Answers are grounded in this paper and cite their sources.
Reviews surveillance frameworks and population trends for obesity, highlighting the necessity of standardized routine measurement to guide public health interventions.
The UK Government has identified obesity as a priority area, as it is thought to present a serious public health problem. Although it is a highly complex multifactorial disease, it has been hypothesized that obesity, in particular morbid obesity, may decrease life by up to 9 years and significantly increase the risk of many other important diseases, e.g. diabetes, heart disease and some cancers. Furthermore, it has been estimated that if obesity continues to rise at the current rate, children could die earlier than their parents (1). A joint report from the Audit Commission, the Healthcare Commission and the National Audit Office estimated that the current cost of obesity to the National Health Service (NHS) in England stands at £1 billion, and that the cost to the UK economy is £2.3–2.6 billion in indirect costs. If this present trend continues, by 2010 the annual cost to the economy could be £3.6 billion a year (2). Obesity cannot be considered in isolation from other lifestyle factors, in particular, physical exercise. Strategies for reducing obesity must therefore consider both sides of the energy equation. The clinical definition of obesity uses the body mass index (BMI, body weight/height2). For adults, ideal weight, overweight, obesity and morbid obesity are classified as a BMI of 20–24.9, 25–29.9, 30–39.9 and >40 respectively. In children, the definition is more complex. Currently, two methodologies are utilized: (i) UK BMI percentiles (3) (which define overweight and obesity cut-offs as the 85th and 95th percentiles of UK reference curves) and (ii) international classification (4) (whereby adult BMI cuts-offs for overweight and obesity are extrapolated to formulate age-appropriate equivalents). Currently, there is no consensus as to suitable cut-offs for other methods of measuring body fatness. The following illustrations are provided to demonstrate the importance of good routine data to understand the problem. Although data can be extrapolated from research projects, it is routine surveillance that will provide the most robust information. Figure 1 demonstrates that obesity is not a problem confined to the UK, some of the biggest increases are being seen in developing countries, which is assumed to be the result of increased adoption of a Western diet. However, the graph clearly shows the predominance of European countries and the position of the UK. Overall trends in obesity in the world; annual increase in prevalence, 1980–1998. Source: extracted from a draft WHO report, Swinburn B et al. Figure 2 shows the year-on-year rise in England taken from the Health Survey for England. The graph also shows that the prevalence of obesity in men is now similar to women, a new situation compared with 20 years ago. Overall trends in obesity in England (19). Source: Health Survey for England, 1993–2004. Figure 3 shows the increasing prevalence of childhood obesity, which appears to have accelerated since 1998. Figure 4 shows the regional diversity in obesity. In England, figures for obesity are higher for women than men. This is the same in the regions of England, with the exception of the north-east, south-east and south-west. Childhood obesity (19). Source: Health Survey for England, 1995–2003. Regional differences in obesity in England (19). Source: Health Survey for England, 1995–2003. There are also important socioeconomic and ethnic differences (particularly among women) in the prevalence of obesity, with more obese individuals arising from the most disadvantaged groups (Fig. 5) and the black ethnic group (Fig. 6) (5), There is also evidence to suggest that individuals with disabilities are at higher risk of suffering from obesity (Fig. 7) (6, 7). Socioeconomic differences in obesity prevalence for the South-East region (19). Source: Health and Safety Executive 2000–2002. Ethnic differences in obesity prevalence (19). Source: Health and Safety Executive 1999. BMI, body mass index. Prevalence of disabling conditions by BMI (19). Source: Health and Safety Executive 2001. BMI, body mass index. Obesity targets were first set by Government in 1992, the targets set then were to reduce obesity in men by 7% from 1986/1987 figures and in women by 6% (8). Policy statements on obesity assume the availability of good data. Sir Derek Wanless, in his report in 2004 (9), not only underlined the need for a government objective to ‘halt the rise in obesity now’, but also pointed out the dearth of locally available information on a range of public health issues such as childhood obesity. In his view, obesity could be equal in importance to smoking as a determinant of future health. . . . to halt the year-on-year rise in obesity among children aged under 11 by 2010 in the context of a broader strategy to tackle obesity in the population as a whole (PSA Obesity Target 2004). In 1998, the Government issued Public Sector Agreement (PSA) targets. PSAs aim to increase effective performance management of centrally determined targets and are delivered locally. In 2004, the reduction of childhood obesity became a PSA target. Three government departments are jointly responsible for this target: the Department of Health, the Department of Culture, Media and Sport, and the Department for Education and Skills. Obesity is identified as one of six key priorities in the White Paper Choosing Health: Making Healthy Choices Easier (10). This draws together a series of other initiatives such as the Coronary Heart Disease (11), Diabetes (12), National Service Frameworks and the National Cancer Plan (13) – all of which set out plans for action (including prevention, identification, management and treatment of obesity). Further information was subsequently provided in the White Paper Delivery Plan (14) and Department of Health document, Choosing Health? Choosing a Better Diet (15), which included promised action on: advertising and promotion of foods to children; simplified food labelling; obesity education and prevention; nutritional standards in schools, hospitals and the workplace. Surveillance has been described as a continuous and systematic process of collection, analysis, interpretation and dissemination of descriptive information for monitoring health problems (16). Surveillance is not screening – this is important when introducing a large surveillance programme. Screening is defined as the examination of asymptomatic people in order to classify them as likely or unlikely to have the disease that is the object of screening – the people who are found are then treated. One of the important issues about obesity is to understand how the problem can be monitored in the future. Surveillance is an essential tool given the epidemic proportions of obesity and has an important role in identifying groups at risk and the efficacy of interventions. However, it is critical to be clear about the meaning of surveillance and, in particular, to be clear about the differences between surveillance and screening. There is no suggestion that a screening programme should be instigated as part of the programme for the management of obesity in the population. Succinctly, the reasons for this are that the problem does not fulfil the usual criteria for the creation of a screening programme described by Wilson and Junger (17): Is it an important public health problem? Is there a detectable preclinical phase? Are there simple and acceptable diagnostic tests? Is the treatment simple and effective? Does treatment prevent onset of complications? The prevalence of obesity is most commonly assessed through the calculation of BMI. However, for some populations such as some ethnic groups, the elderly, individuals with certain disabilities or extreme athletes, simple BMI classifications can be inaccurate. In such populations, additional measures of body fatness, which may include a waist circumference or waist : hip ratio and/or skin-fold thickness measurements, may also be considered. While these anthropometric measurements will track the prevalence of obesity within a population, it may also be desirable to evaluate potential predictors, associated comorbidities and aetiological factors if the data are to identify groups ‘at risk’ and inform the development of appropriate interventions. Table 1 proposes a minimal core dataset, broken down into three levels. It is, however, important to balance the potential benefits of the data to be collected with the economic implications and participant burden that will ensue as levels of data collection increase. It is also essential to assess the quality of data collected, as self-reported data must be interpreted with appropriate caveats. In January 2006, the Department of Health issued guidance to Primary Care Trusts (PCTs) on how to measure childhood obesity, requiring initiation of systems to measure the height and weight of children aged 4–11 years. PCTs are required to measure: all primary school children in the reception year (aged 4–5 years); all primary school children in year 6 (aged 10–11 years). Further guidance from the Department of Health on the system for collation, interpretation and dissemination of the information was issued in April 2006 (18). In December 2006 the Department of Health published initial findings from the National Childhood Obesity Database (19). The report states that the first year of data collection has resulted in a number of practical difficulties, such as the transmission of data from PCTs to the national database, but also, more importantly, some suggestion that overweight children may have opted out, thus giving rise to a possibility that the prevalence of obesity has been underestimated. Nationally, 12.3% of girls and 13.4% of boys were found to be overweight in year R (reception year) and 9.2% of boys and 10.7% of girls in the same year group were obese. In the older age groups (year 6), 13.8% of boys and girls were found to be overweight and 15.4% of girls and 18.9% of boys were found to be obese. The report sets out a number of recommendations for improving the data collection system in the future. Thus, in the context of obesity, the objective is surveillance. Applied to school-age children, this means that measurements will be taken, but results will not be routinely given to parents. In theory, there is no need to measure every child in a school or class. But as selecting pupils randomly for measurement in a school is unlikely to be workable, it has been decided to measure all children. The guidance for PCTs states that information on height and weight will be given to parents on request. Health survey data is one of the very important sources of data on obesity at the present time. The most important source in England is from the Health Survey for England (5, 19). This is carried out annually in a sample of adults and children in the country. Heights and weights are measured and recorded to a detailed protocol. However, the sample size is usually such that the data are only available at the level of government office regions and the (old) strategic health authorities. Data from the Health Survey for England are not available at the level of PCTs or local authorities. Any surveillance of height and weight needs to be measured rather than self-reported. In order to surmount this difficulty, synthetic estimates can be calculated that provide an estimate of the level of obesity at a local level by associating the socioeconomic and demographic features of the population with other populations (usually larger) where the prevalence of obesity is known. The disadvantage with this technique is that it does not allow ongoing monitoring of the trend in that community and is only as good as the parameters used in the methodology. Body mass index and waist : hip ratios have also been collected in the National Diet and Nutrition Survey. This survey has been carried out on two occasions: 1986/1987 and 2000/2001. Included in the survey were 1800 measurements from a sample throughout the UK, but unfortunately this sample size is insufficient for much other than a national picture and a corroboration of the Health Survey for England (20). Recent work carried out by the North East Public Health Observatory and the University of Teesside suggests that maternal obesity is also a growing problem in the number of obese women presenting at antenatal clinics in the past 15 years. Not all maternity units uniformly measure height and weight at booking and record it in electronic format, but this is crucial in order to be able to monitor the problem in the future. Similarly, we have published data from the Health Survey for England that suggest there is a particular problem with obesity in individuals with a disability (21). There needs to be a clear method for collecting collating data on childhood obesity from PCTs, with a much more standardized approach. We need an expansion of the Health Survey for England to permit more local analysis of the data. The new contract in general practice provides new opportunities for data collection from primary care. Of course, some general practitioners (GPs) have collected data for many years, but this has not been particularly systematic. There is currently pressure for the new Quality Outcomes Framework in general practice to include targets for obesity. If this is the case, there will be financial incentives for GPs to monitor height and weight in the future. This will potentially improve the collection of data on obesity from general practice and, as intervention will also be rewarded, this should have an important role in reducing the prevalence of obesity. The NHS local delivery plan 2005–2008 (22) sets out a requirement for future data collection in primary care for monitoring of the obesity status in the 15–75-year age groups, but not many practices can achieve this at the current time. Improving surveillance data is a long-term goal. Many of the surveillance systems described above will take a number of years to show meaningful results, but if effectively implemented will improve our understanding of obesity in the longer term. The development of a single care record from cradle to grave could provide the longitudinal data on obesity and its relationship to health. Issues of disclosure and confidentiality could be a barrier to this, but if techniques such as ‘pseudonymization’ (which allows, in effect, personal data to be handled in a secure manner, preserving the confidentiality of the individual) are put in place, this difficulty should be overcome. There are techniques from the clothing industry to assess body size and shape using scanning devices (similar to those at airports) which could be applied in the future to produce population data about the trends in body size. However, at the present time, the cost of these techniques for monitoring obesity is prohibitive. Obesity is undoubtedly a serious problem in this country today, which, if not tackled, will lead to serious future consequences, costly in terms of both the health of individuals and the UK economy and future prosperity. Despite a clear recognition of the problem, our ability to monitor progress is currently limited. There is no local collection of data on obesity and although it can be argued that this is not necessary everywhere, the measurement of height and weight is such a simple procedure that their routine collection should be given serious consideration, given that such data will allow obesity to be monitored at a local level. Action outside the NHS is important, and the development of PSA targets applicable to government departments is no exception. However, targets need to be set that can be monitored effectively and which do not lead to inappropriate levels (and frequency) of monitoring encouraged by perverse incentives. Local communities need better data at a local level and this should be – and is beginning to be – addressed by the collection of data on childhood obesity. No conflict of interest was declared.
No takes yet. Share an insight, caveat, or question.
Wilkinson et al. (2007) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: