TSH in blood work is thyroid-stimulating hormone, a pituitary signal that rises when thyroid hormone output falls and drops when output is excessive. In adults, most laboratory reference ranges run from about 0.4 to 4.0 milli-international units per liter (mIU/L). Pregnancy, age, and certain medications shift that window, so the number only means something against the range printed on your own report.
The prefix "milli" matters here. TSH circulates at concentrations a thousand times smaller than glucose or cholesterol, which is why a result like 2.1 or 5.8 can look trivial and still be clinically meaningful. A prompt rise in TSH is usually the pituitary's attempt to push a sluggish gland harder, and that feedback loop is the reason TSH typically gets screened before free T4 is checked.
What the TSH Test Measures and What It Cannot Tell You
TSH is a glycoprotein hormone made by the anterior pituitary gland under the direction of the hypothalamus. It binds receptors on thyroid follicular cells and tells them to take up iodine and release thyroxine (T4) and triiodothyronine (T3). When circulating thyroid hormone rises, the pituitary cuts back TSH; when it falls, TSH climbs.
That inverse pattern makes TSH the standard first-line thyroid test. It is also why a TSH result alone cannot distinguish between a thyroid gland that is failing and a pituitary gland that is failing to signal it. A low TSH with low T4 points toward a pituitary problem rather than an overactive thyroid, and only a paired free T4 test separates those cases.
The assay itself is measured by immunometric methods, and different platforms give slightly different numbers on the same sample. The American Thyroid Association has documented that method-related variation can matter when a patient switches laboratories, so comparisons across labs deserve hesitation.
The test cannot tell you antibody status, nodule structure, or iodine intake. Thyroid peroxidase (TPO) antibodies, thyroid ultrasound, and a dietary history answer those questions, and ordering the wrong one first is a common and expensive detour.
Where Standard Adult TSH Ranges Come From and Why They Move
The 0.4β4.0 mIU/L window most people see is a statistical artifact, not a biological law. Laboratories derive reference intervals from the middle 95% of a local healthy population, which means roughly 5% of healthy adults fall outside any given range by construction. The National Academy of Clinical Biochemistry flagged this limitation in its 2002 laboratory guidelines and recommended that reference ranges be population-specific.
Until about 2006, many labs treated 0.5β5.0 mIU/L as normal. A landmark analysis of the National Health and Nutrition Examination Survey by Hollowell and colleagues, published in the Journal of Clinical Endocrinology & Metabolism in 2002, showed that the upper limit in an iodine-sufficient population without thyroid disease or antibodies sits closer to 4.1 mIU/L. Some laboratories shifted, others did not, and two reports from two hospitals can therefore disagree without either being wrong.
| Population | Typical TSH reference interval (mIU/L) | What drives the shift |
|---|---|---|
| Adults, non-pregnant, iodine-sufficient | 0.4β4.0 | Local reference population; assay platform |
| Adults over 70 | Upper limit often 5.5β6.0 | Age-related rise in TSH with maintained T4 |
| First trimester of pregnancy | 0.1β2.5 | hCG stimulation of the thyroid lowers TSH |
| Second and third trimester | 0.2β3.0 | Placental hormones and increased binding proteins |
| Patients on levothyroxine | 0.5β2.5 as a common target range | Clinical convention, not a lab interval |
A TSH of 5.2 mIU/L is elevated in a 30-year-old and unremarkable in an 80-year-old. That single example explains more confusion about thyroid blood work than any other fact, and it is why an endocrinologist reads the whole picture rather than a flagged number.
TRH Stimulation Testing: The Older Dynamic Test Still Used for Central Hypothyroidism
Before the sensitive TSH assays arrived in the 1980s, the standard way to assess the pituitary-thyroid axis was the TRH stimulation test. Protirelin, a synthetic form of thyrotropin-releasing hormone, was injected, and TSH was measured at baseline and again at 15, 30, and sometimes 60 minutes. A healthy pituitary responds with a prompt TSH rise.
The test has largely disappeared from routine practice because modern third-generation TSH assays can detect very low concentrations and distinguish a suppressed from a normal gland. It survives in a narrow niche: suspected central hypothyroidism, where baseline TSH may be normal or low despite inadequate thyroid output. In that setting a blunted TRH response helps confirm pituitary or hypothalamic failure.
According to the Endocrine Society's 2012 clinical practice guideline on hormone replacement in hypothyroidism, central hypothyroidism is diagnosed from a low free T4 with an inappropriately low or normal TSH, and TSH should not guide dosing in those patients. Free T4 becomes the monitoring marker instead, typically kept in the upper half of the reference range.
Expect side effects if you ever undergo the study: nausea, flushing, a transient urge to urinate, and a metallic taste are the commonly reported ones from protirelin administration. The test requires a baseline sample, the injection, and timed draws, so plan on roughly 90 minutes in the lab rather than a quick stick.
Reading the Number: High TSH, Low TSH, and Normal TSH With Symptoms
A high TSH with a low free T4 is overt primary hypothyroidism, and treatment is standard. A high TSH with a normal free T4 is subclinical hypothyroidism, and the decision to treat depends on the degree of elevation, symptom burden, age, and antibody status. The American Thyroid Association's 2014 guideline, chaired by Jonklaas, suggests treatment when TSH exceeds 10 mIU/L and offers it between 4 and 10 mIU/L based on those factors.
A low TSH with high free T4 is overt hyperthyroidism. A low TSH with normal free T4 is subclinical hyperthyroidism, and again observation is reasonable at the milder end. Below 0.1 mIU/L, the risk of atrial fibrillation rises, which is where the conversation changes course.
Normal TSH with persistent symptoms is the hardest category. Sib hashimoto's, poor sleep, anemia, depression, and perimenopause all produce fatigue that mimics thyroid disease. Treating a normal TSH with levothyroxine has not shown consistent benefit in trials, and the Endocrine Society advises against it outside a clinical trial.
A crucial failure case: a patient with a low TSH and low free T4 who is labeled hyperthyroid and started on antithyroid drugs can be harmed, because the real diagnosis may be central hypothyroidism. That is exactly why paired testing exists.
An Elevated TSH Alone Rarely Means Treatment
TSH 4.5β10 mIU/L, no antibodies, no symptoms: repeat in 6 to 12 weeks; most such values do not progress to overt disease. TSH over 10 mIU/L: treat in most adults; the risk of progression to overt hypothyroidism within a few years is substantial. TSH elevated in pregnancy: treat; untreated maternal hypothyroidism is associated with impaired fetal neurodevelopment. TSH elevated while on levothyroxine: dose adjustment, typically an increase, and recheck in 6 to 8 weeks because TSH lags behind T4 changes.
Eleven to Twelve Weeks: Why TSH Monitoring Lags Behind Your Symptoms
The biological half-life of thyroxine is about 7 days in a euthyroid adult, and the pituitary needs several weeks to re-equilibrate after any dose change. This is why the standard advice is to wait 6 to 8 weeks before repeating TSH, and why a test drawn two weeks after starting levothyroxine tells you almost nothing useful.
Compare that to the half-life of T3, which is roughly 1 day, and the design of combination therapy becomes clearer. T3 changes show up fast and disappear fast; T4 changes take weeks to settle. A patient who feels better in week one and worse in week six is describing pharmacokinetics, not a new disease.
If you change the dose and recheck too early, you risk chasing a moving target with a second adjustment and overshooting. If you wait longer than 12 weeks, you prolong undertreatment. The 6-to-8-week window is a compromise rooted in hormone kinetics, and it is the same window used for hyperthyroidism treatment monitoring.
One practical caveat: biotin supplements can interfere with streptavidin-biotin immunoassays and produce falsely low TSH and falsely high free T4. The FDA issued a safety communication about this in 2017. Stop biotin at least 2 days before a thyroid panel, and tell the lab.
Pregnancy, Age, and Medications That Rewrite the Reference Range
Pregnancy is the largest single source of TSH variation. Human chorionic gonadotropin shares a subunit with TSH and weakly stimulates the thyroid, so first-trimester TSH normally falls. The American Thyroid Association's 2017 pregnancy guideline, chaired by Alexander, recommends pregnancy-specific ranges or, where unavailable, a fixed upper limit, and it treats TSH above the pregnancy-specific limit as grounds for levothyroxine.
Levothyroxine requirements commonly rise by 20% to 50% in pregnancy, often within the first trimester, so women with pre-existing hypothyroidism need a TSH check every 4 weeks until mid-gestation. That is one place where the usual 6-to-8-week interval is too slow.
Age is the second big modifier. TSH drifts upward with age without a matching drop in free T4, and the oldest adults often tolerate mild elevation without symptoms. Aggressively normalizing TSH in a 90-year-old risks iatrogenic hyperthyroidism and atrial fibrillation.
Medications matter too. Amiodarone, lithium, interferon alfa, and tyrosine kinase inhibitors can all disturb thyroid function; glucocorticoids and dopamine suppress TSH; and estrogen raises thyroxine-binding globulin, which shifts total T4 without changing free T4. Anyone on more than one of these deserves a full panel, not a lone TSH.
Worked Example: Tracking a Levothyroxine Adjustment From 8.4 to 1.9
Suppose a 45-year-old woman presents with fatigue and a TSH of 8.4 mIU/L, free T4 of 0.9 ng/dL, and positive TPO antibodies. She starts levothyroxine at a weight-based dose, often about 1.6 micrograms per kilogram per day, which for 70 kilograms comes to 112 micrograms. The recheck at 8 weeks shows TSH 3.6 mIU/L. A small increase to 125 micrograms brings the 16-week TSH to 1.9 mIU/L.
The arithmetic matters more than intuition here. A 25-microgram increase moved TSH by roughly 1.7 mIU/L in that example, so reaching 1.9 required two adjustments and about 4 months of calendar time. That is normal, and knowing the pace prevents the disappointment of a "still high" reading at week six.
In pregnancy the same math compresses. A woman on 100 micrograms before conception may need 125 to 150 micrograms by the end of the first trimester, and the 4-week recheck interval exists because fetal thyroid function does not begin until about 12 weeks. Before that point, the fetus depends entirely on maternal hormone transfer.
Cost is worth naming plainly. A generic levothyroxine prescription typically costs a few dollars to about $15 per month in the United States without insurance, while TSH testing runs roughly $20 to $80 depending on the laboratory and whether it is bundled. Over a year of monitoring, testing can exceed the drug cost, which is one reason clinicians avoid unnecessary serial measurements.
What to Do If Your TSH Is Out of Range
If the TSH is high with a low free T4, see a clinician within a few weeks rather than months. If it is low with a high free T4, ask about a beta-blocker for symptoms and request a full panel including free T4, T3, and antibody testing. If it is mildly abnormal and you feel fine, ask whether a repeat in 6 to 8 weeks is appropriate before any prescription.
Bring the actual lab report, not just the number your portal showed. The reference range and assay name matter, and the date of the draw tells the clinician whether a repeat is due. Mention every supplement, especially biotin, and every medication, including over-the-counter ones.
Ask specifically whether free T4 was measured, and whether thyroid peroxidase antibodies were checked if you have never had them. Those two answers shape nearly every downstream decision. If you are pregnant or planning pregnancy, ask whether the lab uses pregnancy-specific ranges.
One boundary bears repeating: this article describes how results are interpreted, not how any individual should be treated. Dosing decisions belong to a clinician who can see the whole record.
FAQ
Can TSH be normal and still mean thyroid disease?
Yes. Central hypothyroidism can present with a normal or low TSH because the pituitary fails to raise the signal despite low thyroid output, so free T4 is the necessary second test. Roughly a small fraction of hypothyroid cases are central rather than primary, and the Endocrine Society's guidance is to monitor those patients on free T4 rather than TSH. Anyone with a normal TSH, low free T4, and symptoms such as headache or visual change deserves pituitary evaluation.
How much does a TSH test usually cost?
In the United States, a standalone TSH often runs $20 to $80 before insurance, and a full thyroid panel with free T4 and antibodies can run $100 to $250 depending on the laboratory. Ordering TSH alone first is cheaper but may require a second draw if it is abnormal, adding a repeat visit cost. Ask for the bundled price before ordering, and note that prices vary by region and by whether the lab is hospital-affiliated.
Does a high TSH always mean I need medication?
No. Values between 4 and 10 mIU/L without symptoms or antibodies are frequently observed and repeated rather than treated, because many normalize on their own. Treatment is generally recommended above 10 mIU/L, in pregnancy, and when symptoms and antibodies are present. Age also matters: an 80-year-old with a TSH of 6 often needs no treatment at all.
Can biotin in a hair supplement distort my thyroid result?
Yes, and the effect is well documented. The FDA warned in 2017 that high-dose biotin can interfere with immunoassays, producing falsely low TSH and falsely high free T4, a pattern that mimics hyperthyroidism. Stop biotin at least 2 days before testing, and tell the lab so they can choose a different method if needed. Multivitamins with smaller biotin doses are less likely to interfere but not risk-free.
Why did my TSH change after switching labs?
Assays from different manufacturers can differ by 10% to 20% on the same sample, and reference ranges are derived from different local populations. That means a result of 3.2 at one lab and 4.1 at another may reflect method, not biology. Stick with one laboratory where possible, and if you must switch, ask the clinician to compare against the new range rather than the old number.
What is a TRH stimulation test, and would I ever need one?
It is a dynamic test that injects synthetic thyrotropin-releasing hormone and measures TSH at timed intervals, typically baseline, 15, 30, and 60 minutes. It is rarely used today because sensitive TSH assays replaced it, but it can help confirm central hypothyroidism when baseline results are ambiguous. Expect about 90 minutes at the lab and transient nausea or flushing from the injection.
Can TSH be too low even when I feel fine?
Yes. Subclinical hyperthyroidism, with a low TSH and normal free T4, is often asymptomatic and is frequently observed rather than treated. Below 0.1 mIU/L the risk of atrial fibrillation and bone loss rises, especially in older adults, which is where treatment discussions begin. A persistently low TSH should still be repeated and paired with free T4 before any decision.
Do I need to fast before a TSH blood test?
No fasting is required for TSH or free T4, and the test can be drawn at any time of day. Timing does matter for levothyroxine users: the standard practice is to draw blood before the morning dose, since a recent dose can transiently affect free T4. If you take biotin, stop it 2 days ahead instead.
This article is general educational information about laboratory testing and is not a diagnosis or treatment plan. TSH results depend on your age, pregnancy status, medications, and the assay your laboratory uses, so interpretation requires a qualified local clinician who can review the full record. Seek urgent care for symptoms such as chest pain, severe palpitations, or confusion, and contact local emergency or crisis services if you are in immediate danger.