Umbilical Cord Blood Biomarkers UCH-L1 and GFAP Reveal Fetal Neural Stress in Maternal Sleep-Disordered Breathing
Key Insights
A prospective study of 85 pregnant women found that shorter apnea-hypopnea duration, rather than overall OSA severity, was associated with significantly elevated umbilical cord blood UCH-L1 (search) levels in neonates.
In non-GDM pregnancies, UCH-L1 (search) concentrations were also significantly higher when maternal mean nocturnal oxygen saturation fell to ≤95%, suggesting even mild hypoxemia triggers fetal cellular stress.
GFAP (search) levels did not differ significantly by maternal SDB status, though a non-significant elevation was observed in REM-OSA (search) cases, which were also associated with lower neonatal birthweight.
A new study published in Scientific Reports has identified that specific features of maternal sleep-disordered breathing (search) (SDB)—particularly apnea-hypopnea event duration and mild nocturnal hypoxemia—may trigger measurable fetal neural stress detectable through umbilical cord blood biomarkers. The research, conducted across three Spanish university hospitals, quantified ubiquitin C-terminal hydrolase-L1 (UCH-L1 (search)) and glial fibrillary acidic protein (GFAP (search)) in cord blood from 85 pregnancies, providing the first assessment of these neuronal injury markers in the context of maternal SDB using an automated immunoassay platform.
Shorter Apnea Events Linked to Elevated UCH-L1 (search)
The study found that cord blood UCH-L1 (search) levels were significantly higher in neonates born to women with shorter apnea-hypopnea duration, defined as AH length below the 50th percentile (18.8 seconds), compared to those with longer events. This finding challenges the conventional reliance on the apnea-hypopnea index (AHI) as the primary metric of SDB severity.
"Although traditional metrics such as the AHI are informative, the duration of obstructive events and hypoxemia may better reflect the biological impact of OSA during pregnancy on fetal cellular stress," the authors note. Bivariate analysis across all participants revealed significant negative correlations between UCH-L1 (search) concentration and both AH length (Rho = −0.31) and mean nocturnal oxygen saturation (mSaO₂; Rho = −0.25).
These results align with prior research showing that women with shorter AH events exhibited higher maternal and cord blood protein carbonyl levels, suggesting that transient but repeated hypoxic episodes can trigger acute oxidative stress responses. The authors propose that "fetal neural cells are more sensitive to shorter respiratory obstructive events."
Mild Nocturnal Hypoxemia at the 95% Threshold
The cohort predominantly presented with mild OSA, with a median mSaO₂ of 96% (IQR 95–97). Using a threshold of mSaO₂ ≤95% as an indicator of nocturnal hypoxemia, the researchers identified 31 women meeting this criterion. In a sub-analysis restricted to non-GDM pregnancies, UCH-L1 (search) concentrations were significantly elevated both in cases with maternal mSaO₂ ≤95% and in those with AH length below the 50th percentile.
This finding underscores the potential biological significance of even modest oxygen desaturation during pregnancy. As the authors state, "even mild hypoxemia may impair fetal development," citing recent evidence that maternal mSaO₂ values of 94–95% independently predicted fetal growth restriction.
REM-OSA (search) and GFAP (search): A Signal That Did Not Reach Significance
Overall, no significant differences were observed in GFAP (search) or UCH-L1 (search) cord blood levels based on maternal SDB categorized by OSA (AHI ≥5 h⁻¹), REM-OSA (search) (REM-AHI ≥5 h⁻¹), or mSaO₂ ≤95%. However, GFAP concentrations were higher in REM-OSA cases compared to those without REM-related OSA (194.2 [130.4–279.6] vs. 163.7 [109.6–237.3] pg/mL; p = 0.12). Although not statistically significant—likely due to limited statistical power—this trend may suggest a specific impact of REM-OSA on astroglial injury.
Notably, median birthweight and birthweight percentiles were lower in newborns from women with REM-OSA (search), consistent with prior evidence linking REM-OSA to lower neonatal birthweight and differential cord blood gene expression.
Study Design and Methodology
The prospective study recruited singleton pregnant women in the third trimester from three Spanish university hospitals: Son Espases, Araba, and Miguel Servet. All participants underwent polysomnography (PSG) at a sleep unit, with standard 30-second epoch scoring. Apnea was defined as airflow reduction greater than 90% for at least 10 seconds; hypopnea as airflow reduction of 30–90% for at least 10 seconds with a ≥3% oxygen desaturation or arousal.
Umbilical cord blood was collected in EDTA tubes at delivery, and plasma was analyzed using the Alinity immunoassay platform (Abbott), which simultaneously evaluates both GFAP (search) and UCH-L1 (search). The authors emphasize that this automated method "offers improved reproducibility and sensitivity compared to conventional ELISA."
The median maternal age was 36 years, with a pregestational BMI of 22.9 kg/m² (IQR 20.9–26.8). Obesity (BMI >30 kg/m²) was present in 7.1% of participants, and 29.4% were pregestational smokers. Fourteen women met criteria for OSA, 25 for REM-OSA (search), and 31 had mSaO₂ ≤95%.
GDM as a Potential Confounder
The differential UCH-L1 (search) expression patterns observed in the non-GDM subgroup, but not in GDM-complicated pregnancies, raise important questions. The authors hypothesize that "neuronal injury markers may already be elevated in GDM, potentially masking any additional effects of SDB on these biomarkers." Intrauterine exposure to GDM has been independently associated with increased risk of neurodevelopmental disorders and cerebral white matter alterations.
Limitations and Future Directions
The authors acknowledge several limitations: a relatively small sample size (particularly only 14 women with OSA), multiple secondary comparisons without formal adjustment for multiple testing, assessment limited to the third trimester, predominantly mild OSA severity, and a majority Caucasian population limiting generalizability. Additionally, circulating UCH-L1 (search) in fetuses may not be exclusively brain-specific, as it can also derive from placental and other fetal tissues.
The study was approved by the institutional Ethics Committee of the Balearic Islands (IB1510/10PI), and all participants provided written informed consent.
These findings position UCH-L1 (search) as a potential biomarker for identifying fetuses experiencing neural stress due to maternal SDB, which could support closer monitoring and timely intervention. However, the authors caution that results should be interpreted as "exploratory and hypothesis-generating rather than confirmatory," and call for longitudinal studies to determine whether these biomarker alterations translate into long-term neurodevelopmental consequences.