The Shenzhen neonatal ARDS cohort study: a multi-omics approach to elucidating regional epidemiology, refined...

Neonatal Acute Respiratory Distress Syndrome (NARDS) is a critical contributor to neonatal morbidity and mortality, with a global health burden that varies significantly by region. The Montreux definition provides a unified diagnostic framework; however, a significant clinical paradox exists. A prospective cohort in China reported a NARDS mortality rate of 12.6%, which is notably lower than the 17%-24% reported in a large-scale international prospective study. The underlying reasons for this discrepancy remain to be elucidated, whether due to differences in etiology, clinical practice, or patient demographics.

The Shenzhen Neonatal ARDS Cohort Study (SZ-NARDS) is a prospective, multicenter observational cohort study spanning from 2025-2028, designed to address this knowledge gap. We will enroll more than 1,000 neonates who meet the Montreux criteria across nine tertiary neonatal intensive care units (NICUs) in Shenzhen, China. Longitudinal data collection includes granular clinical parameters, respiratory support metrics, and multi-modal biospecimens for deep phenotyping and multi-omics profiling. Survivors will undergo rigorous follow-up until 36 months' corrected age, with standardized neurodevelopmental, pulmonary, and growth assessments.

The primary objective of this study is to characterize the epidemiology of NARDS in this regional population and to test the following hypotheses: (1) The true incidence, etiology, and mortality rates of NARDS in Shenzhen will differ from existing international and Chinese cohorts, and these differences can be systematically explained by specific clinical and demographic factors. A multi-modal predictive model that integrates early clinical variables with multi-omics biomarkers has the potential to accurately identify neonates at high risk for severe NARDS [oxygenation index (OI) ≥ 16] and long-term adverse outcomes [Area Under the Receiver Operating Characteristic Curve (AUROC) > 0.85].

The SZ-NARDS cohort is uniquely positioned to resolve a major clinical contradiction in NARDS epidemiology. By integrating deep phenotyping with a longitudinal biobank and advanced machine learning algorithms, this initiative will generate a comprehensive dataset. This dataset will serve to refine existing prognostic models, identify regional disparities in disease biology, and inform the development of precision medicine interventions for this vulnerable population.

Chinese Clinical Trial Registry, identifier ChiCTR2400093854.