Australasian Science: Australia's authority on science since 1938

Spoilt by Choice



By David Raubenheimer & Stephen J. Simpson

Our supermarkets provide a wide variety of foods, so why do more than a billion people worldwide eat more poorly than hunter-gatherers? A study conducted in a Swiss chalet was the starting point to test a theory.

The variety of foods now available is unprecedented in history, yet more than a billion people worldwide are overweight or obese. Considering the immense disease burden that this imposes on society, it seems that humans are not very good at making the most of the privileged times in which many of us live.

The same cannot be said for other animals, even humble invertebrates. For example, when fruit flies and hunting spiders are given combinations of foods that vary in their nutritional balance, they select a diet from these that supports maximal reproduction (which, in evolutionary terms, is the best possible outcome).

Why do people overconsume energy rather than selecting a healthy diet from the great variety of foods available? Much of the research into this question has focused on the two nutrients that are commonly associated with obesity: fats and carbohydrates (notably sugar). Our research has suggested that to understand the curious nutritional behaviour of our species we need to focus on the relationship between these and the third energy-providing macronutrient: protein.

Our first experiment, performed with the help of Oxford undergraduate student Rachel Batley, provided a fascinating clue. We offered 10 of Rachel’s colleagues free board and lodging in a Swiss chalet on the condition that they selected their meals from a prescribed menu and allowed us to record how much of each food they ate. For the first 2 days of the experiment all of the volunteers were given a menu comprised of foods that spanned a wide range of compositions in terms of the proportion of energy contributed by protein, fat and carbohydrate. This enabled them to compose a diet with a macronutrient balance of their choice by eating particular proportions of foods with high and low protein content.

In the second 2-day period – which is the period that interested us most – five of the volunteers were offered a menu containing mainly high-protein foods (HP) such as fish, ham and low-fat yoghurt. The other five were given only lower-protein (LP) foods rich in carbohydrates and fat, including bread, honey, margarine, cashew nuts and waffles.

Without them knowing it, the responses of the volunteers to these nutritionally imbalanced menus enabled us to test whether humans have a stronger appetite for protein or non-protein energy (carbs and fat). If protein appetite is more dominant than the appetite for non-protein energy, we would expect all volunteers to eat until they had gained the same amount of protein as they did in the free-choice phase of the experiment regardless of whether they had the HP or LP menu.

But to do so the LP group would need to eat more carbs and fat than they would usually choose (i.e. in the first phase of the experiment), and the HP group would eat less carbs and fat – and less total energy – than preferred.

Alternatively, if the subjects’ appetite for carbs and fat was dominant they would over-eat protein (on the HP menu) or under-eat protein (on the LP menu) to ensure that dietary imbalance does not interfere with their intake of carbohydrates and fat.

The winner was... protein!

In all three conditions – free choice, HP and LP – the volunteers ate similar amounts of protein but very different amounts of energy from carbs and fat. This suggested that a cause of the human tendency to over-eat energy might be the drive to satisfy a strong protein appetite while eating foods that contain a low proportion of protein. It might also explain why high protein content is consistently associated with effective weight loss diets.

We called this idea the protein leverage hypothesis because it proposes that the protein content of foods leverages the amount of energy that we eat. Since research on the causes of obesity has focused on carbs and fat, the protein leverage hypothesis provides a new perspective that might help us to understand and manage this major global health issue.

But it is a giant leap to generalise the results of a small experiment involving ten volunteers to a global scale. Together with Dr Alison Gosby (University of Sydney), Dr Susan Jebb (University of Cambridge) and Professors Peter Gluckman (University of Auckland) and Terrence Forrester (University of the West Indies) and colleagues, we therefore performed two further studies using a larger number of subjects and human nutrition facilities that were far more sophisticated than a Swiss chalet.

The first study, based in Sydney, involved Australian volunteers, and the second (based in Kingston) involved Jamaicans. An important detail in the Australian and Jamaican studies is that, unlike the Swiss experiment, the diets were disguised in their macronutrient content. This was achieved by carefully designing recipes that could be altered using protein or carbohydrate additives to contain 10%, 15% or 25% of their energy in the form of protein (Fig. 1).

The crucial achievement is that the three versions of each food were prepared such that a panel of tasting judges rated them all to be equally palatable – they looked similar and were equally tasty. This is important because it enabled us to determine whether the results of the Swiss study really were due to protein leverage (a strong drive to eat the right amount of protein) or the less interesting possibility that the students ate more of the LP foods simply because they were tastier than the HP foods.

Both the Sydney and Kingston studies confirmed the protein leverage hypothesis. Protein intake was regulated more strongly than carbs and fat, and consequently the experimental subjects overconsumed energy when eating LP diets.

Given the larger number of subjects, the fact that three populations have now been tested and the rigorous methods applied in these studies, we are confident that the balance of protein in relation to non-protein energy might well play a pivotal role in the human obesity epidemic.

An interesting question that remains to be addressed, however, is why in a world of virtually limitless choice should humans select to eat low protein diets that cause the overconsumption of energy, leading to obesity and associated diseases? We undertook an analysis of supermarket foods with Professor Rob Brooks of the University of NSW, and found that economics might play a role. Gram-for-gram, foods with high protein content are more expensive than low-protein foods, suggesting that our biological sensibilities might be subverted by economic concerns.

This might help to explain why obesity is more prevalent among the socio-economically disadvantaged in our modern western societies. It also demonstrates the economic advantage for the processed food industry to substitute expensive protein with cheap carbs and fat, the extreme being products with no protein at all, such as carbonated sugary beverages.

A second likely reason why humans have a tendency to select low protein diets stems from our evolutionary past. Several sources of evidence suggest that in our pre-agricultural hunter-gatherer past, in which human nutrition-related behaviour and physiology were shaped by evolution, carbohydrates and fats were scarce compared with protein (lean meat). For example, studies have shown that living hunter-gatherers will expend considerable effort and risk to obtain limited sources of concentrated carbohydrates in their foraging environments, such as honey, and will frequently trade protein (meat) with agricultural societies for carbohydrates (e.g. vegetables).

In a world where carbs and fats are seldom available in large amounts, it makes biological sense that we should have evolved to enjoy eating these when the opportunity arises, and there would be little reason why we should have evolved mechanisms that prevent us from over-eating these nutrients. This is especially true considering that surplus dietary energy can be stored as fat for the proverbial “rainy day”.

But in the shelter of our supermarkets there are no “rainy days”: fats and carbs are constantly abundant, and given their low cost compared with protein are likely to be the “staple” macronutrients for most people.

Where to next? Important – even essential – as it is to disentangle the causes of obesity in tightly controlled experiments, we now need to seek evidence that these causes actually do play a role in the everyday environments where people make their dietary decisions: their homes, in restaurants, and in food markets. We have therefore teamed up with Professors Winsome Parnell of the University of Otago and Bernhard Breier of Massey University to address this question using dietary survey data from the New Zealand Adult Nutrition Survey.

While this is still a work in progress, a recent analysis of an American diet survey (the National Health and Nutrition Examination Survey) is suggestive. As predicted by the protein leverage hypothesis, the incidence of obesity has risen in lock-step with a decrease in the protein concentration in the US diet.

Our interest in humans extends beyond the patterns and consequences of human dietary choices to the question of why we evolved a dominant protein appetite that interacts with modern environments to produce adverse outcomes. This is an evolutionary question that is best answered by comparing humans to a range of related species that have evolved in different ecological circumstances.

We therefore teamed up with Dr Annika Felton of the Australian National University and colleagues to study the nutritional choices of Peruvian spider monkeys in a remote Bolivian jungle (Fig. 2). Intriguingly, they also show protein leverage but there is no evidence of obesity, probably because, as was true of our hunter-gatherer human ancestors, availability of carbohydrate- and fat-rich foods in the forest habitats of spider monkeys is restricted.

What about our closer living relatives? Together with Dr Jessica Rothman of Hunter College, New York, and Professor Colin Chapman of Université de Montréal we conducted similar studies on wild mountain gorillas in the Bwindi Impenetrable National Park, Uganda. To our great surprise these apes showed the opposite behaviour to humans and our more distant relatives, spider monkeys. Mountain gorillas over-ate protein to gain the target intake of carbs and fat.

Clearly, therefore, the nutritional biology that makes us so vulnerable in modern environments was not inherited intact as a legacy of our primate ancestry, but a diversity of patterns of nutrient regulation has evolved in different circumstances. By extending the study to a wider range of primates we hope to better understand the relationships between foraging environments and nutritional regulation, and shed light on the origins of our inability to make the best use of the privileged times in which we live.

David Raubenheimer is Professor of Nutritional Ecology and Director of the Bachelor of Natural Sciences degree at Massey University. Professor Stephen Simpson is ARC Laureate Fellow in the School of Biological Sciences, and Academic Director of the Charles Perkins Centre, the University of Sydney.