Less Wrong is a community blog devoted to refining the art of human rationality. Please visit our About page for more information.

Gram_Stone comments on The Thyroid Madness : Core Argument, Evidence, Probabilities and Predictions - Less Wrong Discussion

10 Post author: johnlawrenceaspden 14 March 2016 01:41AM

You are viewing a comment permalink. View the original post to see all comments and the full post content.

Comments (132)

You are viewing a single comment's thread.

Comment author: Gram_Stone 14 March 2016 11:05:47AM *  2 points [-]

Post 2 of 3 was virtually ignored; post 1 got a lot of votes. My guess is no one with a modicum of medical knowledge has looked at it, so they up voted to increase the post's visibility because you're open to criticism and they're curious, then they had no reason to up vote the second time. Like CronoDAS said, it would be cool to have Sarah Constantin look at this. Or maybe Scott Alexander, though he's not an endocrinologist.

This also seems like exactly the sort of thing NancyLebovitz has been looking for.

Comment author: philh 15 March 2016 11:12:11AM *  2 points [-]

I'm pretty sure someone asked Scott on tumblr for his opinions on the first post. I think his reply was something like, "seems plausible but not my area of expertise", but I can't find it now.

edit: I just scrolled through seventeen pages of his tumblr (until I got to before the first one was posted) and didn't find it. I could have missed it, but this decreases my confidence that it happened.

Comment author: pianoforte611 15 March 2016 01:45:03PM 2 points [-]

Yes he said it could be plausible but would require more work to form better thoughts on.

Comment author: johnlawrenceaspden 15 March 2016 06:05:30PM *  0 points [-]

Scott and I were in communication earlier, and he asked some good questions that provoked post 2. Now he's not talking to me. Not sure why.

Just really busy, apparently.

Comment author: pianoforte611 15 March 2016 10:37:38PM *  2 points [-]

I honestly want to know - what do you expect him, or someone here to do? Say you're right? Figure out a way to fund a clinical trial of FM patients treating them with T3 or pig's thyroid? (I admit I didn't read all of your posts from beginning to end, you said that such trials were already done? If that's the case are they good quality? If this is such miracle treatment then were the results not clear?)

If you want to convince anyone, then you need interventional data, not hypothesizing. In other words, you have to pony up, or convince someone to pony up and fund said research.

Comment author: johnlawrenceaspden 15 March 2016 11:58:53PM *  0 points [-]

I honestly want to know - what do you expect him, or someone here to do?

I want him, or anyone else, to show me things that I haven't thought about, or have just missed, that mean I'm wrong. If I wanted disciples, I know exactly where to find them.

It is a core belief of this website that one is very bad at seeing the evidence against one's own ideas. I believe in that.

I am asking people to take me down, because I am probably unable to do that for myself, and I do not want to cause a catastrophe while I am still dithering, but if I just dither secretly, all I will do is find more evidence in favour.

If you want to convince anyone, then you need interventional data, not hypothesizing.

This seems like the entire problem. How to convince people to do the expensive experiments needed to explore the obvious hypothesis, without already knowing the answers?

Physics will spend billions trying to find surprises, for the sheer joy of it.

In medicine, where millions of lives may hang in the balance, no one cares.

Comment author: pianoforte611 16 March 2016 12:22:49PM *  3 points [-]

This seems like the entire problem. How to convince people to do the expensive experiments needed to explore the obvious hypothesis, without already knowing the answers?

Get an MD, prescribe pig's thyroid if you really believe it such a fantastic treatment. If the evidence is clear, start a clinical trial. Admittedly, I don't know if you can do that in the the nationalized UK health system, if you're in private practice in the US I believe you have enough leeway to do that. You'll be under very heavy scrutiny though, and there may be insurance issues but I don't understand them well.

As to what you're missing? It's really simple - you have an interesting hypothesis and not much else. And you have almost no quality data to back it up. Without that, there is nothing there to criticize. Your single case of pig's thyroid working is the most interesting evidence to me, however, personally I would bet on FM and the other somatoform disorders having heterogeneous causes (they aren't the same disease in every person). For example:


I'd be surprised if pig's thyroid cured a mitochondrial myopathy.

Also, the reason why expert' aren't taking you seriously, is that you kind of sound like a crank. A very intelligent crank admittedly, but a crank nonetheless. It's not that you have lots of facts wrong, it's more subtle than that, and I'm afraid it's hard to explain. Statements like this

It seems that one is either forced to accept (CFS/FMS/Hypothyroidism are extremely similar diseases which are nevertheless differently caused), or to believe that blood hormone levels can be normal in the presence of Hypothyroidism.

slightly exaggerated reaction Erm what? No they aren't extremely similar, and that's not even logically correct, rejecting 2.1 doesn't entail accepting 2.2 at all. And why is diseases having similar presentation a grand revelation? Do you know anything about the rest of medicine? Differential diagnoses based on a clinical picture are the rule not the exception.

I'm sorry if that's sounds harsh, but I'm trying to convey the attitude that someone who is in the field might have towards you. You don't quite speak the language. To use an analogy - you know the words but you have funny accent.

Comment author: johnlawrenceaspden 16 March 2016 05:13:36PM *  0 points [-]

No worries about sounding harsh! I declared Crocker's Rules, so I'm explicitly asking you to optimise for communication and not worry about offending me. And I very much appreciate you taking the time to tell me things I don't know.

I'd be surprised if pig's thyroid cured a mitochondrial myopathy.

That's exactly what I'm saying! The action of T3 seems to be to control ATP recycling in the mitochondria. Sarah Myhill's beautiful paper to my mind proves almost beyond doubt that that's the problem in CFS. This is what I mean by 'every time I look for disconfirming evidence, I find new reasons to believe'.

I know that I sound like a crank. That's because I am a crank. I am a member of several at-risk groups for Arrogant Overconfidence Disorder, which I strongly suspect to be related to hypothyroidism in some way. Others have suggested that I am under a certain amount of 'stress'.

CFS/FMS and hypothyroidism are much more similar than most diseases, to the point where out of a fabulous number of possibilities I was trying to fit to what was wrong with me, hypothyroidism looked instantly like what I had, despite the fact that I'd not only had the test for it, but the test was bang in the middle of the normal range. And I think the CDC agree. One of the diagnostic criteria is explicitly that hypothyroidism have been ruled out (haven't checked this, just a memory).

But also, doesn't the fact that all diseases look similar strike you as suspicious? As I understand it that was the whole reason for the 'stress' theory in the first place.

Let me think about the logic for a while, I'll get back to you.

OK, logic looks fine. I really need to know if that bit's wrong. It means my mind is broken.

If they're not differently caused then they have the same cause. And if that's true, then in one case the TSH test is picking it up, and in the other it's not. So the test is not doing what it's supposed to.

Suppose diabetes was diagnosed by insulin levels instead of blood glucose. And there were two sets of patients, who had roughly the same symptoms, but one lot weren't treated because the insulin test showed that their problems weren't diabetes.

Would you not say that the insulin test was broken?

We should be looking for the 'blood glucose' for hypothyroidism. And as a very lot of people have been claiming since 1940, that's 'slow metabolism'.

I really hate arguing by analogy. But it seems people don't understand unless I do, and I'm now arguing to persuade. Not of the truth of the hypotheses, but of their plausibility.

The sciences I trained for would leap on this. Medical Science has left it uninvestigated (to say the least) since 1970. Whether I'm right or not, that's careless.

And if I'm right.... Jesus Christ.

Comment author: pianoforte611 17 March 2016 02:20:04AM 1 point [-]

It seems I somewhat misunderstood your argument and misjudged you; I tentatively pegged you as a pig’s thyroid evangel feigning humility. I apologize. I also apologize because I am not the opponent you are looking for.

Since I apparently didn’t stress this enough, I will conclude by saying again that without interventional data, you have nothing. It doesn’t matter how beautiful your theory is, it doesn’t matter how smart you are, if it disagrees with experiment then its wrong. Repeating your hypothesis again and again, doesn’t help your case, it hurts your credibility. Unfortunately this is all I have to offer that I think is worth offering at this point.

Comment author: Lumifer 17 March 2016 03:16:54PM 1 point [-]

without interventional data, you have nothing

That is not true. You would prefer to have data from randomized intervention trials, but even without them you can look and collect data and come to conclusions.

Comment author: johnlawrenceaspden 17 March 2016 02:42:48PM *  0 points [-]

My dear old thing! That is a perfectly natural assumption to make and there is no need to apologise. If I were convinced of the truth of this idea, that is likely exactly what I would be, here practising the argument before I have to make it as a wild-haired prophet.

But I think I have managed to retain enough sanity to not want to believe it if it's not true. And I have pretty high standards for truth, and they definitely include intervention, cause, randomisation, placebos and control.

At the moment, I think that my hypotheses are probably false (because there is no way that I can see that it can be a widespread problem and yet fibro-turks are hot)

If it's false, then I think it's probably important to refute it properly if only to stop Wilson.

But I don't care very much about that. My own problems seem to be gone, they are or were probably either non-existent or horribly idiosyncratic and no one can help me with them, and I am just going to have to work it out on my own. That's a man's death and I am glad to have found a worthwhile enemy.

But I disagree with you about beautiful hypotheses. If they disagree with experiment then they are wrong, no question.

But they are worth looking at carefully, and a science that does not bother is not a science. And probably not truth-finding, even over the long run.

If you run into any interested opponents, do tell them there is someone wrong on the internet. There is still a mystery to explain! It's just back to being a hobby, for me.

Comment author: ChristianKl 16 March 2016 11:11:18PM 1 point [-]

But also, doesn't the fact that all diseases look similar strike you as suspicious?

Today we have people saying that cancer isn't a single illness and people trying to make distinction within broad categories of illnesses.

I think there's a good chance that depression isn't a single illness and a lot of our present disease categories aren't.

Comment author: Jiro 16 March 2016 10:11:44PM 1 point [-]

Suppose diabetes was diagnosed by insulin levels instead of blood glucose. And there were two sets of patients, who had roughly the same symptoms, but one lot weren't treated because the insulin test showed that their problems weren't diabetes. Would you not say that the insulin test was broken?

No, not in general. It might be for diabetes, but that's fact-specific. Let's try substituting in something else:

Suppose fractured skulls were diagnosed by X-rays. And there were two sets of patients, who had roughly the same symptoms (head pains and bleeding), but one wasn't treated because the X-ray test showed that their problem wasn't a fractured skull. Would I say that the X-ray test is broken? Of course not.

Comment author: johnlawrenceaspden 16 March 2016 11:27:44PM 0 points [-]

Nicely done, thank you! My brain is broken, and this "informal reasoning" is harder than it looks.

In your case, the X-ray test is doing its job perfectly. And if the posited type 2 hypothyroidism needs different treatment from the type 1 version, which it probably would, then the TSH test will be a great way to tell them apart.

What I don't think you're allowed to do is say 'no problem, can't be anything to do with your car crash' when what you mean is 'your skull is not fractured'.

So the TSH test is a great test for TSH, and probably a good test for circulating thyroid hormones (although it doesn't give the whole picture). But I don't think that means that the TSH test is a good test for 'no thyroid hormone-related problem'.

Do we still disagree? Can I phrase my A&B&C=>(D OR E) thing better? Or do I need to abandon it?

Perhaps: Hypothyroidism (by which I mean any failure of thryoid hormones to act on cells).....

Comment author: Lumifer 16 March 2016 05:50:39PM 1 point [-]

We should be looking for the 'blood glucose' for hypothyroidism. And as a very lot of people have been claiming since 1940, that's 'slow metabolism'.

Well, what's the test for speed of metabolism? Usually it's measured by the consumption of oxygen (VO2, high-level athletes do that a lot) and that's not a particularly expensive or difficult test. I am sure there is data on the distribution of VO2 in normal population. Do you think this test would be adequate for your purposes?

Comment author: johnlawrenceaspden 16 March 2016 06:14:38PM 0 points [-]

I think it most definitely would. Broda Barnes didn't like it, but only because the test is stressful and so tends to give false negatives (you're looking for the resting rate). But as long as it's done carefully, it should be fine.

Comment author: Lumifer 15 March 2016 06:16:08PM 2 points [-]

Not sure why.

He's a busy man and you're very enthusiastic -- you have a personal stake and endocrinology is your third-favourite hobby :-/

Comment author: NancyLebovitz 14 March 2016 03:43:39PM 2 points [-]

Not only is it the sort of thing I've been looking for, I've linked it at SlateStarCodex.

However, looking at the lack of response, I think want this sort of well thought-out material, I also want well-thought out material which leads to good conversation. This may be a harder thing to define.

Comment author: johnlawrenceaspden 15 March 2016 03:25:25AM 1 point [-]

I am flattered again. Why have you been looking for this sort of material, and how do you define 'this sort'? I will be happy to optimise for good-conversation-provoking-ness. I am seeking argument. I want people to tell me what I have missed.

Comment author: moridinamael 14 March 2016 02:35:24PM 1 point [-]

Extrapolating from a sample of Myself, I think maybe people don't feel qualified to comment but think this is interesting, and well written, and are just upvoting and then waiting for somebody qualified to show up.

Comment author: Gram_Stone 14 March 2016 04:32:22PM 1 point [-]

Agreed, but that doesn't explain why the second post of three sits at three votes.

Comment author: johnlawrenceaspden 14 March 2016 09:51:46PM 0 points [-]

No, I don't get that either. People liked the question, but didn't care about my attempt to answer it? I was really excited.

Imagine my horror to find that there's a book called 'type 2 hypothyroidism', by a homeopath, which is full of the most ghastly pseudoscience, in spite of which a lot of the author's clinical observations are in fact my own predictions.

Two months ago I would have thrown it away unread. Still, no one can now tell me that I'm making a horribly technical argument that no one can understand!

Comment author: NancyLebovitz 15 March 2016 01:26:46AM 1 point [-]

As I recall, homeopaths do extensive interviews of their patients-- this might give homeopaths a chance to notice true things, even if their treatments are nonsense.

Comment author: johnlawrenceaspden 15 March 2016 01:59:20AM 0 points [-]

Oh, absolutely agree. For years I've had my back sorted out every six months or so by a local chiropractor, and I know that the actual beliefs of her profession are lunacy, but she seems to me to be a really skilled physiotherapist who is extremely good at getting rid of my back pain. And she's really nice.

I've tried things like booking appointments three months in advance to control for regression to the mean, and all I can think is that either what she does works or I'm buying a really good placebo which works whether I believe in it or not.

It's more that I'm a bit annoyed to have come up with a nice theory, which at the time I thought was really clever, and then found it already published, along with a plausible-looking mechanism in 'Medical Hypotheses', of all places (how my friends laughed), and then found an actual book about it by a man who clearly doesn't have an enormously powerful filter for loony ideas.

Comment author: CronoDAS 16 March 2016 01:27:58PM 1 point [-]

Spinal manipulation (of the kind practiced by chiropractors) has been shown to be an effective treatment for acute back pain...

Comment author: johnlawrenceaspden 16 March 2016 05:33:47PM *  0 points [-]

Oh brilliant! Well done medicine! How long have they been claiming that? For twenty years I've been going to see a witch-doctor, and now it turns out her spells really have kept me out of a wheelchair.

Aargh, I've just believed you without checking because it fits. A primary sin. Do you have a reference for that?

Comment author: CronoDAS 17 March 2016 11:05:31PM *  0 points [-]

Never mind. My original source was my father (who's usually pretty reliable when it comes to pseudoscience) but apparently the current Cochrane review says it's no better than placebo. (The subluxation theory of chiropractic is nonsense, but it's not ridiculous that fiddling with someone's back can relieve back pain - massages, for example, tend to feel good.)


Comment author: johnlawrenceaspden 18 March 2016 05:10:22PM *  0 points [-]

The thing you linked is not from Cochrane. They say:


I paraphrase: (it works, a bit, it's as good as anything else as far as we know, we have no clue really, further research is needed.)

That looks like science. I am glad they do what they do.

Comment author: johnlawrenceaspden 18 March 2016 04:46:57PM *  0 points [-]

Your dad might be right! This is from NICE (British recommendations for national treatment, a very sane system in my opinion, they try not to waste public money on things that don't work or are too expensive)

The manual therapies reviewed were spinal manipulation (a low-amplitude, high-velocity movement at the > limit of joint range that takes the joint beyond the passive range of movement), spinal mobilisation (joint movement within the normal range of motion) and massage (manual manipulation or mobilisation of soft tissues). Collectively these are all manual therapy. Mobilisation and massage are performed by a wide variety of practitioners. Manipulation can be performed by chiropractors and osteopaths, as well as by doctors and physiotherapists who have undergone specialist postgraduate training in manipulation.

1.4.1 Consider offering a course of manual therapy, including spinal manipulation, comprising up to a maximum of nine sessions over a period of up to 12 weeks.

I have found over the years that a single session every year or so obliterates a problem that, if left unmessed-with becomes painful and eventually restricts my ability to play various sports.

The first time it was done (I was about 25?), a short-sighted friend blanked me in the street because the way I walked had changed and she didn't recognise me. She said the way I walked had been very characteristic, and now it was more normal!

I don't know what NICE are basing their recommendation on, they could conceivably be going after a cheap placebo effect.

P.S. My nice chiropractor says I have: Ilio-psoas hypertonicity secondary to Sacro-Iliac instability.

P.P.S. Careful. Chiropractors are loonies. Their explanatory theory is rubbish. My personal chiropractor is great. I would be very very suspicious of even her colleagues, let alone the rest of her profession. On the other hand, at least they are trying obvious things, to see if they work.

Comment author: johnlawrenceaspden 14 March 2016 09:42:51PM 0 points [-]

I am pathetically grateful for your flattery, but really! I came here for an argument...