What we really know about nutrition, part 2
by Alan Cohen
In the last installment of this series on nutrition and science, I argued that epidemiological studies on associations between diet and health are extremely difficult to conduct well. But aren’t there other ways we can get information on how to eat well? Yes. We know the biochemistry of molecules such as vitamin E relatively well, and we know a lot of what they do and why they are essential in the body. We have lots of animal studies on nutrition, both in mice as model organisms, and for domesticated animals such as dogs, chickens, cows, and so forth. These animal studies do what human studies can’t: they perform controlled experiments to identify causal relationships.
In addition to these basic science studies, there is also a lot of circumstantial evidence about nutrition. For example, there are studies that measure the nutritional content of hunter-gatherer diets. There are studies comparing the intake of different nutrients across species. And there are some lessons from basic evolutionary principles that can be applied to nutrition.
What can we learn from all this? The take-home message is that the details are very complicated and poorly understood, but there are some broad brush recommendations we can make. In my opinion, most of the basic science does little more than generate confusion about the details, and the most useful lessons come from the circumstantial evidence.
What we can and can’t learn from basic science on nutrition
Basic science has done an excellent job of characterizing the molecular structure of various components of the diet: vitamins, minerals, lipids, proteins, carbohydrates, and so forth. This knowledge is important, and is applicable well beyond nutritional recommendations: it is the foundation for understanding how our bodies work. It is not, however, a good basis for making dietary recommendations. Just because vitamin E binds to receptors on immune cells doesn’t mean you need to eat more of it. The role of lipids in cell membranes doesn’t make them evil. Even the fact that sugars are implicated in insulin signalling which can become dysregulated and lead to diabetes doesn’t really tell you how much sugar is the right amount to eat.
Animal experiments would seem to be a bit more instructive. Countless studies have been done on how to optimize the nutritional content of chicken feed or dog food, often in terms of promoting specific benefits like healthy chickens, tender breast meat, or a shiny coat for dogs. The problem is, one of the things we’ve learned is that different animals are different. Chickens don’t have the same needs as zebra finches or cockatiels. Pigs don’t have the same needs as sheep or cows. So again, all this animal science tells us relatively little about what exactly we should eat.
All of this basic science does, however, help us understand some of the basic principles of nutrition (if not the specific dietary recommendations). One of these principles is that things are complicated. My colleague Kirk Klasing at UC Davis works on bird immunology and nutrition, and he has done lots of experiments on chickens. One very interesting result he has found is that if you feed chickens more vitamin E or more omega-3 fatty acids or more carotenoids, you boost one part of their immune system but repress another part of their immune system. However, if you give them more of any two of these things at the same time, the effect is exactly the opposite! (You boost the other branch of the immune system.)
It is possible that this result only applies to chickens, and more specifically to that specific breed of chickens under the particular lab conditions at UC Davis. But what almost certainly applies more generally is the principle that diet is VERY important for determining different aspects of how our bodies function, but the precise details of the effects depend on lots of other things, including what else is in the diet, genetic background, and aspects of our environment or lifestyle.
Another proof of this complexity is the number of receptors we have for different molecules in our diet. Evolution has bestowed on us genes that encode for proteins that have as their specific function the detection of molecules called polyphenols, a type of antioxidant found in many plants. This is in addition to all the receptors for the various vitamins, carotenoids, and minerals we take in. The presence of a receptor means that something is being regulated in response to the presence of the molecule, and while the details have yet to be worked out, one can imagine the complexity of trying to integrate all of the different signals from all of the different elements in the diet.
What evolutionary principles can teach us about nutrition
One of the key principles of evolution is that natural selection works on organisms in their imperfect environment, rather than on optimizing them for a hypothetical perfect environment. This is why the scarlet gilia flower grows better when deer eat it than when they don’t, and why our immune systems probably require exposure to pathogens in order to develop properly. In the context of nutrition, this means that our bodies were built by evolution to make use of the sorts and amounts of foods our ancestors encountered. Our bodies were not built for the hyper-abundance of the modern world.
Our ancestors likely experienced a wide range of diets. Some communities by the sea ate very large quantities of fish or oysters, which were abundant resources. Modern-day Tibetans and Inuits eat mostly meat and very few vegetables for large parts of the year, and have specific adaptations for these diets. Others, such as the !Kung bushmen in southern Africa, eat diets very close to that currently recommended by the USDA (except with a lot more calcium). We modern humans are likely a mix of various lineages that ate various diets. And conditions changed: sometimes food was abundant, sometimes it wasn’t. Some populations experienced more frequent famines than others.
Based on these conditions, there would have been substantial evolutionary pressure to evolve some flexibility in our diet. There probably isn’t one perfect way to eat right: there are lots of ways to eat well within the general range of diets our ancestors may have experienced. However, diet evolves quickly (as seen by the adaptations among Tibetans and Inuit), so it may be better for most of us to stay close to average diet of our ancestors. There may be some flexibility, but no system can be perfectly tolerant of infinite changes. The further we stray from the diet our bodies have evolved to make use of, the more likely something will go wrong. This is why I think fad diets such as Atkins are likely to do a great deal of harm in the long run. We confuse the idea of slimness with health – yes, some people lose weight on Atkins, but probably because they have put their bodies into a profoundly unnatural state.
Another general principle is balance. Our brains like to think in simple dichotomies: antioxidants are good, fat is bad. Reality, however, does not follow such simple rules. If too much can be bad, so can too little, and vice versa. Just because many antioxidants are important for our health doesn’t mean we should take supplements with abnormally high levels of these substances. This is probably why many studies have shown that some vitamin supplements increase the risk of mortality. It’s simpler to have a clear rule: eat more of this, eat less of that. But the fact of the matter is we need to eat the right amount of this, and the right amount of that.
A final evolutionary principle is that our brains use simple cues to find the foods we need. Because fat, sugar, and salt were often in short supply, we have built in cravings for these things. Now, we have as much of them as we want and we are taught to think of them as “bad” because we tend to eat too much. But of course we need them, in moderation. So we have to learn to avoid cravings and to have these things in reasonable quantities. (My personal craving is often for gummi bears and jellied candies. I am certain that these things substitute in for my natural fruit craving. Needless to say, I should eat real fruit rather than allow myself to be tricked by the gummi bears! But I indulge once in a while…)
All of this leaves us with a conundrum: if we have to eat just the right amount of all these different things (with a bit of tolerance for error built in), and if epidemiological science doesn’t seem able to given us precise prescriptions for how much to eat of what, what should we do?
The answer, I think, is to stop micromanaging our diets. We need to stop thinking that we can count calories, and vitamin content, and fat content, and somehow know enough to arrive at health via our calculators. Instead, we should avoid processed foods and cook from scratch as much as possible.
I like Michael Pollan’s rules, which I paraphrase here: 1) Not too much meat (I’d add that it’s good to have a fair amount of seafood); 2) Lot of variety of fresh fruits and vegetables; 3) Lots of whole grains; 4) Not much processed food (cook from scratch when possible); and 5) Don’t overthink it – you’ll do yourself more harm than good. I’d also add: take pleasure in eating. You’ll release digestive enzymes and absorb more nutrition.
So, next time you hear a study touting the benefits of coffee or the risks of salt, take a deep breath, ignore it, and go to the kitchen to cook a fresh meal. Eat the meal with gusto, and with your family. Relax and enjoy. Remember, some of the healthiest people on earth – Japanese, Koreans, and mediterraneans – take great pleasure in food and do little to micromanage their diets. It hasn’t hurt them too much.