Skip to main content
Choosing Care

A TSH of 3.8 with symptoms: the four tests nobody ordered.

A TSH of 3.8 is inside the range and is the only thyroid test most people ever get. The four that were not run are free T3, reverse T3, and thyroid antibodies, and a positive antibody with a normal TSH is a diagnosis nobody made.

Daniel Tagge, MD5 min read

A TSH of 3.8 mIU/L falls inside almost every laboratory reference range, which typically runs to about 4.5. So the result comes back normal, the thyroid is crossed off the list, and you are left with fatigue, cold hands, dry skin, hair that is thinner than it was, a cycle that has changed, and no explanation.

The TSH is not wrong. It is one test, it was asked one question, and it answered that question correctly. The problem is that it was the only test run.

What TSH actually measures

TSH is a pituitary hormone. It is the signal your brain sends telling the thyroid to work harder. It is not a thyroid hormone at all.

So a TSH result tells you how loudly your pituitary is shouting. It does not tell you whether the thyroid heard, whether it produced enough T4, whether that T4 was converted into the active T3 your cells actually use, or whether your immune system is quietly dismantling the gland.

Those are four different questions and they need four different tests.

The range is the other problem

Reference ranges are built from a population of people who had the test done. Those populations include a meaningful number of people with undiagnosed thyroid disease, which widens the range at the top compared with a range drawn from a healthy population.

Many people with thyroid symptoms report feeling better with a TSH toward the lower part of the range. A 3.8 sits inside the range and is entirely compatible with symptoms. Where a given person does best is a clinical judgment about that person, not a universal target.

One specific case where this stops being a matter of opinion: if you are pregnant or trying to conceive, the targets are lower than the general range, and a 3.8 is a number to act on rather than observe. That is worth raising with whoever is managing your care rather than waiting to be told.

The four that were not run

Free T4. How much of the storage hormone you are actually producing. TSH is the request; free T4 is the delivery.

Free T3. The active hormone, the one your cells use. Most T3 is made outside the thyroid by converting T4, and that conversion depends on selenium, zinc, and iron. Someone with a ferritin of 20 can have a perfectly functional thyroid and still be short of active hormone, which is one of several reasons those two numbers are worth reading together.

Reverse T3. The inactive form. Under stress, illness, caloric restriction, or heavy training, the body shunts T4 toward reverse T3 instead of T3. The panel then looks acceptable while the active hormone is being diverted.

Thyroid antibodies, TPO and thyroglobulin. This is the one that matters most, and it is the one almost never ordered.

The diagnosis hiding behind a normal TSH

Positive thyroid antibodies with a normal TSH means autoimmune thyroid disease that has not yet exhausted the gland.

The immune system is attacking the thyroid. The thyroid is compensating, which is why the TSH still reads inside the range. That compensation does not last indefinitely, and antibody-positive people progress to overt hypothyroidism at a much higher rate than antibody-negative people.

That is a real diagnosis. It changes what gets monitored, how often, what happens in pregnancy, and what other autoimmune conditions are worth having on the radar, because they cluster. And it is a diagnosis nobody made, because the only test ordered was the one that comes back normal until quite late.

I am not arguing that everyone with antibodies needs treatment today. Many do not. I am arguing that knowing is different from not knowing, and that "your thyroid is fine" is a different sentence from "your thyroid is under attack and currently keeping up."

What a chatbot does with this

Reasonably well, actually. Ask about a TSH of 3.8 with symptoms and a current model will usually explain that TSH is a pituitary signal, note that optimal is narrower than the reference range, and suggest free T4, free T3, and antibodies. That is a genuinely good answer and better than the one a lot of people get in person.

Where it stops is the same place it always stops. It cannot order any of them. It cannot palpate your thyroid, which occasionally finds a nodule that changes the plan entirely. It cannot send you for the ultrasound. And it does not know what it was not shown, which in this presentation is nearly everything.

The question a 3.8 raisesWhat answers itCan a chatbot do it?
Is the gland producing?Free T4It can suggest it. It cannot order it.
Is it converting to active hormone?Free T3No
Is T4 being diverted under stress?Reverse T3No
Is this autoimmune?TPO and thyroglobulin antibodiesNo
Are the conversion cofactors there?Ferritin, selenium, zincNo
Is there a nodule?Physical exam, then ultrasoundNo
Does pregnancy change the target?Yes, and it needs managingIt can flag it
Is it moving?Repeat testing over time, same labNo

What I would order

A full thyroid panel rather than a TSH: TSH, free T4, free T3, reverse T3, TPO antibodies, thyroglobulin antibodies. Ferritin with iron studies, selenium, and zinc, because those are the conversion cofactors and the fix is often there rather than in the gland. A morning cortisol, because the stress axis and the thyroid axis are not independent and a diverted T4 usually has a reason.

Then I examine your neck, which sounds old-fashioned and is not replaceable by any panel.

The point

A TSH of 3.8 is the answer to a question you did not ask. You asked why you feel like this. The lab answered whether your pituitary is shouting.

The gap between those two questions is where thyroid symptoms live for years at a time, and closing it costs one broader panel rather than one narrow one. If the antibodies come back positive, you have a diagnosis. If everything comes back clean, you have genuinely ruled out the thyroid and can stop wondering, which is worth something too.

If you have a TSH sitting in the threes and nothing else was ever run, that is what the Precision Health Assessment is for. The symptom picture is at thyroid symptoms and fatigue, the full thyroid panel is part of how I work, the iron half of the story is here, and if you want to talk it through first, apply to work with me.

The Precision Health Report

The science of optimizing your biology, sourced and translated, with no hype. A new issue by email every Friday.

Get the Report by email

Educational only, not medical advice. Unsubscribe anytime. Privacy Policy

Daniel Tagge, MD

Written by

Daniel Tagge, MD

Board-certified family physician and founder of Tagge Health. Certified in Health Optimization Medicine.

About Dr. Tagge

Lab reference

  • TSH: the laboratory range, the evidence, and what moves itThe clear harm thresholds for TSH are at or above 10 mIU/L (coronary events and coronary deaths rose in pooled cohorts) and below 0.1 (coronary deaths and atrial fibrillation rose). Inside the laboratory range the evidence is contested: 85 percent of antibody-free US adults sit below 2.5, the lowest later thyroid dysfunction and lowest mortality sat near 1 to 2 in two cohorts, but TSH rises with age, higher-normal TSH went with longer life in one Dutch cohort, and randomized trials found no symptom benefit from treating a mildly raised TSH.
  • Free T4 and free T3: the laboratory range, the evidence, and what moves itFree T4 and free T3 are the unbound fractions of the two thyroid hormones, and about four fifths of the T3 in blood is made from T4 outside the thyroid by deiodinase enzymes. In 134,346 adults pooled from 26 cohorts, cardiovascular and mortality risk was lowest with free T4 in the 20th to 40th percentiles of the range (about 1.05 to 1.15 ng/dL) and rose in the top fifth, where atrial fibrillation, sudden cardiac death and about three fewer years of life were also seen; a low T3 marks illness and calorie restriction and was the pattern in offspring of long-lived families, so no study supports aiming high in either range, and no trial has targeted a position inside it.
  • TPO antibodies: the laboratory range, the evidence, and what moves itA positive TPO antibody marks autoimmune thyroiditis, not hypothyroidism: about 11 percent of Americans carry it, most with a normal TSH, and in a 20-year British cohort antibodies alone raised the odds of later hypothyroidism 8-fold in women while antibodies plus a raised TSH raised them 38-fold. Selenium lowers the titer in pooled trials without proven clinical benefit, thyroid hormone did not improve live births in 952 antibody-positive women with normal thyroid function, and the cutoff for 'positive' varies several-fold between laboratory methods.

Measure it

Want these numbers on your own chart?

Testing is chosen by judgment rather than protocol, interpreted against your own history, and repeated only when it will change a decision. That reading is where care with me begins.

Explore working with Dr. Tagge