Date read: 26-03-2024

Author: Chris van Tulleken

How strongly I recommend it: 7/10

Buy the book

Read more summaries

Most of what we eat today is ultra-processed food (UPF). They contain things we’ve invented, but are not actual chemicals found in real foods, like emulsifiers, stabilisers, gum, and more. We’ve only now started to look at how bad these chemicals, and therefore “food”, are for us.

My takeaway: Stop eating UPFs and focus on eating real foods.

My notes

UPF has a long, formal scientific definition, but it can be boiled down to this: if it’s wrapped in plastic and has at least one ingredient that you wouldn’t usually find in a standard home kitchen, it’s UPF.

In any event, Paul explained how ingredients like emulsifiers and gums help in the making of UPF – and in cutting costs. First, they make the ice cream tolerant of warmth, which makes the process of moving the ice cream around easier. From factory to truck, truck to supermarket, supermarket to your freezer at home, ice cream will go from –18ºC to –5ºC and back down again many times. The gums, glycerine and emulsifiers all stop ice crystals forming by holding water close to them. This means that ice cream can be made in bulk in one factory and then transported around the country. It allows the supply chain to be a little less rushed at each stage and reduces the need to maintain very low temperatures. ‘Customers like creaminess,’ Paul said, ‘not shards of ice!’ Centralised manufacturing also allows the companies to negotiate a price with a retailer for shops around the whole country, which further cuts their costs.

Chemists realised in the nineteenth century that by chemically modifying starch they could create the exact properties they required. Modified starches, which you will start to notice in so many UPF ingredients lists, can replace fats and dairy, hold water during freezing and bulk out any sauce. With the taming of starch, came the possibility of turning very cheap crops into unimaginable amounts of money.

Once you can modify a starch precisely, there’s very little you can’t do. Thin your starch with acid, and it’s useful for textiles and laundry. Treat it with propylene oxide, and you get that gloopy feel for salad dressings. Mix it with phosphoric acid and you can improve stability through multiple cycles of freezing and thawing – perfect for pie fillings. And maltodextrins (short glucose polymers – a form of modified starch) can do things like giving a surface sheen and creaminess to what people think is a ‘milkshake’. No more need for expensive dairy fats: these starches come from crops that can be grown at vast scales and at a fraction of the cost.

Oil for UPF needs to be bland, plain and flavourless, so that it can be used to make any edible product – thus the use of RBD. So, manufacturers refine the oil by heating, use phosphoric acid to remove any gums and waxes, neutralise it with caustic soda, bleach it with a bentonite clay, and finally deodorise it using high-pressure steam. This is the process used to make soybean oil, palm oil, canola (rapeseed) oil and sunflower oil – four oils that make up 90 per cent of the global market – and any other non ‘virgin’ or ‘cold-pressed’ oils.

Take, as an example, the large study of more than 100,000 people that was published in the British Medical Journal, which suggested a link between UPF and cancer. The teams from France and Brazil looked at the risk of breast, prostate, colorectal and overall cancer, and found that, with a 10 per cent increase in the proportion of UPF in the diet, there was a roughly 10 per cent increase in the overall risk of cancer and the risk of breast cancer.

His idea was founded on the fact that, as well as fuel, Germany was short of edible fat. By the 1930s, the country was consuming around 1.5 million tons of fat per year but was only able to produce about half that amount domestically. They depended on importing linseed from South America, soybeans from east Asia and whale oil from the Antarctic. Imhausen was working on techniques to turn paraffin into soap and realised that, because soap is chemically a lot like fat, if he could make one, he could make the other.

Imhausen partnered with Hugo Henkel, inventor of Persil, and they founded the Deutsche Fettsäure Werke in 1937. This merged with IG Farben, a vast German chemical giant, and by 1938 they were making high-quality fatty acids. From there, it was a simple step to add glycerine, and so produce ‘Speisefett’ – edible fat.

Eating has become, in part, an intellectual rather than a purely instinctive project. Many of us consider calories, portion size, good food and bad food, vitamins, etc. Eating purely by instinct, in the way that a cow does, rather than trying to follow the advice of food packaging or nutritionists is an approach nearly unimaginable to many people. The idea that humans might have an internal system just like Eddie’s cows that allows us to self-regulate and balance our diets seems unlikely considering how little we are trusted by the authorities to eat without guidance. Could humans really leave eating up to instinct?

For the whole of the second age of eating, children of all mammals had eaten more or less what adults ate. There might have been some extra mashing and softening, and perhaps a little less spice, but there was no ‘baby food’ – just milk, then food.

However, by the 1920s, feeding a child had morphed into a quasi-science in the USA.

The plan was simple but quite revolutionary. Davis would let the infants choose their own food and then measure if they could be as healthy as infants who were fed ‘prescribed’ diets using the best nutritional advice of the time. She chose children who had been exclusively breastfed up to the very start of the experiment, so that they had ‘no experience of the food or of the preconceived prejudices and biases about food’.

Her hypothesis was that, since the human body has internal regulatory mechanisms for water and oxygen intake, heart rate, blood pressure, body temperature and every other physiological variable, the same should be true for body composition and nutritional intake.

A further twelve infants were recruited over the next few years, and they all settled into the diet just as enthusiastically. Almost all of them tried everything they were offered at least once, and their appetites were ‘uniformly good’: they often greeted the approaching food trays ‘by jumping up and down in their beds’. Once they were at the table, they devoted themselves steadily to eating for fifteen or twenty minutes, then ate intermittently, ‘playing a little with the food, trying to use the spoon and offering bits to the nurse’.

The experiment was an enormous success. There were just two children who wouldn’t eat lettuce, and one who wouldn’t try spinach. All the infants succeeded in managing their own diets, and all met their nutritional requirements as if they’d been reading all the latest textbooks. Their average calorie intake was found to be within the limits set by nutritional standards of the day, and they were free of all the usual feeding-related problems that are still staples of paediatric practice today.

Perhaps the best argument for internal nutritional regulation concerned Earl’s rickets. He’d arrived with the condition, in which the bones become soft and weak. There are X-rays of his little hands in the paper, taken when he first arrived, good enough to see the reduced bone density and the loss of the hard outer cortex of the bone. The growth plates at the ends of the bones are indistinct and fuzzy, and in an accompanying photograph, Earl looks bowlegged and distressed.

So, Davis immediately proposed a treatment for Earl: ‘Bound by a promise to do nothing, or leave nothing undone, to his detriment, we put a small glass of cod liver oil on his tray for him to take if he chose.’ Cod liver oil was, at the time, the only edible source of vitamin D. Over the first three months of the experiment, Earl drank the little cup of oil ‘irregularly and in varying amounts’ until his blood calcium and phosphorus reached normal levels and his X-rays showed that the rickets were healed, at which point he stopped drinking the stuff entirely. After it went untouched on the tray for more than a fortnight, the nurses stopped offering it.

The other children followed that same pattern, too. Whatever problem they arrived with, once they were allowed control of their nutrition, according to Davis, they all quickly reached optimal health.

But Davis was very clear this should not be the verdict – adults need to teach children what to eat to avoid poisoning and so forth – but she did think that, once safe food is established, we should recognise that children should be learning to self-regulate their eating in response to what they need, sending signals back and forth between the brain and the gut.

Remember Blair saying that there was ‘virtually no compelling evidence’ that fast food and sugary drinks are to blame?

Coke helped Blair establish that non-profit group, the Global Energy Balance Network, which promoted the message that there was no compelling evidence of a significant link between sugar-sweetened beverages and obesity. Coke funded all those papers I listed earlier, by Blair, Hill and Katzmarzyk. Coke even funded an entire national programme, run by the American College of Sports Medicine, called ‘Exercise is Medicine’. Steven Blair has served as the vice president and the president of the American College of Sports Medicine.

A team from Oxford and the London School of Hygiene & Tropical Medicine mapped the universe of Coca-Cola’s research funding, which involves almost 1,500 different researchers (probably not all direct grant recipients), corresponding to 461 publications funded by the brand. The researcher who has published the most articles (eighty-nine) with Coca-Cola funding is Steven Blair. His research institution received a total of around $5.4 million of research funding to study the role of energy balance at high levels of energy intake.

She told me at length about how the plastics from UPF packaging, especially when heated, significantly decrease fertility (and according to some experts, may even cause penile shrinkage.)

Whenever I talked about the ‘food’ I was eating, she corrected me: ‘Most UPF is not food, Chris. It’s an industrially produced edible substance.’

The basic construction materials of UPF are industrially modified carbs, fats and proteins, and the processes they are put through remove almost all the chemical complexity. The intensity of ultra-processing means that vitamins are destroyed (or deliberately removed in the case of bleaching), fibre is reduced, and there’s a loss of functional molecules like polyphenols. The result is lots of calories but very little other nutrition

Whole foods contain thousands more molecules than manufacturers add back in, and it is these molecules’ more subtle health effects that could be responsible for the well-established benefits of eating whole food – protection against cancers, heart disease, dementia and early death.

Dana Small, a neuroscientist at Yale, did a series of experiments in humans which demonstrated that whether we learn to want a particular flavour seems to depend on how much our blood glucose changes when we consume it. The team gave volunteers randomly flavoured drinks, and after a few exposures they learned to want the flavours that had been paired with a tasteless carbohydrate, maltodextrin. The more their blood sugar went up, the more they wanted the flavour.

‘Factory farming and UPFs are two sides of the same industrial food coin,’ Percival said. ‘And then, of course, lots (though not all) of that factory farmed meat is subsequently turned into UPF.’

The result of this is that, of the thousands of different strains of plants and breeds of animals that have been cultivated since the birth of agriculture, just twelve plants and five animals now make up 75 per cent of all the food eaten or thrown away on earth

So dominant is soy as industrial animal feed that the average person in the UK or Europe consumes approximately 61kg of soy per year, largely in the form of animal products such as chicken, pork, salmon, cheese, milk and eggs

Inland rain requires trees. Rain clouds on their own cannot travel more than 400km from the sea, so rain in the centre of a continent – the very rain that creates the central forest of the Amazon for example – requires continuous forest to the coast. Around half the rain that falls on the Amazon comes from its trees

Those trees ‘breathe out’ water vapour, which creates new clouds that travel further inland in so-called ‘flying rivers.

Crucially, this is how water reaches the soy and corn plantations in central and western Brazil. Once you destroy the forest you get less rain. A 2019 study showed that the rainy season in the state of Mato Grosso had become a month shorter in a decade, and many of the major soy farms in Brazil are now suffering from the very drought that they have caused.

A team at Cambridge found that many parents were feeding their babies much more formula than needed. Milk, of course, solves all problems – crying, teething, and so on – and so the babies were consuming hundreds of calories more per day than recommended by the World Health Organization. Babies were overfed so that some were drinking as much as a litre of formula per day. These babies were being taken to the GP due to symptoms such as crying, fussing and vomiting, and then being diagnosed with allergy or reflux before being prescribed expensive specialist formulas. But, when parents reduced the amount of formula they were offering to recommended levels, these symptoms resolved for many babies.