Nutrigenetics—that is, matching your diet to your DNA—is touted as the future of wellness and disease prevention. But do we all really need special, customized eating plans?

Patrick Clinton
September 11, 2019

Way, way back when the human race was young—sometime after the age of dinosaurs, but before the invention of salad forks and pants with pockets—our ancestors never devoted one second to wondering what they ought to eat. They ate what they could. And that depended on where they lived. If there were fruits, leaves and insects, that's what was on the menu. Ditto if there were caribou and walruses in the neighborhood. There was no other choice.

Over time, human beings in different nutritopes (the fancy science term for areas defined by what's available to eat) adapted to get the most out of their diets. A walrus eater would be born with a genetic mutation that kept her heart healthy despite the saturated fat in the blubber she ate. And eventually, through natural selection, almost all the babies in her village would have the same gene variation, called a polymorphism—a mutation that has become well-established in a population. Through a similar process, herders in Northern Europe mostly ended up with a genetic mutation that let them continue to digest lactose as adults, while the rest of the world mostly didn't. And Polynesians who went on long sea voyages developed the ability to function with less vitamin C than the rest of us—helping them fend off scurvy.

But then things changed. Across the planet, people pulled on their pants, put their salad forks in their pockets and left the old neighborhood. They crossed continents, navigated oceans and intermarried with everybody they met along the way. They also killed off game, cut down forests and invented modern agriculture—as well as sliced bread, refrigerated trucks, supermarkets and Big Macs. By the time they got done, the old nutritopes were largely gone or changed beyond recognition, and most of us were a genetic muddle, not necessarily optimized for any traditional diet.

However, those genes we inherited from our ancestors are still in there. So it makes sense that if you could figure out which ones you carry, you could align your diet with the sweet spots they create and get healthier, happier and thinner.

Nutrigenetics and Nutrigenomics

Science has been chipping away at that idea for a long time, arguably since the early days of the 20th century, when Archibald Garrod discovered that alkaptonuria, a disease that prevents the body from processing certain proteins, was inherited. The field he created—"inborn errors of metabolism"—eventually led to the birth of nutrigenetics and nutri­genomics, the twin sciences of how your genes interact with the food you eat and, in turn, affect your health. (What's the difference? Nutrigenetics is more about health impacts; nutrigenomics has to do with the molecular nuts and bolts of how those impacts happen.)

And nutrigenetics has been making pretty remarkable progress recently—­especially considering that each person has between 20,000 and 25,000 genes, consisting of 3 billion base pairs of DNA. For those who need a refresher: Your DNA is a big instruction manual for building and operating a human body written in four chemical bases or "letters"—the molecules known as G, T, C and A. Changing a single letter can be enough to alter how your body works. A change that's well established among humans—like the walrus-eating thing I mentioned—is most commonly a single-nucleotide polymorphism, or SNP. Although there are other types of polymorphisms, SNPs are the primary thing that nutrigeneticists look at, and the average human has 4 to 5 million of them.

Science has made hundreds of links between genes, diet and physical outcomes. We've learned that a variation in a gene called FTO predisposes people to obesity. We've learned that those with a particular alteration in the gene APOA2 lose more weight when they reduce saturated fat in their diet compared to people without the variation. And we know that more than 10% of American women have a variation in the gene MTHFR that can lead to birth defects like spina bifida in their babies, unless the pregnant women supplement their intake of folate.

When you think of how little we knew about this sort of thing even a decade ago, the growth of nutrigenetics has been absolutely explosive. On the other hand, when you think of how much there is still to know, it's another story. There are more than 100 million known SNPs and an estimated 25,000 bioactive substances in food—compounds that aren't essential nutrients, but that can affect the way your body operates, including things like lycopene, resveratrol, tannins and basically every unfamiliar name you see in an article about The Next Miracle Nutrient. It's going to take some time to get through them all and understand how they play together.

Are We Ready for a Personalized Diet Based on Our Genes?

But wait. Can't you already get personalized diet recommendations based on your genes from a zillion different spit-in-a-tube DNA testing companies? Well, it's true that a bunch of them—including Habit, Orig3n and 23andMe—­offer a panel of tests related to health and nutrition (see "What I Learned from My Nutrigenetics Test," below). Typically, they look at SNPs in 20 or 30 genes and tell you things like if you're lactose-intolerant, whether you're prone to a half-dozen vitamin deficiencies, how you metabolize caffeine, alcohol and different kinds of fats, and if you carry a gene that makes cilantro taste weird to you. It's interesting, I guess, but the advice—­especially after the Food & Drug Administration cracked down on the consumer-DNA industry a few years back—tends to be extremely cautious and not really all that personalized. So, yes, it's nutrigenetic advice. But only in the same sense that the Speak & Spell you had when you were a kid is a laptop computer.

Real personalized nutrition still faces some big hurdles. First, there are a lot of SNPs to test to see what they do. Then there's the enormous problem of figuring out which of the discoveries are real and which are experimental error or a statistical glitch. Martin Kohlmeier, M.D., Ph.D., a nutrition research professor at the University of North Carolina, notes in his book Nutrigenetics: Applying the Science of Personal Nutrition that most of the new results that get published won't ultimately pan out.

And a lot of the things we "know" now will unfortunately fall into the same ­category. A few years back, for example, researchers analyzed more than 600 genetic association studies. Only about one study in six had its results confirmed in a follow-­­up paper and just 1% of the 600 were replicated twice or more (a sign that, scientifically, you're really onto something). Let's be optimistic and say that study results will be confirmed half the time. That leaves 500 coin flips. "The industry makes it seem like personalized nutrition is ready for prime time and it's not," says ­EatingWell advisor David L. Katz, M.D., M.P.H, director of Yale University's Yale-­Griffin Prevention Research Center and author of The Truth About Food.

But even when the knowledge base about genes and diet gets more solid—and leaders in the field believe that might only be a few years off—there will be the question of what to do with it. A personalized diet sounds like a wonderful thing, but as Kohlmeier points out, if you go beyond one or two genes, things can get complex. Imagine, he says, that you're trying to adjust your diet to meet nutrition needs based on a handful of gene variants: say, total calories need to come down by 8%, sodium by 20% and saturated fat by 50%. You need to up your total folate and vitamin C by 50%, reduce your folic acid by two-thirds and get your calcium/magnesium ratio under 2.6.

So, what's for dinner?

Can the Recommendations Translate to Real Life?

Most of us, working with real-life ingredients—tomatoes and greens and pork chops—would have a hard time translating that information to our plates. And it's not clear that we even have to deal with ­every message we get about our genes. Many SNPs, taken individually, aren't that impactful. The outcomes that most of us care about—obesity, diabetes, heart disease, cancer—involve dozens of genes interacting in ways we don't yet understand. Genes can also be turned on and off. They do different things at different times of life and in different parts of the body. So how much does a single gene contribute to a person's developing one of these conditions? "Very little," says Cecile Janssens, Ph.D., a professor of epidemiology at Emory University whose research focuses on how genomic studies translate into clinical and public health practice. "It's likely less than 1%."

What's more, experts estimate that, overall, genes only explain about 10% of the risk linked to diet-related diseases like obesity and type 2 diabetes. "It's Homo sapiens arrogance to think we can't know what to eat until we're all individually profiled—­because we're all so unique and special," says Katz. "There are fundamental truths to feeding humans, just like dolphins should eat fish and pandas should eat bamboo. Ninety percent of health is a diet that applies to everyone. Someday we'll be able to use personalized nutrition to ice the cake. But the cake itself is just a generally healthy diet." In other words, your body doesn't demand a perfect diet, just a good one.

What we will be able to do with all the spectacular nutrigenetic data we're going to acquire is use it to make better diagnoses of diseases and recommendations for dietary changes that do really matter. Kohlmeier gives a great hypothetical example: A 51-year-old man has an enlarged liver. He is overweight, has stopped drinking alcohol, and his cholesterol is under control, thanks in part to a diet low in eggs, meat and saturated fat and high in fruits, vegetables and whole grains—still, his condition is getting worse. The textbook medical approach would call for weight loss and exercise. But a DNA test his physician gave him shows that the patient has a gene variant that means his body doesn't synthesize choline well—and he's eliminated perhaps half his intake of this nutrient by eating less meat and eggs. Not getting enough choline can cause fatty liver disease. The dietary Rx: get more choline from food or a supplement.

Connecting the Dots

You'll notice that this isn't the kind of nutrition wisdom you'd expect to get from a website or app. And that's probably going to be true of much of the nutrigenetic advice we receive in the near term. It's going to sensitize physicians to the range of variations they should expect to see in their patients, and it's going to provide a way to connect the dots between symptoms, genes, diet and factors like physical activity and environment. Ultimately, this knowledge could provide the basis for deeply informed health coaching.

Actually, there's already a prototype of what that kind of coaching could look like. A few years back, a team led by ­Leroy Hood, M.D., Ph.D., a super-important figure in the history of genome sequencing, started Arivale, a company aimed at developing an extremely detailed picture of what is going on in the body and using it to deliver health strategies to customers via skilled coaches. They looked at DNA—not just a few genes, but the whole shebang—then did repeated tests of pretty much anything you've ever heard of and more, like blood pressure, cholesterol, height, weight, waist circumference, cortisol levels, microbiome, proteins that affect heart and brain function, and metabolites (the end products of chemical reactions in your body—sugars, lipids, amino acids, fatty acids and the like). Then they threw in a whole lot of computing power to make sense of it all and provided clients with personalized diet and lifestyle plans to improve health outcomes.

The results, by all accounts, were impressive. People who stuck with the program saw their cholesterol, triglycerides, waist diameter and hemoglobin A1C all move in the right direction—enough to have a substantial impact on health. And because Arivale continually analyzed its own data, looking for new connections between genetics and health, it was likely to get even better.

I say "was" because Arivale went out of business this past spring. The problem, according to co-founder Nathan Price, Ph.D., associate director of the Institute for Systems Biology, which developed the concept: "The economics just didn't work out. Basically the program's too expensive for what people can pay." That was still true even after Arivale cut its subscription price from about $3,500 a year to $99 a month.

Not to worry. Arivale or something similar will be back. They've figured out the hard part—how to better people's health based on DNA. Now they just have to figure out how to pay for it, and costs are coming down all the time. Give it a few years.

In the meantime, eat your vegetables, watch your saturated fat, stay active and have a little faith in the ongoing ingenuity of the human race. Despite occasional glitches, we're not so dumb.

I mean, we invented the salad fork.

What I Learned from My Nutrigenetics Test

by Lucy M. Casale

"Your Nutrition Genome Report Is Ready!" After seven weeks of waiting, the email had arrived in my inbox and I wondered what potentially life-altering diet insights I was about to uncover. (I'm an editor here at EatingWell and had volunteered to play guinea pig.) The company that did my report, Nutrition Genome, offers a 50-plus-page analysis ($300; that includes, among other things, a rundown of genetic "strengths" and "weaknesses," plus a personalized grocery list based on your DNA. Cool! And also, honestly, a little bit terrifying. Call it the fear-of-the-unknown factor. Some of my strengths:

Thanks to my APOA2 gene, I'm less likely to gain weight from eating saturated fats. Good news, since I love cheese.

Because of my "improved FTO gene function," I'm more apt to have normal levels of ghrelin (the hunger hormone), decreasing my risk for overeating and abdominal weight gain. Also good news, since I hate crunches and want to still fit in my jeans.

And my NOS1 gene may help lower the inflammatory process from psychological stress. Work deadlines? Bring 'em on!

Next up, my genetic minuses: Because of my BCMO1 gene, I may have a reduced conversion rate of plant-based beta carotene to vitamin A: "This increases your need for foods higher in vitamin A like eggs, cod liver oil, wild salmon oil and organ meats for skin, digestion, healthy eyes, lungs and immunity." OK, good to know.

I may have reduced PON1 gene function "for pesticide detoxification, and HDL and LDL oxidation." Luckily, there are "numerous strategies to improve PON1 including choosing organic foods, adequate calcium and magnesium, broccoli sprouts, high-quality olive oil, and a glass of red wine." Cheers to that!

Due to variants in GATA3, I could have increased sensitivity to processed meats and, as a result, colon cancer risk. I should "reduce processed meat intake, optimize vita­min D levels and increase berries, apples, sauerkraut, broccoli, tomatoes, basil, rosemary, garlic, ­onions and leeks." Adding sauerkraut to my shopping list; scratching off pepperoni.

The DNA-based grocery list I was given based on all of this featured many foods that are already regulars in my diet—bananas, olive oil, yogurt, berries, spinach and wild salmon. But an equal number were totally foreign, like wild boar, tiger nuts, yacon syrup, bilberry, pastured lard (??) and heart.

And while the report gave some fascinating insights into my genetics, it also felt like information overload and I wasn't sure what to do with it. EatingWell advisor David Katz, M.D., to whom I forwarded my report for input, was less impressed. "The recommendations are dubious," he told me. "For example, the possibility of some relative inefficiency in making vitamin A led to the recommendation to eat organ meats—with no consideration of how this would affect health overall, or interact with the many other suggestions. And the magnitude of these effects is entirely unclear. In some cases, the genes in question may exert a major effect, and in many, they could have a very minor effect. There's no basis to differentiate." To me, the results were nice to be aware of, but not worth changing my whole diet for—until nutrigenetics tests advance more, anyway. Maybe I'll take the test again in 2040.

Patrick Clinton is an Ann Arbor, Michigan-based journalist, educator and master science distiller. He has covered food, health, medicine and the bizarre world of food law and regulation for a variety of publications.

Eating for Your Genes: Personalized Nutrition | October 2019 EatingWell