Subclinical hypothyroidism in the elderly

Definition and prevalence. Subclinical hypothyroidism is an elevated serum TSH level in the face of normal or normal-low free thyroid hormone values. Overall prevalence is 4–10% in the general population and up to 20% in women older than 60. Subclinical disorders are more prevalent than overt disease in older populations — a consequence of sensitive TSH assays and more frequent testing.

Confirm before you act. A TSH measurement over 4.5 mU/L should be confirmed within 1–3 months, and then repeated every 6–12 months in asymptomatic patients. If confirmed, consider past radioiodine administration, previous thyroid surgery, thyroid enlargement, and a family history of thyroid dysfunction, explore subtle clinical signs of hypothyroidism, and check the lipid profile. Also exclude non-thyroidal illness recovery, drugs and assay interference.

Natural history. Progression to overt hypothyroidism varies from 3 to 20% but is generally around 5% per year; the risk is greater in patients with goitre or thyroid antibodies, or both. Antithyroid antibodies are useful for improving the prognosis of progression to overt hypothyroidism, but they may not change the management of the patient. Hence patients should be followed and eventually treated rather than treated immediately.

The treatment threshold. If, during follow-up, TSH increases to over 10 mU/L, the patient should be treated with levothyroxine. When TSH is between 4.5 and 10 mU/L, an ex adiuvantibus (therapeutic trial) administration of levothyroxine may be considered on an individual basis to help improve subtle clinical symptoms — with a clear plan to stop if no benefit accrues.

The paradox of advanced age — mild thyroid failure may be protective. In the oldest old (over 85 years), thyroid hormone substitution in subclinical hypothyroidism is not likely to be beneficial and may even be harmful. In the Leiden 85-plus study, none of the patients with subclinical hypothyroidism had progressed to overt disease when retested at age 88, there was no association between TSH and cognition, depression or ADL disability, and increased TSH at baseline was associated with a slower decline in instrumental ADL ability and with lower all-cause and cardiovascular mortality despite higher cholesterol. The explanation offered is that subclinical hypothyroidism seems detrimental in young to middle-aged people, whereas it may be harmless or possibly beneficial at an advanced age.

Individualise, and use a higher TSH target in the old. The decision to treat should be individualised — not based on TSH cut-offs alone, but also considering chronic comorbidities and frailty — and restricted to high-risk patients, to avoid the adverse effects of unnecessary replacement in vulnerable older people. A higher target serum TSH of 6 or 7 mU/L has been suggested in people over 70–80, particularly if they are at risk of cardiac arrhythmias or fragility fractures. Remember too that many older patients started levothyroxine when younger for inappropriate reasons or for transient hypothyroidism and continue it without review — in one nursing-home study thyroid hormone therapy was successfully withdrawn from half the residents studied.

The TRUST trial settled the symptomatic question, and subsequent practice has consolidated around not treating. TRUST was a double-blind, randomised, placebo-controlled trial of 737 adults aged 65 and over (mean age 74.4) with persistent subclinical hypothyroidism, randomised to levothyroxine 50 µg/day — or 25 µg/day if body weight was under 50 kg or coronary heart disease was present — titrated to TSH, versus placebo with simulated dose adjustment. Mean baseline TSH was 6.40 mU/L; at one year it fell to 3.63 mU/L on levothyroxine versus 5.48 on placebo, yet there was no difference in the Hypothyroid Symptoms score or the Tiredness score, and no benefit on executive cognitive function, blood pressure, weight, waist circumference, Barthel index or instrumental ADL scores. Current guidance and the consolidated position is therefore that treatment of minor TSH abnormalities meeting the criteria for subclinical hypothyroidism in the elderly is not recommended. Two honest caveats to state in the exam: the trial was underpowered for cardiovascular outcomes, and patients with higher TSH concentrations, high symptom burden, positive TPO antibodies, and the oldest patients over 80 were under-represented or not assessed, so no definitive conclusion can be drawn for those subgroups. TSH >4.5 mU/L with normal free T4 → confirm at 1–3 months, exclude drugs, non-thyroidal illness, assay error ↓ TSH >10 mU/L TSH 4.5–10 mU/L → Treat with levothyroxine → Repeat every 6–12 months (especially if TPO antibody positive) Consider a time-limited therapeutic trial if symptomatic — stop if no benefit ↓ Age over 85, or frail, or at risk of arrhythmia or fracture → generally do not treat; if treating, aim for a higher TSH target of 6–7 mU/L CURRENT PRACTICE UPDATE Aspect Textbook Current recommendation Symptomatic benefit of Trial of therapy may be considered at TSH TRUST: no symptomatic, cognitive or levothyroxine 4.5–10 functional benefit in those over 65 — do not treat routinely TSH target when treating Normalise TSH Accept a higher target of 6–7 mU/L over age an older patient 70–80, especially with arrhythmia or fracture risk Existing long-term Continued indefinitely Actively review for deprescribing — levothyroxine successfully withdrawn in half of nursing-home residents studied Exam approach: The single sentence that earns the mark is that subclinical hypothyroidism appears detrimental in the middle-aged but harmless or possibly protective in the oldest old. Say it explicitly, then give the TSH 10 threshold.