I HESITATED, then I watched my friends go for it, and finally I shelled out the $200 for 23andMe genetic testing — more curious about my health risks, less about my ancestry. 23andMe sent the kit, and I spat into the tube and mailed it back. Slightly worried that nothing would happen, I waited several weeks for the promised results.
Since its launch in 2007, 23andMe has had an uneven run, starting with test results that included risk factors for everything from obesity and alcoholism to breast cancer and Parkinson’s, along with guidance on how to change negative outcomes. In 2013, charging lack of evidence, the FDA put a halt to the risk analyses and left only the ancestry break-down.
23andMe relaunched in the fall of 2015 with health-risk analyses severely curtailed. For medical information, the FDA allowed only 36 “carrier status” reports, all of which in my case came up “no variant detected,” meaning I was not carrying anything from cystic fibrosis to familial dysautonomia. Not very useful.
But here’s the genius of 23andMe: It’s like a high-tech horoscope with the same appeal — it’s all about ME!!!
And further analysis of your sample is plentiful from other sites that use raw data from 23andMe. For an additional $5, I chose Promethease, which provided results in five minutes. When downloaded and copied into a Word document, they filled 290 pages: too much information. I pulled back, re-sorted for fewer conclusions and have since spent more hours than I like to admit deep-diving the mounds of information.
DNA test results are based on genetic variations — the differences among people — also called “single nucleotide polymorphisms” (SNPs). Promethease ranks your SNPs by magnitude (how much interest each SNP has garnered from the “community” of other Promethease users) and frequency in your population, which for me is Caucasian/European. Promethease results include paragraphs on each SNP, with links to SNPedia (a DNA Wikipedia), where more links lead to published research, and on and on.
Luckily, Promethease also provides colorful borders around each SNP description — red for “bad,” green for “good” and gray for “not set” — and, for each sorting of SNPs, creates a corresponding colored pie chart. Despite my overwhelming total of 19,489 SNPs, my pie looks pleasingly green and gray with just a sliver, 2%, of red.
Genetic test results come with plenty of caveats. Having a variant indicates merely an increased or decreased risk that you could have a specific health issue, and not that you do or will definitely have it. Also, a variant means the specific gene is working at an efficiency level that is altered but not necessarily better or worse. Finally, whether and how much a genetic variant is ultimately expressed depends on epigenetic (lifestyle) variables such as diet, exercise and environment. Many experts believe genetic test conclusions are too poorly substantiated to be of much value. So, maybe not any more reliable than playing the horses.
Currently, the best use of DNA testing may be to maximize the effects of specific medications based on an individual’s rates of metabolism and response for that drug. The indication by one SNP that aspirin can reduce my risk of colon cancer was reassuring, especially when my doctor was not supportive.
Although my recent MyLittleBird article, “The Ideal Diet for your DNA,” cites several experts who tailor individual diet and exercise regimens to DNA profiles, my Promethease diet and exercise results come with the caveat “interesting hypothesis but not well-validated.” Nonetheless, I was interested to find that better weight loss for me could be achieved with high-energy exercise and low-fat diets. (NutraHacker offers more extensive diet information, but at a higher price tag of about $20 for each report on different dietary issues, which could add up quickly.)
The few results now permitted for 23andMe can seem silly, obvious and/or wrong. Among my “physical responses,” for example, 23andMe gave me an only average tendency to sneeze after eating dark chocolate, based on my Neanderthal ancestry, but corroborated my photic sneezing (in response to sunlight). Also well known to me, I have a 75% chance of being able to detect the “funny smell in urine” after eating asparagus.
In the category of wrong conclusions, I am 72% more likely to have wavy, slightly wavy or straight hair — don’t I wish? In fact it’s very frizzy. Also my muscle composition makes me more likely a sprinter than an endurance athlete: not true.
On the other hand, I was intrigued to find that I likely have more “periodic limb movements” per hour of sleep compared to the average of about 10. I do not have the variant linked to “deep sleep,” making my sleep typically but not excessively deep. Also, based on a SNP for a protein that transports fatty substances, I have a 93% chance of wet earwax (compared to dry and flaky) that, more important, is linked to less body odor.
On caffeine, 23andMe found my consumption 37% higher than those who are genetically similar to me, while Promethease labeled me a “fast metabolizer” of coffee — which decreases its effect — but on the other hand, to have “significantly higher anxiety levels after moderate caffeine consumption.” Both sites agree that I have a high likelihood of not tasting certain bitter compounds — in coffee as well as in vegetables like broccoli or Brussels sprouts.
On Promethease, I have a “stronger bones” SNP as well as the conflicting SNP for “increased risk of low bone mineral density disorders” — but my pleasing pie chart showed six green good results vs. only five red bad ones. Most surprising was the long list of SNPs indicating my risk for Type 2 diabetes, which has appeared nowhere in my family but could be affected by my sugar consumption over the years — though that risk pie chart actually looked quite green as well.
Of my two favorite Promethease results, both marked bright green, one indicates “better avoidance of errors,” “women have higher persistence” and “lower obesity due to increased pleasure response to food.” The other is “[worrier] advantage in memory and attention tasks…more difficult to hypnotize…placebo is more effective…more efficient at processing information,” also “enhanced vulnerability to stress,” based on what’s known as the “warrior/worrier hypothesis,” which I intend to investigate further.
What is worrisome to many is the privacy issues of genetic testing sites. 23andMe makes no bones about being out to collect personal information: With a current database of more than one million customers, it is noted that each additional sample helps improve everyone’s results. Among more private alternatives, the Genophen platform can be ordered by private physicians through Base Health at $130 and is reputed to give better information on the interaction of an individual’s genotype with diet, exercise, lab results and family history.
For some, though, genetic testing sites provide anonymity — I am not obliged to discuss the findings with anyone, not even my doctor — along with links for endless ferreting throughout the Web, time to reflect on the results, and experts at the ready for questions: When I emailed the site to challenge my high-risk SNP for male-pattern baldness, the answer came quickly, the error acknowledged.