"I personally started getting very frustrated doing all this research when there was nothing waiting on the other side of the fence waiting to implement it," says Stephan, the Navigenics co-founder.
Get developments in medicine, nutrition and fitness delivered to your inbox with our The Health Report newsletter. Sign up »
Now, with the advent of an enormously successful research tool called the genotyping chip, it is finally inexpensive enough to offer these scans to the public (see accompanying story).
Navigenics, 23andMe and deCODEme all use such chips to analyze DNA in spit or cheek swab samples sent in by customers, using them to detect as many as 1 million DNA variations known as SNPs: single nucleotide polymorphisms.
Very few SNPs have been associated with disease, and even fewer still reliably so. But as science progresses and links more SNPs to diseases, the companies aim to update their databases.
When 23andMe first launched, it offered information on about 20 diseases and traits (including earwax type and bitter-taste perception). Today, that number is 58 (deCODEme offers 26 and Navigenics 18 -- 17 each for men and women, including prostate and breast cancer respectively).
for On 23andMe's website, each disease is listed as either "Established Research" -- findings generally confirmed in at least two studies of 1,000 people or more -- or "Preliminary Research," findings that have yet to be repeated and validated.
"We want to buck the paternalistic attitude," 23andMe co-founder Linda Avey says of the company's rapid addition of new conditions even when the science is tentative. "And if anything, what we are hearing back from our customers is 'We want more.' "
In addition, 23andMe and deCODEme give their customers the opportunity to search their raw results -- all those hundreds of thousands of SNPs, and check them for the latest genetics result. So when a new study is published suggesting that smokers who carry two DNA letter "T"s at position "rs1051730" in the human genome tend to smoke more cigarettes and get more lung cancer -- as was published April 3 in the journal Nature -- customers can head straight to their accounts, type in "rs1051730" and see where they fall.
Navigenics has taken a more conservative approach. If, for example, I want to get my entire genome scan, which I certainly do, I must sign a waiver acknowledging that any of the SNP-disease associations I might Google myself are likely to be spurious. Only then will the company send me the file on an encrypted disk.
"We will take responsibility for the information we believe is real," Stephan says. "For everything else, you can have your genome to play with. But without our blessing, the potential for misinformation is extremely high."
It might not be smart, but I just can't help myself. I want to know everything.
For the conditions with full reports, there are ample resources to help explore results. All three companies' websites are packed with explanatory graphics, background information, the biology of diseases, links to disease associations and expert medical opinions. At every turn, the sites try to make clear that risk profiles are estimates, not diagnoses. "It is really complicated information in a lot of ways," 23andMe co-founder Avey says. "That's our challenge. How do we distill this information down and write it in language that everyday people can understand?"
The Navigenics site emphasizes strategies to reduce your risk and offers free, unlimited access to genetic counselors to make sure that people understand the ocean of numbers in front of them. Also, there are tips for talking to your doctor and advice for taking the printouts to an appointment.
My DNA turned up a heightened risk of glaucoma -- almost three times the average. As it happens, I just got a reminder card in the mail to see my optometrist for a check-up. Even though genetics plays a very small role in development of the disease, I'll be sure to let him know, so he can poke my eye and check for increased pressure. Why not?
Do scans help or harm?
But no amount of transparency, caveats, tutorials or access to genetic counselors can satisfy the concerns of many critics: bioethicists, doctors and health economists who worry that the newness and unreliability of these genetic data make the services, at best, a waste of money and, at worst, a threat to public health.
For one thing, Khoury says, despite the high-tech wizardry of genome scans, most of the advice available so far is just common health sense: Don't smoke, do exercise, eat a healthy diet, maintain a healthy weight, keep your blood pressure and cholesterol in check, and go for your regularly scheduled tests. What's more, most of the risks revealed by these tests are pretty small. The really potent genetic effects are probably obvious from family history alone.
"If someone asks me, shall I take [the test]," Khoury says, "I would say 'No, you don't need it. Tell me about your family history. If you have a family history of skin cancer, stay away from the sun. If you have osteoporosis, take calcium supplements.' We know what to do."