TODAY I searched my kitchen cupboards for dark matter, and found it in a packet of Korean instant noodles. The food label ran to 38 ingredients, many of them additives. But it also listed some real foods, including soy, chilli, sesame, shrimp, cabbage, seaweed, mushroom, anchovy and cuttlefish. And also the one I was looking for, garlic.
To be clear, I’m not suggesting that garlic contains actual dark matter, the 85 per cent or so of material in the universe that physicists say is there but cannot observe directly. But it does contain what has been called “nutritional dark matter”: the thousands and thousands of compounds that are in food but which, until recently, were totally unknown, and which may be affecting our health. Given that eating is one of the big human universals, that’s a mind-boggling oversight.
Advertisement
“Our understanding of how diet affects health is limited to 150 key nutritional components,” says Albert-László Barabási at Harvard Medical School, who . “But these represent only a small fraction of the biochemicals present in our food.” It is time, he says, for nutritionists to go dark-matter hunting, to massively expand our knowledge of what is on our plate and its impact on us.
The idea that food is a rich and complex mix of biochemicals is hardly news. Even the well-known macronutrients – proteins, carbohydrates and fats – are hugely diverse. There’s also a vast supporting cast of micronutrients: minerals, vitamins and other biochemicals, many of which are only present in minuscule quantities, but which can still have profound health effects.
The official source of information on this complex biochemical soup is the for Standard Reference, maintained by the US Department of Agriculture (USDA). It contains information on the composition of hundreds of thousands of foods, broken down into 188 different nutritional components.
A lot to digest
Search for “garlic”, for example, and the database serves up 58,055 foodstuffs that contain it, ranging from whole, raw garlic to processed foods such as instant noodle soup. The lists 67 macro- and micronutrients, some quantified down to micrograms per 100 grams, or concentrations of less than 0.00001 per cent. That may seem detailed, but is far from comprehensive. For example, it omits some of the quintessential flavour compounds in garlic, such as alliin.
This is a general problem across the USDA database, says Barabási. It tracks only common nutritional components in many foods, and so omits many rarer ones. The USDA began to fill this gap in 2003 by adding 38 flavonoids – plant compounds associated with a lower risk of cardiovascular disease – to its existing panel of 150 components. But that is as far as it went.
Around 10 years ago, an international team of researchers decided to compile a more comprehensive database after trying and failing to get detailed information on the composition of various foods. “We could only track down lists of a few dozen compounds, not the hundreds or thousands we expected,” says David Wishart at the University of Alberta, Canada, one of the project’s founders. “The fact that there was so little known about the micronutrients in commonly consumed foods really bothered us.” So they trawled the literature and filled in other blanks through their own chemical analyses in the lab.
The result is a massive database called which Wishart says now holds information on about 70,000 nutritional compounds – nearly 400 times more than the USDA database. For example, the USDA lists 67 compounds in raw garlic, but FooDB has 2306.
By that reckoning, with the USDA as your guide, 99.5 per cent of the components in food are a mystery. Even FooDB is incomplete. Of the 2306 compounds in garlic it lists, for example, only 146 have been quantified, meaning that more than 2000 others are known to be present but at unknown concentrations. This problem is writ large across the whole of FooDB, with a total of about 85 per cent of nutritional components as-yet unquantified.
In other words, when it comes to the make-up of the food we eat, we have only assessed the tip of the iceberg lettuce (which, incidentally, the USDA says contains 11 compounds, while FooDB lists more than 4000). “We’ve severely underestimated the complexity of food,” says Tim Spector at King’s College London, author of the forthcoming book . “Food is incredibly complicated: the chemicals are complicated; when it enters our guts, it interacts with microbes which make it into other chemicals, which also have complicated effects on our body. Because we’ve focused on macronutrients and calories, the whole field has been dumbed down.” Barabási’s research is “opening people’s eyes” to the true complexity of food, he says.
Why complicate things further? Because dietary dark matter may be affecting our health, both for good and ill. “Having more detailed information of what compounds are in food certainly would be helpful to nutrition researchers,” says at the Jean Mayer USDA Human Nutrition Research Center on Aging at Tufts University in Massachusetts.
Dark nutrients
Right now, it is hard to say exactly how helpful, because so much remains uncharacterised. “The health implications are largely unknown,” says Barabási. But there is good reason to believe that some of the neglected compounds have meaningful effects (see “Garlic crush“). “Certainly, some of these dark nutrients are quite important for human health,” says Wishart.
The idea hasn’t gone down well in some established nutrition circles. There is an urgent need to identify confounding factors in nutritional epidemiology, says Mike Gilbey at University College Dublin, Ireland, but gaps in our knowledge of nutrients isn’t one of them. “Nutrients are fully known, and they alone account for almost all of the impact that diet has on non-communicable disease. There is no dark matter.”
Another obvious objection to the claim that understanding the dark matter is important for health is that most of the compounds occur in tiny quantities. But, says Barabási, that raises a misconception. “The concentrations of mapped and unmapped nutrients span about nine orders of magnitude, yet concentration is not always the factor: vitamin E comes at micrograms per 100 gram intake of food, yet its absence has adverse health effects. So it is not true that low concentration chemicals do not have an effect on health.”
To start to understand the scale of our ignorance, Barabási and his colleagues consulted another database – the , an inventory of how thousands of chemicals interact with our bodies.
First, they looked at the 67 “official” components of garlic, and found that 37 were already known to be linked to human health and disease. Then they did the same for the 2306 compounds in garlic listed in FooDB, and discovered a .
These vast tracts of uncharted complexity could be the reason why nutrition science so often produces inconsistent and irreproducible results, says Barabási. This is familiar to anyone who tries to keep up with flip-flopping nutrition advice. One week red wine is bad for you, the next week it isn’t. Ditto red meat, eggs, saturated fat and many more. Without a complete picture of the nutritional composition of food, you can’t be sure you are comparing the right things.
This is also true of individual micronutrients. “Consider beta-carotene,” says Barabási. “It tends to be positively associated with heart disease, according to epidemiological studies, but studies adding beta-carotene to the diet do not show health benefits. One potential reason is that beta-carotene never comes alone in plants; about 400 molecules are always present with it. So epidemiology may be detecting the health implications of some other molecule.”
Another probable cause is the effect of the microbiome on dark nutrients, says Wishart. “Most dark nutrients are chemically transformed by your gut bacteria. That’s probably why studies on the benefits of different foods give relatively ambiguous results. We don’t properly control for the variation in gut microflora, or our innate metabolism, which means different people get different doses of metabolites from their food.”
“It would be foolish to dismiss 99.5 per cent of what we eat as unimportant”
All of which suggests there is a food mountain to climb. But there is a way forward. There are many parallels, says Barabási, between our understanding of nutrition and health and our knowledge of genetic epidemiology before the human genome was sequenced.
In those pre-genome days, only around 1.4 per cent of human DNA – the chunk that codes for proteins – was considered important, and the other 98.6 per cent was considered junk. The idea of sequencing the whole genome was often dismissed as a waste of time and resources, says Barabási. And yet we now know that about two-thirds of sequences linked to disease are in those “junk” regions, many of which are stretches of DNA that control gene expression.
That doesn’t imply that the majority of associations between diet and health will be found in nutritional dark matter. But it would be foolish to dismiss 99.5 per cent of the compounds we eat as unimportant, says Barabási.
There is only one answer, he says: to fully map the dark matter. His team and others are already working on it. That inevitably means having to analyse food the old fashioned way, with slow and laborious lab techniques such as mass spectrometry and nuclear magnetic resonance. To make this a tractable problem, these methods may have to be re-engineered to make them faster, he says.
Diet and disease
But there are other avenues. Barabási and his team, who are physical scientists rather than nutritionists, specialise in using big data and the science of complex networks to probe, among other things, the genetic origin of human disease. This is what led them to nutrition. “Genetic data can only explain about 5 to 20 per cent of disease causation,” he says. “We are interested in finding the rest, and food is the next big opportunity to fill the gap. Also, big data has not yet touched nutrition.”
To test the potential of such approaches, Barabási and his colleagues created a to dredge the scientific literature for scraps of information about the constituents of food. This unearthed some of the missing concentration figures for garlic – for example, it found one for diallyl disulphide, another pungent flavour molecule with possible health benefits. It also discovered a further 96 components of garlic that aren’t even listed in FooDB. Doing the same for cocoa, the researchers found 238 novel compounds. This suggests there is a good deal of information out there already which has yet to be pulled into the databases.
But garlic and cocoa are just two of about 2000 natural ingredients people consume. “We are currently expanding to multiple foods including milk, apple and basil,” says of the Network Science Institute at Northeastern University in Massachusetts, who co-developed the text-analysis tool, called FoodMine. “We aim to cover the most common staple foods, to have a good grasp of the chemical complexity of our diet.”
As if all that wasn’t hard enough, there is the added complication that cooking transforms chemical components into others, sometimes with health implications. When sugars and amino acids react at high temperatures, the result is new molecules that make roasted and grilled food delicious, but the reactions also produce the compound acrylamide, which is a . Cooking and processing nutritional dark matter could produce similar, but unknown, toxins, says Barabási.
Our metabolisms get to work on nutrients, too, adding yet more complexity. “Food compounds are often chemically transformed by various enzymes in the mouth, stomach and the intestine,” says Wishart. “As a result, many of the dark nutrients are biotransformed into a variety of smaller, stranger metabolites that then recirculate back into the blood and other tissues. These biotransformed dark nutrients are the ones that really have the more significant health effects.”
There are also the myriad chemicals added during food processing. What happens to them when they encounter our metabolisms and microbiomes is unknown.
And if just identifying the compounds in food is a challenge, working out how they interact with human biology is an even bigger one. “Ultimately, we need to connect all the food molecules to their molecular effects in the human cell,” says Barabási. That, too, requires laborious lab work.
Once again, new computational techniques can do the heavy lifting. Over at the USDA, for example, a team led by Parnell is developing a way of using artificial intelligence to predict how dietary dark matter acts once we have eaten it.
Its system, called , cross-references FooDB with another database called . This is an inventory of 1.9 million compounds known to have some sort of biological effect, maintained by the European Bioinformatics Institute near Cambridge, UK.
For a pilot project, the team wanted to see whether the tool could figure out which compounds in FooDB might activate a protein that is involved in fat and glucose metabolism, and therefore could be of interest in treating diabetes.
When the AI was set loose to scour FooDB for compounds that are also likely to interact with the protein, it got 10 hits. Two of these compounds were already known to act on the protein. But the other eight weren’t, including sesamin from sesame seeds, a flavone called irigenin from lima beans and compounds from tea, herbs and spices. It will take time, says Parnell, but it should be possible to eventually assess the biological activity of all 70,000 compounds in FooDB.
Ultimately, says Barabási, the goal is to build a complete picture of the interplay between nutrients and human health. That is a colossal undertaking that requires yet more data on individuals’ genomes, microbiomes and other factors, but it is the only way for us to truly understand the complexities of human nutrition and health. “So much to do,” he says. “But we’re making progress.”
The Korean noodles are still in my kitchen cupboard, and I don’t think I will be eating them any time soon. I never kidded myself that they were a health food. But I don’t want to go over to the nutritional dark side.
Garlic crush
Researchers are hunting down tens of thousands of untracked “dark nutrients” in food (see main story), but why do we need to know about what we eat in such fine detail? To get an idea, let’s talk about steak and garlic. Gut bacteria break down a substance in red meat to produce a compound called trimethylamine (TMA). This is then absorbed and transported to the liver, which converts it into trimethylamine N-oxide (TMAO). High levels of TMAO in the bloodstream are linked to significantly higher death rates from coronary heart disease. But one of garlic’s flavour molecules, allicin, is known to inhibit TMA production. This may be one reason why the garlic-rich Mediterranean diet protects against heart attacks. Allicin has also been reported to have anticancer properties, while another component of garlic, luteolin, which isn’t in the standard database of food compounds produced by the US Department of Agriculture, also seems to protect against cardiovascular disease.
