In Wales, where 62% of the population is either overweight or obese, improving food environments and enabling people to live healthier lives, is a critical social challenge. Data on the prevalence of obesity and its drivers is key to the design, trialling and scaling of effective solutions. In an ideal world, an entire data ecosystem - containing accurate, timely and granular data on the prevalence of obesity, people’s diets and the food environment - would support policy makers in their efforts. But what happens when much of that data isn’t collected?

We’ve isolated the missing publicly available data or ‘data gaps’ in this ecosystem which, if filled, could support Welsh Government, Nesta and other stakeholders to reduce the prevalence of obesity in Wales. Through a combination of desk research and conversations with experts, we identified 27 gaps in areas ranging from the available data on the nutritional content of food and drink to data on exposure to food advertising and promotions. Each gap was assigned a priority level to convey the importance of closing the gap given current policy priorities, the effort required to close the gap and potential impact of providing this data.

Seven high-priority data gaps were identified, reflecting their magnitude, impact and urgency. Strikingly, four of these seven data gaps hinder efforts to address childhood obesity. For several of these gaps, comparable data is already available for England and Scotland.

• Childhood obesity: Regular and systematic monitoring of body weight in children from: (i) early years to (ii) primary school age and (iii) adolescence is severely lacking, along with limited data on their (iv) daily diets and energy intake. Filling these gaps, in a sensitive and cautious manner, would allow more timely identification of at-risk children and the design of more effective interventions to reduce prevalence of childhood and long-term childhood-onset obesity.
• Adult diets: Limited data on what we eat represented a significant data gap as did the lack of data on the impact of food consumed outside the home on diets and obesity. Improving data on adult diets would allow more accurate targeting of harmful foods for reformulation, and measures such as restrictions on price and placement promotions; which show potential to be impactful.
• Local affordability: While the Healthy Weight Healthy Wales strategy highlights the importance of improving diets, a lack of localised data prevents tracking the affordability of healthier food for different groups across Wales. The recent rise in food and energy costs has only increased the need for more interventions around affordability. Data on the affordability of healthier food could support the implementation of subsidies, healthy food voucher schemes, or promoting available options that are cheaper and healthier.
• Publicly procured food: The Healthy Weight Healthy Wales strategy calls on Welsh food companies to reduce the energy content of the food and drink they produce. Foods served in locations where the government has greatest control over the food on offer (e.g,. schools, hospitals and other public sector workplaces), would be a sensible starting point. However, we found limited data on publicly procured foods, which would be of great value to target and track such efforts.

The good news is that it is relatively straightforward to begin addressing many of these data gaps. Suggestions for how some of these gaps could be filled include: funding a boosted sample for the National Diet and Nutrition Survey to address critical data gaps on diets in both adults and children, and cautiously extending the Child Measurement Programme to include older primary school children and adolescents.

More challenging will be the creation of new infrastructures and networks for gathering data on the likes of online exposure to food advertising for example. Nevertheless, our findings provide a starting point for strategic and coordinated action to enhance the data available to support obesity reduction in Wales, and Nesta would be keen to explore how we could work with other organisations to start doing just this.

Nearly 1.6 million adults in Wales live with obesity or are overweight, and more children are obese by the time they start primary school in Wales than in any other region of the UK. This project aims to show the data gaps which, if filled, could improve our understanding of this challenge, and in turn support, Nesta and other stakeholders to design, test and scale solutions that decrease the prevalence of obesity in Wales.

Achieving both the Welsh Government's goals of preventing and reducing obesity — and Nesta’s own goal that the UK will have halved the prevalence of obesity by 2030 — will require all organisations who share this goal to have access to the information they need to:

• Understand the drivers of obesity — for example, how do different factors, such as the cost of food, distance to supermarkets and availability of healthier foods, impact access to healthier foods in urban and rural areas?
• Effectively design and target interventions to those most at risk — for example, which foods should be prioritised for reformulation to improve the diets of young people with excess weight?
• Evaluate what works and for whom — for example, if implemented in Wales, what impact would mandatory calorie labelling have on consumers’ food choices and obesity prevalence?

While one-off studies, where information is not collected on an ongoing basis or shared with others, may be enough to achieve ‘one-off policy wins’, we need many such wins over a number of years to reduce obesity in Wales. For this reason, our vision is a whole data ecosystem that provides regular, accurate and sufficiently granular data on obesity and its drivers, which in turn will allow us and others to design, test and scale better solutions.

When the COVID pandemic hit the UK in 2020, we saw a similar data ecosystem rapidly emerge. It provided accurate and open information on infections, hospitalisations, deaths and vaccinations. This information was collected by a variety of organisations and was quickly made available to all those that needed it to make important decisions.

We need a similar ecosystem, ideally created with a similar level of urgency, for tackling obesity in Wales. While we must continue to make progress with the data we do have available, we should also invest in building a better ecosystem. A first step to creating this ideal data ecosystem is understanding where there are currently gaps in access to data.

In this context, a ‘data gap’ is a lack of data that reduces the ability of Nesta, the Welsh Government and other stakeholders to decrease the prevalence of obesity. Data gaps can occur because no data is collected or existing data does not meet the needs of users. Examples include data that is not sufficiently granular (perhaps there are no regional or demographic breakdowns), is prohibitively expensive, very delayed, or of poor quality.

While ‘data’ is often thought of as numbers, it also includes textual information (such as nutritional labels) and images (such as advertising billboards). All these forms of data can inform our understanding and increase the effectiveness, of our efforts to reduce obesity.

In the course of this project, conducted between January and May 2023, we developed an approach to identify data gaps, which we will continue to refine. A more in-depth description of the methodology is provided in our blog titled 'Mind the gap'.

Nesta’s healthy life mission has five focus areas:

1. improving evidence on diets
2. reducing less-healthy food and drink promotion
3. reducing the energy content of food and drink
4. improving access to healthier food and drink
5. understanding the attitudes of decision makers.

In each area, the mission team has set goals. We used these to build a vision of the ‘ideal data ecosystem’ which would allow the team to perfectly track and achieve these goals. On ‘reducing energy content of food and drink’, for example, the ideal data ecosystem might contain a database that captures the nutritional composition of foods over time. While other factors do contribute to obesity — such as genetics or activity levels — these were not considered in this analysis as they are outside Nesta's areas of focus.

We then spent considerable time identifying the aspects of these ecosystems that already exist. This involved extensive desk research and speaking with an array of experts. In areas where some data was found to exist, we assessed its potential to meet our requirements. In many areas, however, we found that virtually no data was available and the gaps were sizeable.

Having identified the gaps, we gave each gap a priority rating of low, medium or high, based on the following criteria:

• Time sensitivity— is the gap related to a policy currently being considered, in which case new data could lead to immediate impact?
• Potential impact — what is the potential impact on obesity that may result from this data becoming accessible?
• Effort required — how large is the gap between the data that is available and what is preferable?

Gaps that have been prioritised as high are those that Nesta would be particularly keen to explore how, in partnership with others, we might begin to address.

Data gaps are subjective, as data that is adequate to meet one organisation’s needs may be inadequate for another’s. Gaps may also be filled (rending information out of date). Finally, given the large number of potential data providers, it is easy to miss a data source and incorrectly assume there is a gap.

If you have any corrections to our map, or would like to discuss how we might be able to work together in the future to improve and enrich data in Wales - please contact Jonathan Bone at: [email protected].

Data on the prevalence of obesity in Wales and its evolution over time, is crucial to ensuring that appropriate funding and resources are directed towards tackling this social challenge. This data can also help to target interventions towards specific geographic areas and groups. Finally, this data is critical for monitoring the progress and effectiveness of interventions, and thereby helping policymakers to understand what works best and for whom.

The most significant challenge in regards to measuring obesity is the potential for bias if we either rely on self-reported data (where respondents tend to underestimate their weight), or if we rely on data from GP visits (which represent a non-random sample of the population). While the Healthy Weight, Healthy Wales 2022 - 2024 delivery plan makes a commitment to ‘improve systematic collection and access to primary care data regarding height and weight’(p.22), no detailed plans on how this will be achieved have been made public.

Overcoming the biases in weight data would require objectively measuring weight for a representative sample; potentially as part of the National survey for Wales (this is already the case for England and Scotland’s national health surveys). However, this process can be time-consuming and costly. Finally, any measurement programme must be carried out sensitively and cautiously, given the potential for perpetuating weight stigma and associated harms.

Other data gaps that have been identified could also potentially be addressed by drawing on existing infrastructure; for example, by extending the Child Measurement Programme (which recurrently only collects BMI data for reception-age children), to include older primary school and secondary school children.

“There is still a lot that we don't know. We know an awful lot about obesity and being overweight and the causes of it, but we don't really have a very clear understanding, in Wales at least, of the point at which children start to become overweight.”

Dr Frank Atherton, Chief Medical Officer, Welsh Government

What we eat and drink is the most important predictor of obesity, and health more broadly. Accurate information on food and drink consumption is critical to identifying the categories of food, and more specifically the ingredients, that are contributing most to daily calorie intakes in Wales. This would enable us to target interventions much more effectively.

However, accurately tracking dietary data presents several challenges including a tendency for individuals to underreport their consumption in dietary surveys. For this reason, we explore the availability of purchase data, which has been used as a proxy for consumption. However, purchase data is typically only available through commercial providers where costs can be prohibitive. Moreover, purchase data has its own limitations such as not knowing how much of the food purchased is actually consumed or by who in a household. Since consumption data is already collected — albeit with relatively small samples of Welsh participants — boosting sample sizes of these surveys, particularly the National Diet and Nutrition Survey, could deliver significant value.

Evidence suggests that consumers who buy more of their food and drink on promotion, purchase more HFSS (high in fat, salt or sugar) foods and are more likely to experience obesity. There are a number of avenues through which improved data on food and drink promotion could accelerate Welsh policy interventions that are aimed at curbing the promotion of HFSS foods. For example, data on Welsh residents' exposure to HFSS products via television and online spaces, could put pressure on the UK Parliament to push through proposed advertising restrictions that are currently postponed (and at risk of being scrapped). Similarly, data on advertising in public spaces (of HFSS products) could help build a case for, and support the effective implementation of, restrictions (such as in train stations or near schools) — actions that the Welsh Government has previously committed to considering. Finally, data on Welsh people’s exposure to price and placement promotions for HFSS foods would support the implementation of restrictions and bans; these measures were a focus of the recent Healthy Food Environment consultation.

The largest hurdle to improving our understanding of promotions and advertising is the large array of mediums, which range from billboards to web pages, and from sponsorships to loyalty apps. A second challenge is the personalisation of online platforms — two visitors to the same website will not necessarily see the same adverts, which means that some groups may be more heavily exposed to HFSS advertising than others. A third hurdle is defining what constitutes HFSS advertising, particularly when a brand sells many products, only some of which may be classed as HFSS. A final challenge is the difficulty of capturing in-store promotions (such as end-of-aisle discounts), which may change frequently and may be unique to each store. The fact that some of this data is available to purchase suggests collection is feasible; finding ways to open this data up for public use could add significant value.

The importance of ‘Welsh businesses increasing the scale of reformulation and developing healthier alternatives’ is highlighted in the Government’s Healthy Weight, Healthy Wales strategy (p. 11). In addition, several measures that could indirectly incentivise reformulation, such as bans on price and placement promotions of less-healthy foods, are currently under consultation by the Welsh Government. The Healthy Weight Healthy Wales strategy even states that ‘Welsh taxation powers on discretionary foods high in fat, salt and sugar to be considered if the pace of reformulation does not meet expectations’ (p. 11). However, the successful evaluation of these measures all rests on having access to data on food reformulation which, at present, does not exist.

A major challenge contributing to the lack of data on the composition of food is that while producers of pre-packaged foods are typically required to list the ingredients and provide nutritional information on their labels, they are not required to register this information with any centralised database; doing so could add huge value by allowing the tracking of reformulations. Moreover, it has been suggested that when policymakers rely on such information to be provided voluntarily by food manufacturers, this can create a transactional relationship, where policymakers can feel obliged to return the favour, which may lead to less impactful food policies.

“In determining the impact of reformulation on population dietary behaviors, policy makers often depend on data provided by these manufacturers. Where such data are “gifted” to regulators, there may be an implicit expectation of reciprocity that adversely influences nutrition policies.”

Compbell et. al. (2021)

A major barrier to improving access to healthier food and drink in Wales is not knowing which factors are having the largest impact on Welsh diets. For this reason, we consider just three basic aspects of access — physical access, online access and affordability. A future study might take a much wider view on access, and consider factors such as the time available for preparing and eating healthier meals, as well as the range stocked in local food and drink outlets.

In addition to the lack of data on access at a nationwide level, there is even sparser data on how access varies across Wales. Regional variation in factors such as the cost of food or the availability of public transport could well create very different pictures of access across the country, particularly given that a high proportion of the population live in rural areas.

A specific area of opportunity to improve access to children’s diets in Wales is the rollout of free school meals to all primary school children. Better data would help us make the most of this.

“There are no public domain data sources of online groceries provision at a neighbourhood level in GB. Whilst retailers make claims about coverage in press releases or annual reports, they publish no definitive list of areas covered by their service, with consumers required to input their postcode on retailers’ websites to check delivery coverage.”

Newing et. al. (2020)

Capturing the attitudes of Welsh decision makers, and tracking the policy and measures implemented by the Government, are critical to mapping progress towards tackling obesity. It is also vital to track the attitudes of those groups that influence the Government, including the food and drinks industry, the media and the public. The media, in particular, plays a key role in shaping the narrative around obesity and measuring the media discourse allows us to detect progress.

A number of challenges prevent the collection of data on the attitudes of these three groups. There is a lack of free programmatic access to most datasets of historical newspapers and other media. While the views of industry and the public might be captured via the likes of industry press releases and social media sites, it is very difficult to obtain unbiased views from these data sources. An alternative approach is to use surveys but these can be prohibitively expensive, particularly if they are carried out regularly, which is necessary to detect changes over time. However, there are ways datasets could be created by extracting information from existing sources that could provide insights on the attitudes of various groups.