Interestingly, however, research by Scobbo et. al. in 2004 showed that TSH tests declined in 97 of 100 of the people studied -- by an average of 26.39% -- when compared to early morning, fasting, TSH test results. This resulted in as many as 6% of patients being reclassified from the diagnosis of subclinical hypothyroidism to "normal."
Researchers concluded that the diagnosis of subclinical hypothyroidism should not be made only on a fasting TSH measurement.
Implications for Patients
Essentially, these researchers retested patients later, when they could get a normal range result and declare the patients to be free of hypothyroidism.
Doing so, however -- retesting a patient without fasting, later in the day, in order to see if an elevated TSH level will then drop down to be "borderline normal" -- is questionable medicine, at best.
It's particularly questionable, given that researchers have not established that the earlier, fasting level results are any less valid than the later, non-fasting, "normal range" results. Researchers also don't even know how fasting affects thyroid function or TSH values, or why TSH values would be different based on fasting or non-fasting.
In some cases, if you are having your thyroid testing done along with other bloodwork -- such as glucose, insulin, or cholesterol levels done -- you will be asked to fast, and have your test done first thing in the morning. In other cases, your doctor may not dictate when to have your TSH test, or whether or not to fast. But be aware that if you test later in the day and you've eaten, you are likely to have a lower TSH level than if you test earlier in the day with fasting.
For more help understanding thyroid blood tests, see:
Key Thyroid Function Tests: Laboratory Values and Interpretation
(Reference: Scobbo RR, VonDohlen TW, Hassan M, Islam S. Serum TSH variability in normal individuals: the influence of time of sample collection. W V Med J. 2004;100:138-142.)

