Evaluating the Link Between Thyroid Function Test Results and Levothyroxine Dose in the Management of Hypothyroidism:...

Over 10 million thyroid function tests (TFTs) are carried out in England each year, most requests coming from primary care. Our previous work showed that only 25% of results for patients being treated with Levothyroxine fell within the TSH/FT4 boundary circumscribing 75% of untreated individuals. This study aimed to investigate further the differences in thyroid hormone levels, taking into account both diagnostic code and amounts prescribed.

Using a city-wide population record, we analysed TSH/FT4 simultaneous results from 47,869 consecutive diagnosed hypothyroid individuals by medication dose and 393,101 untreated/euthyroid individuals over 14 years. For those on medication, we only included those who were diagnosed over 2 years ago, had no more than two tests per year and more than 2 years of test results available. For those not on medication, we included results from those patients who had a single test or two tests with more than 4 years between tests.

The FT4 distribution for Levothyroxine-treated individuals was similar in shape versus untreated individuals but shifted towards higher FT4 even at the lowest dose of Levothyroxine, with an increasing separation of the distributions as Levothyroxine dose increased (F value = 1.5 increasing to F value = 4.2). In contrast, the distribution of TSH was substantially different for untreated individuals versus those on Levothyroxine, where the distribution was massively skewed to low or undetectable TSH with a 'hockey stick' configuration, with increasing skewness as doses of Levothyroxine rose. For those not on thyroid hormone replacement, 90.3% of individuals were within the TSH reference range and of these, 0.8% were recorded with a low FT4. For those on medication, only 43.8% were within the TSH reference range. For men versus women, the median Levothyroxine dose was higher in all decades, with the highest median daily dose at age 50-59 years (men: 107 mcg/day; women 93 mcg/day). Median T4 rose (women > men) and TSH fell progressively (women > men) by age in treated individuals. The levels of TSH in treated and untreated populations were only similar at around FT4 = 20pmol/L: below this treated patients have a higher TSH and above it, treated have a lower TSH for the same FT4.

We have here described that distribution of FT4/TSH is different in people on and off Levothyroxine treatment. For those on Levothyroxine, only 43.8% were within the TSH reference range and the degree of difference increased in treated individuals with Levothyroxine daily dose. The potential implication of our findings is that clinicians must be mindful as they diagnose and treat hypothyroidism that the administration of Levothyroxine, while in most but not all individuals is clinically beneficial does not return the individual to the same balance of TSH and FT4 as seen in euthyroid individuals.