Low-dose Naltrexone – Hashimoto’s Cure or Powerful Placebo?


[HD: This is a shorter-than-usual post – a bite-sized morsel intended to address a recent comment on one of my blog posts about the utility of low-dose naltrexone (LDN) for the treatment of hypothyroidism due to Hashimoto’s thyroiditis.]

One of the commenters on my blog recently left the following comment about her experience with hypothyroidism:

…Coming from a research background I decided to look into things. And there IS a treatment! Low Dose Naltrexone. There is research and evidence that it can lower the immune attack, prevent, and occasionally reverse hypothyoidism. Why didn’t my Endo know about this? Is it because Naltrexone is not sold in the 4.5mg amounts needed, and so not promoted to doctors by pharmaceutical companies?

I am now on it (thanks to a functional Dr) and my chronic fatigue has lifted! I also sleep better at night and have lost weight…

Holy crap, is this true?!  There’s a way to cure Hashimoto’s thyroiditis?  This is amazing!  Unbelievable!  I’m so stoked!  Where can I get this stuff?!  Wait…why have I never heard of this treatment?

Most of you who see “lame”-stream doctors like me have never been offered low-dose naltrexone for your Hashimoto’s thyroiditis because there is no clinical data to support its use in hypothyroidism.  Not limited clinical data.  Not preliminary clinical data.  NO clinical data.  There is not a single study in a peer-reviewed journal demonstrating the clinical efficacy of low-dose naltrexone for the treatment of hypothyroidism due to an autoimmune attack on the thyroid.  If you can prove me wrong, I will print the study and eat the paper.

I realize this is going to come off as me picking on the commenter, which is not my intention.  However, when someone says that she has a background in research and then proceeds to claim that there is evidence to support her assertion that LDN has evidence behind it, I have to challenge her when the evidence is absent.

Let’s do a little experiment.  Go ahead and open up a new tab in your browser.  Go on, now.  I’ll wait.  Got it?  Good.  Search “naltrexone hashimoto’s.”  Scroll down the page.  What do you notice about the composition of the results?  Go 7-10 pages deep into the results.  Notice anything different?  No, you don’t.  It’s all anecdotal “evidence,” mostly from fringe blogs, all citing the same few, historical, miraculous “cures” of other conditions (not thyroid!) that led to the promulgation of LDN as a legitimate therapy for…almost everything.

There is one shining light on page 2 of the results, a well-researched piece by Steven Novella, MD (not ND or PharmD), Low Dose Naltrexone – Bogus or Cutting Edge Science?  Although this post on Science-Based Medicine is from 2010, the concepts hold up well.  And, at least with respect to research on LDN for thyroid dysfunction, not much has changed since then.

Continuing with our experiment, let’s go to one of the most commonly used and respected medical search engines, PubMed.  C’mon, humor me.  Open up the tab.  Good.  Now search “naltrexone and hashimoto’s.”  What do you see?

Zero results?  That can’t be right!  There’s tons of stuff on the Google results!  This must be a conspiracy.  The pharmaceutical companies just don’t want doctors to know about this treatment because then they’d lose money on millions of Synthroid prescriptions!

Right, of course.  The U.S. government (PubMed.gov, people) is in bed with the pharmaceutical companies, conspiring to keep doctors in the dark about this miraculous cure-all.

Moving on, perhaps the original search terms were too restrictive.  Maybe I created a “false positive” just to illustrate a point.  Maybe I’m no more honest than any of the folks out there pushing this treatment.  Well, let’s try a broader search on PubMed, “naltrexone and thyroid.”  That should be a fairer assessment of what kind of data is out there, yes?

Sixteen results.  That’s what you get.  Spoiler alert: none of these papers offers any clinical evidence that LDN can be used to treat, reverse, or otherwise cure hypothyroidism due to Hashimoto’s thyroiditis.

But Google shows 67,300 results for “naltrexone hashimoto’s!”  You can’t tell me there’s nothing to it!  I won’t believe you!  I won’t, I won’t, I won’t, I won’t!

To paraphrase William Shakespeare, “Methinks thou doth protest too much.”  I don’t care how much you want to believe that there is good data supporting a course of action on which you’ve already embarked.  Wanting to believe it’s true does not make it true.

I’m really glad that this issue was raised on my blog, as it is illustrative of a critical concept: Google searches ≠ research, for the most part (Google Scholar notwithstanding).  A web site with “naltrexone” in the title should be viewed with a healthy dose of skepticism, as the bias should be obvious based on the title.  Blog posts written by people detailing their personal experience with LDN do not constitute evidence of its efficacy.  For that matter, blog posts written by clinicians, detailing their “amazing results” using LDN, also do not constitute evidence.

At this point, I’d like to make it clear that I’m not saying LDN can’t possibly have a beneficial effect in people with hypothyroidism due to Hashimoto’s thyroiditis.  What I’m saying is, we don’t know.  To suggest otherwise, no matter how much anecdotal evidence there is on the web, is irresponsible.

But how can you say that all these testimonials don’t constitute evidence?

Easily.  Just as you can offer up blog after blog claiming success with using LDN for hypothyroidism, I can proffer hypotheses as to why it appears to work:

  • There is a powerful placebo effect with any drug that has the type of hype surrounding LDN.  The small community of fringe practitioners promulgating this therapy tends to attract followers who have similar missionary-like zeal.  With all these people extolling the virtues of LDN, do you think you will be more or less likely to believe it’s working?
  • LDN has actually been studied for use in other conditions, where it has shown potential promise.  As it may be useful for fibromyalgia, and the symptoms of fibromyalgia and hypothyroidism overlap significantly, perhaps LDN is treating something other than the hypothyroidism.  In other words, maybe the patient with “uncontrolled hypothyroidism” actually has fibromyalgia (or something else) that responds to LDN.
  • The mechanism of action of LDN is thought to involve chronically increasing endorphin levels through partial blockade of opiate receptors.  Perhaps any positive effects from the drug are due simply to increased endorphin levels, which naturally reduce pain (and possibly improve other symptoms).
  • With respect to weight loss, we already have access to an FDA-approved medication called Contrave, which is a combination of naltrexone and bupropion.  Obviously, there is some utility of naltrexone for weight loss, so we don’t need to invoke a beneficial effect on thyroid function to explain some weight loss (though I should make it clear that naltrexone alone has not been shown to be a powerful weight loss drug).

So what if LDN isn’t making the thyroid any better?  If I feel better on it, isn’t that all that matters?

While I agree that the ultimate goal is to make you feel better, I would simply urge you to consider this: when you use an unstudied thyroid therapy like LDN – despite claims that side effects are minimal to non-existent – you have little idea if there will be any serious short or long-term adverse effects.  In addition, this therapy is said to “boost immune function,” which is a nonscientific, throwaway phrase.  But it does create a contradiction with respect to the myriad conditions LDN is said to improve.  I think Dr. Novella says it best:

Further, there is an inherent contradiction in simultaneously treating diseases that are auto-immune (the immune system attacking the host), and immunodeficiency diseases (like AIDS) and claiming to treat cancer by “boosting” immune activity. Increasing immune activity actually worsens auto-immune diseases, and suppressing the immune system would worsen AIDS. This is a difficult contradiction to resolve.

All I am asking here is for precision when discussing untested and unproven therapies.  If you come to me and say, “I have hypothyroidism, I didn’t feel well, my TSH was normal, and then I started LDN and now I feel better,” I can’t argue with that statement.  I’m glad you feel better.  But let’s not extrapolate your situation to mean that LDN is an appropriate treatment for hypothyroidism or that it can cure Hashimoto’s.

Low-dose naltrexone needs to be studied in the context of a randomized, double-blind, placebo-controlled trial.  I’d be interested in seeing pre-intervention thyroperoxidase antibody levels and post-intervention levels, in addition to the usual thyroid function tests we use to guide thyroid hormone dosing in clinical practice.  Study participants should have low-normal TSH levels at baseline but have reduced quality-of-life (QOL) scores on a standardized questionnaire.  They should have no history of opiate or alcohol-dependence, and they should not be taking any drugs with a central nervous system effect (antidepressants, stimulants, etc.).  They should also be free of other diseases thought to respond to LDN.  If a trial like this shows both benefit and safety of LDN, then consider me intrigued.

One more cautionary tale before I wrap this up.  Selenium, a mineral necessary for proper thyroid hormone metabolism, has been studied in people with Hashimoto’s thyroiditis.  Many in the medical community got very excited when selenium supplementation was shown to lower thyroperoxidase antibody levels – the antibodies that cause thyroid destruction.  We thought that this might be a disease-modifying drug – the holy grail for treating autoimmune hypothyroidism.  Unfortunately, despite lowering antibody levels, selenium was not shown to actually reduce the need for thyroid hormone.  In other words, these patients continued to require the same dose of thyroid hormone as before they started taking selenium.  On top of that, when you look at the trials that have shown decreased antibody levels with selenium use, these trials have been conducted in parts of the world where selenium deficiency is more common (than in the U.S., where it is not common at all).

My point is, there are all kinds of interesting hypotheses out there when it comes to medicine, and many of them sound quite plausible.  Spoiler alert #2: most of these hypotheses will never pan out in real life.  So, be open to new ideas, but be skeptical.  And do not claim that there is research to support an unstudied treatment that you just really, really, really want to believe in.


Have you used LDN?  What’s been your experience?  Why do you think it worked/didn’t work?  Are you a physician or other clinician who prescribes LDN?  What kind of results have you seen?  Are you a researcher currently studying LDN?  Comment below!

By interacting with me in the Comments, you agree to abide my Disclaimer.


18 Replies to “Low-dose Naltrexone – Hashimoto’s Cure or Powerful Placebo?”

  1. Of course, you understand that the response to your emminently reasoned post will be that, “no one will study this because BigPharma can’t make money off of it”. Obviously you are just another Pharma Shill reaping millions by supressing miracle cures that people know work because they did their “own research”. : )
    I wish your commenter had been more specific about his/her “research background”. This seems as vague as the research itself. I wrote some papers in college. I checked out five books on the subject at the library (pre-internet) and skimmed them for “evidence”. The paper promoted a professor’s pet theory, so I got an ‘A’ for my cherry-picked “proof”. I guess I have a “research background”.

    1. Pharma has been late with their last few payments to my account in the Caymans, so I might just come clean and confess that I am, in fact, a shill.

      In the commenter’s defense, she didn’t know she’d be pushing one of my red-hot hot buttons by being vague about her research. It drives me nuts when people come in to my office toting their research, which consists of printouts from the quackiest of quacky websites.

  2. After being misdiagnosed for 25 years and being told by mainstream doctors that nothing was wrong with me, I finally went to a functional doctor that tested ALL of my numbers and found that I have Hashimoto’s. After 6 months and antibodies in the uncountable range, I decided to give LDN a try. It has been a life-changer for me. After 12 months, my antibodies have dropped to 60, that is not a placebo affect. I have cut my thyroid dose in half and I am thriving like I haven’t experienced in my entire adult life. I am 45 years old and finally know what it’s like to do more than just survive.

    1. Glad to hear you’re feeling so much better, Melissa. Whatever caused the turnaround, LDN or otherwise, we can agree that doctors need to place high value on actually helping people feel their best.

  3. It’s STUPID, by-the-book doctors like you that I, and many Hashimoto’s patients avoid and feel contempt for, and whose opinions I throw into the trash immediately. Your blogs are now getting a lot of disrespect from one of the groups I belong to. Going so far as to say, by someone else, not me, that you should be reported. Good luck with your practice.
    You’re a disgrace to the medical profession.

  4. the plural of anecdote is not data, but we are not always privileged to have a double blind, placebo controlled, adequately powered and properly replicated study. people talk about evidence-based medicine, but i find in my practice [psychopharmacology, with a special interest in bipolar disorder] that i usually run out of strict evidence-based illumination by 15 minutes into the evaluation. there are many issues which i would love to see explored in good studies, but i know it will never happen – they would be too complicated, too expensive, take too many years, require too many patients…. real patients have concurrent illnesses [excluded from the studies], sometimes consume alcohol [excluded], take other medicines [excluded], require 2 [almost always excluded] or even 3 or more relevant agents [always excluded], are not drug naive [usually excluded] and so on.

    nonetheless, i’d like to think my decisions are at least informed by the evidence, even if not directly supported.

    i’ve prescribed naltrexone for many years, as an opiate blocker and an alcohol craving reducer, and for the life of me i don’t see how it could help with hashimoto’s. but stranger things have happened.

    i’d be happy to see an open label case series for a start.

    1. Excellent points, jk – thank you for making them. I really like the way you put this:

      “i’d like to think my decisions are at least informed by the evidence, even if not directly supported.”

      I also agree that an open-label case series for LDN in Hashimoto’s would be a more realistic place to start, as we’re not likely to see anyone get a big enough grant to do a proper randomized trial.

  5. given that naltrexone is pretty benign, especially in near-homeopathic doses, i think it could be offered ethically as an experimental adjunct to standard treatment. on the one hand, it risks having one’s colleagues think you’re a quack, otoh it would add some level c evidence to the level z [=Zero] evidence we have now.

    perhaps it might carry some weight with patients who come in with preconceived notions. i.e. it would allow a treater to say “we tried it on N patients and found they did/did not do any better than our other patients.” but i wouldn’t count on that being persuasive.

    1. You’re probably right, jk, that low-dose naltrexone is benign. At least, it appears to be from what I’ve read on blogs ;).

      You’ve also hit on one of the issues that would prevent me from offering it to my patients until there is some degree of evidence to back it up – it would make me look a quack. In addition to looking like a quack to my colleagues, it would likely attract the type of patient that, under other circumstances, would hate me and my practice style. The doctor-patient relationship would likely go south very quickly, especially if the LDN didn’t work and the patient came in asking for all kinds of other tests and treatments promulgated on quackythyroidblog.com.

      This is actually one of the many reasons why I have a personal “never prescribe desiccated thyroid” policy. Not only do I think it is the wrong treatment, but I also do not want to have my name plastered all over the quacky thyroid forums as the only Endo in the area who will prescribe it. The patients would come out of the woodwork and my days would become real painful, real fast.

  6. Given that naltrexone has been used successfully in treating the hyperarousal of post-traumatic stress disorder, I wonder if the success of LDN with some Hashimoto’s patients is simply the result of the overlap between post-traumatic stress and the development of an autoimmune disease. If there really are as many PTSD patients out there as someone like, say, Bessel Van Der Kolk suggests there are (due to the prevalence of child & domestic abuse), it would make sense that clinicians would see them regularly, and that their complaints would not be fully addressed by standard disease treatments.

  7. Finally: a ray of light and a hefty dose of reality in all this hearsay. Just last night I heard of this amazing LDN and wondered why after I had experienced Hashimoto’s for 17 years had I not heard of it. Then I opened up the tab and did the searches and found out why.

    There is no scientific evidence to support this. What I am really amazed by is that the person I spoke with was seeing a ‘qualified doctor’ who advised it was the ‘new treatment’. Me thinks the ‘qualified doctor’ should do a little research too.

    I intend to provide the person I spoke with some information in the hope that they are able to make informed choices about their health and health care. Will also investigate the ‘qualified doctor’ as well.

    Homework is valuable at any age.

    PS. I had my thyroid out earlier this year as it was growing and pushing on my oesophagus to the extent I had difficulty swallowing – no doubt I feel a lot better without it and no LDN to be seen.

  8. In the last para of your post, would it make sense to change “theories” to “hypotheses”? Distinguishing between the two seems important when we are talking about science-related topics.

    1. Fair point. Theory and hypothesis are not synonymous, though I was using them that way. After rigorous testing of a hypothesis, it may then be classified as a theory if it seems to be holding up against scrutiny.

Leave a Reply

Your email address will not be published. Required fields are marked *