Should my hypothyroidism be treated?

Continuing through the ATA/AACE Guidelines for Hypothyroidism in Adults (Endocrine Practice 2012;18(No.6), let's talk about whether or not hypothyroidism should be treated.  While the answer may appear obvious, there are some nuances.

First, how to define hypothyroidism. 

NHANES III (a great-big population study of Americans) found those who were not on thyroid medication and said, "I have a healthy thyroid" can have a TSH of up to 4.5uIU/mL.  When those same patients were more closely sorted out, and removed those with positive anti-thyroid antibodies, pregnant, or on medications such as estrogens (like birth-control pills), androgens (like testosterone), and lithium, the upper limit of normal dropped to 4.12uIU/mL. 

Another group altogether, the National Academy of  Clinical Biochemists, found 95% of patients without evidence of thyroid disease have a TSH less than 2.5uIU/mL.  However, the upper limits of normal in the NHANES group (above) was likely defined by the top 97.5%, so this is a little bit like comparing apples to oranges. . . . .

TSH can vary amongst ages and races, with a TSH as high as 7.8uIU/mL being acceptable for Mexican-Americans >80years old.  It can even vary, within the normal range, in the same person by 50%.  This means someone can have a TSH of 2.5uIu/mL one day, 1.25uIU/mL another day, and 3.75uIU/mL yet another day.

So, now on to the treatment (let's answer the question, yes):

If your TSH is higher than 10uIU/mL, you won't find many folks who will argue with you that medication is the right way to go.  If it is 0.5-2.5uIU/mL, your thyroid is most likely functioning properly, and medication would not necessarily address symptoms. 

How about the grey area?  Let's split it in half.  Some studies have shown those with TSH 2.5-4.5uIU/mL who have high cardiovascular risk factors such as elevated lipoprotein a (Lpa), poor endothelial function, and high intima-medial thickness (all research tests, not widely clinically done), may have benefit from treatment, though it remains controversial.  The other population, better studied, are pregnant ladies (even antibody-negative), in order to decrease risk of miscarriage and stillbirth.

The other half, those with TSH 4.5-10uIU/mL, are not statistically likely to benefit.  However, by the time patients make it to my office (I am, after all, an endocrinologist), they have had enough drive for treatment that I will oblige as long as the patient is willing to take medication daily and have labs regularly.