Direct-to-consumer genetic tests allow anyone with a few hundred dollars to submit a saliva sample and get back genetic information on everything from family ancestry to Alzheimer's disease.
The powerful technology could lead to personalized medical treatments based on individuals' genetic risks. But these tests have long been controversial, with many bioethicists worried they could mislead people about their disease risks.
Now the government is taking a harder look at the tests. A Food and Drug Administration warning earlier this month prompted drugstore chain Walgreens to delay plans to sell gene test kits in stores. Less than a week later, the House of Representatives Energy and Finance Committee launched an investigation of personal genomics companies 23andMe, Navigenics and Pathway Genomics.
And government officials aren't the only ones interested in learning more about how these gene tests tick. In recent years, researchers have started to investigate why people subject their genome to such scrutiny — and what they do when they get the results.
Genetics experts have a standard list of concerns when it comes to direct-to-consumer tests. People might overreact to their results and subject themselves to unnecessary stress. Or they may find out they're at low risk for a particular disease and take it as license to partake in unhealthy behavior. Finding out you're at risk for something nasty and incurable is another concern, as is finding out genetic information that might affect other family members who didn't consent to testing.
But until recently, these concerns have been mostly theoretical, said Barbara Bernhardt, a genetic counselor and professor of medicine at the University of Pennsylvania School of Medicine. Few studies have examined what people actually do once the genetic report is in their hands. Do they understand the results? Do they share them with their doctors? Do they change their behaviors, exercising more if they have a high genetic risk of heart disease — or shrugging and reaching for that second doughnut regardless?
Risks and responses
Bernhardt is a collaborator on one of the only large studies aimed at answering these questions. The Coriell Personalized Medicine Collaborative, a project of the Coriell Institute for Medical Research in Camden, N.J., has scanned the genomes of about 5,000 healthy participants so far. The goal, says head researcher Michael Christman, the institute's president and CEO, is to follow the participants over time to see how the information affects their health. The results should help healthcare professionals integrate genomics testing into the medical system.
The process differs from direct-to-consumer testing in that participants go through a rigorous, in-person informed consent session before the test. During the session, they're told what the test can and cannot predict.
The Coriell study only provides information on "actionable" results, meaning diseases that can potentially be prevented. That means that, unlike many commercial tests, the Coriell study doesn't report genetic risk factors for non-preventable diseases like Alzheimer's.
The good news is that most patients seem to understand basic genetics fairly well, Bernhardt said. They know that one copy of a bad gene variation raises their risk, while two copies raise it even more. Almost all — 96 percent — understand that the diseases being tested for have both genetic and behavioral causes. And so far, no one seems to be using the results as license to misbehave or as reason for undue worry.
"We really haven't found those kinds of results at all, so in a way, our study findings have been fairly reassuring," Bernhardt said.
Interviews with 44 participants up to a year after the tests revealed that one-third had shared their results with a doctor. Half of those people had received behavior or lifestyle recommendations from their doctors based on the data; 25 percent reported that their physicians didn't know what to make of their information. Bernhardt presented these results at a May 19 roundtable discussion at the University of Pennsylvania.
Meanwhile, participants' own grasp on the results was a "mixed bag," Christman said. Most understand whether their genes put them at higher or lower risk for particular diseases, although some people have trouble grasping the magnitude of the risk when compared with the rest of the population. They also tend to cling to some misconceptions, like the belief that genetic risk information is more important than family history or behavioral factors like diet.
"There's something special about genetic information in the eyes of participants," Christman said.
The researchers also asked participants whether they were changing their lifestyles after getting the results. A "sizable group" said that they planned to change their behavior, Bernhardt said, but the plans were often vague.
One person joined a gym; another stopped smoking. Otherwise, she said, "almost no one started to engage in risk-reduction behavior that they hadn't already been doing."
From science geeks to standard medicine
Other studies find similar patterns. A small pilot study published in 2009 in the journal Cancer Epidemiology, Biomarkers and Prevention found that smokers given genetics testing to determine their risk of lung cancer had a good understanding of their results.
The researchers offered free smoking-cessation treatments, and 91 percent of participants availed themselves of the service. But there were no differences in motivation to quit between the genetically high-risk and low-risk respondents, perhaps because all of the participants had a relative newly diagnosed with lung cancer.
In contrast, the Coriell study provides a glimpse into the motivations of the current crop of casual genetics consumers. Coriell's participants match the typical demographic closely, Bernhardt said: mostly white, highly educated, and very curious.
"They're science geeks," she said.
But things could change once personalized medicine becomes more common, Bernhardt said. New demographics will likely get their genomes scanned, and the practice will become less about curiosity and entertainment and more about assigning the right type and dose of drugs based on an individual's genome.
Such advances could be lifesaving, but the evolution of genomic testing may mean that the current research on how people use the tests will no longer apply.
"Five years from now, we'll have a very different type of person that's going to be going into this," Bernhardt said. "The landscape is going to change."