Clinical Chorioamnionitis and Neurodevelopment at 5 Years of Age in Children Born Preterm: The EPIPAGE-2 Cohort Study.

OBJECTIVE: To assess the association between clinical chorioamnionitis and neurodevelopmental disorders at 5 years of age in children born preterm. STUDY DESIGN: EPIPAGE 2 is a national, population-based cohort study of children born before 35 weeks of gestation in France in 2011. We included infants born alive between 240/7 and 346/7 weeks after preterm labor or preterm premature rupture of membranes. Clinical chorioamnionitis was defined as maternal fever before labor (>37.8°C) with ≥2 of the following criteria: maternal tachycardia, hyperleukocytosis, uterine contractions, purulent amniotic fluid, or fetal tachycardia. The primary outcome was a composite, including cerebral palsy, coordination disorders, cognitive disorders, sensory disorders, or behavioral disorders. We also analyzed each of these disorders separately as secondary outcomes. We performed a multivariable analysis using logistic regression models. We accounted for the nonindependence of twins and missing data by generalized estimating equation models and multiple imputations, respectively. RESULTS: Among 2927 children alive at 5 years of age, 124 (3%) were born in a context of clinical chorioamnionitis. Overall, 8.2% and 9.6% of children exposed and unexposed, respectively, to clinical chorioamnionitis had moderate-to-severe neurodevelopmental disorders. After multiple imputations and multivariable analysis, clinical chorioamnionitis was not associated with the occurrence of moderate-to-severe neurodevelopmental disorders (aOR, 0.9; 95% CI, 0.5-1.8). CONCLUSIONS: We did not find any association between clinical chorioamnionitis and neurodevelopmental disorders at 5 years of age in children born at <35 weeks of gestation after preterm labor or preterm premature rupture of membrane.

Early Antibiotic Exposure and Adverse Outcomes in Very Preterm Infants at Low Risk of Early-Onset Sepsis: The EPIPAGE-2 Cohort Study.

OBJECTIVE: To assess the association between early empirical antibiotics and neonatal adverse outcomes in very preterm infants without risk factors for early-onset sepsis (EOS). STUDY DESIGN: This is a secondary analysis of the EPIPAGE-2 study, a prospective national population-based cohort that included all liveborn infants at 22-31 completed weeks of gestation in France in 2011. Infants at high risk of EOS (ie, born after preterm labor or preterm premature rupture of membranes or from a mother who had clinical chorioamnionitis or had received antibiotics during the last 72 hours) were excluded. Early antibiotic exposure was defined as antibiotic therapy started at day 0 or day 1 of life, irrespective of the duration and type of antibiotics. We compared treated and untreated patients using inverse probability of treatment weighting based on estimated propensity scores. RESULTS: Among 648 very preterm infants at low risk of EOS, 173 (26.2%) had received early antibiotic treatment. Early antibiotic exposure was not associated with death or late-onset sepsis or necrotizing enterocolitis (OR, 1.04; 95% CI, 0.72-1.50); however, it was associated with higher odds of severe cerebral lesions (OR, 2.71; 95% CI, 1.25-5.86) and moderate-severe bronchopulmonary dysplasia (BPD) (OR, 2.30; 95% CI, 1.21-4.38). CONCLUSIONS: Early empirical antibiotic therapy administrated in very preterm infants at low risk of EOS was associated with a higher risk of severe cerebral lesions and moderate-severe BPD.

Volume of Neonatal Care and Survival without Disability at 2 Years in Very Preterm Infants: Results of a French National Cohort Study.

OBJECTIVES: To investigate the relation between neonatal intensive care unit (NICU) volume and survival, and neuromotor and sensory disabilities at 2 years in very preterm infants. STUDY DESIGN: The EPIPAGE-2 (Etude Epidémiologique sur les Petits Âges Gestationnels-2) national prospective population-based cohort study was used to include 2447 babies born alive in 66 level III hospitals between 24 and 30 completed weeks of gestation in 2011. The outcome was survival without disabilities (levels 2-5 of the Gross Motor Function Classification System for cerebral palsy with or without unilateral or bilateral blindness or deafness). Units were grouped in quartiles according to volume, defined as the annual admissions of very preterm babies. Multivariate logistic regression analyses with population average models were used. RESULTS: Survival at discharge was lower in hospitals with lower volumes of neonatal activity (aOR 0.55, 95% CI 0.33-0.91). Survival without neuromotor and sensory disabilities at 2 years increased with hospital volume, from 75% to 80.7% in the highest volume units. After adjustment for gestational age, small for gestational age, sex, maternal age, infertility treatment, multiple pregnancy, principal cause of prematurity, parental socioeconomic status, and mother's country of birth, survival without neuromotor or sensory disabilities was significantly lower in hospitals with a lower volume of neonatal activity (aOR 0.60, 95% CI 0.38-0.95) than in the highest quartile hospitals. CONCLUSIONS: These results suggest that lower neonatal intensive care unit volume is associated with lower survival without an increase in disabilities at 2 years. These results could be useful to generate improvements of perinatal regionalization.

Association of Language Skills with Other Developmental Domains in Extremely, Very, and Moderately Preterm Children: EPIPAGE 2 Cohort Study.

OBJECTIVE: To analyze language skills in children born at 24-34 weeks of gestation at 2 years of corrected age and the association between language and other developmental domains. STUDY DESIGN: We included 2424 children (64% of the eligible population) from the French population-based EPIPAGE 2 cohort study. At 2 years' corrected age, children were screened with the French short version of the MacArthur-Bates Communication Developmental Inventories and the Ages and Stages Questionnaire completed by parents. RESULTS: Small lexicon size, <10th percentile of the calibration sample (ie, 28 words in a list of 100) was observed in 135 of 300 children (45%) born at 23-26 weeks, 484 of 1513 (32%) born at 27-31 weeks, and 165 of 611 (27%) born at 32-34 weeks of gestation. Small lexicon size was associated with 2 other language measures: word combination use and the Ages and Stages Questionnaire communication domain score. It was also significantly associated with the Ages and Stages Questionnaire score below the threshold in the other developmental domains (gross motor function, fine motor function, problem solving skills, and personal social skills) for all gestational age groups, after adjustment for potential confounders. Overall, 46% of children with a small lexicon size had ≥1 of these domains below the threshold, as compared with only 22% of children without a small lexicon size. CONCLUSIONS: These results highlight the usefulness of the MacArthur-Bates Communication Developmental Inventories in preterm children, especially those who do not participate in specialized follow-up. A small lexicon size points to developmental difficulties in language and increased risk for other developmental and neurobehavioral functions.

Neonatal Outcomes in Extremely Preterm Newborns Admitted to Intensive Care after No Active Antenatal Management: A Population-Based Cohort Study.

OBJECTIVE: To evaluate the association between active antenatal management and neonatal outcomes in extremely preterm newborns admitted to a neonatal intensive care unit (NICU). STUDY DESIGN: This population-based cohort study was conducted in 25 regions of France. Infants born in 2011 between 220/7 and 266/7 weeks of gestation and admitted to a NICU were included. Infants with lethal congenital malformations or death in the delivery room were excluded. A multilevel multivariable analysis was performed, accounting for clustering by mother (multiple pregnancies) and hospital plus individual characteristics, to estimate the association between the main exposure of no active antenatal management (not receiving antenatal corticosteroids, magnesium sulfate, or cesarean delivery for fetal indications) and a composite outcome of death or severe neonatal morbidity (including severe forms of brain or lung injury, retinopathy of prematurity, and necrotizing enterocolitis). RESULTS: Among 3046 extremely preterm births, 783 infants were admitted to a NICU. Of these, 138 (18%) did not receive active antenatal management. The risk of death or severe morbidity was significantly higher for infants without active antenatal management (crude OR, 2.60; 95% CI, 1.44-4.66). This finding persisted after adjustment for gestational age (OR, 2.08; 95% CI, 1.19-3.62) and all confounding factors (OR, 1.86; 95% CI, 1.09-3.20). CONCLUSIONS: The increased risk of severe neonatal outcomes for extremely preterm babies admitted to a NICU without optimal antenatal management should be considered in individual-level decision making and in the development of professional guidelines for the management of extremely preterm births.

Impact of Latency Duration on the Prognosis of Preterm Infants after Preterm Premature Rupture of Membranes at 24 to 32 Weeks' Gestation: A National Population-Based Cohort Study.

OBJECTIVE: To assess the impact of latency duration on survival, survival without severe morbidity, and early-onset sepsis in infants born after preterm premature rupture of membranes (PPROM) at 24-32 weeks' gestation. STUDY DESIGN: This study was based on the prospective national population-based Etude Épidémiologique sur les Petits Ȃges Gestationnels 2 cohort of preterm births and included 702 singletons delivered in France after PPROM at 24-32 weeks' gestation. Latency duration was defined as the time from spontaneous rupture of membranes to delivery, divided into 4 periods (12 hours to 2 days [reference], 3-7 days, 8-14 days, and >14 days). Multivariable logistic regression was used to assess the relationship between latency duration and survival, survival without severe morbidity at discharge, or early-onset sepsis. RESULTS: Latency duration ranged from 12 hours to 2 days (18%), 3-7 days (38%), 8-14 days (24%), and >14 days (20%). Rates of survival, survival without severe morbidity, and early-onset sepsis were 93.5% (95% CI 91.8-94.8), 85.4% (82.4-87.9), and 3.4% (2.0-5.7), respectively. A crude association found between prolonged latency duration and improved survival disappeared on adjusting for gestational age at birth (aOR 1.0 [reference], 1.6 [95% CI 0.8-3.2], 1.2 [0.5-2.9], and 1.0 [0.3-3.2] for latency durations from 12 hours to 2 days, 3-7 days, 8-14 days, and >14 days, respectively). Prolonged latency duration was not associated with survival without severe morbidity or early-onset sepsis. CONCLUSION: For a given gestational age at birth, prolonged latency duration after PPROM does not worsen neonatal prognosis.

Effect of Intra- and Extrauterine Growth on Long-Term Neurologic Outcomes of Very Preterm Infants.

OBJECTIVE: To determine whether extrauterine growth is associated with neurologic outcomes and if this association varies by prenatal growth profile. STUDY DESIGN: For 1493 preterms from the EPIPAGE (Étude Épidémiologique sur les Petits Âges Gestationnels [Epidemiological Study on Small Gestational Ages]) cohort, appropriate for gestational-age (AGA) was defined by birth weight >-2 SD and small for gestational-age (SGA) by birth weight ≤-2 SD. Extra-uterine growth was defined by weight gain or loss between birth and 6 months by z-score change. Growth following-the-curve (FTC) was defined as weight change -1 to +1 SD, catch-down-growth (CD) as weight loss ≥1 SD, and catch-up-growth (CU) as weight gain ≥1 SD. At 5 years, a complete medical examination (n = 1305) and cognitive evaluation with the Kauffman Assessment Battery for Children (n = 1130) were performed. Behavioral difficulties at 5 years and school performance at 8 years were assessed (n = 1095). RESULTS: Overall, 42.5% of preterms were AGA-FTC, 20.2% AGA-CD, 17.1% AGA-CU, 5.6% SGA-FTC, and 14.5% SGA-CU. Outcomes did not differ between CU and FTC preterm AGA infants. Risk of cerebral palsy was greater for AGA-CD compared with AGA-FTC (aOR 2.26 [95% CI 1.37-3.72]). As compared with children with SGA-CU, SGA-FTC children showed no significant increased risk of cognitive deficiency (aOR 1.41[0.94-2.12]) or school difficulties (aOR 1.60 [0.84-3.03]). Compared with AGA-FTC, SGA showed increased risk of cognitive deficiency (SGA-FTC aOR 2.19 [1.25-3.84]) and inattention-hyperactivity (SGA-CU aOR 1.65 [1.05-2.60]). CONCLUSION: Deficient postnatal growth was associated with poor neurologic outcome for AGA and SGA preterm infants. CU growth does not add additional benefits. Regardless of type of postnatal growth, SGA infants showed behavioral problems and cognitive deficiency.

Breastfeeding duration and cognitive development at 2 and 3 years of age in the EDEN mother-child cohort.

OBJECTIVE: To investigate the dose-response relationship between breastfeeding duration and cognitive development in French preschool children. STUDY DESIGN: In the French EDEN Mother-Child Cohort Study, we evaluated language ability with the Communicative Development Inventory (CDI) in 1387 2-year-old children and overall development with the Ages and Stages Questionnaire (ASQ) in 1199 3-year-old children. Assessments were compared between breastfed and non-breastfed children and also according to breastfeeding duration in multivariable linear models, controlling for a wide range of potential confounders. We tested departure from linearity. RESULTS: After adjustments, ever-breastfed children scored 3.7 ± 1.8 (P = .038) points higher than never-breastfed children on the CDI and 6.2 ± 1.9 (P = .001) points higher on the ASQ. Among breastfed children, exclusive and any-breastfeeding durations were positively associated with both CDI and ASQ scores. The fine motor domain of ASQ was associated with any-breastfeeding duration, and the problem solving domain with exclusive-breastfeeding duration. We did not observe significant departures from linearity. No interactions were found between the child's sex, parental education or socioeconomic status, and breastfeeding duration. CONCLUSION: Longer breastfeeding duration was associated with better cognitive and motor development in 2- and 3-year-old children and a dose-response relationship was suggested.

The influence of fetal growth reference standards on assessment of cognitive and academic outcomes of very preterm children.

OBJECTIVE: To compare 3 methods of identifying small-for-gestational-age (SGA) status in very preterm children as related to cognitive function and academic outcome. STUDY DESIGN: There were 1038 singletons in the Epipage Study, born before 33 weeks in 1997 without severe neurosensory impairment, who were classified as SGA when birth weight was below the 10th percentile according to: (1) birth weight (bw) reference: SGA(bw)/appropriate for gestational age (AGA)(bw); (2) intrauterine (intraut) reference: SGA(intraut)/AGA(intraut); and (3) intrauterine reference customized (cust) according to individual characteristics: SGA(cust)/AGA(cust). Cognitive function was assessed by the mental processing composite (MPC) score of the Kaufman Assessment Battery for Children at age 5 and academic achievement by a parental questionnaire at age 8. RESULTS: Of the children, 15% were SGA(bw), 38% were SGA(intraut), and 39% were SGA(cust). All children SGA(bw) were also SGA(intraut) and SGA(cust). MPC was <85 in 32% of children and 27% had low academic achievement. AGA(bw)/SGA(intraut) children had a significantly increased risk of MPC <85 (adjusted OR 1.74, 95% CI 1.22-2.28) or low academic achievement (adjusted OR 1.64, 95% CI 1.05-2.55) compared with AGA(bw)/AGA(intraut) children. The SGA(cust) group was only slightly different from the SGA(intraut) group. CONCLUSIONS: An intrauterine reference identified very preterm infants at risk of poor cognitive or academic outcomes better than a birth weight reference. Customization resulted in only slight modifications of the SGA group.

Impact of fetal growth restriction on mortality and morbidity in a very preterm birth cohort.

OBJECTIVE: To assess the impact of being small for gestational age (SGA) on very preterm mortality and morbidity rates by using different birthweight percentile thresholds and whether these effects differ by the cause of the preterm birth. STUDY DESIGN: The study included singletons and twins alive at onset of labor between 24 and 31 weeks of gestation without congenital anomalies from the Models of Organising Access to Intensive Care for very preterm births very preterm cohort in 10 European regions in 2003 (n = 4525). Outcomes were mortality, intraventricular hemorrhage grade III and IV, cystic periventricular leukomalacia, and bronchopulmonary dysplasia (BPD). Birthweight percentiles in 6 classes were analyzed by pregnancy complication. RESULTS: The mortality rate was higher for infants with birthweights <25th percentile when compared with the 50th to 74th percentile (adjusted odds ratio, 3.98 [95% CI, 2.79-5.67] for <10th; adjusted odds ratio, 2.15 [95% CI, 1.54-3.00] for 10th-24th). BPD declined continuously with increasing birthweight. There was no association for periventricular leukomalacia or intraventricular hemorrhage. Seventy-five percent of infants with birthweights <10th percentile were from pregnancies complicated by hypertension or indicated deliveries associated with growth restriction. However, stratifying for pregnancy complications yielded similar risk patterns. CONCLUSIONS: A 25th percentile cutoff point was a means of identifying infants at higher risk of death and a continuous measure better described risks of BPD. Lower birthweights were associated with poor outcomes regardless of pregnancy complications.

White matter damage and intraventricular hemorrhage in very preterm infants: the EPIPAGE study.

OBJECTIVE: To evaluate the prevalence of cranial ultrasound abnormalities in very preterm infants as a function of gestational age, plurality, intrauterine growth restriction, and death before discharge. STUDY DESIGN: A prospective, population-based cohort of 2667 infants born between 22 and 32 weeks of gestation in 1997 in nine regions of France, transferred to a neonatal intensive care unit, for whom at least one cranial ultrasound scan was available. RESULTS: The frequencies of white matter damage (WMD), major WMD, cystic periventricular leukomalacia (PVL), periventricular parenchymal hemorrhagic involvement, and intraventricular hemorrhage with ventricular dilatation were 21%, 8%, 5%, 3%, and 3%, respectively. The risk of WMD increased with decreasing gestational age. Mean age at diagnosis of cystic PVL was older for the most premature infants. Intraventricular hemorrhage with ventricular dilatation was associated with a higher risk of cystic PVL. Intrauterine growth restriction was not associated with a lower prevalence of cystic PVL. CONCLUSION: The frequency of WMD is high in very preterm babies and is strongly related to gestational age. The incidence of cystic PVL did not differ between babies with intrauterine growth restriction and babies who were appropriate for gestational age.