By Turna Ray
A study published this week in the New England Journal of Medicine found that genomic data did not lead to any short-term changes in people's psychological health, diet, or exercise behavior.
In the first large-scale behavioral analysis of early adopters of direct-to-consumer genome-wide profiling, researchers from the Scripps Translational Research Institute surveyed 2,037 people who were screened for their genetic predisposition for 23 diseases and conditions through Navigenics' Health Compass. Study participants were then asked to complete questionnaires about their level of stress, exercise, diet, and the use of screening tests after receiving test results.
The authors, led by Eric Topol, director of the Scripps Translational Science Institute, wrote that the study's findings "support the null hypothesis (that provision of the results of a direct-to-consumer genomic risk test does not affect health-related behavior)," though they note that "the potential effects on the population at large are still unknown."
The study revealed that "with regard to what they could do for themselves, which is exercise, they didn't want any part of that," Topol told PGx Reporter this week. "That's really sobering and a downer that we didn’t see any change in that."
The study was launched two years ago under a collaboration between the Scripps Translational Science Institute, Navigenics, Affymetrix, and Microsoft. The original plan was to enroll 10,000 patients and follow them for 20 years to track the behavioral impact of DTC genomic testing (PGx Reporter 10/17/08).
The study was funded by the National Institutes of Health and Scripps Genomic Medicine. "Navigenics did not provide any financial support for the study and was not involved in the study design, the accrual or analysis of data," or the publication of the study, the study authors disclosed.
Primary analysis of the study cohort revealed no significant changes in levels of anxiety, dietary fat intake, or exercise after receiving genetic test results. Secondary analysis showed that test-related distress went up or down in line with the average estimated lifetime risk of the 23 conditions, but 90.3 percent of study participants reported "no test-related distress."
Topol had hoped that some people might make some initial changes if they learned that they were at greater genetic risk for certain diseases that are preventable through better diet and more exercise, such as heart disease and diabetes. "I thought maybe we would see a change for the first few months, the first six months, and then it would go back to the old ways," Topol reflected. "But we didn't even see a hint [of a change in behavior] in the early part of the survey, and this just reflects previously reported data that it's very hard to change lifestyle."
Even though it was evident that genomic data did not encourage behavioral improvements in diet and exercise, Topol said that there were some positive outcomes from the study.
"It provided evidence that genetic testing data of this kind doesn't induce psychological stress or anxiety. That's good, and no one really knew that before," he said. In the years since the launch of the consumer genomics industry three years ago, critics of the DTC model have feared that giving people unvalidated information about their predisposition for diseases and conditions would cause unnecessary anxiety and lead to rash medical and surgical decisions.
The Scripps study does suggest that access to genomic information could potentially impact people's willingness to get follow-up screening for conditions for which they are at higher risk.
However, as the authors note in the NEJM paper, even though half of the study subjects indicated they planned to undergo one or more screening tests more frequently in the future, "the likelihood that reported intentions will translate into actual increases in use is extremely low." This is characterized as "a good thing" by the study authors, since the "majority of the screening tests we assessed are considered inappropriate for asymptomatic persons," and "would probably result in a waste of healthcare resources."
Most study participants — which comprised employees at health and technology companies, as well as their friends and family — had to pay $200 to participate in the study. The companies participating in the study were Scripps, Life Technologies, the energy firm Sempra, and the telecommunications company QualComm.
Of these firms, Sempra decided to cover the full cost of genotyping and genetic counseling for 1,000 of its employees — a decision that ultimately impacted the results of the study because the Scripps researchers found that those who received the test for free were less likely to complete the survey.
When Sempra decided to cover the cost of testing, "it was like a stampede to donate saliva," Topol said. Unfortunately, "there was a trend that the people who got it for free were less like to complete the survey," Topol noted. "The people who paid for testing were more invested."
At the start of the study, participants were given health assessments and received information about what genetic testing can tell them about their health as part of the informed consent process. After participants received their 90-page genome profiling report from Navigenics, investigators followed up with them at three months and then at 12 months. The NEJM paper reports the researchers' findings at the three-month follow-up.
Since the survey was conducted in a self-selected population who wanted to be genetically tested, the findings likely mirror the behavioral patterns of the types of people who are current customers of DTC genomics services. However, the study authors note that "the potential effects of this type of genetic testing on the population at large are not known."
Even without this information, Topol doubts that randomly selected subjects would be more willing to change their behavior based on probabilistic genomic data regarding their health.
"I would guess that no matter how broad the population is we wouldn't see the change in lifestyle that we would hope to evoke," Topol said. "That's going to probably wind up being more disappointing than we had hoped."
From an educational standpoint, those who received testing through DTC firms received a primer on genomics, which is one positive aspect of the services provided by companies like Navigenics and 23andMe.
"These people benefited a lot by learning about genomics in general" through this process, Topol said. "So, it's an anchoring experience. [Initially], almost none of these people knew anything about this stuff, and as a result of being in this study they got oriented about what they are learning and what they're not learning.
"That was a good thing as far as the democratization of DNA," he said.
Topol said that Scripps will no longer enroll 10,000 people as was originally planned because the time commitment from volunteers and the cost are too great for enrolling such a large cohort. However, he and his colleagues will continue to follow the current study population to see whether ancestry and pharmacogenomics data impacts behavior.
Ultimately, Topol's goal is to study how whole-genome sequencing data is received by consumers. However, that study will only be feasible once the cost of sequencing and annotating a whole human genome drops far below what it is today.
While sequencing costs are plummeting quickly and many in the industry believe a human genome will be able to be sequenced for $1,000 within the next year, whole-genome clinical annotation is still a complex and labor-intensive process. Currently, Complete Genomics and Illumina offer whole-human genome sequencing for researchers who order in "bulk" for less than $10,000 per genome, but these firms leave the final analysis to their customers. Knome, meantime, offers a sequencing service for individuals that is priced at $68,500. The service includes information about the customer's genetic predisposition for various diseases, and includes interactions with geneticists, clinicians, and bioinformaticians.
Have topics you'd like to see covered in Pharmacogenomics Reporter? Contact the editor at tray [at] genomeweb [.] com.