The review will consider public health strategies which reduce existing inequalities in the prevalence of obesity as well as those interventions that might prevent the development of inequalities in obesity. The review will also serve as a mapping exercise of the types of interventions that have been evaluated in relation to tackling inequalities in relation to obesity amongst children, thereby highlighting any gaps in the evidence base. The review will also seek to establish how public health interventions which might reduce or prevent inequalities in obesity are organised, implemented and delivered. Context is increasingly recognised as an important factor in the success of public health interventions . However, the assessment of implementation has not really featured strongly in previous obesity reviews. We will therefore develop, refine and apply existing methodological tools which assess the implementation of complex public health interventions .
The study design inclusion criteria in the review are broad, given that whilst trials of individual, and even community, level interventions are likely, we expect a dearth of experimental studies in relation to societal level interventions. Indeed, large evaluations, such as those of Change4Life (England) and Ensemble, Prévenons l'Obésité des Enfants (EPODE; France), have all used a repeat cross-sectional design. This is perhaps because, as Law and colleagues observed , societal level interventions tend not to be easily evaluated using experimental study designs. Furthermore, other recent systematic reviews of the effects of societal level public health interventions on socioeconomic inequalities in relation to health have located few relevant experimental studies .
We anticipate that our extensive search strategy, combined with the inclusive study design criteria, will ensure that a sizeable literature will be located for synthesis. Recent Cochrane Heart Group reviews of interventions that prevent and treat obesity amongst children found 22 and 64 randomised controlled trials, respectively [22, 33]. Whilst we acknowledge that the literature on the effects of interventions on health inequalities is likely to be smaller, we will maximise the likelihood of locating relevant studies by taking a more inclusive approach to study design, contact study authors for unpublished data on health inequalities, and evaluate interventions targeted at deprived groups or areas as well as studies that include comparative data on the effects of interventions on differential impacts across two or more socioeconomic groups. The size of the available evidence base will also be extended, because we will look at different levels of intervention: individual, community and societal. We will also examine the full papers of all studies which fit our population, intervention, design and health outcome inclusion criteria, even if there is no mention of socioeconomic inequalities in the abstract. By adopting this strategy, we will be less likely to exclude studies which undertook subgroup analyses by socioeconomic status but did not publish the findings in the abstract. We will then contact the study authors for possible subgroup analyses and request any additional unpublished data on health inequalities. This will increase the comprehensiveness of the search strategy and therefore the quality of the final synthesis.
Once the evidence has been synthesised, an 'implications for policy and practice' review dissemination workshop will be held with invited NHS commissioners whose responsibilities include obesity, Department of Health policy-makers with responsibilities for obesity and inequalities, user group representatives (for example, community groups, schools and employer organisations, trade union congresses), as well as UK research network representatives (for example, the Faculty of Public Health, Nutrition Society and the United Kingdom Public Health Association) to discuss the results, aid in the write-up and facilitate the translation of the findings into practice. The technical report and executive summary will then be finalised, and a short 'key findings' summary of the research will be sent to relevant stakeholders. The research will be disseminated via national and international academic and/or practitioner cross-over conferences, and a policy-orientated summary paper will be published on an open access basis so that it is freely available to practitioners and the public.