Personal genetic testing, a growing private industry, may not be as reliable as advertised for determining a person's predisposition to common diseases, according to a recent study by Dartmouth Medical School professor Jason Moore and Vanderbilt University professor Scott Williams.
The pair found in a second study, however, that the technique may be useful in gathering ancestral information.
The researchers published their findings in the Sept. 11 issue of the American Journal of Human Genetics.
Companies that provide genetic testing typically offer mail-order kits available on the Internet that claim to measure a person's susceptibility to common diseases like Type-2 diabetes, Moore said. One of them, 23andme.com, advertises a kit that offers to analyze a person's risk for more than 116 diseases for $99, according to its web site.
"[The web site is] very nice and very flashy," Moore, who used the kit himself, said in an interview with The Dartmouth. He added later, "[The test] really doesn't tell you anything about your risk for one particular disease."
The problem lies not with the kits themselves, he said, but with the scientific research on which they are grounded.
The test provides "quite a bit of genetic information" based on a user's sample, he said, analyzing one million variations in the human genome. But, according to Moore, each single variation has a very small effect on a person's risk of disease. Instead of separately examining the variations, he said, the genes must be looked at holistically, which is not currently possible.
"The problem is with the information, not with the tests," he said. "It's an exciting technology, and I think someday it will be very useful."
Testing for some diseases, including cystic fibrosis and Huntington's disease, is more reliable, since those maladies are more directly affected by a single variation in the genome, Moore said.
Personal genetic testing is controversial, as detractors argue that groups like health insurance companies and employers could use its results for discriminatory purposes. The Genetic Information Nondiscrimination Act, signed into law in May 2008 by then-President George W. Bush, prohibits this type of discrimination based on genetic information.
The second study, by Moore, Williams, Chantel Sloan GR '10 and others, examined the effectiveness of personal genetic testing in the context of analyzing ancestral history. The researchers looked at more than 1,000 genetic markers from over 800 people in New Hampshire, Sloan said in an interview with The Dartmouth.
"What we found was that, using a relatively small number of genetic markers, we found that there were genetic subgroups within the population and that these subgroups were correlated with the ancestry that their parents and grandparents had," Sloan said.
Those subgroups can, in turn, be matched with the characteristics of different populations to determine geographical information about a person's ancestors, Sloan said.
While both types of testing involve examining genetic information from DNA samples, Moore said that using the technique to examine ancestral information is more practical than disease screening.
"That's an example where I think personal genetic testing can be useful," he said.
Studies examining genetic testing and how it relates to ancestral information have been conducted before, Sloan said, but this was the first analysis that involved such a large sample of a population within the United States.
The study may also have broader implications beyond determining ancestry, Sloan said. Eventually, she said, the hope is to complete the "triangle" connecting ancestry, genetics and health.
That remains a future goal for now, Moore said.
"This is an example of how personal genetic data can be used to help inform people of their ancestry," Moore said in a College press release. "Informing people of their future health is still out of reach, though."
No contact information was available for 23andme.com, and calls to a similar web site, decodeme.com, were not returned by press time.
