Nexus Connect: Obesity in Preventative Health

The panel was led by Nathan Berry, Head of Collaboration and Health Tech Lead at Nexus.

Together, the panel examined what preventative health really means, the barriers standing in its way, the role of data and AI, and the misconceptions that continue to hold progress back.

Here’s a look back at the conversation.

Preventive health is about more than stopping illness before it starts. Rooted in the NHS’s ambition to reduce disease and improve the health of the population, it encompasses the broader social determinants of health, including education, housing, environment and access to employment, that shape people’s wellbeing long before they ever see a clinician.

The panel outlined three levels of prevention:

• Primary prevention – stopping conditions from developing in the first place
• Secondary prevention – preventing the progression of existing conditions
• Tertiary prevention -managing the comorbidities of conditions like obesity and supporting people across the full spectrum of their health journey

Crucially, the panel emphasised that preventative health must be understood as a whole-system responsibility. Historically, it has been framed as an individual’s duty, but as the discussion made clear, the world is built in ways that make unhealthy choices easier.

Policy is beginning to respond. The UK is leading the way on several fronts, including bans on unhealthy food advertising, calorie labelling on menus, and restrictions on promotional placement, but policy alone will not be enough.

Central to this challenge is the persistent oversimplification of obesity itself. The Foresight project identified over 100 factors contributing to obesity, spanning science, psychology, culture, society, and environment. This complexity is frequently ignored in public discourse, where obesity is still too often reduced to a matter of willpower – a framing that is not only inaccurate but actively harmful. The stigma attached to living with obesity, and to seeking treatment, has real consequences for people’s willingness to engage with support.

With 3.8 million people expected to turn to anti-obesity treatment, the scale of need is clear. Obesity can be the result of metabolic and genetic factors, not simply lifestyle choices, yet those who proactively engage with treatment, including medicated approaches, can still face the accusation that they are “taking the easy way out.” This stigma bleeds into how people using any method of weight management are perceived and treated, and it remains one of the most significant barriers to progress.

There is a real opportunity for technology here. Platforms already exist to support behaviour change on a one-to-one level, but can we use technology to scale these interventions? And if so, can we overlay individual profiling to inform and personalise those interventions at a population level? These were questions the panel returned to throughout the evening.

If technology is part of the solution, data is its foundation. The panel explored both the enormous potential of data-led approaches and the very real risks of getting them wrong.

The ESRC Strategic Network for Obesity, has helped develop an obesity systems map drawing on national surveys conducted annually. The question now is what additional data points could be generated, and from where. Sources such as gym card swipes, supermarket purchase data, and step counts all offer potential insight, but the challenge lies in making this data accessible, representative, and ethically sound.

An example of data-led intervention is the work the University of Leeds did with supermarkets to redesign checkout layouts, resulting in two million fewer unhealthy products being sold per year by October 2022. This demonstrates that food environments can be analysed, understood, and changed, and that the right data, applied well, can drive meaningful population-level outcomes without needing to survey millions of individuals.

However, a significant concern raised was digital invisibility. Those without a digital footprint are systematically underrepresented in data-led approaches, and any system that fails to account for this risks widening health inequalities rather than addressing them. Ensuring that technology is accessible and that data usage is secure and transparent must be prioritised and we need to be designing a toolbox of approaches, not a single solution.

It was also noted that for many people on weight loss journeys, data begins and ends with stepping on the scales, and many naturally resist even that. BMI, while still the metric clinicians tend to reach for, is not always meaningful or motivating. There is a growing case for “non-scale victories”: tracking whether someone is sleeping better, moving more, or making healthier.

Building on the discussion around data, the panel turned to one of the most talked-about tools in modern healthcare, artificial intelligence – and what it could realistically offer in the context of obesity and preventative health.

There is genuine potential in using AI to scale interventions and understand behaviour at a population level. Given the complexity of obesity, AI’s capacity to support truly personalised approaches is significant. Many current interventions treat people the same regardless of their individual circumstances. AI, if implemented well, could change that.

One of the most important threads running through the evening was the need to challenge misconceptions about obesity.

The traffic light system on packaging acts as an example of an ineffective approach. Labels show nutrition for a single product, not a full day’s intake, making it easy to accumulate an overall red without realising it.

The traffic light system on food packaging is one example of an approach that falls short of its intent. Labels show nutrition for a single product rather than a full day’s intake, making it easy to accumulate an overall red without realising it, and doing little to support genuinely informed choices.

For poorer communities, cheap, unhealthy food is often the most accessible option. Cooking from scratch takes time, money, and access to ingredients that not everyone has. The economics of healthy eating are simply not equitable, and any approach that ignores this will disproportionately fail those who are already most vulnerable.

The biggest misconception surrounding obesity is that everyone who lives with this got there through overeating or poor choices. The reality is far more complex. Trauma, metabolic factors, genetics, and environment all play a role. It is not about willpower. It is not about eating less.

Equally, obesity has long been framed through the rhetoric of personal responsibility, “they did this themselves”. This framing is both scientifically inaccurate and socially damaging.

The panel agreed that the very definition of obesity is poorly understood. Obesity is a chronic condition, meaning it can last a lifetime. Interventions that are short-term or immediate are, by definition, inadequate. Managing obesity can take years, and for some people, it is a lifelong process.

• Obesity is a chronic, complex condition shaped by a range of factors, including genetics, metabolism, environment, and lived experience – not a failure of willpower
• Preventative health requires a whole-system approach, addressing the social determinants of health, not just individual behaviour
• Data has enormous potential to transform our understanding of obesity, but must be ethical, inclusive, and representative to be effective
• Digital inequality is a real and present barrier – those with no digital footprint are being left out of data-driven solutions
• AI offers significant opportunities for scaling and personalising interventions, but carries risks around bias, accountability, and data quality that must be addressed
• Stigma remains one of the most damaging barriers to progress, both around living with obesity and around seeking treatment
• Collaboration across disciplines, sectors, and communities is essential – no single intervention, data set, or technology will be enough on its own

Our latest Nexus Connect event highlighted that obesity is not a simple problem, and it will not yield to simple solutions. But it also showed that the tools, knowledge, and will to make progress are there.

From AI-driven behaviour change tools to redesigned food environments, from better data governance to more compassionate public narratives, the ingredients for meaningful change exist. What is needed now is the willingness to bring them together, and to ensure that the people most affected are not left behind in the process.

Stay connected with Nexus to join future conversations and explore how collaborative innovation can transform cutting-edge research into real-world impact.