Assessing the impact of obesity interventions in the early years: a systematic review of UK-based studies

Parents, acting as public contributors, were mixed in their views around the potential impact and effectiveness of obesity interventions in early childhood. Half (n=3) felt that interventions in the early years could be effective, while half were sceptical. All parents thought that preventative interventions had the potential to have the greatest impact and the majority (n=5) were in favour of multicomponent interventions that incorporated dietary and physical activity components, giving that the goal in early life is to encourage healthy lifestyles. Parents reported mixed views in relation to intervention components; three parents favoured educational interventions around living healthier lifestyles, two favoured motivational approaches and one favoured addressing environmental factors to enable healthy growth among children. In terms of the interpretation of findings, all parents were surprised at how few interventions with robust evaluation there were in the UK. All parents reported that the cost of living a healthy lifestyle (eg, buying fresh food) was an important factor and suggested there should be financial help or subsidies for parents of young children, particularly families with low income. There was consensus among parents that interventions should be properly evaluated and better funded to ensure they are robust; for example, by following participants up over longer time frames. There was also consensus among parents that families, early years practitioners (eg, health visitors) and parents of children living with excess weight should be consulted when developing interventions in early childhood.

Discussion

The aim of this systematic review was to identify and evaluate interventions and programmes designed to prevent or treat obesity among children aged 6 months5 years, implemented and evaluated in the UK. Six studies met the inclusion criteria, all of which included a combination of diet and physical activity components. Compared with baseline, three interventions reported a reduction in zBMI immediately after completing the intervention37 38 and four interventions a while after the intervention end,34 36 38 with the effect ranging from −0.02 to −0.9 for the preventative interventions and from −0.3 to −0.5 for the treatment ones. Three of these reductions were significant34 37 38 (two of which targeted children with overweight/obesity).37 38 Compared with the control conditions, five interventions showed a reduction in zBMI,34–36 38 of which one was significant.34

This is the first systematic review that focuses on childhood obesity interventions among children aged 6 months5 years evaluated in the UK. Our findings support existing evidence from international reviews that multicomponent interventions, with both dietary behaviours and physical activity elements, are effective in reducing zBMI among preschoolers.26 39 40 Small effects were found at individual level; however, if the interventions are delivered at population level, the effects can be larger and meaningful.41 The present work contributes to the research by synthesising the evidence of recent interventions and assessing their effects on children of any weight status or on children with overweight and obesity.

The current review highlights a lack of UK interventions with robust evaluation. Only Planet Munch was evaluated through a randomised controlled trial. ToyBox and HENRY were evaluated with feasibility cluster randomised trials without a follow-on full trial being published, and these feasibility trials were not adequately powered to detect zBMI changes. Methodological limitations across the studies identified, such as unclear adherence to the intervention, limit our ability to draw firm conclusions about their impact. Evaluating interventions is complex; it is important not only to ascertain impact, but also when, how and in which conditions and for whom interventions work. A theoretical framework, a process to identify failures in implementation, an assessment of the sample and a range of outcomes, in addition to adaptation to local settings, are important contributory factors to the effectiveness and applicability of an intervention.42 43

Of the studies identified in this review, only two were followed up beyond 12 months34 38 meaning there is very little evidence for longer-term effects. All the studies included in this review were small-scale evaluations and not easily generalisable. While effect sizes give some measure of clinical importance (if adequately powered), statistical significance is affected by the sample size and so has limited importance.44 A reduction in zBMI may be clinically relevant when setting a child on a healthier growth trajectory and improving health outcomes.45

Overall, we found a lack of evaluations of interventions that recorded anthropometric outcomes. There were evaluations that reported indirect indicators of intervention impact including reductions in calorie intake or increases in physical activity; however, these indicators can be prone to bias. Children and parents often underreport their energy intake,46 ,
47 with greater underreporting often observed among children living with obesity.48 Subjective measures of physical activity also often have low validity and may be biased due to poor recall or social desirability.49 There are evaluations of interventions (eg, the HENRY programme) that show effectiveness in changing energy balance behaviours50 but have not been included in this review.

Within a whole-systems approach, local authorities can play a leading role in influencing health behaviour. However, it is important to recognise that financial constraints can limit local action to predominantly statutory services. Owing to these and other limitations, many local authorities have developed interventions relevant to the early years which have not been subject to robust assessment. These are often brief interventions (eg, cooking classes or baby yoga), for which there is no evidence of meaningful and sustained lifestyle changes.

Published studies also had a high risk of bias, limiting the confidence in their results. Additionally, findings were difficult to compare owing to heterogeneity in study designs, intervention approaches and outcome reporting. Some of the reported reductions in zBMI should be interpreted with caution. For example, children that received the HENRY intervention started the programme with higher mean zBMI than those in the control group, which may have influenced the effect size. This highlights the need for appropriately designed RCTs with stated primary outcomes, in order to evaluate obesity interventions. Indicators of eating behaviour, physical activity and other energy balance behaviours are important, anthropometric measures remain the most robust and reliable measure of intervention effectiveness.

Findings from this review suggest that interventions that promote healthy growth in the early years could be effective in reducing the risk of obesity as part of a programme of policies and interventions for young children and families. However, all studies would benefit from evaluation using large-scale RCTs before firm conclusions can be drawn. Unpublished data were provided for two trials relating to the Planet Munch (formerly Trim Tots) intervention, which found reductions in zBMI for children living with overweight or obesity or who were at high risk. Importantly, reductions were maintained at longer-term follow-up, up to 24 months after completing the intervention. Planet Munch was the only preschool obesity intervention in the UK that complied with all NICE recommendations that is, provided advice on how to achieve a healthy diet, encouraged physical activity and included behaviour change strategies in the whole family.

The LEAF Programme was also effective in reducing zBMI but was only delivered to children with severe obesity and was not evaluated in an RCT. Our findings are consistent with previous work,26 which found obesity management interventions to be more effective compared with preventative interventions in young children. This is an expected outcome, considering the different populations in treatment and prevention trials. Treatment interventions aim to slow weight gain and normalise zBMI while preventative interventions aim to maintain a healthy growth trajectory. Both aim to promote healthy lifestyle behaviours.

This systematic review has limitations. The initial search strategy included international evidence without applying limits on geographical location or language and thus yielded a high number of records. We used EPPI-Reviewer software to apply an active learning approach, which greatly reduced screening time but resulted in 12 004 records being excluded without being screened. Though we also conducted a hand search, there may be relevant studies that have not been included in this review. We found but did not include a child-care self-assessment intervention that aimed to improved physical activity, oral health and nutrition for children aged 2–4 years and was delivered in nurseries.51 The study was published but not formally peer reviewed, meaning it did not fulfil our criteria for inclusion. We also excluded studies published before 2011; however, previous systematic reviews have not included any UK studies conducted prior to that point, so it is unlikely that any robustly evaluated interventions had been conducted.

Only a small number of preschool obesity interventions have been implemented and evaluated in the UK and a smaller number have been robustly evaluated in randomised controlled trials. Meanwhile, obesity prevalence and the health inequalities remain high in the UK. Obesity policies and interventions are not always implemented or delivered at scale, and implemented interventions are rarely evaluated robustly. Interventions often target individual behaviour changes without addressing the structural and systemic determinants of obesity.52 The evaluation outcomes of some interventions are encouraging, but more evidence is needed via larger-scale trials. However, scaling-up interventions can be challenging, as adaptations are often required, to meet real-world contexts. Intervention effects can also decrease when applied to the general population, so adaptations should consider the context, implementation and setting of the intervention.50 53 More research is required to understand what support might be most useful to UK preschoolers and their families.

This post was originally published on https://bmjopen.bmj.com