Subclinical hypothyroidism sits in an awkward space between a lab abnormality and a disease. In later life, that awkwardness matters. A mildly raised TSH can reflect early thyroid failure, temporary variation, medication effects, or an age-related shift in the reference range. The evidence does not support reflex treatment for everyone, but it does support careful follow-up and context.
The lab result is not the whole diagnosis
Subclinical hypothyroidism usually means thyroid-stimulating hormone, or TSH, is above the laboratory reference range while free thyroxine, or FT4, remains within range. That is different from overt hypothyroidism, where thyroid hormone itself is low. The distinction is not semantic; it changes how confidently symptoms can be attributed to the thyroid.
Fatigue, low mood, constipation, dry skin, weight change, and cold intolerance are real symptoms, but they are not specific thyroid symptoms. They also travel with poor sleep, depression, anaemia, chronic inflammation, medication side-effects, menopause, low energy intake, kidney disease, and ordinary ageing. A borderline TSH result can therefore feel explanatory before it has earned that role.
This is why repeat testing matters. NICE guidance on thyroid disease recommends considering thyroid peroxidase antibody testing once in adults with TSH above the reference range, and bases treatment decisions on repeated TSH results rather than a single stray number. The practical point is simple: one abnormal result is a signal to interpret, not a diagnosis to medicate automatically.
Age changes the thyroid conversation
Older adults are more likely to have mildly raised TSH, and the clinical meaning of that rise is not identical at every age. A TSH of 6 mIU/litre in a 35-year-old trying to conceive is not the same clinical problem as a TSH of 6 in a 78-year-old with normal FT4, no clear thyroid symptoms, and several other possible causes of tiredness.
That does not mean thyroid disease becomes harmless with age. Overt hypothyroidism still needs proper assessment and treatment. A higher TSH, especially if persistent and above 10 mIU/litre, carries more clinical weight than a small, fluctuating elevation. Positive thyroid antibodies, previous thyroid surgery, radioactive iodine treatment, neck irradiation, goitre, or a strong family pattern also make underlying thyroid disease more plausible.
The mistake is to treat every older thyroid panel as a longevity score. The thyroid is not a simple metabolic dial, and levothyroxine is not a wellness tonic. In later life, the question is less “is this number ideal?” and more “does this pattern predict benefit from treatment, risk without treatment, or a need for monitoring?”
What randomised trials found in older adults
The largest trial changed the tone of this debate. The TRUST randomised trial in the New England Journal of Medicine studied 737 adults aged 65 or older with persistent subclinical hypothyroidism. Levothyroxine lowered TSH, as expected, but it did not improve hypothyroid symptom scores or tiredness scores at one year compared with placebo.
That result matters because it separates biochemical correction from lived benefit. A medicine can move the lab value in the intended direction and still fail to make people feel better on average. For a reader with severe fatigue, that may be frustrating. For clinical decision-making, it is essential.
A 2018 systematic review and meta-analysis in JAMA reached a similar conclusion across randomised trials in non-pregnant adults with subclinical hypothyroidism: thyroid hormone therapy was not associated with meaningful improvement in general quality of life or thyroid-related symptoms. The review also found no clear benefit for fatigue, depressive symptoms, cognitive function, muscle strength, blood pressure, or body mass index.
Why treatment thresholds are cautious
Guidelines differ at the edges, but many share a cautious centre. NICE says clinicians should consider levothyroxine for adults with subclinical hypothyroidism when TSH is 10 mIU/litre or higher on two occasions three months apart. For adults under 65 with symptoms and repeated TSH above range but below 10, NICE allows a six-month treatment trial, with stopping considered if symptoms persist after TSH normalises.
That age distinction is important for this article. The evidence for symptom improvement is weaker in older adults with mild elevations, and the potential downside of overtreatment is more consequential. Too much thyroid hormone can push the body toward a hyperthyroid state, raising concern about palpitations, atrial fibrillation, angina in susceptible people, and bone loss in those already at fracture risk.
None of this argues for ignoring a persistently abnormal test. It argues against using levothyroxine as a low-risk experiment whenever someone feels tired. The people most likely to need treatment are those with overt hypothyroidism, TSH persistently above 10, convincing thyroid disease features, or a clinical situation where an endocrinologist judges that the balance has shifted.
Cardiovascular and bone outcomes are still uncertain
Subclinical hypothyroidism has been linked in observational studies with cardiovascular risk, particularly at higher TSH levels. Observational evidence can identify risk patterns, but it cannot prove that treating every mild elevation prevents heart attacks, strokes, or heart failure. People with higher TSH may differ in many ways beyond thyroid function.
Randomised data in older adults have not shown a clear cardiovascular win. A pooled analysis of TRUST and the IEMO80+ trial found that levothyroxine did not significantly change cardiovascular outcomes in older adults with subclinical hypothyroidism, including those with existing cardiovascular disease. A later systematic review of cardiovascular and bone outcomes similarly found no strong evidence that treatment in older people reduced cardiovascular events or fractures.
The safety issue runs in the other direction as well. If treatment overshoots, an older person may become biochemically over-replaced. That is not a minor bookkeeping error in someone with atrial fibrillation risk, osteoporosis, frailty, or coronary disease. A cautious target and regular monitoring are part of the intervention, not optional paperwork.
Deprescribing is becoming part of the evidence
A new question is now being asked more openly: do all older adults already taking levothyroxine still need it? A 2026 prospective study in JAMA followed adults aged 60 or older through stepwise levothyroxine dose reduction. About one quarter discontinued treatment while maintaining TSH below 10 mIU/litre and normal FT4 at one year.
That is not a licence to stop medication at home. The study used a structured process with monitoring, and most participants did not successfully discontinue. It does, however, challenge the assumption that a thyroid prescription started years ago must always be lifelong, especially when the original diagnosis is unclear or the current dose is low.
For clinicians, deprescribing is not the opposite of care. It is a form of care when the medicine may no longer be needed, when the diagnosis was uncertain, or when the harms of overtreatment are rising. For readers, the safe message is narrower: if the reason for a long-standing prescription is unclear, ask for a review rather than making changes alone.
What this means in practice
- Do not interpret a single mildly raised TSH as proof that the thyroid is causing tiredness, weight change, low mood, or brain fog.
- Ask whether TSH and FT4 have been repeated, whether thyroid peroxidase antibodies were checked once, and whether the result has persisted over time.
- Treat TSH persistently at or above 10 mIU/litre as a different conversation from a mild elevation below 10, especially in later life.
- Do not start, stop, or change levothyroxine without clinical advice; dose changes need follow-up blood tests and symptom review.
- Flag chest pain, palpitations, unexplained weight loss, faintness, severe depression, pregnancy, or trying to conceive as reasons for prompt medical guidance rather than watchful waiting.
What we don’t know
We still do not have a perfect way to tell whose mild subclinical hypothyroidism will progress, whose symptoms are thyroid-driven, and who will feel meaningfully better with treatment. TSH is useful, but it is a pituitary signal, not a direct measurement of energy, cognition, mood, or tissue-level thyroid action.
Trials also have limits. Older adults with very high TSH, severe symptoms, pregnancy, fertility concerns, or complex endocrine disease are not the same as the average participant in a trial of mild subclinical hypothyroidism. Evidence against routine treatment is not evidence against all treatment.
The best reading is therefore neither dismissive nor aggressive. Mild subclinical hypothyroidism in later life deserves context, repeat testing, and careful monitoring. It usually does not deserve panic, self-treatment, or a promise that one hormone tablet will restore youth.
Photo: Governo do Estado de São Paulo via Wikimedia Commons, CC BY 2.0.
.jpg){kind=link}