TSH blood test: what high, low and normal results mean

The hypothalamus signals the pituitary, the pituitary releases TSH, and TSH prompts the thyroid to produce mainly thyroxine (T4) plus some triiodothyronine (T3). Circulating thyroid hormones feed back to the brain and pituitary. Small changes in free T4 can produce larger changes in TSH, which is why TSH is sensitive in primary thyroid disease.^{See reference 22,See reference 77}

Many adult reports use an interval somewhere around 0.4 to 4 mIU/L, but the endpoints vary with assay, laboratory and reference population. A result just outside one interval could sit inside another. The report, not a wellness chart, should supply the comparison range.^{See reference 22,See reference 44,See reference 66}

TSH is commonly reported as mIU/L or µIU/mL. These units are numerically equivalent: both numerator and denominator differ by a factor of 1,000. Do not convert either of them to thyroid-hormone units such as pmol/L or ng/dL.

These combinations describe common biochemical patterns. They do not identify the cause, prove that a pattern will persist or replace review of medicines, illness, pregnancy and assay interference.^{See reference 22,See reference 33,See reference 44}

These nine combinations summarize common and discordant patterns described in current thyroid-testing guidance; they do not assign a cause.^{See reference 22,See reference 33,See reference 44,See reference 1313}

This verification sequence follows current laboratory and clinical guidance on paired testing, repeat confirmation, medicines, acute illness and assay interference.^{See reference 22,See reference 33,See reference 44,See reference 55}

Specimen choice, preparation, timing and repeatability vary by assay and clinical purpose, so the reporting laboratory's instructions take priority.^{See reference 22,See reference 33,See reference 44,See reference 66,See reference 1212}

After a levothyroxine dose or brand change, the American Thyroid Association gives about 6 to 8 weeks as a common adult example before checking TSH. That is different from retesting an untreated symptom question, and pregnancy or specialist care may use another schedule.^{See reference 1818}

Biotin can interfere with some immunoassays and may produce falsely high or falsely low results depending on the test design. The safest advice is to disclose the product and dose, then follow assay-specific instructions from the clinician or laboratory rather than using one universal stop time.^{See reference 22,See reference 33,See reference 55,See reference 77}

These are possible contexts—not diagnoses—and the paired free T4, history, examination and repeat pattern determine the next question.^{See reference 22,See reference 33,See reference 77,See reference 1313,See reference 1414}

TSH itself usually does not cause symptoms. Symptoms come from the thyroid-hormone state or another condition. Seek prompt medical assessment for severe or worsening chest pain, marked breathing difficulty, palpitations or feeling faint; do not wait for an online interpretation.^{See reference 1717}

A high or low TSH does not point to one universal diet, supplement or medicine. Treatment depends on the paired free T4 result, persistence, cause, symptoms, pregnancy, age, heart and bone risk, and whether someone is already taking thyroid hormone.^{See reference 33,See reference 77}

Treatment differs by confirmed cause. In particular, antithyroid medicines reduce new hormone production in true hyperthyroidism but do not treat hormone release from destructive thyroiditis.^{See reference 33,See reference 77,See reference 1010,See reference 1111,See reference 1414,See reference 1616}

Excess iodine can alter thyroid function, and laboratory testing has found measurable T4 or T3 in some products marketed for thyroid support. That finding does not establish the contents of every current product, but it is a reason not to use a supplement as a self-directed response to one TSH result.^{See reference 1414,See reference 1515}

In a randomized trial of 737 adults aged 65 or older with persistent subclinical hypothyroidism, levothyroxine lowered TSH but did not improve the trial's hypothyroid-symptom or tiredness scores at one year. A meta-analysis of 21 randomized trials in nonpregnant adults likewise found no improvement in general quality of life or thyroid-related symptoms. These findings do not apply to overt hypothyroidism, pregnancy or every younger person.^{See reference 1010,See reference 1111}

The 2026 American Thyroid Association pregnancy guideline prioritizes laboratory- and trimester-specific TSH and free T4 intervals. When suitable local TSH intervals are unavailable, it conditionally suggests 0.1–4.0 mU/L for the first and second trimesters, based on low-certainty evidence. That fallback is not a universal pregnancy target and should not be used to adjust treatment without maternity or thyroid care.^{See reference 11}

Pregnancy planning also matters because thyroid treatment and monitoring decisions can change before conception and early in pregnancy. If you are pregnant, trying to conceive or receiving fertility treatment, use the care team's reference information rather than the result explorer on this page.

In older people, healthy-population TSH distributions often rise with age. Age-aware interpretation may reduce unnecessary labels and treatment, while still taking persistent symptoms, free T4 and cardiovascular or frailty context seriously.^{See reference 99}

These mistakes reflect documented biological variation, testing limits, acute-illness effects, medicine timing and assay interference.^{See reference 22,See reference 33,See reference 55,See reference 77}

The evidence labels above summarize current guidelines, laboratory reviews, screening guidance and randomized treatment evidence.^{See reference 22,See reference 33,See reference 55,See reference 88,See reference 1010,See reference 1111}

The USPSTF concludes that evidence is insufficient to recommend for or against screening asymptomatic, nonpregnant adults for thyroid dysfunction. This does not apply to people with symptoms, pregnancy, a thyroid history, relevant medicines or another clinical reason for testing.^{See reference 88}

The American Thyroid Association's laboratory review emphasizes that different instruments can return meaningfully different results, reference intervals must match the method and TSH may take weeks or months to settle after an acute change. Trends are most useful when method and circumstances are comparable.^{See reference 22}