High TSH with a normal free T4: what subclinical hypothyroidism does and does not mean

TSH is the pituitary's instruction to the thyroid. When thyroid hormone output starts to fall, the pituitary raises TSH to push harder, and it does this sensitively enough that TSH rises before free T4 leaves its range. A raised TSH with a normal free T4 is therefore a picture of a thyroid working harder to produce a normal output.

The AACE and ATA guideline is direct that TSH is the right first test: a serum thyrotropin is the single best screening test for primary thyroid dysfunction for the vast majority of outpatient clinical situations. What follows from an abnormal one is where the disagreement lives.

The word subclinical is doing something specific and often misread. It does not mean mild disease, and it does not mean early disease that will inevitably progress. It means the abnormality is visible in the laboratory and not in the clinical picture. Some people with the pattern do have symptoms, but the symptoms are not what defines it, and that distinction turns out to matter a great deal.

The obvious assumption is that if TSH is high because the thyroid is struggling, replacing thyroid hormone should make people feel better. TRUST, published in the New England Journal of Medicine in 2017, tested it properly.

The design was a double-blind, randomized, placebo-controlled trial in 737 adults aged 65 or over with persisting subclinical hypothyroidism, meaning a TSH between 4.60 and 19.99 mIU/L with a free thyroxine inside the reference range. Mean age was 74.4 years and 53.7 percent were women. Half received levothyroxine starting at 50 micrograms daily, titrated against TSH; half received placebo with mock dose adjustment. The two primary outcomes were change in the Hypothyroid Symptoms score and the Tiredness score at one year, each running 0 to 100, with a minimum clinically important difference of 9 points.

The treatment worked biochemically. Mean TSH was 6.40 at baseline and fell to 3.63 in the levothyroxine group against 5.48 on placebo, at a median dose of 50 micrograms. The symptoms did not follow.

Both confidence intervals sit far inside the 9-point threshold the investigators set in advance as the smallest difference that would matter to a patient. This is not an underpowered trial that failed to detect a real effect; it is a well-sized trial that excluded an effect of a size worth having. The authors reported no beneficial effects on secondary outcomes and no significant excess of the serious adverse events they had flagged as of special interest.

The AACE and ATA guideline puts it in one sentence: the decision to treat subclinical hypothyroidism when the serum thyrotropin is less than 10 mIU/L should be tailored to the individual patient. Above 10, treatment is more widely accepted, on the grounds that progression to overt hypothyroidism is more likely and the association with adverse outcomes is stronger.

Age belongs in that judgment. The upper limit of the TSH distribution rises across adult life, so a value that is abnormal for a 30-year-old can sit inside the normal spread for someone in their eighties. We cover that shift, with the age-specific figures, in our guide to the TSH normal range.

Given how common the pattern is, it is reasonable to ask why thyroid testing is not part of a routine check. The US Preventive Services Task Force examined it and concluded that the current evidence is insufficient to assess the balance of benefits and harms of screening for thyroid dysfunction in non-pregnant, asymptomatic adults. That is an I statement, issued in 2015, and a literature scan in May 2024 found a lack of new evidence to support updating the review.

An I statement is not a recommendation against testing. It is an admission that the evidence does not exist to say either way, and it exists partly because of results like TRUST: finding the abnormality is easy, and demonstrating that acting on it helps has proved hard.

Repeat it, with free T4. TSH is not stable. It varies through the day, rises transiently during recovery from illness, and is affected by biotin supplements, which interfere with many immunoassays and can distort thyroid results in either direction. A single value should not be acted on.

Ask about thyroid peroxidase antibodies. Positive antibodies raise the probability of progression to overt hypothyroidism and shift the balance towards treating a borderline value.

Separate the symptoms from the number. Tiredness, weight change and low mood are common and have many causes. TRUST is the reason to be cautious about attributing them to a mildly raised TSH before other explanations have been considered.

Flag pregnancy plans immediately. Thresholds and management in pregnancy and preconception are different and stricter, and this is one situation where a borderline result should not be watched.

If treatment starts, agree what success looks like. The AACE and ATA guideline, and the 2014 ATA treatment guideline, both keep levothyroxine as the standard of care and find no consistently strong evidence that combination or thyroid extract preparations improve health outcomes over levothyroxine alone. A trial of treatment with a defined review point is reasonable; open-ended treatment on a number alone is what the evidence does not support.

Our TSH reference page carries the interval and the subclinical zone with their source and effective date, and the lab result lookup will place a specific value against them.

Book a review for any raised TSH so it can be repeated with free T4, and sooner if the free T4 is low, if the TSH is above 10, if you are pregnant or planning to be, or if you have symptoms that are getting worse rather than staying steady. If treatment is being offered for a mildly raised TSH, two questions are fair: what is my free T4 and antibody status, and what change would tell us in three months whether this is working.