At the advice of many medical experts, I’m leaving the following article, in which I’ll discuss my personal probabilities of disease based upon my genetics, unsigned.
While the 2008 Genetic Information Nondiscrimination Act (GINA) protects individuals from being denied health insurance or employment based upon their genetic code, health insurers can still weigh your genetics against your payment determinations. Plus, the implications to life insurance, disability insurance, long-term care insurance and even the fate of my other family members are yet unknown.
I’m particularly worried about being denied long-term care, as my test results chillingly indicate that I could develop a devastating degenerative disease.
This point of anonymity aside, I intend to share with you the most intimate details of myself—namely, all the forces that might one day kill me (if I’m lucky enough not to wipe myself out before from risky behaviors or sheer clumsiness).
* * * * * *
Several weeks back, I mailed off two vials of my spit. One went to a company called 23andMe, the other to a company named Navigenics. Aside from some differences in branding, both companies are cut from the same mold. Using a specialized chip that’s set to track only specific genes, each examines an infinitesimal portion of your genome, yes, but hopefully one large enough to take advantage of what the world’s researchers have decoded (of importance) regarding human genetics.
(Given that the costs of mapping your entire genome still runs in the six figures, looking at this snippet of code allows 23andMe and Navigenics to keep costs to $420 and $1000, respectively.)
Each sets you up with a login to their site; each offers a handy UI to examine yourself; each calculates the probability of you developing diseases overall and in comparison to your peers; each brings you to the relevant research backing their claims with a few button presses.
For most diseases, it’s not as simple as, “Oh, you’ll have diabetes!” Instead, users are left with something more like, “You’re considered high risk for developing type 2 diabetes. You’re in the top 30% of the population most prone to this risk. Your overall risk is 2%. This data is based upon this paper by Harvard Medical.”
In other words, it’s a game of percentages stacked on percentages, all various levels of true based upon research, all confounding.
So now that you understand how these services work, let’s move on.
This is me—the me according to 23andMe.
I don’t expect you to know what all this means just by looking at it, so allow me to explain a few of the highlights.
On the top is 23andMe’s clinical report—the items for which they have a significant amount of research. They’re worried about a few things. I’m at “elevated risk” for Crohn’s Disease—there’s just a 1.5% chance that I’ll develop this GI tract disorder, but it’s considered elevated risk because, based upon the combination of 12 related markers (or genes) that I have, I’m about 3X more likely to develop Crohn’s than the average person of European descent.
The numbers are similar for Type 1 Diabetes, my only other elevated risk in 23andMe’s clinical reports. Overall, my risk of developing the disorder sometime in my life is 1.3%. For most of the population, it’s 1%.
The second report 23andMe offers is my research report—the items for which the service is reporting results based upon what they’ve deemed less reliable research for one reason or another. I have many more elevated risks in this category than clinical. Ulcerative colitis (great, more butt disease), hypertension, asthma—nothing you want to see, but none of my numbers seem that scary unto themselves.
Heroin addiction. Wow. I’m at elevated risk on that one, too. I’ve stayed clear of the stuff, and given my propensity to its dependence, I’ll be passing on its casual use at parties and clubs. We’ve learned from a study of 309 Swedes and their specific OPRM1 genotype (which seems to affect how the brain responds to morphine), I’m 2.9X more likely to become addicted to heroin than the baseline person.
But so far, so good. 23andMe says I’m a pretty freaking healthy guy, which confirms my lifelong suspicions. I haven’t developed symptoms of Crohn’s, which generally manifests in people around my age group. Chances are, I’m in the lucky 99% there. Type 1 diabetes, asthma, hypertension, ulcerative colitis? Right now, I look to be in the clear. Heroin? I can stick to poppy seed bagels and consider myself a rebel from the ingestion of carbs alone.
But one other high risk factor has haunted me since reading the report. Alcohol dependence.
I read that warning—conveniently highlighted in red for me—and considered the last day I’d gone without having not only one drink, but generally two or more.
The Journal of the American Medical Association defines alcoholism as:
A primary, chronic disease with genetic, psychosocial, and environmental factors influencing its development and manifestations. The disease is often progressive and fatal. It is characterized by continuous or periodic: impaired control over drinking, preoccupation with the drug alcohol, use of alcohol despite adverse consequences, and distortions in thinking, most notably denial .
Does my behavior fit in that criteria of alcoholism? I definitely fall short of its most dire interpretations.
Alcohol has by no means destroyed the relationships of my life, but things I’ve said (drunk) have damaged a few friendships—which I guess fulfill the category of “adverse consequences.” Given that I can be a bit stoic in person, I appreciate its function as a social lubricant. Given that I can be a bit stressed at the end of a workday, I appreciate its use as an over the counter anxiety medication—both of which may qualify as “preoccupation(s)” with the substance.
Alcoholic or not, there’s no doubt that I drink frequently and rely on the substance on more than a casual basis. I knew that before my 23andme results, and I was pretty OK with myself about it.
But something about seeing that trait, one that I formerly snickered at as voluntarily self-destructive, mapped to a specific gene…well, the whole practice stopped feeling so adolescently charming, yet I haven’t curbed my drinking one bit.
Then, after all this introspection, I looked at my results further, checking the actual genotypes I had in question. According to the only results listed by 23andMe, I was at normal risk.
Bamboozled! I’ve searched through every piece of data 23andMe has on me, and I can’t find what caused what I’m assuming was an error on my research report. I made this mistake, and I’m college educated, scientifically inclined and, heck, pretty experienced (at this point) at reading the results of my genetic tests compared to the average person. (UPDATE: 23andMe contacted us, admitting the formatting error in alcoholism and sharing plans for a fix.)
And the great news is, no matter how much I drink now, my genome in hand, I know I’m not dependent!
So that’s the 23andMe me—a guy with stomach problems, diabetes, propensity for substance abuse and “wet” earwax.
This is the Navigenics me:
Navigenics lays out my biggest risks by raw percentage, which is a bit different than how 23andMe presents their data (as well as more clear). Get passed that, and you’ll note some immediate similarities. Crohn’s Disease. Is that double trouble? Two different groups pointed it out!
No. It’s just an overlap in their testing. Navigenics cited the same study with the same 1% overall probability of developing the disease.
Lactose Intolerance. 95% chance. (23andMe actually found this, too.) That’s shocking to me because I suck down more milk and cheese than most cows. A solid 6 inches taller than the rest of my family, I’m basically a walking “milk does a body good” commercial. Screw soy milk. I want to suck on a udder with one side of my mouth while shoving fresh cheese curds into the other.
I’m not lactose intolerant, which brings us to the huge problem with these odds numbers. Human beings generally read a 95% chance—or anything over 85%, really—as a certainty, as 100%. Conversely, you’d read anything below 15% as a 0% chance. Getting the brain to appreciate the true meanings of statistics is part of the challenge here.
Of course, maybe I am lactose intolerant, or at least, I would be if I ever stopped consuming dairy or didn’t so regularly eat probiotic yogurt. Maybe the fact that I was breastfed helped turn the tables. Maybe the science isn’t as simple as we think. Maybe I’m part Asian—my genotype nearly guarantees Europeans lactose intolerance while Asians carrying it can be in the clear. (And 23andMe tracks my maternal ancestry to geography that reaches into central Asia.)
But none of these theories are all that comforting facing what is, for me, the most frightening component of my results.
It’s Alzheimer’s. Or it could be. I don’t know.
There are three specific APOE markers for Alzheimer’s that mankind has unraveled so far. You must have two (since you receive one from mom, one from dad). I have one ApoE3 and one ApoE4. (Navigenics reports them in your results while 23andMedoes not.)
The ApoE3? Not a big deal. You see, you can only possess ApoE2, 3 or 4 alleles (somehow arranged in a pair). An ApoE2 allele would put me at low risk for Alzheimer’s. I don’t have any E2s. The E3 puts me at normal risk for Alzheimer’s. The E4 that complements my E3? That’s the bad one.
The presence of these markers puts me at a 20% risk for developing Alzheimer’s sometime in my lifetime. In comparison to the rest of the population, I’m in the highest 26%. (At least according to the sample population that’s the basis of this particular piece of research, “60 Americans of European ancestry.”
My prognosis could be worse, as my wife, who knows a bit on the topic for reasons I won’t divulge, explained to me. We were in the middle of a long weekend dinner at home involving too much wine and lots of laughter. It was one of those magical yet everyday evenings you might see on an overzealous commercial for furniture or carpeting. Then, out of nowhere, the entire atmosphere changed.
“When you told me you were taking these tests, you never told me they’d be testing for the APOE-4 gene,” she said. “You just don’t realize, do you? If you had two of those, it would have been bad. Really bad. Like, we pick up and change the way we live our lives from here on out, bad.”
The most devastating thing I learned after two DNA samples was that I’d approached this entire experiment as naively as a child.
* * * * * *
The irony of it all, of course, is that aside from an obvious propensity for melodrama, my greatest overall risks are for heart disease, obesity and osteoarthritis—something Navigenics is clever to point out whereas 23andMe skirts the issue—which is also what any doctor would tell me without a physical exam, based upon real world statistics. (Based upon two risk markers, I’m 10% less likely to have a heart attack than the average bear, but there’s still a 32% risk that it will happen some time during my life…possibly ending it.)
In other words, as frightened as I am of the unknown secrets lurking deep within my body’s blueprints, we already know what we should be most worried about:
Heart disease is the number one killer. Infectious disease is the number two killer. Cancer is the number three killer. (2008)
But none of this is of much comfort when it comes to losing your mind to the perplexing games of corporeal probability.
Not to pick on anyone, but 23andMe markets genetics as social networking, a place to be, a hip fun thing you should share. Hey, even host a “spit party”-–your friends, some of whom are about to receive devastating news—will love it!
Aside from the fact that sharing your genetic information could cause so many massive repercussions to your insurance coverage—and still, you’re encouraged to share your results with a community—your genetics are not a fucking novelty for cutesy marketing, just as a video of someone losing his mind in an assisted living facility isn’t something you’d upload to YouTube. Both 23andMe and Navigenics have the power, albeit limited, to give you the worst news of your life. People without medical training (and some statistics skills, too) by and large don’t know how to take this information. To receive it in such a casual, even social manner, seems antithetical to its purpose, not fun or even pro-active, as it’s pitched.
Call me old fashioned, but when I’m experiencing a cold F U from deep in my evolutionary past, I like that to be in a sterile, uncomfortable hospital room that’s devoid of smiles.
All in all, I look at my results, and I’m left with the impression that my DNA is strong, but that’s only based upon a relatively small sample of studies looking at a small sample of markers. There’s just so little we know about our genome. In five or ten years, not only will the services be a whole lot cheaper; they’ll be significantly more detailed, hopefully presenting results in a means that’s simpler to decipher.
Having said that, I feel lucky to have experienced a small, if humbling, peek into what may be in my future—pure science fiction just 20 years ago. And if you are dying to know your DNA better, I’d suggest forking over the $1000 for Navigenics rather than the $420 for 23andMe—mainly because I prefer Navigenics’ interface and along with their lack of persistent pressure to share my private genetic in a sphere that could damage my financial well-being down the line.
* * * * * *
We sat there for a moment. My wife had remained coolly objective regarding my genetic results until Alzheimer’s came up, and for the first time in our relationship, I imagined a life in which I, the proud male, was the physically weaker one, in which we staved off an almost inevitable ravaging of my brain starting who knows when.
But there was no laughing it off, so instead I chose to shrug. I dodged the bullet of one APOE marker only to be hit by another. Whether or not that bullet, irremovable from my DNA, actually injures me is something I’ll always worry about—keeping an ear to the ground on the latest researches of prevention, and hoping that 20% always rounds down to 0%.
Via @ GIz