TSH Blood Test: What It Measures, How to Read Your Results, and What It Means for Athletes
Table of Contents
Key Takeaways: TSH Blood Test
- A TSH blood test is usually the first thyroid test because small changes in free thyroid hormone produce large changes in TSH.
- TSH is most informative when read together with free T4; the combination separates overt, subclinical and central patterns.
- Abnormal results often do not persist: in a five-year follow-up of 422,242 outpatients, 62.1% of mildly elevated and 51.2% of decreased initial TSH values (among patients not taking thyroid-specific medication) were normal when the test was repeated.[6]
- Acute illness and recovery, laboratory interference, marked obesity and pituitary disease can all make TSH misleading.
- In my clinical experience, thyroid tests are worth considering early in the work-up of exhaustion, palpitations, sleep problems and anxiety, because hyperthyroidism can be mistaken for anxiety and somatic causes should be excluded before far-reaching conclusions about psychological factors.
- In athletes, test in a rested, standardised state, treat a mildly raised TSH as a reason for proper assessment rather than an automatic explanation for poor performance, and give a suppressed TSH the full standard work-up.
- In athletes and others with prolonged inadequate energy intake, a low TSH together with a low free T4 can point to an energy deficit, as it can in my patients, but the same pattern also raises the question of central hypothyroidism, and it should not be mistaken for subclinical hypothyroidism.
Introduction: Why the TSH Blood Test Comes First
When thyroid disease is on the list of possible explanations for someone’s symptoms, the TSH blood test is usually the first test requested. A review of TSH measurement describes it as the single most useful test of thyroid function and one that is widely and appropriately used for screening.[1] That position rests on physiology rather than convenience, and the same physiology explains the test’s blind spots.
Thyroid dysfunction is common enough that the question comes up often. The US NHANES III survey measured TSH, total T4 and thyroid antibodies in 17,353 people aged 12 and older. It found hypothyroidism in 4.6% of the population (0.3% clinical and 4.3% subclinical) and hyperthyroidism in 1.3% (0.5% clinical and 0.7% subclinical).[2] In that survey, the subclinical forms were more common than the clinical forms for both hypothyroidism (4.3% versus 0.3%) and hyperthyroidism (0.7% versus 0.5%), which is the range where reading a TSH blood test carefully matters most.[2]
In my own work, especially in occupational health, I order thyroid tests often. In practice, they are usually part of the basic laboratory panels. At the clinic, hypothyroidism is far more common than hyperthyroidism: I see genuine hypothyroidism fairly often, and thyroxine is a common medication on my patients’ medication lists.
This article explains what the test measures, when it is ordered, how to read it together with free T4, when the result can mislead, and why a single abnormal value is rarely the final answer. The last part looks at what changes in athletes. Reference intervals, including the effects of age, time of day, biotin and pregnancy, are covered separately in my article on the TSH normal range, and lower-normal values in trained people in TSH in athletes.
What TSH Measures and Why It Is So Sensitive
Thyroid-stimulating hormone (TSH, also called thyrotropin) is released by the pituitary gland and stimulates the thyroid to secrete thyroxine (T4) and triiodothyronine (T3). Thyroid hormone in turn acts back on the pituitary to restrain TSH release. The loop works much like a thermostat: when thyroid hormone falls, TSH rises to push production up, and when thyroid hormone is in excess, TSH falls.
What makes this loop diagnostically useful is its shape. The relationship between TSH and free thyroid hormone is negative and log-linear, so modest changes in free hormone concentrations produce large changes in TSH, and even subtle shifts in thyroid hormone production are best detected by measuring TSH.[1] In subclinical disease, TSH is already outside its reference range while free T4 (and T3, in hyperthyroidism) remains within it.[4][3] The 2016 American Thyroid Association (ATA) guidelines similarly state that serum TSH has the highest sensitivity and specificity of any single blood test when thyrotoxicosis is suspected, and recommend it as the initial screening test.[3]
That sensitivity comes with limits. The same review criticises the tendency to rely on TSH alone when assessing complicated thyroid disease or patients being treated for thyroid dysfunction.[1] Because TSH depends on a working pituitary, disease of the pituitary or hypothalamus can leave TSH inappropriately non-elevated even when free T4 is low, and a 2024 Lancet seminar on hypothyroidism advises measuring free T4 together with TSH so that this central form is not missed.[4]
When Is a TSH Blood Test Ordered?
Most TSH blood tests are requested to investigate symptoms or to follow known thyroid disease. Testing people without any symptoms is a separate question. In its 2015 recommendation, the US Preventive Services Task Force identified serum TSH as the primary screening test for thyroid dysfunction, but concluded that the evidence was insufficient to judge the balance of benefits and harms of screening nonpregnant adults without symptoms.[5]
Most results come back normal. In a five-year follow-up of 422,242 outpatients tested by primary care physicians in a large health management organisation, 95% of initial TSH values were within the normal limits used (0.35–5.5 mIU/L); 1.2% were decreased (below 0.35 mIU/L), 3.0% were elevated (above 5.5 and up to 10 mIU/L) and 0.7% were highly elevated (above 10 mIU/L).[6] When the first TSH was normal, the likelihood of an abnormal value within five years was low, at 2%.[6]
In my clinical experience, thyroid tests belong in the differential diagnosis of many different situations, including stress, palpitations, exhaustion, fatigue, weight loss, weight gain and sleep difficulties. In my view, thyroid disease is worth considering early, before these complaints are worked up as something else, and burnout and insomnia are two examples.
In my clinical experience, hyperthyroidism can be a tricky diagnosis, because it is often confused in the clinic with an anxious patient whose sympathetic tone is raised. For that reason, thyroid tests are particularly important in the differential diagnosis of anxious and exhausted patients, patients with overtraining, and any patient with a disturbance of the autonomic nervous system or stress system. In my view, thyroid tests are also worth checking early in patients with palpitations, whatever the cause, and an anxious patient is worth testing as well. The symptoms of patients with hyperthyroidism may also be downplayed, and the threshold for getting into care can unfortunately sometimes be high. That is why, in my view, it is important to take the patient’s symptoms seriously and to test and exclude somatic causes before making far-reaching decisions about anxiety or psychological factors.
How to Read TSH Blood Test Results With Free T4
Each laboratory reports its own reference interval, and how those limits are set is explained in my article on the TSH normal range. The more useful step is reading TSH together with free T4 (and, in selected cases, T3), because the combination points to the type of problem:
| TSH | Free T4 / T3 | Usual interpretation |
|---|---|---|
| Above range | Free T4 below range | Overt primary hypothyroidism[4] |
| Above range | Free T4 within range | Subclinical hypothyroidism[4] |
| Low or undetectable | Free T4 and T3 within range | Subclinical hyperthyroidism[3] |
| Subnormal, usually undetectable | Free T4 and/or T3 elevated | Overt hyperthyroidism[3] |
| Inappropriately non-elevated | Free T4 low | Central hypothyroidism (pituitary or hypothalamic)[4] |
| Normal or elevated | Thyroid hormones elevated | Consider inappropriate TSH secretion (such as a TSH-producing pituitary adenoma), resistance to thyroid hormone, or assay interference[3] |
Two points help put the table in context. First, the subclinical hypothyroid pattern is common: a JAMA review estimates that it affects up to 10% of adults and is most often caused by autoimmune (Hashimoto) thyroiditis.[7] The 2024 Lancet seminar likewise identifies chronic autoimmune thyroiditis as the most common cause of hypothyroidism in iodine-sufficient countries.[4] Second, “subclinical” describes the laboratory pattern rather than how a person feels. The same JAMA review notes that middle-aged patients with subclinical hypothyroidism may have cognitive impairment, nonspecific symptoms such as fatigue, and altered mood, although these associations are less certain with mild TSH elevations.[7]
In my experience, TSH works well as a screening test, but each result has to be read case by case together with the other laboratory results. If the value is abnormal, at least free T4 needs to be measured, and sometimes free T3 as well; the ATA guidelines note that total T3 is frequently preferred over free T3 in clinical practice, because free T3 assays are less widely validated.[3]
In my clinical experience, subclinical hypothyroidism became something of a fashionable diagnosis for exhaustion at one point. The typical situation was a TSH at the upper end of the reference range with free T4 also within its reference range. By the definition above, subclinical hypothyroidism requires a TSH above the reference range.[4]
When TSH Can Mislead
A TSH blood test is a reliable first test for most outpatients, but several situations can uncouple it from true thyroid status.
Pituitary or hypothalamic disease. In central hypothyroidism, TSH remains inappropriately non-elevated while free T4 is low, and levothyroxine treatment in these patients is monitored with free T4 rather than TSH.[4]
Acute illness and recovery. During severe illness, TSH can be suppressed, with cytokines implicated. During recovery it can rebound, although rarely above 10 mU/L, and the JAMA review advises deferring a diagnosis of hypothyroidism until the person has fully recovered.[7]
Laboratory interference. Depending on the assay, heterophilic antibodies or circulating macro-TSH can produce artefactually elevated TSH values.[7] The ATA guidelines also list interfering antibodies among the reasons a normal TSH may not reliably exclude thyrotoxicosis.[3]
Marked obesity. Increases in TSH associated with marked obesity, typically with a body mass index above 40, reverse with weight loss.[7]
Everyday pre-analytical factors. Time of day, pregnancy and age can influence TSH or its interpretation, and biotin supplements can interfere with some laboratory assays without any change in thyroid function; these are covered in my TSH normal range article.
Athletes in some sports may run an energy deficit because of the nature of the sport. In my experience this is seen a lot in gymnasts and skaters, and also in weight-class athletes. I remember it well from my own career, when I had to cut weight to drop into a lower weight class, and I could feel my own metabolism slowing down gradually. These situations are worth taking into account when interpreting an athlete’s thyroid tests.
From a clinical perspective, athletes, people with anorexia regardless of sport, and others with prolonged inadequate energy intake can show an unusual pattern on a thyroid panel: both TSH and free T4 may be low. In my clinical experience, at least in my own patients, this pattern may point to an energy deficit, although it is not specific to one. It differs from what many clinicians are used to, because the usual expectation is that either both values are within range or one is high and the other low, which is the logical pattern, so a low TSH together with a low free T4 can confuse the interpretation. It should not be mistaken for subclinical hypothyroidism, or at least one has to be careful and precise, which is why the patient’s history is important. A low free T4 with a low or inappropriately normal TSH is also the pattern of central hypothyroidism, so other causes still need to be considered.[4] This is a point that may differ from the typical sedentary population. The link between low energy availability and TSH is discussed in low TSH in athletes.
Why One Abnormal Result Is Rarely the Final Answer
An abnormal TSH blood test result often does not persist. In the 422,242-patient follow-up, a normal value was found on repeat testing in 62.1% of patients with an initially elevated TSH (above 5.5 and up to 10 mIU/L), 51.2% of those with a decreased TSH and 27.2% of those with a highly elevated TSH (above 10 mIU/L), among the 346,549 patients not taking thyroid-specific medication.[6] For an elevated TSH suggesting subclinical hypothyroidism, the JAMA review therefore recommends confirming the result with a second TSH and free T4 1 to 3 months after the first; if the initial TSH is above 15 mU/L, testing is repeated within 1 to 2 weeks. These intervals should not be applied automatically to other thyroid test patterns.[7]
Thyroid peroxidase (TPO) antibodies help estimate where a mildly raised TSH is heading. People with circulating TPO antibodies have a greater risk of progressing from subclinical to overt hypothyroidism, and the annual rate of progression in mild (grade 1) subclinical hypothyroidism is about 2% to 4% depending on antibody status.[7] The 20-year follow-up of the Whickham Survey pointed in the same direction. Odds ratios for developing hypothyroidism were 8 in women and 44 in men with a raised TSH alone, 8 and 25 with positive antithyroid antibodies alone, and 38 and 173 when both were present, with wide confidence intervals around each estimate.[8] Increasing TSH values above 2 mU/L at the first survey raised the probability of later hypothyroidism, and antibodies raised it further.[8]
None of this means every raised TSH needs treatment. The JAMA review concludes that most people with subclinical hypothyroidism can be observed without treatment, and that treatment might be indicated when TSH is 10 mU/L or higher, or in young and middle-aged people with symptoms consistent with mild hypothyroidism.[7] Autoimmune thyroid disease in women who train is discussed further in autoimmune thyroid disease in female athletes.
When the picture points to hypothyroidism, I often check TPO antibodies, which help estimate how likely a mildly raised TSH is to progress.[7] Hypothyroidism can often be managed at the primary care level, although some situations may warrant an endocrinologist’s assessment. When the picture points to hyperthyroidism, I fairly routinely check TSH-receptor antibodies as well.
TSH Blood Test in Athletes: What Changes
The biology of the test does not change because someone trains hard, and athlete status is not a reason to explain away a clearly abnormal result. What changes is the context. Training load affects when a sample should be taken, performance complaints raise the question of whether a mildly raised TSH matters, and a suppressed TSH raises questions that come up less often outside sport. Lower-normal values linked to heavy training and energy availability are covered in TSH in athletes and low TSH in athletes; this section deals with the other three issues.
In my clinical experience, TSH is part of the basic periodic health examination in athletes as well. Being an athlete is not in itself an indication to measure thyroid tests; in this setting TSH is a screening test. A specific reason to suspect thyroid disease, however, can justify targeted TSH measurement and further work-up.
Timing the Test Around Training
Two small studies with different designs reported different TSH responses to exercise. When 60 well-trained male athletes cycled at 45%, 70% and 90% of maximum heart rate, the most prominent hormonal changes appeared at the anaerobic threshold (70% of maximum heart rate), and T4, free T4 and TSH kept rising at 90% while T3 and free T3 began to fall.[9] In contrast, when eight elite male weightlifters completed one week of very intense strength training, serum TSH, T4 and T3 decreased, while free T4, reverse T3 and thyroxine-binding globulin did not change significantly.[10]
These findings do not show which aspects of training explain the different responses, and they do not define a precise interval between training and testing. In practical terms, they cannot prove an optimal sampling time, but they make a simple practical suggestion reasonable (my synthesis rather than an evidence-based rule): schedule the TSH blood test in a rested, standardised state, not straight after a hard session or at the end of an overload block, and repeat a borderline result under the same conditions. My guide to preparing for a blood test as an athlete covers the wider pre-test routine.
A Raised TSH and Exercise Capacity
The two controlled studies discussed here examined patients with a mildly raised TSH rather than athletes. In one study, 23 patients with subclinical hypothyroidism were compared with 10 matched euthyroid controls. Maximal power output (P = 0.02) and VO₂max (P = 0.04) were lower in the patients, who also reached higher heart rates at the same oxygen uptake.[11] In the double-blind, placebo-controlled treatment phase, six months of levothyroxine improved neuromuscular symptoms but did not significantly change the energy and substrate response to exercise.[11] Another trial randomised 23 patients to levothyroxine (n = 11) or no treatment (n = 12); submaximal cardiopulmonary exercise performance improved after six months of normalised TSH, with no changes in the untreated group.[12]
Neither study involved trained athletes, both were small, and their results only partly agree. The practical message is therefore modest: in an athlete with unexplained performance decline, a mildly raised TSH deserves confirmation and a proper thyroid assessment, but it should not automatically be accepted as the explanation. The wider differential diagnosis is covered in why is my performance declining? and fatigue in athletes.
A Suppressed TSH: Heart Rhythm and Thyroid Hormone Intake
A low TSH can matter even when free T4 is normal. An individual-participant analysis pooled 10 cohorts with 52,674 participants and defined endogenous subclinical hyperthyroidism as a TSH below 0.45 mIU/L with normal free T4, after excluding people taking thyroid-altering medication; 2,188 participants (4.2%) met that definition.[13] It was associated with increased total and coronary heart disease mortality and with incident atrial fibrillation, and the risks of coronary heart disease death and atrial fibrillation were highest when TSH was below 0.10 mIU/L.[13] These cohorts were not studies of athletes, so extrapolation to athletes is uncertain; palpitations or an irregular pulse in an athlete with a low TSH still deserve assessment.
A suppressed TSH with raised thyroid hormones needs the same work-up in an athlete as in anyone else, including for conditions such as Graves’ disease. One additional question is whether thyroid hormone is being taken. The ATA guidelines note that factitious ingestion of thyroid hormone can be distinguished from other causes of thyrotoxicosis by a low serum thyroglobulin, which is interpreted together with other findings because assay interference is possible.[3]
Data on how often thyroid hormone is misused in elite sport are limited. Thyroid hormones are not prohibited under the World Anti-Doping Code, and in 498 serum samples from Australian anti-doping tests, two athletes had biochemical thyrotoxicosis, a prevalence of 4 per 1,000 (upper 95% confidence limit 16).[14] The authors found minimal evidence of thyroid hormone abuse among these athletes, and their estimates were lower than estimated T4 prescription rates in the age-matched Australian population.[14]
From time to time I also see hyperthyroidism, whether Graves’ disease or another cause. These cases usually need at least an endocrinology consultation, so I readily send a consultation request, and in Finland the guidance I receive on their treatment is usually good. Care then often follows its own treatment pathway, but each case still needs an individual assessment.
In my experience, thyroiditis is seen a little less often still, and it may occasionally go unrecognised in the clinic. These patients can, at least in my experience, have neck pain, perhaps raised inflammatory markers, and a hyperthyroid picture. Many of them recover from this state, although thyroid function is worth following until it has normalised, and closer follow-up is needed if the course is prolonged or something else happens. An endocrinology assessment is often not necessary.
Conclusion: Reading a TSH Blood Test in Context
A TSH blood test earns its place as the first thyroid test because the pituitary amplifies small changes in thyroid hormone into large changes in TSH. That same sensitivity is why a single value needs context: free T4 to identify the pattern, a repeat test before a diagnosis, antibodies when TSH is raised, and awareness of illness, laboratory interference and pituitary disease.
For athletes, the additions are mainly practical. Test rested, take a mildly raised TSH seriously without assuming it explains every poor performance, and give a suppressed TSH the same careful work-up as in anyone else.
In my own practice, hypothyroidism is far more common than hyperthyroidism and can often be managed in primary care, whereas hyperthyroidism usually leads me to at least an endocrinology consultation. In my view, thyroid tests are worth checking early in patients with exhaustion, palpitations, sleep problems or anxiety, and taking the symptoms seriously means excluding somatic causes before drawing far-reaching conclusions about psychological factors. In athletes who are not getting enough calories, a low TSH together with a low free T4 may point to an energy deficit, although central hypothyroidism has to be kept in mind, and the patient’s history is important to avoid mistaking it for subclinical hypothyroidism.
References
- https://doi.org/10.1016/S0889-8529(05)70186-9
- https://doi.org/10.1210/jcem.87.2.8182
- https://doi.org/10.1089/thy.2016.0229
- https://doi.org/10.1016/S0140-6736(24)01614-3
- https://doi.org/10.7326/M15-0483
- https://doi.org/10.1001/archinte.167.14.1533
- https://doi.org/10.1001/jama.2019.9052
- https://doi.org/10.1111/j.1365-2265.1995.tb01894.x
- https://pubmed.ncbi.nlm.nih.gov/16380698/
- https://pubmed.ncbi.nlm.nih.gov/1753718/
- https://doi.org/10.1210/jc.2004-2344
- https://doi.org/10.1007/BF03346488
- https://doi.org/10.1001/archinternmed.2012.402
- https://doi.org/10.1210/jendso/bvad027
