|Home | About | Journals | Submit | Contact Us | Français|
Well evaluated complex interventions are still needed
Physically active children and adolescents are at reduced risk of developing risk factors for cardiovascular disease,1 and they are likely to have enhanced mental and emotional wellbeing.2 However, as with other developed countries, in the United Kingdom, three out of 10 boys and four out of 10 girls are estimated not to take the recommended 60 minutes each day of moderate to vigorous intensity physical activity.3 This is important, because in 2003, 28% of children in the UK were estimated to be overweight or obese.4
In this week's BMJ, Van Sluijs and colleagues report a systematic review of interventions to promote physical activity in children and adolescents.5 The review found weak or inconclusive evidence for the effectiveness of strategies to promote children's physical activity. It confirmed lessons from tobacco control6—that at least in adolescents the most effective interventions have many components and are undertaken in multiple settings (school, home, and community).
Randomised controlled trials that focus mainly on education are not sufficient to change behaviour and sustain such changes. This is irrespective of whether interventions target children, adolescents, or parents; low or high socioeconomic groups; or whether they are conducted at school or in the community. Effective interventions are generally those that educate as well as facilitate physical activity by providing opportunities and supportive environments at school, at home, and in the community.
Close examination of the review's findings5 suggests that it is not all “doom and gloom,” however. More than two thirds of the interventions had a positive effect, and just under half had a significant effect. This is despite the use of, at times, crude self reporting or proxy reporting of physical activity.
A recent narrative review of physical activity interventions in children and adolescents found that 64% of studies (n=25) that used an objective measure of physical activity reported significant effects compared with only 38% of studies (n=66) that used survey measures.7 This highlights the need to incorporate valid and responsive (sensitive to change) objective measures of physical activity in intervention trials, particularly in studies of children.
Van Sluijs and colleagues' review also highlights the importance of incorporating a thorough evaluation of the intervention process, which should include measures of fidelity, dose (delivered and received), reach, recruitment, and context. Without this information, it is difficult to determine why an intervention succeeded or failed.
The review5 identifies many gaps in our knowledge about the most effective strategies for promoting physical activity in young people. The authors question whether it is worth pursuing interventions that target boys and girls separately, ethnic minority populations, or those that attempt to change the environment or are delivered via the family or community settings. However, most of the interventions reviewed provided education alone, and these interventions are seldom effective. Moreover, few intervention studies have examined the moderating effects of sex, socioeconomic status, or other potentially important factors. Only five studies reviewed focused on environmental interventions. The mediators of change in physical activity behaviour are also rarely assessed or even targeted in interventions to promote physical activity in children. Overall, most interventions reviewed were delivered in schools—very few in other settings—and most involved only education. This suggests that at this stage, there is a lack of available evidence upon which to draw conclusions rather than evidence of a lack of efficacy for interventions targeting subgroups or conducted in various settings.
Future interventions must include parents and families. The review5 confirms findings from previous reviews7—that school based interventions that involve families are more likely to be effective than those that do not. Parents are the gatekeepers of children's physical activity and facilitate adolescents' physical activity by providing transport to recreational activities.8 They also have an important influence on children's sedentary behaviours.9 Hence, more interventions delivered in the family setting to promote young people's physical activity are needed.
Several potentially important physical activity behaviours were not explored in the review. Encouraging active transport is one way to increase overall levels of physical activity.10 This approach also has potential environmental and social benefits. About 20% of all car journeys during the weekday morning rush hour in the UK are thought to be short journeys undertaken by parents taking children to school.11 Children's independent mobility is greatly influenced by traffic and parents' real and perceived concerns about safety. Thus, creating environments that support local walking and cycling is a priority. More research is needed for a better understanding of the social and physical environmental determinants of young people's active transport.
There has been considerable debate about the limitations of randomised controlled trial designs in complex interventions12 and for complex behaviours. Some people argue that randomised controlled trials of interventions undertaken in microsettings have little relevance for practitioners who need to deliver population-wide effects.6 This view was partly supported by van Sluijs and colleagues' review,5 which found multicomponent interventions more effective in adolescents. Thus, despite the methodological challenges posed, more trials of complex interventions are needed.
Importantly, the interventions themselves need to be subjected to the same level of scrutiny as the study design when assessed for funding and publication, and in systematic reviews. Reviewers need to consider whether the proposed “dose” of intervention is sufficient to produce an effect and how fidelity with the proposed protocol will be (or was) assessed (for example, process evaluation). They also need to consider whether adequate formative research was undertaken or proposed, to ensure that the intervention is suited to the target group and the setting, whether the intervention is based on theory, and whether it included efforts to create a supportive physical or social environment (or both). All of these factors will contribute to the effectiveness of interventions.
In the wake of the obesity epidemic, promising multicomponent interventions need to be disseminated, while the evidence base continues to be built. High quality adequately funded evaluation of programmes based on best practice principles is also needed. Given the complexities involved, partnerships between academics and practitioners are essential.
Competing interests: None declared.
Provenance and peer review: Commissioned; not externally peer reviewed.