Thyroid hormone levels
Why TSH first
- TSH is the single most useful test of thyroid function
- Log-linear relationship to free T4 - a small fT4 change produces a large TSH change
- Only valid when the pituitary-thyroid axis is intact and steady - not in pituitary disease, not in acute illness, not within 6-8 weeks of a dose change
The panel
| Test | Use |
|---|---|
| TSH | First line, screening, monitoring primary disease |
| Free T4 | Add when TSH abnormal, or when TSH cannot be trusted (pituitary disease, amiodarone, recent treatment) |
| Free T3 | T3-toxicosis, amiodarone, assessing severity of thyrotoxicosis. Not useful in hypothyroidism |
| TPO Ab | Autoimmune thyroiditis; predicts progression of subclinical hypothyroidism |
| TRAb | Graves' - diagnosis, pregnancy risk stratification, relapse prediction |
| Thyroglobulin | Post-thyroidectomy DTC surveillance; factitious thyrotoxicosis |
Epidemiology
- TFTs are among the most frequently requested pathology tests, and among the most frequently misinterpreted
- ~5-10% of adults have subclinical thyroid dysfunction
- Around 1 in 20 hospitalised patients has an abnormal TFT from non-thyroidal illness alone
Synthesis and transport
- Iodide trapped by NIS -> organified by thyroid peroxidase (TPO) onto tyrosyl residues of thyroglobulin (a large ~660 kDa glycoprotein, ~120 tyrosines) -> MIT + DIT couple -> T4 and T3 -> stored in colloid -> proteolysed on TSH stimulation
- T4 is secreted exclusively by the thyroid
- ~80% of circulating T3 is generated peripherally by 5'-deiodination of T4
- T3 is the active hormone (~10x receptor affinity)
| T4 | T3 | |
|---|---|---|
| Half-life | ~6.7 days | ~0.75 days |
| Source | Thyroid only | ~20% thyroid, ~80% peripheral conversion |
| Protein bound | ~99.97% (~70% to TBG) | ~99.7% |
- The long T4 half-life is why levothyroxine is dosed once daily, why a missed dose does not matter, and why TFTs are rechecked at 6-8 weeks not 2
Deiodinases
- D1 (liver, kidney, thyroid) - T4 -> T3
- D2 (brain, pituitary, muscle, BAT) - local intracellular T3; sets pituitary TSH feedback
- D3 - inactivating: T4 -> reverse T3, T3 -> T2
Binding proteins
- TBG (~70%), transthyretin, albumin
- Total T4/T3 track the binding protein; free hormone is what matters
- inc TBG: pregnancy, oestrogen/OCP, tamoxifen, hepatitis, heroin, congenital
- dec TBG: androgens, glucocorticoids, nephrotic syndrome, cirrhosis, severe illness, congenital
- Displacement from binding: furosemide, heparin, salicylates, phenytoin, carbamazepine, NSAIDs
Patterns
| TSH | fT4 | Interpretation |
|---|---|---|
| inc | dec | Primary hypothyroidism |
| inc | normal | Subclinical hypothyroidism |
| dec | inc | Primary thyrotoxicosis |
| dec | normal (check fT3) | Subclinical hyperthyroidism, or T3-toxicosis if fT3 inc |
| dec/normal | dec | Central (secondary) hypothyroidism - TSH is NOT low enough to be reassuring; check the rest of the pituitary |
| normal/inc | inc | TSH-oma, thyroid hormone resistance, assay interference, heparin artefact, amiodarone, non-compliance with recent big dose |
| dec | dec (low fT3, inc rT3) | Non-thyroidal illness |
Non-thyroidal illness (sick euthyroid)
- Sequence: dec T3 first (D3 activation) -> then dec TSH -> then dec T4 in severe illness -> rebound inc TSH in recovery
- *Do not test thyroid function in acute illness unless thyroid disease is genuinely suspected*
- Low T4 + low TSH in ICU = poor prognosis marker, not treatable hypothyroidism
Assay interference - the reason for a discordant result
- Biotin (high-dose supplements, MS trials) - biotin-streptavidin immunoassays -> falsely low TSH with falsely high fT4/fT3, and false-positive TRAb mimicking Graves'
- Stop biotin 48-72 h and repeat
- Heparin / LMWH - in-vitro lipase release -> NEFA displace T4 -> spuriously high fT4/fT3 with normal TSH, mimicking a TSH-oma
- Heterophile / anti-mouse antibodies, macro-TSH -> falsely high TSH
- Anti-T4/T3 autoantibodies
- Discordant result -> repeat on a different platform before investigating a pituitary tumour
Thyroglobulin
- Rises with: high TSH, Graves'/autoimmune thyroid disease, thyroiditis, neonatal period, thyroid cancer, goitre
- Undetectable in factitious thyrotoxicosis (exogenous T4) and in athyreosis - useful in the congenital hypothyroidism workup
- *Anti-thyroglobulin antibody falsely lowers Tg* - must be measured with every Tg in cancer surveillance; the antibody trend then becomes the tumour marker
Drug effects to know
- Amiodarone - transient inc TSH early (D1 inhibition), then AIT type 1 (iodine-induced) or type 2 (destructive)
- Lithium - hypothyroidism, goitre
- Checkpoint inhibitors - thyroiditis then hypothyroidism
- TKIs (sunitinib) - hypothyroidism
- Oral oestrogen raises T4 requirement; PPIs, calcium, iron, soy, coffee reduce levothyroxine absorption
Sampling and monitoring rules
- Recheck TFTs 6-8 weeks after any levothyroxine dose change - shorter is uninterpretable
- Take the blood sample before the morning levothyroxine dose (fT4 peaks 2-4 h post-dose)
- Levothyroxine on an empty stomach, 30-60 min before food, separated by 4 h from calcium, iron, PPIs
When TSH cannot be used alone
- Central hypothyroidism -> titrate to fT4 in the upper half of the reference range
- Pregnancy -> trimester-specific reference ranges; requirement rises ~25-50% in the first trimester
- Recently treated thyrotoxicosis -> TSH stays suppressed for weeks to months; use fT4/fT3
- Amiodarone, dopamine, high-dose glucocorticoids, severe illness -> TSH suppressed independently of thyroid status
Screening
- No case-finding in asymptomatic adults
- Test: AF, unexplained weight change, dyslipidaemia, hyponatraemia, infertility/menstrual disturbance, dementia, T1DM and other autoimmune disease, amiodarone/lithium/checkpoint-inhibitor therapy, pregnancy with risk factors, neonatal screening
Antibodies and autoimmune clustering
- TPO antibody - Hashimoto's (80-99%), also 70-80% of Graves'; predicts progression of subclinical hypothyroidism (~4%/year if positive)
- TRAb - Graves' (70-100%); crosses the placenta - measure in the second trimester in any woman with current or past Graves'
- Autoimmune clustering: T1DM, coeliac, Addison's, pernicious anaemia, vitiligo
- Down and Turner syndromes
- Post-partum thyroiditis (~5-10% of women; higher if TPO positive)
Natural history
- Subclinical hypothyroidism: progression to overt disease ~2-4%/year, higher with TSH >10 or positive TPO
- Many mildly abnormal TFTs normalise on repeat - confirm before treating
- Levothyroxine over-replacement (suppressed TSH) -> AF and osteoporosis, particularly in older women
- The commonest reason for a persistently raised TSH on adequate levothyroxine is adherence or absorption, not dose
- Missed doses then catch-up before the visit -> high fT4 with a still-high TSH
🔒
15 more sections, plus exam facts
Premium unlocks every note across every specialty, and the full exam fact library behind it.
Get premium access