Natural thyroid hormones: what is behind NDT

There is one sentence I hear often in my consulting room. It comes quietly, and it comes at the end of the conversation.

I read something on the internet. About natural thyroid hormones.

Then comes a small pause. Inside that pause sits a question that few people ask out loud: do you now think I am being difficult?

No. Anyone who still cannot get out of bed in the morning after two years of stable laboratory values has good reason to keep asking. And this topic is not an invention of the internet. It is the oldest treatment for hypothyroidism there is.

It is just that the matter is more complicated than either camp would like. Many people know the feeling of standing between two very confident opinions and having none of their own.

What an NDT tablet actually contains

Natural. That word carries a lot. It sounds like origin, like something that was around long before anyone rebuilt it in a laboratory. There is something to the word. Just not what most people take it to mean.

Natural thyroid hormones are dried, ground thyroid tissue from pigs, standardised to a hormone content and pressed into tablets. Internationally the group is called NDT or DTE, that is desiccated thyroid extract. Well known products are Armour Thyroid, NP Thyroid, Erfa Thyroid and Thyroid-S. I name these as factual information, because you will read them in forums, not as a recommendation.

The two numbers this really turns on

That tissue contains both thyroid hormones. T4, the storage hormone, and T3, the active form. They sit in there in a ratio that the pig's body set, not yours.

This set ratio is the quiet lead character of this article. Keep it in mind: roughly 4 to 1.

And what about T2, T1 and thyroglobulin?

You will hear one argument almost every time. NDT contains more than T4 and T3, namely T2, T1 and thyroglobulin, and that is where the difference lies. The first part is correct. The second is open.

Mechanistically plausible, human data thin: that is exactly how this point should be described, not as an argument that settles the question.

Why the immune system attacks its own thyroid in the most common cause of an underactive gland is covered in the guide Hashimoto and the immune system.

And now you know why the number 4 to 1 will appear several more times in this text.

More than a hundred years old: the history and the switch

If someone tells you that natural thyroid hormones are an internet fashion, the opposite is true. Animal thyroid preparations were the first pharmacological treatment for hypothyroidism at all and dominated the market for most of the twentieth century. In this story, levothyroxine is the newcomer.

A note on honesty about the year. In forums you almost always read 1892. The scientific reviews are more careful and speak of more than a century, or of the last roughly 130 years. I therefore write end of the nineteenth century, because I do not hold a clean source for the exact year.

A second, independent research group arrives at the same picture. James Hennessey traced how levothyroxine came to prevail, and describes that the use of animal preparations has been scientifically documented for around 130 years and that refinements in manufacturing and regulation continue to this day.

Hennessey JV. Endocrine. 2017;55(1):6-18. PMID: 27981511 · DOI: 10.1007/s12020-016-1199-8 [Review]

Why the switch was a good decision

Something matters to me here: the move to levothyroxine was not a conspiracy against nature but, with the data available at the time, a good medical decision. A synthetic single active substance can be set to a defined amount, while a biological material varies. T4 stays in the blood for about a week and can smooth out fluctuations that way, T3 is shorter lived. And if the body converts T4 into T3 anyway, the storage hormone is theoretically enough.

The last point is the only one where something has moved since. The conversion is not an automatic machine that runs the same way in everyone. It depends on enzymes that are themselves regulated. How that works is covered in the guide From T4 to T3: the conversion. That is the mechanism the entire debate revolves around.

And now you know why I do not set up a battle line on this topic. Endocrinology did clean work here. It just left a remainder group behind.

The decisive question: is NDT better than levothyroxine?

This is the question you came here for. You get it answered in full, including the numbers that do not fit the picture.

What surprises many people: there are exactly two randomised double blind trials of NDT against levothyroxine, both from the same research group.

These three numbers tell you how much weight any statement here can carry.

Why complaints can persist on levothyroxine at all has far more facets than the choice of product. That is covered in the guide Levothyroxine and persisting symptoms.

One note that belongs here: subgroup findings like these are a reason for a medical conversation, not for a switch on your own. Whether someone belongs to one of these subgroups cannot be read off any text, only from history, findings and course under medical supervision. And anyone already taking a thyroid medicine changes nothing about it without a medical prescription.

And now you know why both camps can quote the same studies and still arrive at opposite conclusions.

The preference paradox

Picture two rooms. In the first sit statisticians looking at questionnaire scores. They say: no difference. In the second sit the participants of the same study, asked which product they would like to keep. About half of them say: the other one.

Both rooms are right. That is exactly what makes this topic so hard to grasp.

The genetic attempt at an explanation, and why it does not hold

There is a hypothesis that could explain this paradox, and it is mechanistically elegant. Perhaps some people convert T4 into T3 less well for genetic reasons.

I write this with regret. That finding would have been an elegant explanation.

Three possible explanations, without committing to one

First the study design. In a crossover trial, every person experiences both products one after the other. Expressing a preference is then easy, and order effects cannot be fully calculated out.

Second the measuring instrument. The questionnaires capture symptom lists, general well being and depressive mood. Perhaps they do not capture what people mean when they say that on one product they felt more like themselves. From the perspective of clinical psychoneuroimmunology that is not an odd idea: drive, sense of warmth and resilience arise from the interplay of nervous system, immune system and metabolism, and they are hard to press into 36 points.

Third a genuine subgroup. Perhaps a portion really does benefit, and this effect disappears in the mean because another portion gets worse. That is exactly what the subgroup analyses of 2021 and 2025 point to.

And now you know why this debate so often ends with one side accused of ignoring the studies and the other of ignoring the people. Both accusations miss at the same spot.

Why this debate is fought so emotionally

Hardly any medical topic generates this much heat. Not blood pressure, not cholesterol. Why is that? I believe it comes down to four things that apply at the same time.

First, this is about a remainder group that is not small. The most recent high ranking seminar on hypothyroidism in the Lancet writes that levothyroxine is safe and inexpensive, brings values into the reference range and improves symptoms in the majority. And in the same paragraph, that around 10 percent have persisting complaints despite unremarkable values. For statistics that is a fringe group. For the individual person it is their whole life.

Taylor PN, Medici MM, Hubalewska-Dydejczyk A, Boelaert K. Lancet. 2024;404(10460):1347-1364. PMID: 39368843 · DOI: 10.1016/S0140-6736(24)01614-3 [Systematic Review]

Second, the asymmetry between population and person. A guideline asks what is on average the best decision for the population. That is its job, and it does it well. A person in the consulting room asks something else: what is right for me, with my starting point, in this year of my life. These two questions do not always have the same answer.

Third, the feeling of not being heard. The Mayo analysis of 673 people had exactly this as its core theme, not the product. Anyone who is told for years that the values are fine, and still does not function in the morning, develops mistrust towards the sentence that everything is fine. And then that person finds a website that describes exactly their complaints and offers a solution.

Fourth, there are interests on both sides. Suppliers of NDT products earn money from this topic, manufacturers of levothyroxine do too. That is not an argument against either side, but it is a reason to check texts against their sources, mine included. That is why every study here carries a PMID and a DOI.

Clinical tradition and anthroposophical perspective, explicitly not a study finding: in anthroposophic medicine the thyroid is described as an organ at a border, where thinking passes into the voice and the will passes into the warmth of metabolism. That is a way of looking, not a measurement, and it stands beside physiology, not in its place. What I like about it is a question it raises: why does this organ of all organs so often affect people who endure a lot and say little? I have no evidence based answer to that and I simply observe it. More on the term natural in medicine is covered in the guide Anthroposophic medicine.

And now you know why I do not take their arguments away from either side in this text.

What ATA and ETA say, and why they say it

The professional societies do not reject NDT on principle, but as routine therapy, and they give four reasons for it. Each can be backed with a number. You get all four, then the other side just as fully, and at the end, where I weight things differently.

Reason one: the ratio that does not fit

Reason two: batch variability and recalls

Standardising a biological material to a hormone content is difficult. This is not a theoretical objection: it happened twice within eleven months.

In its 2020 recall notice the FDA names the possible consequences explicitly: signs of an overactive thyroid up to chest pain and cardiac arrhythmia, and for pregnant women miscarriage and impaired fetal development. Fairness requires adding: that these recalls happened also shows that oversight exists and takes effect. Both have since been closed.

Reason three: missing long term data on hard endpoints

Reason four: the risk of a suppressed TSH

This is the point that actually protects people, and it gets its own section in a moment. In short: because NDT contains a proportionally large amount of T3, and T3 can push TSH down more strongly, it is easier to end up in a range with a very low TSH on this product. And that range is not harmless.

And now the other side, just as fully

And now you know why I go into this section in such detail. It is the part of the topic that most German websites skip.

Safety: the point is not the product, it is the dose

If you keep only one section from this article, please make it this one.

The safety question does not turn on whether the raw material comes from a pig. It turns on one number: your TSH. It is the signal with which the pituitary addresses the thyroid. If the hormone level rises, TSH falls. If it is very low, a lot of thyroid hormone is on the move. And that low range is well studied.

A claim I would like to correct, kindly but clearly

On several German language websites it says that a TSH below 0.1 is normal and harmless on NDT, because TSH simply behaves differently on T3 containing therapy. The first part is correct: T3 can push TSH down more strongly. The second part is not supported by the data. The three works above together cover more than 140,000 people and point consistently in the same direction below 0.10.

I do not assume bad intent from anyone. This sentence probably arises from an understandable observation: people often feel better with a low TSH, at least in the short term. It is just that bones and heart rhythm say nothing as long as they are still keeping up. That is precisely why we measure.

Pregnancy: its own framework with its own target values

In pregnancy, different target values and a different care setting apply. There is a dedicated guideline from the American Thyroid Association from 2017 for this, reaching from the preconception phase into the breastfeeding period. The physiological reason is simple: in the first trimester the fetus depends on maternal T4, and T4 is the form that crosses the placenta to a relevant degree. That is why a T3 heavy replacement is not the standard route here. Everything concerning thyroid therapy in pregnancy belongs in medical hands without exception.

Alexander EK, Pearce EN, Brent GA et al. Thyroid. 2017;27(3):315-389. PMID: 28056690 · DOI: 10.1089/thy.2016.0457 [Guideline]

When a T3 containing product is discussed, a TSH value alone is not enough. What makes sense is the combination of TSH, free T4 and free T3, at a defined time interval from intake, because T3 levels fluctuate over the course of the day. Which values say what is covered in the guide Thyroid values: which ones really count.

And now you know why the sentence about the supposedly harmless TSH below 0.1 is the sentence I find most uncomfortable in this field.

The legal and supply situation in Germany

Suppose you now think: fine, that would be worth a conversation. Then in Germany you run into a hurdle that hardly anyone writes about.

No NDT product is licensed as a finished medicinal product in Germany. So there is nothing here that could simply be prescribed and collected at a pharmacy.

The legal route is called single patient import and is set out in section 73 paragraph 3 of the German Medicines Act. This provision allows finished medicines that are not licensed in Germany to be brought in under narrow conditions.

German Medicines Act, section 73 paragraph 3. Federal Ministry of Justice, gesetze-im-internet.de [Agency Document]

The second option is compounded preparations through specialised pharmacies. That too is prescription only and no way around a medical prescription.

It is not a grey area and not a loophole, but a clearly regulated route with high requirements. Anyone ordering NDT online without a prescription is moving outside this framework, with no certainty about origin and content. As a rule it is not covered by statutory insurance and runs on a private prescription.

A look across the Atlantic that surprises people

Many assume that NDT is normally licensed in the United States. That is not the case.

The FDA itself records that these products derived from animal tissue are not FDA approved. It applies a risk based enforcement approach and has announced to manufacturers that it will issue guidance on developing these products towards marketing applications.

This tension between regulatory status and lived practice is perhaps the most honest frame for the topic. Not a fringe phenomenon, and still nowhere a regularly licensed medicine.

And now you know why a serious conversation about this topic starts with the legal framework and not with the product.

For whom the question comes up, and what a responsible framework looks like

That leaves the question from the beginning: who is this actually relevant for?

For whom it comes up: for the roughly 10 percent who have persisting complaints despite well adjusted values. This group is a topic of its own, and it has many possible causes beyond the product. Both are covered in detail in Normal values and still symptoms and in Functional hypothyroidism.

For whom explicitly not: for everyone who feels well on levothyroxine. That is the most important sentence of this section, and it does not come from me but from the post hoc analysis of 143 participants. In the group with few symptoms, the values on levothyroxine were significantly better than on NDT. A switch there would, based on this evidence, not be a gain but a step backwards.

What belongs on the table first

From the perspective of functional medicine the order is decisive: first the building blocks and the context, then the product.

What a responsible framework looks like

If the question remains after all of that, then it belongs in a conversation with clear rules. And these are the rules.

A medical prescription, no going it alone. Monitoring intervals with laboratory checks agreed in advance. What is monitored is TSH, free T4 and free T3, not one value alone. Stopping criteria set in advance and an agreed point in time for an honest interim assessment. For the combination the ETA names three months as the period after which treatment is discontinued if there is no improvement. I consider that clarity one of the strongest points of the guideline.

What this text cannot and should not deliver: a recommendation for or against NDT in your case. That takes a history, an examination and a conversation, and none of it fits into a blog article.

And now you know why I do not consider anyone difficult for raising this topic. It is a good question. It just has a more complicated answer than either side on the internet claims.

Frequently asked questions about natural thyroid hormones

These are the questions that come my way most often on this topic, in the consulting room as well as by email.

The thyroid and the rest of the body

A thyroid hormone is never on the road alone. It hangs on cofactors, on the stress axis and on the energy balance. From here several paths lead onwards.