Survey says: Genetic testing yielded unexpected results for both Yeoman and his dog, Scooter. Credit: Jared Lazarus
Survey says: Genetic testing yielded unexpected results for both Yeoman and his dog, Scooter. Credit: Jared Lazarus

Do-It-Yourself Genetics

The mapping of the human genome inspired not only a flood of scientific research, but also a flurry of commercial genetic tests aimed at the curious consumer. A Duke Magazine writer submitted his DNA and gained access to a trove of information that purported to reveal where his ancestors lived and his risk of becoming obese or developing arthritis or cancer.
Writer: 
November 30, 2010

When we adopted Scooter in 2008, my partner and I were curious about the ancestry of our sleek five-month-old puppy. Clearly she was a Labrador retriever mix, as the Durham County animal shelter had labeled her. She possessed the classic Lab muzzle, along with a dense black coat and an eagerness to please. But what else was she? With that broad forehead and the mouthy way she wrestled at the dog park, might there be some pit bull in Scooter’s background (not that we would mind)?

If Scooter had come into our lives a few years earlier, we would have had to rely on educated guesses. That was before the genetic revolution began flooding the consumer market with DNA tests. Now, in exchange for a blood sample, cheek scraping, or vial of saliva, we are promised insight not only into our own genetic roots, but also those of our dogs. Anyone can gain access to a trove of information that will supposedly reveal where our ancestors lived and whether we possess risk markers for baldness, obesity, arthritis, or cancer.

In 2003, the Human Genome Project completed the first detailed map of our DNA, sparking a frenzy of research. Ever since, that map has been scrutinized by scientists all over the world looking for markers for every imaginable illness and physical trait. But even before geneticists have fully digested these findings, entrepreneurs have begun designing test kits and selling them to the public. You can find out, at least according to the marketing claims, a good many things: your odds of suffering a heart attack or stroke; whether you’re descended from Genghis Khan; and even who might find your body odor alluring.

These are the perfect products for the twenty-first-century curiosity seeker. We are a nation of voyeurs: We watch reality television, read strangers’ personal blogs, and constantly check our friends’ Facebook statuses. As a journalist, I’m inquisitive too, and I have the privilege of listening as people tell me their most intimate stories. So perhaps it was only natural that I turn that curiosity inward; I decided to embark upon my own genetic journey and report my discoveries along the way. Because the best adventures often feature a faithful canine companion (and because we muttowners are relentlessly curious about where our dogs came from), I decided to bring Scooter along with me.

At the vet’s office, I watched as the receptionist sized up Scooter’s sinewy body. “Hello,” she said. “You’re a pit bull.” We would see. Scooter’s blood was drawn and sent to a canine genetics lab in Nebraska. The test, called the Wisdom Panel, identifies 170 breeds and claims an accuracy rate of 90 percent for mutts whose parents are purebreds, albeit from two different breeds. As I waited three weeks for her report, I started the process of submitting to my own DNA tests.


Traditionally, genetic testing has been the province of the medical establishment. Many couples learn about these tests as they embark on having families: It’s possible to screen for certain genetic defects in the parents before pregnancy, the fetus during pregnancy, and the baby immediately after birth. In the 1990s, as the science advanced, doctors rolled out a new type of test, which screens adults for mutations that predict cancer. Having a mutation doesn’t predetermine you’ll get sick, but it puts you at a much higher risk. The best known of these tests looks for alterations in two tumor-suppressing genes called BRCA1 and BRCA2. (The latter gene was located by a team that included Duke researchers.) A woman with a known BRCA mutation faces up to an 85 percent chance of developing breast cancer and a 60 percent chance of developing ovarian cancer. Women who test positive can discuss preventive steps, such as prophylactic mastectomy, with their physicians.

My own family has two BRCA risk factors: We’re Eastern European Jews, a population that carries the mutations in elevated numbers, and one of my great-grandfathers had breast cancer, which is rare in men. So I decided to start my genetic journey old-school—with a doctor-approved test in a hospital setting. I contacted the Duke Comprehensive Cancer Center and scheduled an appointment with genetic counselor Robin Hutchison King ’00.

Before I arrived at the clinic, King took my medical history by phone. Using a formula based on ethnicity and family disease, she calculated my chances of having a BRCA mutation at 6.5 percent—low enough not to panic but still twenty times higher than normal. “That’s telling us that it’s worth testing,” she said.

“If I were not doing an article, would you still recommend it?” I asked.

“Yes,” King replied. “When there is breast or ovarian cancer in a Jewish family, then it’s definitely worth testing.”

King walked me through the science, and we talked about what all the possible outcomes might mean. She told me about the Genetic Information Nondiscrimination Act, which protects patients from insurance and employment bias. And we discussed the price. For $460, covered partly by insurance, the test would search for three mutations commonly found in Eastern European Jews. I signed a consent form, and King led me down the hallway to have my blood drawn.

Over the eleven days it took for the results to come back, I tried to keep my anxiety levels manageable. The odds of being fine were overwhelmingly in my favor, but when I mentioned the male breast cancer in my family during telephone interviews with genetics experts, I noticed some uncomfortable pauses at the other end of the line. Each time I detected such a pause, I felt a little less like a detached journalist. I kept pacing the hallway between my living room and kitchen, asking my partner repeated questions that began, “What if…?” I wondered how I would break the news to my female relatives, who would likely bear the brunt of the mutation if they had inherited it. I began worrying that I’d acted irresponsibly by not first consulting with them. It was a great relief, then, when King called with the results: The DNA analysis showed no abnormalities in my BRCA genes.

Feeling more assured on life-and-death issues, at least for the moment, I turned to a less urgent, but more intriguing, question: Where in the world, literally, did I come from? Like many Americans, my family tree has invisible roots: My great-grandparents emigrated from Eastern Europe 100 years ago, leaving behind all records of our collective history. Without the benefit of genealogical records, perhaps my genes would tell me something about my ancestral past.

About forty companies and one nonprofit organization promise to help consumers recover some of that history. By analyzing your DNA, they claim, they can provide a general sense of where your forebears originated. There’s a catch, though. The two most popular kinds of tests—mitochondrial and Y-chromosome DNA—trace only a limited number of your ancestors. Mitochondrial tests follow your direct maternal line. They look at genetic material that’s passed from mother to child unchanged inside the cellular power plants called the mitochondria. Y-chromosome tests, available only to men, do the same for the paternal line. The problem is, if you go back ten generations, you have 1,024 ancestors, only two of whom stay genetically intact down these same-sex lines.

With that in mind, I ordered two separate tests. The first came from the National Geographic Society’s Genographic Project, which for $107 will analyze either your Y-chromosome or your mitochondrial DNA from a scraping of the inside of your cheek. The Genographic Project is a five-year effort to collect 100,000 DNA samples from around the world and construct a history of human migration. (Some Native-American organizations oppose the project, fearing that the new data could be used to deny their historic land claims.)

The other test came from 23andMe, a Silicon Valley company that, for $499, takes a deeper look into your genome. After its technicians analyze your saliva (the kit provides a plastic vial and miniature funnel for spitting), you get your personal report on its password-protected website. There you’ll find reams of information, not just about your roots, but also about possible disease risks (more on that later). Unlike most companies, 23andMe looks at ancestry markers throughout your genome—not just on the Y-chromosome or inside the mitochondria—capturing at least vague signals from your entire clan.

Making of a Man: Yeoman’s mitochondrial genome, above, shows a comparison of his DNA to that of the Cambridge Reference Sequence, a forerunner of the Human Genome Project that serves as a baseline of sorts for genealogical sequencing. Highlighted differences are mutations that cause the replacement of one nucleotide with another. For example, a pyrimidine base (C) is exchanged with another pyrimidine (T) or with a purine (A or G). These differences allow researchers to reconstruct the migratory paths of Yeoman’s ancestors.

Taking two tests, it turns out, is more confusing than taking one. When the results came back, National Geographic and 23andMe agreed on one thing: My mitochondrial DNA put me in a population subgroup, or haplogroup, called N9a. But their online presentations were strikingly different.

National Geographic showed me an interactive map and video describing what it called my maternal ancestors’ early migration. From East Africa, where they hunted antelope on the savannah, my forebears migrated to the Middle East about 50,000 years ago, then “went on to inhabit most of western Eurasia.” It seemed like a generic explanation, and on closer inspection, I understood why: National Geographic’s narrative applies to everyone in the enormous N population group, which has been described as “one of the two major trunks emerging from the original African root.” The Genographic Project’s director, Spencer Wells, explained to me that N9a has not been studied enough to warrant its own migration map. “We don’t want to overstate what we know,” he said.

23andMe gave me far more targeted information, focusing on where members of N9a live today. My subgroup is an “extremely rare” branch of that large trunk, the report said. “Its only concentrations are among indigenous inhabitants of the Malay peninsula.” I learned that I have mitochondrial cousins sprinkled throughout Asia, from southern China to Kazakhstan. “Ancient China’s famous Terracotta Army”—8,000 clay soldiers guarding Emperor Qin Shi Huang’s tomb—“was constructed by men bearing haplogroup N9a,” 23andMe’s website offered as a tidbit.

The Asia connection fascinated me. It didn’t surprise the geneticists I interviewed. Wells, who has worked extensively in Central Asia, speculated that several Jewish groups from the region—including the Khazars, a medieval people whose kings adopted Judaism—might have migrated to Europe and blended into the local gene pool. Researchers recently have been studying why some Jews possess this unusual Asian mutation set. “It’s very exciting that you got this result,” he said. “Your DNA is going to help us piece together this new story.”

Even so, Wells did note that DNA cannot pinpoint my origins with certainty. His colleagues agree. Charmaine Royal, an associate research professor at Duke’s Institute for Genome Sciences & Policy (IGSP), says genetic research has great potential to help reconstruct the human story. But she wants the public to understand more clearly the limits of today’s tests. “Genetics does tell us something about our ancestry,” she says. “My main concern is that the companies need to be clear with consumers about what they are and are not getting. There are people who think they’re getting a whole lot of information when they do lineage tests. The companies, for the most part, don’t say anything about limitations.”

The tests that have stirred up the most controversy are the DNA-based health kits sold directly to consumers over the Internet, offering results without ever involving a doctor or genetic counselor. These DIY tests promise to calculate your risk for all kinds of ailments, including heart attack, Alzheimer’s, and various forms of cancer.

The new at-home tests are a direct outgrowth of the Human Genome Project, a thirteen-year effort to identify the 20,000-plus genes and determine the sequence of the 3 billion chemical base pairs in human DNA. With that mapping now considered complete, scientists are able to scan the genome quickly, looking for tiny differences between people who have certain illnesses and others who don’t. These variations, which occur at specific locations on the DNA, are called single nucleotide polymorphisms or SNPs (pronounced “snips”). Many of the latest studies by geneticists studying specific diseases ferret out small associations. For example, someone with a certain SNP or set of SNPs might be 10 percent more likely than the general population to have a particular disease: Instead of a 24 percent chance of developing Type 2 diabetes, that person might have a 26 percent chance.

Based on identifying high-risk SNPs, the most ambitious DNA kits scan for many diseases at once. The two big players in this arena, 23andMe and deCODE genetics, calculate their customers’ risks and post the information to websites accessible by password. 23andMe also sells customers’ genetic data, along with other personal information that customers self-report, to researchers. The information is aggregated, and identifying details are removed.

Anne Wojcicki, who cofounded 23andMe after working as a health-care investor, says her company is helping create “a radical transformation” in medicine, giving customers information about their risk factors, then allowing them to make their own lifestyle decisions. “They’re going to be able to say, ‘Oh, I love wine, but I see I’m at higher risk for breast cancer. Should I modify my wine consumption?’ ” she says.

But critics call these DNA tests poor predictors. Many of the diseases in the 23andMe panel are caused by multiple genes working together—some we know of, but many we don’t—along with factors like diet, stress, and environmental contamination. Scientists have discovered interesting links, but their findings are preliminary and not well understood. Nor are the genetic variations necessarily causal. “For the vast majority of diseases that most of us will get, we remain more in the dark than illuminated,” says David Goldstein, director of the Duke Center for Human Genome Variations. There’s a reason these consumer tests are often called “recreational genomics”—according to skeptics, they titillate more than educate.

A major concern among scientists and physicians is that the private companies are putting test results directly into the public’s hands without the consultation of a doctor and the expertise of a genetic counselor—think of ordering and interpreting your own mammogram. Experts worry that consumers might make unhealthy decisions based on their results. “I’m most concerned with those who get their tests and believe they’re home free,” says Susanne Haga, an assistant research professor at Duke’s IGSP.“They may say, ‘That’s great, I’m not at risk for diabetes. I can eat more fat and sugar than I would have before.’ That’s what we definitely don’t want to encourage.” On the flip side, Haga notes, someone deemed at high risk for a disease might undergo unnecessary and invasive testing.

Superstitiously, I waited a week after 23andMe notified me by e-mail that my reports were viewable online. I wanted to brace for any alarming news. When I finally clicked through the links that took me to various disease reports, most of them brought amusement or relief. I learned that I have a “greatly decreased” likelihood of male pattern baldness—a surprise given my deeply receded hairline—and a propensity to dislike Brussels sprouts, my favorite vegetable, based on my inherited ability to taste bitterness. As for more serious conditions, my SNPs pointed to a lowered risk for heart attack, rheumatoid arthritis, and celiac disease, and only the slimmest elevation for obesity and colorectal cancer.

Nothing remarkable, I told myself. Then I realized there was one disease risk I had forgotten to check.

I logged back into my results—and felt my stomach contort as I read the computer screen. According to 23andMe’s calculation, I had a 31 percent chance of developing prostate cancer over my lifetime, twice the average for European males.

Prostate cancer is the most common cancer among men in the U.S. It kills almost 30,000 people a year, and the surgery can leave patients with urinary incontinence and sexual impotence. In 2008, the New England Journal of Medicine published an article linking five particular SNPs to prostate cancer. I found out that I have four of those five markers.

By the time I read my results, I knew that SNPs were not considered reliable disease predictors, and that 23andMe’s calculations excluded diet, family history, and undiscovered genetic factors. I also knew that prostate malignancies often grew so slowly that most patients die with the cancer rather than from it.

That didn’t stop me from panicking. The Washington Post had just reported that deCODE’s chief scientific officer, Jeffrey Gulcher, had received results similar to mine and was shortly thereafter diagnosed with aggressive prostate cancer. Gulcher and I were both slightly shy of fifty, the age at which screening is recommended. It didn’t matter that his illness was a statistical fluke. The coincidence was too creepy to shake off.

I had dinner with friends that night, but I didn’t say a word during the conversation and can’t remember what we ate. I was pulled deep into myself, imagining the specter of a tumor growing inside me, contemplating my mortality, engaging in a bit of supernatural bargaining. I e-mailed my friend Duane Culler, a former Duke genetic counselor and now a senior genetic associate at University Hospitals Case Medical Center in Cleveland. “Should I be worried?” I asked, downplaying the fact that I already was. Culler responded immediately. “The first thing to do is take a deep breath,” he wrote. “Tell yourself this test isn’t used clinically because it hasn’t been validated.”

Over the next week, several physicians pointed out that I’d fallen into a trap of at-home tests: I had learned about a risk that may, or may not, exist without a genetic counselor on board to help me interpret the science. I had gained enough information to feel anxious, but not enough to do anything useful. “It’s, frankly, exactly the kind of situation I worry about,” said James Evans, a professor of genetics and medicine at the University of North Carolina at Chapel Hill. “You and I both know what most people are going to do with these results. It’s hard to sleep at night with somebody telling you that you have an increased risk of prostate cancer.” Patients will ask their doctors to screen them earlier than recommended—and the results, Evans said, could take them down a path toward inconclusive biopsies and pointless prostate removal.

Much later—after I had consulted with King, the genetic counselor at Duke’s Cancer Center; talked with my own physician; learned more about my family history; and decided to defer screening until I turned fifty—I told the story to Misha Angrist, an assistant professor at Duke’s IGSP and author of Here Is A Human Being. “I would draw the exact opposite conclusion from Jim Evans,” he said. “This spurred a fact-finding mission, which turned up useful information.” Once I survived my initial “freak-out,” Angrist said, I was able to make a better informed decision about my health.

But some consumers lack the knowledge, or the stomach, to handle scary genetic data. For that reason, Angrist says, not everyone should purchase over-the-counter tests. For those who do, he says, “this is one of the challenges: to take this information in perspective, and to be more accepting of the freak-out.”

The results are in: Yeoman’s test results indicated mitochondrial cousins “sprinkled throughout Asia, from China to Kazakhstan”; Scooter’s suggested she wasn’t the Lab mix she seemed to be. Credit: Jared Lazarus

When I told friends about my genetic journey, they listened politely to my tales of cancer terror and Asian ancestry. But their ears often pricked up when I talked about my dog’s DNA test. Durham, it seems, is a city of mutts, and all the dog owners I know have a nagging curiosity about their pets’ true origins.

It turns out that the test Scooter took, the Mars Wisdom Panel, emerged from serious human medical research. At Seattle’s Fred Hutchinson Cancer Research Center, geneticist Elaine Ostrander and genetic statistician Leonid Kruglyak recognized that some breeds are plagued by specific conditions. Scottish terriers, for example, have astronomically high rates of bladder cancer, and Doberman Pinschers are prone to narcolepsy. To help locate the disease markers, the two scientists mapped the genetic architecture of the canine kingdom, looking for the “DNA fingerprints,” as Ostrander calls them, that distinguish one breed from another. Their research was then licensed to Mars Veterinary, a British firm whose corporate parent also sells candy and pet food. (Ostrander and Kruglyak have since moved on to other jobs.)

The Wisdom Panel is not a medical screening. “There’s very little evidence to suggest that the single-gene diseases that are found at high frequency in some of the breeds are actually that common in mixed-breed dogs,” says Neale Fretwell, the R&D director for Mars Veterinary. Instead, the tests mostly serve to satisfy the curiosity of owners like me.

Two weeks after we submitted Scooter’s blood, her results came back: Our “Lab mix” had absolutely no markers for Labrador retriever. Instead, she showed traces of boxer, chow chow, and American Staffordshire terrier (a type of pit bull), along with less reliable signals for Chinese shar-pei and bearded collie.

“She looks like a black Lab,” I protested to Paul Jones, a geneticist who helped design the panel.

“A lot of people think they’ve got a Lab,” he replied. “What you often achieve by mixing different guard-dog types—which is virtually all of what’s in your dog—is a mixed, outbred dog that ends up looking very much like a Labrador. It’s a story we’ve heard a fair few times.”

This has made for good conversation at the dog park. While Scooter tussles playfully with the other dogs—many of them mutts with Lab-like faces—I explain to the owners why their pets might not be Labradors after all. But then Scooter will break from the pack, find an errant tennis ball, and bring it back with the hope that I’ll throw it. She can barely control herself from springing off the ground in anticipation. DNA is not always destiny: Scooter might have zero retriever in her, but she still fetches with the best of them.