Protein Leverage

These highly acclaimed research doctors and Michael have a fascinating discussion about how protein drives our appetite and how the ratios of proteins to other nutrients can lead to a longer or shorter life and have a direct impact on obesity and other health issues. All animals have a built-in system to detect and manage levels of nutrient intake, and this system has been intentionally and radically broken by the food industry.

Share this interview:
Subscribe to our YouTube channel!
Don’t miss out on any of our content.


Michael: My guest today are doctors Stephen Simpson and David Raubenheimer who, in the 1980s, were doing research in how arthropods and other animals balance and prioritize different macronutrients — like proteins and carbohydrates, and they made a discovery that really revolutionized the way that we think about and understand nutrition.

I remember some years ago hearing from a researcher that: regardless of the diet that we eat, about 15 to 18% of what we eat in the diet is protein. And that really surprised me at the time — and didn’t really make sense to me why that would be. So, when I first came across your research, it was great. It suddenly made it all make sense to me, in what you call protein leverage.

So, I just want to kind of give my sort-of simplistic understanding, and you guys can correct me. But as I understand it, your theory started with this idea that different animals will eat different food foods that have different ratios of, for example, proteins and carbohydrates, because they’ll have a ratio — I think you started with slime molds and locust that, if I remember right, both had a 2:1 ratio of protein to carbohydrate that they were looking for in their diet, so if they ate something that had a higher ratio of carbohydrate, for example, they might eat something else that has a higher ratio of protein. And so, they’re trying to get to this point where both the ratio of 2 to1 protein to carbohydrates is correct and, also, the amount of nutrition that they need is correct.

And if there’s a case where they don’t have access to foods that would allow them to kind-of zigzag up to that point, then they might — they would, in most cases… I think most animals prioritize protein, so therefore they might make the choice — not consciously, but biologically  make the choice — to eat more carbohydrates so that they can meet their protein target. And, so; that has ramifications for animals and humans as well, because that’s, as we’ll talk about later, one of the things that can lead to obesity in people if they’re eating foods that, in order to get that satiety of protein, gives them more carbohydrate than they should be eating.

Stephen: Exactly as you’re saying, Michael. The key discovery was in our early work on insects — particularly locusts. They were our model system at the time: an animal that is well-known to eat vast quantities; devastate agriculture; be a terrible pest… and we thought from the outset that if we wanted to understand an animal that really didn’t appear to care that much about how much it ate, we should study locusts. So, we began by exploring exactly their capacity to regulate — not only their intake of food — but their intake of specific nutrients. And [we] discovered, as you say, that they, like everything else that we’ve studied since, have specific nutrient appetites. They don’t just feel hungry or full, they feel hungry for protein or for carbohydrate or for salts (they were the three that we explored in our locust experiments), and that those appetites help the animal to navigate its nutritional world in a way that it selects foods to balance its diet… making up a short-fall in one nutrient by selecting foods that are rich in that nutrient and avoiding foods that contain a higher proportion than it needs of that given nutrient.

So that was the first real insight that animals have these specific appetites, and the second thing was that if you force those to compete with one another by putting the animal in a fixed-food environment where it can’t choose, then, in the case of locusts, we found that the protein appetite was dominant. It was the one that they cared more about, if you like. So, the animal would eat too much of everything else to get enough protein — or too little of everything else to get the right amount of protein, depending on the concentration of protein in the diet and that led on pretty well to everything else.

Michael: Okay; so, I know that there are several factors that this goes down to, in terms of you studied prioritization that they make over fecundity versus longevity, for example. You talked in your first book also about how animals will make dietary decisions as well about  — I think it was seabirds that were had carotene in their diet. And so, they would eat a certain amount of that and the female birds, I guess, would like them because, it’s like, “Oh, well, this guy knows where to / how to forage right!” So that would be a factor. Or there was another example you gave where the fish diet… they would wait until the fish were very high in… I don’t remember the fatty acid… but it was it was best for their offspring for their brain development, right? So, they would wait for that diet. So there’s lots of different factors that can drive diet, but primarily, for most animals, it’s going to be protein and then they — I think in humans there’s protein, carbohydrate, fat, sodium, and calcium, right. So, I’m very interested in how we, if the protein is met, … what we’re prioritizing next [and] what we do when it’s not met. I know that there are different ways, if the animal has to make a decision (and again … not a conscious decision), but about… do they care about the protein or do they [care about] the carbohydrates? Do they try to get closest to the target? Or, you know, how do they do that and under what conditions (which may change in their life or may change under the conditions that they’re in)? You talked at one point about — I don’t remember if it was grasshoppers – who, under the fear of spiders, would eat more carbohydrates because of the stress, right? So, if you could talk a little bit about that. I know it’s a very complex subject, but if you can try to simplify it and explain how that works.

David: Well, it might be best to step back and look at first principles here a bit. It’s very complex because there’s so many nutrients and so much complexity in a food environment where animals forage. But what it all boils down to is that the body needs different nutrients at different levels at different times to satisfy a particular function. So, you spoke for example about reproduction versus keeping warm with thermo-regulation — all of the requirements that the animal’s body has. And what we discovered, as Steve said, is that animals have specific nutrients linked specific appetites linked to these different nutrients that the body requires and the function of those appetites is to tell the animal’s brain, for foraging, what nutrient it needs at what levels at a particular time. So, when it’s reproducing, typically it needs more protein, for example. In primates that we studied in the wild, when they lactating they need more energy. So, what an animal prioritizes at a given time will depend what it most needs at that particular time.

Now proteins are kind of a special nutrient in that respect because we need it at a certain level (and animals need it at a certain level) on a daily basis to satisfy protein needs. But they can’t store protein, and that means that they need to ensure that every day they get the right amount of protein. [There’s] no use eating too much because they can’t store it. And when there’s too little, they can’t draw on stores to meet that protein need in the way that they can, for example, with energy. They store energy. So, in different circumstances different animals will prioritize different things, but generally, protein is the one that is prioritized most strongly for many species — but not all species, as Steve said.

Michael: Right. And that’s because of the nitrogen that, you know, … animals can basically clip off the amino group and convert proteins to carbohydrates or fats through gluconeogenesis or lipogenesis, but they can’t go in the other direction if they don’t have the nitrogen. Right? Yeah, so that’s I think kind of what drives that. But there are some species that are predators, I think you found, that didn’t prioritize protein.

Steve: Yes, indeed we found predatory species also, of course, require a sufficient amount of protein, but they’re better able to tolerate excesses because their diet is habitually rich in protein and nitrogen. And the thing that actually is of greater urgency for them, if you like, is energy-yielding nutrients. And that’s a function of trophic level. As you eat more — and more protein-rich — things; as you move up the chains in the food system, you become more and more limited by nonprotein energy. And so, that became more of a priority to predatory species, and we showed that in spiders and predatory beetles and predatory mammals as well — a whole range of different species.

David: and one of the things that enable them to do that, is that, as you mentioned earlier, on being able to deaminate amino acids and use the glucose in energy metabolism. Now, predators have a very-well-developed capacity for doing that known as gluconeogenesis. And other species that live in very high protein food environments do the same thing. They’ve evolved specific mechanisms for using amino acids in energy metabolism. For example, the mountain gorillas [Gorilla beringei] that I study with Jessica Rothman in New York — they live in a high-elevation, tropical environment where fruits are very scarce, so much of the food that they eat are high-protein leaves. And the way that they’ve evolved to cope with this is to develop these accentuated pathways for gluconeogenesis that enables them to satisfy both their protein and energy requirements from amino acids, and that’s the same as what predators do.

Michael: So, on that… one of the, I think, really important things in your research that we talked about a little ago is the longevity and the choice of simplifying sort of fecundity versus longevity. But I think you also had a study that showed that if the diet was optimal that you could sort of optimize both. And I think part of the reason for that is because of the… there’s a toxic element to the deamination of proteins in gluconeogenesis, for example, that can lead to disease or other things that affects the lifespan.

Steve: Yeah, so, it’s well known that fecundity (or lifetime reproduction) and longevity are… they tend to trade off. So you can either have lots of babies and not live as long or live a lot longer and have fewer babies. That’s very well known in in life history theory across all manner of species of animals. And there’s been a whole range of explanations for that — many of them relying on the idea that there’s some sort of a tradeoff. You can do one, but you can’t do the other… either because you’re trading off result resources, so you can spend your resources on one but not the other. Or, there’s a cost to reproducing that shortens your lifespan. And, actually, what we found in our research — initially in flies and then in a range of other insect species and then in mice — was: it’s not so much an inevitable tradeoff because of the costs of one versus the other. It’s just that reproduction and longevity have different nutrient requirements. The optimal diet for reproduction is a different combination of macronutrients than that which supports longest lifespan. And you can then start to delve into what that means, biologically, by looking at the nutritional responses of different markers of aging and reproduction. And we’ve spent quite a lot of time doing that — really trying to understand what are those differences. And the question is to whether you can achieve maximal lifespan and maximal reproduction on the same diet. Only under really special circumstances, it seems. If you titrate amino acids precisely in the protein part of the diet, you can reduce the difference between the optimum for lifespan and reproduction, but you can’t ever really make them exactly the same. There is an inevitable trade-off, and it’s because different biological processes have different nutritional needs. And one of the very interesting things in the evolution of nutritional phenotypes (the way animals respond to their nutritional world) is how they manage those tradeoffs. And that management changes with all sorts of things: your risk of being eaten by something else, disease risk, all sorts of other things, your life history strategy more broadly.

Michael: And did you find that with methionine specifically, if you raise levels of that, that didn’t have an impact on, you know… that sort-of allowed you to optimize both. And I assume there are other amino acids… I know you talk about branch-chain amino acids as well, and how they impact these factors.

Steve: Yes, so amino acid balance is a really significant part of the relationship between aging and health and diet, and it’s well known that that, say, the branch chain amino acids are linked to short shortening of lifespan, increased rates of protein synthesis, and one question is: is it branch-chain amino acids particularly, or is it something to do with their ratio with respect to other amino acids in the diet? Turns out to be a bit of both, but the optimal mixture of amino acids both essential and non-essential across the 20 common protein amino acids turns out to be unexpectedly very directly linked to the ratio of amino acids in the exome in the… so, the coding of amino acids in in our genome has a particular ratio of one to the other to the next. And if you get that ratio and you provide it in the diet… Matt Piper — a colleague who’s down at Monash University (down in Melbourne) — he showed in his work that in flies and mice if you can match that, you can get the best sorts of outcomes. It sort of minimizes the difference between longevity and reproduction. You can get a diet that sort of supports both of them better than if you have a less well-balanced complement of amino acids.

Michael: So, you’re talking about what nutritionists call “quality proteins,” right? Which means that they’re mapped to our needs or our human protein distribution as opposed to how they’re naturally in the food. So, meats are going to be a closer match to us than, you know, like we mentioned methionine… so rice will have more than that and legumes will have less of that, right? So, in different species it’s whatever is matched to what their particular protein needs are for how they construct proteins.

David: Right, that’s what the exome does, is that it matches availability to requirements. So, it’s actually using the specific amino acids that it requires for biological functions.

Steve: So, it’s a sort of 20-dimensional stoichiometry you have to… you want some things, you want them in the appropriate balance. And that speaks actually to the whole body of theory that and I have developed over the last 30, 40 years which is what we call Nutritional Geometry. And it, at its very heart, takes this principle: the fact that nutrient balance is a multi-dimensional concept. And achieving nutrient balance requires animals to mix many different foods in their diet to attain that (typically, anyway), and certain food environments won’t allow you to get to a balanced diet, and you have to therefore trade things off against one another. And different aspects of your life history will have different, what we call “targets,” and it just provides this way of conceiving and then translating into experimental and then observational research — the whole issue of what nutrition is and what does a balanced diet mean. It becomes not a sort-of fluffy, qualitative concept. It becomes a quantitative concept that you can explore using Euclidean geometry and detailed experiments.

Michael: Right, and it gets super complicated when you think about… I mean, just in humans, for example… I think you mentioned in one of your books, when humans are born, they have like a 7% protein target, and at different stages in their life it changes. And later in life, lifespan is optimized if we have a lower protein ratio. However, because of, I assume, sarcopenia, people have a need to get a certain amount of protein later in their life.

Steve: Yeah.

Michael: So, how you balance that and it’s… with all the factors, it’s just a lot.

Steve: Yeah. And it means that the one-size-fits-all dietary recommendation doesn’t quite capture these changes as we go through our lives. And in the book, we talked about the analogy of a leaky bucket. If you need to fill the bucket to a certain level of protein, and that level will change throughout your life course, … if you put holes in the bucket, you’re going to have to eat more to top the bucket up. And our bucket becomes leaky as we get into our older age as we start to break down protein unnecessarily. It also gets leaky when we start to suffer, for example, from insulin resistance, because insulin normally prevents the breakdown of protein and prevents gluconeogenesis — or at least it inhibits both. But if it’s not being responded to through insulin resistance, then the bucket becomes leaky and you got to eat more. And there are other key transitions in human life where this happens as well. So, another one is during perimenopause when, due to the hormonal changes, women start to break down protein at a higher rate and that requires a higher content of protein in the diet. Their target effectively goes up.

David: You know the same, Michael, it sounds as Steve said, it is horrendously complex, but actually, in terms of the way that biology solves this — it’s not that complex at all, because if you think about the number of computations that would be needed to optimize in an engineering sense across that many nutritional dimensions — it’s horrendously complex. But actually, biology simplified it such that if an organism pays attention to a few key factors within its environment, everything else comes along by correlation. Which is why the macronutrients appear to be so important for our species and other species as well. It’s that those that are the things that for most species that need to be paid attention to. But because they’re foraging in whole food environments, if they’re pegged to get the right amounts of macronutrients from the foods that they eat, they automatically they get the right amount of micronutrients as well. So, these kinds of changes that take place through the lifetime — that might take place with respect to lots of different nutrients — but paying attention to the key ones enable animals and people to solve these problems automatically. We’re not performing calculations, of course, that’s what our appetites do.

Steve: it was the reason why we studied such a wide diversity of things from locusts and slime molds through to mountain gorillas and others. It’s to seek inspiration from the natural world as to how Evolution has solved this horrendously complex problem independently so many times and in different nutritional environments. The idea was that if we could understand how apparently diverse and simple things solve this horrendous complex problem of nutrition without requiring apps and food diaries and access to books… if they could do it how did they do it and how has natural selection tamed the complexity and does that give insights into The Human Condition. And the contention is that, yes, it’s been solved beautifully endless numbers of times, and there are some really fundamental insights that can help us understand our own relationship with our modern diet.

David: Of course, the big issue here is that our biology is just the same as the biology of all of the species that we’ve studied. We also have this capability. What’s happened is that the food environment within which that biology operates has been that drastically changed. That those correlations that I mentioned… they no longer work. So, we do need to do computations if we’re eating in a food environment that our biology is not attuned to. But it’s not that difficult to change that. In a whole food, environment people following their appetites can regulate to a healthy diet without doing those calculations.

Michael: Right. One thing that you I heard you say, David, which I love, is that you shop with your brain, and you bring home good foods, and then you eat with your appetite, right? So, you know the big problem, as we all know, is this you know the Group Four NOVA foods that people eat, which are designed to make us eat more of them. And that’s one thing I wanted to ask you about later, or you could just address it now, is if there are [monosodium] glutamates or [disodium] guanylates or [other] things in them that sort-of trick us into thinking that we’re getting protein when we’re not… and so, we are trying to satisfy our protein craving and getting a lot more carbohydrates and trans fats or long-chain fatty acids or [other] bad stuff that affects us then like you say that kind of messes with our biology not working properly to have these cravings of foods that we need so that we don’t have to think about it.

David: Well, that’s a fantastic example where the correlation has been broken. Because in a natural food environment, those umami flavors — they signal protein. You don’t get umami flavors (except in very rare circumstances) where they’re not associated with protein. So, our biology has evolved to respond to those when we need protein, but the processed food industry has broken the correlation by providing those flavors in a diet that is cheaper to make, in a food that’s cheaper to make, because protein is expensive. For example, barbecue flavored crisps are virtually all fat and carbohydrate, but they taste like protein. So, the response is that we eat more of those when we’re craving protein. And what it does is it exacerbates our need for protein rather than satiates it, right?

Steve: That’s an excellent example. It’s what we’ve called “protein decoys,” and we’ve actually started to really begin to understand, both in our own research and colleagues elsewhere around the world working on this problem, that the control of protein appetite involves a series of key hormones. The preeminent among them, at least so far as we understand now, is the thing called Fibroblast Growth Factor 21 (or FGF21) — it’s released principally from the liver, and it’s released under low-protein circumstances. So, it’s a protein hunger hormone, and it turns on the appetitive behavior leading to wanting and liking Umami flavors. so it makes you seek savory flavors. And if the nearest savory flavor is, as David said, a packet of barbecue-flavored crisps, then you’re being “dudded” — you’re being hacked. That’s a protein decoy, and FGF21 is leading you astray, because — not because the appetite system isn’t working — it’s working perfectly — it’s just you put it in the wrong environment.

David: And it’s a fantastically clever commercial strategy, because what it does… in eating those foods is it actually intensifies the need for protein, so it increases protein hunger rather than satiates out the protein hunger — because we become more imbalanced towards excess fat and carbohydrate when we respond to those cues.

Michael: So, I’m curious if, you know: a lot of this model works — obviously we’re going to see foods, and smell them, and then next is we’re going to taste them. And then we’re going to have the post-ingestive effects that, I think, are what lead us, if I understand it, to, like, our calcium need — we’re going to associate that with certain foods, and so we’re going to maybe crave those foods in order to satisfy that need. So, I’m curious if after a time by eating umami-flavored crisps, as you mentioned, if the post-ingestive effect of that — if our body is going to go, “Wait a minute… this is not giving me the protein I need.” If it’s going to then alter that, or if it’s so ingrained in us that that that just doesn’t work

Steve: Look, that’s a really good question. We know… and we’ve actually done our own studies on insects showing that they’ll be fooled for a while but they’ll the post-ingestive feedbacks will help them to regulate properly. Whether that’s the case for humans in such a massively complex food environment where we’ve got such a variety that it becomes rather difficult to make specific associations… I think the evidence is that we aren’t getting better at ignoring these protein decoys. But why we’re not better at recognizing that we’re not getting the post-ingestive feedbacks from real protein that we should be associating with those flavors, I don’t know. It’s not something I know anybody’s looked.

David: So, there is evidence, of course, as you say, Michael, we do compensate: the physiological regulatory mechanisms compensate for those imbalances. But there’s evidence the compensation is incomplete, so there’s a kind of a ratchet effect; and, you know, one place we’ve seen this is in the analysis of population data. If you compare… if you look at people’s eating over the day in relation to the percentage of protein they’ve had — in the first eating period (typically in breakfast), those who have a low-protein breakfast tend to have a higher-protein lunch and dinner than those who have a high-protein breakfast. So, what happens is that they sort-of converge, but they don’t converge completely. So, people with a low-protein breakfast make up for it, but not entirely. And this shows both things happening. There’s compensation, but it’s incomplete compensation.

We did wonderful experiments on domestic cats showing this, where we flavored the foods. We messed up the macronutrient ratios, and we flavored the foods with different flavors: either things that they particularly like, like fish flavor, or things that they didn’t particularly like, like citrus flavor. And initially they responded just to the flavors, but over time they overcame those flavor responses and started balancing their diet again. And there wasn’t evidence for incomplete compensation. It was pretty good compensation.

Steve: And even in humans, we sort of think we’re going terribly badly wrong, and we are!, but we’re making a — there’s small errors that accumulate, rather than being as bigger set of errors as you might expect, so if you if you actually went into supermarkets and shut your eyes and randomly pull things off the shelves, … we regulate far better than that, so we’re not completely subject to our food environment. Our appetites are trying to get us to the right place, and… but making accumulating errors and the problem is that those errors because they translate into issues with our metabolic health and our stores of body fat, they in turn start to change our protein target through the mechanisms we spoke about earlier. You become less efficient at using protein as you become insulin resistant and obesity starts to set in. And so, you get a vicious cycle happening but the errors that actually lead you to that are — it’s pretty impressive how well we’re regulating in our nutritional environment at the moment, given the vast preponderance of these ultra-processed foods or highly industrially processed foods which are breaking all the correlations in our expected nutritional environments relative to whole foods.

David: And there are massive constraints in relation to what food environment different people are foraging in. For example, financial resources determine, hugely, what access you have to quality protein. So you, in many circumstances in food or nutrient deserts that are becoming increasingly common in places like the United States and in areas in Australia as well, … is that to follow those signals in the way that biology intended becomes very, very difficult, and it’s not an issue of small increments. It’s an issue of being chronically confined to diets that generate obesity. No wonder the stats are going in the direction that they are — not because of biology, but because of the food environments we have access to.

Michael: Right. Well, that’s obviously a huge problem worldwide, but as you mentioned, there’s food apartheid and there’s economic … there are all kinds of reasons why people maybe not have access to food or choose from this [particular] set of foods. I was going to mention: I’m sure you guys are familiar with these studies, but Anthony Sclarfani did a study with rats, I think, where he fed them grape and cherry [flavored drinks] and studied the post-ingestive effects of the sugars in both. And then Dana Small did a study with humans with maltodextrin and, basically, [was] to try to do the same thing to see if we can learn. And she made these flavors that were new flavors that no one would have tasted before, so we wouldn’t have associations already. And [she] found that there was a statistical difference in terms of people favoring the one that had the maltodextrin in it, but not a huge difference. But when she looked at the nucleus accumbens, in the brain scan, it was sort of lighting up. So, there was a big positive effect there. And which I found interesting — the whole experiment — but particularly that the effect in the brain was more pronounced than people’s favoring… speaking to which one they liked more.

Steve: Yep. Which speaks to the sub-conscious control of what we’re eating, and so you see this… there’s some beautiful work being done by the Dutch group in Wageningen, in Kees de Graaf’s team, looking at exactly the same brain areas in brain scanners to see how the taste of umami lights up reward centers in relation to protein status, and it’s only when people are short of protein that those flavors have such a massive reward effect in the brain. So, you’ve got this nutrient-specific association in that case, but you can have other associations which can be non-nutritional that can lead to persevering effects in reward centers.

Interestingly, our early work on things like ants and flies showed that they’re actually pretty good initially if you take, for example, a high concentration of sugar or a dilute concentration of sugar and you give them a choice, they’ll always pick the concentrated one — the tastier one. If you confine them on either a concentrated or a dilute sugar solution initially (and with nothing else), initially they’ll eat more of the concentrated one. But over a few days, they decline in their intake of the concentrated one, and they [the other group] up their intake of the dilute one, if that’s all they’ve got, to compensate. And they [the two groups] end up eating exactly the same amount of sugar or carbohydrates. So, the regulatory system can be tricked initially in a way that makes sense: if you’re selecting among available foods, pick the one that’s tastiest. But over time, the regulatory feedbacks kick in, and you get homeostatic control.

David: What the marketing industry targets is exactly those associations, but often they associate food with things beyond foods, like you know setting suns, and all the things that provide rewards to people in other contexts and they come to associate food choices with those. It’s very complex.

Michael: Yeah; and they’re spending a lot of money to trick us into buying and eat [their foods].

David: And it works. That’s why.

Michael: Yeah; that’s a big problem. This is a little off topic, but something you said reminded me of it. I wanted to ask you… there was something in one of your books where you talked about, I think, how — I don’t know if it was locusts or grasshoppers — would eat novel foods unless they saw another locust eating that food or if they cannibalized the locust who had eaten that food.

David: Yeah. It was testing one specific theory of what leads animals to a balanced diet, and that is novelty when they’re eating something that is obviously not right because of those post-ingestive feedbacks. There are essentially two mechanisms for correcting that. One is to stop eating the thing that you’re eating: obviously there’s a problem with that. And the other is to eat something different. And the novelty hypothesis proposed that if you’re eating something bad, then [you should] randomly just eat whatever’s different in the environment. Which has some sense, because it provides an opportunity for animals to learn what are the appropriate foods in that context.

Then that’s the other thing that we’ve worked on more: is not eat something randomly because it’s novel, because it’s new, but specifically target the thing that is complimentary that is required to correct the imbalance that is being set up by the thing that you’re eating. And that’s what I was looking at with Liz Bernays in Arizona: the relationship between novelty as a mechanism novelty, per se, as a mechanism in diet balancing versus targeted selection of complimentary foods that are going to rebalance. We were doing this in an ecological context, trying to understand how this specific species did it in the Arizona desert.

Steve: And it comes back to the point we were raising earlier about there being a relatively small number of specific nutrient appetites; and one of the ways that you make sure you get enough of everything else or in the right ratios with everything else is to rely on correlations in natural foods and food environments. But if they fail, then there are these sort of general mechanisms that can serve many nutrients without requiring a special pathway to the control of that nutrient in the brain. And there are things like neophobia and neophilia, so if you’re feeling generally crummy because you’ve got a shortage of selenium — rather than having a specific appetite for selenium which has access to particular brain circuitry and hormones and so forth — you can just have a response which is, “Oh, I’ll go somewhere else and eat something novel,” because that’s likely to solve the problem without requiring another dimension in the regulatory system which just complicates the whole engineering challenge.

Michael: I wanted to ask you guys a little bit about obesity as well; and one of the things, as we talked about, at different times in your life you’re going to have different protein targets. And I think that you mentioned that a mother’s diet can have an impact on the protein target of her offspring, which can then, therefore, lead to obesity. And there are other things that can also kind-of mess with our protein targets. So, I was hoping you guys would talk about that a little bit. And, also, are there things that we can do to change our own protein targets, if that’s a factor that we know about.

Steve: Okay; so, yes… the idea that: if you enter the world with a higher protein target, or you acquire it very early in life, a higher protein target — in other words, you need to eat more protein before your protein appetite it is satiated, then the argument went that that, we predicted, would put you at greater risk of obesity, particularly in a western food environment. Because you got to eat more of everything else to get enough protein — it’s just, the protein leverage effect will be worse for you. And there’s some observational and even experimental data that sort-of supports that idea. For example, infants that have been fed on a higher-protein infant formula, rather than a lower-protein, one have a higher risk of developing overweight and obesity in childhood and adolescence. And that, there’s a whole set of sort-of correlations that suggested that there might be something about that idea. So, we’ve explored it recently in two ways: one is in pre-clinical studies in mice, where we have changed the protein content of the mother’s diet and shown, indeed, that the pups when they wean choose a higher protein intake and are more at risk of overweight and metabolic disregulation as a result of that. So, there’s some experimental evidence that that idea has some substance. And the other thing with been doing recently is working with colleagues in Copenhagen on the Danish Birth Cohort — which is 60,000 pregnancies, and looking at the association between the mother’s diet and the percent of protein in the diet. …Her risk of weight gain during pregnancy — so gestational weight gain and weight retention after birth — and the risk of her children developing or her child developing overweight and obesity over the first 5 years of life. And we saw exactly the signature you would predict. So, a higher protein diet during pregnancy indeed reduced the mother’s risk of weight gain, and that is [the result], as you might predict from the typical protein-leverage theory (if you concentrate protein, you eat less to gain your protein target, and hence you would be less at risk of weight gain during pregnancy). However, higher protein intakes in mothers were associated with higher risk of overweight in their children. So, it looks as if there’s, again, a tradeoff between what the mother eats and how that impacts both her own health and weight and what impact it has on her children. And that’s an idea we’re in the process at the moment of planning a major three-site international study to test experimentally.

Michael: a longitudinal study?

Steve: Yes. It’s an experiment – well, it’s going to be an experimental manipulation with the mother’s diet: a 2×2 factorial experiment; and then we’ll prospectively follow the children. So, that’s very early days. We’re in the process of starting to plan that experiment with colleagues in in Boston and in Copenhagen and here at Sydney at the Charles Perkins Center.

David: So, I just wanted to say… the infant formula data are actually particularly revealing. Because the correlation between being fed infant formula and susceptibility to obesity later in life applies specifically to high-protein infant formulas, so there’s a relationship between the protein content and the risk of obesity, which is exactly what we would predict from the modeling and evidence that Steve has spoken about.

Michael: You guys also mention that there is a “good obesity” — a more healthful obesity versus a bad one do that have to do with the fatty acid complement or some other factor?

Steve: No; that was from a very large mouse study where we confined mice from weaning throughout their life on one of 25 diets varying in the ratio of protein, fat, and carbohydrate and the total energy density of the diet as well. And we found that we could dial up by changing those ratios you could dial up healthy or unhealthy obesity essentially. And so, one of one of the peculiar things about the data from that experiment — and this has been replicated now many times around the world by ourselves as well — is that if you have a lower-protein, high-carbohydrate, low-fat ratio as a mouse across the life course, that will support long lifespan [with] good markers of mid- to early-late-life metabolic health, but a degree of overweight, high body fat as a result of mice eating more of a low protein diet. So, you’ve got, if you like, an obese mouse, but a long-lived healthy one. But if you substitute fat for carbohydrate in an equivalently low-protein diet, then you end up with a much-shorter-lived, unhealthier mouse [that is] equally fat. And so, the difference between them was: you could have the same level of body fatness, but you could have very different health outcomes according to the ratio of fat to carbohydrate, diluting the protein. So, it allowed us to — you could quite literally dial up different metabolic outcomes and dissociate some outcomes from others, suggesting here that obesity, per se, is not necessarily coupled inevitably to poor metabolic health. Typically it is, but you can, under certain circumstances, dissociate them. And that begs a whole series of really interesting mechanistic questions that we’re delving into, including what about the type of carbohydrate? And we ran some really equivalently large mouse studies with where we looked at different types of carbohydrates; we looked at glucose, fructose, and varying ratios relative to native wheat starch; relative to digestion-resistant starch, relative to sucrose; and titrated those against protein at fixed fat in the diet and then after that ran a series where we looked at carbohydrate and fat types and relationships. And it turns out that, overarchingly, there’s a signature that you see for protein, but the health consequences are very much determined by the type of carbohydrate. And we found that the killer mixture was a carbohydrate intake which had a high content of a one:one ratio of glucose to fructose, which is effectively —

Michael: Sucrose.

Steve: Well, it was worse than sucrose as a disaccharide. So, if you had the two monosaccharides in the same one:one ratio, it was even worse than sucrose, and that’s high fructose corn syrup.

Michael: So, invert sugar.

Steve: Exactly… which is a principal sweetener, particularly in the US diet — less so here — but it’s catastrophe as a carbohydrate source.

Michael: That’s interesting. I mean, I think that’s like 45:55 [percent ratio of fructose to glucose] or something [like that]. Honey also is very high in invert sugar, so I assume you find the same thing with honey.

David: That’s an interesting point, I was going to say, Michael. It’s another example of industrially breaking the correlations, because that particular mixture is exceedingly rare in nature, except in the form of honey, as you’ve pointed out now. You don’t get it commonly in fruits and nectars and other things… high sugar foods in the wild.

Michael: So why isn’t, I mean, because our bodies are going to break down sucrose into one-to-one fructose and glucose, so why is it different if you’re ingesting them pre-hydrolyzed.

Steve: It’s a good question, and it may speak to exactly that: the… even having to break a disaccharide down into its monomers may confer a little bit of a benefit over not having to do that. There could be all manner of things, including simply slowing things down a little bit, but there’s a whole range of biochemical explanations that might be involved there, but we don’t know as the answer. It was worse one:one than either 100% fructose or 100% glucose. So, we did a series of — we titrated it one against the other to the same fixed amount. And the ratio mattered. So it’s not as if fructose is all bad or glucose is all bad, it’s the one:one or near one:one ratio that really seem to tick all the boxes for the worst possible carbohydrate source.

David: It’s a great research question. I’d love to know the answer to it.

Michael: Yeah, I mean I know that you know retrograde starches for example are going to be better for you than, you know gelatinized starches, right? And, but I mean, that kind of makes sense because they — it takes so long for them to get… and they’re also going to feed the microbiome.

Steve: …microbiome, yeah, and all sorts of things yeah… slows rates of digestion, bulks the calories in your diet, feeds the microbiome, there’s all sorts of obvious reasons, but the sucrose, fructose plus glucose comparison is a really intriguing one. Neither of them really good.

Michael: Yeah, it’s super interesting. The other thing that’s kind-of interesting about it just in terms of, you know, foods and diet, is because invert sugar is sweeter than sucrose, you would think that there would be less of it needed for the sweetness level that manufacturers are trying to get, so you’re eating — you’re consuming less sugar by having this 45:55 ratio of high-fructose corn syrup. So, I would expect that it would be — I mean aside from the glyphosates that might be in it — better for you. So, it’s really interesting to hear that in your studies that it’s not.

Steve: Yeah, it seems that whenever any of us from outside the US travel to the US, everything always tastes way sweeter. So, I wonder whether it hasn’t, so much, reduced the concentration of total sugar — it’s just habituated the entire population to sweeter food.

Michael: [sighs] Yeah, that could be. So, we talked a little earlier about toxicity of converting proteins, and I wanted to also ask you about this idea of toxicity in foods in general, and – it not necessarily being a bad thing — I think you mentioned the [saying] “The dose makes the poison,” right? And, David, if I remember, you studied cyanogenic glycosides early on…

David: Yeah.

Michael: …and the insects that that ate them. so I’m wondering if you guys can talk a little bit about this whole concept and how that factors into diet. And, we all know plants are always trying to kill us, right? — in every possible way — but it’s not all bad, and you need, sometimes, a little bit of poison to be healthier, right?

David: Well, there two sides to that: two contexts in which we discussed it. One is maybe what you’re referring to: the concept of hormesis which is…

Michael: that’s exactly it, yeah.

David: Like, physical exercise — it actually damages your body, but it’s constructive damage because triggers pathways that are, in the longer term, beneficial. And there’s this whole area of science in toxicology known as hormesis which, exactly as you say, exposure to low levels of toxins can often provide benefits. Now, that’s different to the butterflies that I studied. Those butterflies have become adapted specifically to not being vulnerable to the effects of cyanide in the way that other species are, [and] that opens up a whole ecological opportunity of feeding on things that other species can’t feed on. So, there’s no competition. And what it also does is it provides them with a potted source of — readily available source of toxins that they sequester and store in their body and use as a defense against other animals that are vulnerable and haven’t evolved the immunity to cyanogenic glycosides.

But the other context in which we discussed this concept is to say, well, actually nutrients and toxins aren’t that different. Because the background that we came from was the assumption that energy was the limiting thing, and the more of energy and nutrients you can get in the environment, the better. But we pointed out that too little is bad (as well as too much is bad). We’ve discussed that a lot in relation to eating high levels of — excess levels of any nutrient amino acids or sugars. We’ve discussed, pointing out that a nutrient is something that’s beneficial over a certain range of intakes, but becomes toxic beyond that — in the same way as what’s classicly known as toxins through hormesis can be beneficial at very low levels but toxic at higher levels. So, the categorization isn’t as clearcut and distinct as we or people tend to think that it is between and nutrient and toxins.

Michael: Yeah; it’s always amazing to me when people do studies and they inject mice with whatever that’s like 600 times the amount that you would eat, and they’re surprised that it causes disease or cancer or whatever. And I was like, “Well, you know any… I mean, salt will do that, right? Yeah.

Michael: Well, it’s been super interesting discussion. If you guys can maybe just talk a little bit about all of these things: what you’ve learned and how that affects how you perceive how we should eat. I know, for example, you know I’m sure you’re going to say eat more fiber and eat less junk food — is kind of the upshot of it — and let your body figure out the rest. But also, anything that you’re working on now — new studies and things that that are really interesting that you’re excited about.

Steve: Yeah, well it pretty much comes down to: “shop with your brain and eat with your appetites,” as you said earlier. That… we are still possessed of these extraordinarily sophisticated pieces of biology: our specific nutrient appetites. They’re still there; they’re still working…. We need to engage them with a whole-food environment. It’s kind of simple. Everything else will follow if we get that right. How do we do that? How do we do that individually? We’ve got some agency over our own food environment, but that’s highly constrained according to how much we can afford and where we live and all sorts of other things. The more challenging question, and David will talk about this, is how we shift the entire food system.

What we’re doing next — I think we’ve given a bit of a flavor of that. We’re going in multiple directions. Here at the Charles Perkins Centre we’ve got this extraordinary place where we’ve brought together more than a thousand researchers who span the humanities all the way through the physical and life and environmental and health and medical sciences. And it gives us the possibility to go deep into the molecular biology and the physiology of appetite control to really try and understand the nuts and bolts of how these systems work. And we have a program doing that. It gives us the opportunity to engage at the level of experimental and clinical trials with human subjects, and we’re engaged very heavily in doing that. And we talked a little bit about one of the upcoming studies of the pregnancy study that that will explore some of the ideas that are generated straight out of protein leverage and nutritional geometry. There are whole areas, too, where we’re starting to take nutritional geometry as a way to help tame the multiple dimensions of data that are involved with what’s currently called “precision medicine” or “multi-omics,” where you take literally tens of thousands of measurements. How do you unite those, integrate those, and relate them one to the other within the frame of nutritional geometry. How can you link metabolic changes all the way through to behavioral and social changes within the same model? And that’s something that we’re really working really intensively at the moment with our colleagues who are mathematicians and computer modelers. So it’s going in very many different direction. Most significantly is at the level of how you shift food systems and that’s very much David’s area.

David: So, that that’s a nice summary of one side of it and Steve’s emphasized the systems approach. What’s different here is taking as systems and metabolic systems approach in the Charles Perkins Centre. But at the same time, we’ve spoken about another system which is our ecosystem: our environment; our food environment. And that’s the other thing we’re doing, is we’re taking the research in the direction of applying an ecological approach which is a systems approach to understanding the dynamics in that food environment and how it is that, the way that human societies — particularly industrial societies — have evolved, is towards this state where the food environment is distorted for needs… for purposes other than human health and environmental sustainability. And to understand that and to identify points where you can intervene to change it, is critically important. You need to apply a sort of ecological systems approach. So, we are now developing another center equivalent to the Charles Perkins Centre that is centered on the ecological systems in the same way as the Charles Perkins Centre is centered on the metabolic systems. But they interact richly in the same way as our biology interacts richly with the food environments in which we forage. And that’s the field of nutritional ecology that we work within — is looking at that interface between ecological systems and metabolic systems. The other direction that I’m looking at is trying to understand the ways that other species respond in interacting with our altered food system. So, I’ve studied a lot of primate species in the wild in undisturbed contexts, and I’m now working in China with a group in China led by Zhenwei Cui to understand how a species of primate that’s particularly important in biomedicine, which is the rhesus macaque, how that responds when it comes into contact with human foods, particularly via contacts with tourists, which is an increasingly important ecological pressure in China and elsewhere in the world. And it’s remarkable how similar their responses are when they encounter our food environment voluntarily. These aren’t animals that are locked in cages; these are animals that can come and go from the wild. But as soon as they get the flavor of the kinds of foods that are designed in our food environments to distort that relationship with natural environments, that goes in the same direction.

Michael: It’s fascinating. Well, you gentlemen have been very generous with your time. I really appreciate this. This is great to get to talk to you. Your work is super interesting, and, I think, extremely relevant for all kinds of things in understanding nutrition and how we eat. And how we can live better lives and help other people live better lives. So, I just want to thank you guys again. Safe travels, Stephen.

Steve: And thank you, Michael

Michael: Yeah; I look forward to seeing what you guys come up with in future.

Stephen: Fantastic. thank you!

David: Thanks!



Stephen J. Simpson & David Raubenheimer

University of Sydney

Dr. David Raubenheimer is the Leonard P. Ullman Professor of Nutritional Ecology in the School of Life and Environmental Sciences, and Nutrition Theme Leader in the Charles Perkins Centre at the University of Sydney. He previously spent 17 years at Oxford, initially as a doctoral student then as a Research Fellow and Departmental Lecturer in Zoology and Fellow of Magdalen College. He heads the Sydney Food and Nutrition Network and is a member of the Australian National Committee for Nutrition.

Stephen J. Simpson is Academic Director of the Charles Perkins Centre and Professor in the School of Life and Environmental Sciences at the University of Sydney. He spent 22 years at Oxford, beginning as a postdoctoral fellow in the Department of Experimental Psychology then moving to the Department of Zoology, where he became Professor of the Hope Entomological Collections and Fellow of Jesus College. He is a Fellow of the Royal Society of London, a Companion in the Order of Australia and has been prominent in the media and television both in the UK and Australia, having appeared on National Geographic, Animal Planet and History Channels in the UK.