Tag Archives: David Dobbs

The Downside of the Downside of Resilience: A New York Times Oped Ventures Into Dangerous Territory

Although we take seriously the threat of genetic discrimination, there aren’t a lot of examples you can offer. In my ethics class, I discuss the Burlington Northern Santa Fe Railroad case — everyone discusses the Burlington Northern Santa Fe Railroad case, not because it is such an interesting precedent but because it is all we’ve got. BNSF secretly tested their employees for genetic liability to carpal tunnel syndrome. The fact that it was genetic testing was almost beside the point. Can ever you secretly test your employees? No, you cannot. But the genetic testing angle made it extra creepy. Why? Because we are primed to worry about genetics. It is too new and too powerful not to carry with it the seeds of some unspecified disaster. We just don’t know what it is yet. We are heading out into the wilderness here, the wilderness within. How can we set about to tinker with the machinery of life without wondering if we run the risk of turning our tears acid and drowning our good intentions in our own rising tide?

Sometimes I wonder if genetic discrimination is a Yeti, a word we whisper around the campfire to give shape to our fears of the great unknown. After all, formlessness does not diminish fear, it makes it worse. If you don’t know what you are looking for it could be anything. It leads us into a state of vigilance that is both laudable and incredibly annoying, since every step forward is met by cheers and then, at the back of the crowd, a sideways glance and a muttered, “what could possibly go wrong?”

This is why I was so struck by Jay Belsky’s article, the Downside of Resilience, published in the New York Times Sunday Review this past week. Belsky points to work, his own included, that suggests some genes that may predispose children to do badly under stressful conditions – abuse, trauma, etc – are not so much “bad” genes as “responsive” genes – and that the same genetic inheritance makes them equally responsive to good parenting or helpful interventions. It is called the orchid and the dandelion theory, with the idea being that some kids do fine in all circumstances – the dandelions, growing like proverbial weeds – while others are hothouse flowers, dying in adverse conditions and blooming in the right hands. If this interests you, read more in this article from the Atlantic by the inimitable David Dobbs (and really just read anything the man writes; you can’t go wrong).

Belsky goes on to propose that we identify children with this genetic predisposition to responsiveness and target them – a good use for our “scarce intervention and service dollars.” We’re not ready to do that, he concedes. But, he asks, “if we get to the point where we can identify those more and less likely to benefit from a costly intervention with reasonable confidence, why shouldn’t we do this?”

Well, okay. A few reasons. First of all, the proposal implies a level of genetic determinism that is unsupported by the facts and fundamentally misleading when it shows up in a place like the NY Times. These are population-based observations, very interesting as to the nature of the genes and how they work, but not valid predictors of individual performance. There are too many confounding variables in the lives and the genetic makeup of individuals. As genetic counselors could tell him, even when you have the same variant in the same gene in the same family, outcomes may vary wildly.

However*, as I said in a response to the Belsky editorial, arguing the science suggests that if we could get that right it would be a good idea. History, on the other hand, suggests that creating classes of people based on what genes they carry is a dangerous proposition and not something to which scientists should lend credibility. The Belsky proposal is obviously well intended. He talks about benefitting the children who have the genes to respond, not disadvantaging the others. But, as he says himself, intervention dollars are scarce. Scarce resource are a zero-sum game. To give to one, you take away from others. You designate certain people as more worthy based on their genes. You incorporate genetics into social policy in a way that is ripe for abuse and prejudice masquerading as scientific facts. We have been down this road before. We know where it leads. It’s not a pretty place.

What does genetic discrimination look like? It looks like this.

*This is what I wrote but not what they published, because the NY Times doesn’t like sentences that start with ‘however’ and changed it to ‘but’. Whatever, NY Times.

Follow me on twitter! 


Filed under Laura Hercher



On Monday, the FDA celebrated the start of the holiday season by sending a letter to 23andMe, informing the direct-to-consumer personal genomics service that they must cease and desist offering their signature test.  The proximal cause of this action, as described by the FDA and not disputed by the company, was that 23andMe had ceased to participate in a process of establishing their PGx test as “validated for its intended uses.”  They had, the FDA suggested, dropped the ball – well, not just dropped the ball, but kicked it out of bounds, an old soccer trick for delay of game, which Mya Thomae and Dylan Reinhardt suggest might have been exactly what the company had in mind, playing for time while they attempted to accumulate better data than what-all they have right now.

The FDA move prompted a vast twitterlanche of commentary, ranging from indignant outrage to smug satisfaction (Dietrich Stephan, founder of the erstwhile DTC competitor Navigenics, said, “Engaging the FDA as a partner to bring the most robust and safe new type of test to market is diagnostics 101”).  Genetic counselors might be suspected of indulging in a bit of schadenfreude, since the relationship between 23andMe and the GC community has inclined in the direction of mutually suspicious, if not downright frosty.   The company, which advocates for access to one’s DNA information with almost a religious fervor, sees GC’s as gatekeepers, as a self-anointed coterie of priestesses guarding the oracle at Delphi.  Genetic counselors, for their part, tend to perceive the very existence of 23andMe as an affront, as though the possibility that a subset of people might benefit from genetic testing without access to counseling was insult and injury — an existential threat.

To be fair, nobody reacts well to the suggestion that their chosen profession is a cabal that threatens the freedom and well-being of fellow citizens – not even investment bankers, and at least they get to soothe their wounded souls with lots of material goods.  But pettiness is unbecoming and unproductive, and we would all do well to remember that a groundbreaking organization like 23andme is a part of the energy and excitement of the field – an expression of an explorer’s mentality that draws people to the potential of genetics in 2013.  That’s not only fun and sort of cool but also incredibly powerful because it attracts the kind of intelligence and curiosity that makes big new ideas possible (David Dobbs does a thoughtful and balanced job making the case for 23andMe in this piece for the New Yorker).

So minus any animus toward 23andMe, was this a reasonable move by the FDA?  There are two main questions that have been raised: 1. can they regulate? and 2. should they regulate? (a third question, HOW DARE THEY?, has also gotten a lot of play but I am going to ignore that one because, c’mon guys get over yourselves this isn’t ONE STEP FROM TOTALITARIANISM).  The first one takes up the issue of whether or not a personal genomics test falls under the FDA jurisdiction.  I am going to say yes, but will not rehash those arguments here, since they have been more ably covered elsewhere – I particularly recommend this piece by Hank Greely at the Stanford Law School Center for Law and the Biosciences blog.

So, should they regulate?  The rationale for regulatory action in the letter to 23andMe is a risk of harm to customers, including the possibility that a customer might alter his medication without medical advice or misunderstand her risk for breast cancer and have an unnecessary prophylactic mastectomy.  While theoretically true, it seems wildly unlikely that very many people would insist on a mastectomy without getting more information than mail-order genetic results – and those cases might be more indicative of out-of-control anxiety issues and irresponsible medical practice than the power of a PGx report.  More commonly, misunderstanding the limits of the test in terms of risk reduction might empower a customer to skip out on appropriate preventative measures. Either way, this is nothing new — a rehash of concerns genetic counselors have had about DTC testing since its inception.  In practice, perhaps the best summary of the clinical impact to date comes from Anders Nordgren, who called it “Neither as harmful as feared by critics nor as empowering as promised by providers.”  Having spent hours poking through the generally well-written and thorough 23andMe reports, and spoken to some of their customers, I would suggest that misunderstanding the results that come from 23andMe could pose some risk — real risk, to be sure, but limited risk.

However, it is possible to envision a scenario where a genetic testing sold DTC did pose a significant danger to consumers, with inaccurate results, irresponsible advice, tests used to market scam treatments or preventatives.  None of this is farfetched, and some of it has already been documented.  For this reason alone, the possibility of FDA action is an important deterrent.  A company like 23andMe, which makes real efforts to be thoughtful and responsible, will ultimately benefit from the restraint on less scrupulous entities.  And of course, it is possible that 23andMe would have been less thoughtful and less responsible if they had not been motivated by the threat of FDA action.  So arguments against regulation in general based on the fact that 23andMe is well-intentioned are misguided.

But despite a bias in favor of showing some muscle, I have questions about how much time and energy the FDA should spend cracking down on the likes of 23andMe.  Is it, I wonder, the best use of resources?  For one thing, attempts to stop the free flow of information in 2013 are fingers in the dike. Razib Khan at Slate expands on this argument, suggesting that companies pushed by the FDA could simply move offshore, away from any regulation.

And more importantly from my point of view, the emphasis on negative action diverts us from the possibility of doing something positive.  Rather than keeping consumers away from tests we think are insufficiently documented, how about providing a resource to the general public that endorses tests that are ready for prime time?  After all, a few bold individuals may be excited at the prospect of downloading Promethease to query their own exome data but most people would rather not, thank you very much.  Most people would be happy to have some guidance.  They can get that guidance from the company, but even the classy companies have a vested interest in hyping the significance of their results – that’s what they’re selling, right?  There is an opportunity here for the government and the genetics community to create a trusted source of information that is neutral, unbiased and supports a best-case scenario use of genetic testing by those eager to take the plunge.  Hell, you could imagine tying in such a resource to something like ClinVar or GenVar, so that early adapters could contribute to publically accessible databases rather than giving it to 23andMe to sell.

After years of running up and down the pitch, the FDA has demonstrated that it knows how to blow the whistle – that’s good.  I’m pretty sure 23andMe will be back – and that’s good too.  But if we really want something great to come out of this discussion, let’s stop doing color commentary on the FDA action, and imagine what it could be like if we changed the game.

Follow me on twitter!


Filed under Laura Hercher