Summary and recommendations for practice

• There are complex evidence needs for preventing childhood obesity.
  
• The International Obesity Taskforce has developed an evidence framework for obesity prevention and it poses five sets of questions that need different types of evidence to answer them.
  
• The evidence on the burden of obesity is sufficient to warrant action and the evidence on the determinants of obesity is also informative on what to do.
  
• The priority target groups [who] are mainly children and adolescents and high-risk adults, and schools are the favoured setting [where] although multiple settings are preferable. The strategies [how] also need to be multi-pronged with communications, programs and policies being the main approaches.
  
• The evidence on effective interventions is quite limited although it is growing rapidly, especially for whole-of-community intervention approaches.
  
• A priority setting process needs to use all available evidence to define a portfolio of promising interventions.
  
• There are still very large evidence gaps from too few “solution-oriented” studies.

Introduction

In an effort to clarify the role of evidence in obesity prevention, the International Obesity Taskforce (IOTF) published a framework¹ that identified the key questions to be answered, the types of evidence needed and outputs produced, and the role of contextual factors (Figure 6.1). In the process of building this framework, a number of general concepts and specific issues emerged about evidence as it applies to obesity prevention.

Figure 6.1 International obesity taskforce obesity prevention evidence framework.

Definitions and hierarchies of evidence

Evidence, in its widest sense, is information that can provide a level of certainty about the truth of a proposition.² This is a very broad definition, more along the lines of the legal, rather than the medical, concept of evidence and implies that this breadth of information is important and valid for decision making. For the purposes of addressing the questions on obesity prevention, the IOTF framework grouped evidence into observational, experimental, extrapolated and experience-based sources of evidence and information.¹ Examples of these are outlined in Table 6.1.³

Table 6.1 Types of evidence and information relevant to obesity prevention.³

Each type of evidence has its own strengths and weaknesses. Each can be judged on its ability to contribute to answering the question at hand. In practice, there is wide variation in the quantity and quality of information available in respect of different settings, approaches and target groups for interventions to prevent obesity. There is virtually no evidence concerning the potential effects on obesity of altering social and economic policies, such as agricultural production policies or food pricing policies, while much more evidence is available on localized attempts to influence the consumer through educational and program-based approaches.

Traditional hierarchies of evidence are based on rankings of internal validity (certainty of study conclusions). These tended to be less valuable in the IOTF framework because of the tension between internal validity and the need for external validity (applicability of study findings). The importance of context on evidence and the need for external validity is greater in some areas than others (left-hand bar in Figure 6.1). It is especially important at the priority setting stage (issue 5), and this is where the informed opinions of stakeholders is paramount.

Modeled estimates of effectiveness and informed stakeholder opinion also become important sources of information where the empirical evidence is complex, patchy and needs to be applied to different contexts. This means that assumptions and decisions must be made explicit and transparent. The acceptance of modeled estimates of effectiveness and informed opinion in the absence of empirical evidence means an acceptance of the best evidence available not just the best evidence possible (as occurred in systematic reviews with strict inclusion criteria).

Evidence on the burden and determinants of obesity

These are the first two issues in the IOTF framework (Figure 6.1). In general, the size and nature of the obesity epidemic has been well enough characterized to have created the case for action. Of course, many gaps and debates still remain, such as the prevalence and trends in poorer countries, the psycho-social impacts of obesity in children, and the effect of the epidemic on life expectancy.

Evidence on the determinants of obesity is very strong in most areas, although to date, most of it is focused on the more proximal biological and behavioral determinants rather than the more distal, but very important, social and environmental determinants. One poorly researched but very obvious set of determinants are the socio-cultural attitudes, beliefs, values and perceptions that may explain the very large differences in obesity prevalence rates seen across different cultures. These socio-cultural determinants may help to explain why both rich and poor countries have obesity prevalence rates in women of less than 5% (e.g., India, China, Yemen, Ethiopia, Japan and Korea) and greater than 40% (e.g., Samoa, Tonga, Qatar and Saudi Arabia).⁴

Knowing a lot about the determinants of obesity should help to guide interventions but, as Robinson and Sirard point out,⁵ “problem-oriented” evidence (what is to blame?) is often quite different to “solution-oriented” evidence (what to do?) It is the solution evidence that decision makers are urgently seeking and it is this evidence that is currently lacking.

Opportunities for action—who, where, how?

This is the third issue on the IOTF framework (Figure 6.1). Many countries have now created strategic plans for action on obesity either as an issue by itself or as part of promoting physical activity or health eating or reducing chronic diseases. Classic frameworks for health promotion specify “who” in terms of target groups (e.g., children, adolescents, pregnant women, minority ethnic groups, those on low incomes), “where” in terms of settings or sectors (e.g., schools, the commercial sector, the health sector), and “how” approaches or strategies (e.g., school education, community development, the use of mass media, environmental change, policy and infrastructure change).³ These issues are addressed in other chapters in this book but a few brief comments about target groups are warranted.

A potential limitation of identifying target groups is that they become too much the focus of the action (e.g., by encouraging them to make the healthy choices) rather than the players that influence the environments that determine those behaviors (those who can make the healthy choices easier for the target group). In this respect, the definition of target groups may need to be widened to include the providers of the determinants of health, such as the providers of health information—the health services, schools, the media, commercial producers—and widened still further to include those that set the policies which shape access to healthy lifestyles through, for example, pricing, distribution and marketing. In this sense, target groups may include shareholders in companies, professional groups, policy-makers and public opinion leaders, including politicians.

Prevention strategies targeting adults make economic sense (if they work) because the consequences of obesity occurring in middle aged and older adults generate economic costs—especially through Type 2 diabetes and cardiovascular diseases.⁶ Adults, especially those with other existing risk factors, are at high absolute risk of these diseases; therefore, they have the potential for high absolute gains. In addition, there is now very strong efficacy evidence that individual lifestyle interventions in high-risk adults prevent diabetes and heart disease.^7,8 However, despite this evidence and logic, children have risen as the priority target group for most action on obesity and this has occurred for a number of reasons—some based on evidence, some based on societal principles, and some based on practicalities.

Excess weight, once gained, is hard to lose but children who are overweight do have a chance to “grow into” their weight. Children’s behaviors are also much more environmentally dependent than adults’ behaviors and most of the evidence on obesity prevention has been in children. However, far more powerful than the sum of the evidence are two other drivers of children as the priority target group: societal protection of children and access for interventions—especially through schools. Policy-makers have been especially sensitive to children because society has an obligation to protect them from ill-health. For adults, the societal obligation shifts towards protecting personal choice—even if that choice is for unhealthy foods and physical inactivity.

Effectiveness of potential interventions

This is the fourth issue in the IOTF framework (Figure 6.1) and asks: “What are the potential, specific interventions and what is the evidence for their effectiveness?” This question is covered in many other chapters in this book and only some overview comments are be made here.

In general, it is possible to generate quite long lists of potential interventions to help prevent childhood obesity, although some of the intervention areas, such as improved parenting skills, are more readily identified in generalities than in specific interventions. However, adding effectiveness evidence to any more than a few of them is very difficult because of the absence of intervention studie. For the interventions that have been studied, most concern primary school children, most are in school settings, most are short term, most are not sustainable, and most have shown relatively modest effects on anthropometry (although they were often able to show improvements in eating and/or physical activity). The number of reviews in the area is starting to outnumber the number of studies.^9–29

As interventions move towards more whole-of-community interventions, there is an increased complexity in the study design, the interventions and the evaluations, although some positive results are starting to emerge from these studies, which is very encouraging.^30–32 More sophisticated multi-level modeling will be needed to tease out intervention effects in these more complex interventions.

For policy-makers considering strategy options, the distinction between effectiveness and cost–effectiveness is critical. If a policy objective is to be pursued with no limitation on spending, then effectiveness (the beneficial effect of a strategy in practice) is the primary consideration. But when cost limitations apply (as they inevitably do), an evaluation of cost–effectiveness is essential if rational decisions are to be made.³³

A remarkable feature of the evaluations and systematic reviews of interventions described above is that they rarely mention the costs of the various programs they examine, and make no estimates of cost–effectiveness. For child obesity prevention, only one study has explicitly examined the costs of an intervention program, the US Planet Health Program.³⁴ Planet Health’s estimated cost–effectiveness ratio gives a value of $4305 per quality-adjusted life year gained, which compares favorably with interventions such as the treatment of hypertension, low-cholesterol-diet therapies, some diabetes screening programs and treatments, and adult exercise programs.³⁵

Creating a portfolio of interventions

Evidence needs

Stay updated, free articles. Join our Telegram channel

Join