BACKGROUND: Neonatal encephalopathy (NE) is a major predictor of death and long-term neurologic disability, but there are few studies of antecedents of NE.
OBJECTIVES: To identify antecedents in a large registry of infants who had NE.
METHODS: This was a maternal and infant record review of 4165 singleton neonates, gestational age of ≥36 weeks, meeting criteria for inclusion in the Vermont Oxford Network Neonatal Encephalopathy Registry.
RESULTS: Clinically recognized seizures were the most prevalent condition (60%); 49% had a 5-minute Apgar score of ≤3 and 18% had a reduced level of consciousness. An abnormal maternal or fetal condition predated labor in 46%; maternal hypertension (16%) or small for gestational age (16%) were the most frequent risk factors. In 8%, birth defects were identified. The most prevalent birth complication was elevated maternal temperature in labor of ≥37.5°C in 27% of mothers with documented temperatures compared with 2% to 3.2% in controls in population-based studies. Clinical chorioamnionitis, prolonged membrane rupture, and maternal hypothyroidism exceeded rates in published controls. Acute asphyxial indicators were reported in 15% (in 35% if fetal bradycardia included) and inflammatory indicators in 24%. Almost one-half had neither asphyxial nor inflammatory indicators. Although most infants with NE were observably ill since the first minutes of life, only 54% of placentas were submitted for examination.
CONCLUSIONS: Clinically recognized asphyxial birth events, indicators of intrauterine exposure to inflammation, fetal growth restriction, and birth defects were each observed in term infants with NE, but much of NE in this large registry remained unexplained.
- CNS —
- central nervous system
- CP —
- cerebral palsy
- NE —
- neonatal encephalopathy
- NER —
- Neonatal Encephalopathy Registry
- SGA —
- small for gestational age
- VON —
- Vermont Oxford Network
What’s Known on This Subject:
Most term and late preterm infants with neonatal encephalopathy have not had recognized asphyxial birth events. Several nonasphyxial risk factors for neonatal encephalopathy have been identified in previous studies.
What This Study Adds:
In a large sample, we confirm the association of several nonasphyxial factors with neonatal encephalopathy, including markers of intrauterine exposure to infection or inflammation, intrauterine fetal growth restriction, and birth defects. We identify steps that would improve studies of neonatal encephalopathy.
Neonatal encephalopathy (NE) is a syndrome of neurologic dysfunction, often accompanied by seizures, presenting in the early hours of life in term and late preterm infants. Mortality in NE is substantial.1 Two-thirds of cerebral palsy (CP) arises in infants born at or after 35 weeks’ gestational age,2 with ∼30% of CP occurring in survivors of NE.3
A majority of births of mature infants occur in facilities without research capabilities. In the past, relatively decentralized care in ill term infants caused our knowledge of neonatal neurologic disorders and their treatment in term infants to lag behind that of our knowledge regarding very preterm infants. With the potential for hypothermia and newer interventions to improve outcome, many more neurologically ill term neonates are now cared for in specialized facilities that provide opportunities to characterize the clinical features of illnesses in such infants, to examine patterns of perinatal care, and to expand our understanding of causes, natural history, and optimal treatment.
Some infants with NE have clinically recognized asphyxial (“sentinel”) events at birth. Controlled studies of antecedents of NE in representative populations have found that most infants who have NE did not have recognized asphyxial birth events, however.4–9 Such studies and others related to them4,10,11 have identified some additional unsuspected risk factors, pointing to a broader set of causal factors than previously recognized. Still, known risk factors for NE, singly or together, do not account for most cases of NE. Our lack of knowledge about causes of NE seriously impairs the progress of efforts to prevent NE and to treat it optimally.
The Neonatal Encephalopathy Registry (NER), a project of the Vermont Oxford Network (VON),12 gathers data on infants with NE who were at least 36 weeks’ gestational age or who received hypothermia therapy. One of the registry’s strengths is its use to examine patterns of perinatal care and to identify changes in those patterns of care. Although the registry is not a representative population and does not have healthy controls, it provides an opportunity to examine selected clinical antecedents of NE. We describe here the frequency with which recognized antecedents of NE occurred in a large sample of encephalopathic term newborns. These findings have implications for future studies of the etiology of NE.
Eligibility and Enrollment
VON maintains 2 registries of infants admitted to NICUSs. Any infant born or transferred into a VON member center at ≥36 weeks’ gestation who displayed evidence of NE within 3 days of birth was eligible to be enrolled in the NER. NE was defined as presence of seizures and/or altered consciousness (eg, stupor, coma). To capture all infants potentially affected by NE, infants with a 5-minute Apgar score of ≤3 or who received neuromuscular blockade extending through the first 72 hours of life were eligible. Regardless of neurologic status or gestational age, any infant who received hypothermia therapy was eligible. Infants born with central nervous system (CNS) birth defects were excluded from the NER. Because multiple gestations pose special risks for adverse neurologic outcome, only singleton infants were examined in this study.
The registry did not require any interventions or protocols for treatment and only de-identified data were submitted, precluding the necessity for informed consent. The institutional review board at the University of Vermont and the institutional review boards at each participating hospital reviewed and approved registry participation.
The NER database collected information on obstetric and prenatal history, neurologic indicators, neuroimaging, diagnoses, hypothermia therapy, and discharge status. Acute asphyxial events of birth (sentinel events) were defined as perinatal events capable of interrupting oxygen supply or blood flow to the fetus, such as antepartum hemorrhage including placental abruption, uterine rupture, cord prolapse, tight nuchal cord, or maternal shock or death. Fetal bradycardia was included as a possible asphyxial indicator. Antepartum hemorrhage, cord prolapse, and uterine rupture were routinely recorded. Abruption, tight nuchal cord, and maternal shock or death were not systematically reported but were included as write-ins for other birth traumas. Inflammatory factors were maternal fever in labor of ≥37.5°C, a clinical diagnosis of chorioamnionitis, fetal tachycardia, prolonged rupture of membranes, early bacterial infection in the infant, or toxoplasmosis, other infections, rubella, cytomegalovirus infection, and herpes simplex infection in the infant. Small for gestational age (SGA) was defined as birth weight below the 10th percentile within categories of gender, race, and multiple gestation based on smoothed curves from the US Natality data set, 2001 and 2002.13 Birth defect included congenital heart defects, gastrointestinal defects, genitourinary defects, chromosomal abnormalities, pulmonary abnormalities, and other defects from the VON Birth Defects Codes List.14 Other antecedents, including descriptors of maternal and infant conditions, were derived from the VON NER database.15
The tables include the number of infants and the number of cases, the unadjusted percentages for categorical measures, and the mean ± SD for continuous measures. The χ2 test was used to evaluate the association between SGA and asphyxial events, and SGA and inflammatory events. All analyses were conducted by using SAS version 9.3 (SAS Institute, Inc, Cary, NC).
Between 2006 and 2010, a total of 4165 singleton births were registered in the NER. Clinically recognized seizures, present in 60%, were the most commonly identified criterion for eligibility (Table 1). Reduced level of consciousness was reported in only 18% of infants considered to have NE and was the sole positive criterion in <1%. In 38%, hypothermia was initiated before admission to the NER. Demographic and other characteristics of these mothers and their infants are provided elsewhere in reports that include multiple births.16
Maternal conditions predating labor that might be relevant to risk of NE were maternal hypertension, diabetes, hypothyroidism, lack of prenatal care, and assisted reproduction (Table 2). At least 1 of these maternal conditions was noted in 27% of infants. Hypertension (16%) and diabetes (10%) were reported most frequently. Maternal hypothyroidism was reported in 2.5%.
Of conditions intrinsic to the infant, SGA was the most frequently reported, present in 16% (Table 2). Non-CNS birth defects were noted in the newborn period in 8%. One-quarter of infants had ≥1 of these conditions predating the onset of labor.
Mean ± SD birth weight of infants who had NE was 3309 ± 616 g, and mean gestational age was 38.7 ± 1.6 weeks (Table 3). Birth was by spontaneous vaginal delivery in 34% of infants. Surgical delivery, performed in two-thirds of these infants, was without labor in 19% and after failure of vacuum or forceps in 2%. Fetal heart rate monitoring was considered nonreassuring in 60%, and 5% experienced skull or limb fracture or other birth injury (excluding cephalohematoma). Cord blood was sampled in 53% of infants. Of those, 54% had pH levels <7.09 and 46% had a cord blood base deficit >12. Slightly more than half (54%) of placentas were sent for pathologic examination.
In infants with NE, 15% had at least 1 sentinel asphyxial event. The most common was antepartum hemorrhage, including placental abruption (11%) (Table 4). Fetal bradycardia of unknown time of onset, severity, and duration was recorded in 34% of infants. If fetal bradycardia was included as an asphyxial indicator, 35% of infants with NE had at least 1 such indicator. Infants who were SGA (n = 672) were no more likely to experience sentinel asphyxial events than those who were not growth restricted (n = 3492); 15% of each group experienced at least 1 such event.
Maternal temperature in labor was recorded for 37% of infants. Of those recorded, 27% of mothers had a temperature of ≥37.5°C. Overall, 24% of neonates had at least 1 inflammatory indicator. Infants who were SGA were significantly less likely to experience inflammatory indicators than infants who were not growth restricted (19% vs 24%; χ2 (1, 7.92), P < .005).
About one-third of these term infants who had NE had sentinel asphyxial events only (13%), inflammatory indicators only (21%), or both (2%). When bradycardia was included, more than one-half had asphyxial indicators only (29%), inflammatory indicators only (17%), or both (6%).
NE is associated with an increased risk of death and is part of an important pathway to long-term neurologic disability.1–3 Infants who develop CP as a result of asphyxial births regularly experience NE in the newborn period,17 as do infants with placental infarction18,19 or intrauterine exposure to inflammation.20 Sorting out the precursors of NE and NE-associated CP is an important task toward developing more effective strategies for primary prevention of these serious disorders.
Several characteristics reported to be risk factors for NE in previous studies were observed with considerable frequency in the VON NER: notably, maternal fever in labor in 27% of those recorded, fetal growth restriction in 16%, and birth defects in 8%. The NER did not include within-study controls; we therefore looked for comparisons with the values in control infants from prospective, controlled, population-based studies of NE in term infants in industrialized countries, as included in the studies of Adamson et al,4 Badawi et al,5,10 and Blume et al21 (Table 5). The severity of illness in VON NER infants, one-half of whom had 5-minute Apgar scores of ≤3 and two-thirds with neonatal seizures, was approximately comparable with encephalopathic term infants in these controlled, population-based studies.
Acute Asphyxial Indicators
Of the encephalopathic term singletons in the current study, 15% experienced a clinically recognized sentinel event such as antenatal hemorrhage (presumably, often placental abruption), uterine rupture, or cord prolapse, all of which are capable of compromising oxygen supply. The frequency of hemorrhage and cord prolapse exceeded the frequency in control populations (Table 5), but most infants who had NE did not have clinically recognized asphyxial birth events. Similarly, sentinel events were identified in a fairly small minority of infants with NE in controlled studies in representative populations: 7.9%5 and 25%.6 In a referral sample of 500 term infants with NE evaluated for therapeutic hypothermia, 48 (9%) had had a sentinel birth event.22
Fetal bradycardia (its onset, severity, and duration unspecified) was recognized in 34% of infants. In the VON NER, a majority of cord bloods tested (54%) were not severely acidotic, and 46% did not have a base deficit >12.
The VON NER did not capture all possible acute asphyxial events or markers. Some infants who had NE may have had undocumented events, such as intermittent occlusions of the umbilical cord in utero, to account for neurologic depression and acidosis. It is not known whether such events were common or rare. Ischemic occlusion, sometimes injurious, can be protective under some circumstances.23,24
Maternal Fever and Inflammation
A relatively common intrapartum complication documented in infants who have NE was intrauterine exposure to fever/inflammation. In the NER, 27% of women with data on temperatures recorded in labor had temperatures ≥37.5°C (18% of women had temperatures >38.0°C) compared with 2.2% to 3.2% in infants free of NE in population-based controls (Table 5). Maternal fever was associated with a trebling of risk of NE in both the studies of Badawi et al5 and Blume et al21; it was also associated with an adjusted odds ratio of 4.7 (95% confidence interval: 1.3–17) in a prospective cohort study.25
Modest elevation of maternal temperature in labor is, according to a substantial and consistent literature,20,26,27 robustly related to adverse outcome in term and late preterm infants in the delivery room and newborn nursery, and later. Elevated maternal temperature is associated with low Apgar scores, respiratory depression, neonatal seizures, and with CP. Experimental studies indicate potential interaction of inflammatory and asphyxial risk factors28 and of those with disorders of coagulation and other potential pathobiologic mechanisms.
Spencer et al29 found that maternal fever in labor was strongly associated with NE and more predictive of NE than abnormalities of fetal heart rate patterns. Although maternal fever in labor can be the result of dehydration and the physical exertion of labor, it is also associated with duration of labor, number of vaginal examinations during labor, and with epidural analgesia.30 Maternal fever in labor is probably often an indicator of inflammation or infection. Clinically diagnosed chorioamnionitis was documented in 11% of VON NER infants compared with 5.4% of controls in other studies.31
Markers of inflammation are consistently associated with adverse neurologic outcome in term and late preterm infants, but few studies from neonatology units have included maternal fever in labor as a predictor of prognosis for NE. The chain of associations of inflammatory placental lesions, microbiologic findings, brain lesions on neuroimaging in the neonatal period, and later neurologic disability have been demonstrated in infants born extremely preterm.32 No such studies have yet been performed in more mature infants.
Fetal Growth Restriction
In the VON NER, 16% of infants were SGA. In the major controlled study of NE, in which 16% of infants with NE and 1.2% of term infants without NE were growth restricted, growth restriction was the strongest predictor of NE examined, associated with a 30-fold increase in risk.10 In a regional study of moderate or severe NE in term infants, 17% were SGA.6 SGA has consistently been associated with risk of CP.33–35
Genetic, infectious, and nutritional factors can contribute to growth restriction, and defective placentation and disorders of the placenta also seem to be major factors. Co-occurrence of growth restriction and placental infarction was linked with CP risk in 2 studies22,23 and was linked with NE-associated CP in 1 study.23
Maternal hypertension, associated with growth retardation in many studies, was observed in 16% of the VON NER. Additional studies in human populations are needed to examine differing pathways to growth restriction and to elucidate the relationship of these factors with NE and NE-associated CP.
It is sometimes assumed that infants SGA at term are more vulnerable to asphyxial events of birth than well-grown infants. We found no evidence of such an effect: Among these encephalopathic infants, those who were SGA had no more asphyxial indicators than those who were not growth restricted.
Infants with non-CNS birth defects recognized in the newborn period comprised 8% of the VON NER and 11.1% of infants with NE in an uncontrolled population-based study.7 In 1 controlled population-based study, none of 89 controls but 5 of 89 cases (5.6%) had birth defects,4 whereas in another population-based study, these numbers were 4% and 10%, respectively.10 Within each relevant report, birth defects were observed more frequently in infants with NE than without NE. Many significant birth defects are not detected until after the newborn period; in controlled studies that included information ascertained in the first year of life or later, case-control differences were especially large.36,37
Because the majority of the infants in the VON NER experienced neonatal seizures, it is relevant to note that in the National Collaborative Perinatal Project, major CNS malformations were observed in the first year of life in 0.5% of controls and in 11.3% of infants with neonatal seizures, non-CNS malformations in 6.9% of controls and 24.7% of infants with neonatal seizures, and any malformation in 14.8% of controls and in 36.5% of infants with neonatal seizures.37 It is likely that structural malformations contribute to NE. The nature and timing in development of birth defects in children who experienced NE warrant further investigation.
Associated with heightened risk of NE in 3 previous prospective studies of NE,4,10,11 clinical hypothyroidism was observed in 3% of mothers in the VON NER. For comparison, a recent review of the literature of overt hypothyroidism during pregnancy cites its prevalence as ≤1.0% in developed countries, and in most studies as 0.2% to 0.3%.38
Strengths and Limitations of the NER
The VON NER is a rich repository of information about infants who have NE. It is a registry, however, and did not have standardized definitions, a standardized approach to diagnostic or therapeutic procedures, or a control group. Inferences drawn from the NER are limited by these factors and by the absence of controls and an incomplete range of items included as potential antecedents of NE. This study, like most studies of NE, included infants who had neonatal seizures regardless of whether they had other criteria of NE. Such infants constitute a substantial subgroup of the NER. Because there was no uniform protocol for neuroimaging, or metabolic or genetic disorders, the incidence of perinatal stroke or other etiologies in these infants with seizures is unknown.
Alteration of consciousness is a key feature of encephalopathy, but the most common indication for inclusion of an infant as having NE in the VON NER was clinically recognized neonatal seizures. Depression of consciousness was reported in only 18%, although one-half of the infants in the NER had 5-minute Apgar scores of ≤3 and other markers of serious neonatal illness. These observations suggest underrecognition of depression of consciousness in these newborn infants and a need for training of caregivers in the reliable recognition of neurologic depression and its severity. For future descriptive studies and clinical trials, characteristics such as persistently low Apgar scores (with and without marked acidosis) might be more reliably ascertainable entry criteria.39
Although most of these infants showed marked compromise in the delivery room, only about one-half of placentas were submitted for examination, a limitation shared with many other studies of NE. Failure to incorporate information from placental examination is unfortunate, as placental lesions are common in encephalopathic term infants.40,41 Findings in the placentas of ill neonates can often contribute to an understanding of the underlying pathobiology and can sometimes influence clinical management. Wintermark et al40 suggest that inflammatory placental pathology reduces the efficacy of therapeutic hypothermia in encephalopathic term newborns, a possibility suggesting that stratification in the analysis for presence, type, and severity of placental lesions should be considered in future trials of therapeutic cooling.
Observations in the NER support the importance of inflammation, aberrant fetal growth, asphyxial birth events, birth defects, and maternal thyroid disorder in NE. In this study, as in previous studies, most infants who met study criteria for NE did not have recognized asphyxial birth events. Much remains unknown about the antecedents of NE.
Given the importance of neonatal neurologic compromise on the causal pathway to long-term neurologic disability, the etiology of NE is remarkably underresearched. More studies are needed regarding the antecedents of NE to enable better etiologic diagnoses and more rapid and specific treatment.
Drs Nelson, Bingham, Horbar, Inder, Raju and Soll are members of the VON NER steering committee.
APPENDIX Hospitals Registering Infants in the VON NER, 2006–2010
- Accepted June 14, 2012.
- Address correspondence to Karin B. Nelson, MD, 5524 Charles St, Bethesda, MD 20892. E-mail:
All authors made substantial contributions to the conception and design of this study and acquisition of data. Drs Nelson and Edwards and Mr Kenny were primarily responsible for analysis and data interpretation. Drs Nelson and Edwards were responsible for drafting the article. All authors helped revise it critically for important intellectual content and give final approval of the version to be published. Each author has participated sufficiently in the work to take public responsibility for appropriate portions of the content.
FINANCIAL DISCLOSURE: Drs Horbar and Soll are employees of Vermont Oxford Network; the other authors have indicated they have no financial relationships relevant to this article to disclose.
FUNDING: No external funding.
- ↵Australian Cerebral Palsy Registry Group. Report of the Australian Cerebral Palsy Register, Birth Years 1993–2003. December 2009. Available at: http://secure.cpregister-aus.com.au/pubs/pdf/ACPR-Report.pdf. Accessed August 14, 2012
- Adamson SJ,
- Alessandri LM,
- Badawi N,
- Burton PR,
- Pemberton PJ,
- Stanley FJ
- Badawi N,
- Kurinczuk JJ,
- Keogh JM,
- et al
- Pierrat V,
- Haouari N,
- Liska A,
- Thomas D,
- Subtil D,
- Truffert P,
- Groupe d’Etudes en Epidémiologie Périnatale
- Badawi N,
- Kurinczuk JJ,
- Keogh JM,
- et al
- Ellis M,
- Manandhar N,
- Manandhar DS,
- Costello AM
- ↵National Center for Health Statistics. Natality public-use tape and CD-ROM. Hyattsville, MD: National Center for Health Statistics. Annual Products. 2001 and 2002
- Vermont Oxford Network
- Vermont Oxford Network
- ↵Pfister RH, Bingham P, Edwards EM, et al. The Vermont Oxford Neonatal Encephalopathy Registry: rationale, methods, and initial results. BMC Pediatr. 2012;12(1):84
- ↵Blair E, de Groot J, Nelson KB. Placental infarction identified by macroscopic examination and risk of cerebral palsy in infants at 35 weeks gestational age and over. Am J Obstet Gynecol. 2011;205(2):124.e1–7
- ↵Grether JK, Nelson KB. Maternal infection and cerebral palsy in infants of normal birth weight [published correction appears in JAMA. 1998;279(2):118]. JAMA. 1997;278(3):207–211
- Okereafor A,
- Allsop J,
- Counsell SJ,
- et al
- Durukan A,
- Tatlisumak T
- Rouse DJ,
- Landon M,
- Leveno KJ,
- et al.,
- National Institute of Child Health and Human Development, Maternal-Fetal Medicine Units Network
- ↵Wintermark P, Boyd T, Gregas MC, Labrecque M, Hansen A. Placental pathology in asphyxiated newborns meeting the criteria for therapeutic hypothermia. Am J Obstet Gynecol. 2010;203(6):579.e1–9
- ↵Chang T, Reyes C, Teng J, Placette J, Massaro A, Nelson KB. Neonatal encephalopathy, sentinel events, and the placenta. J Neonatal-Perinatal Med. 2012;5:1–8
- Copyright © 2012 by the American Academy of Pediatrics