Number crunching: why ultra-processed foods have a calorie problem | Food & drink industry

A calorie is a calorie, right? Fixed and unchanging, like a gram, or a mile? Well … no, not necessarily. You see, what a straightforward calorie count on a restaurant menu or food packet can’t tell you is how your individual body will use those calories. This comes down to multiple factors including genetics, gender, age, hormones, gut microbes, sleep patterns, the time of day we are eating, how active or sedentary we are, our body fat and muscle mass, and – crucially – what sort of food the calorie is in. Our bodies are much better at absorbing the energy from a calorie of low-fibre, processed food (like a potato chip) than they are at taking in calories from whole foods, like an apple.

Calories are a measure of the heat (energy) given off when a food is completely burned away in a pressurised bomb calorimeter. “But we don’t eat calories. We eat food,” says Dr Giles Yeo, an obesity researcher at Cambridge University and author of Why Calories Don’t Count. That energy is used differently by different bodies.

“On average, in high-income countries, we get 50% of our calories from ultra-processed foods (UPFs),” says Yeo. There is no legally binding or agreed upon definition of UPFs, and according to a 2019 review article in the journal Current Developments in Nutrition, definitions used since 2012 vary hugely. For some food scientists, they are foods which are mainly made from additives of various types; for others, they are foods containing few or no wholefood ingredients, made with ingredients consumers can’t usually buy in shops (such as stabilisers), which are also often fortified and which take the place of freshly made meals or snacks. For still others, they are industrial formulations with five or more ingredients; or foods containing additives designed to mimic unprocessed foods or disguise any undesirable qualities in the finished product. Things can get very confusing: according to one definition, a tin of beans in tomato sauce is ultra-processed, but the same beans, canned in water, are merely moderately or minimally processed.

But they aren’t just fried chicken, crisps and ready meals: croissants, hummus, pesto, biscuits, breadsticks, gnocchi, cereal bars, ice-cream, fish fingers and curry paste are all UPFs. And in some UPFs, their original ingredients – whether they’re being processed to be longlife, cheap or just conveniently moreish – become a blank (or bland) canvas for what’s added after. “Ultra-processing strips out flavour in food – just because of the way it’s done – and flavour comes from the holy trinity of sugar, salt, and fat, which you have to add back in,” says Yeo. “So, on average, they are high in sugar, salt and fat, and low in protein and fibre, which makes the food very much more calorically available: you get a lot more out of a calorie in an ultra-processed food.”

The energy from 100 calories in a high-fibre food like chickpeas won’t behave in the same way in the body as 100 calories in a low-fibre food, such as an iced bun. The energy from the chickpeas will take longer to be digested, be slower to reach our bloodstream – and, unlike the energy from the ultra-processed iced bun, is far less likely to cause an unwelcome rush of glucose and insulin or be stored as fat.

Rather than overall calorie content of foods, the more useful thing to know would be the caloric availability of a food to our individual bodies. But you can’t easily turn that into a neat number on a label or restaurant menu.

Those neat numbers come from work by Wilbur Olin Atwater, a chemist born in 1844 in New York. His painstaking, years-long analysis involved vaporising a huge number of foods in ruthlessly efficient bomb calorimeters, which unwittingly resulted in our modern calorie-counting obsession. He calculated that fat gave us nine calories per gram, and carbohydrate and protein four calories each per gram. Once you know how much fat, carbohydrate and protein are in a food, using Atwater’s data, you can figure out how many calories it contains – which is exactly what most food manufacturers – and now many restaurant chefs – have to do. It’s maths, not lab work.

It’s alluringly simple, but Atwater’s methods weren’t foolproof, and he probably never intended his data to be used as it is today. “Atwater rounded everything up and took averages – so errors were baked in,” says Yeo. Atwater also lived in a very different food environment and based his averages on diets which were more likely to feature mutton than avocados, and lard rather than olive oil – and, crucially, diets in which UPFs didn’t feature. “Which is where the margin of error begins.” Atwater also misunderstood how many calories a body can gain from protein – for every 100 calories in protein, we can actually only take in 70. “Protein calories are 30% wrong everywhere,” says Yeo, because every time the Atwater data is used the error is repeated. “Carb calories are about 10% wrong for anything with fibre, and 5% wrong for white flour or sugar. Fat is the only one for which the Atwater factors are still accurate. [Food manufacturers] also don’t empirically determine how much protein is in something. It’s estimated. So on top of the baked-in errors, there’s that additional error, too.”

A 2021 meta-analysis suggested that diets higher in UPFs weren’t just higher in calories (a diet made up of 75% UPFs contains around 200 more calories per day on average than a diet with 15% UPFs) but, more importantly, were also higher in sugar and lower in fibre and most other micronutrients. “The problem [with UPFs] is we eat too much of them rather than them being dangerous, per se,” says Yeo.

Does this mean calorie counts are pointless? Not entirely. Clare Thornton-Wood is registered dietitian and spokesperson for the British Dietetic Association. “In clinical practice, we use calorie calculators to work out how much somebody needs if you’ve got someone in ICU and you’re feeding them on a fluid feed,” she says. “We use calories as a starting point. You might say that the average man needs 2,500 calories and the average woman needs 2,000 – but who is this average man or woman?”

“The calculation for how many calories you should be eating a day depends on a billion different things,” says Yeo. Is there any way to work out what any one person needs? “We could, but that would mean sticking you in a chamber calorimeter.” Unfortunately, these are rare, very expensive and mean living in a sealed room for three days, with scientists measuring your every breath, movement, consumption and excretion.

There is some data from the Medical Research Council Epidemiology Unit in Cambridge that suggests showing calorie counts in places like coffee shops may reduce the calorie content of purchases by about 8%. “But is that the goal?” asks Yeo. “Are we trying to get people to eat 8% less of everything? You can have a really unhealthy diet, but eat less of everything. Or should we be trying to get people to eat less unhealthy stuff and more healthy stuff? That is where calories are useless. Because calories don’t tell you about the health of the item.”

There are, for example, 678 calories in a Pret a Manger hummus salad, 684 in three Mars bars, and 708 in a portion of a Sainsbury’s fish and chips ready meal. These very similar numbers don’t tell us that the salad provides a third of our recommended daily fibre and half our daily fat; that the fish and chips contain almost half our daily salt but also half our daily protein; or that the chocolate bars would bust our sugar allowance. By only looking at calories – as on restaurant menus – we lose other, much more helpful information.

“Thinking in calories trains us to view more caloric meals as bad and low-calorie meals as good,” says Virginia Sole-Smith, an anti fat-phobia campaigner, author of The Eating Instinct and Substack anti-diet newsletter Burnt Toast. It reinforces the misunderstanding that food choices are moral choices. “There are so many reasons why a high-calorie meal can be the right choice for someone. Maybe you skipped breakfast and lunch is your first chance to put food in your body. Maybe you’re getting ready to do a long run. Maybe you’re out to dinner with friends, an experience that offers tremendous mental health benefits. Choosing the low-calorie salad in any of those scenarios won’t necessarily serve you. But we’re conditioned to believe it’s always the right choice.”

Henry Dimbleby led the National Food Strategy, an independent review for the UK government. He thinks a more powerful policy than calorie labelling would be a reformulation tax, dealing with the least healthy of our ultra-processed foods. “We have to make it less attractive for companies to sell the stuff that makes us sick. It’s very politically difficult when you have a cost of living crisis, but I’m convinced it will happen – it’s either that or the NHS falls over.”

Yeo is sanguine about the chances of us giving up or even cutting down on UPFs, but would like manufacturers to reformulate them, with more fibre added. “If you can make a plant-based burger bleed, I think you can add a bit more fibre into a chocolate bar,” he says. “We need to work with the people who are making the food. Whenever I say this out loud, the ‘Real Food’ brigade pop up: I am not trying to compare a chocolate bar – however much protein or fibre is in it – to a banana. Clearly, a banana is a banana, but sometimes, I want a treat. If you feel the need for a chocolate bar, or a lasagne, or a burger, rather than a banana, could there be a slightly healthier version of that burger or chocolate bar or frozen lasagne that you could buy?”

He would like to see healthier food subsidised, too – not just carrots, but healthier processed foods as well. “At the moment, the cheapest choice is typically the most unhealthy choice.”