ETA2022 Poster Presentations Miscellaneous (10 abstracts)
1Aalborg University, Department of Clinical Medicine, Department of Clinical Medicine, Aalborg, Denmark; 2Aalborg University Hospital, Aalborg University, Department of Clinical Biochemistry, Aalborg, Denmark; 3Aalborg University, Department of Clinical Medicine, Aalborg, Denmark; 4Aalborg University Hospital, Geriatric Medicine, Aalborg, Denmark; 5Dept Of Endocrinology, Aalborg Universityhospital, Dept. of Endocrinology, Aalborg, Denmark
Background: Morbidity and mortality are associated with thyroid hormone levels in populations. This has invited a debate on what is the better parameter for assessment of thyroid function: the controlling hormone thyrotropin (TSH) or peripheral thyroid hormone measurements (thyroxine (T4) and triiodothyronine (T3)).
Aim: To assess the ability of TSH and total T4 (TT4) to discriminate between subtle differences in thyroid function.
Methods: Monthly collection of blood samples over one year in 35 subjects leading to 420 samples in 15 euthyroid and 20 mildly hypothyroid subjects. The latter group was classified as subclinically hypothyroid (SH) by two independent tests of thyroid function prior to inclusion. None of the participants received current treatment for thyroid disease. We measured TSH and total T4 in serum. Reference ranges with our assays were 0.4-4.5 mU/l for TSH and 60-140 nmol/l for TT4.
Results: The true thyroid state was confirmed by the mean of 12 repeated measurements, which was 1.19/7.23 mU/l for TSH in the euthyroid/SH subjects and 106/85 nmol/l for TT4. No single TSH measurement was above the upper limit of the reference range for TSH in the euthyroid group, and thus 100% of test results conformed to the euthyroid state. In the SH group, 86% of test results were in keeping with the true thyroid state while 14% were within the reference range. For TT4, 99% of test results were within the reference range among euthyroid subjects while this was 96% for the SH group. An overlap between the two groups was markedly more pronounced for TT4 compared to TSH. Bootstrap estimates based on 1000 replications showed an estimated area under the curve of 0.999 (95%-CI: 0.995; 1.000) for TSH and 0.853 (0.736; 0.935) for TT4. There was no confidence interval overlap between participant groups for TSH, and hence markedly better performance of TSH compared to TT4.
Conclusion: The two groups differed more clearly when evaluated by TSH than by TT4. The TSH measurements were outside the reference range for 86% of individuals with SH compared 4% of TT4 measurements. Thus, our findings point to a higher diagnostic power for TSH compared to TT4 for separating individuals with mild hypothyroidism, and TSH is the most sensitive and accurate index of thyroid status at an individual level.