TSH test: fasting, timing and reflex testing

A TSH test is a standard venous blood draw. Modern labs use third-generation immunoassays that can measure values near 0.02 mU/L, low enough to tell a suppressed TSH from a low-normal one [7].

The AACE and ATA call TSH the single best screening test for primary thyroid dysfunction in most outpatient situations [4]. NICE recommends testing when there is clinical suspicion of thyroid disease, and offering it to people with type 1 diabetes, other autoimmune diseases or new atrial fibrillation [3]. Routine screening of people without symptoms is not recommended [8].

In 57 adults tested fasting and 2 hours after food, TSH fell after eating in every participant while free T4 did not change; 15 of 20 people classed as subclinical hypothyroid on the fasting sample were in range after the meal [2]. It was a small study, but it shows why conditions matter for borderline values.

If your blood draw also includes fasting tests, such as glucose or fasting insulin, fasting covers both. If you take levothyroxine, ask your clinician whether to take it before or after the draw, and keep that consistent.

In 465,593 TSH results from one lab, the upper reference limit changed with the hour of the day, and the nighttime TSH surge grew with age [5]. Surks and colleagues note that early morning values are higher than later ones, and that sleep deprivation, strenuous exercise and night or evening shift work accentuate the effect [1].

Practical points:

• Book repeat tests at a similar time of day.
• Avoid a test straight after a night shift or a sleepless night when you can.
• Keep a regular sleep schedule in the days before; the sleep consistency protocol shows how.

Even with perfect timing, a person's TSH varies from month to month; a change of about 0.75 mU/L was needed to be confident a repeat result had really changed [9].

NICE describes the same approach as cascading: measure TSH alone for adults when pituitary disease is not suspected, then add free T4 in the same sample if TSH is high, and free T4 plus free T3 if TSH is low [3].

Reflex testing has one weak spot. When the pituitary itself is the problem (central hypothyroidism), TSH can look normal while free T4 is low. NICE therefore recommends measuring both TSH and free T4 when pituitary disease is suspected, and in children [3]. TSH alone can also mislead in critical illness [10].

If you take levothyroxine and still have symptoms, NICE suggests measuring free T4 as well as TSH [3]. Read how the two hormones fit together in TSH vs free T4.

• Biotin: after 7 days of 10 mg a day, 9 of 23 biotin-based immunoassays gave wrong results; competitive assays such as free T4 read falsely high and sandwich assays such as TSH falsely low [6]. NICE asks clinicians to check biotin intake [3]. Ask your clinician or the lab how long to pause biotin before a draw.
• Acute illness: NICE advises not testing during an acute illness unless the illness may be thyroid-related [3]. TSH can rise above normal during recovery [11].
• Medicines: glucocorticoids and dopamine agonists can lower TSH [12]; lithium raises it [13].
• Levothyroxine changes: TSH takes weeks to settle, so it is usually checked 6 to 8 weeks after a dose change [8].
• Different assays: labs can differ by about 1 mU/L near 4.0 [14].

• After a mildly raised result: repeat TSH and free T4 with TPO antibodies after 2 to 3 months [15].
• Untreated subclinical hypothyroidism: NICE suggests TSH and free T4 once a year if there are features of thyroid disease, and once every 3 years if not [3].
• Untreated subclinical hyperthyroidism: TSH every 6 months [3].
• On levothyroxine: every 3 months until stable, then yearly [3]; 6 to 8 weeks after any dose change [8].
• New or worse symptoms: NICE suggests a repeat test, but no sooner than 6 weeks from the last one [3].

Your clinician may choose a different schedule for pregnancy, heart disease or other conditions.