01OverviewDefinition, clinical context and the essential points that orientate the chapter.
The label describes biochemistry, not a symptom syndrome. TSH can rise temporarily during recovery from illness, after thyroiditis, because of medicines, after missed levothyroxine doses or through assay interference. A normal free T4 distinguishes this pattern from overt primary hypothyroidism, while central hypothyroidism generally does not present with a clearly raised TSH proportional to hormone deficiency.
Treatment decisions balance progression risk and plausible benefit against lifelong tablet burden and iatrogenic thyrotoxicosis. TSH at least 10 mIU/L is more likely to persist and progress than a mild elevation. Autoimmune evidence, prior destructive treatment and symptoms strengthen the case, but symptoms such as tiredness and weight change remain nonspecific.
A time-limited levothyroxine trial is an experiment with a predefined outcome, not automatic permanent therapy. Agree which symptoms will be reviewed, normalise TSH safely, and stop if there is no meaningful benefit. This approach prevents dose escalation from replacing investigation of sleep disturbance, anaemia, mood disorder, menopause or other common explanations.
Key points
- Subclinical hypothyroidism is a persistently raised TSH with free T4 inside the assay reference range after transient and analytical causes are excluded.
- Confirm the abnormality on two samples separated by about three months before treatment in a stable non-pregnant adult, unless a different urgent context applies.
- NICE advises considering levothyroxine when TSH is 10 mIU/L or higher on two measurements three months apart.
- For symptomatic adults under 65 with TSH above range but below 10, NICE supports considering a six-month therapeutic trial after repeat confirmation.
- Age, symptoms, previous radioiodine or surgery, goitre and thyroid peroxidase antibody positivity modify the chance that a raised TSH reflects persistent thyroid disease.
- If symptoms do not improve once TSH is within range during a treatment trial, consider stopping levothyroxine and look for alternative causes rather than escalating indefinitely.
- Untreated patients with features suggesting underlying thyroid disease usually need annual TSH and free T4; those without such features may be checked every two to three years.
- Pregnancy, planned conception and fertility treatment use separate obstetric-endocrine thresholds and should not be managed from the ordinary adult algorithm.
- In older adults, mild TSH elevation is more often transient and treatment can cause harm through over-replacement, atrial fibrillation and bone loss.
02AetiologyUnderlying causes, associations and risk factors, with why each one matters.
Early autoimmune thyroid failure
Progressive autoimmune loss of thyroid reserve raises TSH before free T4 falls outside its reference interval.
Previous thyroid injury
Surgery, radioiodine or destructive thyroiditis can leave limited gland reserve and a persistent compensated biochemical pattern.
Transient or treatment-related elevation
Recovery from illness, missed levothyroxine, medicines, assay interference and physiological variation can raise TSH without enduring gland failure.
03PathophysiologyThe causal sequence from the underlying abnormality to symptoms and harm.
- 1Thyroid reserve declines
The gland needs stronger pituitary stimulation to maintain free thyroid hormone within the laboratory range as the process continues.
- 2TSH rises through feedback
A small reduction in thyroid-hormone effect produces a larger compensatory TSH increase while circulating free T4 remains normal.
- 3The pattern may persist or resolve
Autoimmune or destructive disease can progress, whereas illness recovery and temporary interference may normalise on repeat testing.
- 4Treatment can overshoot
Levothyroxine given without clear persistent need may suppress TSH and create atrial, skeletal or symptomatic thyrotoxic effects.
04Clinical features and red flagsSymptoms, examination findings, patterns of presentation and time-critical warnings.
TSH remains above the laboratory upper limit while free T4 is normal on appropriately spaced samples, with no acute illness, major medicine change or credible assay problem explaining the finding.
TSH at least 10, positive TPO antibodies, goitre, previous thyroid surgery or radioiodine, and a rising trend make continuing primary thyroid failure more likely than an isolated minor elevation.
Cold intolerance, constipation, dry skin, menstrual change, cognitive slowing and fatigue may support a trial in an eligible younger adult, but symptoms must be documented and competing causes explored.
A free T4 below range means overt hypothyroidism, not subclinical disease. Marked clinical deterioration, hypothermia, bradycardia or reduced consciousness requires urgent evaluation regardless of an earlier mild profile.
Known pregnancy, trying to conceive or assisted reproduction changes the risk-benefit calculation and reference ranges; arrange prompt obstetric or endocrine advice instead of waiting three months routinely.
Palpitations, tremor, insomnia, heat intolerance or a suppressed TSH during therapy suggests over-replacement, which is especially hazardous in older people and those with cardiac or osteoporosis risk.
05InvestigationsWhat to request, why it matters and how to interpret it.
Read from the initial assessment onwards. Tests may run in parallel in urgent care; first-line, preferred, confirmatory, definitive and gold-standard labels appear only when the chapter explicitly states them.
- 01
Repeat TSH and free T4First step - Why
- Confirm persistence and exclude evolution to overt deficiency.
- Interpretation and limitations
- For stable non-pregnant adults, two abnormal TSH results about three months apart underpin treatment decisions; a low free T4 reclassifies the condition as overt hypothyroidism.
- 02
Thyroid peroxidase antibodies - Why
- Estimate whether autoimmune thyroid disease underlies the raised TSH.
- Interpretation and limitations
- A positive test supports persistent autoimmune failure and closer follow-up, but repeating antibody concentrations adds no useful control information.
- 03
Medication and supplement review - Why
- Identify reversible physiology, adherence problems or analytical interference.
- Interpretation and limitations
- Lithium, amiodarone, iodine exposure and some immunotherapies can affect the axis; missed replacement doses and high-dose biotin require a tailored repeat plan.
- 04
Pregnancy testing where relevant - Why
- Prevent use of non-pregnant thresholds in early gestation.
- Interpretation and limitations
- A confirmed pregnancy triggers gestation- and assay-specific assessment with prompt specialist input; routine adult surveillance intervals are inappropriate.
- 05
Focused alternative-cause screen - Why
- Investigate symptoms that may not be thyroid-mediated.
- Interpretation and limitations
- Use history to select full blood count, ferritin, B12, coeliac serology, renal or liver tests, sleep assessment and mental-health review rather than ordering a universal panel.
- 06
Lipid and cardiovascular risk assessment - Why
- Quantify modifiable vascular risk when TSH elevation persists.
- Interpretation and limitations
- Dyslipidaemia can coexist, particularly with higher TSH, but treat cardiovascular risk on its own merits and do not promise levothyroxine will replace standard prevention.
06Differential diagnosisRealistic alternatives and the features that help distinguish them.
Transient illness recovery
TSH can rebound after acute illness; repeating when physiologically stable avoids assigning a chronic thyroid diagnosis prematurely.
Poor levothyroxine adherence
Missed tablets followed by recent resumption can produce raised TSH with a misleadingly normal free T4.
Assay interference
Heterophile antibodies, macro-TSH or another assay-specific interference can produce a TSH result inconsistent with symptoms and free thyroid hormone measurements.
Central hypothyroidism
Low free T4 with a non-raised or biologically inappropriate TSH indicates pituitary disease, not the subclinical primary pattern.
07ManagementImmediate care, first-line treatment, alternatives and escalation.
01ConfirmationValidate the subclinical labelFirst stepA stable adult has raised TSH with free T4 within range.+
- 1Review acute illness, pregnancy possibility, medicines, supplements, prior thyroid treatment and the magnitude and chronology of TSH elevation.
- 2Repeat TSH and free T4 after approximately three months in the routine pathway, adding TPO antibodies once when autoimmune disease status is informative.
- 3Reclassify promptly if free T4 becomes low, and refer when results suggest pituitary disease, assay interference or another complex explanation.
02TSH ten or moreConsider sustained replacementTSH is at least 10 mIU/L on two tests three months apart.+
- 1Discuss symptoms, age, comorbidity, pregnancy plans, progression likelihood, daily administration and the uncertainty of individual symptom benefit.
- 2When treatment is chosen, start levothyroxine using age and cardiovascular-risk principles and titrate TSH into the reference range without suppression.
- 3Review response and safety at appropriate intervals; if treatment is declined, establish surveillance and clear triggers for earlier retesting.
03TSH below tenRun a defined symptom trialAn adult under 65 has persistent mild elevation plus symptoms plausibly related to hypothyroidism.+
- 1Agree two or three specific symptoms and baseline severity, while checking anaemia, sleep, mood, menopause and other likely contributors.
- 2Offer a six-month levothyroxine trial after shared decision-making, adjusting carefully until TSH is within range and avoiding biochemical excess.
- 3If symptoms persist despite a normal TSH, consider stopping treatment, monitor thereafter and redirect evaluation to more likely diagnoses.
04SurveillanceObserve without losing follow-upPreferredTreatment is not indicated, not beneficial or not preferred.+
- 1Check TSH and free T4 annually when TPO antibodies, previous thyroid intervention or another feature indicates underlying thyroid disease.
- 2Use a two-to-three-year interval when there are no underlying-disease features, with earlier testing if symptoms evolve or medicines change.
- 3Provide an explicit pregnancy-planning instruction and advise urgent review for marked deterioration rather than waiting for the scheduled blood test.
Key medicines and prescribing safety1 treatment · regimens, roles and cautions+
Levothyroxine for persistent subclinical hypothyroidism
Select an individual once-daily starting dose using age, body size, TSH elevation and cardiovascular disease; titrate in small steps against TSH and the current BNF rather than using a universal regimen.Benefit is uncertain for mild asymptomatic disease, particularly in older adults. Separate from absorption interactions, avoid suppressed TSH, and use a specialist pregnancy pathway.
08ComplicationsImportant consequences, why they occur and why they matter clinically.
Progression to overt hypothyroidism
Persistent autoimmune loss can eventually lower free T4 and produce clearer metabolic, cardiovascular and reproductive consequences.
Adverse lipid profile
Reduced thyroid effect may worsen atherogenic lipids, particularly when TSH elevation is sustained and other vascular risks coexist.
Pregnancy-related risk
Thyroid physiology and treatment thresholds differ in pregnancy, so an unreviewed outpatient label can lead to under- or overtreatment.
Iatrogenic thyrotoxicosis
Unnecessary or excessive levothyroxine can suppress TSH, causing palpitations and increasing atrial-fibrillation and bone-loss risk over time.
09Monitoring and follow-upTreatment response, safety checks and longer-term review.
- After starting or changing levothyroxine, follow TSH at an interval long enough for steady state and then approximately every three months until stable.
- Once replacement is stable, annual TSH is generally sufficient unless symptoms, adherence, interacting medicines or pregnancy change the situation.
- During a six-month therapeutic trial, track the agreed symptom outcomes as well as biochemistry; normal TSH without improvement argues against continuing solely for symptoms.
- When untreated but autoimmune or structurally damaged thyroid disease is likely, measure TSH and free T4 annually to detect progression.
- In low-risk untreated cases, arrange thyroid tests every two to three years and give the patient a route for earlier clinical reassessment.
- Watch for treatment-related palpitations, atrial arrhythmia and bone-risk implications, especially when TSH falls below range.
10Special situationsVariants, exceptions and circumstances that change the usual approach.
A threshold is not a mandate
TSH at least 10 strengthens the case for treatment, but confirmation, age, comorbidity, patient priorities and special settings still shape the final decision.
Trials need stopping rules
Naming target symptoms before levothyroxine begins makes it possible to judge benefit; otherwise normal day-to-day variation can turn a trial into permanent therapy.
Antibodies predict, not titrate
TPO antibody positivity signals autoimmune substrate and progression risk, while serial titres do not tell clinicians how much levothyroxine to prescribe.
Age changes net benefit
Minor TSH elevation is common in older populations, while overtreatment carries immediate rhythm and longer-term skeletal hazards; restraint can be active care.
Pregnancy is another algorithm
Maternal and fetal considerations, gestation-specific intervals and time-sensitive targets mean even a mild abnormality needs prompt review outside the general-adult pathway.
11Common pitfallsFrequent interpretation and management errors.
- 01
Starting lifelong treatment from one mildly raised TSH after an intercurrent infection.
- 02
Using nonspecific fatigue as proof that the biochemical finding caused symptoms.
- 03
Forgetting to measure free T4 and thereby missing progression to overt disease.
- 04
Escalating levothyroxine until TSH is suppressed because weight has not changed.
- 05
Repeating TPO antibody titres as though they were a treatment target.
- 06
Applying the under-65 six-month trial recommendation to every older adult.
- 07
Waiting three months in a newly pregnant patient instead of seeking prompt specialist advice.