The Establishment of Reference Intervals for Thyroid Hormone Tests in the Korean Population: Using a Direct Selection Technique.

Background/Objectives: Thyroid-stimulating hormone (TSH), free thyroxine (FT4), and total triiodothyronine (TT3) are biomarkers for evaluating thyroid function. Although hormone levels are affected by many biological and environmental factors, most laboratories use manufacturer-provided reference intervals (RIs) without considering these factors. Thus, in this study we assessed RIs for TSH, FT4, and TT3 in a Korean population, using a direct selection technique. Methods: Serum samples from patients without a history of thyroid disease, medication, family history, or antibody-positive test results were collected after a review of medical records. TSH, FT4, and TT3 levels were measured using the Cobas e801 analyzer (Roche Diagnostics GmbH, Mannheim, Germany) with dedicated reagents. RIs were then established using a non-parametric method, using values at the 2.5th and 97.5th percentiles as reference limits, which were then verified in a validation cohort. Results: A total of 618 subjects were enrolled in this study. Because the distribution of reference values for the four subgroups divided by sex and age (65 years) showed insignificant differences, combined RIs were determined, with the established RIs being 0.38-5.46 mIU/L for TSH, 12.28-22.40 pmol/L for FT4, and 0.94-2.32 nmol/L for TT3. When compared to manufacturer-claimed RIs, the Korean RI for TSH showed higher upper limits, while that for TT3 showed lower upper limits. Additionally, when newly established RIs were applied to the validation cohort, the rate of test-positive results decreased significantly. Conclusions: Significant differences in RIs for TSH and TT3 in the Korean population, compared to manufacturer-claimed values, highlight the need for population-specific RIs. Thus, interpreting the results for the Korean population requires caution, and Korean population-based RIs are necessary.