Teresa brought three lab printouts to her first visit, each from a different office, each circled in the same place. TSH: 2.1. TSH: 2.8. TSH: 3.4. Normal every time. Call her Teresa, a composite drawn from the patients who arrive with the same printout and the same question, not one person. She was thirty-four. Exhausted by two in the afternoon most days. Hair coming out in the shower drain. Eighteen pounds heavier in a year without eating any differently. Her doctor had told her, more than once, that her thyroid was fine.
Her thyroid might have been fine. It might not have been. The number on that page cannot tell you which, and almost nobody explains why.
TSH stands for thyroid-stimulating hormone. It doesn’t come from the thyroid. It comes from the pituitary, and it rises when the pituitary senses too little thyroid hormone circulating, and falls when it senses too much. That makes it a reasonable first screen, and for most people it’s also the last one they need. Running every marker on every patient would create more noise than it resolves, and a doctor who stops at TSH is usually making the right call, not a lazy one. The visits that go wrong are the ones where TSH doesn’t close the question and nobody looks further.
What “normal” actually means on that page
A lab’s reference range is built from a wide population, usually somewhere near 0.4 to 4.5 milliunits per liter. Fall inside the line, and the result reads normal. Fall outside it, and it gets flagged. But a 2002 study in the Journal of Clinical Endocrinology & Metabolism followed sixteen healthy men for a year, drawing blood every month, and found something a population range cannot capture. Each man had his own narrow set point for TSH and thyroid hormone, with an individual range roughly half the width of the lab’s reference range (Andersen et al., J Clin Endocrinol Metab, 2002). All sixteen were men. Nobody has run that same year of monthly blood draws on a cohort of women, and I’d rather say so than let the study carry more than it measured. The pituitary-thyroid loop producing that result is the same loop in everyone. The exact numbers, measured only in men, are the part still unconfirmed in women. A TSH of 3.4 can sit comfortably inside the lab’s normal window and still be nowhere near normal for the person carrying it.
That isn’t a loophole. It’s how the feedback loop actually behaves. The range on the printout describes a population. It was never built to describe Teresa.
The gap most visits never check
Subclinical hypothyroidism, a mildly elevated TSH with the hormone itself still testing in range, showed up in roughly one to two percent of U.S. adults depending on the threshold used, in a 2022 analysis combining NHANES survey data with insurance claims through 2019 (Wyne et al., J Endocr Soc, 2022). Scaled to the adult population, that is millions of people. Most never get flagged, because where “elevated” starts and “in range” ends is a line a lab drew, not a law written into the body.
Then there’s the rarer problem I don’t want to overstate, because overstating it is exactly what a worse version of this article would do. Central hypothyroidism, where the thyroid itself works fine but the pituitary signal driving it is blunted, can show up with a TSH that reads completely normal, sometimes even low, while the actual hormone level the tissue depends on, free T4, sits below where it should. It is genuinely uncommon. Estimates run from roughly one in sixteen thousand people to one in a hundred twenty thousand (Lania, Persani, Beck-Peccoz, Pituitary, 2008). I am not telling Teresa she has this. Almost no one reading this does. But rare is not never, and a TSH alone cannot rule it out, because a normal-looking TSH is precisely what this condition produces.
What a fuller workup actually looks like
A complete picture needs more than one number. Free T4 measures the hormone actually available to tissue. Free T3 measures the active form the body converts T4 into, and that conversion step fails in some people for reasons a TSH will never show. Thyroid peroxidase and thyroglobulin antibodies catch Hashimoto’s, the autoimmune process that can be quietly damaging the gland years before TSH ever moves. None of this is exotic. All of it is drawn from the same visit, one needle.
What it costs and how fast it comes back depends on your insurance and your lab, so I won’t hand you a national number I can’t stand behind. What I can tell you from the exam room: results land in days, not weeks. The harder part isn’t the blood draw. It’s the conversation after, sorting out whether a borderline free T3 or a positive antibody actually explains the fatigue and the hair and the weight, or whether something else does too, or does more. Sometimes that fuller panel comes back clean too, free T3 fine, antibodies negative. That doesn’t mean nothing is wrong. It means the thyroid answered and the real question moved somewhere else. Perimenopause can produce this same pattern of normal labs and real symptoms, for its own reasons. So can testosterone that reads normal on the wrong test. A normal TSH closes one door. It does not close the room.
This is also why standard blood work misses what it misses in the first place, and why a fuller functional workup starts by asking which test actually answers the question in front of you, not which test is fastest to order.
Teresa, and what actually happened
Teresa’s TSH, on the fourth draw, was still 2.6. Her free T3 was low. Her antibodies were positive. None of that would have shown up if the only question anyone asked was whether her TSH crossed a line on a page.
She is not cured. She is not transformed. She sleeps better than she did in the spring. The hair loss slowed by March. The two o’clock wall is mostly gone, not entirely. That’s what a fuller answer to one number can actually do. It doesn’t fix everything. It stops the wrong question from standing in for the real one.
If your labs keep coming back normal and you don’t, the lab was probably not lying to you. It was probably just never asked the right thing.