Do Interventions to Increase Walking Work? A Systematic Review of Interventions in Children and Adolescents

The majority of identified studies from this systematic review were delivered in an educational setting (school/college). Since this age group spends a significant proportion of their waking time within the educational setting, it is a highly suitable setting for the promotion of PA and reduces dependence on family input [58].

Seven studies focused on increasing walking through active travel to school, with five of these reporting desirable intervention outcomes [46, 49‐51, 55]. These studies employed a range of different methods to assess changes in walking, making it difficult to directly compare which approaches produced the greatest changes in walking levels in children.

An active travel to school day, which was a one-off event, found no significant effect on levels of walking [48], probably because of the short-term nature of the intervention [48]. A longer intervention (6 weeks), also involving active commuting to school, observed decreases in PA at follow-up for both intervention and control groups [47]. A number of factors may have contributed to this decline, including weather and parental influence on choice of travel to school, coupled with a lack of intervention fidelity [47].

Although the studies included within the present review often did not include health outcomes within their study design, a previous review on the association between active travel to school and health-related fitness found this form of PA was associated with healthier body composition outcomes and improved levels of cardiorespiratory fitness in young people [45]. Furthermore, the present review focused solely on walking and thus interventions which also included cycling as a form of active transportation were not included. Based on the available evidence, active travel to school by walking may contribute to increased levels of walking in children but this effect may be dependent upon environmental factors and parental decision making.

A smaller number of studies (n = 3) targeted at adolescents included intervention content that incorporated the provision of unsealed pedometers with individualised [52] or group discussions [54, 56]; all of these studies reported desirable intervention outcomes. No walking interventions in adolescents focused on active travel to school as a main intervention component, which is in contrast to the majority of studies targeted at children. Although there is a general consensus that PA levels decline as children move into adolescence, there is inconsistency within the literature about the context of this decline in PA, particularly as children transition from primary to secondary education, and towards adolescence. Studies have reported increases in active travel amongst boys and girls after this transition [59], while evidence in girls has shown that rates of active travel to school (walking/cycling) have been shown to remain constant, and not increase, as girls move to secondary school [60]. In contrast, longitudinal data has shown increases in the likelihood of active travel to school until the age of 10 years, when the likelihood begins to decrease [61]. Given that walking to school can contribute up to 34 % of total daily MVPA in children aged 11–12 years [62], the potential for walking to school to increase PA should be further examined by practitioners and policymakers alike as a key approach to tackling the problem of inactivity, particularly amongst adolescents [62].

School-based interventions are known to be important for the promotion of PA levels [63]; however, there is a need for further evaluation of such interventions to assess the long-term benefits on levels of PA [63]. In the present review, the majority of studies did not include a longer-term follow-up post-intervention, making it difficult to draw conclusions on the longer-term effects of such interventions on walking levels. Short-term changes in behaviour are feasible through the implementation of effective interventions [64], with the literature suggesting that the shorter-term nature of interventions may play a role in maintaining the interest and motivation of participants, particularly children [64]. However, similar to the present review, less is known about how effective such interventions are at maintaining changes in PA behaviour.

Parental involvement may form an integral component of school-based interventions [63], with strong evidence indicating that interventions involving schools in combination with family members could increase PA in adolescents [34]. Only one school-based study reporting desirable intervention outcomes involved family members as part of the intervention [55]. Parental support and direct help in facilitating PA have been shown to positively correlate with PA in adolescents [21]; however, in the present review, parental involvement was not outlined in the included studies targeted at adolescents. In addition to social support from parents, peer relationships can play a role in PA involvement at this stage of the lifecycle [65]; however, only one study explicitly mentioned participating in walks with friends as a means of increasing walking behaviours [56].

Targeting PA interventions to sub-groups of the population remains a key focus for research as young people transition from childhood into adolescence. Although ten included studies were targeted at both sexes, only two of these studies provided comparisons between sexes on intervention outcomes [48, 56]. Shimon and Petlichkoff observed no differences between sexes for daily steps post-intervention [56]. An active travel day did not result in desirable intervention outcomes for walking; however, significant differences were observed between sexes for active travel to school, with girls significantly increasing their active travel to school compared with boys, with walking cited as the most common mode of active travel [48]. The lack of evidence on the effectiveness of interventions in relation to sex limits conclusions on how best to promote walking amongst sexes.

Given that girls show a greater decline in PA than boys [14], this may be a suitable target subgroup for walking interventions. Within this review, two studies targeted girls only [52, 54], with both achieving increases in PA. These interventions did not involve structured walking sessions or walking school buses; instead participants were encouraged to set goals and increase their levels of walking by changing smaller behaviours [52, 54]. Given the preferences of adolescent girls, in particular those who are least active, for activities that are non-competitive and informal in nature [29], such intervention content may represent a more suitable means of increasing walking in this population. Furthermore, given that girls are less likely to be active during school recess periods [66], coupled with a decline in PA during recess as girls move towards adolescence [60], other opportunities to promote walking outside of active travel should be examined; for example, opportunities to walk around school grounds has been highlighted as a potential means of increasing walking and subsequent PA [67]. A meta-analysis of interventions to increase PA in young people (<11 years old) highlighted that larger effects were found for studies that were targeted at girls only [64]. Given the limited number of studies considered within this review, it is not possible to determine whether interventions that were sex-specific or mixed sex were more effective in relation to walking levels.

A number of studies within this review aimed to increase walking via multiple small bouts throughout the day rather than a single longer bout. Accumulated walking (in smaller bouts) has been shown to be as effective at increasing PA in adults as walking for a single, longer bout [68]. This may represent a more feasible means of increasing PA within paediatric populations where spontaneous PA is known to be intermittent rather than continuous in nature [69]. Adverse outcomes as a result of participating in walking were not identified within included studies, suggesting that walking may present a low risk of injury and increase its acceptability as a form of PA.

BCTs employed within interventions were multiple and varied. Social support, repetition and substitution, feedback and monitoring and goals and planning were the most frequently employed BCTs within interventions producing desirable outcomes on walking. Successful active travel interventions in children incorporated social support alongside other BCTs, suggesting that the support of others (e.g. researchers, parent volunteers and other children) may be important in the adoption and maintenance of walking to school over car use or other methods of transport. The three interventions aimed at adolescents employed similar BCTs, namely goals and planning, and feedback and monitoring, suggesting that such techniques may be effective at eliciting changes in walking levels in this age group. Pedometers have been shown to increase awareness of how behavioural choices impact upon subsequent PA by providing live feedback, which can assist individuals in making changes to increase their PA [44]. The provision of pedometers for self-monitoring, coupled with individualised or group information sessions, may therefore be effective at increasing walking in adolescents although there is a need for further studies in this age group. Such techniques may be more suitable for adolescents, who possess greater cognitive skills to employ and utilise these techniques, than children. A previous review of BCTs to promote walking and cycling in adults highlighted that self-monitoring and intention formation may be effective tools in promoting these PA behaviours [40]. Although self-monitoring was also identified in the current review as a potential BCT for inclusion in future interventions, there is a lack of evidence on how the use of BCTs translates from adult studies to children [70]. The identification of feedback, monitoring, and goals and planning as potentially successful intervention components to promote PA in children and adolescents within the present review is in contrast to the BCTs identified as effective in the management and prevention of obesity in children [70]. Techniques including prompt practice, environmental restructuring and the provision of information on individual consequences of a behaviour were recommended for use in future obesity prevention/management interventions [70]; however, these BCTs were not identified within the present review. From the available data, incorporating BCTs may be effective in combination with other intervention components in increasing walking behaviours in both children and adolescents.

Of the nine studies that reported a desirable outcome in relation to walking behaviours, seven used an objective measure to assess walking levels. Three studies used sealed pedometers, with information downloaded by study personnel rather than relying on participants to self-report their daily steps [52, 54, 56], reducing the chance of participants misreporting steps and tampering with the devices. The use of pedometers for measurement, as well as an intervention technique to facilitate self-monitoring, has been cited as a potential confounder in studies [54]. While pedometers are beneficial at detecting increases in steps over an intervention period, they do not enable researchers to measure the intensity of activities performed [44]. Despite observed increases in daily steps, it is therefore unclear if such increases were at a sufficient level of intensity to meet current MVPA guidelines and receive the associated health benefits. Two studies that used accelerometers demonstrated that increased walking contributed to increased levels of MVPA in children [46, 50]; however, the limited number of studies makes it difficult to draw associations on the contribution of walking to levels of MVPA. Previous work has highlighted the contribution walking to school can have on levels of moderate intensity PA [32], with validation studies suggesting brisk walking is sufficient to achieve moderate intensity PA in children and adolescents [71, 72].

The level of randomisation varied within studies from individual to class- and school-level randomisation. Randomising by school may be problematic as the chance of school-level factors such as a special event confounding upon results is increased. On the other hand, recruiting both intervention and control participants from the same schools [46, 52, 53, 56] may have introduced contamination with children, parents and teachers communicating with each other about intervention content. Within these studies, two interventions reported increases in walking levels for both intervention and control groups at follow-up [53, 56]. Although increases in control groups were not significant, such findings may suggest that the school environment is a fertile ground for spreading messages and promoting PA.

This is the first systematic review to focus on the impact of interventions aimed at promoting levels of walking in children and adolescents. Although RCTs are regarded as the gold standard in intervention design, this review included other study designs to ensure that all available evidence was captured. Lack of randomisation decreases the validity of a study but can provide an important insight into which intervention components may or may not work when designing programmes for a given population.

This systematic review has some limitations. The first of these was the small number of identified studies eligible for inclusion. Due to limited resources, only papers published in English were eligible for inclusion. In order that the effect of the intervention on walking could be determined in isolation, this review focused only on studies that had reported changes in walking behaviours. Other interventions that employed walking in combination with other activities, such as cycling and running, were not included as data for walking alone was either unreported or unpublished. It is also acknowledged that the majority of studies reported desirable intervention outcomes, which may be due to publication bias [73].

Given the small number of studies identified from this systematic review, it is evident that further research on the promotion of walking among children and, in particular, adolescents is warranted. Owing to sex-related differences in PA, further interventions tailored to girls are justified. From currently available evidence, it appears that walking interventions may be an effective means of increasing walking levels and subsequent PA in this population. Few studies provided a translation of walking outcome measures in relation to intensity, making it difficult to draw conclusions on the effectiveness of such interventions in helping children and adolescents move closer towards meeting current PA recommendations. Research has highlighted that walking may provide further opportunity for overweight and obese participants in relation to meeting these PA recommendations, owing to the increased intensity of the activity in this population [23]. Consequently, the effectiveness of walking interventions in overweight/obese populations warrants further study. In addition, only one study was identified that examined the effectiveness of walking interventions outside of the school setting; this may present opportunity for further research in other/novel settings.