TSH and free T4: what does each test actually show?
TSH is the pituitary signal to the thyroid. Free T4 estimates circulating unbound thyroxine. Read together—and in the right physiologic and analytical context—they distinguish patterns that neither test can safely explain alone.
The quick answer
TSH asks what the pituitary is requesting. Free T4 asks how much unbound thyroxine is available.
In a stable, nonpregnant adult with an intact pituitary, TSH is usually the most sensitive first test for primary thyroid dysfunction. Free T4 helps classify the pattern as overt or subclinical and becomes indispensable when pituitary or hypothalamic disease, pregnancy, acute illness, medication effects, or assay interference may break the usual TSH–T4 relationship.
TSH is not thyroid hormone
It is secreted by the pituitary in response to hypothalamic and thyroid-hormone feedback. A small change in free T4 can produce a larger, nonlinear change in TSH.
Free T4 is an estimate
Routine free-T4 immunoassays estimate the unbound fraction. Binding proteins, pregnancy, severe illness, medications, and method-specific interference can affect the result.
The pair is the unit of interpretation
The TSH/free-T4 pattern, prior values, symptoms, timing, medication and supplement history, clinical setting, pregnancy status, and local reference intervals belong in the same reading.
The central ruleA normal TSH is reassuring only when the pituitary–thyroid feedback loop is intact and the clinical setting makes the result reliable.
Side by side
Different measurements answer different questions.
Pattern recognition
Five common combinations—and the questions they raise.
These are biochemical patterns, not self-contained diagnoses. Confirm unexpected results and first ask whether pregnancy, illness, medication, timing, or an analytical problem could explain the combination.
Overt primary hypothyroid pattern
Primary thyroid failure becomes likely. Review persistence, cause, symptoms, medications, pregnancy status, and urgency.
Subclinical biochemical pattern
Confirm persistence and interpret magnitude, age, symptoms, TPO antibodies, pregnancy, medications, recovery from illness, and cardiovascular context before a treatment decision.
Overt thyrotoxicosis pattern
The tests show hormone excess but not its cause. Graves disease, thyroiditis, nodular disease, medication, and assay interference remain different pathways.
Subclinical or transient pattern
Repeat and review drugs, illness, pregnancy, age, autonomous thyroid disease, and assay effects. Persistence and degree of suppression matter.
Central or nonthyroidal pattern
In pituitary disease this can indicate central hypothyroidism. Severe illness, medication effects, and assay problems can mimic it; TSH alone cannot settle the question.
MERHI ONE Evidence Map
What the evidence supports—and where interpretation fails.
Each rating applies to the precise claim. Diagnostic evidence includes physiology, assay studies, cohort data, and professional guidance; it is not graded as though a laboratory test were a medication trial.
TSH is the preferred initial test for primary thyroid dysfunction in a stable adult when pituitary function is intact.1,2,4,12
The logarithmic pituitary response makes TSH sensitive to small changes in circulating free hormone. If TSH is abnormal—or if the clinical context makes it unreliable—free T4 adds the information needed to classify the pattern.
Free T4 distinguishes overt from subclinical primary dysfunction and is essential in suspected central hypothyroidism.2,3,10,12
High TSH with low free T4 supports overt primary hypothyroidism; high TSH with normal free T4 is a subclinical biochemical pattern. In pituitary disease, low free T4 with a low, normal, or mildly elevated TSH can confirm central hypothyroidism in the appropriate context.
Acute illness can create thyroid-test patterns that do not represent primary thyroid disease.1,4
Nonthyroidal illness and recovery alter hypothalamic–pituitary signaling, deiodination, binding, and assay behavior. The pattern varies with timing and severity and can resemble central or subclinical disease.
Biotin and other immunoassay interferences can produce a convincing but false thyroid pattern.1,6,7,8
In susceptible platforms, biotin commonly lowers measured TSH and raises free T4 or T3, mimicking hyperthyroidism. Heterophile antibodies, anti-reagent antibodies, thyroid-hormone autoantibodies, and macro-TSH can produce other discordances.
Pregnancy requires gestational-age-, population-, and assay-appropriate interpretation.1,9,11
hCG can suppress TSH early in pregnancy, estrogen increases thyroid-binding proteins and total T4, and free-T4 immunoassays vary by method. Adult nonpregnant reference intervals should not be imported uncritically.
Thyroid antibodies answer an etiologic or risk question—not the same question as TSH and free T4.2,9,10,12
TPO antibodies support autoimmune thyroiditis and can inform progression risk in a compatible biochemical setting. TRAb or TSI helps selected Graves and pregnancy questions. Antibody concentration does not measure thyroid function or symptom severity.
Medication history is part of the test, not an afterthought.1,2,10
Glucocorticoids and dopamine can suppress TSH; amiodarone changes synthesis and conversion and can cause hypo- or hyperthyroidism; lithium affects hormone release; heparin can artifactually raise free T4; anticonvulsants, rifampin, estrogen, and thyroid-hormone dosing or timing can alter the pattern.
Routine universal screening of asymptomatic, nonpregnant U.S. adults improves clinical outcomes.5
The USPSTF finds insufficient evidence to determine the balance of benefits and harms. This does not apply to people with symptoms, pregnancy, pituitary disease, thyroid medication use, or other defined indications.
A broad ‘complete thyroid panel,’ reverse T3, or repeated antibody titers are routinely necessary for every symptom or abnormal TSH.1,2,4,12
More measurements can add false positives and analytical discordance without changing care. T3 has a defined role in selected thyrotoxicosis questions; reverse T3 has no established routine role in diagnosing outpatient hypothyroidism.
Treatment should be started from one borderline result without confirmation or context.1,2,4,5,10
Biological variation, recovery from illness, pregnancy, supplements, medications, age, and assay interference can move results across a reference boundary. Overtreatment and undertreatment both have cardiovascular, skeletal, reproductive, and symptomatic consequences.
Two native pathways
The physiology is shared. Screening policy and pregnancy pathways differ.
United States — ATA / AACE / USPSTF / FDA
- TSH is usually the first test for suspected primary disease; free T4 classifies an abnormal TSH and is required when central disease or discordance is possible.
- For asymptomatic nonpregnant adults, the USPSTF concludes that evidence is insufficient to recommend for or against universal screening; clinically indicated testing is a different question.
- The ATA published new preconception, pregnancy, and postpartum guidance in 2026. Pregnancy results require the current gestational and assay-specific pathway rather than a standard adult interval.
- FDA and laboratory guidance emphasize biotin disclosure because supplement exposure can distort susceptible immunoassays.
Brazil — SBEM / FEBRASGO
- The SBEM position uses TSH as the central marker for primary disease and emphasizes special interpretation in older adults, pregnancy, obesity, kidney disease, medications, and other contexts.
- The published FEBRASGO/SBEM pregnancy position recommends early TSH screening in pregnancy and uses free T4 to distinguish overt from subclinical biochemical patterns.
- Brazilian practice should use local laboratory reference intervals and current national specialty guidance; imported U.S. cutoffs are not automatically Brazilian cutoffs.
- Biotin, acute illness, pituitary disease, and medication effects are analytical and physiologic issues in both markets, even when the laboratory platform or care pathway differs.
Situations that change interpretation
The same number can mean something different in a different setting.
Acute or critical illness
Low T3 is common; TSH may be low or normal, free T4 may fall in severe disease, and TSH can rebound during recovery. Avoid diagnosing chronic thyroid disease from the inpatient pattern alone.
Biotin and assay interference
Hair-and-nail supplements and high-dose products can create false low TSH/high free T4 results in susceptible systems. The required washout depends on dose, kidney function, and assay; follow the laboratory’s instructions.
Glucocorticoids, dopamine, amiodarone, lithium, heparin, anticonvulsants, and estrogen
These agents act at different levels—TSH secretion, hormone synthesis or release, conversion, binding proteins, metabolism, or the assay itself. There is no single generic medication correction.
Thyroid-hormone use
Dose adherence, brand or formulation changes, absorption interactions, blood-draw timing, and recent dose changes influence interpretation. A high free T4 soon after a dose is not equivalent to a stable trough pattern.
Pregnancy
hCG and rising binding proteins change TSH and T4 physiology. Use gestational-age and assay-appropriate ranges; assess relevant antibody and prior thyroid history through a pregnancy-specific pathway.
Pituitary or hypothalamic disease
Low free T4 with a TSH that is not appropriately high is the key warning pattern. TSH is not a reliable dose target in confirmed central hypothyroidism, and adrenal status may need assessment before treatment.
TPOAb, TgAb, and TRAb/TSI
TPOAb most often supports autoimmune thyroiditis; TgAb is useful in selected contexts; TRAb/TSI supports Graves-related questions. None replaces hormone measurement, and positivity alone does not prove current dysfunction.
Common interpretation errors
The thyroid panel is especially vulnerable to shortcuts.
The main risks are a false lifelong diagnosis, missed pituitary disease, and treatment of an analytical artifact.
‘Normal TSH excludes hypothyroidism’
Not when central hypothyroidism, severe illness, medication suppression, or assay problems are plausible. Free T4 and context are essential.
‘Positive TPO antibodies mean the thyroid is failing now’
They support autoimmunity and risk, but current function is determined by the hormone pattern, not antibody positivity alone.
‘Low TSH always means Graves disease’
Thyroiditis, medication, pregnancy, autonomous nodules, illness, and assay interference can create low TSH. Free T4/T3 and targeted etiologic testing are needed.
‘A complete thyroid panel is always better’
Unfocused testing increases incidental and discordant findings. TSH plus reflex or indicated free T4 answers most first-line primary-disease questions.
‘Adult reference ranges work during pregnancy’
Gestation changes TSH, binding proteins, total T4, and free-T4 assay performance. Pregnancy-specific interpretation is required.
‘Biotin is only a vitamin and cannot affect the result’
Biotin can change the analytical signal without changing thyroid physiology, potentially mimicking hyperthyroidism.
What a wrong interpretation can cause
- Unnecessary lifelong thyroid medication
- Missed central hypothyroidism or pituitary disease
- False Graves or Hashimoto labeling
- Maternal or fetal risk from nonpregnancy interpretation
- Unnecessary imaging, antibodies, or repeated testing
- Cardiovascular or skeletal harm from over- or undertreatment
A practical next step
Turn two numbers into a structured thyroid assessment.
- 01
Confirm the exact test, unit, reference interval, collection date, trend, and reason it was ordered.
- 02
Read TSH and free T4 together; add T3 primarily when a low-TSH or thyrotoxicosis question requires it, not as a routine hypothyroidism screen.
- 03
Record pregnancy status, acute illness, pituitary history, medications, supplements—especially biotin—and thyroid-hormone dose and blood-draw timing.
- 04
If the biochemical pattern and the person do not match, pause: repeat appropriately and ask the laboratory about platform interference or an alternative method.
- 05
Use TPOAb, TgAb, or TRAb/TSI only when the result answers a defined etiology, prognosis, Graves, or pregnancy question.
- 06
Escalate low free T4 with an inappropriately low or normal TSH, severe symptoms, pregnancy-specific abnormalities, or a rapidly changing pattern through the appropriate clinical pathway.
Frequently asked questions
Direct answers to the questions people actually ask.
Is TSH enough to assess the thyroid?+
Often as the first test for primary thyroid disease in a stable, nonpregnant person with an intact pituitary. It is not enough when TSH is abnormal, central disease is possible, pregnancy or severe illness changes physiology, or the result is discordant.
Does a normal TSH rule out hypothyroidism?+
It makes primary hypothyroidism less likely in the usual outpatient setting. It does not safely exclude central hypothyroidism, where free T4 may be low without an appropriately high TSH.
Do positive TPO antibodies require treatment?+
Not by themselves. They support thyroid autoimmunity and may change risk or monitoring, but treatment decisions depend on thyroid function, pregnancy status, symptoms, and the applicable guideline.
How long should I stop biotin before testing?+
There is no universal interval for every dose and assay. Some ATA and product materials advise at least two days for common use, but high doses, kidney impairment, and laboratory platforms can require a different plan. Follow the ordering clinician and laboratory.
Should thyroid tests be ordered during an acute illness?+
Usually not as routine screening. Testing is appropriate when thyroid dysfunction could be causing the illness or when a known thyroid condition requires urgent assessment. Otherwise the acute pattern may be misleading.
Why are pregnancy reference ranges different?+
hCG can lower TSH and estrogen increases binding proteins and total T4. Free-T4 methods also behave differently. Gestational age, population, and assay therefore matter.
Do I need a complete thyroid panel?+
Usually not. Start with the smallest evidence-based set that answers the clinical question—often TSH, with free T4 when indicated—and add T3 or antibodies selectively.
Scientific sources
References are part of the reasoning—not decoration.
We prioritize professional guidance, official public-health recommendations, and laboratory-method reviews. Claim-level citations show exactly which source supports each statement.
- 01Open source ↗
American Thyroid Association · 2023
TSH and Thyroid Hormones: ATA-Commissioned Review of Current Clinical and Laboratory Status
- 02Open source ↗
AACE / American Thyroid Association · 2012
Clinical Practice Guidelines for Hypothyroidism in Adults
- 03Open source ↗
Endocrine Society · 2016
Hormonal Replacement in Hypopituitarism in Adults
- 04Open source ↗
National Institute for Health and Care Excellence · 2019; updated 2023
Thyroid Disease: Assessment and Management
- 05Open source ↗
U.S. Preventive Services Task Force · 2015; current statement
Thyroid Dysfunction: Screening
- 06Open source ↗
U.S. Food and Drug Administration · Updated 2022
Biotin Interference with Laboratory Tests
- 07Open source ↗
Association for Diagnostics & Laboratory Medicine · 2020
AACC Guidance Document on Biotin Interference in Laboratory Tests
- 08Open source ↗
Endocrine Reviews · 2018
Interferences With Thyroid Function Immunoassays: Clinical Implications and Detection Algorithm
- 09Open source ↗
American Thyroid Association · 2026
Guidelines for Thyroid Disease in Preconception, Pregnancy, and Postpartum
- 10Open source ↗
Sociedade Brasileira de Endocrinologia e Metabologia · 2022
Brazilian Position Statement on Primary Hypothyroidism in Special Situations
- 11Open source ↗
FEBRASGO / Sociedade Brasileira de Endocrinologia e Metabologia · 2022
Screening, Diagnosis and Management of Hypothyroidism in Pregnancy
- 12Open source ↗
American Thyroid Association · Current patient resource
Thyroid Function Tests
Editorial record