Journal of Pediatrics最新文献

Objective: To investigate the effect of fetal growth restriction (FGR) on cardiovascular structure and function in childhood in a growth-discordant identical twin model, thereby adjusting naturally for genetic, obstetric, and environmental confounding factors.

Study design: This study is part of the LEMON cohort study of growth-discordant identical twins born at the Leiden University Medical Center, The Netherlands. Echocardiography was prospectively performed in 42 twin pairs aged between 5 and 17 years. Growth discordance was defined as a birth weight discordance ≥20%. Outcome measures were compared between the smaller and larger twin and included structural measures (left ventricular [LV] dimensions and carotid intima-media thickness) as well as functional measures (LV systolic and diastolic function using Doppler and LV strain imaging, and aortic pulsed wave velocity).

Results: Median gestational age at birth was 34.1 weeks (IQR 31.6-36.0) with median birth weights of 1471 g (IQR 1114-1876) and 2025 g (IQR 1623-2705) for the smaller and larger twin, respectively. Median age at echocardiography was 11 years (IQR 9-14). Smaller twins presented with significantly lower LV global longitudinal strain (-19.6% vs -20.7%), and higher mitral peak E and A velocities, and mitral E/e' ratio compared with larger co-twins. No within-pair differences were found regarding structural cardiovascular parameters, carotid intima-media thickness and aortic pulsed wave velocity.

Conclusions: In a growth-discordant identical twin model, FGR is associated with subclinical changes in myocardial structure and systolic and diastolic LV function in the smaller growth-restricted twin compared with the larger appropriately-grown co-twin. This finding indicates that adverse cardiovascular remodeling after FGR persists into childhood, potentially influencing long-term risk of cardiovascular disease.

Trial registration: International Clinical Trial Registry Platform ID NL9833 (https://trialsearch.who.int/).

Objective: To quantify the impact of nonhemodynamically significant patent ductus arteriosus (non-hsPDA) on pulmonary function in neonates.

Study design: We analyzed a retrospective cohort of very low birthweight (<1500 g) infants born <32 weeks' gestation admitted to a single tertiary neonatal intensive care unit between 2019 and 2023 who underwent both echocardiography and pulmonary function testing at approximately 34 weeks postmenstrual age. Infants with ventricular or atrial septal defects were excluded. Passive respiratory compliance (Crs) was compared among infants with hsPDA, non-hsPDA, and age-matched controls without PDA using one-sided t-tests and ANOVA, with multivariable regression controlling for birth weight and length.

Results: Twenty-four infants were analyzed (8 hsPDA, 8 non-hsPDA, 8 controls). Mean Crs was significantly reduced in both hsPDA (0.83 ± 0.34 mL/cmH₂O/kg) and non-hsPDA (0.92 ± 0.21) groups compared with controls (1.23 ± 0.34; P < .05). The presence of PDA demonstrated a stepwise inverse relationship with Crs that persisted after adjustment for covariates.

Conclusions: Even non-hsPDAs are associated with measurable reductions in pulmonary compliance among preterm infants. These findings suggest that current binary classifications of PDA significance may underestimate pulmonary impact and support re-evaluation of management thresholds for "non-significant" shunts.

Objective: To build a time-series machine learning (ML) model that improves bronchopulmonary dysplasia (BPD) prediction compared with published online calculators.

Study design: We used a single-center, extremely low gestational age newborn cohort (inborn, birth year 2016-2021, n = 438). The primary outcome was a 5-level class outcome for BPD as defined by the Neonatal Research Network (NRN) in 2019. Flowsheet data were extracted from the electronic medical record. Time-series data were generated from birth onward, with 14 static and 35 dynamic input attributes. Iterative static (regression) and dynamic (ML) modeling was performed, comparing model performance with the NRN BPD calculator at several time points (postnatal day 1, 3, 7, 14, and 28) and ranking feature leverage at each time point.

Results: Of the original cohort, 92 infants met all inclusion criteria (gestational age 25.6 ± 1.4 weeks). Static models performed comparably with the NRN BPD calculator (area under the curve = 0.7460), improving to 0.7978 with forward/backward selection. In contrast, dynamic long short-term memory (LSTM) models outperformed static models at all time points, reaching a peak area under the curve of 0.8400 on postnatal day 28. LSTM models performed best for no BPD and severe disease/death. Principal component analysis revealed that respiratory support, ventilator settings, supplemental oxygen requirements, medications, and prenatal/postnatal growth were major factors driving BPD severity.

Conclusions: LSTM-based ML time-series analysis substantially outperformed static approaches for predicting BPD and death among extremely low gestational age newborns. Integrating ML methods into clinical applications holds promise for enhancing real-time BPD trajectory mapping.