Our DNA, Our Selves

by on November 26, 2013 at 7:20 am in Current Affairs, Economics, Law, Medicine, Uncategorized | Permalink

At the same time that the NSA is secretly and illegally obtaining information about Americans the FDA is making it illegal for Americans to obtain information about themselves.

In a warning letter the FDA has told Anne Wojcicki, The Most Daring CEO In America, that she “must immediately discontinue” selling 23andMe’s Personal Genome Service, more affectionately known as the spit kit.

As I wrote when this issue first surfaced in 2010:

The ability of genetic tests to predict diseases is currently limited; if the FDA were simply to require firms to acknowledge this point, say with a clear statement of probabilities, that would be one thing (although this task is better met by the FTC under advertising regulation). But the FDA is brazenly overreaching in trying to regulate genetic tests as medical devices. First, there is no question that these tests are safe–safer than brushing your teeth!–and also effective in identifying genetic markers. Thus, DNA-Test-Tube-300x300there is no medical reason whatsoever for regulation.

Moreover, genetic tests provide information, personal information about our bodies and our selves. The FDA has no standing to interfere with the provision of such information.

Consider, I swab the inside of my cheek and send the sample to a firm. The idea that the FDA can rule on what the firm can and cannot tell me about my own genes is absurd–it’s no different than the FDA trying to regulate what my doctor can tell me after a physical examination or what my optometrist can tell me after an eye examination (Please read the first line.  “G T A C C A…”).

The idea that the FDA can regulate and control what individuals may learn about their own bodies is deeply offensive and, in my view, plainly unconstitutional.

Let me be clear, I am not offended by all regulation of genetic tests. Indeed, genetic tests are already regulated. To be precise, the labs that perform genetic tests are regulated by the Clinical Laboratory Improvement Amendments (CLIA) as overseen by the CMS (here is an excellent primer). The CLIA requires all labs, including the labs used by 23andMe, to be inspected for quality control, record keeping and the qualifications of their personnel. The goal is to ensure that the tests are accurate, reliable, timely, confidential and not risky to patients. I am not offended when the goal of regulation is to help consumers buy the product that they have contracted to buy.

What the FDA wants to do is categorically different. The FDA wants to regulate genetic tests as a high-risk medical device that cannot be sold until and unless the FDA permits it be sold.

Moreover, the FDA wants to judge not the analytic validity of the tests, whether the tests accurately read the genetic code as the firms promise (already regulated under the CLIA) but the clinical validity, whether particular identified alleles are causal for conditions or disease. The latter requirement is the death-knell for the products because of the expense and time it takes to prove specific genes are causal for diseases. Moreover, it means that firms like 23andMe will not be able to tell consumers about their own DNA but instead will only be allowed to offer a peek at the sections of code that the FDA has deemed it ok for consumers to see.

Alternatively, firms may be allowed to sequence a consumer’s genetic code and even report it to them but they will not be allowed to tell consumers what the letters mean. Here is why I think the FDA’s actions are unconstitutional. Reading an individual’s code is safe and effective. Interpreting the code and communicating opinions about it may or may not be safe–just like all communication–but it falls squarely under the First Amendment.

The FDA also has the relationship between testing and clinical validity ass-backward. The FDA wants to say no to testing until clinical validity is established but we are never going to discover clinical validity until we have mass testing. 23andMe is attempting to leverage individuals thirst for knowledge about themselves into a big data project that will discover entirely new connections between genotype and phenotype. But personalized medicine, just like personalized movie recommendations, only works with databases of millions. In the 20th century we took on many of our common diseases but it is now time to take on the uncommon diseases. There are some 7,000 known diseases and only about 500 have a treatment. Individual and disease heterogeneity is so large that even the diseases that we can treat are often not treated well. New approaches are necessary for progress. The collection of large amounts of DNA data is not the last step of personalized medicine but the first and by pushing back against the first steps the FDA is delaying the promise and progress of personalized medicine.

Full Disclosure: The FDA’s threat to regulate genetic tests in 2010 made me spitting mad so I put that spit to good use and became a 23andMe customer. Well worth it, if only to point out to my wife that contrary to all evidence I am in fact only 2.2% Neanderthal.

Tyler Cowen November 26, 2013 at 7:35 am

Here is a good piece by a doctor on the decision: http://slatestarcodex.com/2013/11/26/a-letter-i-will-probably-send-to-the-fda/

Spitter November 26, 2013 at 7:38 am

The regulatory challenge here is that technology has transformed a service (advice about health & risk) into a product (affordable spit kits) so FDA has jurisdiction. That said, I bring this controversy back to its core: should the right of a palm reader to give medical advice also be protected by the first amendment? What if the palm reader promises to get better with practice?

mofo. November 26, 2013 at 8:22 am

Or your mother? Eat Chicken soup and get some rest is probably the most common medical advice in the world.

Bill November 26, 2013 at 8:31 am

+1 Furthermore, Alex’s objection is “the FDA wants to judge not the analytic validity of the tests, whether the tests accurately read the genetic code …but the clinical validity, whether particular identified alleles are causal for conditions or disease. The latter requirement is the death-knell for the products because of the expense and time it takes to prove specific genes are causal for diseases…”

So, imagine that I have this machine (call it a aluminum wrist bracelet) that accurately measures galvanic skin response (similar to: “It accurately reads the genetic code”). I make the claim that galvanic skin response predicts or is associated with the likelihood of a seizure, but have not proven the claim because, as Alex says, the “latter requirement is the death-knell for the products because of the expense and time it takes to prove specific {galvanic skin responses} genes are causal for diseases.”

Sounds silly, doesn’t it.

It is even sillier to say this is constitutionally protected speech. Otherwise, consumer protection statutes would all be unconstitutional.

That said, you could probably have a doctor order a spit test. The doctor reads the tests and the literature, and gives you advice. I am much more willing to have human intervention mediate information and analysis. In that case, the fact that allieles are correctly identified as defective is relevant, and would probably be the limit of FDA jurisdiction.

Bill November 26, 2013 at 8:48 am

Oh, and for the audience members who are not aware of this, we regulate “commercial speech” differently than ordinary speech.

“Commercial speech is speech done on behalf of a company or individual for the intent of making a profit. It is economic in nature and usually has the intent of convincing the audience to partake in a particular action, often purchasing a specific product. Generally, the United States Supreme Court defines commercial speech as speech that “proposes a commercial transaction.” Additionally, the Court developed a three factor inquiry in determining whether speech is commercial in Bolger v. Youngs Drug Products” http://en.wikipedia.org/wiki/Commercial_speech

Dan Weber November 26, 2013 at 8:52 am

For once I’m with Bill.

23AndMe markets this thing as, among other things, a health product. While I’m generally in favor of a much different FDA, given the FDA we have now, this is the right thing for them to do until 23AndMe complies with the regulations dealing with health products.

Z November 26, 2013 at 9:38 am

The elephant in the room missed by Bill and Alex is who benefits from this rabid assault on 23AndMe? The FDA allows all sorts of hucksters to peddle junk science to the public. The harm potential of supplement makers is exponentially higher than this firm. Why the interest? I think that answer is far more important because it underscores the central defect of the custodial state.

prior_approval November 26, 2013 at 10:03 am

‘The elephant in the room missed by Bill and Alex is who benefits from this rabid assault on 23AndMe?’

Apparently, test labs that help 23andMe conform to the necessary regulatory concerns. See below from somone who apparently actually has expertise in the field.

Otherwise, the answer is the beneficiaries are those wishing genetic testing of a medically useful level, compared to the current entertainment value of 23andMe results.

Z November 26, 2013 at 10:35 am

@prior_approval: It is OK if you don’t know the answer. Screaming out the wrong answer is not helpful.

Larry Hamsing November 26, 2013 at 10:11 am

Medical advice can be found in all sorts of print media and the internet. If webmd.com says it’s a good idea to take garlic supplements to reduce blood pressure, is that advice also illegal?

In all instances of so-called “advice” – from what PC to buy to how to fix your car, do your research and talk to someone you trust. In this case it would be your doctor.

Mark Thorson November 26, 2013 at 10:52 am

Yes, it would be illegal if they’re also selling garlic supplements and can’t back up their claims with evidence to the satisfaction of the FDA. If you claim your garlic supplement will treat high blood pressure, you are making a drug claim and you must be held to the same standard as any other drug.

The way the dietary supplement industry gets away with what they do is most manufacturers make no claims or just nebulous structure/function claims (“supports a healthy immune system”). They rely on third-party magazines, books, etc. to educate customers about what to buy.

Alternatively, firms may be allowed to sequence a consumer’s genetic code and even report it to them but they will not be allowed to tell consumers what the letters mean. That’s exactly what they should do. Perform the tests, report the data, and leave interpretation of the data to others.

Dan Weber November 26, 2013 at 10:55 am

I’m pretty sure WebMD avoids saying anything like “take garlic supplements to reduce blood pressure.”

Right now we live in a world where medical and legal advice are considered different than mechanical or home-improvement advice. There are some good reasons for that, even though a very good case could be made to deregulate the former to act like the latter.

Cliff November 26, 2013 at 10:46 am

Well you and I are both lawyers and it seems plain under your definition that this is not commercial speech.

Randall Parker November 29, 2013 at 12:06 am

Finding out details of our genetic code is commercial speech because we have to pay to get the information? Sounds like the Supreme Court did a lousy job coming up with a commercial speech doctrine.

I find immoral the idea that the US government should have the power to prevent me from finding out facts about my biology without going thru a medical doctor.

Keith November 26, 2013 at 9:56 am

Poor analogy. 23andme is not trying to say if you have these alleles you will get this disease. They say that other researchers have found that an allele is associated with an x percentage greater or lesser chance of having a disease. A more appropriate way to state your analogy is to say the provider of the galvanic skin response bracelet include a sheet of paper with their product that explains that reports published by scientists have connected x reading with an x percentage increase in having a disease. This doesn’t sound so silly does it.

I am not sure what galvanic skin disease is so I think the better analogy is that of a thermometer. Temps in the green are normal but those in the red may indicate fever. What is wrong with this?

Bill November 26, 2013 at 7:49 pm

So, Keith, what do you say to the parent of the 23 year old who commits suicide because he/she does not understand probability or statistics?

Keith, either this test tells something or it doesn’t. If you are saying it doesn’t say something, then it is a fraud.

If you say that it is dispensing medical advice in connection with its service, then it has to be regulated.

And, you seem to ignore Alex’s statement, which I take as true, that it is a burden they cannot meet: “the clinical validity, whether particular identified alleles are causal for conditions or disease. The latter requirement is the death-knell for the products because of the expense and time it takes to prove specific genes are causal for diseases…”

Do you give them a oija board?

My advice, go to a doctor who can advise you on genetics, and let them test.

Keith November 26, 2013 at 8:46 pm

The 23 year old can misunderstand information from any number of sources and commit suicide. In fact I would rather they get the service so they get the real story.
The data is not fraudulent, it is collected from current science which means at worst it is incomplete. I guess that is choice C in your list of choices.
Can we also consider the lives that could be saved if people change their behavior in a positive way based on these tests? What about the lives saved in the near future because this technology is allowed to progress?

Bill November 26, 2013 at 9:07 pm

“At worst it is incomplete.” I’m glad you do not make parachutes.

Marian Kechlibar November 27, 2013 at 7:28 am

“… what do you say to the parent of the 23 year old who commits suicide because he/she does not understand …”

This argument can be used to suppress free speech, free press and freedom of religion. All of them have already had some youth’s suicide as a consequence.

The proponents of freedom do not generally claim that freedom is an universally beneficial, unalloyed good. The argument is that freedom in total brings more benefits than damages.

Bill November 27, 2013 at 5:39 pm

Marianne,

You can shout fire in a theatre by your definition of free speech.

Some prefer to live in social societies which curb people shouting fire in the theatre,

Or bringing automatic weapons in the theatre to kill the audience under the guise of protecting the right to bear arms.

Freedoms are never limited..if you live in a society.

Bill November 27, 2013 at 5:39 pm

substitute “always” for “never”

Careless November 27, 2013 at 9:44 pm

” what do you say to the parent of the 23 year old who commits”

“Unfortunately, the Darwin Award committee doesn’t hand out actual physical rewards you can remember your son’s achievement by”

Randall Parker November 29, 2013 at 12:09 am

A doctor who can advise you on genetics: The amount of genetic data is huge. The amount of physician knowledge is tiny. These doctors are rare to non-existent. Plus, the costs are too high.

The FDA has not met a reasonable burden of proof that 23andme is a threat to the public’s health.

Arun November 27, 2013 at 11:34 am

I make the claim that galvanic skin response predicts or is associated with the likelihood of a seizure, but have not proven the claim because, as Alex says, the “latter requirement is the death-knell for the products because of the expense and time it takes to prove specific {galvanic skin responses} genes are causal for diseases.”

But that is not what they are doing. They simply cite the scientific literature, e.g., for Primary Biliary Cirrhosis:

Hirschfield GM et al. (2010) . “Variants at IRF5-TNPO3, 17q12-21 and MMEL1 are associated with primary biliary cirrhosis.” Nat. Genet. 42(8):655-7.
Mells GF et al. (2011) . “Genome-wide association study identifies 12 new susceptibility loci for primary biliary cirrhosis.” Nat. Genet. 43(4):329-32.
Krausgruber T et al. (2011) . “IRF5 promotes inflammatory macrophage polarization and TH1-TH17 responses.” Nat. Immunol. 12(3):231-8.
Jones DE et al. (2011) . “The genetics of primary biliary cirrhosis: the revolution moves on.” Hepatology 53(1):362-4.

They say something like, “0.11 out of 100 men of European ancestry who share this genotype will develop PBC between the ages of 20 and 79″ compared to “0.08 out of 100 men of European ancestry will develop PBC between the ages of 20 and 79″.

(That, incidentally, is a 43% increase in risk).

Show me the scientific literature for your aluminum wrist bracelet and galvanic skin response, and I’ll admit these to be the same.

Why should the market be limited because of general innumeracy and scientific illiteracy?

Bill November 28, 2013 at 9:42 pm

Arun, Citing a medical journal with regressions, t-tests, levels of significance, etc. is facially responding but in fact is not. It is jibberish to the recipient, who then goes off the deep end worrying, or not. This should be mediated through a doctor. Moreover, what you cited is not what Alex said, which was that the FDA was requiring them to prove their claims, which would be too costly and difficult.

Randall Parker November 29, 2013 at 12:10 am

The doctor can no better explain probabilities and probably doesn’t explain them well.

The assumption that medical doctors can interpret genetic tests is false.

Tracy W November 27, 2013 at 4:42 pm

Why does the aluminium wrist bracelet sound silly?

As for the doctor, let me relate a conversation I had once with my doctor’s holiday cover (a a presumably fully-qualified doctor).
“Doctor I’ve just found out that my 6-week old baby’s been exposed to whooping cough!”
“Don’t worry, it’s not very contagious.”
“Are you sure? I’ve just been looking at the NHS website and it says it’s highly contagious!”
“Hmmm, I’ll get back to you.”

Engineer November 26, 2013 at 9:55 am

> That said, I bring this controversy back to its core: should the right of a palm reader to give medical advice also be protected by the first amendment?

Of course.

I fail to understand the paternalistic obsessions of supposedly liberal minds.

prior_approval November 26, 2013 at 10:04 am

Give medical advice or sell a, at least putatively, medical service? A distinction which most people find remarkably easy to make, actually.

Z November 26, 2013 at 10:05 am

http://www.nytimes.com/2013/11/15/nyregion/5-to-15-year-prison-sentence-for-manhattan-fortuneteller.html?hpw&rref=nyregion&_r=0

Since we throw fortune tellers in jail for being wrong, I don’t see why would not throw palm readers in jail for dispensing medical advice. Maybe next we can toss people in jail for telling us we can keep our insurance, if we like our insurance.

chuck martel November 26, 2013 at 10:28 am

What a story! But it doesn’t indicate if she was actually wrong or not. Are only incorrect predictions illegal or are any predictions? Can I send my stock broker to jail for having recommended K-Mart?

Marian Kechlibar November 27, 2013 at 7:31 am

What, random fortune tellers are thrown into prison, and Paul Krugman isn’t?

It probably depends on whom you fortune-tell, and whether you tell him precisely what he wants to hear, or no.

stubbs November 26, 2013 at 11:31 am

Really, isn’t the “core” whether the report constitutes “medical advice” and not just information to be used or not?

Noah Yetter November 26, 2013 at 11:46 am

Yes, it should be protected. Why is this even a question?

Adrian Ratnapala November 27, 2013 at 2:23 am

Because allowing people to do things simply because they have a formal well known important right to it would institute dangerous principle. Society has worked hard to create precedents for ignoring those rights, and in certain circumstances (i.e. those that involve human bodies, and the products of human work). It would be a grand shame to upset those precedents.

ahduth November 26, 2013 at 8:05 am

These are the people who ignored FDA information requests for five years, right? My heart bleeds with sympathy.

Rahul November 26, 2013 at 9:18 am

I’m sympathetic to the libertarian excess regulation aspect of this. OTOH, I’m skeptical 23AndMe will discover entirely new connections between genotype and phenotype.

23AndMe makes money whether they demonstrate clinical validity or not, so why try very hard. It’s a great business model. BigData makes it cool. And when the data is big enough you can be sure to fish out some correlations at least which is icing on the cake.

What it boils down to is whether & how far idiots need to be protected from being parted from their money.

Keith November 26, 2013 at 9:37 am

Rahul,
They have found new genotype-phenotype connections. If you subscribe to their service you will see these indicated. They also published their findings in peer reviewed journal articles and presented them at conferences.

I agree with Alex on this issue. The company lists the probability of suffering from a disease next to the allele. If you want to change your behavior or seek medical treatment based on the results, that is your own business. FDA butt out.

Rahul November 26, 2013 at 10:03 am

@Keith:

Having spent a fair bit of time in academia I’ll say that just having published in peer reviewed journals is a fairly low standard of proof. Having presented at conferences is even lower. I can’t comment about the quality of FDA’s internal scrutiny process but I’d sure hope it was better on average.

Really though, this is not specific for medicine. I worked on refinery process design for a while and if we trusted data from every darn peer reviewed article we’d be having explosions every other day.

Like I said, whether the FDA ought to interfere I don’t know, but I’m skeptical if 23AndMe is adding value or just a bunch of trivial / mostly spurious correlations. A list of disease probabilities means nothing unless I can trust the causal reasoning behind it.

Keith November 26, 2013 at 10:35 am

But their model is that they also provide surveys to their users asking them if they have certain phenotypes (baldness, brown eyes etc.) so they are constantly testing the assertions made in the journals. What more can you ask for from a company that just charges 99 bucks? I am happy with them and so are lots of other people.

Cliff November 26, 2013 at 10:53 am

So, what, doctors can’t give advice based on peer-reviewed studies because those just aren’t reliable? What the hell are we going to use to give advice?

wrparks November 26, 2013 at 10:43 am

Big data basically guarantees they will find new associations between genotypes and phenotypes unless the quality of the data is suspect. I would be concerned about any self-reported phenotypes.

Especially if they can start pushing people to include a pedigree and getting their parents/family involved with testing as well. At some point in there, they may have crossed an ethical guideline between providing a service and using the data for other purposes. Rules are in place in human research that google/apple/amazon don’t have to abide. I’m unsure if these apply.

Makes me glad I don’t work with human genetics!

Randall Parker November 29, 2013 at 12:13 am
charlie November 26, 2013 at 8:12 am

Sell a device as a genetic screening for health and expect the FDA to ignore you?

Guys — she got dumped. And a bad prenup. There is no cash or grants for sticking up for them. next.

Guthrie November 26, 2013 at 8:16 am

Another day, yet another small loss of Americans’ personal liberty and privacy under government rulers.
Constant small losses sum to big losses; once free citizens eventually become subjects and serfs.

Yawn. At what point do you notice the larger trend and seek its cause and cure ?

Thomas Jefferson sagely noted that “the natural progress of things is for liberty to yield and government to gain ground”…
That observation is confirmed everyday in the news.

Brandon November 26, 2013 at 10:26 am

Thomas Jefferson also owned and traded other human beings as chattel. The natural progress in that regard was for white supremacy to yield and equality to gain ground using government as the main tool of change and enforcement. Pithy quotes don’t always tell the full story.

Plus I’m not sure why we should really lament the loss of “liberty” that we had back when patent medicines were all the rage.

Keith November 26, 2013 at 10:36 am

This isn’t snake oil.

mike November 26, 2013 at 11:27 am

Liberty = white supremacy, thanks for broadcasting how the liberal mind works

Brandon November 26, 2013 at 12:15 pm

Actually, if anything, that sort of reinforces my point. Liberty in America in Jefferson’s time meant the liberty of whites to engage in slavery and all sorts of white supremacy. Over time, we ‘lost’ that ‘liberty’ at the hands of government. We gained real, actual liberty at the expense of the ‘liberty’ of the white supremacists. The example was meant to illustrate that Jefferson’s quote is not universally applicable and isn’t really an argument. That it was Jefferson making it just adds irony.

Over a century ago, we had the “liberty” of patent medicines and all sorts of snake-oil remedies. Uranium was sold as a medical supplement. I do not regret the loss of the “liberty” to purchase untested, unverified products making explicit medical claims. I’m not claiming that 23AM is or isn’t snakeoil, but Guthrie’s comments was generalized itself.

mike November 26, 2013 at 1:26 pm

thanks, again

Brandon November 27, 2013 at 10:46 am

Mike,

Thanks for what? I’m pointing out that “liberty” means very different things to very different people, and at certain times and to certain people, their definition of liberty included the freedom to very explicitly deprive others of liberty. Pointing this historical fact out is not equivalent to agreeing with that viewpoint. I’m not sure where that disconnect is coming from.

More explicitly, I completely reject the notion of liberty held by white aristocratic slavers, or at least those parts of their notion of liberty (not everyone is wholly evil, including/especially Jefferson). I reject definitions of liberty that would give us Gilded Age-era “freedoms.” The liberal mind, for however much one single person can represent it, does not equate liberty with white supremacy. It recognizes that some people have definitions of liberty that encompass or include white supremacy, and it recognizes that therefore there is no universally agreed upon standard of liberty.

Which is really just a tl;dr way of saying that “pithy quotes from 18th century aristocrats about liberty and freedom aren’t universal truths”

biomuse November 26, 2013 at 4:39 pm

Mike, do you have an argument to make, or not?

Chip November 26, 2013 at 8:18 am

I became a 23andMe customer a few years ago. Only an ameba – or a bureaucrat at the FDA – would consider it high risk.

I take that back. An ameba would have the innate sense to seek light, whereas the FDA obviously prefers us in the dark.

Claudia November 26, 2013 at 8:24 am

In principle, I am sympathetic to the arguments post, though I suspect there are some less than commendable details on both sides of this story. One thing I am bit puzzled by here: why should I be in favor of 23andMe and against GMO labeling? http://marginalrevolution.com/marginalrevolution/2012/10/michael-pollan-on-proposition-37.html Why is info about my body more important than info about what I put into it? Or this simply about who is in control and less about the benefits to people?

Damien November 26, 2013 at 8:37 am

No-one is against voluntary GMO labeling. If organic farmers want to certify that what they produce is GMO free, let them do it and let them be sued if they’re lying. If 23andMe wants to sell information to willing buyers, let them do it as well.

I fail to see what puzzles you. GMO labeling is about forcing people to provide information that may or may not be relevant to customers, which the market can take care of. If I want to purchase GMO-free food, I can do so, but at a premium. The 23andMe case is about letting people provide information to customers who want to purchase it. A better parallel would be if someone started a service that tells people for a fee whether their food contains GMOs, and the FDA and USDA were trying to shut them down.

ahduth November 26, 2013 at 8:40 am

There is the vague scent of polemics in the air, yes.

I’m just mystified as to why these people thought it was a good idea to ignore the FDA? http://www.bloomberg.com/news/2013-11-25/fda-tells-google-backed-23andme-to-halt-dna-test-service.html

Even if we grant this techno-libertarian point, that they’re providing us with information we already “own,” since it relates to our bodies… the omission simply becomes more confusing. If they thought they were on the right side of this thing, why ignore the undisputed regulator in their market? It reeks of incompetence, or worse.

Axa November 26, 2013 at 9:25 am

Thanks for the Bloomberg link, it seems the only risk that FDA tries to avoid is unintended quackery. Even a well researched decease as Parkinson, it’s not fully understood yet. http://ghr.nlm.nih.gov/condition/parkinson-disease Read to find certain % of cases of Parkinson is related to some gene variations. Also, if you’ve read papers or scientific literature in the last 10-15 years, you’ll remember the causality relationship between genes and deceases is far from settled. Results that one researcher publishes this year, contradict results from other researcher in 2012, but confirm results from 2009 from another researcher……endless.

I don’t know the specific language 23andme uses to communicate results to clients, that may be the problem. It would be great if Alex provided more detail since he is a client. If 23andme sends you a big DAT file containing the G T A C C A code in the standard format scientist use today, fine. Claiming causal relationships to deceases is another story that enters in FDA territory. The gene-decease relationships found up to day are valid, with some degree of error, for the studied population not for individuals. The technical advance 23andme has made in gene sequencing is marvelous, but converting that into medical assertions could be just an educated guess.

Ps. Show me the confidence intervals and/or error estimation when you assert the relationship between my genes, not “my population” genes, and diabetes or Alzheimer.

Keith November 26, 2013 at 9:44 am

23andme list the increase or decrease in risk of having a disease based on the allele. They also have a summary of the latest research. I don’t see the problem; they aren’t advocating any course of action based on the results. I don’t see what is so bad.

biomuse November 26, 2013 at 4:45 pm

When someone walks off a bridge as a result of a poor intuitive grasp of statistics, I think the case for justifiable regulation will become clearer.

widmerpool November 27, 2013 at 4:06 am

How many of the tens of thousands of people actually diagnosed with a fatal or debilitating disease by doctors every day walk off a bridge?

Tracy W November 27, 2013 at 4:45 pm

Biomuse: if someone walks off a bridge after receiving medical news from a fully regulated source, would that make the case against regulation clearer?
Or does the ratchet only go one way?

Cliff November 26, 2013 at 10:56 am

They did not ignore the regulator.

Ryan Vann November 26, 2013 at 11:32 am

It reeks of Google is more important than the FDA and everyone knows it. If Google tells the FDA to go f- itself, then the FDA damn well responds with “how large of a dong should I use.”

Steven Kopits November 26, 2013 at 11:35 am

The two are not comparable. One is voluntary, the other is mandatory.

As for 23andMe: It’s $99, about what you’d spend to take the family to dinner at On the Border, and I can assure you the margaritas and salsa will do you more damage than a gene test. This is not a big deal. And if people don’t want it, they don’t have to buy it.

GMO labeling: This implies that you have to trace the inputs into your products all the way through. Do you buy wheat? Can you prove there’s no GMO seed in there? Further, what constitutes the limits of GMO? Do you mean “gene spliced”, or do you mean “a hybrid plant developed by mankind sometime in the last 5,000 years”? If it’s the former, why is gene splicing special? If the latter, then pretty much every single plant-based food you’re eating is a GMO.

The proper analogy would be that you allow companies to state on their labels things like: “Contains no GMO; no hormones; and fully organic”. I believe companies are in fact allowed to say that, no? Is anything preventing companies for declaring that they do not use GMO?

Claudia November 26, 2013 at 6:49 pm

Steven, I did not say it was apples-to-apples comparison just that the information principle is not being used symmetrically in the arguments on MR. I began mildly pro-GMO labeling as a matter of access to information (those genetic tests aren’t perfect either) but I could see the costs were too large. Maybe the FDA views the costs right now too high for 23andMe’s test results, it is their job to protect “responsible for protecting the public health.” Again with 23andMe in principle, I see getting the information out as beneficial to the advancement of science, as Alex argues, but I just don’t think you can use the ‘give the people the information they want’ in one case and not in another. That’s all.

Brandon November 27, 2013 at 3:53 pm

Does the FDA’s jurisdiction and the legitimacy of their concern rest on the price of the service or product in question?

ap November 26, 2013 at 8:27 am

I’m 2.6% neanderthal, my wife is 3.1%. It’s made winning domestic arguments a lot easier. 23andmeFTW

TMC November 26, 2013 at 11:36 am

You mean you better give in, or you’ll get your a** handed to you?

Tommy November 26, 2013 at 8:31 am

The risks they’re likely concerned about are the disruptions caused by widespread genetic testing, rather than any faults of this specific test. They’re trying to keep the cat in the bag.

FC November 26, 2013 at 8:45 am

Another day, another reason biotech is moving offshore.

Dan Weber November 26, 2013 at 8:57 am

What are your data points for how free Europe is for genetic testing?

FC November 26, 2013 at 9:06 am

Why would you bring Europe into this?

Dan Weber November 26, 2013 at 11:00 am

Genetic testing is a rich people product. The market for it is going to be very limited in Africa, Russia, India, China, or Australia.

Where do you think biologists are fleeing to from the United States so they can do their genetic stuff freely?

Adrian Ratnapala November 26, 2013 at 9:12 am

Because obviously that which is not America is Europe.

Michal November 26, 2013 at 9:20 am

I am afraid he didn’t mean Europe.

prior_approval November 26, 2013 at 9:31 am

As a note – in Germany, it is the U.S: that is viewed as the paradise of unrestricted genetic research, and it is German researchers that are often seen as being force to leave when various regulatory frameworks are found to be too restrictive.

But then, Germany is the place that indelibly showed what happens when unregulated human genetic experimentation runs rampant.

However, this is particular case can be pretty well summed up this way (from https://www.23andme.com/you/community/thread/25706/3/ via metafilter) -

“A little perspective, folks (I work in the medical device industry in Regulatory Affairs):

1. In order to market and sell a medical product in the US, there must be sufficient evidence of both safety & efficacy. The concern here is that 23andMe has not fully validated their processes to limit false positive and false negative results. The FDA letter lays out some examples of adverse outcomes from false readings, and they are worrisome (unneeded medical procedures or incorrect medications can do immense harm). 23andMe doesn’t have to eliminate those false rates, but they will need to demonstrate what they are and (at a minimum) provide adequate warning and education around them to protect the consumer.

2. This isn’t a cash grab. A 510(k) filing currently costs about $5000 plus a few person-weeks of work to complete. Since the letter indicates that 23andMe has submitted two filings and received feedback on each, the issues appear to be around the lack of safety/efficacy data and quality controls. Any money spent on that will go to the test firms, not to the Agency. So if anything, this will promote the business of test labs, not line the pockets of the FDA.

3. The outcome (cease & desist, seizure, etc.) will hinge largely upon what plans the company puts in place and how it responds to the Agency within 15 days. Unfortunately it would appear that they haven’t handled things well so far. I’ve seen many of these letters, and this one suggests an unusual degree of poor behavior on the part of the company.

4. In no way does this represent an attempt to shut down genetic testing or services like 23andMe. It does demonstrate that the regulators are paying attention and expect an adequate level of safety & efficacy data. Given how long the file was in process (~15 months, compared to 3-6 for a lot of other devices), I would suggest the Agency has spent a lot of time trying to negotiate a resolution before resorting to the Warning Letter.

All that said, it would have been nice if 23andMe had provided ongoing guidance and warning to its customers, and I personally believe they should now immediately provide guidance as to what will happen with our data and samples that are currently in process.”‘

derek November 26, 2013 at 10:17 am

THE FDA HAS TO PROTECT AGAINST NAZIS!!!!

I suggest you go take one of your well regulated pills to settle down. But be careful. 100,000 deaths occur in the US per year due to interactions between regulated drugs. Close to 20,000 due to regulated drug side effects.

How many folks has 23andMe killed?

Urso November 26, 2013 at 10:24 am

Did you seriously just Godwin’s law this thread?

The Anti-Gnostic November 26, 2013 at 9:00 pm

I signed up for 23andme, and got a coupon good for 15% off all purchases of Zyklon-B.

Careless November 28, 2013 at 10:27 am

In a particularly insane way, at that.

Keith November 26, 2013 at 10:56 am

I agree they could have hired a few more of you regulatory affairs folks and probably avoided all of this. That said, I am curious about your “safety and efficacy” statement. They only provide your allele and then quote from the literature about what the allele’s effects may be. Are false positives and negatives possible with genetic information? Are you referring to the correct base call? For instance, calling a base a “T” when it is really an “A”? Doesn’t that then go back to the instrument provider (Illumina I think) to get their instrument FDA approved?

Even if that were the case, and the instrument, and the lab have to become FDA approved, I still feel uncomfortable about it. These base readers are the most advanced available, the data presented by 23andme is the best out there, these aren’t diagnoses, and yet, the FDA can still shut it down?

David Wright November 26, 2013 at 2:29 pm

The quoted post gets to the heart of the matter. Alex claims that the regulator’s concerns are not about the accuracy of the genotype determination, just 23andMe’s claims of correlations with phenotypical (health) outcomes. This post contradicts that claim, saying that the regulator’s concerns are about the accuracy of the genotype determination (e.g. when 23andMe says you have BRCA variant, how often do you actually have it?) I would really like to see some documentation indicating who is right about what the FDA’s concerns are. Can anyone provide a link?

David Wright November 26, 2013 at 2:48 pm

Looking at the linked C&D letter, it looks to me like the quoted poster is right and Alex is simply wrong. This is about potentially false genotype readings, i.e. “help[ing] consumers buy the product that they have contracted to buy”.

Brandon November 27, 2013 at 10:47 am

It seems to me that many of the critics of the FDA’s letter, both here and over at Volokh, have not actually bothered to read the letter.

The Anti-Gnostic November 26, 2013 at 9:03 pm

First they tell the Ashkenazim they should be tested for Tay-Sach’s, next thing you know they’re putting them in the cattle cars.

Tracy W November 27, 2013 at 4:27 pm

Germany is the place that indelibly showed what happens when unregulated human genetic experimentation runs rampant

Uh, the Nazis were enthusiastic proponents of eugenics. They are an excellent illustration of some of the dangers of regulating human genetic experimentation.

chuck martel November 26, 2013 at 9:03 am

Where was the FDA when all those police labs were providing bogus drug, alcohol and DNA evidence in criminal cases?

buddyglass November 26, 2013 at 9:04 am

Crap. My parents were going to buy me a 23andme analysis for Christmas, and get one themselves. Guess that’s not happening now.

PD Shaw November 26, 2013 at 9:10 am

Same here. My in-laws usually send me a nice check for Christmas; last year I decided to get new hiking shoes and put-off 23andme until this year. While I’m tempted to go ahead and buy now, I don’t want to get caught with an empty vial.

anon November 26, 2013 at 9:24 am

Try the National Genographic Project kit instead. On sale now for $160
https://genographic.nationalgeographic.com/

And there are many other DNA kit sellers, including
https://dna.ancestry.com/
http://www.familytreedna.com/

john personna November 26, 2013 at 10:30 am

One interesting thing about the Genographic Project is that they explicitly try to avoid any gene test associated with heritable disease. They look for population/migration markers which are not, coincidentally, disease markers. Why? Well, they are avoiding all this controversy, but also avoiding the risk of any lawsuits along the lines of “you told me I had X, and I don’t have X after all!”

Urso November 26, 2013 at 11:22 am

Seems like a good idea. One reason I wouldn’t use 23andme is because I have no interest in knowing any of that.

Anecdotally, I have a relative with a heritable (nonfatal) disease, and he got the 23andme done last year. He was showing me all his genetic markers, and kind of laughed at the one showing a very low chance of inheriting the disease that he already knew he had. Of course, “low” does not equal “zero.”

anon November 26, 2013 at 3:31 pm

This reminds me of this New Yorker cartoon

http://www.condenaststore.com/-sp/Thought-balloon-above-man-reads-Less-Cholesterol-Regular-Checkups-No-Nic-New-Yorker-Cartoon-Prints_i8575250_.htm

An updated version could include the words: “No high risk alleles”

The unexpected still happens.

Careless November 28, 2013 at 1:38 pm

Oh good, more expensive and do less.

Jared November 26, 2013 at 9:15 am

Thanks for the healthy dose of realism, Alex. As a 23andMe data-loving customer, I’m just as outraged.

dearieme November 26, 2013 at 9:20 am

The NSA must be spitting (!) tacks. It was about to get DNA data on zillions of Americans and now that hope is dashed.

Unless the point is just to delay things until they’ve got company’s computers thoroughly hacked.

Rahul November 26, 2013 at 9:23 am

“it’s no different than the FDA trying to regulate what my doctor can tell me after a physical examination or what my optometrist can tell me after an eye examination”

It’s different. Your optometrist would probably get into trouble if he used your eye exam to predict your fortune. Or if your doctor started reading your palm.

The underlying point is the line separating fraud from accepted science.

Axa November 26, 2013 at 10:01 am

Here is the problem, in 23andme the marketing department is stronger than the scientific department. Certainly, “By determining your carrier status for mutations linked to 40+ inherited conditions, we’ll give you and your family knowledge to help you prepare for the future.” and similar texts along the page does not invite you to think in science, statistics or the classic “results may vary”.

https://www.23andme.com/health/benefits/#risks

Compare it to a commercial pregnancy test like this one: http://www.firstresponse.com/early-pregnancy-test.asp You get those boring footnotes that say ” Women’s Health Journal November ’09 Publication: Embryonic Development and Pregnancy Test Sensitivity” a reference to scientific journal and “*In one study, in 68% of the samples tested, pregnancy could be detected up to 6 days before the day of the missed period.” so, the advertised product has a 68% efficacy, sadly not making a distinction between false positives or negatives.

Going back to 23andme, the sold product may be awesome and science backed, but the way they sell it is like 100% true for 100% of people, in other words, a LIE.

john personna November 26, 2013 at 10:08 am

Related, I noticed this headline on Hacker News: My deadly disease was just a [23andme] bug

john personna November 26, 2013 at 10:09 am

(Come on folks, doing a proper html a-tag is not hard.)

Cliff November 26, 2013 at 11:00 am

Totally wrong. The actual product provides way more detail then your pregnancy kit, including links to the actual studies.

john personna November 26, 2013 at 11:17 am

Is this a case of detail not equaling accuracy? I mean, take the “percent Neanderthal” mentioned in this thread. As I understand it, the underlying science has changed, and the percent is not that straight-forward.

Tracy W November 27, 2013 at 4:50 pm

But the underlying science changing is likely to always be true, no matter how much regulation there is. (Well unless you manage to regulate so much you stop all future relevant scientific discoveries.)

edwardseco November 26, 2013 at 10:00 am

Me too:)..

albert magnus November 26, 2013 at 10:02 am

It’s weird you would call her ‘daring’ when she is (or was) married to one of the richest men in the world. Is 23andme even worth 1% of Brin’s Google net worth?

derek November 26, 2013 at 10:18 am

I wonder if the FDA has it’s own swat team yet.

Brandon November 26, 2013 at 10:29 am

SWAT? They’ve got their own, better version of SEAL Team Six. It’s like Team Six and Delta combined, but only more awesome.

derek November 26, 2013 at 11:40 am

Would you change your mind about the FDA if they did have swat teams?

Brandon November 26, 2013 at 12:16 pm

No, I would not, but I would think they would be better off relying on other agencies (state, local or federal) rather than having a redundant function.

Careless November 28, 2013 at 1:52 pm

Department of Education does, why not the FDA?

Seb November 26, 2013 at 10:41 am

No joke, the FDA does have armed agents which they use to intimidate Amish farmers.

http://communities.washingtontimes.com/neighborhood/omkara/2011/dec/20/fda-escalates-war-against-amish-dairy-farmers/

Brandon November 26, 2013 at 10:56 am

That article doesn’t specify that any FDA agents were armed, and I can’t find anything but nutty and/or right-wing paranoia over “Armed FDA raids!” on google.

Seb November 26, 2013 at 11:27 am

Brandon, you did not read very carefully

“After a two year expensive, exhaustive undercover operation, including multiple armed raids on Allgyer’s farm, FDA agents and a team of ten federal lawyers ….”

Brandon November 26, 2013 at 12:20 pm

You should have finished that sentence:

“After a two year expensive, exhaustive undercover operation, including multiple armed raids on Allgyer’s farm, FDA agents and a team of ten federal lawyers amassed over two hundred and seventy-six pages of evidence allegedly proving what Allgyer openly admits, that he is selling fresh (raw, unpasteurized) milk to customers who knowingly carry the milk across state lines.”

FDA agents and a team of ten lawyers amassed a bunch of documentation. Prior to that, armed raids were conducted. What isn’t clear from this story is who conducted those armed raids. The results I got from google, even though they all seemed to be more conspiracy-minded sites, all included language like “FDA agents and local police” or “FDA agents and US Marshals.” I couldn’t find anything that clearly indicated that FDA agents themselves are armed. I will also note that based on both the publication and the language within the article, your source appears to be far from impartial.

Seb November 26, 2013 at 1:55 pm

Seriously Brandon? It’s not a secret that the FDA has armed agents. Look at their own web page

http://blogs.fda.gov/fdavoice/index.php/2013/07/fdas-special-agents-on-the-job-to-protect-the-public/

“Each special agent undergoes specialized training to be effective in their job, including firearms and personal safety training, advanced Special Agent training, and training in FDA law.”

Brandon November 27, 2013 at 10:50 am

I didn’t say it was a secret, I asked for confirmation because google just turned up a bunch of fringe/obviously ideologically slanted websites. The article you posted could be read to imply that they were armed FDA agents, but a few of the other stories I came across indicated that there were definitely other armed agencies involved and didn’t make it clear whether the FDA was armed.

Thank you for the link to the FDA’s website that makes it explicitly clear. I don’t know why the FDA needs its own armed enforcement agents, but perhaps it has to do with jurisdictional or budgeting issues–is the FBI really going to want to spend their resources to support the FDA’s needs, or are they more likely to funnel it into their own operations.

Either way, thanks again.

Eric Falkenstein November 26, 2013 at 10:18 am

I’m 2.7% Neanderthal!

It’s worth noting that our understanding of how genes relate to phenotypes is so limited, even something as objective as height, which is 90% genetic, they can only explain 16% of the variance using about 100 genes. So, if the FDA tells them they need to be really sure before they tell you anything about, say, Alzheimer’s risk, or risk of uterine cancer, they won’t be able to say anything but your sex and eye color. It’s hard for a market to generate incentives if the product has to be that much better than it currently is to be legal.

Brandon November 26, 2013 at 10:30 am

How well does 23andMe communicate the uncertainty over the predictive power of the results they give you? Isn’t that really the central problem?

Keith November 26, 2013 at 11:00 am

They say things like “your chance of being obese is 4% higher based on studies on people with this allele”.

Eric Falkenstein November 26, 2013 at 11:17 am

for health risks, they say stuff like the following. Here’s my biggest risk

“Age-related Macular Degeneration”
confidence: 4 stars
your risk: 20.8%
avg risk: 6.5%
compared to average: 3.18

I click and they note 5 markers (SNPs?) relevant to this diagnosis, and show in a bar chart the individual odds associated with each

I can click for further information and get even more info on each marker. In this case, my biggest bad SNP was : “rs1061147″ where I have genotype AA, and this increases my odds by 2.76. 6 studies are presented below. They have this for all 5 markers.

I don’t really know how they are combining markers to get a total result.

I think they are giving simple stats averaged from a variety of peer-reviewed studies.

In another case, ALS, they note that the average risk is 1 in 100k, and my risk is 30% higher! I’m not too frightened by this, but I’m sure the FDA might think I’ll now kill myself to avoid a life in a wheelchair.

For Alzheimer’s and Parkinson’s they have extra screens telling you to ‘calm down this isn’t definative’ before you see your result, and it’s all just annoying like those first 3 pages on warnings in any Ikea assembly guide, but I assume the FDA will make that standard.

mulp November 26, 2013 at 10:36 am

23andme was pushing the limits by claiming to offer medical diagnosis.

It was clear to everyone this was the case.

One can only assume they wanted to establish the basis for court cases on free speech grounds, or similar.

If not, then they entered the late night get rich quick and cure baldness infomercial realm to get rich quick.

wjca November 26, 2013 at 2:48 pm

It seems the testers have a very simply work-around: Don’t say “You have xxxx gene which causes yyyy medical condition.” Just say “You have xxxx gene which is correlated with yyyy medical condition.” Proving causation is, as noted, extremely expensive. Demonstrating correlation is pretty simple — and hard to argue.

Careless November 28, 2013 at 1:55 pm

Which is exactly what they did.

Arun November 27, 2013 at 11:37 am

As a customer of 23andMe, I strongly disagree with the claim that 23andMe was claiming to offer medical diagnosis.

Abelard Lindsey November 26, 2013 at 10:57 am

The FDA should be abolished. It is a crime against humanity. The bureaucrats running it should be prosecuted for crimes against humanity in “Nuremburg”-style trials.

Brandon November 26, 2013 at 11:01 am

can’t tell if serious or…

Randall Parker November 29, 2013 at 12:28 am

Milton Friedman on the US FDA:

“The FDA has done enormous harm to the health of the American public by greatly increasing the costs of pharmaceutical research, thereby reducing the supply of new and effective drugs, and by delaying the approval of such drugs as survive the tortuous FDA process.”

The argument is that the FDA has therefore killed many more people than it has saved.

jmo November 26, 2013 at 11:12 am

Snake oil for everyone!

Yancey Ward November 26, 2013 at 11:00 am

Under no rational argument can the code itself be embargoed by the government. A simple solution that should satisfy the supporters of the FDA, but probably won’t, is to allow the people to get their results, but get the interpretations elsewhere.

mavery November 26, 2013 at 12:21 pm

On the contrary, I think this would go beyond what the FDA is looking for.

Yancey Ward November 26, 2013 at 1:45 pm

Then what is the FDA looking for?

Yancey Ward November 26, 2013 at 11:04 am

Every single academic paper published linking certain genetic characteristics to disease should have to be approved by the FDA prior to publication if you really do believe that the FDA can prevent 23&Me from interpreting the results of the genetic code sequencing.

Brandon November 26, 2013 at 11:16 am

How does that follow?

Yancey Ward November 26, 2013 at 1:25 pm

23&Me is decoding and interpreting the results for customers. I gather the main contretemps is over the interpretation part. How can the FDA prevent 23&Me from doing this while allowing anyone else to do this?

David Wright November 26, 2013 at 4:56 pm

That is the way Alex presents it, but if you read the FDA C&D letter he links to, it’s pretty clear that for the FDA it is very much about the decoding part, namely that 23andMe has not been responsive to the FDA’s attempts to determine its level of reliability.

Now perhaps this is just a front for a regulatory power-grab, but Alex has provided no evidence of that I’ve got to say his misrepresentation of the content of the C&D letter doesn’t help.

Brandon November 27, 2013 at 10:52 am

You can read the FDA’s C&D letter to find out exactly what their problem is. If your understanding of their claims would lead to a result where basic scientific research would need FDA approval before publication, that should be a big, red flag that your understanding is wrong.

Ryan Vann November 26, 2013 at 11:24 am

So a bunch of Google Glass donning dorks want to get their genome mapped out… What is the controversy now?

“Under no rational argument can the code itself be embargoed by the government. A simple solution that should satisfy the supporters of the FDA, but probably won’t, is to allow the people to get their results, but get the interpretations elsewhere”

The rational argument is “it’s the FDA’s prerogative man. Legal positivism, broski.” Now, that may seem nonsensical to you, but a lot of people see it and think “it makes perfect sense.” Your suggestion while perfectly legally deft, doesn’t address the prime motivator here, which is that Uncle Sam Corleone the local FDA quality assurance don wants his cut. A much simpler solution would be for Google to tell the FDA to pound sand, which they have the power to do.

“The FDA should be abolished. It is a crime against humanity. The bureaucrats running it should be prosecuted for crimes against humanity in “Nuremburg”-style trials .”

Eh, no need for such extremes really. I mean the FDA is already smacking something of a nanotech-beehive here. I see waves of DDOS attacks on the FDA incoming soon. If I’m a schoolyard bully in meat world, I kick sand in the face of the dork getting 23andME testing. If I’m a bully in modern tech world, I don’t piss off hordes of dorks getting 23andMe testing.

Doug November 26, 2013 at 11:31 am

This is a hit job by United Health Group. They are scared. The FDA only took action after United Health sent a very direct letter to them asking them to, ahem, look into the company.

Bill November 26, 2013 at 12:03 pm

No, it was the Bolshevik’s riding blue helicopters.

The FDA was totally unaware of this product, even though they had to file. You know, those lazy bureaucrats wouldn’t read a filing, even if they were required to.

If it weren’t for Alex pointing this out, the sun would not rise tomorrow morning.

Ray Lopez November 26, 2013 at 12:04 pm

Boring. What you folks don’t realize is that government extends into every facet of your life already. For years now the Patent Office has refused to issue patents that claim magnetic bracelets ‘cure’ or otherwise alleviate human disease. And the FDA will not allow you to market the same. It’s bureaucracy run amok, in the name of protecting the sub-80 IQ crowd. Ditto that in Maryland it’s a crime for a fortune teller to say that a hex or curse has been put to you, and that the fortune teller can lift the curse. Again, protecting the imbeciles of society from themselves, and clamping down on the liberties of everybody as a result. As for 23andMe, as another poster has said upstream, their tests are not all that accurate anyway, so no big loss. The rich will get their DNA tested regardless of what the FDA says, even if they have to travel outside the USA; it’s the proletariat that get regulated.

prior_approval November 26, 2013 at 12:28 pm

‘clamping down on the liberties of everybody as a result’

So let me get this straight – you want to have the right to pay someone to say a curse is lifted?

I’ll do it for free – the curse is lifted. There you go – freedom on the march. With an extra helping of free lunch to boot, since I don’t charge for such a valuable service.

Careless November 28, 2013 at 3:55 pm

Such bracelets are marketed and sold.

FC November 26, 2013 at 2:14 pm

So acupuncture is covered under ACA but mail-order genetic testing is forbidden to everyone. Right.

Ethan November 26, 2013 at 2:36 pm

The government is a giant busy body baby sitter that has no idea what it’s doing.

David Wright November 26, 2013 at 2:45 pm

If you read the FDA’s cease and desist letter to which Alex’s post links (http://www.fda.gov/ICECI/EnforcementActions/WarningLetters/2013/ucm376296.htm) you will see that Alex’s claim that “the FDA wants to judge not the analytic validity of the tests, whether the tests accurately read the genetic code as the firms promise (already regulated under the CLIA) but the clinical validity, whether particular identified alleles are causal for conditions or disease” is simply wrong. The FDA letter very clearly states two examples of their concerns: “For instance, if the BRCA-related risk assessment for breast or ovarian cancer reports a false positive, it could lead a patient to undergo prophylactic surgery, chemoprevention, intensive screening, or other morbidity-inducing actions, while a false negative could result in a failure to recognize an actual risk that may exist.” and “For example, false genotype results for your warfarin drug response test could have significant unreasonable risk of illness, injury, or death to the patient due to thrombosis or bleeding events that occur from treatment with a drug at a dose that does not provide the appropriately calibrated anticoagulant effect.” Both of these scenarios concern false genotype readings, not false claims of phenotypical correlations.

Perhaps Alex has some additional information that leads him to believe that the FDA’s concerns and corresponding demands go beyond what is expressed in their letter. If so, he should provide that information, because as it stands it really looks like he is just trying to stir up the “regulator-as-jack-booted-thug” pot (and from some of the comments here it appears he has been successful).

Anthony November 26, 2013 at 3:06 pm

So instead, a patient may take Warfarin in complete ignorance of whether they will suffer the serious side effects? Or do doctors routinely order their own tests before actually prescribing Warfarin or scheduling prophylactic surgery? If doctors don’t, why not?

David Wright November 26, 2013 at 3:44 pm

If read up on it, you will Warfarin dose is usually determined by a quite onerous series of tests which measure the body’s response to various trial doses. This is done because different people need very different doses and getting the dose wrong can be very dangerous (see e.g. Ariel Sharon). There is some hope that genetic testing could make the dosing easier.

Suppose this hope panned out, and for simplicity of argument imagine it panned out in the simplest possible way: there are two genetic variants; if you have variant one you need the high dose, if you have variant two you need the low dose. Suppose furthermore, again for simplicity of argument, that the consequences of getting the wrong dose are equally simple and stark: if you wrongly get the high dose, you will die of an aneurism; if you wrongly get the low dose, you will die of a heart attack. Now someone offers a test for the two gene variants. Don’t you think it’s reasonable for the medical regulator to want to know what the false determination rate is?

It’s probably true that no physician is going to abandon the standard Warfarin dosing protocol, and no surgeon is going to schedule a double mastectomy, just because of what a 23andMe report says. But that is a different (and probably much better) argument than the one Alex is making.

Anthony November 26, 2013 at 3:04 pm

Those of you supporting the FDA position: would you support the FDA regulating hardware stores if they sold scales and tape measures along with a fact sheet about how to calculate BMI and the various health risks at different BMI levels? If not, how is what 23andme doing any qualitatively different?

David Wright November 26, 2013 at 3:49 pm

To try to pursue the analogy, if people built houses floors to just barely support their weight, then the accuracy of scales would be an important concern of health and safety regulators. The likely consequences of getting it wrong does make a difference in determining level of regulation justified.

mike November 26, 2013 at 4:04 pm

I don’t understand your analogy, but I think it successfully conveys your worldview of government as an omnipotent, omnipresent, and omniscient entity that micromanages the universe down to every last subatomic particle for the ostensible benefit of humanity. The only question is, do you literally worship it?

Urso November 26, 2013 at 4:19 pm

Wow, you got all that from two sentences? Impressive. Now do me!

David Wright November 26, 2013 at 4:23 pm

And your mis-reading of me — a card-carrying small-government libertarian who will talk endlessly about how the FDA drug approval process and the AMA cartell kill people and increase medical costs — conveys to me your elevation of mood affiliation over careful argument.

Alex (and several other libertarian commentators, e.g. Nita Farahany at the Volokh conspiracy) claim that the FDA wants to regulate this because of what consumers might do if they know their true genetic profile. But all evidence is that the FDA wants to regulate this because of what consumers might do if they get a false genetic profile.

mike November 26, 2013 at 6:37 pm

no, I think what it means is that the small-government libertarian club needs to take away your membership card

Brandon November 27, 2013 at 10:53 am

Mike,

Do you oppose building codes that mandate minimum structural capacity?

David Wright November 26, 2013 at 4:18 pm

And your mis-reading of me — a card-carrying small-government libertarian who will talk endlessly about how the FDA drug approval process and the AMA cartell kill people and increase medical costs — conveys to me your elevation of mood affiliation over careful argument.

Alex (and several other libertarian commentators, e.g. Nita Farahany at the Volokh conspiracy) claim that the FDA wants to regulate this because of what consumers might do if they know their true genetic profile. But all evidence is that the FDA wants to regulate this because of what consumers might do if they get a false genetic profile.

David Wright November 26, 2013 at 4:19 pm

(Sorry, that was supposed to be a reply to Mike in my thread above.)

andy weintraub November 26, 2013 at 6:56 pm

I haven’t read all the comments, but the FDA’s decision could be part of a plan to reduce the demand for medical attention or care under Obamacare. If we don’t know to what diseases we’re pre-disposed, we won’t be asking our medical professionals for advice and counsel, preventing the demand for those services from growing under a regime of price controls and rationing.

Therapsid November 27, 2013 at 12:19 am

Thanks andy, this is one of the most perceptive comments in this discussion.

We’re definitely seeing a push from the administration, the FDA, the Department of Health and Human Services, etc. against aggressive and expensive treatments and high-tech medicine.

Effectively they want the federal government to ensure a Great Stagnation in medicine.

Robert Reich has been blunt about this. A few years ago at UC Berkeley he gave a hypothetical speech which a candidate for president would give if he were brutally honest, in which he said: “Also I’m going to use the bargaining leverage of the federal government in terms of Medicare, Medicaid—we already have a lot of bargaining leverage—to force drug companies and insurance companies and medical suppliers to reduce their costs. What that means, less innovation and that means less new products and less new drugs on the market which means you are probably not going to live much longer than your parents. Thank you.”

http://www.youtube.com/watch?v=IT7Y0TOBuG4

Bufflars November 27, 2013 at 7:47 pm

I’m not sure I agree with your hypothesis. In nearly all cases that I can think of, early detection and/or early awareness of disease risk factors will almost certainly result in a lower long term demand and cost for medical care. The vast majority of health care costs are spent trying to treat already-occuring chronic conditions, so it absolutely doesn’t follow that the government would want to try to purposely prevent early warning/detection techniques if they were worried about catching flak for “rationing”.

In fact, if you think that “rationing” is the end goal of the ACA, you would think the government would actively be encouraging early detection of risk factors for chronic ailments, so that the long term incidence rate of (demand for) such ailments goes down… Am I missing something here?

Hein November 26, 2013 at 9:08 pm

My family shipped our kits in the mail yesterday.

Hopefully we got in under the wire.

This is very sad news.

dve November 27, 2013 at 3:41 am

My advice (apologies to Vaclav Smil) : go offshore and let them regulate sending spit through the mails.

Side benefit: increase earnings at USPS, UPS, Fedex.

KevinH November 27, 2013 at 12:36 pm

I don’t know if my comment will be effective, but you’ve gone off the deep end Alex.

“Alternatively, firms may be allowed to sequence a consumer’s genetic code and even report it to them but they will not be allowed to tell consumers what the letters mean. Here is why I think the FDA’s actions are unconstitutional. Reading an individual’s code is safe and effective. Interpreting the code and communicating opinions about it may or may not be safe–just like all communication–but it falls squarely under the First Amendment.”

So, I could tell you that my magic potion (just $99.99) would cure your cancer and the FDA would have no justification for shutting me down? It is clear, well precedented and sensible that the FDA has the power to regulate speech once you are claiming a fact that effects your health. Right now, people are choosing NOT to get mammograms specifically because 23andMe claims they have a reduced chance of getting breast cancer, based upon gene correlations that no member of the scientific or medical community has access to. These recommendations by 23andMe are not publicly accepted science, but rather fall under ‘proprietary information’. Please show a clear line between this and snake oil.

By your own words you admit that 23andMe is still able to provide the true underlying value, the identification and dissemination of someones personal genetic code. This is a public good, and not prohibited by the FDA. However, that isn’t 23andMe’s business model. Their business model is to provide medical advice. Just because they use a new technology doesn’t mean they should follow the same rules and regulations that everyone else has to follow.

Doug November 27, 2013 at 3:30 pm

“So, I could tell you that my magic potion (just $99.99) would cure your cancer and the FDA would have no justification for shutting me down?”
23andme does not claim that their service cures anything.
“Right now, people are choosing NOT to get mammograms specifically because 23andMe claims they have a reduced chance of getting breast cancer”
“These recommendations by 23andMe are not publicly accepted science, but rather fall under ‘proprietary information’”
Where has 23andme ever recommended to anyone that their test is a suitable replacement for getting mammograms? Or that if they have a certain results they do not need mammograms? I don’t think they are making any recommendations at all, they are only providing information.
“Please show a clear line between this and snake oil.”
Your the one who seems to be claiming it is the same as snake oil, so you are the one who needs to show the clear line. So far, I don’t see it.
“Their business model is to provide medical advice.”
What advice do they provide? Based on Eric Falkenstein’s comment above, it looks like they only provide information regarding which genes are linked with increased risks of certain diseases. I have not seen or heard of them offering any medical advice as to what treatments people should seek or not seek based on that information.

Brandon November 27, 2013 at 3:56 pm

their marketing materials describe their service as an important first step in determining what conditions you’ve got an increased likelihood to have. It’s marketed as a tool to guide your health care decisions. They may not be offering treatment, but they seem to be sufficiently close to explicit diagnosis that it’s a legitimate FDA concern. They have a disclaimed with the usual language, but their marketing elsewhere for their product gives a very different idea of what the purpose of the product is.

Doug November 27, 2013 at 5:29 pm

How do you define “sufficiently close to explicit diagnosis that it’s a legitimate FDA concern.” What about any of the thousands of websites out there that talk about how non-genetic symptoms indicate an increased likelihood of particular diseases? The bottom line is if you are not actually diagnosing anyone has having or not having any particular disease, and you are not advising anyone to take or forego a particular treatment, then aren’t your words free speech?

Randall Parker November 29, 2013 at 12:36 am

The web is full of web pages with information about how to diagnose whether you have some disease. This information is used by millions of people to guide their health care decisions. Should these web pages be allowed to remain?

Since we can’t get drugs to lower our blood pressure or double masectomies without going to doctors I’m unclear on what the damage is supposed to be from using 23andme genetic testing results. If anything the results are more likely to make one consult a doctor. But then the doctor still has all the legal power to dole treatments or not. So where’s the concern?

Ape Man November 27, 2013 at 4:09 pm

The big irony that is being missed by everyone is that if the FDA held doctors as a group to same standards they are asking of 23andMe, they would have to shut down all doctors. Most doctors do not provided very accurate advice even allowing for the fact that the best medical advice changes over time. http://www.nytimes.com/2012/06/05/health/views/essay-urging-doctors-to-do-less-may-fall-on-deaf-ears.html?_r=0

Jason Braswell November 28, 2013 at 2:25 pm

I can’t believe anyone would defend the FDA here. Makes me angry and sad.

Blogospheroid November 28, 2013 at 11:01 pm

According to Balaji Srinivasan’s lecture, Silicon valley has been strikiing at the dominance of the other cities, including Washington DC. Well, looks like Washington DC just struck back.

Ingrid December 9, 2013 at 6:22 am

This paragraph will assist the internet visitors
for building up new blog or even a blog from start to end.

Comments on this entry are closed.

Previous post:

Next post: