TSH, T4, T3: Reading the Panel
A thyroid panel is three numbers pretending to be one question. This page walks the panel in the order clinicians read it — TSH first, free T4 second, free T3 with context — then takes apart the reference interval itself: where the 0.45–4.5 TSH range came from, why it stretches with age, and what a TSH of 4.5 to 10 is — a grey zone, not a diagnosis.
What the evidence supports
- TSH is the most sensitive single line on the panel: the pituitary amplifies small changes in circulating T4, so TSH drifts first.
- Free-hormone measurements (free T4, free T3) track the biologically active fraction and are preferred over totals.
- The 0.45–4.5 mIU/L interval derives from the NHANES III reference sample after excluding antibodies and abnormal biochemistry — and TSH rises with age even in healthy people (Surks & Hollowell, 2007).
What remains uncertain
- Where the upper TSH limit should sit for older adults is genuinely debated; age-specific ranges are not routine practice.
- Free T3 assays are less standardized than TSH and free T4; results vary more between labs.
- Long-term consequences of a TSH in the 4.5–10 zone with normal free T4 appear modest, and treatment trials in older adults have been largely null.
Evidence last reviewed: September 17, 2026. Conclusions may change as new research is published.
one butterfly, three lines
The Feedback Loop That Writes the Panel
The thyroid is a butterfly-shaped gland, and its lab report is written by a committee. The hypothalamus signals the pituitary, the pituitary releases TSH (thyroid-stimulating hormone), and TSH tells the thyroid to release hormone — mostly T4, plus a small direct share of T3, the more active form; T4 converts to T3 in the liver, muscle, and brain. When circulating hormone rises, the pituitary eases off; when it falls, it leans in. The parent topic, Thyroid & Metabolism: The Master Dial, owns the whole axis; what matters here is its geometry.
That geometry is logarithmic: small dips in free T4 produce large counter-moves in TSH, which is why screening leads with it — it moves earliest and furthest. It reads inversely: a high TSH usually means the gland is underperforming (the pituitary shouting louder), a low TSH means overactivity or suppression. Get that arrow right and the panel stops being alphabetical soup.
Free vs Total: The Distinction That Matters
More than 99.9 percent of circulating T4 rides attached to carrier proteins, mainly TBG; the unbound sliver is the part cells can use. A total T4 counts bound plus free hormone, so it moves whenever the carriers move — and they move in pregnancy and with estrogen-containing contraception, with no change in thyroid status. Free T4 estimates the unbound fraction directly and is the preferred line. The same logic applies to T3, with one caveat: free T3 assays are the least standardized of the three, so treat a surprising free T3 as a retest-and-compare candidate, not a verdict.
| Line | What it measures | Typical span (assay-dependent) | Reading note |
|---|---|---|---|
| 🔍 TSH | Pituitary signal to the gland | ~0.4–4.5 mIU/L | First line; most sensitive to change |
| 📦 Total T4 | Bound + free hormone | ~5–12 µg/dL | Shifts with pregnancy and estrogens |
| 🔓 Free T4 | Unbound, active fraction | ~0.8–1.8 ng/dL | Preferred over total in most cases |
| ⚡ Free T3 | Active hormone, unbound | ~2.3–4.2 pg/mL | Least standardized — compare to your baseline |
| 🧬 TPOAb | Antibody marker, when ordered | Negative / positive | Points at autoimmunity — the Hashimoto's page owns it |
The spans above are typical textbook values, not scripture: assays differ between labs, and your own lab's printed range is the one your numbers should be read against.
How to Read the Numbers in Order
Read TSH, then free T4, then the rest — never one line alone. Common patterns:
- 📈 High TSH, low free T4 — the classic primary hypothyroid pattern; the gland underdelivers and the pituitary pushes back.
- 🌫️ High TSH, normal free T4 — the "subclinical" or mild pattern; a grey zone unpacked below.
- 📉 Low TSH, high free T4 or T3 — the hyperthyroid pattern; overactive gland or overtreatment.
- 🧠 Low TSH, normal free T4 — mild hyperthyroid pattern, but also common with non-thyroidal illness or suppressing drugs.
- ↔️ Normal TSH — status is likely normal for most; persistent symptoms deserve a different investigation, or a repeat test.
Where the 4.5 Cut-off Came From
The number on your lab printout has a specific biography. Hollowell and colleagues analyzed the NHANES III survey (1988–1994, over 16,000 disease-free participants) and derived the interval from a curated subsample of 13,344 — excluding anyone pregnant, taking estrogens or androgens, positive for thyroid antibodies, or showing abnormal biochemistry (Hollowell et al., JCEM, 2002). In that filtered population, the 2.5th to 97.5th percentile interval landed at roughly 0.45 to 4.5 mIU/L, around a geometric mean of 1.40 — and quietly became the default printout range.
Two honest critiques follow. First, the exclusions are aggressive: trimming antibodies and abnormal biochemistry carves the upper tail off the "normal" population, so the range describes a curated healthy cohort, not everyone walking around undiagnosed. Second, age: Surks and Hollowell re-analyzed NHANES III by decade and found the whole distribution shifts upward — the 97.5th percentile was 3.56 mIU/L for people in their twenties but 7.49 for the 80-plus group, and 70 percent of older adults with TSH above 4.5 sat inside their own age-specific range (Surks & Hollowell, JCEM, 2007).
A third critique belongs to individuals rather than populations. Andersen and colleagues sampled 16 healthy men monthly for a year: each person holds a narrow personal set-point, and the population range is far wider than any one person's ordinary swings (Andersen et al., JCEM, 2002). A TSH that has traveled from 1.0 to 3.4 can be a meaningful shift for you while never leaving the printed range; the sibling topic reference ranges vs optimal ranges works through that problem in general.
The Grey Zone: TSH 4.5 to 10
A mildly elevated TSH with a normal free T4 earns the label "subclinical hypothyroidism" — mild thyroid failure in lab language. Three pieces of reading discipline keep it honest. Retest first: TSH drifts with time of day and with recovery from illness, and a single mild elevation often settles when repeated weeks later. Add TPOAb if not measured: antibody-positive grey zones are more likely to progress, the autoimmune layer the Hashimoto's page owns. And respect the upper edge: above 10 mIU/L the case for treatment strengthens; below 10 — especially in older adults — trials are less impressed. The largest, TRUST, found no symptomatic benefit of levothyroxine over one year (Stott et al., NEJM, 2017). Whether you should treat turns on age, antibodies, symptoms, and goals — the treat-or-not page prices it fully. Pregnancy is the standing exception — clinician territory from the first line.
⚠️ When one number is not an answer
Never rebuild your health plan around a single thyroid line drawn on a single morning. If a result is clearly abnormal, or red-flag symptoms ride along — racing pulse, unexplained weight change, a swelling in the neck, severe heat or cold intolerance, or pregnancy — route to a clinician promptly rather than re-reading the printout. And no self-directed starting, stopping, or dose-fiddling of thyroid medication: the panel is information for the conversation, not a prescription pad.
What Can Skew a Reading
- 🕐 Time of day — TSH runs higher overnight and softer by late afternoon; a morning draw keeps your numbers comparable.
- 💊 High-dose biotin — biotin can distort several thyroid immunoassays; mention it to your clinician and pause per their guidance.
- 🤒 Illness — severe non-thyroidal illness transiently rewrites the whole panel; testing during or just after it answers the wrong question.
- 💊 Other medications — amiodarone, lithium, and corticosteroids shift thyroid numbers in known directions; bring the full med list.
- 🏥 Assay-to-assay differences — labs use different platforms with different ranges; compare to your own lab's range and baseline. The testing cadence page builds that routine.
Questions, Answered Briefly
- 🍽️ "Do I need to fast?" No — thyroid lines do not require fasting. If you are drawing glucose or lipids the same morning, the fasting rule is theirs, not the thyroid's.
- 🔍 "Is TSH alone enough?" For screening, mostly yes — it is the sensitive line. Add free T4 when TSH is off, in pregnancy, or when symptoms argue with the screen.
- 🧾 "Why did my lab print 0.4–4.0 and my friend's 0.5–5.0?" Different assays, different calibrations; each lab derives its own range. Read against the paper your result came on.
- 🌫️ "TSH of 6, free T4 normal — am I sick?" You are in the grey zone: retest in a few weeks, check TPOAb, and read the treat-or-not page before worrying. Most grey zones are stable.
- ⏱️ "I do time-restricted eating — does it matter?" Draw-time effects are small, but thyroid conditions and prolonged fasts interact — the hormones and fasting page covers who should be careful.
The Bottom Line
- Read in order, never alone — TSH first as the sensitive signal, free T4 second; patterns, not single lines, carry meaning.
- Free beats total — binding proteins move totals in pregnancy and on estrogens; the unbound fraction reflects status.
- The 4.5 cut-off has a biography — derived from a curated NHANES III sample, and healthy TSH climbs with age; a mildly elevated older adult may simply sit near their own percentile.
- Treat 4.5–10 as a grey zone — retest, add TPOAb, weigh age and goals with a clinician; above 10 the tone changes.
Related Topics
- Hollowell J.G., et al., "Serum TSH, T(4), and thyroid antibodies in the United States population (1988 to 1994): National Health and Nutrition Examination Survey (NHANES III)," Journal of Clinical Endocrinology & Metabolism (2002)
- Surks M.I., Hollowell J.G., "Age-specific distribution of serum thyrotropin and antithyroid antibodies in the US population: implications for the prevalence of subclinical hypothyroidism," Journal of Clinical Endocrinology & Metabolism (2007)
- Surks M.I., et al., "Subclinical thyroid disease: scientific review and guidelines for diagnosis and management," JAMA (2004)
- Andersen S., et al., "Narrow individual variations in serum T(4) and T(3) in normal subjects: a clue to the understanding of subclinical thyroid disease," Journal of Clinical Endocrinology & Metabolism (2002)
- Stott D.J., et al., "Thyroid Hormone Therapy for Older Adults with Subclinical Hypothyroidism," New England Journal of Medicine (2017)
- Panicker V., et al., "Common variation in the DIO2 gene predicts baseline psychological well-being and response to combination thyroxine plus triiodothyronine therapy in hypothyroid patients," Journal of Clinical Endocrinology & Metabolism (2009)