Subclinical Hypothyroidism: treat or wait?

Maybe you are standing at the same point right now. You are holding a lab report with a small upward arrow. And you hear two voices giving you opposite advice.

That is unsettling. But it is not a sign that anyone is doing a poor job. The situation is simply not straightforward. A mildly elevated TSH can be an early sign of Hashimoto's. It can also be a chance result, the aftereffect of an illness or a reflection of your age.

In this article we go through the criteria one by one, with the studies they are based on. Afterwards you will know which questions change your decision. And you can make it together with your doctor, instead of standing between two opinions.

What “subclinical” means and what it does not

Subclinical sounds like: a little ill, but not really. Some people hear it as an all-clear. Others hear a quiet reproach that they are making a fuss. Neither is accurate.

Imagine the feedback loop as a heating system with a thermostat. Your pituitary gland constantly measures how much thyroid hormone is in your blood. If there is too little, it turns up the dial with TSH. So TSH is not a thyroid hormone. It is the signal to the thyroid: more, please.

In subclinical hypothyroidism, this dial is already turned up higher than usual. But the home is still warm, because free T4 is within the normal range. The thyroid is getting its work done, just with more drive from above. Another term for this is latent hypothyroidism.

The European Thyroid Association (ETA) distinguishes two levels. A mildly elevated TSH lies between about 4 and 10 mU/l. A markedly elevated TSH lies above that. We will come across this threshold of 10 several more times, because that is where risks and recommendations shift.

I explained how the upper limit of the reference range is set and why it is disputed in the article on the TSH reference range debate. You can find out which values belong alongside TSH in TSH alone is not enough. Here we turn to the next question: What should happen if the value actually lies above the limit?

Seen through the KPNI lens, an elevated TSH is a junction of several systems. In the hormonal system, it is the adjusting screw of a feedback loop. In the immune system, there is often Hashimoto's thyroiditis, which slowly remodels the tissue. In metabolism, building blocks such as iron or selenium are sometimes missing, or too much iodine arrives. And the nervous system can help determine, through sleep, stress and recovery, when and how strongly TSH swings.

And now you know why the same word can stand for very different situations.

Two recommendations, two good reasons: what the large studies show

Imagine you ask a hundred people with mildly elevated TSH whether they feel better on a tablet. Many would probably say yes. But would it be the tablet or the hope? Large studies asked exactly this question, with a dummy pill and without anyone knowing who received what.

From these data, an international panel derived a strong recommendation in the BMJ in 2019: most adults with subclinical hypothyroidism should not receive thyroid hormones. What is interesting, though, are the exceptions the panel itself names.

And the other side? There are smaller studies that were able to measure something in younger groups. They are not counterproof, but they explain why some doctors lean toward treatment when symptoms are present.

Supported by large RCTs and meta-analyses: with mildly elevated TSH, especially in older age, levothyroxine brings no noticeable gain in well-being on average. Still open: whether smaller groups benefit, such as younger people with pronounced symptoms. Clinically, I observe that the question of why often changes more than the question of the tablet. That is experience, not proof from studies.

And now you know why both doctors in the story had good arguments.

Criterion 1: Is the value stable at all?

Have you ever measured your blood pressure twice in a row and got two different numbers? TSH is quite similar. Hardly anyone knows that, though.

TSH is a living value. It follows a day-night rhythm, with a rise during the night. And it responds to serious illness and to the recovery afterwards. A single value is therefore a still frame from a film that keeps on running.

Another observation is even more important. Many mildly elevated values come down on their own. Very large data sets from everyday practice show this.

The guidelines draw the same conclusion from this. The European guideline recommends repeating an elevated TSH with normal fT4 after about two to three months, together with fT4, and measuring TPO antibodies at the same time. The new 2026 guideline of the American Thyroid Association advises a recheck after four to six weeks when you want to conceive. It notes that in older people up to 80 percent of mildly elevated values may be back in the normal range when measured again.

For the repeat test, comparable conditions are what count. Same lab, similar time of day, no recent infection, no major surgery shortly before. And if you take biotin, tell the lab. I will explain why in a moment.

There are situations in which waiting is not the first choice. If fT4 is already low, this is not subclinical but overt hypothyroidism. And in pregnancy, separate, stricter rules apply. Both belong in medical hands promptly.

And now you know why a second look at TSH is often the most valuable first step.

Criterion 2: Is it the thyroid or something else?

Imagine the oil warning light comes on in your car. Would you top up the oil right away without checking? Maybe there really is oil missing. Maybe it is a leak. Maybe the sensor is acting up.

An elevated TSH is a warning light like that. Most of the time it does point to the thyroid, often to early Hashimoto's thyroiditis. But not always. Before hormone is added from outside, it is therefore worth looking at the surroundings.

Recovery after a serious illness

Whether milder infections also shift TSH noticeably has been studied less well. But a gap of a few weeks after an illness does no harm. That is exactly why the timing of the blood draw mattered so much in the story above.

When the lab is misled

Lab values come from test procedures, and these procedures have weak spots. Rarely, TSH in the blood binds to antibodies and forms large complexes. This is called macro-TSH. The test then measures a lot of TSH, even though little of it is biologically active.

Better known is biotin, also called vitamin B7. It is found in high doses in many supplements for skin, hair and nails. Many lab tests use biotin as a building block. If you take a lot of it yourself, it can interfere with the measurement.

Whether and for how long you pause biotin before a blood draw is something to clarify with your doctor or the lab. What matters is that nobody overlooks it.

Weight, medication and iodine

Metabolism as a whole also has a say. With severe obesity, TSH is often somewhat higher. Whether that is a cause, a consequence or simply a companion has been debated for years.

Some medications can also raise TSH, such as lithium or amiodarone. The same applies to large amounts of iodine, for example from iodine-containing contrast agents. Please never stop such medications on your own. Raise the question with your doctor.

How to assess your iodine supply is explained in Putting iodine in context. You can find the controversy around Hashimoto's in Iodine with Hashimoto's.

Iron and selenium: the quiet building blocks

The thyroid is a workshop. It needs not only iodine as a raw material, but also tools. The most important tool is called thyroid peroxidase, TPO for short. It builds iodine into the hormone. And this enzyme needs iron to work.

Selenium has a narrow window between too little and too much. That is why it should be measured, not guessed. You can find out how ferritin and selenium are sensibly interpreted in Selenium, zinc, iron and vitamin D and in Ferritin: what is normal?.

From the perspective of functional medicine, this is the core of this criterion. The thyroid is rarely the first link in the chain. Recovery after illness, iron, iodine, selenium, weight and medication can all shift TSH. Only once these questions are answered does it become clear whether the feedback loop needs long-term support.

And now you know why the question of why comes before the question of the tablet.

Criterion 3: How high is the TSH? The threshold at 10

Is a TSH of 6 the same as one of 12? On your lab report, both sit behind the same small upward arrow. For your body, though, they mean different things.

Imagine a bridge carrying too much traffic. It copes well with a few extra cars. Beyond a certain load, cracks appear. For TSH, based on current knowledge, this load limit lies at roughly 10 mU/l.

The level of TSH also says something about the future. The higher it is, the more likely subclinical hypothyroidism is to become overt hypothyroidism, in which fT4 falls as well.

Most professional societies draw a similar line from this. The European guideline recommends treatment for a TSH above 10 in people under about 65 to 70, even without symptoms. The JAMA review by Biondi and colleagues also sees a possible indication from 10. The BMJ panel, by contrast, only explicitly exempts values above 20 from its recommendation against hormones. So between 10 and 20, the guidelines do not fully agree.

And now you know why the number 10 appears in almost every guideline.

Criterion 4: Antibodies show where the trend might be heading

Positive antibodies sound threatening. As if your body were attacking itself. That picture is not entirely wrong. But it is incomplete.

TPO antibodies are directed against the enzyme the thyroid uses to build its hormone. They are the typical hallmark of Hashimoto's thyroiditis. Many people carry them for years without any change in function. In others, the gland slowly remodels.

This fits with the numbers above. With antibodies, TSH returned to the normal range much less often in the Cardiovascular Health Study, and in Huber's study progression was more than twice as common with antibodies. The European guideline therefore recommends measuring TPO antibodies as part of the repeat test.

Seen through the KPNI lens, antibodies are above all a trace of the immune system. The new ATA guideline considers it plausible that they reflect a more general tendency toward autoimmunity. If you want to go deeper into the question of why the immune system targets the thyroid, read Hashimoto's: causes in the immune system. What can measurably change antibody levels is covered in Lowering Hashimoto's antibodies.

One distinction is important. Antibodies are a reason to look more closely and more regularly. With normal function, they are not on their own a reason for hormone therapy. According to the new data, this even applies while you are trying to conceive, as you will see further below.

And now you know why the same TSH number is read differently with and without antibodies.

Criterion 5: At 35, the same value reads differently than at 85

At first this sounds unfair. Why should a different number apply to older people? The answer is physiology. The feedback loop shifts over the years.

These randomized trials now exist, and you already know them: TRUST in people aged 65 and over and the pooled analysis in people aged 80 and over. Neither found a benefit for symptoms. In younger people, the picture is less clear.

The European guideline therefore recommends age-specific reference ranges. For very old people from about 80 to 85, it advises watchful waiting with a TSH up to 10. The JAMA review also warns that levothyroxine can trigger an artificially overactive thyroid, especially in older age.

In the anthroposophic view, the thyroid sits between head and trunk and sets the rhythm for the metabolism of the whole body. The idea that this rhythm may become calmer in old age fits well with this view. This is a perspective from clinical tradition, not a study finding. More on this in The thyroid from an anthroposophic view.

And now you know why your year of birth counts when your TSH is read.

Criterion 6: Take symptoms seriously and check them honestly

Tired, always cold, no drive. And then one of two sentences. Either: That is not coming from your thyroid. Or: That is surely your thyroid. Both sentences can be wrong.

The typical symptoms of an underactive thyroid are nonspecific. Tiredness, feeling cold, weight gain and a sluggish mind have many possible sources. Iron deficiency is one of them, as are poor sleep, chronic stress and depressive phases. One possible explanation for why Feller's meta-analysis showed no gain on average: in many participants, the symptoms may not have come from the thyroid at all. That is an interpretation, not a study result.

At the same time, the guidelines take symptoms seriously. The BMJ panel itself writes that its recommendation may not fit people with severe symptoms or young adults. The JAMA review regards treatment as worth considering in younger and middle-aged people with matching symptoms. And the European guideline describes a path in between.

For such a trial to be honest, it needs a yardstick. Write down beforehand what burdens you most, and give each symptom a number from 0 to 10. That way you can compare later instead of relying on memory alone. And ask whether iron, sleep and stress have already been looked at.

Seen through the lens of lifestyle and KPNI, exhaustion is almost always a joint effort. The nervous system can be stuck in chronic stress, the cortisol hormone axis can be shifted, metabolism can be missing building blocks, and the immune system may be busy with Hashimoto's thyroiditis. I describe how these threads come together in Normal values, symptoms anyway and in Tired despite levothyroxine.

And now you know why symptoms are an important criterion, but not proof.

Criterion 7: Wanting a baby and pregnancy change the calculation

As soon as a child is planned, every number feels heavier. It is no longer just about you. It is also about a brain that depends on your thyroid hormone in the first weeks.

The unborn child mainly makes its own active T3 in the brain, from the T4 that comes from the mother. That is why T4 matters most during this time. And that is why separate rules apply when you want to conceive and during pregnancy. In some places they are stricter, in others surprisingly cautious.

Where does this caution come from? Mainly from randomized trials of recent years. Let me introduce two of them.

So why are higher values treated anyway? Observational data suggest that the benefit grows with the level of TSH. But they also show that treatment is not automatically harmless.

The previous ATA version from 2017 had already set pregnancy-specific standards (Alexander et al., Thyroid 2017). The 2026 version builds on it, but places more weight on timing and on confirming a finding. You can read how your cycle, the wish for a child and the thyroid interact in The thyroid and female hormones. Why iron matters so much during this time is covered in Iron deficiency in pregnancy.

And now you know why the same number is read differently when you want to conceive, and why not every deviation needs a tablet right away.

The decision grid: all criteria at a glance

Maybe your head is spinning now. Seven criteria are a lot. So here is everything on one page, the way I go through it with people in my consultations.

Do you see what happened in the story from the beginning? One doctor looked mainly at the wish for a baby. The other mainly at the large studies. Both criteria are in this grid. Only together with the repeat test, antibodies and context did they form a picture.

And now you know why there is no one-size-fits-all answer for subclinical hypothyroidism.

What is well supported and what remains open

I want to show you where the data offer solid ground and where we rely on observations or clinical experience.

Three levers for your next appointment

You do not have to make this decision alone. But you can prepare for it well. You can put these three steps into practice right away.

And now you know why “treat or wait” is not a matter of belief, but a weighing of options that you can help shape.

Frequently asked questions about subclinical hypothyroidism

Further reading