I’m a sucker for information. So when I heard that 23andMe could provide personal genetic reports for a mere $99 (plus a requisite $9/month per year for updates on discoveries made about your DNA over time), I was all over it. Who wouldn’t want to exchange a saliva sample and a couple hundred bucks for a personalized DNA analysis on health and ancestry?

Based in Mountain View, California, 23andMe was named “Invention of the Year” by Time magazine in 2008. The company’s been featured on Oprah. It’s been in the New York Times. And I’ve personally heard it mentioned everywhere from medical conferences to fitness message boards.

The Details

23andMe provides more information to incorporate into one’s understanding of risk of disease. But what do you actually get in a report? It’s divided into several sections. The health section includes disease risks, carrier status, drug response, traits, health labs, and family health history. Disease risks show your percentage of risk compared to average: elevated, decreased or typical risk. There’s a four-star confidence system—four stars are the highest level of confidence, indicating more established research: multiple studies with 750+ participants. In many instances, there is even a percentage of average risk compared to your individual percentage of risk. You can click to get more information (and in some cases, reports) on each disease, and a snippet on whether you’re at higher or lower risk. In my case, there were no surprises in the diseases I was at higher risk for based on my knowledge of family history… but again, they are just increments. (For example, if the average risk of a disease is 15%, and you’re at 25%, then you’ll show ‘elevated risk’—but an increased risk is certainly not necessarily a determinant.)

The carrier status information is interesting, especially for those planning on having children. It’ll show you if you have a genetic variant of any of a list of diseases. You may have other variants not detectable by 23andMe, and the presence of a variant does not necessarily mean you are affected by the condition or will even necessarily pass it on. A genetic counselor might be a good resource for those concerned with results here.

Next on the report is your drug response. This provides lots of information for curiosity’s sake, but not all of it is practical. How would you respond to Abacavir if you had HIV and were prescribed it? What are your odds of myopathy while on statin therapy? There are huge implications for personalized medicine—imagine a doctor not prescribing a drug or even changing dosage based on your genetic profile—but a lot of items here aren’t really all that useful. Example: I don’t have Hepatitis C, so knowing that I’d have a typical response to treatment doesn’t really have any impact on my life. And even though preliminary research for one reported marker indicates that my genotype does not have higher-than-average odds of heroin addiction, I don’t intend to shoot up anytime soon to find out. I did find it interesting to learn that I am a slow metabolizer of caffeine, and that (again, based on preliminary research for a single reported marker) drinking coffee could increase my heart attack risk. This wasn’t a huge surprise for me, though. I save coffee for emergencies because a single cup may keep me up all night.

The next section was traits, which ranges from the obvious (eye color, hair curl, earwax type) to the bizarre (whether you’d be likely to smell asparagus in your urine), but also has some interesting tidbits—whether you are likely to be lactose intolerant, whether you have the gene ACTN3 (which makes you more likely to be a sprinter at the Olympic level), and some information on memory, odor detection and the effect of monounsaturated fats in your diet. One study in particular got my attention, and that was on the ability to learn from past mistakes. “Learning involves a range of behaviors and skills. These include developing a tendency to prefer rewarded (positively reinforced) choices while avoiding those that are either punished or negatively reinforced. Past studies have shown that the neurotransmitter dopamine is involved in such trial-and-error learning; that means variation in genes related to dopamine signaling may also affect a person’s ability to learn.” Apparently, a 2007 study showed that people with my genotype did not learn as well to avoid negatively reinforced choices. Of course, this trait only had one star on the confidence scale, since it was based on a single study with only 26 people, but I did experience a momentary panic until I remembered that I’ve demonstrated my ability to learn from past mistakes—despite my genotype.

“We only include [1 star reports] because they’re kind of fun for our readers. We don’t encourage them to take them seriously,” said Joanna Mountain, Senior Director of Research for 23andMe. “As far as we know, no other group has replicated that finding, so it’s a very low-confidence report,” more for amusement than for drawing solid conclusions, she said. Whew.

In the “health labs” section, you can also find out your probable blood type, heart attack risk (if you know your cholesterol and blood pressure) and how much weight you can blame on your genes (just over 6 pounds for me.)

Then there’s the ancestral information. I could not trace my paternal line since I was not gifted with a Y-chromosome, but I was able to view the geographic distribution of my maternal line, a haplogroup that traces originally to Africa but has also been commonly found among Ashkenazi Jews in Europe. Upon ordering my profile, I was inundated with requests from “potential relatives” wanting to make contact—but sharing a maternal haplogroup isn’t enough of an incentive for me to want to talk to strangers, so I ignored them—but your mileage may vary.

The Good and the Bad

There have been some criticisms of 23andMe, one from our very own Robb Wolf. “23andMe is the dumbest application of a smart idea I can think of,” he says. “Genetic screening played across standard epidemiological risk factoring. Where the heck is the evolutionary perspective? I think if you put some EV-bio into the risk analysis and recommended implementation, you'd have something much more powerful.”

And then there’s the fact that genetic information is, of course, not the only factor to contribute to the development of disease—so a report could be misleading. It is difficult to determine what one’s true level of risk really is, and what does percentage of risk really mean, anyway? We’re talking about teeny tiny increments or decrements in relative risk of diseases—how useful is that, really? Perhaps over time, more predictive algorithms will be developed between genomic and phenotypic data, but for now, it seems clear that services such as the one 23andMe provides are mere educational tools, as they’re not necessarily useful clinical
tools for evaluating individual risk.

Whether or not genetic testing companies provide educational services or diagnostic tools is a matter of legal debate as well, as the latter would require tests to be ordered by physicians in many states. And it’s unclear how a physician would actually use the information, or how much more helpful it would be than family health history (if available). 23andMe provides a database environment to help you visualize your data, kind of like Mint.com does for financial information. It takes literature or published information gathered by researchers, associates it with the genomic data of clients, and puts it in a format that’s easier to understand. So basically, it’s useful to geek out on—so long as you’re aware that even if you do not carry a genetic predisposition for a certain disease, it only means there’s no known, testable genetic risk for it.

Critics of 23andMe have asserted that knowing one’s genetic information could be anxiety-producing, leading to unnecessary testing, screening, and even medical procedures. Proponents have said that information is power—providing it directly to consumers would lead to healthier lifestyle choices and perhaps more compliance with health screening practices. Interestingly enough, a study in the New England Journal of Medicine showed that neither of these things seemed to be true— “…we found no short-term changes in psychological health, diet and exercise behavior, or use of screening tests,” they concluded, adding that “potential effects of this type of genetic testing on the population at large are not known.”

Lifestyle Changes?

Did I make any changes because of my results? Knowing that I’m a slow metabolizer of caffeine has made me hesitate more before ordering that coffee drink, and I’m interested in specifically exploring the effects that lactose has on my body more closely. But for the most part, my results did little other than to satiate my curiosity.

Was it worth it? I believe it’d be very useful for someone without access to family health history, and I find the information fascinating to geek out on: navel gazing with a scientific flavor. And it’s even possible to share results with other people, which could provide many more hours of entertainment (if they’re willing to dish out $200+ for it) and help put one’s own information in context.

But I agree with Robb’s criticism—my report indicated that I was at low risk for celiac disease, for instance, so if I simply relied on a 23AndMe profile to decide whether to avoid grains, I would be doing myself a disservice—much in the same vein as someone who gets an allergy profile rather than trying an elimination diet. At the same time, I found the information in the report interesting for educational purposes. As Krista Scott-Dixon, of stumptuous.com fame, pointed out, you shouldn’t be afraid to know what’s in your DNA, especially since the worst gene-determined diseases “will take you out before birth or shortly afterwards.” This makes a higher percentage of relative risk a possibility, not a destiny—and even if it were a very probable destiny, knowledge is powerful. As Krista says, “…even if you discover you have only a few short years left, wouldn't you want to know that? Then you can book that trip to Tuscany or finally ask that girl/boy to dance—you have nothing to lose! If you know you have an elevated risk—and remember, it's rarely an inevitability—then you can see it as a wake-up call to improve your health and approach to life wherever possible.” (23andMe also requires an extra step to unlock sensitive information, so you can always avoid certain areas and access others.)

Krista didn’t make any changes based on her results either, but did gain “some interesting context and a few surprises.” It’s always fascinating to learn that genetically, you have a lower risk for a disease that affected many people in your family (based on the research currently available), for example, or perhaps you have a high risk for something completely off your radar based on your family history. Or you can have a lower than average risk of a disease where the average is very high, so looking at the information out of context might be misleading. We all know that health is an intricate web of genetic variability and environmental factors. 23andMe just helps provide some more personal context for it.