A novel analytic framework to investigate differential effects of interventions to prevent obesity in children and young people

BackgroundRecent systematic reviews and meta-analyses on the effects of interventions to prevent obesity in children aged 5 to 18 years identified over 200 randomized trials. Interventions targeting diet, activity (including physical activity and sedentary behaviours) and both diet and activity appear to have small but beneficial effects, on average. However, these effects varied between studies and might be explained by variation in characteristics of the interventions, for example by the extent to which the children enjoyed the intervention or whether they aim to modify behaviour through education or physical changes to the environment. Here we develop a novel analytic framework to identify key intervention characteristics considered likely to explain differential effects. ObjectivesTo describe the development of the analytic framework, including the contribution from school-aged children, parents, teachers and other stakeholders, and to present the content of the finalized analytic framework and the results of the coding of the interventions. Design and methodsWe first conducted a literature review to find out from the existing literature what different types of characteristics of interventions we should be thinking about, and why. This information helped us to develop a comprehensive map (called a logic model) of these characteristics. We then used this logic model to develop a list of possible intervention characteristics. We held a series of workshops with children, parents, teachers and public health professionals to refine the list into a coding scheme. We then used this to code the characteristics of each intervention in all the trials which aimed to prevent obesity in children aged 5 to 18 years. FindingsOur finalized analytic framework included 25 questions across 12 characteristics. These addressed aspects such as the setting of the intervention (e.g. at school, at home or in the community), mode of delivery (e.g. to individuals or to groups children), whether the intervention targeted diet and/or activity, complexity (e.g. focused on a single swap of juice for water or aimed to change all aspects the diet), intensity, flexibility, choice, mechanism of action (e.g. through participation, education, change in the social environment, change in the physical environment), resonance (e.g. credibility of the person delivering the intervention), commercial involvement and the fun-factor (as perceived by children). We coded 255 interventions from 210 randomized trials. ConclusionsOur evidence-based analytic framework, refined by consulting with stakeholders, allowed us to code 255 interventions aiming to prevent obesity in children aged 5 to 18 years. Our confidence in the validity of the framework and coding results is increased by our rigorous methods and, especially, the contribution of children at multiple stages. FundingThis article presents independent research funded by the National Institute for Health and Care Research (NIHR) Public Health Research programme as award number 131572. Plain language summaryMore children and adolescents worldwide are developing overweight and obesity. Being overweight at a young age can cause health problems, and people may be affected psychologically and in their social life. Children and adolescents living with overweight are likely to stay that way or develop obesity as adults and continue to experience poor physical and mental health. It is important to understand whether attempts to help children and young people modify their diet or activity levels (or both) reduce the chance that they develop obesity. In previous work we found that over 200 randomized trials have been done in people aged 5 to 18 years. These examine different strategies to try and prevent obesity. Whilst we found that these strategies have small beneficial effects on body mass index (BMI) on average, a notable finding was that there was a lot of variation in their results across the studies. We want to understand what causes some strategies to be more effective than others. To do this we need to re-analyse the results of the studies. To inform this analysis, we developed a list of key characteristics that we and others thought would be likely to explain the variability in effects. We used this list to code over 250 strategies that have been studied. The development process included review of literature and patients/public involvement and engagement (PPIE) that is extensive consultation with children, young people, parents, schoolteachers and public health professionals. Our final list included features such whether the strategy was based at school or in the home, whether the strategy targeted diet or activity, how long and intense the strategy was and how flexibly it could be implemented. We also included the fun-factor of engaging with the intervention, for which we invited children and young people to help us out with the coding.