Urinary 6-sulfatoxymelatonin as a predictive biomarker for brain injury in very preterm infants

Brain injury in preterm infants (BIPI) remains a significant clinical challenge with limited diagnostic biomarkers. This study aimed to investigate urinary 6-sulfatoxymelatonin (6-SMT) levels as a potential noninvasive biomarker for brain injury in very preterm infants. A prospective cohort study was conducted with 127 very preterm infants admitted to our hospital from January to December 2024. Infants were categorized into brain-injury and control group based on neuroimaging findings. Urinary 6-SMT concentrations were measured on postnatal days 1, 3, and 7 using ELISA. Clinical parameters and perinatal risk factors were evaluated. Receiver operating characteristic (ROC) curve analysis was used to assess diagnostic efficacy, and conditional logistic regression analysis was performed to evaluate the association between urinary 6-SMT levels and brain injury. Compared with the control group (n = 97), infants with brain injury (n = 30) exhibited significantly lower gestational age (GA: 29.36 vs. 30.43 weeks, p = 0.028) and birth weight (BW: 1.21 vs. 1.34 kg, p = 0.030), significantly lower rates of acidosis (10.0% vs. 12.4%, p = 0.023) and antenatal magnesium sulfate (MgSO₄) exposure (30.0% vs. 60.8%, p = 0.003), and significantly higher rates of early infection (46.7% vs. 21.6%, p = 0.007) and asphyxia (30.0% vs. 1.0%, p = 0.014). Crucially, urinary 6-SMT levels were markedly lower in the brain injury group on all measurement days (day 1: 558.51 vs. 813.86 pg/mL, p = 0.015; day 3: 722.62 vs. 938.48 pg/mL, p < 0.001; Day 7: 796.81 vs. 1034.48 pg/mL, p = 0.014). ROC analysis identified day 3 urinary 6-SMT (cut-off 762.46 pg/mL) as the best single marker (AUC = 0.714, sensitivity 73.2%, specificity 70.0%), while a combined model integrating levels from days 1, 3, and 7 achieved superior diagnostic performance (AUC = 0.764, sensitivity 78.4%, specificity 66.7%). Urinary 6-SMT levels progressively increased during the first postnatal week and correlated positively with GA and BW on days 1 and 3, but correlations weakened by day 7. After 1:1 matching on GA and BW, conditional logistic regression analysis confirmed a significant protective association between higher day 3 urinary 6-SMT levels and reduced brain injury risk (adjusted OR = 0.996, p = 0.004). Reduced urinary 6-SMT concentrations in very preterm neonates are significantly associated with brain injury. Serial urinary 6-SMT measurements, particularly when combined across multiple time points, demonstrate promising diagnostic potential as a noninvasive biomarker for BIPI. These findings suggest that melatonin deficiency may contribute to the pathophysiology of preterm brain injury and warrant further investigation for clinical translation.