Is Health Data Ever Harmful?

In yesterday’s post I described how searching the medical literature helped me avoid a dangerous surgery with no obvious benefit. The surgeon I consulted, who recommended the surgery, said that published evidence backed her up. I could not find that evidence, however. Others found evidence that contradicted her recommendation.

Among the comments on that post were similar stories: Searching/reading the medical literature had been helpful. Learning what had happened (in research studies) was better than relying on an expert (a doctor). Here is an example:

A little over two years ago, I was “depressed”. My psychiatrist wanted to prescribe an SNRI [serotonin-norepinephrine reuptake inhibitor]. I related, once again, my poor experience with an SSRI and asked for evidence that an SNRI would be any more effective. He said there was evidence that SSRIs [selective serotonin reuptake inhibitors] worked. I pointed out the 2004 meta-analysis that showed no meaningful difference between SSRIs and placebos. Then I asked whether there was any better evidence for SNRIs. Since he wasn’t able to provide any, I told him that since we know that extremely low Vitamin D blood levels, poor diet, no exercise, and no social life can cause depression (all things I had at the time), I’ll try fixing those things first and then resort to drugs if that fails. It did not fail and I quit seeing him.

None of the stories in the comments described the opposite outcome: Knowing the data made things worse.

Are there exceptions? Is it always helpful (or at least not harmful) to know what happened (i.e., know research outcomes)? Has anyone reading this had an experience where knowing health research data was harmful?

16 Responses to “Is Health Data Ever Harmful?”

  1. Alex Chernavsky Says:

    Well, there’s the issue of cyberchondria:

    Cyberchondria (or cyberchondriasis) refers to the unfounded escalation of concerns about common symptomology based on review of search results and literature online.[1][2] Articles in popular media position cyberchondria anywhere from temporary neurotic excess to adjunct hypochondria. Cyberchondria is a growing concern among many healthcare practitioners as patients can now research any and all symptoms of a rare disease, illness or condition, and manifest a state of medical anxiety.

  2. Joseph Buchignani Says:

    I can’t cite peer reviewed studies that harmed me, but certainly amateur medical science has slowed me in my progress towards a personal cure, because it was so convincing, yet wrong.

    During my search for a cure for my dietary issues, I tried and discarded IBS theory, then extended IBS theory (SCD and FODMAPS et al), then paleo diet, then non-conventional paleo, then ketogenic diets, then Aquatic Ape theory, before finally arriving at a diet of
    1. white rice and
    2. ultra lean seafood, including shrimp, lean fish, and scallops

    Other influential models included food reward theory, various elimination diet theories, etc.

    Each paradigm cost me a great deal of time to finally falsify and reject, and clouded a clear interpretation of the evidence staring me in the face. The apparent persuasiveness made me spend much more time and be much more meticulous in my falsification than I needed to be. After each model finally inarguably failed me, I felt betrayed and angry.

    I eventually adopted a solution that fit no known models.

    Had I pursued a strict data-driven approach, I might’ve found a solution much faster. Or maybe not. The paradigms were all useful in interpreting and constructing the final theory.

    I think it is very difficult to see data for what it is when we have persuasive, compelling paradigms suggesting incorrect interpretations and approaches.

    The solution would be to conduct more rigorous and analytical and intelligent experimentation. Unfortunately this was not possible for me, because strict elimination diets are difficult to adhere to, especially when sick, and often don’t provide sufficient food reward or even sufficient basic nutrition.

    I guess the major lesson I take away from this is that one should always focus on interpreting your own data properly, and never fall into the cognitive trap of assuming that a model must be correct simply because there is no better model available. But that’s very, very hard to do.

    I think the vast majority of medical studies don’t even rise to the level of amateur health models, because they are poorly constructed and therefore suggest false or useless paradigms.

    The only health info I truly value at this point is rigorous self-experimentation. Everything else is merely for purposes of theoretical suggestion and inspiration. And it’s a good thing to try things you “know” won’t work, and compare the results to things you “know” should work, to see whether one significantly outperforms the other. That’s a rapid way to test a model’s validity, instead of testing every possible permutation before junking it.

  3. q Says:

    well, there’s the story, i don’t know how accurate it is, that steve jobs died much earlier because he chose a dietary therapy over a surgical one. was that a decision driven by data in any way?

  4. Seth Roberts Says:

    Thanks, Alex. I consider diagnostic categories theory, not data.

  5. Seth Roberts Says:

    Joseph, I agree that theories can be harmful. Since the SCD diet has helped many people with IBS (cases where it helped are data, not theory), I don’t see why it was a bad idea to try it.

  6. Adam StJames Says:

    About 12 years ago, my Dr. prescribed Lipitor for high cholesterol. After some research, I decided that there was more risk than benefit and I refused. After many weeks of arguing with him, I told him that if he didn’t stop, I’d change doctors. He grudgingly told me that there is a diet that “sometimes” works. He prescribed, then, a very high fat diet. Lots of animal fat, fatty meat, butter, avocados, cream, things like that. Within six months, I lost 40 lbs. and my cholesterol was back into the normal range. That easy. And free.

  7. Jim Purdy Says:

    For me, the funding sources of the published data are all important.

    For example, there are many dozens, if not hundreds, of published studies extolling the health virtues of California walnuts. All the studies are paid for by the California Walnut Board. If walnuts are so great, why do they need heavy research ad advertising subsidies from the industry?

    That’s just for an inexpensive commodity like walnuts. When it comes to BigPharma or BigSurgery, the billions o dollars involved inevitably distort the studies and their conclusions. As far as I’m concerned all funded studies need to be held to an extremely high standard. Or ignored.

    But there is one source of data that I consider the gold standard: self-experimentation. I trust my body to tell me when something is working, or not. And my body has told me repeatedly that it does not like BigPharma chemicals, while it does like simple paleolithic foods.

    Self-experimentation! Hmmm … who is the Berkeley professor who has written extensively about self-experimentation? :)

    Maybe the government should find him and make him the head of the FDA.

  8. by Says:

    re: J.Buchignani’s frustration:

    Maybe the emotion of having put so much work into testing a theory and having it fail, is necessary to solidifying your will to stick to a white rice and lean seafood diet for the rest of your life.

    Pain seems to be a necessary precursor to lasting habit change, in anything.

  9. Ashish Says:

    It seems to me that any time your specific health situation/need turns out to be the opposite of the usual case, knowing the “health data” would be harmful. But that’s really Monday-morning quarterbacking, isn’t it?

  10. Scott P. Says:

    My only thought is that information can be paralysing either through volume or being unable to establish information’s validity especially in light of contradictory information, which there is a lot of on any given subject.

  11. Seth Roberts Says:

    Any time your specific health situation/need turns out to be the opposite of the usual case, knowing the “health data” would be harmful. But that’s really Monday-morning quarterbacking, isn’t it?

    Sort of. It’s not a convincing argument for ignorance, I think we agree.

  12. David Harmon Says:

    The big problem with self-research is that it’s not always obvious which articles are (1) peer-reviewed medical science, (2) alt-med craziness, or (3) paid shills for drug companies (or for that matter, alt-med companies).

    It is possible to learn enough to tell the difference, but it’s not necessarily easy — and some of the ringers feature deception and pressure tactics at cult-like levels. The Mercury Mafia is a good example there — they not only denigrate the vaccines that have turned so many deadly diseases into rarities, but promote dangerous chelation therapies which have already outright killed several children. But if you disagree with them, they’ll tell you you’re endangering your child.

  13. Joseph Buchignani Says:

    I’ve just learned that my needs are unique due to Accutane side effects. I’ve been diving deep into that community. It’s an interesting example that has some bearing.

    The data showed it could cure acne quick and permanently. The initial clinical trials didn’t show THAT bad side effects, I guess.

    Unfortunately due to its mechanism of shortening telomeres for rapidly dividing cells, the damage shows up years later. So they put it on the market and a ton of people got delayed poisoning.

    Also, due to the way that studies are done, it’s very hard to establish causation for compensation, because the general nature of the mechanism creates widely different effects in people, that are often hard to diagnose. And the delayed and gradual nature of onset makes causation even more difficult to establish in the legal sense.

    So, the way we collect health data in official studies created this disaster by its formal structure – the link between Accutane and curing acne was easy to establish, while the link between subsequent damage was much more vague, easy to miss if you put on the typical “evidence-driven medicine” blinders.

  14. Joseph Buchignani Says:

    By the way, kombucha and apple cider vinegar both failed for me, and now I know why – it’s the same reason I can’t digest fat. Accutane can impair bile production.

    So I need a non-acidic fermented food. Would yoghurt qualify?

    And, if anyone has clues on reversing epigenetic damage, especially to telomere lengths, I’m all ears.

  15. Aileen Kennedy Says:

    Really interesting. What I wondered when I read this was why the commenter, knowing that those were all issues that might contribute to his/her depression, did not take those steps before seeing a doctor. Why wait until you are sitting in the doctor’s office before deciding to address them? I agree that such issues should be explored by a doctor first, before medication is suggested, but I was just puzzled by the attitude.

  16. Ben Says:

    Cyberchondriasis: Fact or fiction? A preliminary examination of the relationship between health anxiety and searching for health information on the Internet:

    http://dx.doi.org/10.1016/j.janxdis.2011.11.005