A reference system for how filling food really is
Walk down a supermarket aisle and the packaging will tell you a food’s calories, fat, sugar and salt. What it won’t tell you is how likely that food is to leave you satisfied - or how likely it is to make you overeat. Despite decades of nutrition science lurching from one dietary villain to the next - fat, then carbs, then ultra-processing - nobody had ever systematically modelled the roughly 132,000 real foods people actually eat, in all their physical, nutritional and sensory complexity, against how they affect appetite.
Professor James Stubbs, working within the wider SATMAP research programme, used University of Leeds Impact Acceleration Account (IAA) funding to build exactly that: a representative, evidence-based reference system for how satiating everyday foods and meals really are.
Impact
- Knowledge discovery: created a sampling method to obtain a representative sample of commonly available, ready-to-eat foods. This created a dataset covering hundreds of everyday foods. They then obtained a representative sample of 2000 members of the UK population. Analyses are revealing that the relationship between a food’s energy density and how filling it is isn’t a straight line, as long assumed, but a curve with a distinct “satiety sweet spot.”
- Economic and societal impact: established a robust, nationally representative baseline that can be used by the weight-management industry, digital health providers and food manufacturers to help people understand which foods are the most filling, to explain exactly what it is about those foods that makes them more satisfying and to help people navigate through the complex food environment, to select foods that are filling, nutritious and enjoyable.
Key information
- Major funders: BBSRC/MRC Impact Acceleration Account (IAA), within the wider SATMAP research programme; commercial weight-management sector funding supported the programme’s earlier phases
- Partners and collaborators: wider SATMAP research team at the University of Leeds, including related work by Professor Graham Finlayson on food reward
- Disciplines: nutrition science, psychology, behavioural science
- Investigators: Professor James Stubbs, Dr Clarissa Dakin, Professor Graham Finlayson.
The nutrition field has spent decades dissecting individual components of food - fat content, carbohydrate type, energy density, degree of processing - and asking what each one does to appetite and body weight in isolation.
But people don’t eat isolated nutrients; they eat a banana, a sandwich, a bowl of spaghetti bolognese. Nobody had modelled real, whole foods in the way pollsters model a population: by taking a genuinely representative sample and asking how people actually respond.
For most people, health is lower down the list of priorities than cost, convenience and pleasure — so if we want to help people eat well, our tools have to work within that reality, not against it.
– Professor James Stubbs
The wider SATMAP project set out to answer three questions: can you rank a representative sample of the food population by how satiating those foods are; can you explain that ranking using foods’ nutritional, physical and sensory characteristics; and can that explanation then predict how new or unseen foods will affect appetite and weight management, well enough to help people navigate a food supply of well over 100,000 items?
Solving the sampling problem
Answering those questions meant solving a basic statistical problem first: how do you get a genuinely representative sample from such an enormous range of foods? The IAA-funded work did this using a “ternary plot” approach - plotting a sample of 66,000 foods across a triangular matrix by their protein, carbohydrate and fat content, with separate triangles for higher- and lower-energy foods. This allowed the team to select foods representing every part of that matrix, in the same way a political poll selects a representative cross-section of voters.
Those foods, presented as standardised, unbranded photographs of ready-to-eat meals and snacks, were then tested against a nationally representative sample of the UK population - a crucial comparator. Much of the field’s existing data on food “fillingness” comes from studies run through commercial weight-management programmes, whose members tend to be a fairly narrow slice of the population.
The IAA funding gave the wider programme its baseline: a properly representative sample of both UK foods and the UK population, against which any other group’s results can be meaningfully compared.
What the data is starting to show
Where earlier landmark research, such as a well-known 1995 study, ranked the satiating power of 36 foods, the SATMAP programme is now studying 312 foods through surveys and 48 through detailed laboratory work, drawing on data from around 16,000 people in total. That scale is already overturning some long-held assumptions.
Energy density and satiety, long believed to be linearly and inversely related, instead trace a curve - with a cluster of unusually filling foods sitting in a “sweet spot” partway along it, combining both volume and nutrient density. Foods high in sugar and salt, meanwhile, appear to suppress satiety, plausibly by driving up palatability and reward rather than through their nutrient content alone - a finding directly relevant to current debates about ultra-processed foods and high-fat, high-salt, high-sugar foods.
The team argues that current systems for categorising foods, including those used to flag ultra-processed foods, sort foods into broad bins without explaining why they affect appetite the way they do - of limited use to weight-management providers, food manufacturers or the people trying to eat well within real-world constraints of cost, convenience and enjoyment. As Stubbs puts it, most people’s food choices are driven first by cost, convenience and pleasure, with health further down the list - so any tool meant to help people eat better has to work within that reality, not against it.
With analysis of the representative national dataset now underway, the team’s goal is to turn this evidence base into practical tools: helping people navigate the food environment towards choices that are genuinely filling, nutritious and enjoyable, and giving industry a scientifically grounded way to reformulate foods and predict how new products will affect appetite before they ever reach the shelf.
This project was sponsored by the UKRI Impact Acceleration Account.