Ever since direct-to-consumer genetic tests—which analyze a set of genetic markers across an individual’s genome in order to predict his or her risk of developing diabetes, heart disease, prostate cancer, and other common diseases—became available in 2007, experts have debated the consequences. The hope was that the tests would encourage consumers take steps to avoid the maladies for which they were predicted to be at high risk. But skeptics doubted that people would respond differently to genetic-risk information than they do to other, better validated measures, such as blood pressure and cholesterol levels.
The tests, which can be ordered over the Internet, also raised concerns that consumers told they were at increased risk for certain diseases could become overly anxious about the results. Another concern was that the results would lead doctors to order unnecessary follow-up tests, taxing an already overloaded health-care system.
The first large-scale analysis of the behavioral and psychological impact of these tests suggests they have little impact at all.
Eric Topol, director of the Scripps Translational Science Institute, and collaborators surveyed 2,000 people from several San Diego-area companies who were offered a genetic test from Navigenics, one of the first companies to offer genome-wide tests. The researchers found that the vast majority didn’t become overly anxious about the results, nor did they change their behavior by exercising more or altering their diet. The findings were published today in the New England Journal of Medicine.
“For those who hope that genetic results would be the magic bullet to induce people to live healthier lifestyles, I think this is sobering but utterly predictable,” says James Evans, a geneticist at the University of North Carolina, who was not involved in the study. “We already know that giving people all kinds of medical test results doesn’t seem to have profound behavioral impact. I think it was a naïve hope that there would be a something magic about genetic information that would inspire people to exercise and lose weight and eat better.”
On the flip side, people seemed to cope well with the information. “At least, we know for the first time that people hit with this much data did much better in dealing with it than the medical community had predicted,” says Topol. “Many people were concerned that [consumers] would go out and get every test known to mankind.”
Topol cautions that the group studied was not representative of the population as a whole—they worked at health and technology companies in San Diego. “It is representative of those who would buy this test,” because they chose to take it, he says. “But we can’t extrapolate to all comers.”
“This was a highly self-selected group,” says Evans. “The truth is that hordes of people are not availing themselves of this, and I am sure they are not representative of the population as a whole.”
The risk information detected by these tests is in most cases modest at best, and that may help explain both the lack of anxiety over the results and the failure to make healthier lifestyle choices. (One of the major criticisms of the tests has been that they offer little useful information.) “I’m not surprised by the lack of apparent effect on behavior,” says Michael Christman, president of the Coriell Institute for Medical Research, a nonprofit institute that is running a similar study. “We understand only a tiny fraction of total genetic risk for complex diseases, but that will change with advent of whole-genome sequencing. Soon we will be able to tell people the net genetic risk of disease X, and that will have more meaningful impact. I predict that will have more impact on behavior as well.”
About a quarter of the survey respondents reported sharing their tests results with their doctors, and this subset did make healthier exercise and diet choices than did the group as a whole. However, it’s unclear if this group was more motivated to get healthy to start with, prompting them to visit their physicians, or if the doctor’s consultation triggered the changes.
The findings show how important it is to have physicians involved in the process, says Vance Vanier, the president of Navigenics. Unlike other tests, the one from Navigenics is available only through physicians or corporate-wellness programs, rather than directly online. However, most physicians are ill equipped to deal with the results of genetic testing. “Surveys show 90 percent of physicians don’t feel comfortable using genetic data to guide therapy,” says Topol.
Only 10 percent of respondents availed themselves of free genetic counseling, which was offered together with the test. Unlike general practitioners, genetic counselors are specially trained to help patients interpret and use genetic tests. Vanier says Navigenics has changed its approach to counseling as a result of these findings. The company now actively follows up with users to offer counseling, rather than waiting for them to call in.
Given the state of the health-care system, one of the greatest concerns about direct-to-consumer tests has been that they would tax the system. However, people did not report undergoing more medical procedures as a result of the tests, a finding that is confirmed by the Coriell study. “That suggests that concerns about overuse of the medical system may be premature,” says Erynn Gordon, director of genetic counseling at Coriell. “I don’t think anyone has enough data at this point to say the question can be put to rest. But there’s a lot more concern expressed than evidence to support that it’s a problem.”
People predicted to be at higher risk of certain cancers did report higher intent to get follow-up screening tests, such as colonoscopies and mammograms. However, it’s still unclear whether they will follow through on their intentions. Topol says his team continues to follow participants and will try to answer this question.
Both Topol’s team and the Coriell group now plan to look at the effects of pharmacogenetics information, genetic mutations that predict how well certain drugs, such as Plavix, will work in an individual, or how susceptible that person will be to side effects. Because this information has more obvious consequences—physicians may choose a different drug or prescribe lower doses—these tests may have a greater impact than ones predicting genetic risk of disease. “People ought to take that information more seriously now,” says Christman. “For example, Plavix, the second most prescribed drug in the world, doesn’t work for a quarter of the people who take it.”
This new data poisoning tool lets artists fight back against generative AI
The tool, called Nightshade, messes up training data in ways that could cause serious damage to image-generating AI models.
Rogue superintelligence and merging with machines: Inside the mind of OpenAI’s chief scientist
An exclusive conversation with Ilya Sutskever on his fears for the future of AI and why they’ve made him change the focus of his life’s work.
Data analytics reveal real business value
Sophisticated analytics tools mine insights from data, optimizing operational processes across the enterprise.
How to fix the internet
If we want online discourse to improve, we need to move beyond the big platforms.
Get the latest updates from
MIT Technology Review
Discover special offers, top stories, upcoming events, and more.