The review will follow the same procedure carried out for the systematic review of the effectiveness of public health interventions at reducing socioeconomic inequalities in obesity among children
. The review will be carried out following established criteria for the good conduct and reporting of systematic reviews
[20, 21]. A Study Steering Group comprising key stakeholders from the UK policy and research communities, international representatives, a statistician and a health economist will guide the research. The review is registered with the PROSPERO International Prospective Register of Systematic Reviews (registration number: CRD42013003612).
This project has two objectives:
To systematically review the effectiveness of public health interventions (individual, community and societal) in reducing socioeconomic inequalities in obesity among adults.
To establish how such public health interventions are organized, implemented and delivered.
The review will examine public health interventions at the individual, community and societal level that might reduce inequalities in obesity among adults aged 18 years and over, in any setting, in any country. The review will utilize the intervention framework (Table
1) and group interventions as individual, community or societal with acknowledgement that some interventions might be cross-cutting. For example an individual level intervention could be a relatively complex weight management program providing dietary and physical activity education with behavioral counseling delivered individually (as opposed to group-based) or could simply be a gift card given to low-income families for the purchase of fruits and vegetables. An example of a community level intervention may be access to a community fitness center (with/without the addition of physical activity education), or a workplace health promotion campaign involving dietary and physical activity education. A societal level intervention may also be a workplace health promotion campaign but one that involves an environmental change such as healthy food choices in the work canteen. The review will consider public health strategies which might reduce existing inequalities in the prevalence of obesity as well as those interventions which might prevent the development of inequalities in obesity. However, clinical interventions such as those using drugs or surgery and laboratory-based studies will be excluded. We will also exclude studies of interventions designed for adults with a critical illness or severe comorbidities (such as diabetes, metabolic syndrome, hypertension and mental health disorders).
A rigorous and inclusive international literature search will be conducted for all randomized and nonrandomized controlled trials, prospective and retrospective cohort studies (with/without control groups) and prospective repeat cross-sectional studies (with/without control groups) of the effectiveness of public health interventions at reducing inequalities in childhood obesity. Studies with a duration of at least 12 weeks (combination of intervention and follow up) will be included; an inclusion criterion used in previous Cochrane reviews of interventions aimed at preventing obesity in children
 and of the effectiveness of exercise for weight loss in adults with overweight or obesity
The search strategy (Additional file
1) will include the following electronic database searches (host sites given in parentheses) MEDLINE (Ovid), EMBASE (Ovid), CINAHL (EbscoHost), PsycINFO (EbscoHost), Social Science Citation Index (Web of Science), ASSIA (CSA), IBSS (EBSCO), Sociological Abstracts (CSA), and the NHS Economic Evaluation Database (NHS CRD). The skills of a trained information scientist (HJM) will be used to develop and implement the electronic searches. All databases will be searched from start date to present. We will not exclude papers on the basis of language, country or publication date.
We will supplement the electronic database searches with website and gray literature searches. We will hand search the bibliographies of all included studies and request relevant information on unpublished and in-progress research from key experts in the field. In addition, we will hand search the last two years of the most common five journals revealed by the electronic searches as well as journals identified by experts in the subject area.
In terms of outcomes, we will only include studies if they include a primary outcome that is a proxy for body fatness (weight and height, body mass index, waist measurement/waist to hip proportion, percentage fat content, skin fold thickness, and ponderal index in relation to childhood obesity). Data on related secondary outcomes (such as physical activity levels; dietary intake; blood results, such as cholesterol and glucose levels) will also be extracted from those studies which have a primary outcome. We will include both measured and self-reported outcomes. Studies will only be included if they have examined differential effects with regard to socioeconomic status (education, income, occupation, social class, deprivation, and poverty) or the intervention must have been targeted specifically at disadvantaged groups (for example, unemployed, low SES and low income) or deprived areas. Data on the organization, implementation and delivery of interventions will be extracted using existing methodological tools which assess the implementation of complex public health interventions
, adapted and refined for the purposes of this review. Examples of the implementation components that will be examined include: theoretical underpinning; implementation context; experience of intervention team (planners and implementers); consultation/collaboration processes (planning and delivery stages); and resources (for example, time, money, staff, and equipment).
Data extraction and quality appraisal
The initial screening of titles and abstracts and full paper inclusion will be conducted by one reviewer (FCH or JMC) with a random 10% of the sample checked by a second reviewer (HJM or FCH). Data extraction and methodological quality appraisal of the included studies will be conducted by one reviewer (FCH, JMC, CLB, or CDS) using established data extraction forms
[20, 25–30] and will be checked by a second reviewer (FCH, JMC, CLB, or CDS). Any discrepancies will be resolved through discussion between the authors and, if consensus is not reached, with the project lead (CLB). The methodological quality of the included studies will also be appraised using the Cochrane Public Health Review Group recommended Effective Public Health Practice Project Quality Assessment Tool for Quantitative Studies
, which includes, among other things, an examination of sampling strategy, response and follow-up rates, intervention integrity, statistical analyses and assessment of adjustment for confounders. We will use the quality appraisal criteria for descriptive purposes and to highlight variations between studies.
Analysis and synthesis
Where possible, meta-analysis will be used to synthesize data using Comprehensive Meta-Analysis (Biostat, Englewood, NJ, USA) based on the primary outcomes. A fixed effect model will be used for the meta-analysis, unless there is evidence of heterogeneity between studies, in which case a random effect model will be used. The presence of heterogeneity will be investigated with the use of likelihood ratio test statistic, while funnel plot will be considered to explore publication bias. However, where meta-analysis is not possible, narrative synthesis will be conducted. We will report our analyses in accordance with PRISMA guidelines
. The main analysis will examine the effects of 1) individual, 2) community and 3) societal level public health interventions on socioeconomic inequalities in obesity, using the multi-dimensional framework outlined in Table
1. Where data permits, we will conduct demographic subgroup analysis by age, gender and ethnicity.