The FDA and Personalized Genetic Testing

The company 23andMe provides personal genetic testing from a convenient home saliva sample kit. Their home page indicates that their $99 genetic screening will provide reports on 240+ health conditions in addition to giving you information on your genetic lineage. The benefits, they claim, are that you will learn about your carrier status and therefore the risk of passing on genetic diseases to your children. The information will also inform you about health risks so that you can change your behavior to manage them. Finally the genetic information will tell you about how you might respond to different drugs so that you can “arm your doctor with information.”

The home page also contains a link to testimonials about how their DNA testing has changed people’s lives.

This all sounds great – the promise of genomics that we have been hearing about now for about two decades. Isn’t this exactly what we were led to expect once we mapped the human genome?

Why, then, has the FDA recently sent a warning letter to 23andMe instructing them to discontinue marketing their Personal Genome Service (PGS)?

The primary reason appears to be a lack of documentation for the validity and reliability of the tests, but I have concerns that go even deeper.

The FDA is concerned mostly with whether or not the genetic testing offered by 23andMe works as advertised. False positive and false negative testing could have significant medical consequences. The company is marketing its PGS specifically to aid in medical management, including drug management. Despite the fact that the website says “arm your doctor,” the FDA fears people will use the information to self-manage, for instance by adjusting their own dose of critical medication.

The FDA is also particularly concerned about genetic tests such as those for BRCA, which conveys risk for breast and ovarian cancer.

In other words, there are serious medical consequences to the genetic information being provided by 23andMe and they have not performed the due diligence required to make sure their tests are valid and reliable. The FDA letter documents their attempts over the last five years to work with the company to help them come into compliance, but to no avail.

I can only speculate that the company is not complying with the FDA because they cannot – they do not have the scientific information the FDA requires and they are dragging their heels on producing it. They may simply be trying to avoid the cost of such research, and hoping the FDA will eventually back off (perhaps they were hoping to hold out for an administration change). The FDA, however, is being tenacious.

I certainly like to see the FDA aggressively doing its job of protecting the public from unsubstantiated health claims and from products and services that are not adequately backed by scientific studies and proper quality control. The public expects that this is the case.

However, I also always marvel at the irony such cases highlight. Articles discussing this and similar cases often include quotes such as the following:

“Our society regulates medical products to protect public health. Without strong public oversight, we’re back to the era of snake oil,” said executive director Marcy Darnovsky. (of the Center for Genetics and Society)

I completely agree. Now let’s talk about the entire homeopathy industry, the lack of even basic quality control in the supplement industry, and the pathologically false health claims routinely made within the world of CAM (complementary and alternative medicine).

While I agree with the FDA going after 23andMe for noncompliance, I have to wonder why it has picked this company out of all the possible targets out there. I know that much of it has to do with their regulatory scope. They cannot take on homeopathy, for example, without Congress changing the law. But I can think of many other suitable targets that seem to be getting a free pass. Perhaps it simply comes down to the fact that they are not adequately funded to do their job given their regulatory limitations.

As I alluded to above, I have issues with personalized genetic testing that go beyond the stated concerns of the FDA. Even if the company provided data showing that its genetic tests were accurate, valid, and reliable, that does not mean that they are necessarily a good idea. The deeper question is – is the information itself useful, or will it perhaps have negative unintended consequences?

We have discussed the pitfalls of screening tests on SBM before. Superficially, screening tests seem like a good idea with no downside. More information is always better, right? That is certainly the approach that 23andMe is taking (of course, their business model involves providing that information).

However, there are potential issues with any screening test. Some do relate to sensitivity and specificity – how reliable are the tests. No test is 100% sensitive or specific, which means that even very good tests will have occasional false positives and false negatives. If the entity you are screening for is rare enough, then even a very accurate test may still provide more false positives, for example, than true positives.

This issue aside, meaning even if we had a 100% accurate test, we have to ask the further question – what is the net effect of providing this information? Potential negative consequences of accurate screening information include the unnecessary worry of learning that you have a predisposition or risk of developing a disease in the future. This may seem like a small thing, but it can have a significant impact on quality of life, even to the point of people being crippled by their fear of impending disease.

Further we have to consider what, exactly, will be done with the information. The implication is that drug treatment can be modified, and risk factors for disease can be addressed. But – is genetic testing (with our current knowledge base) the best way to approach such treatment? Will the genetic information cause people to change their behavior in a way that will have a net negative effect on their health?

Such behavioral changes may also include a false sense of security in those with negative screening for certain risk factors. If someone finds that they are in a low-risk group for heart disease, will they then feel free to engage in an unhealthy diet that will still be a risk factor for them?

Screening indicating a high risk may also motivate people to seek out unnecessary health care, which may lead to unnecessary tests that have their own false positive and negative rates.

There are downstream consequences to providing health information to large numbers of people. The 23andMe website is providing only handpicked testimonials with smiling faces about how their lives were improved. We have to wonder about cases of negative effects from the information provided. The only way to know net health effects is to properly study not just the accuracy of the test, but its effect on the people who receive the results. The FDA seems concerned only about the former, and not the latter (which may be appropriate given its scope).

What we have to consider as a society is how to regulate private companies directly marketing health services to the public, bypassing medical professionals. This is part of a larger issue with the public having increased access to information in general, including their own medical information. In general I am in favor of increased information access, but we can’t do this blindly. We need to consider consequences and how to put into place systems that can help put such information into proper context and use.

For example – electronic medical records systems mean that my patients can often now directly access their test results. This is convenient, and they certainly deserve to have access to their own information. But what often happens is that the test results contain misleading noise – numbers that are just outside of the normal range, or MRI findings that sound scary but are actually insignificant. I have seen this cause unnecessary anxiety on the part of patients, and take up more staff and physician time than if the results were filtered through the physician in the first place.


The FDA’s current problem with the PGS of 23andMe is concerning, as it indicates that the company has not provided proper documentation that the services they provide are valid and accurate. There are problems with the service, however, that go deeper than simply accuracy.

Our society has increasing information and public access to information. While it is difficult for me to think that this isn’t a good thing overall, we have to thoughtfully consider the possible unintended negative consequences. This case is part of a larger pattern of sacrificing quality-control filters for the sake of open access. This increasingly puts the burden on the public to make sense of sometimes complex and technical information.
Everyone, now, can be their own geneticist.

Posted in: Public Health

Leave a Comment (79) ↓

79 thoughts on “The FDA and Personalized Genetic Testing

  1. daedalus2u says:

    The FDA is right to do this. Genome analysis doesn’t provide the kind of health information that people pushing it have promised. Differential genetics are simply not as important as other things in differential incidence of health and disease.

    Many people really “want” genetics to be all-important, but that desire is due to human hyperactive agency detection and the ego-driven want to “other” those who are different and to attribute inferiority to the “agent” that is imputed to make those “others” different, their genes.

    These two things have been driving eugenics and racism for millennia. Now that we have the technology to actually determine the actual importance of differential genetics in differential phenotype properties, the data says that genes are mostly not important in determining phenotype properties such as intelligence and neuropsychiatric disorders.

    1. windriven says:

      I share your revulsion at the misuse of genetics to fuel racism. But the issue this morning involves disclosure to the individual, not public disclosure.

      DNA testing and genetics in general are likely to claim an increasing share of the public dialog. Attempts to obscure or withhold are likely to ultimately fail. As with many issues, education and vigilance against those who would twist information to fuel hatred is the best approach – IMHO, of course.

    2. Julian says:

      Sorry, genes are not important in properties such as intelligence? This is ridiculous. There is a huge amount of twin studies (identical twins vs fraternal twins). The ONLY real statistical difference in large numbers between these pairs is genetics. And you can clearly see that genes contribute a huge amount to traits such as intelligence. We might not yet know which exactly, that does not change the fact that we can clearly show genetics has a huge impact.

  2. windriven says:

    “Despite the fact that the website says “arm your doctor,” the FDA fears people will use the information to self-manage, for instance by adjusting their own dose of critical medication.”

    I’m all for the FDA insisting that 23andMe and every other provider of products and services prove the safety and efficacy of their products. But I am deeply opposed to FDA or Dr. Novella or anyone else telling me what information I am entitled to have about my own body; this cuts right to the heart of agency.

    Moreover, testing information generally requires the participation of a physician when action is to be taken. Assume for a moment that I show a BRCA2 mutation that is associated with a high risk of breast cancer. The chances of me sharpening up the kitchen knife and slicing off my breasts are remote.

    True, there are test results that can create a false sense of security or risk but that is what IFUs are for. And if someone doesn’t read the IFU and decides that it is OK to eat a stick of butter for breakfast each day, well, we’re back to the issue of agency. How stupid is too stupid?

    There are lots of activities that people ‘shouldn’t do’ based on a particular perspective. People shouldn’t smoke. We’ve made it difficult and expensive for those who do, but they still are able if they insist.

    Just as there are dangers in giving people information, there are dangers in stripping away personal responsibility with the assumption that some one or some agency will protect them or decide for them.

    1. MadisonMD says:

      I’m all for the FDA insisting that 23andMe and every other provider of products and services prove the safety and efficacy of their products. But I am deeply opposed to FDA or Dr. Novella or anyone else telling me what information I am entitled to have about my own body; this cuts right to the heart of agency.

      Perhaps we could agree on this:
      (1) You are entitled to your DNA sequence from any company that can legitimately demonstrate that they have quality controls in place such that the sequence information you receive is accurate and is yours.
      (2) If you want the sequence information medically interpreted, there should be scientific evidence demonstrating that the interpretation is accurate.

      So, the FDA would have to regulate both. Lower bar for #1, but there wouldn’t be much of a market to sell raw sequence information. Even #1 requires FDA regulation. Otherwise some miscreant can just throw out your cheek swab and send you a DNA sequence; or, less maliciously, mix up samples because of poor quality control. How would you know?

      So Windriven, I think the FDA preventing your access to misinformation is in the same league as preventing you from receiving quack medicine. I’m for it.

      1. windriven says:


        We would agree on those two items completely. The problem, of course, is with number 2.

        ” there should be scientific evidence demonstrating that the interpretation is accurate”

        You and I would agree on a framework for establishing accuracy. But the Universal Society of Quantum Homeopaths might argue that their interpretation, while wildly different from anything that you or I might consider coherent, is just as valid. And if Orrin Hatch or Tom Harkin had anything to say about it (and they would), we would likely have to live with it.

        On a more serious note the line between what is and is not misinformation is not always sharp. If I (accurately) tell you that you have a particular allele but do not tell you that you have another allele elsewhere that impacts the expression of the first, have I given you misinformation? What if the second gene was not in the ‘package’ of info that you purchased?

        1. MadisonMD says:

          You and I would agree on a framework for establishing accuracy.

          There is such a framework for testing by the FDA. Usually it is a lower bar than the framework in place for testing drugs–and perhaps rightly so because the latter is an intervention. That framework would probably weed out such things as electrowindrivenology, but would allow direct-to-consumer tests that are validated with known accurace (e.g. pregnancy test). I guess we could quibble about exactly where to draw the line.

          On a more serious note the line between what is and is not misinformation is not always sharp.


          If I (accurately) tell you that you have a particular allele but do not tell you that you have another allele elsewhere that impacts the expression of the first, have I given you misinformation? What if the second gene was not in the ‘package’ of info that you purchased?

          Interesting questions. But right now, except for a few highly penetrant single-gene diseases, the science of interpreting results is rather crude– I don’t know of any examples of this today. Do you?

          My impression is that most of the highly penetrant genes (where science is not so crude) are rare enough that false positives will be common if an unselected (screening) population is tested. So, it is an open question whether the testing should be done on anyone who wants it (accepting a high false positive rate), or if specific tests should be done only in the setting of suspected disease (increasing baseline prevalence of tested population, decreasing false positive rate). This problem would be a non-issue if 23andme or another companies demonstrate an outstanding receiver-operator curve.

          1. windriven says:

            @Madison MD

            Not much daylight between your position and mine. In a perfect world I would advocate for nearly unbridled access to this sort of information but we live in a far from perfect world.

            “But right now, except for a few highly penetrant single-gene diseases, the science of interpreting results is rather crude– I don’t know of any examples of this today. Do you?”

            I don’t. And this is very far indeed from a field in which I would claim any expertise.

    2. Michele Manion says:

      I don’t think anyone is suggesting you shouldn’t have access to info about your own body, but 23andMe is performing diagnostic testing and providing medical information, activities clearly under the purview of the FDA. The reason this is important is that you also deserve to have accurate info about your own body. With no oversight, the quality of data being generated by these companies is suspect.

      Here is a link to the GAO report from 2010 addressing the quality issue that you might find interesting:

      1. windriven says:

        @Michele Manion

        “I don’t think anyone is suggesting you shouldn’t have access to info about your own body”

        I was reacting to Dr. Novella’s:

        “Even if the company provided data showing that its genetic tests were accurate, valid, and reliable, that does not mean that they are necessarily a good idea.”

        Read that as you will.

  3. Calli Arcale says:

    io9 recently had an interesting article from a contributor who actually used one of 23andme’s tests and was given terrible news . . . which turned out actually to be a screwup on 23andme’s part, a software bug. It’s definitely worth a read.

    My deadly genetic disease was just a bug

  4. Chris Irwin Davis says:

    I am actually a member of 23andMe. Having now seen my own results, I have serious doubts about the about the ability of 99% of their customers to properly interpret their data.

    I have a Ph.D. in a science field that involves statistical modeling, so I don’t believe that I’m making a bold claim that I have above average skills in interpreting statistical data. And I have had some difficulty trying to objectively assess my own genetic information. After having my results for one full year, the only changes I’ve made are those that I could have (and should have) made irrespective of any genetic data. I’m eating better. I’m exercising more.

    1. David Gorski says:

      I am actually a member of 23andMe. Having now seen my own results, I have serious doubts about the about the ability of 99% of their customers to properly interpret their data.

      Only 99%? You have a far more optimistic view of people’s ability to interpret such data than I do. Very few are qualified to do it, and even among them many issues are controversial and they don’t always know what to do with the results.

      1. Michele Manion says:

        In our experience with a genetically complex rare disorder, it is not just 99% of the public who have trouble interpreting genetic reports. A majority of medical professionals unfamiliar with the disorder do, as well. The notion that genetic reports are black and white and risk can be easily gauged by looking at them is simply not true. This is a reality that calls into question what actual service these companies are providing.

    2. Denise says:

      I am also a member, and no scientist, with no PhD and no background in statistics. But I feel well able to understand how to use and not use the information. I joined out of ordinary curiosity and have found it very interesting, and it has prompted me to learn some things about genetics that I probably would not have otherwise. This has become an ongoing interest.

      The health information has given me a little more incentive to pay attention to a few things that, as you say, I should have been paying attention to anyway; as a practical matter it only found one elevated risk that I was not aware of, of blood clotting, and I think that is useful to know. It did not identify me as having above average risk for the autoimmune disease that I in fact have, but I am neither surprised nor confused by this; I understand that the disease has many genetic contributors, most of which have not been identified yet, as well as many environmental factors.

      I would certainly like the FDA to verify that the information given to me is correct. But I strongly resent any suggestion that I shouldn’t be allowed to have it. I think I should have any information I want about my own body, as long as I pay for it myself.

      1. Andrey Pavlov says:

        The issue Denise is that you think you understand how to use it and you think you learned something from it. But this is simply incorrect. Some people like Deepak Chopra also think they understand quantum mechanics, but nobody really understands it. To claim you do is simply ignorance of your own ignorance. And it is the same with these genomic data. You can’t possibly actually understand what they really mean and learn something from them because nobody actually understands what they really mean.

  5. Windriven – it’s not about what information you are entitled to have, it is about the responsibility of professionals who are giving you information.

    What if I order an MRI scan on a patient of mine who has severe anxiety. The MRI scan shows a terminal brain tumor. Should I call them up at home when they are alone and give them the full information, including a hopeless prognosis? They have a right to that information. At the very least I should impart that information in an ethical way, in a controlled setting, where I am available to answer questions and offer counseling, with some family support, etc.

    At present if I obtain a DNA test on a patient as part of research I am ethically not allowed to disclose the results of that test to the patient. That would violate regulations and I would get in trouble. The reason is because as a researcher it is not part of my role to make a clinical decision about the implications of the DNA test and to put it into that context for the patient.

    I can, however, disclose the results to the patient’s physician (with the patient’s approval), and they can decide if and how to disclose the information to their patient.

    There is a thoughtful and elaborate scheme for how to disclose genetic information to patients. This is meant to protect the patient – to protect their agency. All of this is bypassed when a private company simply sends genetic test results straight to consumers.

    1. windriven says:

      @Dr. Novella

      “I can, however, disclose the results to the patient’s physician (with the patient’s approval), and they can decide if and how to disclose the information to their patient.”

      Certainly you can see how one might read this and conclude that physician paternalism is alive and well.

      I have not followed the ethical arguments surrounding DNA testing closely. There is never enough time to explore everything that is interesting. But at first blush I have a difficult time wrapping my mind around the notion that the “thoughtful and elaborate scheme for how to disclose genetic information to patients” somehow protects their agency.

      If you or any of the SBM community can suggest some sources on the ethics of DNA testing I would be most appreciative.

      1. Here is an example – not everyone would want to know that they are destined to develop a horrible genetic disease later in life (like Huntingtons). Patients have a right to not know that information. So – before you tell a patient the results of a genetic test, you have to inform them about the implications of the test, and then let them decide if they want the information. Otherwise their free choice not to know the information (their agency) is violated.

        This is the primary reason why I can’t just disclose genetic information to study subjects, without being in the role of their treating doctor.

        1. windriven says:

          Excellent example, thank you. One can’t choose to ‘un-know’ something. I’ll certainly think through the implications.

        2. Johan Viklund says:

          Funny thing that. In Sweden, a doctor is required by law to give the patient all information they know. And I am pretty sure that the patient don’t have a right to not know this, the relevant laws don’t mention it, and I have never heard it in a swedish context (there are some exceptions, but a patients right to not know is not one of them).

        3. MS, MT(ASCP) says:

          There is also the problem of violating someone else’s privacy through disclosure, and I am in the middle of that ethical mess. I carry a known genetic feature that can cause autism, but I am not and never have been (as far as I know) on the spectrum. Three generations of people on one branch of my family have been tested, and a handful of individuals with the feature have been identified. I believe that there are others in another branch of the genetically related family who carry the feature, based on visible features associated with the genetic feature. My simple act of identifying that I have an inheritable genetic feature associated with autism, along with a basic understanding of Mendelian genetics allows me (or someone else) to make fairly accurate inferences about who else might have the feature. Anyone can do this despite their explicitly expressed desire to not know their status (we asked, and were given a curt “No.”). If I even casually communicate that 1) I have the inheritable feature, and 2) this is a physical trait associated with it, their privacy and agency have been violated. As a result, I have to be very careful and aware of what I say, and to whom, so that I do not violate the privacy and agency of those who have made it very clear that they do not want to know their status. The issues of 23andMe testing validation are significant, but do not represent the full problem. What to do with the information and how to protect others from implication is equally significant, and is like walking through a minefield.

          1. windriven says:

            Interesting ethical dilemma; another that I hadn’t considered.

        4. Lumen2222 says:

          I’ve taken the 23nMe tests, and this response you’ve given makes me wonder if you have yourself Dr. Novella. With specific test results the results are sealed and hidden precisely for this reason. You have to acknowledge that you do want to know the results for that particular terrifying disease. The results are designed so you can’t accidentally stumble across horrible information that you actually didn’t want to know.

          If you haven’t actually gone through the process of the test yourself then I’m a bit disappointed. You aren’t going to have a particularly accurate understanding of the flaws or potential just by reading someone else’s information.

          1. Sawyer says:

            You have to acknowledge that you do want to know the results for that particular terrifying disease.

            But this isn’t solving the fundamental ethical dilemma; it’s just kicking the problem up a level. That choice of withholding genetic information is still going to be made by someone with extremely limited knowledge of how to act on that information. While I like what 23andme is doing I think this is a pretty obvious problem, and you don’t need to go through the test to understand why it happens.

            As an example, check out Sam Keane’s book The Violinist Thumb. He goes through 23andme’s process, and is surprised how his attitude about receiving and withholding information changes over time. And this isn’t a random person on the street, it’s an author who probably has access to a lot of experts.

            On The Media did a segment on this topic a few weeks ago that also covers this dilemma.

        5. T. says:

          I am sorry if I am mistepping, but please bear with me.

          I still think people have an absolute right to know that kind of things if they wish. If you are trying to say that the results should be interpreted with the aid of a trained phisician, then I agree wholeheartily (say, raising the price to 500 $ and including a complete rewiev with a doctor with the correct specialization).

          But this is as far as I am going to go, in all honesty.

          (Apologize for my english, I am Italian :) )

    2. ravingdesi says:

      I agree that for your scientific study, a desire of the participants to know about derived information can be concluded. But why can’t they get the same data the physician would get if they insist to get it?
      The only explanation I can think of is because it’s not in a readily understandable format. But if it’s completely raw data, the standard physician will not be qualified to interpret it either, so my guess it you already interpret the data when giving it to the physician (I assume a primary care physician here).

      I know, most people will not be able to accurately handle that information, and I certainly know that my ability also has its limits. But it would hurt my soul to see that a woo embracing physician is judging what I am to be told (and yes, I have yet to found one primary care physician here in Germany who does not embrace some kind of woo!).

    3. Nathania says:

      As a cancer patient with severe anxiety, tell me the results (1) ASAP and (2) over the phone. My anxiety is wreaking havoc on my life as I wait for test results. The answer -even if it is a bad one, is better than uncertainty. Also, the second you tell me to come over to your office to discuss the results, I know it’s terrible. You want me to get in a car and drive to your office with that much anxiety? No, thank you.

      I was given my original diagnosis over the phone. I was able to cry in the privacy of my own home, and didn’t drive home like a deer in the headlights.

      Now, every year I’m waiting on results. Is this the year the cancer recurs? I click refresh on my digital records all day, despite the fact that I know I won’t get them until after the doctor calls me. Many times, I’ll need to call them. The results are in. They’re just too busy.

      The phone call lasts five minutes. I don’t remember half of what they say. Finally the record is available in my digital records. I have to do due diligence because doctors do miss stuff. (My dad had a nice chunk of brain tumor missed by one major hospital but picked up at another. He died less than a year later.)

      So with all due respect, let me have my results, the sooner the better. I’ll spend far more time with them than you will.

      As for the FDA, when they ruled that generic levothyroxine was just as stable as brand despite tons of scientific evidence to the contrary, they lost all my respect.

      I don’t believe for a second that they’re 23andme action was based out of concern for public health. They’re lobbied just as heavily as Congressmen, the White House, etc. They have tons of conflicts of interest.

  6. Harriet Hall says:

    I agree that people have the right to know their own medical information, but I am not a fan of these tests, because:
    1. If I had a high risk of a disease where there was nothing I could do to reduce the risk, all it could do is make me worry.
    2. If I had a high risk of a disease where I could take preventive action, that action is almost always just what all of us should already be doing (healthy lifestyle, exercise, diet, weight control, tobacco avoidance, etc.)
    3. If I had a low risk of a disease, that might falsely reassure me and I might be less motivated for healthy lifestyle actions.
    4. The tests don’t analyze your entire genome, only selected SNPs.
    5. All they actually do is compare your SNPs to those of other people and tell you your risk compared to theirs (like “your risk of Parkinson’s is 30% higher than average”). This may not accurately reflect your risk, especially if you are different from the test population.
    6. Low risk doesn’t mean no risk. High risk doesn’t mean you will get the disease. The presence of a gene doesn’t mean that gene will necessarily have an effect. What matters is the gene expression, which varies with environmental factors and with the effects of other genes. It’s complicated.
    7. “30% higher than average risk” is meaningless unless you know the base rate, the prevalence of the condition in the population. If the condition is rare, a 30% difference may be insignificant.
    8. The test can be misleading. In one example it said the person’s eyes were brown, when in fact they were blue. In another example, an African American wanted to know what part of Africa his ancestors came from, and he learned that he didn’t have a single drop of African blood: he was a mixture of Caucasian, Asian, and Indian ancestry.

    I simply don’t see any point in knowing the kind of information they provide. It mostly encourages navel gazing and it amounts to a sort of astrology with delusions of grandeur. It’s an example of premature marketing taking financial advantage of people’s excitement about a new development in science.

    1. David Gorski says:

      All they actually do is compare your SNPs to those of other people and tell you your risk compared to theirs (like “your risk of Parkinson’s is 30% higher than average”). This may not accurately reflect your risk, especially if you are different from the test population.

      I haven’t looked into this, but I wonder if companies like 23andMe report absolute risks in addition to relative risks. The implication of a 30% increase of the chance of developing a disease is much different if the baseline chance of developing it is, say, 20% than it is if the baseline chance of developing it is 0.1% In the former case, it’s the difference between a 20% chance and a 26% chance. Your chance is 6% higher. In the latter it’s the difference between a 0.1% chance and a 0.13% chance. Your chance is 0.03% higher, which for all practical purposes is indistinguishable from 0.1%.

      1. Josh says:

        Hi Dr Gorksi

        23andme reports both, for example on the summary ‘Health Risks’ page for each disease it reports it like this:

        Colorectal Cancer: Your Risk, 4%; Avg Risk: 5.6%; Compared to Avg: 0.71x

        Then if you click through to the page for that disease it reports it as a natural frequency:

        “4.0 out of 100 men of European ethnicity who share your genotype will develop Colorectal Cancer between the ages of 15 and 79.”

        “5.6 out of 100 men of European ethnicity will develop Colorectal Cancer between the ages of 15 and 79.”

        Then it estimates the relative contribution of genes vs environment:
        “The heritability of colorectal cancer is estimated to be 35% …”

        Then has information about screening tests, lifestyle changes recommended by the American Cancer Society, other information, and provides pubmed links to the genome wide association studies that the risk estimates for each SNP are derived from.

        1. Harriet Hall says:

          The information is there, but how do people interpret it? A prominent geneticist, Francis Collins, wrote a book “The Language of Life” in which he described the results he got from 3 different screening tests, one of which was 23andme. He seemed to fixate on the relative risks, and said that learning he had something like a 30% above average risk convinced him to change his lifestyle to decrease the risk. And he knows far more about this field than most scientists.

    2. Denise says:

      I joined, and I assure you it is not astrology or navel gazing to me – nor has it been a source of anxiety or of false reassurance. I understand the risk information and what it doesn’t mean. I’m not a child and I’m not stupid.

      I don’t see why you or anyone else needs to see the point in my having information about myself.

      1. Harriet Hall says:

        Just curious: if you understand what it doesn’t mean, what benefit has it been to you?

        1. Lumen2222 says:

          Isn’t curiosity and an interest in learning more about genetics enough? I find it baffling that so many doctors in this thread seem to be missing what seems so obvious to me. Genetics, just like any other subject becomes far more interesting when it’s personalized. 23 and me is attractive to a lot of curious people for the same reasons that some would want to keep our X-rays of our broken bones, or install iPhone apps that let us listen to our own heart beats. You don’t have to be a doctor to be curious and fascinated by how your own body works. So you don’t entirely understand all or much of the information…so what? When has that ever stopped people from wanting to try? The genie is out of the bottle on this one already and there is no way you’re going to stuff it back in with vague arm waving about how people might misuse the information or make bad choices out of ignorance. And really when has the answer to bad choices EVER been to restrict information?

          Frankly I find it disheartening that so many people I consider to be science educators appear to be hopelessly unprepared to tackle this topic. The arguments made for not allowing people free access to their genetic information are pretty much guaranteed to backfire into your faces. It smacks of paternalism and is precisely the kind of attitude that drives so many people into the arms of alt med screaming about patient choice. Props to Dr. Novella for trying to delve into the subject, but the only convincing concern I saw raised was the question regarding the accuracy and quality control of the tests. The rest of the article came across as: “You shouldn’t be allowed to easily obtain this information because it might upset you” coupled with “let someone like me decide what you need to know” I’m not sure when in human history those types of statements might have worked, but right now in THIS culture I’m pretty sure its probably going to have the opposite effect than what you are intending.

          1. Harriet Hall says:

            Curiosity is good, but misunderstanding is bad. The companies are giving the customers what amounts to misinformation about the meaning of these tests even if they are accurate and validated. No one advocates preventing people from getting information about their health. But SNPs that are chosen differently by different companies and that give different risk numbers are not as helpful as people would like to believe.

            1. Lumen2222 says:

              And I agree that this has great potential to be a problem. A company like 23nMe should have a moral obligation to ensure that the information being presented is clearly labeled and that they are not misleading people with regards to it’s practical use.

              HOWEVER, that is very different than saying that the information itself is dangerous, or to imply that anyone who is still interested in the service is irrational or does not understand enough to be allowed unfettered access and need to have their hands held lest they have a panic attack, or decide they are invincible and can eat twinkies all day while blaming 23nM. These are consequences that should be considered, however I haven’t seen any actual evidence presented that any real people actually responded to their results this way, other than Dr. Novella’s implication that the glowing marketing testimonials are somehow masking a dark (yet unproven) underbelly of negative consequences. So let’s be clear shall we? These are concerns. But there is not yet any evidence of a real effect.

              Perusing through the comments I see many negative reactions from people who have actually used the 23andMe service, and I have to say I often felt similarly while reading, and had to take a few breaths before I hit the reply button. I think it’s very important for you (as science educators) to consider why lay people might find these tests interesting, and not dismiss the excitement with statements about how they couldn’t possibly understand the information, absolutely don’t need it, and scathingly equate their curiosity to “navel-gazing”. (Them’s fightin’ words there).

              And for the record I appreciate that many people would not want to know if they had an incurable disease. However, I most certainly would. Some of us would rather fall face down and meet the end with eyes open. Another person’s desire to live in ignorance should not impede my desire to know. Surely there are solutions so we both might have our desired result?

              I had an absolute blast using 23nM, and after all my results were in I spent several hours pawing through the results, the vast majority of which were entirely unremarkable except for a very confusing (to me) mtDNA result. This in turn inspired several days of excited research into Maternal Haplogroups and my own family tree. Research I frankly wouldn’t have been inspired to do except it was suddenly directly interesting to me personally.

              In the end I was extremely pleased with the service, and I felt that my $100 was very well spent. Ultimately it turned out to be mostly educational entertainment accompanied by an uncomfortable reminder that I should be taking better care of my health (as Diabetes does in fact run in the family, a fact I already knew). Should I have been doing that anyway? Of course. However let’s be realistic. Many of us are complacent, and we don’t do what we know we should. Seeing a genetic test confirming what I in fact already knew was a useful little slap on the face. Are we all really so jaded here at SBM that no one can understand this rather common human reaction?

              Services like these have tremendous potential as teaching tools, and it saddens me to see SBM writers overlook that potential. It would be far better to push for clear regulation, and to make recommendations for how to improve the service, provide genetic counseling, and more clearly provide the tools for people to understand what the results do and do not mean. Dismissing the curiosity and interest in favor of “you don’t need to know” rhetoric is a poor tactic, and one doomed to failure. Harnessing that interest seems to me to be a more worthy pursuit.

      2. WilliamLawrenceUtridge says:

        Considering it is cutting-edge science to determine what the results mean, considering the entire field is still undeveloped, I’m really curious whether your statement that you understand the information is a genuinely true statement, or the Dunning-Kruger effect in action. It’s possible you are one of the few dozen PhDs who might have the background to understand how a specific SNP realistically effects your health, in which case I wonder what you think of people who don’t have a topically-relevant PhD having access to the information.

    3. Julian says:

      Its all fine what you write. You are not a fan of these tests for certain reasons. But does that give you the right to forbid OTHER people from making their own choice? I for example am not a fan of eating Pizza. Does it mean that no one gets to eat Pizzas? Obviously we are in a society where we treat people as actual adults that are able to make their own choices, even if they differ from our world view. No one forces you to take such a test.

      Its really fine if you don’t see a point in the information etc… And its great that you educate people about the pros and cons. Thats what actually should happen. What will happen though is that people are robbed of their choice to execute an idea that differs from what the FDA wants.

      What I would do: Forbid 23andme to make false advertisement. If they claim they got science, they need to back it up. Or otherwise not be allowed to advertise that what they do is scientifically backed. Maybe even make them hand out brochures about the pros / cons of getting the test and then let people wait for at least 2 days to think about it. If I got you wrong and you don’t think the FDA should somehow make this service unavailable to people, sorry.

      1. Harriet Hall says:

        “does that give you the right to forbid OTHER people from making their own choice?”

        Of course not, and I never suggested any such thing. I do, however, think society has the right to protect consumers by regulating false advertising.

  7. Michele Manion says:

    There was a GAO report published in 2010 on an investigation into these direct-to-consumer genetic testing companies. The report was titled: ” DIRECT-TO-CONSUMER GENETIC TESTS: Misleading Test Results Are Further Complicated by Deceptive Marketing and Other Questionable Practices” so you get an immediate feel for what they found. In direct comparison of four companies using 15 fictitious patients, they very frequently got inconsistent results across the companies, calling into question the technology used by these outfits.

    Even if the testing technology were good, the interpretation and reporting functions are extremely sloppy. They are providing disease info, in the form of fact sheets that could be found for free on the Internet, with little ‘personalized’ guidance at all and interpreting genetic testing results is a complicated process in many cases–not one to be handled by providing a few fact sheets.

    As a person who has worked in rare disease advocacy for decades, The other major red flag I see with 23andMe is their connection to Google. Just like stem cell therapy providers and genetic testing providers attempt to abuse the gray area that currently exists in regulating these activities, marketing firms and for-profit outfits are taking advantage of loopholes in regulation related to protected health information. Some do it through ‘share your story’ opportunities online, avoiding the need for informed consent about how the data will be used and never mentioning that these stories are data mined and sold to the highest bidder looking for patient information they are not legally entitled to. It is becoming a huge problme and one that is of grave concern to patient advocacy orgs. The nature of the relationship between Google and 23andMe is not entirely clear, but given Google’s past misdeeds related to providing private information to government agencies, the fact that they may have access to and the ability to mine millions of dollars worth private health data should be enough to give anyone pause.

  8. goodnightirene says:

    23andMe was founded by Anne Wojcicki, wife (or former wife) of Sergey Brin, the Google guy. She has a biology degree and worked in healthcare investments before co-founding this company. It all seems part of the modern “get rich off technology” culture we now live in. I’m not questioning her motive, just the whole idea that anyone needs to SELL such things in this way. If you need to know something about your genetics, it seems you should consult with your doctor and have it done in the same way as any other medical test, with proper counseling before–to understand the implications, and after–to deal with any negative outcome. It seems more like getting those useless allergy test kits Scott G writes about, rather than going to a clinic and getting proper allergy tests.

    1. windriven says:


      I haven’t checked my facts on this this morning but I believe this is the outfit that has a broader goal of building a more complete picture of the variations in the human genome. I raise this not to condone their marketing tactics or the validity or value of their product. Just an observation.

    2. Carl says:

      “It all seems part of the modern “get rich off
      technology” culture we now live in.”

      Someone call Guinness. I think we just found the oldest living person on the planet.

    3. Lumen2222 says:

      The business model is not the individual test fees. It’s the larger database that they are attempting to compile and selling that information. I’m pretty sure the reason they are able to offer the tests so cheaply is precisely because they are not focused on profit from the fees. This is all information that is on their website and is explained multiple times through out the purchase and testing. It’s also precisely modeled on Google which offers free products in exchange for meta information about your habits.

      23nme is basically a free market attempt at advancing knowledge about variations in the human genome. I realize that these days it’s quite fashionable to reject icky icky profit as a motivation for scientific research, however last I checked public funding was increasingly limited, and hysteria about genetic privacy ensures there can be no government driven solution for gathering this kind of data. So frankly I’m all for these private companies stepping up and convincing people to contribute to the pool of knowledge, so long as they are well regulated.

  9. Frederick says:

    Totally agree with Dr Novella.
    For exemple if i take myself, Academically speaking, i have no real degree in anything, right now i’m back to college. but i have work in a factory for 11 years. But even there i can say that i have a lot more knowledge than average on lots of thing, and in science. Since i’m 7-8 years old i always been interest in science. I know lots of Science people.
    I have a good understanding of how genetic and statistic Work. Be even that being said, i’m pretty sure i would not be able to really understand what is those in report and info. i’m conscious of my Bias. But some people like me might think that given the relative knowledge being huge, someone can be overconfident and think they can analyze those. I know very brilliant people with Science knowledge but somehow believe in woo-woo, i suspect for this kind of reason, overconfidence about your own intelligence and knowledge.

    Personally, I think that if you order those test, the result should go to your doctor FIRST, and than when he had analyze then and had a meeting with you, you could access them. Of course this is if the test are accurate and reliable, witch seem to be far for the case here. The FDA is totally right, I would prefer the FDA to go after some real woo-woo, but hey, we take what we can FDA FOR THE WIN!! lol :-)

  10. Ed Whitney says:

    Serves ‘em right for not marketing 23andme as a diet supplement “not for the treatment of any disease.” Maybe with a statement to the effect that interpreting the results exercises the lobes of the brain known to be involved in cognitive abilities. Any reader of this website could have told them how to do it.

    1. Yes, 23andMe boldly crossed the line of suggesting their services may be of use in treating medical conditions, thus leaving their butt open to the Federal Death Authority thugs.

      Should have put a disclaimer in – this is like a genealogy service, to find out who you are. And thats all.

  11. SkepticalOfPaternalism says:

    If a company makes fraudulent claims then the government should appropriately be involved (they should be cracking down on many unscientific medical treatments but aren’t). If 23andme hasn’t made fraudulent claims the FDA should instead focus on other companies where far more harm is being done in the alternative medical realm. I haven’t used 23andme but I should have the freedom to do so and to evaluate the information however I choose, turning to scientific experts that I trust and taking into account the potential for error. I see no reason to let a perhaps misguided political process determine what I should be able to see.

    We now have governments that lend the air of credibility to questionable alternative medical practices by licensing them. The more we let politicians legislate medical choices (or information flow) based on the pretense that government knows best, the more we risk the political process pushing questionable things like unscientific alternative medical views. Legislators and regulators don’t necessarily do what is best for the public. George Stigler won his nobel prize in economics for work on regulatory capture theory which explains why it is to be expected that regulatory agencies will tend to be “captured” by special interest groups (like alternative medical practitioners). It is to be expected they may not necessarily represent the public’s best interest despite wishful thinking that they will. The fewer tasks a regulatory agency engages in, the greater the ability for the public to try to keep an eye on it.

    Yes, not all patients will do the right things with this information since they aren’t skeptical enough and don’t understand the possibility of error well enough and that much of the research to determine genetic links to problems is still a work in progress. The answer should be to encourage the public to turn to credible scientific medical sources for their information, rather than blindly trusting a political process that often approves of unscientific medicine to do the right thing.

    There is no reason to hand the government power to dictate the flow of information in a way that could be easily abused. Despite our wishes it could be the FDA winds up being captured by alternative medical types who then prevent credible medical information from being offered, or require alternative medical claims be published in addition to scientific information.

  12. 23andMe just sent the following letter to their clients:

    Dear 23andMe Customers,

    I wanted to reach out to you about the FDA letter that was sent to 23andMe last Friday.

    It is absolutely critical that our consumers get high quality genetic data that they can trust. We have worked extensively with our lab partner to make sure that the results we return are accurate. We stand behind the data that we return to customers – but we recognize that the FDA needs to be convinced of the quality of our data as well.

    23andMe has been working with the FDA to navigate the correct regulatory path for direct-to-consumer genetic tests. This is new territory, not just for 23andMe, but for the FDA as well. The FDA is an important partner for 23andMe and we will be working hard to move forward with them.

    I apologize for the limited response to the questions many of you have raised regarding the letter and its implications for the service. We don’t have the answers to all of those questions yet, but as we learn more we will update you.

    I am committed to providing each of you with a trusted consumer product rooted in high quality data that adheres to the best scientific standards. All of us at 23andMe believe that genetic information can lead to healthier lives.

    Thank you for your loyalty to 23andMe. Please refer to our 23andMe blog for updates on this process.

    Anne Wojcicki
    Co-founder and CEO, 23andMe

    1. Harriet Hall says:

      They “believe” that genetic information can lead to healthy lives, but they don’t have any evidence that the information they provide has led to healthier lives. That’s something that could be tested.

  13. Harriet Hall says:

    One other thing: the various genetic testing companies test for different SNPs and often come up with very different assessments of risk. How do we know which is most accurate?

    1. MS, MT(ASCP) says:

      The pharmacogenetics industry is facing the same sort of conundrum. While they are testing a limited array of genes (the CYP450 family being the major one) and SNPs known to reduce the metabolic capacity in vitro, the in vivo story has turned out to be more complicated. Add to that the lack of drug prescribing information that is acceptable to physicians. The idea of pharmacogenetics is a good one (identifying the most effective drug based on genotype and reducing adverse reactions and interactions), but there is still a lot of work to be done as far as FDA clearance of testing methods, quality assurance, interpretation and application goes.

  14. Peter James Moran says:

    There is no question that there will be substantial downsides for many of those who unthinkingly and without sufficient forewarning undergo genetic testing.

    Nevertheless, how do we doctors reconcile our caring instincts with our much-vaunted respect for patient autonomy in cases like this?

    We are all in favour of openness and fully informed consent when it suits us, and despite the downsides and many failings they possess. We criticise some well-intended medical activities for being deceitful, too patronising or authoritarian, yet here we are having considerable nervousness about patients having “too much information” about themselves, when at their own cost and risk, but ,– critically –, when outside of our direct control.

    By all means work for the highest standards of information and for technical accuracy, but I am not sure that we can or should even think of denying genetic information to anyone who wants it. That must come eventually, I think. The public will demand it as their right. And it should have considerable benefits in some ways, most surely by reducing the incidence of genes which have serious medical consequences for families.

    1. windriven says:

      Well said.

  15. Bruce Lindgren says:

    Suppose for a moment that the test is (or becomes) valid. Should I not be able to have access to the information because I might mis-interpret them? I cam mis-interpret any number of things about my health just by watching Dr. Oz or reading certain magazines or websites. There are a number of issues that I discuss with my doctor based on what I read or questions I have about my most recent lab work. If I were to have a high risk of developing Alzheimer’s Disease, I might be more inclined to buy long term care insurance than if my risk were low. No genetic test predicts the future perfectly, but it does provide more information rather than leaving me with less.

    1. Ken Pidcock says:

      If I were to have a high risk of developing Alzheimer’s Disease, I might be more inclined to buy long term care insurance than if my risk were low.

      So what do you think that will do to long term care insurance premiums? What 23andMe provides now might be pretty useless, but that can’t last. Eventually, there will be risk pool effects on health, disability, long term care and life insurance. I can’t wait to hear the market-based solutions to that problem.

  16. MadisonMD says:

    P. Moran: Well said and agreed–once accurate, everyone should have access to good genetic information. I suppose the only question in my mind is whether people will really receive informed consent prior to testing. They probably wouldn’t get it from a company selling the test. (Maybe I missed it or interpreted this post differently; did Dr. Novella say that they should not have access to accurate information?)

    This all reminds my of residency practice in NY State. In order to run HIV test, by law, the patients had to be first counseled about all the ramifications of the test–including risk of discrimination, etc. They then had to be told the results in person with post-test counseling. Today, I work with genetic counselors who do the same prior to any genetic testing. It seems a lot of work but perhaps useful for people to stop and think a bit about whether they want to have the results.

  17. Carl says:

    I think it would be awful politics to completely forbid private DNA sequencing just for the sake of wanting to do it, but these people are clearly promising more than that. With the current state of genetic screening as it is, even saying nothing more than “take this to your doctor” might still be a problem because it implies that the information provided has more value than it likely has.

    I think they might be hesitant to subject themselves to too much investigation because if their current service is found not to hold up to their dreams and promises, they might not be able to do anything about it. I suspect that it would be either impossible or impractical to do any more than they currently do for anything close to the same price, and even a complete genome would have little practical value at the moment.

  18. Andrey Pavlov says:

    To address a couple of points by Peter Moran and MadisonMD:

    The issue at hand is not a simple denial of access to personal information. If you want a (very long) piece of paper that lists every nucleotide base pair in sequence of your entire genome, so be it. The issue is taking that data and saying it means something. If you want a list of your particular SNPs then great. But if they come with an opinion as to what that means, then it should be accurate. And while the actual polymorphisms are accurate, the interpretation of what they mean is not. We simply don’t know enough yet.

    Even then if it were just purely an issue of isolated personal use, then it wouldn’t be an issue. But it isn’t. People and (inevitably) corporations/insurance companies will act on that data. Otherwise what’s the point? So while you may wish to increase your long term health cover if you have the SNPs to predispose for Alzheimer’s, your insurance company, employer, or other entity may wish to charge you more for that cover on the exact same basis. Now, if that data were accurate, once again it wouldn’t be an issue. But it isn’t.

    By analogy what if there was a test sold that claimed to tell you your risk of your house catching on fire, but it was in no way validated. Now people will act in a manner concordant with the concern that their house is at extra risk of fire; they will waste money on prevention, be charged more for home owner’s insurance, have their property values (and that of their neighbors) drop, all for nothing. And that is the real rub with these private genomic profiling companies. We protect free speech and access to information, but we also agree that yelling fire in a crowded theatre is a criminal act. In a much more subtle way the supposed meaning of the genomic profiling is exactly like that.

    1. Kathy says:

      When I last took out health insurance I was asked if I had any pre-existent conditions. If I had and failed to reveal it, then became ill, the insurance would lapse. I wonder if knowing one is vulnerable to some genetically-based condition later in life would count in a case like this?

  19. Harriet Hall says:

    I don’t think the FDA or anyone else wants to prohibit genomic testing. I certainly don’t. I just want truth in advertising, and I think 23andMe has led consumers to think they are getting more than they really are.

  20. Angora Rabbit says:

    I have many Silicon Valley friends who have participated in 23andMe and are amazed that I, the scientist, haven’t done so. They pull out their wallet cards and I just chuckle b/c the “health” information they’re given is not much different from generic health advice. As FDA points out, just knowing sequence tells us very little. My lab is in the midst of such a project right now, and oh, I would be sooo happy if we could point to which transcript alterations contribute to disease risk! (It would make the manuscript reviewers happier, too.)

    Phenotype is so much more than a DNA sequence, let alone a SNP analysis. I tell my students, “Genes may load the gun, but often it’s environment that pulls the trigger.” For starters, SNPs are associations but may not be causative; maybe you had a rearrangement near the SNP that altered risk?

    Or what if you’re chimeric? Turns out lots of people are. That cheek swab may not be the same genome that’s expressed in blood or liver – it all depends on when a mutation occurred during embryogenesis. The biggest chimerism is in women, who only express one X-chromosome and individually differ on when the silencing takes place. Microchimerism is more common than we’d ever thought.

    And then there’s your epigenome. Maybe it’s not the SNP that’s important but the imprinted methylation pattern in the gene’s promoter?

    My point is that, like a lot of medicine translated for layfolk, genomic sequencing is way more complicated than 23andMe or other companies are marketing it. Frankly, it’s naive. Plus it’s not ethical for the excellent reasons that other posters have discussed above. We simply don’t know enough to say intelligent things about the vast majority of SNPs and allelisms. The metaphor of DNA sequence as a book is not really accurate – there is a lot more information than just sequence, which is why phenotype may not match genotype.

    For those who want to learn more, written at lay level, I point to an excellent News and Analysis in the Sept 13, 2013 issue of Science (vol 341, issue 6151, pg. 1163). It discusses why blood screening of metabolites (phenotype) is still far superior to DNA sequencing for screens of newborn diseases.

  21. Johan Viklund says:

    I’ve considered doing the 23andMe test for a long time. Mostly because of the ancestry information you can get. And of course it would be fun to know how much “Neanderthal” I am (though, that part I am a bit suspicious about). Another thing that would be fun is to compare the SNPs to my relatives, to track how much from whom and where (which one of us siblings is most similar to our grandmother for example).

    The disease information I see as an extra, mostly for entertainment purposes. If we had any severe genetic diseases in my family I would have known about them. Besides, I’m not the worrying kind (and have never really understood the urge of not knowing, I’m glad I don’t have to get into the dilemmas described above). There is one trait that I would like to find out about though. My skin is very elastic (some collagen “dysfunction”), and if I remember the information in OMIM correctly, this is “caused” by a repeat structure somewhere. No luck with 23andMe there unfortunately.

    I don’t think that “benefit” is a good criteria for evaluating this. It is a good criteria for a medical doctor to evaluate what should be done with respect to a certain individual, but that is a very different situation. For me it would satisfy some curiosity, provide some entertainment and it would give me an opportunity to teach my family about genetics.

    Given all this I find it really sad that 23andMe don’t seem to take their responsibility with regards to the FDA, it can’t be that difficult. The picture I get from reading the letter from the FDA and the response from 23andMe is not encouraging. They should do better than this.

  22. I never said people should not have access to their own genetic information, or any health information. I am essentially making two points:

    1 – We should consider the public health consequences of any mass program, including direct-to-consumer medical testing. Health-care costs are out of control, resulting in rationing of care. We should at least know if the services being provided by a company will result in a net increase in unnecessary costs and unintended negative health outcomes.

    2 – While people have a right to any information about themselves, this does not equate to professionals being alleviated of the responsibility of providing that information in a proper medical context.

    As an example – should we allow private companies to offer direct-to-consumer MRI scans, with all the same kind of claims and justifications as 23andMe. Arm yourself (or your doctor) with information, detect disease early, etc. The MRI results, which are done without any medical consultation, are sent directly to the consumer to do with what they will. Shouldn’t we know what the consequences of this will be? Is it wrong to argue that an MRI scan (let’s ignore any direct physical risk, which is minimal, for the sake of argument) should be done in the context of a medical evaluation? Is the individual’s right to information about themselves enough to justify such a service? I think this is an exact analogy.

    1. Kathy says:

      Steve Novella wrote: “While people have a right to any information about themselves, this does not equate to professionals being alleviated of the responsibility of providing that information in a proper medical context.”

      I once did some visiting at the Spinal Unit of a large hospital in the city where I lived. It seemed that the tetraplegics in the ward had their food plonked down on the bed-table and were left like that by the nurses. They were too busy to sit and feed each individual. So we volunteers were called to come and help feed them.

      And yet the hospital could say truthfully that the patients were fed … had good balanced meals … so why were they still hungry?

      There’s rather more to “informing” someone than just plonking down a bundle of raw data in front of them. Surely they need some help in understanding what it means and the implications. I don’t think this is patronising or paternalistic, just realistic.

  23. RMJ says:

    I am a 23andMe customer. I did it for entertainment. I also have a PhD in pharmacology and do not panic easily about test results. Before doing it I asked my siblings if they wanted me to tell them if anything significant to family health turned up and they both (PhD and MD) said no.

    I shared my health risk results with my doctor. They were all pretty boring (slight increases or decreases in risk) except for celiac, where my risk was listed as 16x normal. I didn’t think anything of it, since I have no typical symptoms. My doctor, however, knowing my health history and knowing far more about atypical celiac symptoms than I do, thought it would be worth pursuing. I was tested (standard anti-tTg test) and results were highly positive. I’m now gluten-free, feeling great, and my anti-tTg results are normal. I would never have been tested for celiac without 23andMe.

    I also order my own standard blood panel each year ( which I share with my doctor. It surprises me that I can order the tests, but I do it that way because it is very inexpensive. I enjoy having more control over my healthcare, but understand the limits in my ability to interpret results.

    p.s. First time commenter; love SBM!

  24. ChrisL says:

    Sorry if I missed anyone’s prior reference to publications on the issues around genetic testing, but for those interested this 2011 NEJM review paper seems a reasonable overview: (Free full text access linked from PubMed.)

    It is interesting that expanding access to genetic testing direct to consumers is occurring just as efforts have been made to encourage physician/patient reflection on the necessity of a gamut of established medical tests through the ABIM Choosing Wisely initiative:

  25. June Hua says:

    wow, i can be my own geneticist! very insightful and informative!

    1. WilliamLawrenceUtridge says:

      June, you have the stupidest website I’ve ever seen. Most of the crap spam comments I’ve seen have included the words “insightful and informative” in them. I think you’re a robot.

  26. woody says:

    I have already encountered a problem with misinterpretation of direct-to-consumer genetic testing in my personal life. A friend explained that he had done such testing (not sure if it was 23andMe), and was probably going to get Alzheimer’s disease based on his APOE genotype included in the report. I asked for the details, and it turns out he has one copy of the APOE4 allele which does increase his risk of eventually developing Alzheimer’s disease, but is by no means a guarantee that he will get the disease, even if he lives to an advanced age. He also didn’t realize that even if he had not been a carrier, he could still get the disease – the distinction between genetic causes and risk factors for disease.

    Someone more devastated by that information might take drastic action, which is at least part of the reason there are genetic counseling protocols within medical institutions before proceeding with testing for incurable debilitating diseases like Huntington’s disease.

    Does anyone who has used 23andMe know whether they omit certain results with potential devastating impact, like causative mutations for Huntington’s disease, ALS, Fragile X, etc.?

  27. Boldizar says:

    I part ways ways with you here. As someone who spent half my life traveling and living in third-world countries, where I had to self-diagnose, get my own tests at local labs, and self-medicate for things like malaria, dengue, and typhoid, I found it extremely frustrating when I moved back to Canada and needed a prescription for something as simple as an antibiotic. In, say, Indonesia, the fact that something like oxycontin was available without a prescription never tempted me to start popping random pills for fun.

    To move this line down to information itself shows why the alternative quacks have gained as much of a foothold as they have — the paternalistic mindset of complete patient disempowerment pushes patients away from good science.

    I’ve found that the only times I’ve ever pulled out the fact that I have a doctorate from Harvard — something that is as socially clumsy as it gets in normal circumstances — is when talking to doctors here in North America. Far too many doctors assume that all patients are idiots.

    My doctor is there as my consultant, not as my father/mother. My own body should be nobody’s business, whether the issue is a woman’s right to an abortion or anyone’s right to self-medicate as they find best. And when the issue is simple information, it is really offensive to be told that I should not be allowed to obtain that information — information about my own body — because it might make me “anxious.”

    1. Harriet Hall says:

      Don’t attack a straw man. No one has suggested you should not be allowed to obtain that information, only that the companies should not be promising more than they can deliver, and that the information is not as useful as most consumers think. Also consider that if everyone could buy antibiotics, not everyone is as well informed as you are, so a lot of people would be taking antibiotics they didn’t need, and that would increase antibiotic resistance to the detriment of all of us.

    2. WilliamLawrenceUtridge says:

      Boldizar, considering your training is not in any fields remotely related to medicine (law and economics according to your website), I would strongly question your ability to apply medical knowledge in all necessary circumstances. You may be a smart person, but you still lack the training of an actual doctor. I used to think I could simply demand medicines from my doctor – until my medical problems became more complicated, at which point I realized my understanding was orders of magnitude too simplistic compared to what I needed. Not to mention, you don’t have the base rate appreciation for adverse effects, interactions, and the natural progression of diseases, because your n is always 1 (versus a doctor, whose n can wander into the hundreds, or even thousands). Yeah, it’s inconvenient – but is it really that inconvenient? Not to mention, there’s a good chance you might not appreciate the risks of attempting to treat one’s self. There’s a reason doctors aren’t supposed to treat themselves or their families.

      Your doctor is indeed a consultant, an expert consultant. And in the case of the information provided by 23andMe, even expert consultants can’t actually translate it into meaningful information. It could not only provoke anxiety, it could also provoke unjustified complacency. It’s simply useless information, beyond entertainment.

Comments are closed.