Why does this number matter?
A TSH of 3.8 is normal. It falls within the reference range printed on every lab report in the country. No flag. No follow-up. No conversation.
Here is what that reference range doesn't tell you: it was built from a population that included people with undiagnosed autoimmune thyroid disease. When researchers excluded those people, 95% of genuinely healthy adults had a TSH below 2.5 [1]. That 3.8 isn't a clean bill of health. It might mean your body is already working harder than it should to keep your thyroid on track. TSH comes from your pituitary gland, and when it rises, even slightly, it means the pituitary is raising its voice, asking the thyroid to do more.
TSH is the most sensitive thyroid test available. It moves first, before thyroid hormones themselves drop out of range. But it's only one number in a story of three: TSH tells you what the brain is asking for, Free T4 tells you what the thyroid is producing, and Free T3 tells you how much active hormone your cells are actually getting. Testing TSH alone is like reading the subject line of an email without opening the message.
What is actually happening?
Picture a great clock tower that keeps time for a whole city. Inside, falling weights drive the gears, and the gears swing the bells. When the bells ring, the city runs by them. Those bells are your thyroid hormones, setting the pace of nearly everything: heart rate, temperature, energy, metabolism.
A control room watches the clock and listens for the bells. When they fall behind, it sends a winding crew to haul the weights back up, and the harder the clock struggles, the more crew it sends. TSH is that crew. When your thyroid lags, your pituitary sends more TSH. So the number tells you how hard the system is straining, not why.
Here's the catch the control room can miss. The gears don't ring the bells directly; a small mechanism between them turns the motion into an actual strike. If that mechanism jams, the gears spin all day and no bell ever sounds. The control room sees movement and relaxes. The city hears silence. That is exactly what happens when TSH looks normal but your body isn't converting T4 into the active T3 your cells need.
TSH (thyroid-stimulating hormone, also called thyrotropin) is produced by specialized cells in the anterior pituitary gland, a pea-sized structure at the base of the brain. It's released into the bloodstream and travels to the thyroid gland, where it binds to receptors on thyroid follicular cells and stimulates them to produce thyroid hormones, primarily T4 with smaller amounts of T3 [2].
The entire system runs on a feedback loop. When thyroid hormone levels in the blood are adequate, the pituitary senses this and reduces TSH output. When thyroid hormone levels drop, the pituitary increases TSH, effectively turning up the volume on its request. This feedback relationship is highly sensitive and follows a log-linear pattern: a 50% drop in Free T4 can produce a 90-fold increase in TSH. That amplification is what makes TSH the most sensitive detector of thyroid dysfunction and the first number to move.
The practical result is a set of diagnostic patterns. High TSH with low Free T4 is primary hypothyroidism: the thyroid is failing. High TSH with normal Free T4 is subclinical hypothyroidism: the thyroid is struggling but compensating, for now. Low TSH with high Free T4 is hyperthyroidism: the thyroid is overproducing and the pituitary has backed off. Low TSH with low Free T4 is central hypothyroidism, where the pituitary itself isn't sending enough signal.
TSH is a 28-kDa glycoprotein heterodimer composed of an alpha subunit (shared with LH, FSH, and hCG) and a unique beta subunit that confers thyroid specificity. Its release is stimulated by hypothalamic TRH (thyrotropin-releasing hormone) and inhibited by thyroid hormones through both direct pituitary feedback and indirect hypothalamic modulation. Dopamine, somatostatin, and glucocorticoids also inhibit TSH secretion, explaining why critically ill patients or those on high-dose corticosteroids often have suppressed TSH [3].
The reference range debate is grounded in NHANES III data. Hollowell et al. analyzed 16,533 individuals and found a population TSH range of 0.45–4.12 mIU/L (2.5th to 97.5th percentile). When they excluded individuals with positive thyroid antibodies or detectable thyroid disease, the distribution was markedly right-skewed and 95% fell below 2.5 mIU/L [1]. The Colorado Thyroid Disease Prevalence Study found abnormal TSH in 9.5% of 25,862 participants, the majority of whom were previously undiagnosed [4].
Subclinical hypothyroidism, defined as TSH above the reference range with normal Free T4, affects 4–10% of the general population and up to 20% of women over 60. Biondi and Cooper's review established that subclinical hypothyroidism is associated with LDL elevation, diastolic dysfunction, increased cardiovascular risk (particularly when TSH is above 10), and progression to overt hypothyroidism at roughly 2–5% per year [5]. Surks et al.'s joint scientific review proposed a nuanced approach: the 4.0–10.0 range is controversial, and individual factors including symptoms, antibody status, age, and pregnancy plans inform the decision [6].
Reference & Optimal Zones
mIU/L
Standard lab reference ranges are wider than the longevity-optimal zone, and on this marker both ends of the scale carry risk. Context matters: family history, other biomarkers, and inflammatory markers all modify interpretation.
How TSH connects to everything else
TSH does not exist in isolation. It is a downstream signal of several converging metabolic processes, which is why treating it effectively means understanding its inputs.
When this number moves
TSH has one of the most pronounced circadian rhythms of any hormone. It nadirs in the late afternoon, begins rising in the evening, and peaks between midnight and 4 AM. The nocturnal surge can be 50–100% above the afternoon trough. Morning draws (8–10 AM) provide the most reproducible values for comparison over time.
Biotin (vitamin B7, commonly found in hair, skin, and nail supplements) interferes with the immunoassay platform used for TSH measurement, causing falsely low TSH that can mimic hyperthyroidism when nothing is actually wrong. This is the most common cause of spurious thyroid test results.
One night of sleep deprivation can reduce the TSH peak by 30–50%. Chronic sleep disruption may contribute to HPT axis blunting over time, independently of any thyroid pathology.
Prolonged fasting (beyond 24 hours), acute illness, and physiological stress can all transiently suppress TSH via cytokines and cortisol. A TSH drawn during acute illness may be misleadingly low. Confirm any borderline result with a repeat 6–8 weeks later.
Human chorionic gonadotropin (hCG) has weak TSH-like activity, causing physiological TSH suppression in the first trimester. Trimester-specific ranges apply, with a general target below 2.5 mIU/L in the first trimester.
TSH increases modestly with age in iodine-sufficient populations, appearing to reflect a physiological shift rather than disease. Age-adjusted reference ranges should be used for patients over 70, and a slightly elevated TSH in an older adult may not warrant action.
TSH tends to be slightly higher in winter months, possibly reflecting increased thyroid hormone demand for thermogenesis.
What you can actually change
Listed by strength of evidence, not by how loudly they're sold.
Your TSH is a conversation between your brain and your thyroid, a constant, quiet negotiation about how fast your body should run. When it tips out of balance, the effects are pervasive and often misattributed: fatigue blamed on stress, weight gain blamed on aging, brain fog blamed on too many screens. The most important takeaway may be this: "normal" on a lab report means your number falls within a range that includes people with early, undetected disease. It doesn't mean optimal. It doesn't mean you feel well. And it doesn't mean there's nothing to investigate. TSH is the starting point, not the whole story. Follow it with Free T4 and Free T3, and you'll have a picture your pituitary alone could never show you.
TSH is available as a standalone, direct-access test. No doctor's order required. Prices verified March 2026. NY, NJ, and RI residents face restrictions at most services.
No fasting is required. But timing matters: draw in the morning (8–10 AM) for the most consistent result, and stop biotin supplements 48–72 hours beforehand to avoid a false low.
Yes. TSH can look normal when there is a conversion problem between T4 and T3 (because the pituitary measures T4 satisfaction, not T3 delivery to peripheral tissues), or during non-thyroidal illness. This is why TSH alone is not a complete thyroid assessment. You need Free T3 to know what your cells are actually receiving.
Confirm the result with a repeat measurement 6–8 weeks later before drawing conclusions. A single borderline result can reflect a transient cause (illness, stress, sleep disruption, biotin interference). If a second draw confirms elevation, add Free T4 and TPO antibodies to understand the cause.
Once you have a stable baseline, once or twice per year is reasonable for monitoring. After any meaningful dietary or lifestyle change (iodine, selenium, stress management, weight loss), retest in 6–8 weeks to see whether TSH has shifted.
Yes. TSH rises modestly with age as a normal physiological trend. In adults over 70, a TSH between 4.5 and 7.0 mIU/L may be physiological rather than pathological. What warrants attention in a 35-year-old may be normal variation in a 75-year-old.
- 1.Hollowell JG, Staehling NW, Flanders WD, Hannon WH, Gunter EW, Spencer CA, Braverman LE. Serum TSH, T(4), and thyroid antibodies in the United States population (1988 to 1994): National Health and Nutrition Examination Survey (NHANES III). J Clin Endocrinol Metab. 2002;87(2):489-499. doi:10.1210/jcem.87.2.8182 doi:10.1210/jcem.87.2.8182
- 2.Brent GA. Mechanisms of thyroid hormone action. J Clin Invest. 2012;122(9):3035-3043. doi:10.1172/JCI60047 doi:10.1172/JCI60047
- 3.De Groot LJ. Non-Thyroidal Illness Syndrome. In: Feingold KR, Anawalt B, Blackman MR, et al., editors. Endotext [Internet]. South Dartmouth (MA): MDText.com, Inc.; 2000-. Updated 2015 Nov 1. PMID: 25905425
- 4.Canaris GJ, Manowitz NR, Mayor G, Ridgway EC. The Colorado thyroid disease prevalence study. Arch Intern Med. 2000;160(4):526-534. doi:10.1001/archinte.160.4.526 doi:10.1001/archinte.160.4.526
- 5.Biondi B, Cooper DS. The clinical significance of subclinical thyroid dysfunction. Endocr Rev. 2008;29(1):76-131. doi:10.1210/er.2006-0043 doi:10.1210/er.2006-0043
- 6.Surks MI, Ortiz E, Daniels GH, Sawin CT, Col NF, Cobin RH, Franklyn JA, Hershman JM, Burman KD, Denke MA, Gorman C, Cooper RS, Weissman NJ. Subclinical thyroid disease: scientific review and guidelines for diagnosis and management. JAMA. 2004;291(2):228-238. doi:10.1001/jama.291.2.228 doi:10.1001/jama.291.2.228