Flavor in the context of ancestral human diets

• Main Author: Richard Wrangham
• Format: Article
• Language: English
• Published: Frontiers Media S.A. 2014-07-01
• Series: Frontiers in Integrative Neuroscience
• Subjects: Cooking; evolution; flavor; great apes; digestive tract; hunter-gatherers; Ancestral Diets
• Online Access: http://journal.frontiersin.org/Journal/10.3389/conf.fnint.2015.03.00009/full

Given that nothing in biology makes sense except in the light of evolution, to understand the evolutionary biology of human flavor perception we need to know what kinds of foods have been sufficiently important in the human past for natural selection to favor specific mechanisms for perceiving and digesting them. Humans share with great apes a long prehistory of specializing on eating ripe fruits. Wild ripe fruits have much less sugar and more fiber than domestic fruits, but are similar in tending to offer two main tastes, sweet mixed with sour. While a preference for sweetness is easily explained, the attraction of a sweet-sour combination is still uncertain. A plausible explanation is that because mild acidity inhibits microbial growth, it signals a low probability of toxins. Whatever the explanation, the human preference for a combination of sweet and sour tastes appears to be a strong response reflecting our frugivorous ancestry. However for at least 2 million years fruit-eating has been less important for humans than it is for most other primates. Humans specialized dietarily in two respects, composition and processing. First, though composition varies widely, for their body size humans select items of unusually high caloric density. Thus compared to great apes, hunter-gatherers consume less fiber and more starch and lipids. They do so by eating much less foliage and fruit than great apes do, and more roots and animal-derived foods including both meats and honey [1]. Although meat is often regarded as important because it provides protein, great ape diets provide more than enough protein from fruits and foliage alone: fat is a more critical component of meat. Honey from honey-bees Apis mellifera has a surprisingly large role in the human evolutionary diet, i.e. for African hunter-gatherers. It is a strongly preferred item which can be the predominant sources of calories: hunter-gatherers eat as much as 1 kg per day for weeks at a time. Evidence that honey has been a major part of the diet for thousands of generations comes from a co-evolved symbiosis between humans and an African bird, the greater honeyguide Indicator indicator. The relationship depends on honeyguides being born with a tendency to guide humans to honey, and on humans extracting honey from the hives to which the honeyguide leads them. Large intakes of fructose and glucose by humans are thus an ancient phenomenon, suggesting that humans have had the opportunity to at least partially adapt to this aspect of a diabetogenic diet [2]. The second dimension of difference between humans and other primates is food processing. Humans are the only species that processes their food with heat, i.e. by cooking. Cooked meat or tubers provide more energy than the same items eaten raw [3]. Studies of human raw-foodists (individuals who choose to eat their diets raw) indicate that most adherents suffer chronic energy shortage and in the case of women, amenorrea [4]. Yet raw-foodist diets are high-quality compared to what a hunter-gatherer could eat raw. Raw-foodist eat agriculturally derived foods, so plant items are low in fiber and rich in sugar compared to wild foods, and meats (since not all raw-foodists are vegan) are fat-rich. Raw-foodists live in industrial societies with access to global markets and no serious seasonal shortages of food. They often process their food non-thermally and tend to take less exercise than hunter-gatherers. The fact that despite all these advantages, raw-foodists experience severe energy shortages and impaired reproduction indicates that hunter-gatherers would not be able to survive on raw diets. Humans have a different kind of digestion from other animals: we are not biologically adapted to eating our food raw. Little is still known about how the human digestive system is functionally constrained to including a substantial component of cooked food in the diet but compared to all other primates humans have smallest relative sizes of guts and molars. Multiple specializations in human physiology can likewise be expected. By identifying such differences between human and non-human digestive capacity we can expect to make much better use of animal models. Given that humans are evolutionarily specialized on diets composed of items of high caloric density that have been cooked, how should we expect sensory perception of food to have evolved? With regard to composition, our sensory mechanisms with regard to food perception should be largely similar to other omnivores, such as rodents and primates that eat similarly high-quality items whenever possible, even though they obtain them less often than humans do. For example mechanisms including taste, flavor and physical perception all presumably help animals identify attractive features of cooked foods, such as reduced bitterness, fatty odors and softness. However, some of the cooked-food flavors that we find attractive are evolutionarily novel [5]. Two processes are particularly intriguing. Caramelization produces sweet and bitter tastes and a range of nutty flavors generated by several thousand pyrolized sugars. It requires temperatures of at least 105oC, and was therefore an unknown phenomenon prior to cooking. Caramelized sugars seem unlikely to have been significant in human diets until pots allowed relatively elaborate cooking only ~ 20,000 years ago. Adaptive responses to caramelization therefore seem unlikely. The Maillard process is another source of multiple flavors derived from cooking. It again produces thousands of compounds, in this case from sugars and amino acids, some of which can be carcinogenic and many of which are strongly preferred flavors. Unlike caramelized sugars, Maillard compounds are produced abundantly on the surface of cooked meats even when roasted in very simple ways on a wood fire. Humans have therefore had rich opportunity to adapt to them. The Maillard process offers many opportunities to test the hypothesis that the value of all flavors is learned from experience. Identification of the evolved systems of human food perception is a critical step in understanding the flexibility and malleability of our motivations to eat. It will also yield clear guideless about how best to use animal models that did not adapt to eat cooked foods.