Key points
The full panel shows what a single TSH number misses; the table shows what each part of it adds.
| Situation | How it helps | What human studies showed |
|---|---|---|
| TSH + free T4 and T3 together | catch what TSH alone misses: impaired T4 → T3 conversion, central hypothyroidism, thyroid hormone resistance | — |
| Anti-TPO and anti-Tg antibodies | markers of an autoimmune process that may precede a TSH shift by years; identify who needs regular monitoring | Whickham, 20-year follow-up: antibodies + elevated TSH — odds ratio for overt hypothyroidism 38 in women and 173 in men (Vanderpump, PMID 7641412) |
| Narrow TSH upper bound — 2.5 mIU/L | with symptoms, TSH 2.6–4.0 is already subclinical hypothyroidism; when planning pregnancy, TSH > 2.5 is a reason for further work-up, in the first trimester — to consider therapy | without anti-TPO carriers and ultrasound signs of thyroiditis in the reference sample, the upper bound is 2.5 instead of 4.0 mIU/L, NACB position (Wartofsky & Dickey, PMID 16148345) |
| Selenium in autoimmune thyroiditis | cofactor of T4 → T3 conversion and modulator of the immune response; lowers anti-TPO titer | RCT, sodium selenite 200 mcg/day for 3 months: anti-TPO fell by 36% in patients with baseline titer above 350 IU/mL, normalized in 9 of 36 vs 2 of 34 on placebo, ultrasound echogenicity improved (Gärtner, PMID 11932302); in pregnant women with anti-TPO — significantly less postpartum thyroiditis and permanent hypothyroidism (review Drutel, PMID 23046013) |
| Ultrasound by ACR TI-RADS | points for 5 nodule features give a TR1–TR5 class and a decision: biopsy or follow-up | — |
| Iodine status (urinary iodine) | before prescribing iodine-containing drugs and supplements, shows whether iodine is needed | — |
What TSH Measures: HPT Axis and Feedback Inertia
Standard thyroid screening in many countries reduces to a single marker — thyroid-stimulating hormone (TSH). Convenient for population screening, but not for a decision in an individual patient. TSH is a pituitary hormone, not a thyroid hormone. It reflects not thyroid function directly, but the negative feedback of the hypothalamic-pituitary-thyroid (HPT) axis.
A normal TSH in a patient with hypothyroid symptoms is a reason to look wider, not to close the visit.
TSH is synthesized by thyrotrophs of the anterior pituitary under control of hypothalamic thyrotropin-releasing hormone (TRH). Low circulating free T4 → TRH ↑ → TSH ↑. This is negative feedback along the hypothalamus → pituitary → thyroid loop.
A key property of this loop is inertia. TSH responds not to acute changes in function, but to chronic imbalance spanning weeks to months. Acute states — stress, infection, fasting, early pregnancy — shift TSH without changing thyroid function. Conversely, early Hashimoto thyroiditis may show normal TSH with already elevated antibodies and ongoing tissue destruction (Hennessey, PMID 26200184).
The principled consequence: TSH is a late marker. It reports the result of accumulated dysfunction, not its cause or its onset.
The Narrow Range: 2.5 vs 4.0
Lab reference ranges of 0.4–4.0 mIU/L are wide population intervals, calculated on mixed cohorts. The problem: these cohorts traditionally included undiagnosed subclinical hypothyroidism patients and anti-TPO carriers — which artificially elevated the upper bound.
When anti-TPO carriers and ultrasound-confirmed thyroiditis are excluded from the reference population, the upper bound shifts to 2.5 mIU/L. This is the position of the National Academy of Clinical Biochemistry (NACB) and successive ATA revisions (Wartofsky & Dickey, PMID 16148345).
| Category | TSH, mIU/L |
|---|---|
| Population reference | 0.4–4.0 |
| Narrow upper bound (anti-TPO excluded) | ≤ 2.5 |
| Subclinical hypothyroidism (with symptoms) | 2.6–4.0 |
| First-trimester pregnancy | < 2.5 |
| Second–third trimester pregnancy | < 3.0 |
Clinical consequences:
▸ TSH 2.6–4.0 mIU/L with symptoms = subclinical hypothyroidism, not "normal"
▸ TSH > 2.5 in first-trimester pregnancy = consideration for treatment (fetal neurodevelopment risk)
▸ TSH > 2.5 while planning pregnancy = grounds for full workup, not waiting until 4.0
When a Single TSH Is Not Enough: Six Situations
TSH can read "normal" when the thyroid is in fact dysfunctional — and vice versa.
▸ Non-thyroidal illness syndrome (NTI / euthyroid sick syndrome) — severe systemic illness, sepsis, major trauma can suppress TSH while conversion is impaired and fT3 is low. Standard reference ranges do not apply.
▸ Estrogens and oral contraceptives — raise thyroid-binding globulin (TBG), shift the balance of free and bound fractions, can produce a reassuringly normal number despite tissue-level deficit.
▸ Thyroid hormone resistance (RTH) — rare but real, with normal or elevated TSH alongside high fT3/fT4. Without an expanded panel, this is missed entirely.
▸ Interfering medications: biotin (>5 mg/day) — falsely low TSH on immunoassays; corticosteroids, lithium, amiodarone, dopamine, metformin — each distorts results differently.
▸ Isolated pituitary dysfunction (central hypothyroidism) — low or normal TSH with low fT4. Single-marker TSH will miss the diagnosis.
▸ Circadian rhythm: TSH is 30–50% higher in the morning. A 2 PM draw and an 8 AM draw give different numbers in the same patient.
Full Panel: Minimum Diagnostic Set
What I consider the minimum sufficient panel for initial thyroid evaluation in a symptomatic adult:
▸ Hormones: TSH + free T4 + free T3 (always as a trio, not individually)
▸ Antibodies: anti-TPO + anti-Tg (even once — determines management trajectory)
▸ Imaging: thyroid ultrasound with echogenicity, nodules, vascular pattern
▸ Cofactors: ferritin, selenium (red blood cell, not serum), zinc, copper (zinc:copper ratio)
▸ Context: 25(OH) vitamin D, B12, homocysteine, total protein, liver enzymes
▸ Symptom profile: questionnaire + basal morning temperature (3–5 days), resting heart rate
With a tight budget, priority order: TSH + fT4 + fT3 + anti-TPO + ferritin + vitamin D. Six items that close 80% of clinical decisions.
Detailed breakdown of cofactors and why ferritin 70–100 is the clinical minimum: Ferritin, liver and iron.
What the full panel adds to TSH alone:
▸ Anti-TPO and anti-Tg antibodies — markers of autoimmune destruction that may precede TSH shift by years
▸ Free T4 and free T3 — what is actually available to tissues
▸ Thyroid ultrasound — nodules, hypoechogenicity (Hashimoto pattern), vascular pattern
▸ Cofactors of synthesis and conversion: ferritin, selenium, zinc, vitamin D, B12, homocysteine
▸ Symptom profile and basal temperature — the clinical layer without which the lab "floats in air"
Particularly important not to miss low fT3 with normal TSH and fT4 — the classic picture of impaired T4 → T3 conversion, where TSH does not react because T4 is normal. This is thyroid "functional hypothyroidism" — a state where the patient complains of fatigue, hair loss, cold extremities, and the lab is "normal".
Reference Ranges for fT4, fT3 and Reverse T3
The ranges I rely on. Every lab issues its own reference range — check it first, not "generally accepted" values.
Free T4 (fT4). Most immunoassays use a range of 9–22 pmol/L (or 0.8–1.8 ng/dL). The clinically meaningful "functional normal" is the middle third of the range, roughly 14–18 pmol/L in a symptomatic non-pregnant adult. fT4 below 12 pmol/L with TSH above 2.5 mIU/L is no longer the "low end of normal" but a combination that requires repeat measurement and an expanded panel. In pregnancy, fT4 is physiologically higher in the first trimester (human chorionic gonadotropin stimulates the thyroid directly) and declines by the third trimester; trimester-specific laboratory reference ranges are mandatory.
Free T3 (fT3). Typical range — 3.1–6.8 pmol/L (or 2.0–4.4 pg/mL). Functional normal is the upper half of the range, roughly 4.5–6.2 pmol/L. fT3 in the lowest quarter with normal fT4 is the main marker of impaired T4 → T3 conversion, which this article describes as "functional hypothyroidism". Two contexts are critical here: fasting and severe systemic illness physiologically reduce conversion (NTI pattern), so fT3 in the lowest quarter in a patient in the acute phase of pneumonia or after surgery is not an indication for T3 therapy.
Reverse T3 (rT3). A metabolically inactive T3 isomer formed via the alternative deiodinase D3 pathway. The range in most labs is 8–25 ng/dL. High rT3 with low fT3 is the biochemical fingerprint of conversion shunted toward the inactive side under chronic stress, a low-calorie diet, severe illness, high cortisol, or iron and selenium deficiency. According to the consensus on subclinical thyroid disease (Surks et al., PMID 14722150), when symptoms and labs diverge, rT3 adds valuable information, although it is not part of routine screening.
fT3/rT3 ratio — a calculated conversion index. A value below 0.2 (nmol/nmol) or below 6 (pg/ng) points to impaired conversion even with normal absolute values. It is not a diagnosis but a trigger to look for the cause: ferritin, selenium, cortisol, a low-calorie diet, a chronic inflammatory background.
TRAb (TSH receptor antibodies) — the basis for diagnosing Graves' disease; normal is below 1.75 IU/L (method-dependent). They are ordered only when thyrotoxicosis is suspected or to tell hashitoxicosis from Graves' disease. Thyroglobulin (Tg) — not a screening test: after thyroidectomy for papillary or follicular cancer it serves as a marker of recurrence.
Conversion Biochemistry: DIO1, DIO2, DIO3 and Why fT3 Can Be Low
The thyroid secretes mostly T4 (≈ 80 nmol/day) and only 20% of active T3 (≈ 6 nmol/day). The rest of T3 is formed peripherally from T4 by three isoforms of selenium-dependent deiodinases:
▸ DIO1 — mainly in the liver and kidneys; supplies circulating T3 to the blood. Sensitive to propylthiouracil and beta-blockers, including propranolol.
▸ DIO2 — in the brain, pituitary, brown fat and skeletal muscle: local T4 → T3 conversion for the tissue's own intracellular needs. This is the key mechanism for the pituitary: in thyrotrophs T4 is converted to T3 locally via DIO2, and it is intrapituitary T3 that suppresses TSH secretion. Williams and Bassett described DIO2 not merely as an "enzyme" but as a local autonomous regulator of tissue status (Williams & Bassett, PMID 21292729).
▸ DIO3 — in the placenta, fetal brain and skin. Inactivating: converts T4 to reverse T3 (rT3) and T3 to T2. Protects tissues from hormone excess and is expressed under stress and systemic inflammation.
Clinical implications:
1. Selenium deficiency lowers the activity of all three DIOs (all of them are selenium-dependent).
2. Systemic inflammation, sepsis, chronic illness activate DIO3 → T4 is shunted into reverse rT3 instead of T3. The patient is "euthyroid by the lab" but has tissue hypothyroidism — this is the non-thyroidal illness syndrome.
3. DIO2 gene polymorphism (Thr92Ala) — ≈ 12–16% of the population; associated with worse quality of life on T4 monotherapy and a better response to T4 + T3 combination. Not a routine test, but clinically relevant when a patient on adequate T4 still has symptoms.
Bottom line: "normal TSH + normal T4 + low T3" is not lab noise but a signal of peripheral deiodinase dysfunction that needs its own therapeutic response (correcting selenium, zinc, ferritin; considering combination therapy).
Anti-TPO and Anti-Tg: Titers, Progression, Clinical Thresholds
Anti-TPO. Most labs consider values below 35 IU/mL negative, 35–60 IU/mL a gray zone, and above 60 IU/mL clearly positive. Clinically it is useful to distinguish three levels: up to 100 IU/mL (weak positivity, moderate prognosis), 100–500 IU/mL (moderate autoimmune background), above 500 IU/mL (active autoimmune thyroiditis with a high probability of overt hypothyroidism within 5–10 years). The 40% titer drop at baseline values above 1200 IU/mL cited in Gärtner's study was achieved over 3 months of selenium (Gärtner et al., PMID 11932302).
Anti-Tg. Negative is usually below 40 IU/mL. Isolated anti-Tg positivity with negative anti-TPO is less common but clinically significant — especially in female patients with symptomatic hypothyroidism and a normal ultrasound.
Prognostic value. The 20-year Whickham follow-up (Vanderpump et al., PMID 7641412) gave the following odds ratios for developing overt hypothyroidism: positive antithyroid antibodies alone — OR 8 in women and 25 in men; elevated TSH alone — OR 8 and 44; antibodies combined with elevated TSH — OR 38 in women and 173 in men. A logit model showed that TSH above 2 mIU/L at the first examination already significantly increases the probability of hypothyroidism over 20 years, and adding positive antibodies multiplies the risk.
Practical takeaway: positive anti-TPO with TSH 2.0–4.0 mIU/L is not "asymptomatic carriage" but a state with a high 10–20-year probability of becoming overt. This is a patient who cannot be "released with a normal TSH" — they go on regular monitoring every 6–12 months with a focus on cofactors, lifestyle, selenium and vitamin D.
Titer dynamics and interpreting repeat measurements. Antibodies show marked within-person variability: fluctuations of 20–30% between measurements are possible without any real change in immune status. Therefore a titer change of more than 50% on repeat testing after 3–6 months is considered clinically significant. A fall in anti-TPO with selenium, vitamin D, normalized iron, and gluten elimination in patients with celiac disease is a documented phenomenon. The goal of therapy is a lower titer and stable echogenicity on ultrasound, not "zero antibodies".
Selenium in Autoimmune Thyroiditis: Doses, Forms, Expected Dynamics
Mechanism. Selenium is part of selenoproteins — glutathione peroxidases and deiodinases. All three deiodinases (D1, D2, D3) are selenoproteins, so without selenium T4 → T3 conversion is impaired. In parallel, selenium modulates the thyroid immune response, reducing the production of pro-inflammatory cytokines and the activity of autoantigen-specific T lymphocytes (Duntas, PMID 20810577).
Evidence in autoimmune thyroiditis. The main results are in the "Key points" table: Gärtner's 2002 RCT (Gärtner et al., PMID 11932302) and the review by Drutel et al. (PMID 23046013), which confirms the effect is reproducible.
Practical protocol. Dose 100–200 mcg/day, form — selenomethionine or sodium selenite, course length — 3–6 months with repeat anti-TPO assessment. Selenomethionine is better absorbed and incorporated into tissue stores; selenite gives a faster antioxidant effect. Recheck anti-TPO no earlier than 8–12 weeks from the start (antibodies have their own inertia).
Other RCTs. In an RCT by Hu et al., 6 months of selenium yeast lowered anti-TPO, anti-Tg and TSH in patients with Hashimoto's; for the TSH decrease, the subgroup with subclinical hypothyroidism appeared to benefit more (Hu, PMID 33650299). Myo-inositol 600 mg × 2 plus selenium 83 mcg restored euthyroidism in Hashimoto's patients with TSH 3–6 mIU/L (Nordio & Basciani, PMID 28724185).
Status marker. Red blood cell selenium, not serum selenium, is the correct status marker; serum selenium is too sensitive to inflammation and to current supplement intake.
Regional context and food sources. Selenium content in soil, and hence in food, varies 50–100-fold geographically. Eastern Europe, Finland, parts of Germany and the UK are zones of moderate deficiency; a usual diet there does not provide 100 mcg of selenium per day. Among food sources, Brazil nuts lead (50–90 mcg per nut — but the content depends strongly on the batch and the soil of origin), followed by sea fish, seafood, liver and eggs. The "2 Brazil nuts a day instead of a supplement" approach works only with a stable source; with fluctuating content it means either underdosing or overdosing.
Zinc, Iron, Vitamin D: Doses and Timing
Cofactors are not just a list: each has a dose, a duration and a target. All doses are starting guides for an adult without serious comorbidities.
▸ Zinc — citrate or picolinate 15–30 mg/day, taken in the evening, for 3 months; then check plasma zinc and the zinc:copper ratio. If the ratio is < 0.7, add copper 1–2 mg/day. Zinc is a cofactor for TRH synthesis and 5′-deiodinase activity.
▸ Iron — ferritin target 70–100 ng/mL in a symptomatic person, not "> 30" as in abbreviated lab ranges. Ferritin < 50 ng/mL with normal hemoglobin is latent iron deficiency that impairs T4 → T3 conversion and oxygen delivery to muscles. Products: iron bisglycinate 28 mg elemental iron once or twice daily, or intravenous iron sucrose if tablets are not tolerated. Recheck in 8–12 weeks.
▸ Vitamin D 25(OH) — target 40–60 ng/mL (100–150 nmol/L). For deficiency < 20 ng/mL, start cholecalciferol 5000 IU/day for 8 weeks, then a maintenance dose of 2000–4000 IU. Low D is associated with higher anti-TPO and progression of Hashimoto's.
Iodine Status: The Missed Screening Step
Physiology. Iodine is an essential substrate for T4 and T3 synthesis; a T4 molecule contains 4 iodine atoms, T3 — 3. The recommended intake for adults is 150 mcg/day, in pregnancy 220–250 mcg, during lactation 250–290 mcg (Zimmermann, PMID 19460960). Globally, about 2 billion people live with insufficient iodine intake, and about 50% of Europe remains in the mild deficiency zone.
Deficiency. Chronic iodine deficiency raises TSH with a preserved thyroid (thyrotropic hyperstimulation to increase iodine uptake), causes endemic goiter and nodule formation, and in severe forms in pregnant women impairs fetal neurodevelopment. In adults, mild deficiency shows up not as acute symptoms but as a cumulative decline in function — the classic "normal TSH with fatigue and low energy".
Excess. Excessive iodine intake (above 500–1000 mcg/day long term) in predisposed individuals provokes the opposite — induction or flare of autoimmune thyroiditis via the Wolff–Chaikoff effect, increased immunogenicity of thyroglobulin and activation of anti-TPO.
Assessment. The marker of population status is urinary iodine (iodine concentration in a spot or 24-hour urine sample). The norm in adults is 100–200 mcg/L, in pregnant women 150–250 mcg/L. Serum iodine is not used. In an individual patient, a single urinary iodine has high within-person variability; the average of 2–3 samples on different days is more accurate.
Practical takeaway. Before prescribing any iodine-containing drugs or supplements (including kelp, spirulina, iodized salt in high amounts), assess iodine status. A patient with autoimmune thyroiditis and normal iodine does not need extra iodine; for a patient with mild deficiency and no autoimmune background, iodine at a physiological dose via iodized salt or moderate supplements is justified. Pregnant women and those planning pregnancy are a separate category: 150 mcg of potassium iodide per day is recommended by most professional societies in regions at risk of deficiency, and here the risk of fetal hypothyroxinemia outweighs the theoretical risk of activating autoimmune thyroiditis in the mother.
ACR TI-RADS: Reading the Nodule Ultrasound Report
Here is the ACR TI-RADS system (Tessler et al., PMID 28372962) — the standard used in most thyroid ultrasound protocols.
Principle. Each nodule is scored on 5 ultrasound features: composition (cystic/mixed/solid), echogenicity (hyper-/iso-/hypo-/markedly hypoechoic), shape (wider-than-tall or taller-than-wide), margins (smooth/ill-defined/infiltrative), echogenic foci (macrocalcifications, peripheral calcifications, punctate echogenic foci). The sum of points gives the TI-RADS class:
▸ TR1 (0 points) — benign (e.g., a purely cystic nodule). No biopsy, no follow-up needed.
▸ TR2 (2 points) — not suspicious. No biopsy.
▸ TR3 (3 points) — mildly suspicious. Biopsy recommended at ≥25 mm, ultrasound follow-up at ≥15 mm.
▸ TR4 (4–6 points) — moderately suspicious. Biopsy at ≥15 mm, follow-up at ≥10 mm.
▸ TR5 (≥7 points) — highly suspicious. Biopsy at ≥10 mm, follow-up at ≥5 mm.
Key warning features: solid composition, markedly hypoechoic pattern, taller-than-wide shape, ill-defined or infiltrative margins, microcalcifications. A combination of two or more is a reason for fine-needle aspiration biopsy regardless of size in borderline cases.
What matters for the patient. TI-RADS is not a diagnosis but a biopsy decision system. An 8 mm TR3 nodule requires neither biopsy nor worry — only follow-up in 12–24 months. A 6 mm TR5 nodule is a reason to discuss biopsy with an endocrinologist and an ultrasound specialist, even if TSH is perfect. A TSH value neither treats nodules nor protects against them — it is a parallel diagnostic track that lab tests cannot replace.
With multiple nodules, the biopsy target is not the largest but the most suspicious by TI-RADS. This error — biopsying a 25 mm TR2 and missing an 8 mm TR5 — happens regularly and is costly.
Ultrasound quality as a variable. TI-RADS works reproducibly only with a standardized description. If the ultrasound report does not describe echogenicity, shape, margins, calcifications and composition for each nodule separately, it is an incomplete report, from which a TI-RADS class cannot be assigned. A patient asking for a repeat ultrasound from a specialist who reports nodules in TI-RADS format is, in this situation, not "overcaution" but a necessary step. An additional plus is elastography and vascular assessment in CFM/PDI mode, which are not part of the TI-RADS scale itself but improve differentiation accuracy.
Follow-up. Repeat ultrasound for TR3–TR4 nodules — in 12–24 months. Nodule growth of more than 20% in two dimensions or new suspicious features are reasons to revisit the biopsy decision, even if the initial class was low. Stable size and a stable picture over 3–5 years of follow-up is a criterion for reducing monitoring to once every 2–3 years.
Decision Principle
The decision to start thyroid therapy must not rest on a single lab value without assessment of symptoms, antibodies, and ultrasound. This is the position of the American Thyroid Association and the Endocrine Society in updated clinical guidelines (Garber et al., AACE/ATA Hypothyroidism Guidelines, PMID 23246686).
In practice this means:
1. Symptoms + full panel + ultrasound — assemble the picture
2. Reconcile clinic with lab: where they match, where they diverge
3. Find the source of divergence: cofactors, conversion, pituitary level, interferents
4. Decide on therapy only after integrating all three data layers
If natural desiccated thyroid (NDT) or T4 monotherapy is on the table — Natural desiccated thyroid (NDT) covers the form choice. For high antibodies and active autoimmune thyroiditis — Low-dose naltrexone (LDN) as immunomodulation.
Levothyroxine: Doses, Intake, Target TSH
▸ Starting dose in an adult < 60 without cardiovascular disease — 1.6 mcg/kg ideal body weight per day, a full replacement dose (Jonklaas, PMID 25266247). For a 60 kg woman that is ≈ 100 mcg/day, for an 80 kg man ≈ 125 mcg/day.
▸ Starting dose at ≥ 65 years — 1.0–1.1 mcg/kg actual body weight, a third less than in younger people. In the Baltimore Longitudinal Study of Aging, the average dose on which people 65+ were euthyroid was 1.09 mcg/kg actual weight (Gavigan, PMID 37391043). Exceeding this dose is associated with a risk of iatrogenic thyrotoxicosis, atrial fibrillation and osteoporosis.
▸ Starting dose with coronary artery disease or a history of atrial fibrillation — 12.5–25 mcg/day, increased by 12.5–25 mcg every 4–6 weeks under ECG and clinical monitoring.
▸ Intake — on an empty stomach, 30–60 minutes before breakfast, with water. Calcium, iron, proton pump inhibitors, coffee — no sooner than 4 hours later. Soy protein reduces absorption.
▸ First TSH check — 6–8 weeks after starting or changing the dose (earlier, feedback inertia means the value has not yet plateaued).
▸ Dose adjustment — in steps of 12.5–25 mcg; a change of more than 25 mcg/day at once risks "overshooting" into hyperthyroidism.
Target TSH on replacement therapy:
▸ Adults < 60: 0.5–2.5 mIU/L
▸ 60–75 years: 1.0–3.0 mIU/L
▸ Over 75: 1.0–4.0 mIU/L (with caution near the lower bound)
▸ Pregnancy, first trimester: < 2.5 mIU/L
▸ Thyroid cancer: the target is set by the oncology protocol (often 0.1–0.5 mIU/L)
Adherence is critical: levothyroxine has a narrow therapeutic index. Missing 2–3 doses a week makes the follow-up TSH unrepresentative.
Subclinical Hypothyroidism: When to Treat
TSH 4.5–10 mIU/L with normal free T4 is a grey zone with the most divergent recommendations. The algorithm I use brings together AACE/ATA and Endocrine Society positions and cardiology data.
▸ TSH ≥ 10 mIU/L — treat always, regardless of age. In Thyroid Studies Collaboration cohorts, TSH ≥ 10 mIU/L was associated with increased coronary heart disease mortality (HR 1.58) and heart failure risk (Floriani, PMID 28329380).
TSH 4.5–10 mIU/L — the decision is stratified:
1. Age < 65 + symptoms + positive anti-TPO → start levothyroxine 25–50 mcg and recheck in 8 weeks. If symptoms resolve and TSH normalizes — continue; if not — reconsider the diagnosis.
2. 65–75 years, no symptoms and no cardiovascular disease → observation, reassessment in 3–6 months. In 30–50% of cases TSH normalizes on its own (transient subclinical hypothyroidism after infection, stress).
3. Over 75, no symptoms → with TSH < 10 therapy is not indicated; moreover, excess levothyroxine in this group raises the risk of atrial fibrillation and osteoporotic fractures (Duntas & Yen, PMID 31482381).
4. Dyslipidemia and TSH 4.5–10 → a 3-month trial of levothyroxine: lipid profile improved — continue; if not — stop.
In subclinical hypothyroidism with positive anti-TPO, progression to overt hypothyroidism is 4.3% per year (Vanderpump, PMID 7641412). This is not "watch forever" but watch until the trend becomes obvious.
A separate note on subclinical hyperthyroidism (TSH < 0.1 mIU/L, normal fT3/fT4): treatment is indicated over 65 (atrial fibrillation risk, HR 1.68), with coexisting osteoporosis and clinically significant symptoms. Young people without symptoms — monitoring every 6–12 months.
Monitoring and Reassessment
After any protocol change — therapy initiation, dose change, cofactor addition — reassess at 6–8 weeks for TSH and 8–12 weeks for antibodies. Earlier — TSH has not yet caught up due to feedback inertia, antibodies have not shifted yet.
▸ Stable picture: monitor every 6 months
▸ Active autoimmune thyroiditis or dose titration: every 8 weeks
▸ Pregnancy: every 4 weeks in first trimester, then per protocol
▸ After acute illness: do not retest within 4–6 weeks of recovery (avoid NTI pattern)
Seasonality: patients with autoimmune thyroiditis often have winter worsening (low vitamin D, low sun exposure, higher inflammation). If reassessment falls in January–February — that is not grounds for immediate dose change, that is grounds to check vitamin D and cofactors.
Who should see a doctor first
▸ Pregnancy — a separate regime. Trimester-specific reference ranges, risks of both hypo- and hyperthyroidism. Self-directed dose changes are unsafe.
▸ Biotin — discontinue at least 72 hours before TSH/fT3/fT4 testing. This is the cause of many "paradoxical" results.
▸ Amiodarone, lithium, corticosteroids — chronic use requires TSH interpretation in the context of drug and phase.
▸ Thyroid nodules TI-RADS 4–5 — a separate oncologic pathway (fine-needle aspiration), not treated "by TSH".
▸ Aggressive correction in elderly with coronary artery disease — risk of arrhythmia or ischemia. Start at minimum doses, titrate slowly.
▸ Selenium — long-term supplementation above 300 mcg/day was associated in cohort studies with a higher risk of impaired glucose metabolism and type 2 diabetes, so a course should have a start and an end rather than become lifelong. Selenium does not replace levothyroxine in overt hypothyroidism and does not cancel ultrasound monitoring of large nodules. If serum selenium exceeds 130 mcg/L, stop taking it (risk of chronic selenium toxicity).
▸ Iodine in Hashimoto's and with positive antibodies — be especially careful: high doses of iodide (e.g., 12.5 mg and above from "therapeutic" protocols) can worsen the course.
▸ Persistently high anti-TPO despite adequate cofactors and therapy is not a reason to escalate selenium doses or add immunosuppressants.
▸ Reference ranges for fT4, fT3 and rT3 are not targets for drug correction: they show where the patient sits within their own normal range, not a reason to shift the levothyroxine dose "for the numbers".
▸ Cofactor doses in this article are starting guides for an adult without serious comorbidities. In pregnancy, kidney failure or on anticoagulants — individual recalculation.
This article is not a substitute for consultation. Any therapy changes — only with the treating endocrinologist who has access to full clinical context.
Bottom line
Treat the patient, not the number: TSH is one of six to eight inputs to a clinical decision. The full panel sees early autoimmune thyroiditis, impaired T4 → T3 conversion and central hypothyroidism where TSH is "normal".
The correct stance: full panel + ultrasound + symptoms + cofactors → decision; a consultation helps put it together.
About the Author
I am Dr. Vladimir Pereligyn, endocrinologist and researcher. I specialize in endocrine, metabolic, and autoimmune protocols with a holistic approach and individualized lab diagnostics. Book a consultation — universum.earth/consultation. Daily clinical breakdowns — @md_pereligyn_thyroid.
Sources
- Hennessey JV. Diagnosis and Management of Subclinical Hypothyroidism in Elderly Adults: A Review of the Literature. J Am Geriatr Soc 2015. PMID 26200184
- Wartofsky L, Dickey RA. The evidence for a narrower thyrotropin reference range is compelling. J Clin Endocrinol Metab 2005. PMID 16148345
- Garber JR et al. Clinical Practice Guidelines for Hypothyroidism in Adults: cosponsored by AACE and ATA. Endocr Pract 2012. PMID 23246686
- Alexander EK et al. 2017 Guidelines of the American Thyroid Association for Diagnosis and Management of Thyroid Disease During Pregnancy and the Postpartum. Thyroid 2017. PMID 28056690
- Jonklaas J et al. Guidelines for the Treatment of Hypothyroidism (ATA Task Force). Thyroid 2014. PMID 25266247
- Biondi B, Cooper DS. Subclinical Hyperthyroidism. N Engl J Med 2018. PMID 29924956
- Williams GR, Bassett JHD. Deiodinases: the balance of thyroid hormone: local control of thyroid hormone action: role of type 2 deiodinase. J Endocrinol 2011. PMID 21292729
- Hu Y et al. Effect of selenium on thyroid autoimmunity and regulatory T cells in patients with Hashimoto's thyroiditis: A prospective randomized-controlled trial. Clin Transl Sci 2021. PMID 33650299
- Nordio M, Basciani S. Myo-inositol plus selenium supplementation restores euthyroid state in Hashimoto's patients with subclinical hypothyroidism. Eur Rev Med Pharmacol Sci 2017. PMID 28724185
- Gavigan C et al. Levothyroxine Dosing in Older Adults: Recommendations Derived From The Baltimore Longitudinal Study of Aging. Endocr Pract 2023. PMID 37391043
- Floriani C et al. Subclinical thyroid dysfunction and cardiovascular diseases: 2016 update. Eur Heart J 2018. PMID 28329380
- Vanderpump MP et al. The incidence of thyroid disorders in the community: a twenty-year follow-up of the Whickham Survey. Clin Endocrinol (Oxf) 1995. PMID 7641412
- Duntas LH, Yen PM. Diagnosis and treatment of hypothyroidism in the elderly. Endocrine 2019. PMID 31482381
Source: TSH Is a Pituitary Hormone, Not a Thyroid Hormone: The Single-Test Blind Spot — universum.earth. md_pereligyn — Перелыгин Владимир Игоревич.
