Article Text

Download PDFPDF

The placenta in infants >36 weeks gestation with neonatal encephalopathy: a case control study
  1. Breda C Hayes1,
  2. Sharon Cooley2,
  3. Jennifer Donnelly2,
  4. Elaine Doherty3,
  5. Andrea Grehan1,
  6. Cathy Madigan3,
  7. Cliona McGarvey4,
  8. Siobhan Mulvany1,
  9. Stephanie Ryan5,
  10. John Gillian6,
  11. Michael P Geary2,
  12. Tom G Matthews1,2,
  13. Mary D King3
  1. 1Department of Neonatology, Rotunda Hospital, Dublin, Ireland
  2. 2Department of Obstetrics and Gynaecology, Rotunda Hospital, Dublin, Ireland
  3. 3Department of Neurology and Clinical Neurophysiology, The Children's University Hospital, Dublin, Ireland
  4. 4Irish National Sudden Infant Death Register, The Childrens University Hospital, Dublin, Ireland
  5. 5Department of Radiology, Rotunda Hospital, Dublin, Ireland
  6. 6Department of Pathology, Rotunda Hospital, Dublin, Ireland
  1. Correspondence to Breda C Hayes, Department of Pediatrics, Rotunda Hospital, Dublin 1, Ireland; breda.hayes{at} or bredahayes{at}


Objective To determine placental characteristics associated with neonatal encephalopathy (NE) and correlate these with short- and long-term neurodevelopmental outcome.

Design Case/control study.

Setting Neonatal Intensive Care Unit, Rotunda Hospital, Dublin, Ireland.

Patients Newborns ≥36 weeks gestation, with NE (cases). Healthy term newborns (controls).

Interventions Placental pathology was obtained from the official placental report. Brain MRI was blindly reviewed. Children were assessed using a variety of standardised assessments. Data were analysed using multinomial logistic regression analysis.

Main outcome measures RRR for grade of encephalopathy. OR for neurodevelopmental outcome.

Results Placental reports were available on 141 cases (76 grade 1; 46 grade 2; 19 grade 3) and 309 control infants. Meconium phagocytosis, haemorrhage, raised placental to birth weight ratio and/or markers of infection/inflammation were independently associated with NE and showed a synergistic effect, when combined, for short- and long-term impairments.

Conclusions Evaluation of the mechanisms leading to the placental characteristics identified may help to characterise the causal pathway of NE.

View Full Text

Statistics from

What is already known on this topic

  • The presence of a placenta reported as abnormal is associated with a doubling in risk of neonatal encephalopathy (NE).

  • Few controlled studies of the placenta are reported.

What this study adds

  • High placental to birth weight ratio is a marker for infants at risk of poor tolerance of the labour process.

  • The presence of haemorrhage, meconium phagocytosis, markers of infection/inflammation and increased nRBCs on placental pathology is associated with an increased risk of NE.

  • A synergistic relationship between placental lesions exists for both grade of encephalopathy and longer term neurodevelopmental outcome.


The placenta is an important yet often ignored source of information. In 1892, Ballantyne1 wrote ‘…..a description of a foetal malady unless accompanied by a notice of the placental condition is incomplete…’ More recently Badawi et al found that the presence of a placenta reported as abnormal was associated with a doubling in risk of neonatal encephalopathy (NE).2

Previous studies have identified placental lesions significantly associated with neurodisability independent of clinical indicators of birth asphyxia.3,,5 However, few controlled studies of the placenta in NE are reported.

The objective of this study was to determine placental characteristics associated with NE and to correlate these with grade of encephalopathy and neurodevelopmental outcome.


Neonatal Intensive Care Unit (NICU), Rotunda Hospital, Dublin, Ireland.


Case selection

Two hundred and forty-five newborns were admitted with NE within the first 48 h after birth due to presumed hypoxia–ischaemia from January 2001 to December 2008. Cases born between January 2001 and July 2005 (98 infants) were identified retrospectively. Those admitted between July 2005 and December 2008 (147 infants) were identified prospectively. NE was graded as grade 1, 2 or 3, according to Sarnat and Sarnat staging.6 Maternal records of eight case infants could not be located within the timeframe of data collection. Placental examination was performed in 141 (76 grade 1; 46 grade 2; 19 grade 3) of the remaining 237 infants. Exclusion criteria for cases were outborn infants, gestation <36 completed weeks, major congenital anomaly or any primary cause for encephalopathy other than hypoxia–ischaemia. An Apgar score of ≤5 at 10 min and/or continued need for resuscitation, including endotracheal or mask ventilation, at 10 min after birth and/or acidosis within 60 min of birth (defined as any occurrence of umbilical cord, arterial or capillary pH ≤7.10) was present in 93/155 (60%) of newborns with grade 1 NE; 47/61 (77%) of newborns with grade 2 NE and 21/21(100%) of newborns with grade 3 NE. Supplementary table S1 outlines important antenatal, labour and delivery characteristics of enrolled cases and controls.

Control selection

As part of a research project (‘The impact of Acquired Thrombophilia on maternal and fetal wellbeing in a low-risk primigravid population’), placental histology of low-risk pregnancies was studied.7 ,8 Inclusion criteria for this study were Caucasian primigravida women, 18–40 years who booked at <20 weeks gestation, with a singleton pregnancy. Exclusion criteria were any fetal anomaly, prior pregnancy loss, presence of a chronic medical condition requiring ongoing specialist review, personal/family history of thrombosis, current use of prescription/recreational drugs, a positive HIV/hepatitis B virus/hepatitis C virus or venereal disease reference lab test or poor understanding of English.

Placentae from a subset of these pregnancies were chosen as controls. This subset was randomly chosen from newborns with a gestation >36 weeks without NICU admission and a cord pH of ≥7.0.

Data acquisition

Maternal and neonatal histories were obtained from medical chart review. Placental to birth weight ratio (PW/BW) (ie, placental weight in grams divided by infant birth weight in grams) was calculated. PW/BW percentiles were determined based on infant gender and gestation using percentile curves.9

Examination of placentae

Both case and control placentae were prepared for examination in the same manner. All examinations were performed by a single pathologist (J G) who reported on the placentae in the newborn period without knowledge of the clinical outcome. For the purpose of this study, all information was obtained from the placental pathology report as dictated at the time of placental examination. For a detailed description of placental examination please refer appendix A.


S R (consultant paediatric radiologist) and M D K (consultant paediatric neurologist) blindly reviewed brain MRI. These images were scored as described by Barkovich et al.10 Some infants had injury which did not fit into the above scoring system and were grouped together as ‘other’ pattern of injury.

Neurodevelopmental assessments

All case infants were invited to attend for detailed neurodevelopmental assessments. Children with cerebral palsy were assessed using the Gross Motor Functional Classification Scale11 and manual ability classification system.12 Children >42 months of age without cerebral palsy were assessed using the NEPSY 2 (A Developmental NEuroPSYchological Assessment),13 Movement Assessment Battery for Children 2 (MABC2),14 Behaviour Rating Inventory of Executive Function (BRIEF)15 and Child Behaviour Checklist (CBCL).16

Children <42 months of age without cerebral palsy were assessed using the Bayley Scales of Infant/Toddler Development 3 (BSITD3).17

Statistical methods

A number of statistical methods were used (birth weight – analysis of variance for groups of unequal variance; gestation – t test for two groups of equal variance and one way analysis of variance of multiple groups; placental weight – Mann–Whitney U test for non-parametric data; PW/BW t test of two means (equal variance). RRR of variables for grade of encephalopathy was performed using multinomial logistic regression analysis. Criteria for inclusion in the multivariate model were achieving significance at 0.05 level in the univariate analysis. Variables for gestation and PW/BW were included as continuous variables. Each factor was adjusted for all other factors within the final model. Logistic regression analysis was used to determine the relationship between the number of placental variables present and outcome.


Ethical approval was obtained from the research ethics committee at The Rotunda Hospital.


Placental reports were available on 141 cases (76 grade 1; 46 grade 2; 19 grade 3) and 309 control infants. The mean gestational age of cases was 39.41±1.66 weeks and of controls was 39.74±1.48 weeks. Twelve case infants (8.5%) and seven control infants (2.3%) were between 36 and 37 weeks gestation (p<0.05). Eighty-one case infants (27.7%) and 108 control infants (35.0%) were ≥41 weeks gestation. Eighty-nine case infants (57.4%) and 147 control infants (49.0%) were male. Thirty-one (22%) case infants (16 with grade 1 NE; 9 with grade 2 NE and 6 with grade 3 NE) had a documented sentinel event (defined as presence of uterine rupture, placental abruption or shoulder dystocia). Placental weights were available on 138 case infants (75 grade 1; 45 grade 2; 18 grade 3 infants) and 307 control infants. The median placental weight of cases was significantly higher than in controls (p<0.05). PW/BW was a significant factor for grade 1 and 2 encephalopathy but not for grade 3 (supplementary table S1). When PW/BW percentile (adjusted for gestation and gender) was calculated, cases, regardless of grade of NE, had a tendency towards a higher PW/BW percentile which was more evident in infants with a history of grade 1 or 2 NE. The presence of villous dysmaturity, accelerated villous maturation, perivillous fibrin deposition or thrombosis was not significant on univariate analysis.

Multinomial regression analysis

Evidence of haemorrhage (including all locations) was significant across all grades of encephalopathy. Higher PW/BW was a risk factor for grade 1 and 2 NE. Maternal age ≥30 was no longer significant when analysis was adjusted for parity (table 1 and supplementary table S2). Although a history of maternal diabetes, hypertension and pre-eclampsia were statistically significant, associated CIs were wide. The presence of infection and/or inflammation was significant in infants with a history of grade 1 NE. Replacing the variable ‘markers of infection/inflammation’ with the individual components used to define this variable showed that chorioamnionitis, vasculitis and funisitis were significant in infants with grade 1 NE. For infants with grade 2 NE only villitis was significant and with grade 3 NE only funisitis remained significant (table 2). The presence of increased nRBC had a RR of 10.18 (95% CI 3.95 to 26.15) p<0.001. Intervillous haemorrhage appeared to be the location most strongly associated with subsequent development of NE (supplementary table S3). Frequency of haemorrhage (p=0.03) and nucleated red blood cells (p=0.037) differed significantly between newborns with versus without a sentinel event. The frequency of villitis (p=0.118), funisitis (p=0.625), meconium phagocytosis (p=0.286), accelerated villous maturation (p=0.105) and villous dysmaturity (p=0.807) did not differ significantly between these groups. Excluding infants with a sentinel event did not significantly affect the main findings on multinomial logistic regression (supplementary table S4).

Table 1

Multinomial logistic regression (parity=primigravid included in model)

Table 2

Multinomial logistic regression – individual markers of infection/inflammation

The odds of NE increased as the numbers of placental lesions present increased (table 3).

Table 3

Number of placental lesions* present and logistic regression analysis of case control data

Placental findings and pattern of injury on neonatal brain MRI

Fifty-six infants had both placental examination and brain MRI in the neonatal period. There was no correlation between placental histology and pattern of injury seen on brain MRI. This may reflect the small sample size with each category of injury (table 4).

Table 4

Pattern of injury on neonatal brain MRI and placental characteristics

Placental findings and outcome

A total of 15/141 (10.6%) infants died in the neonatal period. A total of 95/126 (75.4%) remaining cases attended for follow-up assessments. Children lost to follow-up were more likely to have a history of grade 1 NE and be discharged from clinical care. Cerebral palsy was diagnosed in 13/95 (13.7%). All children diagnosed with cerebral palsy were greater than 2 years at the time of diagnosis. In the remaining 82 children, 30 were <42 months of age and 52>42 months of age.

In children assessed with BSITD3, age at the time of assessment ranged from 15 to 41 months (mean 28.4 months, median 28 months). In children assessed with NEPSY2, MABC2, CBCL and BRIEF age at the time of assessment ranged from 3 years 6 months to 8 years 9 months. In the 82 children without cerebral palsy, 45 children (54.9%) scored above or within the normal range on all assessments. Scores from the remaining 37 children (45.1%) fell below the normal range on at least one assessment. The presence of increased numbers of placental lesions (ie, haemorrhage (all locations), infection/inflammation, nRBC, infarctions, thrombosis, meconium phagocytosis, accelerated villous maturation or villous dysmaturity) was associated with a poorer outcome (table 5). This remained significant after adjustment for grade of encephalopathy. In the presence of four or more placental lesions, 11/30 (36.7%) infants on whom outcome is known died or developed cerebral palsy.

Table 5

Number of placental lesions and long-term follow-up


This study identified a number of placental lesions which were independently associated with an increased risk for NE. Furthermore, a synergistic effect was seen between these lesions and the development of encephalopathy and long-term neurodevelopmental outcome.

In accordance with previous studies,3 ,4 ,18 haemorrhage and/or meconium were significantly associated with NE. Meconium causes vasoconstriction in isolated umbilical venous tissue.19 ,20 Meconium-induced vascular necrosis has been associated with fetal distress, lower Apgar scores and arterial pH less than 7.19.20 Intervillous haemorrhage was the commonest site of haemorrhage identified in cases and was not seen in control placentae. Previous reports have described intervillous haemorrhage as one of the strongest placental findings associated with perinatal asphyxia.18 Intervillous haematomas have been directly linked to a reduction in placental blood flow, placental insufficiency and fetal growth restriction.21 In this study, a trend towards significance remained if intervillous haemorrhage and presence of intervillous thrombi were combined.

A recent study identified PW/BW as a predictor of short-term health risks for newborns.22 In the current study, the finding of lower infant birthweight with a heavier placenta was significantly associated with NE, suggesting that a higher PW/BW is a marker for infants at risk of poor tolerance of the labour process. Heavy placentae were also noted in a recent study by Chang et al reviewing placental reports of newborns with moderate to severe NE admitted for therapeutic hypothermia.23 The mechanism underlying this is unclear; even after adjustment for markers of infection/inflammation PW/BW remained significant. Oedema due to poor placental perfusion may be a factor but villous oedema was noted in only five case infants. The incidence of villous dysmaturity, which can be associated with reduced placental efficiency, did not differ significantly between cases and controls. As the final analysis was adjusted for gestational age, the higher proportion of infants with a gestational age of 36–37 weeks with grade 3 NE does not explain why PW/BW was not significant in this category. This may reflect small sample size as there are only 18 infants in this group. Alternatively, it may be that placental hypertrophy in response to adversity is protective against grade 3 NE.

Early reports on PW/BW found an association between a PW/BW >0.1 (10%) and a significant risk of an Apgar score <6.24 Literature reported mean PW/BW values range from 0.1425 to 0.17.26 The median PW/BW found in this study are within this range. More recently, Almog et al9 published PW/BW percentile curves for a Canadian population and found higher mean PW/BW: 0.2 (girls) and 0.19 (boys) at 40 weeks gestation. Thus the population in that study had either larger placentae or lower birthweight or a combination of both. Despite differences in placental weight, when PW/BW and short-term perinatal outcomes were correlated for that population, a higher PW/BW was predictive of poorer outcome22 similar to our findings. Fetal growth restriction is a well-recognised risk factor for term asphyxia. High PW/BW may be a marker for infants with fetal growth restriction which is not apparent when birth weight is examined in isolation.27

A recent study examined the placental characteristics of infants eligible for hypothermic cooling and found 48% (11/23) had a placental weight <10th percentile.28 However that study did not comment on PW/BW and control placentae were not examined.

Edwards et al26 found that increased PW/BW in rat fetuses was associated with lower activity of placental 11 β-hydroxysteroid dehydrogenase (11β-OHSD). 11β-OHSD converts glucocorticoids to inactive products. Fetuses with low birthweight and high-placental weight had the lowest 11β-OHSD activity and presumably the greatest exposure to maternal glucocorticoids.26 Therefore, raised PW/BW may reflect alterations in placental enzymatic activities which affect the fetal maternal environment.

The role of perinatal infection and/or inflammation is of considerable interest.29 ,30 Markers of infection/inflammation were more strongly associated with lower grade encephalopathy but when the individual components of the dummy variable ‘infection/inflammation’ were analysed, villitis and funisitis (indicative of long-standing infection/inflammation) were significant risk factors for higher grade NE. Therefore, increased duration of infectious/inflammatory exposure may increase the risk of higher grade of encephalopathy. Chang et al23 report a significantly lower frequency of inflammatory lesions in newborns with a history of a sentinel event. In the current study, the only significant difference between newborns with and those without a sentinel event was the frequency of haemorrhage and nRBCs The broad definition of an acute sentinel event, (including tight nuchal cords) in the study by Chang et al may explain the differences observed.

Elevated circulating fetal nRBC is a marker of intrauterine stress.3 A significant increase in circulating nRBC has been noted between 6 and 12 h after the onset of hypoxia suggesting that irrespective of events at or around the time of delivery a finding of significantly increased nRBC in the fetoplacental circulation of cases indicates the presence of antenatal stress of at least 6–12 h duration.3

A synergistic relationship between placental lesions has previously been reported.31 The present study confirms this finding. However, there was no relationship between placental characteristics and pattern of brain injury seen on MRI. A recent study32 examined placental characteristics and MRI findings in 23 newborns and noted that the presence of acute chorioamnionitis associated with fetal vasculitis and chorionic plate meconium was significantly associated with brain injury on MRI, but did not look at specific patterns of brain injury. Larger studies are needed to determine if there is an association between specific patterns of brain injury and placental findings in NE.

Cases and controls were not matched for gestational age and gender, but adjustment was made for these factors and for maternal age in the final analysis. Rates of maternal diabetes, another possible confounding factor,29 ,33 were similar in cases and controls. Reliable data regarding maternal smoking habits were not available and all control placentae were from Caucasian women. However, despite the fact that both smoking and ethnicity have a significant affect on placental weight,34 previous studies have shown that smoking and ethnicity do not appear to affect the mean (SD) PW/BW.25 ,35 Smoking and ethnicity36 may have similar effects on both the infant and placental weight thus leaving the PW/BW relatively unchanged. Without total population data specific to this population, any bias in socio-economic status or other characteristics of controls introduced by the requirement for antenatal care prior to 20 weeks cannot be identified. Also the possibility of negative bias of ascertainment in the control group exists, as lesions (eg, intervillous haemorrhage) may not be mentioned in the final report of otherwise normal placentae.

Placental characteristics are independently associated with NE and show a synergistic effect, when combined, for short- and long-term impairments highlighting the importance of requesting placental examination in any newborn with NE. In addition, as much research is focused on the pathogenesis of NE in order to target inflammatory and/or brain repair pathways,37 further evaluation of the placental characteristics associated with NE (increased PW/BW, villitis, funisitis and increased nucleated red blood cells) may help in the identification of novel therapies for the treatment of NE.


The authors acknowledge the Friends of The Rotunda for their support and also the children and their families who participated in this project.


View Abstract

Supplementary materials

  • Supplementary Data

    This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

    Files in this Data Supplement:

    • Web Only Data - This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
    • Web Only Data - This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
    • Web Only Data - This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
    • Web Only Data - This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
    • Web Only Data - This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.


  • Funding Funding for this study was provided by the Friends of The Rotunda. The Friends of The Rotunda is an official fundraising arm and registered Charity (CHY240) of The Rotunda Hospital.

  • Competing interests None.

  • Ethics approval Rotunda Research ethics Committee.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Data sharing statement Data sharing from the corresponding author at breda.hayes{at}

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.