Introduction

Epidemiology of childhood overweight and obesity

In England, 9% of 4- and 5-year-olds are classified as obese and a further 13% are classified as overweight. By age 10 and 11 years, the prevalence significantly increases, with 20% being obese and 14% being overweight.1 Ethnic differences in childhood obesity prevalence exist, with the prevalence in 10- to 11-year-old South Asian children (those originating from the Indian subcontinent, including Pakistan and Bangladesh) at 25% compared with 18% in white children.1 Specifically, Pakistani and Bangladeshi children have higher obesity levels than white British children; at ages 10 and 11 years, 28% and 33% of Pakistani and Bangladeshi boys were classified as obese in 2014/15 compared with 19% of white British boys, and 22% of Pakistani and Bangladeshi girls were classified as compared with 16% of white British girls.2

Obesity is associated with a range of physical, psychological and social consequences in childhood,3 and up to 80% of obese children will remain obese in adulthood.4 Obesity in adulthood is a risk factor for a variety of health consequences, including certain cancers, cardiovascular disease and type 2 diabetes mellitus;5,6 thus, it is a major contributor to morbidity and premature mortality. The economic costs associated with overweight and obesity-related ill-health are considerable. Costs to the NHS in England were estimated to be £5.1B in 2006/7.7 In addition, there are societal costs associated with a reduction in years of disability-free life, increased absences from work and early retirement, increased morbidity for those of working age and reductions in productivity.8

South Asians are the largest minority ethnic population in the UK and are particularly susceptible to the health consequences of obesity. South Asians have higher levels of body fat and central obesity than white European populations9 and are more vulnerable to the cardiometabolic consequences of obesity than other ethnic groups in the UK.10 Markers of increased cardiovascular risk are seen in South Asians, even in childhood.11 Thus, South Asians represent an important target group for obesity intervention, particularly in childhood, given the disparity in obesity prevalence in children.

Effectiveness of current programmes addressing overweight and obesity in children

Evidence to date for effective childhood obesity behavioural treatment programmes is limited. In the Cochrane review conducted by Luttikhuis et al.12 in 2009, a meta-analysis of a small subset of high-quality studies showed that behavioural intervention programmes have a small but clinically significant effect on weight status at 6 months post intervention [a reduction in body mass index (BMI) z-score of 0.06 and 0.14 compared with standard care in preadolescents and adolescents, respectively]. However, the review highlighted a range of methodological issues within the included studies, such as insufficient power, lack of allocation concealment, high attrition rates and lack of intention-to-treat analysis. Therefore, it was difficult to draw firm conclusions. Clinically meaningful effects on anthropometric measures have been reported in more recent randomised controlled trials (RCTs) of behavioural intervention programmes13–16 and in a comprehensive evidence review that informed the 2013 National Institute for Health and Care Excellence (NICE)’s guidance on children’s weight management [public health guideline (PH) number 47],17 which included meta-analyses to estimate the pooled intervention effect of behavioural programmes targeting children and their parents or families. A 0.2 reduction in BMI z-score at the end of the intervention and a 0.1 reduction at the 6-month follow-up compared with standard care was estimated from 8 and 11 studies, respectively.17 Evidence from studies suggests that even such small reductions in BMI z-score could lead to clinical improvements.18 The absence of data on the cost-effectiveness of childhood obesity behavioural treatment interventions and the lack of data on longer-term outcomes have been highlighted.12,19

Critical elements of effective programmes

Owing to limitations of the evidence, the most effective intervention components of behavioural child weight management programmes are, as yet, undetermined. In most studies to date, the measurement of processes and an assessment of the fidelity of programme delivery have not been undertaken or reported,12,20 which limits the interpretation of outcomes and assessment of which elements of the programme are likely to have had the most influence. However, the available evidence suggests that, in the preadolescent age group, interventions that address both diet and physical activity, include behavioural elements and involve parents are the most promising,12,20,21 although the specific nature of this parental engagement still needs to be determined.21,22

Evidence of the clinical effectiveness of overweight and obesity management programmes in minority ethnic children

Evidence evaluating the clinical effectiveness of childhood obesity treatment programmes in minority ethnic populations is also limited and studies were undertaken mainly in the USA.23–26 Existing studies have evaluated programmes with adaptations such as the delivery of materials in different languages, the tailoring of nutritional content and the ethnic matching of programme providers. Two RCTs that evaluated culturally adapted interventions, one targeting Chinese American children aged 8–10 years24 and the other targeting a mixed population of Hispanic, black and white children aged 8–16 years,25 have reported small to moderate reductions in BMI z-score in the intervention compared with the control groups (controls received no intervention and a low-intensity clinic-based intervention, respectively). Effects appeared to be sustained for at least 8 months. Another uncontrolled study in a Latino community in the USA reported positive effects on dietary and physical activity behaviours but no effect on BMI.26 Furthermore, there is evidence to suggest that interventions delivered in multiethnic populations may have a differential effect; a recent RCT of a multidisciplinary intervention targeting obese adolescents in the Netherlands reported a 0.35 reduction in BMI z-score over 18 months in participants of Western descent but no BMI z-score reduction in participants of other ethnicities.27

In the UK, one small RCT (n = 72) has been undertaken to evaluate the clinical effectiveness of a family-based behavioural treatment programme targeting obese children in an ethnically and socioeconomically diverse community of 8- to 12-year-old children (of whom 43% were non-white European). The programme, originally developed in the USA,28 was not culturally adapted and, although acceptable to the target population,29 it did not have a significant effect on weight.30

Retention in children’s weight management programmes

Previous studies suggest that there is an association between better programme attendance and the weight loss achieved in children’s weight management programmes;14 however, retention of participants has been highlighted as a problem. Programme characteristics that are associated with greater dropout include having large group sizes and logistical barriers to attendance.31,32 Lower retention in programmes has also been reported to be associated with socioeconomic disadvantage32 and ethnicity.33,34

Development of culturally adapted children’s weight management programmes

To date, theoretical approaches to the cultural adaptation of child weight management programmes have been lacking,23 and little has been done in the way of evaluating the success of cultural adaptation of obesity treatment and health promotion programmes in general. In a comprehensive evidence synthesis review on the adaptation of health promotion programmes for minority ethnic groups, Liu et al.35 stated the requirement for future trials to compare culturally adapted programmes with standard programmes.

Provision of childhood obesity treatment services in the UK

Although in the last 15 years there has been widespread provision of children’s weight management services across the UK,36 no programmes have been specifically designed to meet the needs of ethnically and culturally diverse communities, such as those living in many large cities in the UK. In Birmingham, the UK’s second largest city, 42% of the total population and 59% of the 0- to 15-year-old population are from minority ethnic communities. More than half of these residents are of South Asian ethnicity, although the great majority were born in the UK.37

Since 2010, a child weight management programme, First Steps (a weekly programme delivered over a school half-term that targets parents), has been available in Birmingham. Participants are referred through health professionals, schools or the National Child Measurement Programme (NCMP) or are self-referred. Over one-third of the referrals received by the service have been for children and families from Pakistani and Bangladeshi communities, but an analysis of routine service data has highlighted differences in completion rates between different ethnic groups. Pakistani and Bangladeshi families are as likely to start attending the programme as other families but are less likely to complete the programme, which suggests that it is less well suited to families from these communities. Of those starting the First Steps programme, 40% of Pakistani and Bangladeshi families completed it compared with 65% of white British and African-Caribbean families.

Although there may be a number of reasons for this difference in completion, it may imply that aspects of the current programme are not culturally relevant to Pakistani and Bangladeshi families. The existing weight management programme in Birmingham has developed over time, based on research evidence and service provider experience, to try to make it more acceptable to the community that it serves. Prior to the introduction of First Steps in Birmingham, established and evidence-based child weight management programmes, such as Mind, Exercise, Nutrition . . . Do it! (MEND)13 and Watch It,38 had been commissioned. However, the intensity and duration of the courses resulted in very low uptake rates and high attrition, and service providers identified that there was a lack of flexibility to tailor the programmes to accommodate cultural and language requirements. The First Steps programme was designed to be more suited to the local population by placing a greater focus on parental engagement, increasing the visual programme content, including culturally appropriate foods in the programme materials and providing interpreters. Elements of the Watch It and MEND programmes that were observed to have worked in the local population were retained in First Steps (specific behaviour change strategies, and some nutrition content). Other elements that were incorporated were consistent with the evidence base. Routinely collected data at the first and last sessions of First Steps indicated that those who completed the programme achieved an average reduction in BMI z-score of 0.1 at programme end, which is comparable to effect sizes reported in clinical trials.12 Given the evolution of the First Steps programme in the local ethnically diverse population, it provided a good foundation on which to develop a new programme that incorporated further cultural adaptations.

Rationale for the CHANGE study

The focus of the Child weigHt mANaGement for Ethnically diverse communities (CHANGE) study was to further develop and culturally adapt the current First Steps programme, using a theoretically informed approach, such that it better met the needs of the Pakistani and Bangladeshi communities within Birmingham. However, a programme delivered to families from only these communities would not address the needs of the wide range of cultural communities found in a city such as Birmingham. Communities across Birmingham provide an example of super-diversity, a term used to characterise the complexity of modern communities in Britain, in which there are dynamic relationships between multiple variables, including country of origin, ethnicity, language, religion, regional/local identities, migration history and experience, and immigration status.39,40 This complexity creates a challenge to health-care providers to meet the health needs of all members of society. Thus, this study aimed to develop a programme that was culturally adapted to meet the needs of Pakistani and Bangladeshi families but that would also be flexible enough to accommodate the needs of all families and could be transferred to communities with a different cultural and ethnic composition. To ensure that the developed programme was suitable for all, we evaluated it in a feasibility study involving families from different ethnic and cultural backgrounds.