By Peter Jaret, November/December 2008
After living together for more than three decades, my partner Steven and I couldn’t be more alike. We eat the same meals. We trudge off to the gym on the same schedule. Each of us could fit, with a bit of tugging, into the jeans we wore in college. Steven has put up with the brief appearance in our kitchen of all passing health-food fads. (Remember broccoli sprouts?) I like to think the best choices have stuck around—rolled oats, low-fat yogurt, nuts, plates piled high with a colorful mix of vegetables, a piece of dark chocolate for dessert. Our cholesterol levels should be perfect. And, at a recent doctor’s visit, mine were—so good that even the doctor was impressed.
Steven’s lurked deep in the danger zone.
“Genes," the doctor breezily explained, when Steven objected to the injustice of it all. Then he wrote Steven a prescription for a cholesterol-lowering drug.
Maybe we aren’t so alike after all. Did Steven’s “bad" genes curse him with high cholesterol, despite following the healthiest of diets? And did “good" genes mean I could feast on junk food, if I chose to, and still stay healthy? What about the effects of specific foods? Should we be following different diets, less of this for him, more of that for me?
My questions led me to the emerging science of nutritional genomics (a.k.a. nutrigenomics)—the study of interactions between genes and diet. The latest studies suggest that the interplay between DNA and diet may have a powerful influence on what we like to eat (bitter vegetables or sweets), why some people get fat and others stay thin, why some develop diabetes and others don’t, why certain people get a buzz from coffee and others aren’t affected—and yes, why two people following the same diet can end up with cholesterol levels at opposite extremes. There are already more than 30 companies that offer some type of genetic screening and personalized nutrition advice. Should Steven and I take one of these tests, which typically involve swabbing a cheek for DNA and filling out some questionnaires, to find out? Before shelling out anywhere from $300 to $2,500 per test, I decided to do some more digging.
Like Sweets? Blame Your Grandparents
In 2002, the Human Genome Project completed an awesome undertaking: the first comprehensive map of the tens of thousands of genes, those tiny bits of DNA, that make us human. Yet as towering as that achievement was, what researchers have discovered in the years since may be even more important. Although all of us have very similar genes, hundreds of thousands of tiny differences exist. These genetic variations—what scientists refer to as single nucleotide polymorphisms, or SNPs—make each of us unique in all sorts of ways, including how our bodies respond to different types of foods. For example, one gene discovered in May of this year appears to stimulate the desire to eat sweet foods. People with a particular form of this gene consistently consume more sugar and sweet foods and beverages than those with a slightly different one.
Could the discovery of this “sweet tooth gene" help explain why some people are more likely to become overweight? “It’s easy just to say that people get fat because they eat too much," David M. Mutch, Ph.D., told me when I rang him up in his office at INSERM, the French Institute for Health and Medical Research, in Paris. “But it’s much more complicated than that."
When I reached Mutch he was making plans to return to Canada, where he grew up, to start a nutritional genomics research group at the University of Guelph, in Ontario—one of dozens of such groups being established around the world.
“The numbers are pretty staggering: worldwide, 1 billion people are overweight and 300 million of them are obese," he said. “The problem is spreading and so far we haven’t been able to do much to stop it. The hope is that nutritional genomics may help explain why some populations are especially susceptible to obesity, and perhaps offer better ways to prevent or treat it."
For more than two decades, researchers have speculated that so-called “thrifty genes" may make some people more likely to put on pounds than others. According to the prevailing theory, certain populations evolved during times of frequent famines to have genes that are especially thrifty, storing every extra calorie of energy in the form of body fat. Those thrifty genes are life-saving when food is scarce. When food is abundant, they may increase some people’s risk of becoming fat. The Pima Indians of Mexico, for instance, are lean and have very little risk of diabetes when they continue to eat their traditional diet, which is abundant in beans, seeds and whole grains. But when Pima Indians moved to the southwestern United States, where fast-food restaurants flourish, they quickly began to suffer exceptionally high rates of obesity and diabetes.
Indeed, new research suggests that the level to which a poor diet leads to weight gain depends partly on specific genes. Last year, data from the Framingham Heart Study, which has followed thousands of people spanning two generations in the town of Framingham, Massachusetts, revealed a variant gene that makes people who carry it especially susceptible to weight gain when they eat a high-fat diet.
But it’s far more complicated than identifying just one gene to understand and solve the obesity epidemic. “So far, we’ve identified more than 300 genes that influence body weight and body fat," Mutch explained. Some affect appetite and satiety. Others appear to influence how efficiently people burn fat for energy.
My Metabolism, Your Metabolism
Researchers are still a long way from understanding the whole picture. But nutrigenomics is starting to explain several baffling mysteries. Consider coffee. A number of studies offer evidence that drinking java lowers heart-disease risk, most likely as a result of antioxidants and other beneficial compounds in coffee beans. But a few studies show heavy coffee drinkers having a higher than average risk of heart disease—leaving scientists scratching their heads. Nutrigenomics suggests an explanation. In people with the genetic variant that causes sluggish metabolism of caffeine, the stimulant sticks around in the bloodstream longer than usual, where it may disrupt normal heart rhythms and boost blood pressure, overwhelming any benefit. Quick metabolizers, on the other hand, clear the caffeine fast from their bloodstreams but still enjoy the benefits.
Type 2 diabetes offers another example. Several studies suggest eating too much sugar and refined carbohydrates—foods linked with big jumps in blood sugar—can lead to type 2 diabetes. But when researchers look at large groups of subjects, no clear link emerges. The reason may be that only some people are genetically sensitive to the effects of these foods on blood sugar.
When researchers do large studies and pool the data, treating all the subjects as if they’re genetically alike, such differences typically get lost in the averages. What’s more, most prevailing diet recommendations are based on such studies. So the advice may work for most people but not all. “Dietary recommendations are based on averages across large populations," says Jose Ordovas, Ph.D., who directs the Nutrition and Genomics Laboratory at Tufts University in Boston and has published more research on diet and gene interactions than almost anyone in the field. “What nutritional genomics teaches is there is no one-size-fits-all diet that works for everyone."
The ideal heart disease–prevention diet may differ significantly among different people, says Ordovas, who studied the biochemistry of cholesterol before he got “hooked" on genetic-related research after attending an American Heart Association seminar on the topic in 1983. Preliminary research suggests that some people reduce their risk for heart disease most significantly by following a low-fat diet; others fare better on a higher-fat diet (assuming they choose healthy fats). Similarly, studies show that some people shed pounds on low-fat, high-carb regimens; others seem to do best on low-carb, higher-fat regimens. Genetic screening may eventually be able to pinpoint who’s who, allowing doctors to offer individualized advice.
And what about the mystery under my own roof? Can nutrigenomics explain why some people’s cholesterol levels respond to a healthy diet and others’ don’t? Here, too, there are plenty of clues. Scientists have detected one gene variation that seems to enhance the health benefits of polyunsaturated fats, for example, giving people who possess it a bigger boost in good cholesterol when they eat a diet rich in plant oils. Another appears to make bad-cholesterol levels more likely to soar when people eat a high-fat diet. “Variants in a gene called APOE, which controls cholesterol metabolism, seem to be especially important," Ordovas told me. People with one genetic pattern see a big drop in cholesterol levels when they switch to a healthier diet. Those with a slightly different pattern get almost no benefit at all.
To Test or Not To Test?
Findings like those Ordovas described could someday allow doctors to know in advance who can control their cholesterol levels through a healthier diet and who, like Steven, will also need medication to keep their numbers in check. But almost all the researchers I spoke with said it was too early to offer specific nutrition advice based on nutrigenomics.
Prowl the Internet, though, and you’ll find plenty of companies already doing just that. “Identify your inherited genetic variations and understand how they influence your Health & Well-being," proclaims Sciona, a Denver-based company, on its splashy website. “Learn whether variations in your own genome have been associated with a genetic risk for a number of common complex diseases," promises another, called deCODEme.com, based in Iceland.
Most of these direct-to-consumer companies use mail-order test kits that allow customers to collect a sample of their own DNA with a cheek swab and ask them to complete questionnaires that ask about diet, exercise and other lifestyle behaviors. The DNA is then analyzed for genetic variants that have been linked to health problems, food preferences, athletic performance and other traits. To find out more, I called Rosalynn Gill, Ph.D., a molecular biologist and founder of Sciona. The company, she told me, had recently published a study that suggests that genetic screening and personalized nutrition advice may already help people shed pounds more successfully.
Researchers at the National Technical University of Athens, in Greece, conducted the experiment with 93 overweight or obese volunteers who had tried and failed to lose weight. Everyone in the study was given the same general weight-loss advice—to follow a Mediterranean-style diet based on fish and poultry, legumes, fresh fruit, vegetables and whole grains. (They were also asked to follow a recommended exercise routine.) In addition, 50 of the study’s participants received Sciona’s genetic test and additional personalized nutrition advice based on the results. For example, those with a gene variation that makes it harder for the body to metabolize folate, a B vitamin associated with lower risk for heart disease and some cancers, were encouraged to take a supplement.
The aim of the study was to determine whether the company’s genetic test would help ensure that dieters get adequate nutrients while they are cutting back on calories, not whether genetic testing would boost weight loss, Gill explained. And, in fact, none of the personalized advice from the genetic screens related to weight loss. But when researchers ran the numbers, something unexpected popped up. The volunteers who received genetic screening lost more weight. After almost a year, they had lowered their BMI by 5.6 percent, on average, compared to a 2.2 percent gain among the control subjects. “We think genetic screening may have encouraged people to stick a little more closely to a healthy diet," Gill said. “If you think the recommendations are absolutely about you, you’re much more likely to take them to heart." That may be. But anyone forking over $299 or more for a DNA test is likely to want more than just a motivational boost.
A test offered by Interleukin Genetics, another direct-to-consumer genetic-screening company, analyzes genes linked to inflammation, which is believed to be a risk factor for heart disease and other health problems. "If you carry those variants, you may be able to reduce your risk through diet or a nutritional supplement that reduces inflammation," Kenneth Kornman, Ph.D., the company"s founder, told me. To test that idea, Interleukin Genetics recently collaborated with a company called Nutrilite to create such a supplement. (The pills included rose hips, powdered blueberry and blackberry, grapevine extract and other substances linked with reducing inflammation.) Seventy-nine volunteers who were screened using the company"s genetic test took the pills—or placebo tablets—daily for 12 weeks. After taking the supplement, some of those who tested positive for the gene variant linked to increased risk of inflammation saw a drop in inflammatory markers. For those who tested negative for this gene variant, however, taking the supplement had no effect on inflammation levels. The DNA test, in other words, successfully identified people who would get the biggest benefit from taking a supplement.
"If you"re investing in dietary supplements to prevent chronic illnesses like heart disease," Kornman told me, "a test like this can help you make the best choice."
So has the era of nutrigenomics already arrived? I put that question to Jim Kaput, Ph.D., who directs the Food and Drug Administration"s Division of Personalized Nutrition and Medicine. The FDA created the new division in 2006 to advance research in nutritional genomics. Its existence is as good a measure as any of how much excitement the new field has generated. But when I asked Kaput if nutrigenomics screening was ready for prime time, his answer surprised me: "My personal opinion? Not yet."
One of his concerns is that most companies currently test only a few dozen gene variants. Research already shows that hundreds, even thousands, may impact how diet affects your risk of heart disease or diabetes—and the genes that companies screen for today may not even prove to be all that important once the whole picture becomes clearer. "To study nutritional genomics, you need to have good data about what people actually eat, and you have to have a good understanding of how genes work," Kaput said. "We have a lot of work to do on both counts." Part of the task of the new division, he told me, is to improve the quality of data by encouraging better studies. "In five years, I think we"ll begin to have the kind of solid data we really need to make sense of the connections. It"ll take longer than that to offer useful advice based on what we learn."
The Ethics of It All
Sara Katsanis, a research analyst for the Genetics & Public Policy Center at Johns Hopkins University, worries that the risks of the at-home genetic-testing kits go beyond wasting money on something unsubstantiated by hard science. "Along with genetic tests, most direct-to-consumer companies also dispense information and advice," she explained. Some firms will tell you how much certain genetic variations increase your risk of heart disease, diabetes, macular degeneration or other conditions. They'll also give you recommendations on how to lower your risk. "But at the moment there is almost no federal oversight of the claims these companies make," Katsanis said. There's no way to know, in other words, how reliable these tests are. And because the companies sell their services directly to consumers, doctors aren't usually involved in interpreting the results.
Not long ago, in fact, California banned direct-to-consumer genetic screening and the State of New York sent out warning letters, arguing that tests that provide medical information must be ordered, or at least supervised, by a health-care professional. More recently, California licensed two companies that use in-house doctors to review orders for the test. Health officials there say additional companies are likely to win approval as the state reviews their applications.
Even with appropriate medical supervision, there's the issue of privacy. A genetic-screening test contains a vast amount of information about you, including variants that may put you at risk of serious illnesses. Because direct-to-consumer companies aren't medical entities, they are not required to abide by the Health Insurance Portability and Accountability Act (HIPAA), although some states may have regulations that help protect consumer privacy. "Many of these companies do everything they can to protect your privacy," Katsanis said. "But that's no guarantee." Indeed, most companies explicitly say on their websites that they can't guarantee privacy. The danger is that if the results get into the wrong hands, people with genetic-disease risks could suffer discrimination in the workplace or when applying for health insurance. Several sites offer the option of e-mailing your genetic results to friends and family, I discovered. How secure is that?
Steven and I toyed with the idea of getting tested. It would be interesting to compare our genomes, to see if any gene variations popped up to explain the mystery of our divergent cholesterol numbers. As some of the websites promised, it would even be fun. "Get to know your friends and family through genetics," declared a company called 23andMe.com, one of the leading direct-to-consumer firms. "Add some excitement to your family reunion."
But the more we thought about it, the less enthusiastic we became. The issue of privacy spooked us. The price—$1,000 per test, on average—didn't seem worth it after what I'd learned. We don't need a genome scan to tell us that Steven's cholesterol numbers won't drop any lower no matter how many more bowls of oatmeal he eats. And when we glanced over the personalized recommendations many of the companies offer, much of the advice looked a lot like standard nutrition guidance.
Eventually, as researchers get a better handle on which genetic variants are most important and precisely how they affect health, nutrigenomics will be able to offer much more specific advice. For now, though, we agreed to forgo genetic testing and go on eating the way we always have. That clearly won't be enough to rein in Steven's cholesterol. But odds are it offers us a slew of other benefits, from lower blood pressure to less risk of heart disease, diabetes and certain forms of cancer. Eating well also keeps our weights in a healthy range, which subsequently makes it easier to stay active—two important factors in warding off disease.
And who knows? Nutrigenomics research may yet turn up some food that's especially potent for someone with Steven's particular genotype.
I wonder if anyone's looked at broccoli sprouts.
Peter Jaret's story "The Search for the Anti-Aging Diet" (November/December 2007) won a James Beard Foundation journalism award. His most recent book is Nurse: A World of Care (Emory University Press, 2008).