Associations between Early Thyroid-Stimulating Hormone Levels and Morbidities in Extremely Preterm Neonates

Introduction: High-end cutoffs of thyroid-stimulating hormone (TSH) have been emphasized for hypothyroidism therapy in extremely preterm infants, but the significance of low TSH levels remains unknown. This study hypothesized that the spectrum of TSH levels by newborn screening after birth signifies specific morbidities in extremely preterm neonates. Methods: The multicenter population cohort analyzed 434 extremely preterm neonates receiving TSH screening at 24-96 h of age in 2008-2019. Neonates were categorized by blood TSH levels into group 1: TSH <0.5 μU/mL, group 2: 0.5 ≤ TSH <2 μU/mL, group 3: 2 ≤ TSH <4 μU/mL, and group 4: TSH ≥4 μU/mL. Neonatal morbidities were categorized using the modified Neonatal Therapeutic Intervention Scoring System. Results: The four groups differed in gestational age, birth weight, and the postnatal age at blood sampling so did the proportions of mechanical ventilation usage (p = 0.01), hypoxic respiratory failure (p = 0.005), high-grade intraventricular hemorrhage (p = 0.007), and periventricular leukomalacia (p = 0.048). Group 1 had higher severity scores for respiratory distress syndrome (RDS; effect size 0.39 [95% confidence interval [CI]: 0.18-0.59]) and brain injury (0.36 [0.15-0.57]) than group 2, which remained significant after adjusting for gestational age, birth weight, dopamine usage, and the postnatal age at TSH screening (RDS: mean + 0.45 points [95% CI: 0.11-0.79]; brain injury: +0.32 [0.11-0.54]). Conclusions: Low TSH levels in extremely preterm neonates are associated with severe RDS and brain injuries. Studies recruiting more neonates with complete thyroid function data are necessary to understand central-peripheral interactions of the hypothalamic-pituitary-thyroid axis.