Do you need to fast before a thyroid blood test?
Short answer: for a standard thyroid panel, no. Most labs do not require fasting for TSH, Free T4 or thyroid antibodies.
If that is all you came for, you can stop reading.
But fasting is the thing people ask about, and it is not what moves the numbers most. Three other things do, and almost nobody is told about them before they walk into the lab. One of them can make a perfectly normal thyroid look overactive.
Why people ask about fasting at all
Because most blood tests do require it, and nobody separates the panels for you. If a thyroid test is drawn at the same appointment as cholesterol or glucose, which it often is, then you are fasting for those, and the thyroid test comes along for the ride.
Fasting is not entirely irrelevant. Eating lowers TSH somewhat, and for someone sitting near the top of a reference range that is occasionally enough to change which side of the line they land on. If your result is borderline and a decision is being made on it, being consistent, always fasting or always not, removes one source of noise.
Consistency matters more than fasting. That is true of everything below.
The thing that actually matters: biotin
If you take a hair, skin and nails supplement, a prenatal vitamin, or a B complex, read this part twice.
Biotin interferes with the way many thyroid tests are run. The FDA issued a safety communication about it in 2017. It does not change your thyroid. It changes what the machine reads, and it can push results in either direction depending on the assay, often making TSH look falsely low and Free T4 falsely high.
That particular combination is the pattern of an overactive thyroid. In somebody being treated for an underactive one, it can look like their dose is too high.
How long should you stop for? The AACC guidance is a minimum of 8 hours after a 5,000–10,000 mcg dose, and up to 72 hours may be needed for some assays. Higher doses and reduced kidney function both mean longer. Different analysers have different thresholds, so the lab running your sample is the authority on its own equipment. It is a reasonable thing to ring and ask.
The dose on a typical hair and nails supplement is 5,000 or 10,000 mcg. For comparison, the adequate intake for adults is 30 mcg a day, so these products contain a few hundred times what anyone needs from food. People take them for months without ever connecting one to a strange lab result.
What time of day you go
TSH is not a flat number. It peaks overnight, somewhere between about 11pm and 5am, and reaches its lowest point in the late afternoon and early evening. Over 24 hours the swing is roughly twofold.
That is large enough to change what a result is called. In a study of patients with subclinical hypothyroidism, the median TSH was 5.83 mU/L in the morning and 3.79 mU/L in the afternoon, and half the cases identified in the morning would not have met the same threshold if the blood had been taken later in the day. A separate study comparing 7 to 8am draws with 2 to 3pm draws found 6.2% of patients reclassified at the upper end of the reference interval.
This is one reason a result can appear to move between draws when nothing about you has changed. If one was taken first thing and the next was squeezed in after work, some of that difference is the clock.
There is no correct time. There is a consistent one. Going at roughly the same hour each time makes the comparison between draws mean something.
Whether to take your levothyroxine first
Levothyroxine raises Free T4 for several hours after you swallow it. Free T4 peaks a few hours after the dose, so a sample drawn shortly afterwards can read higher than your usual level. A result caught at that peak can look like too much medication when nothing has changed.
The usual approach is to have the blood drawn before the day's dose and take it straight after, which also means not taking it in the waiting room. TSH is much less affected either way, because T4 has a half-life of roughly a week and TSH responds over days rather than hours.
Whether that applies to you is a question for whoever prescribed it. What matters here is that it is a variable, and most people have never been told it is one.
What to actually do
- Fasting is usually not required. If you are fasting for other tests anyway, that is fine. Be consistent either way.
- Stop biotin and B-complex supplements at least 2 to 3 days before, and check with your lab if you take a high dose.
- Go at the same time of day each time. Morning is the conventional answer.
- Ask whether to take your levothyroxine before or after. If nobody has told you, after the draw is the common advice.
- Write down which of these applied, and when. It is the only way to tell a real change from a timing artefact six months later.
That last one is the whole point. One result is a number. A run of results taken under the same conditions is something you and your doctor can actually read.
Where this fits with Haloscan
We built the app around this problem, and the biotin alert exists because of it. If you have logged a supplement containing biotin and a blood draw is coming, it tells you when to stop, 48 and 24 hours out, rather than expecting you to have read an FDA communication.
It also keeps a note of your draws so that when a number moves, you can see what else was different that week.
Sources
Biotin:
- FDA: Biotin Interference with Troponin Lab Tests and Other Assays
- AACC Academy Guidance: Biotin Interference in Laboratory Tests
- Quest Diagnostics: Biotin Interference with Laboratory Assays
- Assessment of biotin interference in thyroid function tests
- NIH Office of Dietary Supplements: Biotin
Time of day:
- Clinical significance of TSH circadian variability in patients with hypothyroidism
- The impact of time of sample collection on the measurement of TSH values in the serum
Medication timing:
Written by a software developer whose family lives with this condition, not by a clinician. Sources are linked so you can check them. It is not medical advice, and nothing here should change what you do without talking to your doctor. Who writes this · Editorial policy
← All posts