear whether standing interruptions can prevent potential negative health effects of excessive sedentary behaviour in children.
Surprisingly, although in a number of studies children were involved in implementing the intervention (e.g. setting personal goals, discussing non-sedentary activities), no studies collaborated with children and/or parents in the development of interventions. It is likely that active participation of both children and their parents in the choice and development of intervention strategies may lead to more acceptable and attractive strategies and thereby more effective interventions [50]. We aimed to identify effective intervention strategies to reduce sedentary behaviour. Unfortunately, strategies were not always clearly described. Besides being able to reliably extract intervention strategies for systematic reviews, clear descriptions of behaviour change techniques may also benefit accurate replication, faithful implementation and well-designed development of interventions [51]. We therefore recommend future studies to make use of clearly described standardized behaviour change techniques when designing and reporting intervention content.
Only four out of eight moderate quality studies reported significant intervention effects. This is in contrast with previous reviews and meta-analyses on interventions targeting sedentary behaviour in combination with physical activity promotion, in which predominantly small but significant effects and effect sizes were reported [15,17,18,[21][22][23]25]. One explanation may be that a combined focus on reducing sedentary behaviour and increasing physical activity is more effective. Secondly, in contrast to previously published reviews, we performed a methodological quality assessment and excluded the studies with weak quality from our evidence synthesis, of which a number of studies reported some significant findings.
Only one previous review additionally summarized interventions targeting solely sedentary behaviour [19]. In contrast to our findings, Leung et al. [19] concluded that sedentary behaviour interventions significantly reduced sedentary time. Their review was based on only three studies, whereas we included 21 studies in our review. Additionally, Leung et al. [19] did not include the methodological quality of the studies in their evidence synthesis.
We found no studies of strong methodological quality. Representativeness of the study sample, controlling for relevant confounders, blinding and measurement of sedentary behaviour were issues that limited the quality of the included studies. We recommend that future studies keep these potential sources of bias in mind when designing a trial. For example, analyses should preferably adjust for baseline levels of sedentary behaviour. Regarding blinding, assessors should be blinded to group assignment, and participants should preferably be blinded to the research question or the authors could speculate on the effect of any suspected modifying factors, such as belief in the intervention, in the discussion. Finally, sedentary behaviour should be assessed through accurate, valid, reliable and responsive measures.
Strengths of this systematic review include the focus on interventions exclusively targeting sedentary behaviour. This is important when examining the true effectiveness of sedentary behaviour interventions and optimizing future interventions. Another strength of this review is that we not only assessed the methodological quality of included papers, but also took this into account in our evidence synthesis. A limitation of every review is that our findings may suffer from publication bias. Additionally, we did not calculate effect sizes as the number of included studies with a moderate or high quality was limited (i.e. eight studies) and these studies were rather heterogeneous, i.e. intervention durations ranged from 4 weeks to 2 years, age of included children ranged from 3.1 to 11.2 years and outcome measure of sedentary time reflected parent-reported total screen time, TV time, computer time, parent-and childreported total sedentary time and objectively measured sedentary time (ActivPAL, Actigraph). Finally, due to the wide age range of included samples in the moderate quality studies, we could not draw age-specific conclusions.

Conclusions
We conclude that to date there is unconvincing evidence for the effectiveness of interventions targeting exclusively sedentary behaviour. Based on the eight moderate quality studies that found a significant intervention effect, encouraging a TV turnoff week and implementing standing desks in classrooms seem promising. As all included studies applied multiple intervention strategies, it is impossible to distinguish which strategies are most promising. We recommend that future studies explore which strategies are most effective, by applying mediation analyses. Moreover, in order to increase the effectiveness of interventions, knowledge of children's motives to engage in sedentary behaviour is required as well as their opinion on potentially effective intervention strategies.