This scoping and mapping exercise identified where there were policies in place in England to target risk factors for obesity, and whether the methods employed were appropriate according to a behavioural science perspective, and where there was scope for additional actions. A substantial amount of policy activity was identified aiming to address childhood obesity and strong coverage of policies to target many of the energy balance-related risk factors from the Foresight systems map. A total of 115 relevant policy actions were identified and over half of the potential opportunities for addressing these risk factors had appropriate actions in place. This indicates that Government has implemented many actions in England to address early years obesity.
The mapping work in the current study was able to provide specific information about whether the policy action in place to target the Foresight variables used appropriate approaches, as identified by COM-B and the BCW. There were policy actions targeting all of aspects of the model but we also identified gaps. The COM-B model identifies two types of capability- physical and psychological. Physical capability refers to physical skill, strength or stamina and is best acted upon with training or enablement, whilst psychological capability refers to knowledge and psychological skills and is best acted upon with education, training and enablement. The majority of Foresight variables identified as being amenable to change through increasing physical capability related to physical activity (e.g. transport activity) and we identified good coverage of policy actions targeting these across most of the relevant Foresight variables. The majority of the Foresight variables related to psychological capability were psychological (e.g. stress, food literacy) or dietary (e.g. portion size) and we identified numerous policy actions targeting this based on education but gaps for actions based on training and enablement.
Within the COM-B model, opportunity comprises physical and social aspects. Physical opportunity relates to time, resources, locations and cues, whilst social opportunity relates to interpersonal influences, social cues and social norms. Both can be targeted using restriction, environmental restructuring and enablement; physical opportunity can also be increased with training and social opportunity can be increased with modelling (i.e. setting a ‘good’ example). The majority of the Foresight variables across several domains (diet, physical activity, diet, economic) were related to opportunity and we identified reasonable coverage of policy actions across all of these. The Foresight variable ‘market price of food’ was a notable exception with no policy actions identified for either physical or social opportunity. The COM-B model goes onto identify reflective motivation (including self-conscious intentions and making evaluations) and automatic motivation (including emotional responses, impulses, and inhibitions). Both reflective and automatic motivation can be increased using multiple approaches from the BCW. The majority of the Foresight variables across several domains (diet, physical activity, diet, economic, physiology) were related to either reflective or automatic motivation. We identified numerous gaps in the coverage of policy actions to target these Foresight variables, with less than half having any actions. In particular, there were few policy actions relating to motivation for the following Foresight variables: stress, food advertising, self-esteem, demand for indulgence, tendency to graze, and market price of food.
Looking across the components in COM-B and the Foresight variables, the most common policy approach (as per the BCW) that we identified was education, along with a focus on guidelines targeting environmental restructuring and policies encouraging modelling opportunities (indirectly acting on the child via parent/carer behaviour change). We identified opportunities to further develop policy actions focused on enablement, persuasion, incentivisation, restriction, and coercion. For example; restriction is a possible approach for increasing physical and social opportunity, so policies based on restriction could be developed to promote physical activity (including both recreational and transport activity). Potential examples of this are restrictions on car use near schools to promote active school journeys or adding restrictions to tablets to limit their use to encourage active recreational time. Another example; since incentivisation is a possible approach for increasing reflective and automatic motivation, consideration could be given to developing policies based on incentivisation in relation to the Foresight variable portion size, an example could be incentivising purchasing of smaller packaged snack foods. Together, this highlights the focus on education and indicates that there are opportunities to build upon efforts for upstream change. In particular, there are numerous opportunities for further developing policies which act via increases psychological capability, reflective motivation and automatic motivation. Strengthening policies which increase the latter, such as people’s desires, emotions and inhibitions may be particularly powerful as they go beyond a reliance on people’s self-conscious ‘choices’.
There were a many policies addressing environmental change, with regards to both the food and physical activity environment. However, despite the UK’s good record of developing evidence-based policy guidelines, implementation of guidelines (especially public health policies) has often been poor (27). A study of implementation, using the Food Environment Policy Index (Food EPI) to map out and rate policies targeting childhood obesity in England, included a rating of implementation by experts (28). Implementation was rated highest for monitoring (of obesity, risk factors, diet), nutrient declarations on labels, access to information, availability of dietary guidelines, school food standards, and population level targets. Implementation was rated lowest for food subsidies, planning policies to encourage fruit and vegetables, and systems-based approaches. This supports our findings that upstream policies are particularly challenging to implement. A focus on strengthening existing policy recommendations to facilitate implementation, especially those targeting upstream actions, may be useful. An example of such an upstream action is the price of food which can have a huge impact on people’s purchasing decisions, with less healthy foods typically costing less and being consumed more by lower SES groups than healthier foods (29); a lack of activity here may result in other policies having limited influence.
The emphasis on education and the limited action targeting automatic motivation indicates a reliance on policies primarily focusing on individual level change. A recent study found that, for addressing obesity, governments from developed countries tended to concentrate on policy levers addressing individual-level change rather than the environment, even in countries (such as England) with a strong policy focus on childhood obesity (19). A recent systems-mapping exercise examined how local authorities in England address obesity using the ‘Action Mapping Tool’ (18). Consistent with our findings, this work found that whilst only a small proportion of the causes of obesity were coded as ‘individual lifestyle factors’, nearly 60% of the actions around obesity targeted individuals. This suggests that an individual-orientated approach is a common theme throughout both national and local obesity policy. Interventions based only on individual choice have limitations. They require families to perceive change as important and be in a position to make such changes. This is likely to be challenging for many families but particularly difficult in families from lower SES backgrounds; this may act to further widen the health inequalities apparent with obesity (30). A recent review indicated that all intervention types risk widening health inequalities but complex interventions which are targeted at multiple levels (systems, community, individual) and in multiple settings (school, health, population) appear to have less negative effects, and fiscal measures may even bring benefits (31). Successive UK governments’ policies highlight obesity as a serious problem; however there is a political tension between state and individual responsibility. Health choices are assumed to be the individual’s responsibility to control even though the behaviours leading to excess weight gain are acknowledged to be greatly influenced by the environment (32). Consistent with the findings in the current study, previous government policy documents have focused on information provision to change behaviour (33). One example comes from an analysis of the pledges within the Public Health Responsibility Deal which found that most pledges focused on providing information for consumers, rather than structural changes (e.g. reformulation) (34). Consumer views echo this, with analyses of online reactions to news stories about obesity policy finding either contradictory views around responsibility (35, 36) or that blame is attributed to the individual (37, 38). This discourse is at odds with the evidence for the important role of environmental factors in contributing to obesity (39). Of note, a greater emphasis on restricting of unhealthy food advertising was observed with the mapping of COP2 policy actions suggesting a move to more upstream action.
Strengths and limitations
This study systematically identified national policies on childhood obesity using an authoritative system analysis of risk factors for obesity (Foresight) and took a behavioural science approach to first describe risk factors and policies, and then to conduct mapping work. We believe that this is the first time that a comprehensive mapping of obesity policies has been conducted using this approach. This allowed behavioural targets and policies to be systematically described in detail, allowing the identification of gaps and opportunities for further policy development. These gaps and opportunities were specifically characterised (according to the type of intervention, the method and the target) providing explicit information to inform the strengthening of current policy and future policy development. The work has potential to be built on and could be applied at a local level and used to inform needs assessments.
Our work is subject to a number of limitations. England is a populous country with high childhood obesity levels and a history of strong public health actions on obesity, thus findings are not necessarily generalisable to other countries. We focused on national-level policy and recognise that in most countries, including England, local or community policy actions may also be in place. We were unable here to include policy activity by the 152 upper-tier local authorities in England, but the approach used in the current study could be applied in that context. This analysis recorded actions and recommendations, not how well they were being implemented, which was beyond the scope of the study. Estimating the expected impact of policy actions on behaviour and weight was also beyond the scope of this work but could be useful since the number of policies available does not necessarily correspond to their expected impact. In particular, insight into the implementation of NICE guidelines would be useful, especially for public health guidelines where there may not be the same accountability as the clinical guidelines, which are included in service commissioning processes. Tools to support implementation include the Food EPI, an established method using expert consensus to provide policy ratings and identifies gaps and policy priorities (28) and surveillance plans (19). The Foresight systems map was used to identify risk factors for childhood obesity as it is a comprehensive review of the evidence; however it was developed in 2007 so may not capture recent research. In addition, this is a fast moving field, the Childhood Obesity Plan (Chap. 3) was published as part of the Prevention Green Paper after the mapping work was completed and therefore not included (40). There is also considerable interest in the role of the pre-conception period for later obesity risk (41); however, the life-course stage for this work was restricted to pregnancy and early life to ensure that the study was feasible. The policy scoping was done via online searches, it is possible that there are additional policy actions not identified with this approach.