7.1 Introduction

The West Midlands Key Health Data 2005/06 ⁽¹⁾ included a chapter on childhood obesity focussing on the causes of obesity and issues surrounding the use of Body Mass Index as an indicator of obesity in children. This chapter continues to explore and expand on the theme by presenting up to date regional data based on recent schools-based obesity monitoring and outlining a Birmingham based prevention research study.

7.2 Monitoring Obesity in School Children: Findings from the 2006 Data Collection from West Midlands PCTs

In July 2004 the Government responded to the challenge of childhood obesity by developing a public service agreement (PSA) target “to halt the year on year rise in obesity among children aged under 11 by 2010”. In November 2004 the White Paper “Choosing Health” ⁽²⁾ charged the Department of Health and the Department for Education and Skills with developing appropriate systems for recording lifestyle measures among school age children. In order to be able to monitor trends in childhood obesity, and track progress towards this target, we need high quality, up to date information on children’s height and weight. In 2006 the Department of Health directed all Primary Care Trusts (PCTs) to introduce routine annual height and weight monitoring for primary school children. Guidance issued prior to the implementation of monitoring stated that the purpose of gathering height and weight data is to enable population monitoring and not to screen individual children or initiate any intervention for individuals ⁽³⁾. Children in Reception year (aged 4 or 5) and Year 6 (aged 10 or 11) are included in the process. The first round of monitoring took place in the summer term of 2006. At it’s outset this process was termed “Childhood Obesity Monitoring” but has recently been renamed the “National Child Measurement Programme”. 

Defining Childhood Obesity for Schools-based Monitoring

In adults the Body Mass Index (BMI) is used to identify whether someone is underweight, normal weight, overweight or obese. The BMI is calculated using a person’s height and weight (kg/m²). For an adult, a BMI of 18 - 24.9 is considered normal, 25 to 29.9 overweight and 30 or more as obese. However, for children the BMI is harder to interpret because they are still growing. In 1990 a database of information about a large sample of UK children was constructed including height, weight, age and sex. The BMI for each child was calculated. Children in this dataset with a BMI in the top 5% for their age and sex were considered obese and this 5% percentile sets the cut-off point against which children’s BMIs are compared for epidemiological purposes. A BMI which falls above that 95% cut-off point (i.e. is in the top 5%) is indicative of obesity. Although obesity is the focus of the PSA target, monitoring the prevalence of overweight in children is also important. Children in the 1990 dataset with a BMI between the 85^th and 95^th percentile were considered overweight. In the 2006 monitoring process, children with a BMI above the 85% 1990 percentile and below the 95^th percentile were identified as overweight.

The 2006 Monitoring Process

Organising the 2006 monitoring process presented a considerable challenge to PCTs due largely to the short time available in which to complete the data collection. Local monitoring arrangements differed between PCTs, but the same underlying principles applied.  In most cases, all Local Authority primary schools were invited to participate. Schools could opt out of the process if they wished, although this was generally rare in the West Midlands. Parents and children could also opt not to participate. Clearly the process of measuring children must be handled sensitively. In the majority of cases, the height and weight monitoring was carried out by School Nurses on school premises although a number of other staff were involved. The data collected in schools was collated by the relevant PCT before being uploaded onto the National Childhood Obesity Database (NCOD). The national database assigned a BMI to each child which allowed children to be categorised as normal weight, overweight or obese. NCOD is a secure restricted-access database which allows PCTs to download summaries of the results of the monitoring. These summaries do not contain any identifying details or results for individual children, but provide the proportions of obese and overweight children according to a range of factors including school year and sex. It should be noted that, subsequent to the analysis of the data from the first year of monitoring, improvements have been made to data collection processes including changes to the national database and to guidance provided to assist PCTs with data collection.

Reorganisation of West Midlands PCTs

Prior to October 2006 there were 30 PCTs in the West Midlands, and it was these organisations that organised and carried out the 2006 monitoring. In October 2006 an NHS reorganisation meant that a number of PCTs merged to create a total of 17 PCTs across the region. Although monitoring data were obtained according to the former PCT areas, in order to facilitate the use of the data in the context of current organisational structures, data have been collated to provide figures for the current PCTs.

Data Completeness and Caution in Interpreting Data

Achieving adequate levels of data completeness is essential if reliable figures are to be obtained. Tables 7.1 and 7.2 describe the completeness of data for Reception and Year 6 children in West Midlands PCTs. One PCT did not undertake monitoring of Reception Year children and 5 PCTs did not undertake monitoring in Year 6.

Data completeness was calculated based on the number of children identified as eligible for monitoring submitted to NCOD. Where data were available the levels of completeness vary widely from 16.2% to 88.1% for Reception Year (average 73.9%), and from 3.8% to 90.1% in Year 6 (average 56%). Children may not have been measured because their school, parents or they themselves opted out, they were absent on the day of measuring, or if measurement could not be carried out for another reason. There is anecdotal evidence that overweight or obese children are less likely to participate in the monitoring than those at a normal weight ⁽⁴⁾. It is not possible to distinguish the different causes for not being measured from the available data.

Given the variation in completeness, the low levels of completeness achieved by a number of PCTs, and the possible participation bias, rates of obesity and overweight calculated from the figures must be viewed with caution.

This cautious approach is recommended in a national analysis of 2005-06 childhood obesity data ⁽⁴⁾. At national level, obesity data were available for 57% of eligible Reception year pupils and for 42% of Year 6 pupils, with wide variation between Strategic Health Authorities and PCTs.

Table 7.1: 2006 Childhood obesity Monitoring data for West Midlands PCTs
Children Measured and Data Completeness: Reception Year

2006 Childhood obesity Monitoring data for West Midlands PCTs Children Measured and Data Completeness: Reception Year

Table 7.2: 2006 Childhood Obesity Monitoring data for West Midlands PCTs Children Measured and Data Completeness: Year 6

2006 Childhood Obesity Monitoring data for West Midlands PCTs Children Measured and Data Completeness: Year 6

Rates of Obesity and Overweight

Table 7.3 shows obesity rates in Reception and Year 6 children for all West Midlands PCTs from the 2006 data collection exercise

Care must be taken in any comparison of obesity rates between areas, and especially if attempting to draw conclusions based on comparisons of obesity rates at small area level due to the low levels of completeness and small numbers involved. Particular care must be taken when attempting to use obesity rates at individual school level, as the numbers of children involved can be very small and large variations in obesity rates are common. One approach that avoids the pitfalls of releasing school-level obesity data was piloted in South Staffordshire, where Control Chart methodology was used to analyse school-level rates, with the small number of schools with outlying obesity rates being followed up ⁽⁵⁾. This approach allowed the PCT to utilise all the obesity data available to contribute to local public health action planning, without the potential negative effects of releasing school-level obesity rates. 

Across West Midlands PCTs the 2006 monitoring data suggest that obesity rates in the region are in line with the national pattern:-

• 10.1% of Reception Year children in the West Midlands are obese compared to 10.0% for England (Chi Square p = 0.38)
• 17.6% of Year 6 children in the West Midlands are obese compared to 17.3% for England (Chi Square p = 0.36)

Table 7.3: 2006 Childhood Obesity Monitoring data for West Midlands PCTs Obesity rates in Reception and Year 6

2006 Childhood Obesity Monitoring data for West Midlands PCTs Obesity rates in Reception and Year 6

Table 7.4: 2006 Childhood Obesity Monitoring data for West Midlands PCTs Overweight rates in Reception and Year 6

2006 Childhood Obesity Monitoring data for West Midlands PCTs Overweight rates in Reception and Year 6

Table 7.4 shows that, across West Midlands PCTs the 2006 monitoring data suggest that rates of overweight in the region are also in line with the national pattern:-

• 12.5% of Reception Year children in the West Midlands are overweight compared to 12.8% for England (Chi Square p = 0.10)
• 14.3% of Year 6 children in the West Midlands are overweight compared to 13.8% for England (Chi Square p = 0.08)

Conclusions

The 2006 childhood obesity monitoring process was the first time that this exercise was carried out at a national level. Despite the challenges that this presented, in particular due to the short timescale available for organising and carrying out monitoring, the majority of PCTs were able to contribute data. Despite the limitations of the data described, this national dataset provides a valuable initial picture of the prevalence of obesity and overweight in children in England. The West Midlands displays prevalence levels in line with the National findings.

The experience from the 2006 data collection showed that there are things that could be done differently to make the monitoring process easier to carry out and to improve the reliability of the data. The Department of Health has stated that a minimum of 80% coverage in both year groups will be required. In light of this PCTs are looking at ways to encourage more schools to participate, have fewer children opt out, and ensure that the practical arrangements made with schools enable as many children to participate as possible. As well as being an important National exercise, the Child Measurement Programme provides an opportunity to strengthen local working relationships between PCTs, schools and other partners in promoting healthy lifestyles for children.

The second round of the monitoring process will be completed by the end of the summer term 2006/07. As improvements are made to the data collection process and further monitoring data are obtained this resource will become increasingly important in tracking how obesity rates change at national and local level.

7.3 Preventing Obesity in Childhood

Systems of monitoring trends in childhood obesity and tracking progress towards the Government’s PSA target are now in place, but in order to achieve the target, effective ways of preventing obesity need to be identified and implemented across the country. This section outlines a current research study into the development of an effective childhood obesity prevention intervention that is based in Birmingham.

Many interventions aimed at preventing childhood obesity have been developed and evaluated. These range from relatively simple to multi-faceted complex interventions, and encompass components aimed at improving diet, increasing physical activity, reducing sedentary behaviours and improving general health. A variety of settings have been used for delivering prevention interventions, including family, pre-school, school, community and health care settings ⁽⁶⁾.

Despite the extensive research into interventions aimed at preventing childhood obesity, there is minimal evidence to support their effectiveness. The National Institute for Health and Clinical Excellence (NICE) published a clinical guideline entitled ‘Obesity: the prevention, identification, assessment and management of overweight and obesity in adults and children’ in December 2006 ⁽⁶⁾. NICE reported that many studies looking into obesity prevention had an inadequate duration of follow up, were conducted outside of the UK and were poorly reported. The guideline also stated that there was a dearth of evidence looking at obesity interventions aimed at key ‘at-risk’ groups in the UK, such as young children and certain ethnic minority groups. NICE concluded that there was a further need for well designed trials of interventions to tackle obesity, with a period of follow up of at least 12 months.

Birmingham Healthy Eating and Active lifestyle for Children Study (BEACHeS)

The University of Birmingham Department of Public Health and Epidemiology are currently undertaking the Birmingham healthy Eating and Active lifestyle for Children Study (BEACHeS), a research study funded by the National Prevention Research Initiative (a multi-disciplinary initiative aimed at supporting high quality research into disease prevention).

The purpose of the BEACHeS study is to develop and pilot an intervention package aimed at preventing obesity in children aged 6-8. The study is focussing on this relatively young age group as obesity appears to be established in the pre-pubescent years, and once established, is likely to persist into adulthood ⁽⁷⁾.

The study aims to develop an intervention that will have an impact on children from all ethnic backgrounds but there is a focus on South Asian children, as the UK, and in particular, Birmingham, has a substantial South Asian population (4.1% of the UK population and 18.5% of the Birmingham population were Indian, Pakistani or Bangladeshi in 2001 ⁽⁸⁾). South Asians are particularly vulnerable to the health consequences of obesity (heart disease, type II diabetes) ⁽⁹⁾ and so it is important to ensure effectiveness of obesity prevention interventions in these ethnic groups.

BEACHeS Study Methods

The study commenced in September 2006 and is being conducted in two phases over a 3 year period. These phases are based on the Medical Research Council framework for the development and evaluation of complex interventions ⁽¹⁰⁾. Eight primary schools in Birmingham with a greater than 50% proportion of South Asian pupils have been recruited.

The aim of the first phase is to develop an intervention package using a combination of information from the scientific literature, ‘lay knowledge’ from stakeholders related to the participating schools and surrounding communities and expert input from a range of professional stakeholders. Various stakeholders are attending focus groups so that their views on potential interventions can be explored. The intervention will be developed and delivered during the next school year.

The second phase comprises an exploratory trial to evaluate the feasibility, acceptability and effectiveness of the intervention package. The developed intervention will be delivered to half of the participating schools and communities during the next school year.

Children in years 1 and 2 from the eight participating schools have been assessed in a variety of ways during 2006-7 to provide a baseline, and the follow up measures will be undertaken the year after delivery of the intervention (2008-9). Assessments include BMI, waist circumference, skinfold thickness, bioimpedance analysis, blood pressure, physical activity monitoring, 24 hour dietary intake assessment, and measures of self perception and health related quality of life.

Preliminary Data from the BEACHeS Study Stakeholder Focus Groups

Several of the stakeholder focus groups expressed that, while schools are an important setting for obesity prevention in the target age group, schools are becoming saturated with healthy eating and physical activity initiatives. Teachers, community representatives and some parent groups in particular felt that family and community settings are important to target for prevention interventions.

Several common themes for obesity prevention interventions came up in the focus group discussions. These include; developing parenting skills, providing activities for families, improving children’s self-esteem, providing daily physical activity in school, improving healthy food provision in school and rewarding children for healthy behaviours.

Interestingly, the focus groups also yielded much data on potential barriers to the successful implementation of obesity prevention interventions. Some of these barriers are specific to the South Asian population. For example, many children from Islamic families spend every evening at the mosque, and so do not have the opportunity for physical activities or attendance at clubs during this time.

The BEACHeS study team are currently in the process of detailed analysis of data from the stakeholder focus groups, which will be used in the development of the intervention package.

Preliminary Data from the BEACHeS Study Baseline Measures

Response Rates

Of 1090 pupils eligible in the 8 schools, 574 have undergone baseline assessments (52.7%).

Body Mass Index and Prevalence of Obesity and Overweight

BMI was used to categorise children into 4 groups; underweight, healthy weight, overweight and obese according to the age and gender specific UK National 1990 BMI percentiles reference data. The 5^th, 85^th and 95^th percentiles were used as cut-offs for the categories above.

The proportion of overweight and obese children in the study population was 22.5% (males 20.3%, females 24.9%). Table 7.5 shows the proportions of children in each weight category.

Table 7.5: Number and proportion of children in BEACHeS study population in each weight category

Number and proportion of children in BEACHeS study population in each weight category

Overweight, Obesity and Age

There was little variation in the prevalence of overweight and obesity across the age group in the BEACHeS study population, as illustrated in Figures 7.1 and 7.2. The age band studied is quite narrow (5-7 year olds), so one would not necessarily expect to see a large variation in prevalence.

Figure 7.1: Overweight and obesity prevalence in BEACHeS study population by age - Males

Overweight and obesity prevalence in BEACHeS study population by age - Males

Figure 7.2: Overweight and obesity prevalence in BEACHeS study population by age - Females

Overweight and obesity prevalence in BEACHeS study population by age - Females

Overweight, Obesity and Ethnicity

Figure 7.3: Overweight and obesity prevalence in BEACHeS study population by ethnicity - Males

Overweight and obesity prevalence in BEACHeS study population by ethnicity - Males

Figure 7.4: Overweight and obesity prevalence in BEACHeS study population by ethnicity - Females

Overweight and obesity prevalence in BEACHeS study population by ethnicity - Females

Comparison of Regional Data from National Monitoring and BEACHeS Study Data on Prevalence of Overweight and Obesity

Figure 7.5 compares the prevalence of overweight and obesity in the BEACHeS study population (5-7 year olds) with the regional prevalence in reception age children (4-5 year olds) and year 6 children (10-11 year olds). We would expect the prevalence to increase with age. The regional prevalence of combined overweight and obesity in reception age regionally and in the BEACHeS population is similar, but there is a greater prevalence of obesity in the BEACHeS population. Again, these comparisons need to be made with caution due to the data quality of the national monitoring programme and the response rates within the BEACHeS study.

Figure 7.5: Comparison of regional data and BEACHeS study data on prevalence of overweight and obesity

Comparison of regional data and BEACHeS study data on prevalence of overweight and obesity

Next Steps for the BEACHeS Study

The BEACHeS study will combine the analysis from the stakeholder focus groups with the existing evidence base on childhood obesity prevention interventions to develop an intervention package over the next 3 months. An implementation planning phase will then take place in Autumn 2007 and delivery of the intervention package will commence in 4 of the 8 participating schools in January 2008 for 6 months. During this various process measures will be undertaken. The final year of the study will comprise of follow up measures in 2008/9 in years 3 and 4 children.

7.4 Further Information

Further information about the National Child Measurement Programme is available via the Department of Health website:
www.dh.gov.uk/en/Policyandguidance/Healthandsocialcaretopics/Healthyliving/DH_073787

Further information on the BEACHeS study can be found at: www.pcpoh.bham.ac.uk/publichealth/beaches/index.htm

References:

1. University of Birmingham Department of Public Health and Epidemiology. West Midlands Key Health Data 2005/06. September 2006.
   
   
2. Department of Health. Choosing Health: Making healthier choices easier. November 2004
   
   
3. Department of Health. Measuring Childhood Obesity – Guidance for Primary Care Trusts. January 2006
   
   
4. Association of Public Health Observatories. Analysis of the National Childhood Obesity database 2005-06: A report for the Department of Health by the South East Public Health Observatory on behalf of the Association of Public Health Observatories. January 2007.
   
   
5. Teale A, Jones S, Hansford K and Harrison W. Exploring school-level obesity rates in South Staffordshire. British Journal of School Nursing July/August 2007;2(4):156-159.
   
   
6. National Institute of Health and Clinical Excellence. Obesity: the prevention, identification, assessment and management of overweight and obesity in adults and children (CG43 Obesity NICE Guideline). December 2006.
   
   
7. Wardle J, Brodersen NH, Cole TJ et al. Development of adiposity in adolescence: five year longitudinal study of an ethnically and socioeconomically diverse sample of young people in Britain. BMJ 2006;332:1130-1135.
   
   
8. Office for National Statistics http://www.statistics.gov.uk/census/GetData/datasources.asp, accessed 16th July 2007.
   
   
9. McKeigue PM, Shah B, Marmot MG. Relation of central obesity and insulin resistance with high diabetes prevalence and cardiovascular risk in South Asians. Lancet 1991;337 :382-6.
   
   
10. Campbell M, Fitzpatrick R, Haines A, Kinmonth AL, Sandercock P, Spiegelhalter D et al. Framework for design and evaluation of complex interventions to improve health. BMJ 2000;321:694.