STELLA lived on the outskirts of Cape Town, South Africa. It was a beautiful, rural setting just below Table Mountain, surrounded by vineyards, trees, wild fynbos heathland and scattered settlements.
In 2010, Caley Johnson, a graduate student of anthropology at City University of New York, arrived to study Stella. For 30 consecutive days she followed her, watching and recording exactly what, and how much, she ate.
Stella’s diet was extremely diverse: almost 90 different foodstuffs over that time. On the surface, she didn’t appear particularly discerning. And indeed, the ratio of fats to carbohydrates in her diet varied widely from day to day.
But when Johnson crunched the numbers, something interesting popped out. When she looked at the ratio of combined daily calories from carbs and fats to calories from protein, she always got close to 4:1. This happened every day, regardless of what Stella ate. Even more interestingly, this ratio was very similar to what is considered nutritionally ideal for a female of Stella’s size. Far from being indiscriminate, Stella was a meticulously healthy eater.
How did she calibrate her diet so precisely? Doing so is difficult, and even professional dieticians have to use computer programs to do it. But Stella didn’t have access to a program because she was a wild Cape baboon.
The Stella study is one of many that we have been involved with over the course of our 30-year scientific collaboration. As a result, we think we have discovered something profoundly important about human nutrition, which changes how we understand appetite, explains the obesity epidemic – and suggests a way of solving it.
Our journey began in 1991, when we were colleagues at the University of Oxford. We set out to answer two questions. First, how do animals choose what to eat? And second, what happens if they fail to follow a healthy diet? To find out, we designed a huge experiment using the most voracious and indiscriminate eaters we could think of: locusts.
We put 200 young locusts in individual plastic boxes and prepared 25 different foods containing various proportions of protein and carbohydrates, the main nutrients the insects eat. The foods ranged from high-protein/low-carb to high-carb/low-protein, and everything in between.
Each locust was fed just one of the 25 formulations, in unlimited quantities, until they reached adulthood and shed their skin. This took a minimum of nine days and up to three weeks. We meticulously recorded how much each locust consumed each day, plus their weight and how much fat and lean tissue they had put on.
Once all the locusts had reached adulthood or died, we worked out which diet was closest to ideal. For that, we identified the mixture of protein and carbs that allowed locusts to grow and survive best. This turned out to be approximately 300 milligrams of carbs and 210 milligrams of protein a day.
Then we looked at what the locusts had actually eaten. Obviously they were restricted by their diet, but what was striking was that all of them managed to get close to the ideal amount of protein, even if that meant missing the carbs target by miles.
The locusts that were given a low-protein diet, for example, hugely overate carbs, consuming more than double the target amount. And that came at a cost. They took much longer to reach adulthood and they got fat. Granted, it is hard to tell that a locust is fat because of its exoskeleton. But it is chubby on the inside, like an overweight knight wedged into a small suit of armour.
In contrast, the locusts on a high-protein diet ate too few carbs and were unhealthily lean. They were less likely to survive to adulthood, and those that did had too little body fat to survive in the wild.
This experiment documented for the first time the battle between two nutrients: protein and carbs. When the locusts’ food didn’t allow them to eat a balanced diet, they prioritised protein over carbs at great cost to growth and survival. In fact, we later realised that what we were seeing wasn’t so much a competition between nutrients as between two appetites – one for protein, the other for carbs. Locusts had two separate appetites.
Up to that point, appetite had always been viewed as a single entity, an all-consuming drive that compels animals (including us) to eat our fill. This was the first hint that there was more to it.
The next question was whether the two appetites worked together to help the locusts achieve a balanced diet. So we performed another experiment where each locust had unlimited access to two different formulations differing in their protein and carb content. They were free to eat as much of the two foods as they liked. Regardless of which foods they had available, they combined them in precisely the right proportions to always eat an identical – and ideal – balance of protein and carbs.
This demonstrated that when locusts have a wide choice of foods, their two appetites collaborate so they consume an optimal diet. But when they are given imbalanced foods, as in our first experiment, the appetites for protein and carbohydrate compete, and protein wins. That suggested that, more so than carbohydrate, protein has to be carefully calibrated in the diet. We were later to learn why. If an animal has too little, it can’t grow and reproduce, and too much protein speeds up ageing.
“We had found two appetites in locusts, but was the same true in humans?”
This raised a much bigger question. We had discovered two appetites in locusts. Was the same true of other animals?
That was the purpose of the Stella study and many others that we have done. These have shown that appetite-driven nutrient balancing is common across the animal kingdom. It has been documented in life forms as diverse as slime moulds, cockroaches, beetles, spiders, cats, dogs, mink and non-human primates. Some turn out to have not two, but five appetites, three for the main macronutrients (protein, carbohydrates and fat) and two for specific micronutrients – sodium and calcium. Given a range of foods to eat, they will always precisely calibrate their intake.
This naturally got us thinking: do humans also have several appetites?
Answering this question wasn’t going to be easy. Human nutrition science has always been bedevilled by the difficulty of getting an accurate record of what people eat. Most research relies on study subjects self-reporting. The trouble is, people forget.
Ideally, you want to treat your human subjects like locusts: keep them in isolation with only the food you provide, all weighed and measured. However, this doesn’t get people banging down the doors to volunteer as participants.
Fortunately, we found a compromise. One of our students had access to an isolated chalet in the Swiss Alps, far from shops or restaurants. She recruited a group of 10 friends and family and took them there to spend a week as human locusts.
For the first two days, participants chose whatever they wanted to eat from a highly varied buffet. Everything they ate was weighed, and their intake of calories, protein, carbs and fat was recorded (caffeine, alcohol and chocolate weren’t available).
On days three and four, the volunteers were divided into two groups. One group got a high-protein buffet, the other a low-protein, high-carb and high-fat buffet. For the final two days, they returned to the original diet.
In phase 1 of the experiment, our human locusts reliably got about 18 per cent of their calories from protein, in keeping with studies that show people typically need 15 to 20 per cent.
In phase 2, everyone maintained their absolute protein intake. But to do so, those on the low-protein diet had to eat 35 per cent more total calories, while those assigned the high-protein diet ate 38 per cent fewer calories. Our volunteers responded like locusts, with their appetite for protein dominating, and determining the total consumption of food.
Your five appetites
We later did two bigger and more sophisticated versions of the chalet experiment, in Sydney and Jamaica, and found essentially the same thing: people on a low-protein diet consume more calories.
The explanation for this is that humans also have more than one appetite. In fact, we have the five that our earlier research found in some other organisms: protein, carbs, fats, sodium and calcium. It is a mistake to think of appetite as a single, powerful drive to eat. We need separate appetites to keep track of various nutrients, and hence to construct a balanced diet.
Those five have been singled out by evolution for good reasons. One is that there is a limit to how complex biological systems can get and still operate efficiently. We couldn’t have specific appetites for dozens of nutrients. Another is that these nutrients are needed in very specific quantities. Third, some components, like sodium, were often rare in our ancestral environments and we needed dedicated machinery to seek them out, for example in mineral deposits.
What about vitamins and the other essential minerals? We probably didn’t evolve specific appetites for them because our natural diets are rich in these nutrients, and by eating the right amounts of the big five, we automatically get enough of the rest.
As a result of our discoveries on the ways in which nutrient appetites interact – the dance of the appetites, as it were – we were confident in putting forward another hypothesis: in a food environment that is protein-poor but energy-rich, people will overeat carbs and fats as they strive to reach their protein target.
If true, the implications would be huge. It may come as a surprise, but we do actually live in a protein-dilute, energy-rich food environment. According to the UN’s Food and Agriculture Organization, between 1961 and 2000, the proportion of protein in the average US diet fell from 14 per cent to 12.5 per cent, with the balance made up of fats and carbs. Given that shift, the only way people in the US could have maintained their target protein consumption was to increase total calorie intake by 13 per cent – more than enough to create an obesity epidemic.
Intriguingly, in our experiments with people, we found that most of the extra calories eaten by those on a low-protein diet came from savoury snacks, especially those that tasted of umami, the signature flavour of protein. Protein-deprived subjects were craving things that tasted like protein, even though they were made of carbs. Our food environment is awash with such umami-flavoured carbs and fats, which we call “protein decoys”: crisps, instant noodles, crackers and so on.
These are also known as ultra-processed foods, which we now see as the main cause of the obesity epidemic. We are hardly the first to make that claim, but our research suggests we were looking at the problem of overconsumption the wrong way. It has less to do with these foods being full of fat and carbs than with them being depleted in protein.
Ultra-processed foods are industrial creations designed to be irresistible. They include such common fare as pizzas, crisps, breakfast cereals, sweets, bread, cakes, mayonnaise, ketchup and ice cream. More than half of the typical US and UK diet is made up of ultra-processed foods, and some people eat them almost to the exclusion of everything else.
The thing about ultra-processed foods is that they tend to be low in protein – which is expensive – and high in cheap carbs and fats. It is these foods that have largely been responsible for the dilution of protein in Western diets since the 1960s. And the more ultra-processed foods people eat, the more calories they need to consume to get the target intake of protein, with disastrous consequences.
Ultra-processed foods make us fat, but not because we have strong appetites for the fats and carbs they contain, as is often thought to be the case. Rather, it is because our appetite for protein is stronger than our ability to limit fat and carb intake. So, when protein is diluted by fats and carbs, our appetite for it overwhelms the mechanisms that normally tell us to stop eating fats and carbs.
Ultra-processed foods also contain very little fibre, which is filling and so puts a brake on appetite. Their frequent flavouring with umami, which our protein appetite craves, only makes matters worse. As a result, we eat way more than we should.
This realisation set us up to tackle the biggest challenge of all. Can this new view of appetite help us to fix our problems? The answer is yes. Here is how to take charge of your food environment and help your appetites work for rather than against you.
How much protein?
The initial step is to calculate your protein target. First, look up the daily energy requirement for your age, sex and level of activity. You can do this with something called the Harris Benedict equation calculator, available on numerous websites.
Next, work out the portion of those calories that should come from protein by multiplying it by roughly 0.15 (that is, 15 per cent of energy from protein; this multiplier varies depending on age: 18 to 30-year-olds require 18 per cent (0.18), people in their 30s need 17 per cent and those over 65 should get 20 per cent). Then divide the resulting number by 4 to get the number of grams of protein per day you should eat (a gram of protein contains 4 kilocalories of energy).
Finally, work out how to obtain that from protein-rich foods such as meat, fish, eggs, dairy, pulses, nuts and seeds. This is slightly complex, but the protein content of all these foods is available online and on food labels.
Everything else flows from this. It will satisfy your protein appetite and automatically ensure that you don’t overeat carbs and fats. In fact, you don’t need to keep track of these at all, as your protein appetite will manage them for you. Just make sure you supplement the high-protein foods with mostly wholefoods, mainly plant-based, which will also supply the fibre you need.
Most important, avoid ultra-processed foods. Keep them out of the house. You will eat them if they are there. They are designed to be irresistible.
If you follow these steps, the rest should be easy. All you have to do is listen to your appetites – they will guide you towards a healthy and satisfying diet. That is what they evolved for: to work for you, not for processed food companies.
newscientist.com, 20 May 2020