Highlights
1. The TSH-FT4 relationship is segmented, not uniformly log-linear, with distinct slopes identified at breakpoints of 0.22 and 4.25 mIU/L.
2. The upper breakpoint closely aligns with harmonized reference limits, while the lower breakpoint shows discordance, suggesting potential overestimation of the lower reference interval.
3. FT4 remains stable between TSH values of 0.22–4.25 mIU/L, indicating a physiological homeostatic plateau.
Background
Thyroid function tests are cornerstone diagnostics in endocrinology, relying heavily on the presumed log-linear relationship between thyrotropin (TSH) and free thyroxine (FT4). However, this assumption may oversimplify the hypothalamic-pituitary-thyroid axis’s feedback mechanisms. Misalignment between empirical data and reference intervals can lead to overdiagnosis or missed cases of thyroid dysfunction. This study aimed to empirically characterize the TSH-FT4 relationship and evaluate its concordance with current reference limits.
Study Design
The retrospective analysis included 782 paired TSH-FT4 measurements from untreated outpatients (median age 43 years), excluding those with TSH receptor antibodies, thyroid hormone treatment, or pregnancy. Segmented regression models were applied to identify optimal breakpoints in the TSH-FT4 relationship, with model fit assessed via the Akaike Information Criterion.
Key Findings
The two-breakpoint model provided the best fit, revealing significant shifts in the TSH-FT4 slope at 0.22 and 4.25 mIU/L. The FT4 plateau between these breakpoints suggests a homeostatic range where thyroid function is tightly regulated. Notably, the upper breakpoint closely matched assay-specific and harmonized reference limits (4.20–4.27 mIU/L), while the lower breakpoint (0.22 mIU/L) fell below conventional lower limits (0.27 mIU/L), prompting questions about the clinical validity of current thresholds for subclinical hyperthyroidism.
Expert Commentary
Dr. Kasahara’s findings underscore the complexity of thyroid axis regulation and highlight the limitations of one-size-fits-all reference intervals. The study’s segmented regression approach offers a nuanced framework for interpreting borderline thyroid function tests, particularly in cases where small TSH deviations may not reflect true pathology. However, the retrospective design and single-center data limit generalizability, warranting validation in diverse populations.
Conclusion
This study challenges the dogma of a uniformly log-linear TSH-FT4 relationship, proposing a segmented model with clinical implications for refining reference intervals. Empirically derived breakpoints may complement population-based standards, especially in equivocal cases. Future research should explore the impact of these findings on diagnostic accuracy and patient outcomes.
Funding
Not reported.
References
Kasahara T. Empirical Thyrotropin-Free Thyroxine Relationship Across the Thyrotropin Spectrum and Implications for Reference Intervals. Thyroid. 2026;36(4):10507256261445291. PMID: 42011873.
