Why has Thyroid Stimulating Hormone (TSH) become the Rodney Dangerfield of hormone tests? If I had a nickel for every time I’ve seen a practitioner of some stripe claim that we can’t rely on TSH, I would be drowning in nickels. But why do people defecate on TSH? Is it really as inaccurate a reflection of your thyroid status as many an Alternative Medicine aficionado would have you believe? Does TSH only tell you what your pituitary thinks about your thyroid hormone levels, as opposed to how the rest of your body feels about said levels? Must you discount TSH results in favor of T3 and reverse T3 levels? Do these questions even have clear, unambiguous answers?
A Brief History of TSH
In the mid 1910s, scientists figured out an easier way to measure basal metabolic rate (BMR). Recognizing that a low BMR was associated with hypothyroidism, physicians used the presence of a low BMR and other signs/symptoms of hypothyroidism to make the diagnosis. They also used BMR to titrate the dose of thyroid hormone.
By the 1940s, serum protein-bound iodine (PBI) emerged as both a diagnostic and therapeutic marker; this was the only way to quantitate thyroid hormone status at the time. Unfortunately, T4 monotherapy would often increase PBI above normal, while T3 monotherapy would normalize BMR but not increase PBI. Combination T4/T3 therapy, however, had the advantage of normalizing PBI.
At the time – like today – it was well-recognized that things like cholesterol, deep tendon reflexes, and symptoms were just not sensitive or specific enough to guide thyroid hormone replacement therapy.
In 1971, the first TSH radioimmunoassay was developed. This turned out to be a game changer. When scientists studied patients who had previously been treated to normalize symptoms, BMR, and PBI, they found that most of these patients had a suppressed TSH.
One could argue that a suppressed TSH by itself could not, at this stage of the research, be considered conclusive evidence that patients were being over-treated. After all, perhaps pituitary secretion of TSH just wasn’t a good marker for overall thyroid status, right? Wrong. Studies at the time found that T4-treated hypothyroid patients could be rendered biochemically and clinically indistinguishable from healthy controls, using a T4 dose just high enough to normalize the TSH. These doses turned out to be about 1/4 to 1/2 the doses previously used to normalize BMR and PBI – two markers that were nowhere near as sensitive or specific as TSH.
And thus it became accepted that hypothyroid patients had been overmedicated for decades, standard T4 replacement doses dropped dramatically, and everyone lived happily ever after…
But I Still Don’t Feel Well
As anyone reading this site is well-aware, there is a sizable minority of hypothyroid folks who do not feel normal after normalizing the TSH. So what gives?
In many cases, it’s just not your thyroid. Having access to a diagnostic marker as sensitive as TSH is a double-edged sword. On one hand, people with very subtle perturbations of thyroid function can now be diagnosed and treated early, before they spend years feeling like crap. On the other hand, a large number of people will have mild elevations of TSH in the presence of “hypothyroid symptoms,” but those nonspecific symptoms are often not due to hypothyroidism. Unfortunately, once someone is labeled as hypothyroid, it can be difficult for both patient and clinician to look past that label. This results in constant tinkering with the dose or preparation of thyroid hormone, embarking on a Quixotic quest to find the magical balance of T4, T3, and TSH that will restore the person to “normal.”
In case the preceding paragraph did not make my position clear, let me state this very plainly: if you have been tinkering with your thyroid hormone dose and/or preparation for more than a year, and you feel little to no improvement in your symptoms, your symptoms almost certainly are not caused by your hypothyroidism. In fact, you may not even have hypothyroidism – especially if you were diagnosed by a naturopath, functional medicine doctor, or chiropractor. Let me remind you of an axiom commonly stated or implied throughout this blog: if you enter an Alt Med provider’s office wondering aloud if you have hypothyroidism, you have a better than 95% chance of coming out with that diagnosis. Given that the prevalence of hypothyroidism in the US is about 1 in 300 people, the likelihood of being incorrectly labeled as hypothyroid is exceedingly high.
No, Really – Pretty Sure it’s My Thyroid
But what about folks who have absolutely no other plausible explanation for their symptoms? This is where things start to get as murky as a bayou swamp, and where I hope to provide at least a modicum of clarity. I think the best way to achieve this is to debunk Alt Med’s claims about TSH one by one:
Claim: There are myriad conditions (stress, depression, chronic fatigue syndrome, etc) in which T4 and T3 levels drop, but TSH levels do not rise accordingly. Therefore, TSH cannot be used as a diagnostic or monitoring tool in these settings.
HD: Mostly False. Alt Med loves to dredge up the well-described phenomenon of non-thyroidal illness syndrome (NTIS) and make it the poster child for discrediting the utility of TSH. The problem with this is that NTIS is usually a transient entity seen primarily in severely ill patients. In NTIS, thyroid function studies can be variable and incredibly challenging to interpret, but the relevant pattern for this post would be slightly low/low-normal TSH, T4 and T3.
The literature on this subject is confusing and conflicting. The consensus opinion among experts in the field, however, is that most instances of NTIS do not require treatment with thyroid hormone. It is thought that the lower thyroid hormone levels in periods of critical illness may be a protective response, guarding against excessive tissue catabolism (breakdown). This is plausible, given that a body under that much stress has more important metabolic priorities – like staying alive.
It is also recognized that the lower T4 and T3 may be a maladaptive response in some circumstances, in which case the patient might benefit from treatment. An in-depth discussion of how to make this distinction is outside the scope of this post, but I will tell you that there is no elegant way to make the diagnosis or determine the “right” treatment. It’s really, really, really hard.
What I can say, however, is that most free-living people (non-ICU dwellers) who have stress, depression, or any number of other conditions will not have NTIS. If you walk into my office under your own power, and your thyroid function tests appear fairly normal, you almost certainly do not have clinically meaningful hypothyroidism.
While there are – for example – studies showing that some patients hospitalized for severe depression may have lower T3 levels than controls, these patients with normal TSHs generally do not fulfill any other clinical/biochemical criteria for making a diagnosis of hypothyroidism. In addition, when randomized, double-blind trials have looked at giving T3 to euthyroid, severely depressed patients, there has been no benefit.
The bottom line: equating conditions like fibromyalgia, chronic fatigue syndrome, or PMS with NTIS is a misapplication of our understanding of the hypothalamic-pituitary-thyroid axis. This oft-employed strategy by Alt Med is yet another way they can diagnose absolutely anyone with hypothyroidism, embarking on a treatment journey that ultimately will prove to be fruitless.
Claim: The pituitary is different from all other tissues in the body. In some circumstances, T3 levels in the pituitary may be normal, while T3 levels in the rest of the body are low. Therefore, it is possible to have clinically meaningful hypothyroidism in all the body’s tissues except the pituitary gland. These normal pituitary T3 levels tell the gland that the body’s thyroid hormone levels are fine when they aren’t, leading to a normal TSH.
HD: Partly False. Unfortunately, the part that is false is the most important part. In order to explain, this may get a bit dense. But my readers tend to be pretty smart, so here goes:
Whoa!!! System overload, HD!
Bear with me, y’all; I promise I’ll do my best to not hurt your brains here. Let’s walk through the process of how TSH-producing cells in the pituitary decide whether to make more TSH (thyroid hormone levels aren’t high enough, so the thyroid needs to be encouraged to make more) or less TSH (thyroid hormone levels are too high, so the thyroid doesn’t need any more encouragement).
As you may remember from my post about reverse T3, the body has a few types of deiodinase enzymes that do many things, including conversion of T4 to T3 and local regulation of thyroid hormone bioactivity.
The type 2 deiodinase (D2) is found in multiple tissues throughout the body, including the pituitary. Because our bodies like homeostasis, they are very good at defending T3 production when T4 is a bit low and decreasing T3 production when T4 is a bit high. This means that D2 activity is up-regulated when T4 is a little low and down-regulated when T4 is a little high.
If that was the only mechanism by which T4 to T3 conversion was controlled, then TSH would not be a good thermostat for thyroid function, right? Think about it: let’s say the thyroid is overactive and T4 levels rise, but T3 conversion is kept constant through down-regulation of D2 activity, then the pituitary would think that thyroid hormone levels were just fine, and the TSH would not be suppressed. But that’s not what happens.
It turns out that – in the above example – the down-regulation of D2 activity is more than offset by a rise in reactivation of previously inactivated D2 and an increase in D2 synthesis.
If you want a visual representation of this process, go back to the pretty diagram with the colors. Notice the bidirectional arrows between the yellow (active) D2 and pink (inactive) D2. You see that the inactive D2 can be reactivated by yet another enzyme that we won’t discuss. Also, if the pituitary is making more of the active D2, it will lead to accurate transduction of the high T4 levels, by making more T3 inside the pituitary cell.
Let’s press pause for a moment and put this all into context. Alternative Medicine claims that the pituitary is special. That is correct. Although we don’t fully understand the mechanisms that allow this system to operate, we think that massive D2 synthesis in the pituitary exceeds the maximal rate at which higher T4 levels cause inactivation of D2. Therefore, the net effect is that higher T4 levels in the body will lead to higher T3 levels in the pituitary and lower T4 levels to lower T3 levels. Basically, we’re saying that the feedback system between the thyroid and pituitary is appropriate, and we’ve postulated how it works.
Alternative Medicine also claims that the pituitary will remain in a state of blissful ignorance when the body’s thyroid hormone levels are too low. They claim this is why the TSH is unreliable – it won’t rise in the setting of “tissue hypothyroidism,” because T3 levels in the pituitary are peachy, thanks to the turbocharged conversion of T4 to T3 by D2. Although there is D2 activity in other tissues besides pituitary, D1 does much of the heavy lifting elsewhere, when it comes to converting T4 to T3. Since D1 is not up-regulated to the same extent as D2, then the argument goes that the body can suffer low thyroid hormone levels without an appropriate rise in TSH.
Unfortunately, this logic is based almost entirely on research about NTIS, which I already told you is seen in critically ill, hospitalized patients. You simply can’t apply NTIS data to someone with a common “outpatient” disease like fibromyalgia or depression. To the best of my knowledge, there are no credible data proving that patients with any of these outpatient diseases suffer from some nebulous form of chronic NTIS.
Under normal circumstances – which apply to just about everyone reading this – if the pituitary is making normal amounts of TSH, you should infer that thyroid hormones levels in the rest of the body are adequate.
Claim: Deiodinase enzyme activity and nuclear thyroid hormone receptor/transporter activity vary depending on physiological conditions. These processes contribute to regulation of tissue thyroid hormone levels. Therefore, serum thyroid hormone levels and – by extension – TSH may not adequately predict tissue thyroid hormone levels.
HD: Mostly True. This is a prime example of Alt Med taking science and manipulating it to draw the conclusion they want. It’s true that deiodinase enzymes will have different levels of activity in different tissues depending on the needs of that particular tissue at that particular moment. It is also true that thyroid hormone receptors, as well as the transporters that carry thyroid hormone into the cell, will be up- or down-regulated depending on the needs of the cell at that moment. These are complex processes that cannot be measured in patients.
It is also true – for example – when treating hypothyroidism with levothyroxine, serum T3 levels are often low-normal or even slightly low. As the vast majority of people on levothyroxine with normal TSHs feel fine, the clinical significance of the T3 levels is unknown. It is often assumed that deiodinase and cellular transport activity acts to maintain cellular T3 levels, which would explain why most of these folks feel just fine.
The derailing of this train of thought occurs by asserting that TSH does not adequately predict tissue thyroid hormone levels. I apologize for sounding like a broken record, but it is only in NTIS and other rare circumstances (e.g. central hypothyroidism) that this is true. There are copious data showing that euthyroid people, people with well-treated hypothyroidism, and people with well-controlled hyperthyroidism have normal TSHs. There are also copious data showing that most people with uncontrolled hypo- or hyperthyroidism have abnormal TSHs. I don’t really know how to say it any clearer than that.
Sure, the local regulation of thyroid hormone conversion, transport, and action is a complex process. It is true that we can’t measure tissue thyroid hormone levels in any meaningful way. What you need to remember is that the body is constantly aiming for homeostasis, and it’s generally excellent at achieving that. But Alt Med wants you to believe that humans have evolved a spectacularly dysfunctional thyroid hormone production and signaling system. It is so messed up, in fact, that almost any chronic illness/disease can throw the thyroid system completely out of whack. What’s more, it’s impossible to accurately measure the degree of disarray, so the obvious solution is simply to treat everyone who feels poorly with thyroid hormone. Titrate the dose to resolution of symptoms or atrial fibrillation, whichever comes first. Ridiculous.
Claim: Some people have a genetic mutation that affects the pituitary’s conversion of T4 to T3. Therefore, a hypothyroid patient with this mutation may have a normal TSH, even when they clearly need a higher dose of thyroid hormone.
HD: True. This is the last Claim I’ll address today, as I’m pretty sure I’m burning out your attention span. There are many known single nucleotide polymorphisms (SNPs) in the deiodinase genes. The functional relevance of most of these is unknown, but there is one that has been reasonably well-researched. It’s called Thr92Ala – a SNP in the D2 gene.
The Thr92Ala change in the coding region of D2 does seem to be associated with slower T4 to T3 conversion. Studies in people with the SNP have shown a few interesting things:
- A correlation between the SNP and TSH in healthy people, but not between the SNP and thyroid hormone levels.
- A lower TSH in people with one copy of the SNP, when compared to people without the SNP and when compared to people with two copies of the SNP (surprising!).
- Thyroidectomized people with the SNP may require a higher T4 dose to normalize the TSH.
- One study showed the SNP was not correlated with general condition, neurocognitive functioning, or response to combination T4/T3 therapy; a second study showed an impairment in well-being that was mildly improved with combination T4/T3 therapy.
So what should we make of this? First, know that if you are euthyroid, you probably have nothing to worry about. Remember that the body is great at achieving homeostasis. If there is slightly sluggish T4 to T3 conversion in tissues that use D2, the body has many available tools to mitigate that. It can inactivate D2 more slowly, reactivate D2 faster, increase thyroid hormone transport into the cell, etc. Bottom line: if you ordered one of those DIY genetic profiles online and it says you have the SNP, relax.
Second, if you’re hypothyroid because your gland doesn’t work great (as opposed to having had surgery to remove it), much of the above will still apply to you. But if you feel poorly on levothyroxine despite a normal TSH, then you may wish to broach the idea of adding in a small dose of liothyronine. Just keep your expectations low, as T3 is not the cure-all as Alt Med portrays it.
Third, if your thyroid has been surgically removed and you still don’t feel well despite having “optimized” your TSH on levothyroxine, there is some credible evidence that you might feel better with the addition of T3 to your T4 therapy.
The regulation of thyroid hormone conversion, transport, and action at the tissue/cellular level is extraordinarily complex. At this time, we do not have accurate ways to measure thyroid hormone levels in individual tissues. We do, however, have the TSH test. When T4/T3 levels drop a tiny bit within the normal range – assuming that’s not normal for your tissues – TSH will shoot way up above the upper limit of normal. When T4/T3 levels rise a tiny bit within the normal range – assuming that’s not normal for your tissues – TSH will drop way below the lower limit of normal.
Is it a leap of faith to assume that TSH reflects the tissues’ thyroid status, as opposed to merely the pituitary’s T3 content? Perhaps. But in my opinion, Alt Med has not offered a more plausible theory. I simply do not believe that there are millions of people with undiagnosed hypothyroidism, for whom replacement therapy will solve their problems.
Questions? Comments? Let me know below! By using this site and interacting with me in the Comments, you agree to abide by my Disclaimer.