Thyroid Test Ranges Vary Widely Between Affluent and Slum Clinics
A thyroid-stimulating hormone result of 6.0 mIU/L can mean two different things depending on which clinic ordered it. Where the lab's upper limit sits near 4.2, that number reads as abnormal and invites a prescription. Where the limit sits near 6.5, the same number reads as normal and invites a repeat test in three months. The patient never sees the difference.
The Same TSH, Two Different Normal Ranges
Thyroid-stimulating hormone, or TSH, is the first-line blood test for thyroid dysfunction. It is cheap, stable, and sensitive to small shifts in thyroid output. A high TSH usually means the pituitary is pushing a sluggish gland harder. A low one usually means the gland is overactive or the patient is over-replaced.
What counts as high depends on the reference range printed beside the result. Laboratories set these ranges themselves, usually from a sample of local healthy people. Two labs running the same assay can publish different upper limits because their reference populations differ.
In practice, clinics serving wealthier catchment areas tend to work with tighter upper limits, often somewhere in the low-to-mid 4 mIU/L band. Clinics serving poorer populations may work with limits that run higher, sometimes past 5 or 6 mIU/L, reflecting older kit inserts or a reference group with more undiagnosed disease in it.
How Reference Ranges Get Built
A reference range is a statistical interval. Laboratories recruit a group of people who appear healthy, measure the analyte, and take the central 95 percent of results. By construction, one in twenty truly healthy people falls outside the range.
The trouble sits in who gets recruited. A reference population drawn from a well-screened, iodine-sufficient, largely non-pregnant group will produce a lower upper limit than one drawn from a catchment where goitre, iodine deficiency, and chronic illness are common. Undiagnosed thyroid disease in the reference group drags the curve upward.
Iodine status shifts the whole distribution. Populations with mild deficiency tend to run higher TSH values across the board, so a range derived there will call fewer people abnormal. Pregnancy does the same thing in the other direction, since TSH normally falls in the first trimester and normal ranges are trimester-specific.
A range is not a diagnosis. It is a comparison tool, and it is only as good as the people it was built from. When the catchment and the reference group diverge, the printed interval stops describing the patient in front of you.
What the Evidence Actually Shows
Guidelines are consistent on overt hypothyroidism: a clearly raised TSH with a low free thyroxine level warrants treatment. That is not in dispute, and the benefit is well established for symptoms and for pregnancy outcomes.
Subclinical hypothyroidism is a different problem. Most mild TSH elevations, roughly in the 4 to 10 mIU/L band with a normal free T4, do not progress to overt disease. Many revert to normal on repeat testing without any intervention at all.
Trials of levothyroxine for mild TSH rises in older adults have generally shown little benefit on symptoms, cognition, or cardiovascular endpoints. The evidence is weak enough that several guideline bodies now advise against routine treatment above a certain age, and against treatment at all when TSH is only slightly raised.
Overdiagnosis has a cost. Once levothyroxine is started for a borderline result, patients often stay on it for years, with repeat blood tests, dose adjustments, and a small but real risk of over-replacement and atrial fibrillation. This site has argued before that depression care differs sharply between poor and wealthy clinics, and the thyroid pathway shows a similar pattern of unequal thresholds.
Why Clinics Choose Different Cutoffs
Laboratory contracts and assay platforms differ by setting. A well-funded clinic may run an automated immunoassay with a manufacturer's reference range validated on a large, screened population. A smaller or under-resourced lab may still use an older insert range that was never locally validated.
Affluent clinics can afford confirmatory tests, thyroid antibodies, and repeat sampling. That capacity makes a tighter cutoff safer, because a borderline result can be clarified rather than acted on. The extra test costs money, and that money is available.
Slum clinics face cost, transport, and follow-up constraints. A wider range reduces referrals and repeat visits, which matters when a patient may not return for three months. The trade-off is real: fewer false positives, but also more true disease left untreated.
A wider range is not simply worse care. It can be a rational response to a system where an unnecessary referral consumes a slot that a sicker patient needs. The objection comes from the other direction: a patient with genuine hypothyroidism whose TSH sits at 5.8 may be told everything is fine and never tested again.
The Patient in Front of You
Symptoms matter more than a single borderline number. Fatigue, cold intolerance, weight gain, constipation, and dry skin are nonspecific, but their pattern and duration change the interpretation of a mildly raised TSH. A patient with no symptoms and a TSH of 5.5 is a different case from one with clear symptoms and the same value.
Iodine deficiency and pregnancy change everything. A pregnant woman needs trimester-specific ranges and prompt treatment if TSH is raised, because the fetal thyroid does not work independently in early gestation. In iodine-deficient areas, a raised TSH may reflect the whole population's status rather than individual disease.
Comorbid illness and medications alter thyroid tests. Acute illness, recovery from surgery, and drugs such as amiodarone, lithium, and high-dose biotin can all shift TSH without primary thyroid disease. A result drawn during a hospital admission is often not the result to act on.
A repeat test in six to twelve weeks is frequently the answer. It costs one blood draw and resolves most borderline cases without committing anyone to lifelong medication. The related piece on formulary tiers routing the same diabetes drug to two prices makes a similar point about how system design, not biology, decides who gets treated.
Practical Steps for Clinicians
Check your laboratory's TSH reference range and ask how it was derived. If the upper limit came from an unvalidated insert or a reference group that does not resemble your catchment, treat the number with more caution than the printout suggests.
Treat overt hypothyroidism, and observe mild TSH rises with a normal free T4. Repeat the TSH before starting levothyroxine for a borderline result, ideally at least six weeks later and outside any acute illness.
Ask about iodine intake, pregnancy status, and current medications at the point of ordering the test, not after the result comes back. These three questions change the interpretation more than the assay platform does.
Document the range you used and the reason for any treatment decision. If you start levothyroxine for a TSH of 6.2, write down why, so the next clinician can review the choice rather than inherit it.
Where a wider local range is in use, flag patients whose results sit near the top of it for a repeat rather than discharge. The range protects the system from unnecessary referrals; the flag protects the patient from being missed.
When the Range Itself Becomes the Diagnosis
There is a quieter problem: the range can start to function as a diagnosis in its own right. A patient whose TSH sits at 4.8 in a clinic with an upper limit of 4.2 may be labelled hypothyroid and started on treatment, even though the value is well within the range used by another clinic down the road. The label sticks. It follows the patient into insurance forms, employment questionnaires, and future consultations. A borderline result becomes a chronic condition, not because the biology changed, but because the threshold did.
This matters most for women in their reproductive years, where a mildly raised TSH can trigger a cascade of investigations and treatment that may not be warranted. It also matters for older adults, where the threshold for treatment should arguably be higher, not lower, given the weak evidence of benefit and the real risk of over-replacement. The same number that prompts action in a 30-year-old may be best left alone in a 75-year-old, yet the reference range does not adjust for age unless the laboratory has built age-specific intervals.
Clinicians can push back by asking a simple question: would I treat this result if the upper limit were 6.0 instead of 4.2? If the answer is no, the decision should probably be to repeat, not to prescribe. That question does not require a new test or a new guideline. It requires only an awareness that the number on the page is not a fact about the patient. It is a fact about the laboratory.
This article is informational and does not constitute personalised medical advice. Thyroid test interpretation depends on individual history, examination, and local laboratory standards. Consult a qualified clinician about your own results.
