You are what you eat—and what you eat may be encoded in your DNA. Studies have indicated that your genetics play a role in determining the foods you find delicious or disgusting. But exactly how big a role they play has been difficult to pin down. “Everything has a genetic component, even if it’s small,” says Joanne Cole, a geneticist and an assistant professor at the University of Colorado School of Medicine. “We know that there is some genetic contribution to why we eat the foods we eat. Can we take the next step and actually pinpoint the regions in the genome?”
New research led by Cole has gotten a step closer. Through a large-scale genomics analysis, her team has identified 481 genome regions, or loci, that were directly linked to dietary patterns and food preferences. The findings, which have not yet been peer-reviewed, were presented last month at the American Society for Nutrition’s annual flagship conference. They build on a 2020 Nature Communications study by Cole and her colleagues that used data from the U.K. Biobank, a public database of the genetic and health information of 500,000 participants. By scanning genomes, the new analysis was able to home in on 194 regions associated with dietary patterns and 287 linked to specific foods such as fruit, cheese, fish, tea and alcohol. Further understanding how genetics impact how we eat could reveal differences in nutritional needs or disease risks.
“One of the problems with a lot of these genomics studies is that they’re very small. They don’t have enough people to really be able to identify genes in ways that are credible. This study had a huge cohort of people, so that’s really powerful,” says Monica Dus, an associate professor at the University of Michigan, who wasn’t a part of the new research but studies the relationship between genes and nutrition. “The other thing that I thought was really great is that they have so many different traits that they’re measuring in respect to diet. They had cholesterol, the body, socioeconomic backgrounds.”
As the research advances, Dus says such genome analyses could possibly help health care providers—and even policymakers—address larger issues that affect food access and health. “Instead of trying to obsess over telling people to eat this or that, a more powerful intervention is to link it to making sure there aren’t ‘food deserts’ or to make sure that there’s a higher minimum wage—things that have a broader impact,” she says.
Scientific American spoke with Cole about the new research and what it tells us about the relationship between our genetics and food.
[An edited transcript of the interview follows.]
What do we know about the link between genetics and diet?
There has been research on it for more than 100 years. Some of the earliest studies are heritability studies. Heritability is the amount of genetics that contribute to a trait. So traits such as height, which is very genetic, can have a 50 to 80 percent heritability. But diet probably has a really small heritability, or genetic component, because it’s influenced by so many other important things such as socioeconomic status, culture, upbringing and all these other factors that have nothing to do with how you taste or like foods. It wasn’t until newer methods came out in the past 15 years or so that we could use thousands of distantly related individuals [to study genetic associations with foods]. It helped remove that messy environmental component, and we could get more accurate estimates of heritability.
In 2020 I did a heritability analysis where we scanned the genome to find regions that were statistically associated [with foods and diet] and narrowed those regions down. I found that the median of the genetic component of most dietary traits was only about 5 percent. So it meant that they are super environmental traits—but it doesn’t mean that that 5 percent is nothing. Because we have those big data now, we can study that 5 percent.
What did you find in your recent genome association analyses?
We found regions in the genome that are likely influencing a variety of dietary traits, including things like fruit, poultry or fish intake or different kinds of coffee or alcohol intake. We were also able to see some genetic associations with more complicated dietary patterns, such as if you eat a healthy versus unhealthy diet.
Some of the genes that had the strongest effect on diet, which I find most compelling, are the taste receptor, olfactory receptor and digestive enzyme genes. This includes the bitter taste receptor gene—most prominently known for influencing intake of cruciferous vegetables, such as Brussels sprouts. In most of our diets, we eat various foods together, so the foods we eat are often connected to each other. But the olfactory receptor genes we identified were so specific. One was associated with cheese and nothing else. There’s one very specific to fruit. There’s one very specific to vegetables. There’s one very specific to coffee.
There was one region in the genome for an olfactory receptor associated with how much tea someone drinks—that has a really fascinating story. Researchers have found a genotype, or a different version of a genetic variant, that explains someone’s ability to smell a compound called beta-ionone. If you have one version of this olfactory receptor gene, then you can smell beta-ionone, and if you have the other version, you [are less sensitive to the smell or] can’t smell it at all. It actually turns out that beta-ionone is in a bunch of things—such as tobacco, grapes, orange juice, papaya, peaches, raspberries, spearmint and tea. So this one version of this genotype may dictate—a little bit, to some degree—whether you smoke, because it’s in tobacco, or whether you eat grapes or drink orange juice or tea.
How were you able to tease out other influences, such as environmental and social factors?
It’s an extremely hard problem in the field of genetics. One of the major limitations with these genetic associations with diet is that they’re so influenced by your environment, educational attainment, socioeconomic status and income.
How do we also know whether a gene is actually impacting dietary intake or if it’s another trait? A good example would be a gene that increases your risk for diabetes. When you have diabetes, you tend to change your diet in order to manage the disease. If we find a gene that links to diabetes, it’s going to look like it’s also associated with diet because diabetes is associated with diet.
I did an analysis that essentially takes a genetic variant that you know is strongly associated with, say, fruit intake and see if it also has an association with other traits—whether it’s a health trait, such as diabetes, or a lifestyle trait, such as socioeconomic status. If the genetic variant has a stronger association with something related to the environment, health conditions or all the other outside factors that influence dietary intake, it’s probably not associated with a direct mechanism, such as digestion or taste.
My goal is to find the genes that are really closely linked to dietary intake and not other conditions because I want to see if we can act on those genes’ biological pathways. For instance, we could alter the flavor compound that binds to [specific receptors] to elicit a different brain response. Perhaps that could improve people’s nutritional adherence to healthier foods for disease management.
Where would you like the research to go next?
I think there’s a whole slew of studies and analyses that we can do. This one is just scratching the surface. Now that we have these regions in the genome, let’s figure them out and identify those genes that have a direct mechanism impact on dietary intake.
The sensory mechanisms are one of my favorites because I think that we could really use them as a tool. They tend to be these lock-and-key receptors, and if we create a different key, we might be able to alter the bindings. So I think that there’s potential, scientifically, to intervene with them. One of the things I’m particularly interested in is understanding how having different genotypes, particularly of these sensory genes, changes the activation of pleasure and reward regions in the brain. I wonder if we can start creating synthetic or even natural compounds that could alter someone’s pleasure reaction either to healthy or unhealthy foods—if we can use this almost as a supplement to change how people like food or not. Flavor is actually the number-one driver for food choice, even among all the other factors that influence whether someone eats certain foods or not. So can we modify the perception of flavor by understanding the biology?
There’s also some room to assess whether certain foods are actually impacting different health conditions. Unfortunately, in the nutrition field, there have been a lot of studies based on correlation. They’re studies that say a food is associated with lowering cholesterol, but many of those studies aren’t successful in a randomized controlled trial in humans—which is really hard to do and expensive. There’s a method in genetics called Mendelian randomization that mimics a randomized controlled trial, so we can test for causal and not correlative associations between different foods and different diseases. By identifying these regions in the genome that are influencing dietary intake, we can plug them into this method and get more information on what diseases certain foods actually causally influence.