Defining age partitions for thyrotropin (TSH) reference intervals in infancy: a robust statistical approach

Age-specific thyrotropin (TSH) reference intervals (RIs) during the first year of life remain inadequately defined, posing a significant challenge given the essential role of thyroid hormones in normal growth and neurodevelopment. Therefore, applying age-specific TSH RIs in the paediatric population is essential for accurate interpretation and to prevent diagnostic uncertainty and unnecessary testing. This indirect RI study aimed to establish statistically derived age partitions for TSH RIs that reflect thyroid physiology during the first year of life. De-identified TSH results from infants were extracted from a tertiary-care public hospital in Victoria, Australia. After excluding results from infants with repeat testing, extreme TSH values, and conditions affecting thyroid physiology, the remaining data were analysed using the Mann-Whitney U test, applied sequentially across 3-monthly, weekly, and daily intervals to identify physiologically relevant inflection points for age partitioning. The derived partitions were then validated using the proportion criteria described by Lahti et al., and TSH RIs were calculated as the 2.5th-97.5th percentiles. Of 17,525 TSH results from 11,045 infants, 7708 met the inclusion criteria. Five age partitions were established: days 0-1, 2, 3, 4-89, and 90-365. TSH peaked on days 0-1 (RI: 1.9-40.1 mIU/L), declined sharply by day 2, and decreased more gradually to near-adult levels by days 90-365 (RI: 0.7-5.7 mIU/L). No clinically significant sex differences were observed. This study presents a robust, data-driven framework for defining age-specific TSH RIs during the first year of life. The proposed methodology enhances diagnostic accuracy for thyroid disorders in infants and provides a reproducible approach for defining physiologically meaningful age partitions for other age-dependent analytes.