European Journal of Paediatric Neurology 27 (2020) 78e85

Contents lists available at ScienceDirect

European Journal of Paediatric Neurology

Minor neurological signs and behavioural function at age 2 years in

neonatal hypoxic ischaemic encephalopathy (HIE)
Caroline J. Edmonds a, b, Suzannah K. Helps c, Denise Hart d, Anna Zatorska b, d,
Neelam Gupta d, Rina Cianfaglione b, Brigitte Vollmer b, d, *
a
School of Psychology, University of East London, London, UK
b
Clinical and Experimental Sciences, Faculty of Medicine, University of Southampton, Southampton, UK
c
School of Health Sciences and Social Work, University of Portsmouth, Portsmouth, UK
d
Paediatric and Neonatal Neurology, Southampton Children's Hospital, University Hospital Southampton NHS Foundation Trust, Southampton, UK

a r t i c l e i n f o a b s t r a c t

Article history: Background: Neurodevelopmental follow-up in Neonatal Hypoxic Ischaemic Encephalopathy (HIE)
Received 20 October 2019 typically focusses on major neuromotor (cerebral palsy, CP) and severe cognitive impairment. Outcomes
Received in revised form in those without major neuromotor impairment are less well explored.
15 March 2020
Objectives: To examine behavioural, cognitive and neurological outcomes after neonatal HIE, in a clinical
Accepted 7 April 2020
cohort of children without CP, at age 2 years.
Methods: Clinical routine outcome data from children admitted to a tertiary centre with neonatal HIE for
Keywords:
hypothermia treatment between 05/08/09e30/05/2016. Children were assessed for neuromotor status e
Neonatal hypoxic ischaemic
encephalopathy (HIE)
particularly minor neurological signs (MNS), with Bayley Scales of Infant and Toddler Development III
Therapeutic hypothermia (Bayley III) or Ages and Stages Questionnaire-3 (ASQ), Child Behavior Checklist 1.5e5 (CBCL), Quantitative
Minor neurological signs Checklist for Autism in Toddlers (Q-CHAT).
Neurodevelopmental Results: Of 107 children, 75.5% had normal neurology, 12.1% CP, 12.1% MNS. Children with CP were
Neuromotor excluded from analyses. For those without CP, Bayley-III scores were in the average range for the ma-
jority; mild cognitive delay observed in 5%, 4.2% language, 1.3% motor development; severe delay in 1.3%
for cognitive, 4.2% for language. More than in the normative population scored in clinical ranges for CBCL
externalising, sleep, and other problems. No significant difference was seen for Q-CHAT. Children with
MNS were significantly more likely to have impaired Bayley-III scores, parent-reported internalising,
sleep, and other problems.
Conclusions: In this clinical cohort, the majority of children had favourable outcome at 2 years. However,
children with MNS were at risk for cognitive and behavioural difficulties and will benefit from enhanced
clinical follow-up and support.
Crown Copyright © 2020 Published by Elsevier Ltd on behalf of European Paediatric Neurology Society.
All rights reserved.

1. Introduction high income countries [1]. Children surviving neonatal HIE are at
increased risk of adverse outcomes including severe neuromotor
Neonatal Hypoxic Ischaemic Encephalopathy (HIE) as a conse- impairment (cerebral palsy, CP), global cognitive impairment, vi-
quence of peripartum asphyxia is a major cause of neurological sual and/or hearing impairment, and epilepsy.
injury, affecting 1.3e1.7 newborns per 1000 live births in middle to Therapeutic hypothermia (TH) has now become standard care

Abbreviations: HIE, Hypoxic Ischaemic Encephalopathy; CP, Cerebral Palsy; TH, Therapeutic Hypothermia; Bayley-III, The Bayley Scales of Infant and Toddler Develop-
ment; 3rd edition, ASQ; Ages and Stages Questionnaire, CBCL; Child Behavior Checklist 1.5, 5 year; Q-CHAT, Quantitative Checklist for Autism in Toddlers; MNS, Minor
neurological signs; MND, Minor Neurological Dysfunction; M-ABC, Movement Assessment Battery for Children-2.
* Corresponding author. Clinical and Experimental Sciences, Faculty of Medicine, University of Southampton, University Road, Southampton, SO17 1BJ, UK.
E-mail addresses: [email protected] (C.J. Edmonds), [email protected] (S.K. Helps), [email protected] (D. Hart), [email protected] (A. Zatorska),
[email protected] (N. Gupta), [email protected] (R. Cianfaglione), [email protected] (B. Vollmer).

https://doi.org/10.1016/j.ejpn.2020.04.003
1090-3798/Crown Copyright © 2020 Published by Elsevier Ltd on behalf of European Paediatric Neurology Society. All rights reserved.
 C.J. Edmonds et al. / European Journal of Paediatric Neurology 27 (2020) 78e85 79

for infants with moderate or severe HIE. Several large randomized showed signs of moderate to severe encephalopathy. Severity of
controlled trials [2e4] have shown that TH reduces both mortality encephalopathy was classified using the modified Sarnat and Sar-
and severe neurodisability. These effects appear to continue to nat [15] staging; altered state of consciousness (reduced or absent
school age [3e5], although there are some inconsistent results response to stimulation), abnormal tone, and abnormal primitive
between studies with regards to neurodevelopmental outcomes. reflexes (weak or absent suck or Moro response). In line with
The British Association of Perinatal Medicine and The National clinical protocol of this centre which is set to avoid delay in initi-
Institute for Health and Care Excellence (NICE) currently recom- ation of TH, amplitude integrated EEG (aEEG) was not used to
mend that a formal neurodevelopmental assessment is carried out determine initiation of TH.
at around 2 years of age in children surviving neonatal HIE. These
assessments typically comprise a developmental assessment such 2.2. Neurodevelopmental assessment
as the Bayley Scales of Infant and Toddler Development, but in-
formation about behavioural impairments or social and emotional Children were assessed in the follow-up clinic using a stand-
functioning is rarely reported. After this, the children are typically ardised protocol carried out by a Paediatric Neurologist or Neona-
discharged from clinical follow-up unless they have global devel- tologist with experience in neurological and developmental
opmental or severe neuromotor impairment (CP). assessments, together with a Physiotherapist.
Studies examining long term outcomes have typically focused
on outcomes of HIE such as CP and global cognitive impairment 2.2.1. Structured neurological examination
There is little information [6,7] on whether there are specific Neurological examination included assessment of cranial nerve
groups of children surviving HIE without major neuromotor function, movements, posture, reflexes, and muscle tone. Neuro-
impairment that may be at heightened risk of cognitive and/or logical status was categorised as normal (completely normal
behavioural impairment, and the majority of information dates neurologic status), minor neurological signs (gross or fine motor
from the period prior to TH becoming routine clinical practice. coordination difficulties, muscle tone imbalance, without definite
However, some studies have shown that even in the absence of signs of cerebral palsy [CP]), or abnormal (signs of CP present as
such global impairments [8e10] children with HIE may also exhibit defined by the Surveillance of Cerebral Palsy in Europe Working
motor, cognitive and behavioural impairments [11,12], and a recent Group, SCPE, 2000 [16]). Assessors were not blind to the neonatal
systematic review [13] indicates that in the absence of CP, a high course since the children were assessed in a follow-up clinic for
proportion of survivors of HIE remain at risk of general and/or infants who had neonatal HIE.
specific cognitive impairments, even after TH. Current evidence for
behavioural problems is limited [13]. 2.2.2. Developmental assessment
Thus, this indicates that it is important to assess cognitive and 2.2.2.1. Bayley Scales of infant and Toddler Development eIII.
behavioural outcomes of children with HIE with and without CP The Bayley Scales of Infant and Toddler Development eIII (Bayley-
separately. In the present study, which investigates a clinical single III) [17] is a standardised assessment which consists of a series of
centre cohort, we therefore focus on those children who have developmental play tasks. Composite scores are derived for
survived without developing severe neuromotor impairment. The cognitive, language, and motor development and scaled to a metric,
aim of the current study is to describe the neurological, cognitive, with a mean of 100, standard deviation of 15, and range of 40e160.
and behavioural outcomes at a developmental age 2 years of a Mild impairment was defined as a composite score
1e1.5 SD
clinical sample of children who were admitted to a Neonatal below the mean, and severe impairment as a composite score
2
Intensive Care Unit (NICU) for consideration of hypothermia SD below the mean.
treatment for HIE.
2.2.2.2. Ages and Stages Questionnaire-3 (ASQ). Where the Bayley-
2. Methods III could not be completed due to non-compliance, parents
completed the Ages and Stages Questionnaire-3 (ASQ) instead. The
Secondary analysis of anonymised routinely collected clinical ASQ is a questionnaire that screens children for development in the
data on this sample of children was approved by the University of areas of communication, gross and fine motor development,
Southampton Research Ethics Committee (Ethics ID: 26356) and personal-social development, and problem-solving skills. It can be
the HRA and Health and Care Research Wales, HCRW (Reference ID used for children aged between one month and 5½ years. The ASQ
20/HRA/0260; IRAS project ID 278072; University Hospital South- has been found to have similar predictive value for follow-up de-
ampton R&D protocol number RHM CHI1047). cisions to the Bayley-III [18].

2.1. Study population 2.2.3. Behaviour including attention and autism spectrum disorder
symptoms
Infants were enrolled in the study if they had been admitted to 2.2.3.1. Child Behavior Checklist 1.5e5 (CBCL). The parental version
the neonatal unit at University Hospital Southampton between 05/ of the Child Behavior Checklist 1.5e5 (CBCL) was used to screen for
08/09 to 30/05/2016 for consideration of hypothermia treatment behavioural difficulties. The CBCL 1.5e5 uses parental and/or
and enrolled in the clinical follow-up programme. Exclusion teacher behavioural ratings to detect emotional and behavioural
criteria were causes for encephalopathy other than perinatal problems in preschool-aged children. It provides normed scales of
asphyxia, genetic or syndromal disorders and not receiving the full Internalising, Externalising, and Total Problems derived from sub-
72 h of TH. Our criteria for TH are: gestational age
36 weeks scales including emotionally reactive, anxious/depressed, with-
(however, in this clinical cohort, there were 3 infants who were drawn behaviour, somatic complaints, for the internalising scale;
born <36 weeks, 1 at 34 weeks, and 2 at 35 weeks of gestation, who attention problems and aggressive behaviour for the externalising
received TH) and at least one of the following: Apgar score of 5 or scale; sleep and other problems [19]. As recommended by Achen-
less 10 min after birth; continued need for resuscitation, including bach & Edlebrock (1993) [20], we used raw scores in the analysis as
endotracheal or mask ventilation, 10 min after birth; or acidosis they are more precise and uniform than t-scores. The normative
(defined as pH < 7 or base deficit >15 mmol/L, or both, in umbilical sample of the CBCL is based on data from parental report of 700
cord blood or any blood sample) within 1 h of birth [14], and healthy children from 40 different states in the United States.
80 C.J. Edmonds et al. / European Journal of Paediatric Neurology 27 (2020) 78e85

2.2.3.2. Quantitative Checklist for Autism in Toddlers (Q-CHAT).

The Q-CHAT is a 25-item parental report of autistic behaviour
which assesses domains of joint attention, pretend play, language
development, repetitive behaviours, and other aspects of social
communication. The psychometric properties and predictive value
of Q-CHAT have been previously examined [21] and the tool has a
range of scores which approximate normal distribution [22]. The Q-
CHAT is able to discriminate between toddlers with and without a
diagnosis of autism, however, to date, a cut-off point above which
children would be invited for further assessment is yet to be vali-
dated. To evaluate the Q-CHAT scores in the present study, the
mean þ2SD for the normative population reported in Allison and
colleagues’ paper (mean for normative population 26.7, SD 7.8; boys
27.5, SD 7.8; girls 25.8, SD 7.8) were used as a comparison [22].

2.3. Statistics

Outcomes are first reported across the whole sample of tested

children, then comparisons are made between children with minor
neurological signs (MNS) and normal neurology. Children with a
diagnosis of Cerebral Palsy are excluded from these comparison
analyses.
Statistical tests were conducted using IBM SPSS Statistics 25.
Differences between groups were assessed with independent
samples t-test where data were normally distributed and contin-
uous and Mann-Whitney U test where the data were non-normally
distributed and continuous, and with Fisher's Exact tests for cate-
gorical data (as some cells contained fewer than 5 samples).
Comparisons between group means and published norms were
made using a one sample t-test. Where categories were determined
from continuous data e.g. mild or severe delay on assessment with
the Bayley Scales, or borderline or clinical range in CBCL, reference
values given in the test manuals were used.

3. Results

As shown in Fig. 1, 194 newborns were admitted for TH between

05/08/09 to 30/05/2016. Seventeen were excluded from the present
analysis as they had a primary diagnosis other than HIE due to
perinatal asphyxia; 17 (8.8%) infants were excluded since they did
not receive TH or TH was not administered for the full 72 h since the
Fig. 1. Study population. Children eligible for this study were admitted to Neonatal
treating clinician either felt after review that TH was not indicated
Unit for consideration of hypothermia treatment (TH) for neonatal HIE, University
at all, or could be stopped before 72 h of treatment. Of the 140 Hospital Southampton from 05/08/09 to 30/05/2016.
infants who had a primary diagnosis of perinatal asphyxia as a
cause for encephalopathy and who received 72 h of TH, 13 (9.3%)
died. Of the 127 surviving newborns, 20 (15.7%) were lost to follow 3.2. Developmental outcomes
up, 107 (84.3% of the survivors) received follow-up at age 2 years. Of
those, 87 (81.3%) were assessed according to our standardised Analyses of developmental outcomes were performed on the
neurodevelopmental follow-up protocol (see below), and infor- sample of children who had not developed CP by the age of 2 years
mation on neurological outcome was available for further 20 (n ¼ 94). 71 of the 94 children (75.5%) were assessed with the
(18.7%) children who were seen by their Neonatologist or Paedia- Bayley-3 Scales and in 10 (10.6%) developmental screening was
trician for follow-up but were not formally assessed by the Paedi- performed with the ASQ-3. For 13 (13.8%) of the children without
atric Neurologist or according to the standardised protocol. The CP only information on neurological status was available.
mean age of the tested children was 25.9 months (SD 2.6 months;
min 21 - max 32.1 months). The baseline characteristics of the
surviving infants for whom follow-up data were available did not 3.2.1. Bayley-III Scales
differ significantly from those who were lost to follow-up (Table 1). For children (without CP) tested with the Bayley-III Scales
(n ¼ 71), mean Bayley composite scores were broadly average, see
3.1. Neuromotor outcomes Table 2. When considering the presence of delay, 94.4% of tested
children had normal cognitive scores (>85), 4.2% mild cognitive
Of the 107 children with information on neuromotor outcome, delay (70e85), and 1.4% severe cognitive delay (<70); 92.2% had
81 (75.7%) had normal neurology, 13 (12.1%) had minor neurolog- normal language scores (>85), 3.1% mild language delay (70e85)
ical signs, and 13 (12.1%) had CP. The children with a diagnosis of CP and 4.7% severe language delay (<70); and 98.5% had normal motor
were excluded from subsequent analyses, leaving a final sample for scores (>85), 1.5% mild motor delay (70e85), none had severe
analysis of 94 children. motor delay (<70).
 C.J. Edmonds et al. / European Journal of Paediatric Neurology 27 (2020) 78e85 81

Table 1
Demographic data for tested and untested surviving newborns with perinatal asphyxia and those treated with TH; note that this table excludes newborns who had other
diagnoses than perinatal asphyxia as primary cause for encephalopathy and includes the children who developed Cerebral Palsy (CP).

Whole Sample of Died in neonatal period Lost to Follow-up Follow-up age 2 years Comparison between
newborns N ¼ 13 N ¼ 20 N ¼ 107 group of
with perinatal those lost to Follow-up
asphyxia and those
and hyopthermia who had Follow-upe
treatment
N ¼ 140

Birth Weight (g), mean (SD) 3451.6 (680.5) min 3951.0 (687.7) min 2950- 3361.1 (708.7) min 1540- 3375.2 (653.8) min 2200- 0.5
1540-max 4980 max 4930 max 4625 max 4980
Gestational age (wk), mean 39.7 (1.7) min 34-max 39.7 (1.1) min 37- max 41 39.5 (1.9) min 34- max 42 39.7 (1.6) min 35 -max 42 0.36
(SD) 42.1
Sex, n male/female 67/73 7/6 10/10 50/57 0.16
Mode of delivery n (%) 2 (15.4) 5 (25) 49 (45.8) 0.10
Spontaneous vaginal 56 (40) 1 (7.7) 1 (5) 10 (9.3)
Forceps 12 (8.6) e 2 (10) 3 (2.8)
Ventouse 5 (3.6) e e 4 (3.7)
Planned Caesarean 4 (2.2) 10 (76.9) 12 (60) 41 (38.3)
Emergency Caesarean 63 (45)
a
Apgar at 10 min, mean (SD) 5.04 (2.3) min 0-max 3.1 (2.8) min 0-max 9 5.1 (2.5) min 1-max 9 5.3 (2.1) min 0-max 10 0.21
10
b
Cord PH (arterial), mean (SD) 6.9 (0.2) min 6.5-max 6.9 (0.1) min 6.6 -max 7.01 6.7 (0.2) min 6.6- mx 7.3 6.9 (0.2) min 6.5 e max 7.3 0.51
7.3
c
Cord Base Excess (arterial), -15.2 (7.6) min -30 e -16.1 (7.5) min -28.6 - max -13.4 (9.9) min -30 e max -2 -15.4 (7.3) min -30 - max þ4 0.12
mean (SD) max þ4 -2.4
d
Neonatal Seizures, n (%) yes 65 (46.4) 10 (76.9) 6 (30) 49 (45) 0.29

Percentages given are percentages based on available data for each variable in each group.
a
Information missing for 1 newborn who died.
b
Information missing for 11 newborns overall (for 2 newborns who died; for 3 newborns who were lost to FU; for 6 newborns who were seen at age 2 years).
c
Information missing for 11 newborns who were seen at age 2 years.
d
Information missing for 4 newborns overall (for 3 newborns who died; for 1 newborn who was lost to follow-up).
e
Comparison between the group who was lost to follow-up and the group for which were follow-up information available: Chi-Square and Fisher's Exact test for categorical
data, Mann-Whitney U test for continuous data.

Table 2
Results from testing with the Bayley-3 Scales for the 71 children who underwent hypothermia treatment and did not have Cerebral Palsy, categorised by neuromotor status.

Neuromotor status n (%)

All tested children Normal Minor neurological signs p -valued; comparison between group
n ¼ 71a (100%) n ¼ 61 (85.9%) n ¼ 10 (14.1%) with normal neurology and group
with minor neurological signs
a
Cognitive composite score, 106.1 (16.4) min 65-max 145 107.1 (16.8) min 65-max 145 99.5 (11.7) min 80-max120 0.14
mean (SD), min-max
b
Language composite score, 102.9 (15.7) min 59-max 141 104.4 (15.2) min 68-max 141 91.3 (16.3) min 59-max112 0.05
mean (SD), min-max
b
Language receptive scaled score, 10.8 (2.6) min 4-max17 10.9 (2.6) min 4-max 17 8.9 (2.1) min 5-max 12 0.04
mean (SD), min-max
b
Language expressive scaled score, 10.2 (3.1) min 1-max17 10.4 (3.1) min 2-max17 8.1 (3.3) min 1-max12 0.06
mean (SD), min-max
c
Motor composite score, mean (SD), 105.0 (12.5) min 106.0 (12.3) min 97.5 (12.2) min 82-max 115 0.07
min-max 82-max 136 88-max 136
c
Fine motor scaled score, mean (SD) 10.9 (2.9) min 6-max19 10.9 (2.3) min 6-max19 11.2 (3.5) min 7-max 18 0.84
c
Gross motor scaled score, mean (SD) 10.7 (2.9) min 6-max19 11.1 (2.8) min 7-max19 8.0 (2.2) min 6- max 12 0.003
a
Data available for 71 children.
b
Data available for 65 children.
c
Data available for 69 children.
d
Group comparison between normal neurology and minor neurological signs performed with: Mann Whitney U- Test.

3.2.2. Ages and Stages Questionnaire-3 3.2.3. Assessment of behaviour

Ten children had developmental screening with the ASQ. The
majority of children in this subgroup had scores in the normal
3.2.3.1. Screening for autism spectrum symptoms with the QCHAT.
range for all areas that were assessed (for communication 10/10;
Information from the QChat was available for 71 (75.5%) of the 94
gross motor development 8/10; fine motor development 9/10,
children without CP. The mean score across all tested children was
problem solving skills development 9/10, and personal-social
28.0 (SD 9.4; min 12-max 56). This was not significantly different
development 9/10). Eight of the 10 children had normal
from the published norm of 26.7 (Allison et al., 2008); one sample t-
neurology, and 2/10 had MNS. Visual inspection of the date indi-
test, t (0.97), mean difference ¼ 1.05, p ¼ 0.33). There was weak
cated that there was no obvious difference between those with
evidence (p ¼ 0.07) for those with MNS (n ¼ 10; mean score 32.7,
normal neurology and those with MNS (see Table 3). However, it
SD 10.4) showing more autistic spectrum disorder symptoms than
has to be kept in mind that the sample size was very small.
those with normal neurology (n ¼ 61; mean score 26.9, SD 8.6).
82 C.J. Edmonds et al. / European Journal of Paediatric Neurology 27 (2020) 78e85

Table 3 language assessments. CBCL total scores in children with HIE prior
Number of children scoring as normal, borderline or delayed on ASQ, shown by to TH becoming standard clinical care report similar overall rates of
neuromotor status.
total problems to those observed in our study, but the breakdown
Neuromotor status, n of total problems score into the subcomponents was not reported
Normal n ¼ 8 Minor neurological [23]. Our data add to the literature on the particular types of
signs n ¼ 2 behavioural problems that children with HIE treated with TH might
ASQ Communication Skills experience. In our sample, the proportion of children experiencing
Normal 8/8 2/2 internalising problems was similar to that observed in the general
Borderline 0 0 population (when compared to test norms); our data suggest that
Delayed 0 0
high rates of overall problems on the CBCL total problems score are
ASQ Gross Motor Skills
Normal 7/8 1/2 specific to difficulties with externalising problems and sleep. The
Borderline 1/8 0 pattern of findings for performance on the Bayley Scales were
Delayed 0 1/2 similar, with composite scores similar to those reported in the
ASQ Fine Motor Skills general population. It is tempting to speculate that TH may not be
Normal 7/8 2/2
protective with regards to more subtle behavioural problems.
Borderline 0 0
Delayed 1/8 0 However, current evidence is limited, and more research is required
ASQ Problem Solving here.
Normal 7/8 2/2 In our cohort, 75.7% of children were neurologically normal,
Borderline 1/8 0
12.1% had CP (and were excluded from our analyses) and 12.1% had
Delayed 0 0
ASQ Personal-Social Skills
minor neurological signs. The rate of CP is lower than that reported
Normal 7/8 2/2 for toddler age from the large RCTs on therapeutic hypothermia
Borderline 1/8 0 [24] but similar rates have been reported from observational
Delayed 0 0 studies [25]. Comparison of the rate of MNS is difficult, since this
has not been previously reported in children with HIE treated with
TH at toddler age. This, however, appears to be an important group
3.2.3.2. Screening for internalising, externalising, sleep and other since they were found to consistently experience more difficulties
problems with the CBCL. Information from the CBCL was available across multiple domains than the children with HIE and normal
for 74 (78.7%) of the 94 children without CP. Table 4 shows the neurology. Children with MNS differed from children with normal
mean scores for total problem score, internalising and externalising neurology in multiple domains, with significantly lower cognitive,
problems, sleep problems, and other problems. On the CBCL, 72.9% language and motor scores, as well as slightly higher ratings of
of children had total problem scores in the normal range, 10.8% in autistic traits (Q-CHAT). Children with MNS scored higher on the
the borderline range, and 17.6% in the clinical range. For internal- total problem and internalising scales and had higher scores for
ising problems, 85.1% scored in the normal range, 5.4% were in the parent reported sleep problems on the CBCL than children with
borderline clinical range and 9.5% in the clinical range. For exter- normal neurology. Analysis of the individual subscales of the CBCL
nalising problems, 74.3% of children scored in the normal range, showed that children with MNS had higher scores in the domain of
6.8% in the borderline clinical range, and 18.9% in the clinical range. anxiety/depression, and this difference was approaching signifi-
Clinically relevant sleep problems were reported for 8.4% of the cance for emotionally reactive and somatic complaints subscales.
children, borderline scores for 7.3%, and 84.3% had scores in the In the context of those children with MNS having lower cogni-
normal range for sleep behaviour. tive scores, and screening for behavioural and sleep problems
In comparison, normative published means show that, for total indicating more difficulties than for those with normal neurological
problems, externalising and internalising scores, 82% of children examination, the concept of Minor Neurological Dysfunction
scored in the normal range, 8% in the borderline range, and 10% in (MND) is of interest. Minor neurological dysfunction, usually not
the clinical range. For sleep problems, normative means report that diagnosed before school age, describes a neurological profile that
92% of children score in the normal range, 5% in the bordeline range includes difficulties with muscle tone regulation, posture, balance,
and 3% in the clinical range [20]. coordination, mildly abnormal reflexes and cranial nerve function
Comparing the group with MNS (n ¼ 10) with those with normal [26]. Whilst the complex form of MND has been linked to learning,
neurology (n ¼ 64), total problem scores were significantly cognitive and motor problems in school aged children born very or
(p ¼ 0.04; Table 4) higher in the group with MNS. Looking at the extremely preterm [27,28], both the presence of MNS or MND, and
component parts of the overall score, there was a significant dif- whether it is linked to other aspects of cognitive and behavioural
ference between groups for reported sleep problems (p ¼ 0.04), development, has not been previously reported in children with
“other problems” (p ¼ 0.01) and internalising problems (p ¼ 0.03). HIE. It has to be noted that the presence of MNS at toddler age does
There was no difference between those with MNS and normal not necessarily mean that these children will go on to have MND
neurology for externalising problems (p ¼ 0.09). The observed later in life. The children in our cohort are too young for a formal
differences between the two groups on the internalising scale were assessment of MND with the currently available tools (i.e. Touwen
most pronounced for the subscales screening for anxious/ Neurological Examination) and therefore it is not possible to assign
depressed behaviour (p ¼ 0.04), somatic complaints (p ¼ 0.07), a diagnosis of MND to the children in our cohort in whom MNS
emotionally reactive behaviour (p ¼ 0.06). For the externalising were seen. Currently, there is work going on for validation of an
scale, there was weak evidence that aggressive behaviour was more assessment for MND that can be used at toddler age (Hempel
frequent that in those with normal neurology (p ¼ 0.06). Assessment) to identify MND, but validity and reliability data are
not yet available for the populations of infants with HIE or born
4. Discussion preterm. However, it seems reasonable to infer from our findings of
associations between MNS and poorer performance on cognitive
When the whole sample of 2 year old children with HIE who assessment and behavioural difficulties screening that those with
survived without developing CP are considered, the majority of MNS need enhanced surveillance and specific assessment for MND
children scored in the normal range on cognitive, motor and once they reach school age.
 C.J. Edmonds et al. / European Journal of Paediatric Neurology 27 (2020) 78e85 83

Table 4
Results from the CBCL 1.5e5 parental questionnaire for children who had TH and did not develop Cerebral Palsy, categorised by neuromotor status.

Neuromotor status n (%)

All children with CBCL data Normal Minor neurological signs p -valuea; comparison between group
n ¼ 74 (100%) n ¼ 64 (86.4%) n ¼ 10 (13.6%) with normal neurology and group
with minor neurological signs

ˆCBCL Total Problem Score 33.4 (28.6) min 0-max 117 29.1 (26.4) min 0-max 117 56.7 (37.3) min 4-max 108 0.04
ˆCBCL Internalising Score, mean (SD) 7.3 (7.5) min 0-max-34 5.9 (6.1) min 0-max 31 14.4 (11.9) min 0-max 34 0.03
^CBCL Internalising - Emotionally reactive 2.1 (2.7) min 0-max15 1.7 (2.1) min 0-max 8 4.8 (4.9) min 0-max 15 0.06
^CBCL Internalising - Anxious/depressed 2.1 (2.7) min 0-max 13 1.8 (2.3) min 0-max 13 4.4 (4.1) min 0-max 13 0.04
^CBCL Internalising - Somatic complaints 1.8 (2.3) min 0-max 10 1.5 (1.9) min 0-max 8 3.5 (3.4) min 0-max 10 0.07
^CBCL Internalising - Withdrawn 1.2 (2.1) min 0-max 12 0.97 (1.5) min 0- max 7 1.7 (2.7) min 0-max 8 0.7
ˆCBCL Externalising Score, mean (SD) 13.1 (10.9) min 0-max 41 12.0 (10.4) min 0-max 41 20.3 (13.3) min 0-max 38 0.09
^CBCL Externalising - Attention 2.8 (2.7) min 0-max 10 2.5 (2.7) min 0 e max 9 3.9 (3.2) min 0-max 10 0.15
^CBCL Externalising - Aggression 10.1 (8.5) min 0-max 33 9.4 (8.2) min 0 e max 33 15.8 (10.1) min 0-max 28 0.06
ˆCBCL Sleep problems 3.01 (3.7) min 0-max 14 2.6 (3.4) min 0-max 14 5.9 (5.0) min 0-max 14 0.04
^CBCL Other problems 10.1 (8.9) min 0-max 36 8.6 (8.5) min 0-max 36 17.0 (9.9) min 3-max 31 0.01

CBCL, Child Behavior Checklist, raw scores used; data available for 74 children.
a
Group comparison between normal neurology and minor neurological signs performed with Mann Whitney U- Test.

Comparison with other studies is difficult since, to our knowl- cognitive, emotional and behavioural difficulties should be fol-
edge, no data on MNS at toddler age in the context of HIE have been lowed up in future studies.
reported. Some information is available for school age. Marlow et al Comparison with another group of infants with early brain
(2005) [29], in a regional cohort of school aged children with injury or atypical brain development, i.e. children born very or
neonatal encephalopathy (not treated with TH), using the Touwen extremely preterm, shows a similar pattern of associations be-
assessment for MND, did not find an increased prevalence/incidence tween neuromotor signs in the absence of CP and cognitive and/or
of MND in the absence of CP. However, this is the only study looking behavioural function as in our cohort [27,28]. This might suggest
at MND in children with neonatal HIE and more research is needed that, despite injury to the brain occurring at different develop-
to either replicate or disconfirm this, in particular for children who mental stages in these two groups, subsequent subtle widespread
were born in the era of TH. Furthermore, at older ages, outcome alterations to similar brain networks involved in motor control may
assessments vary for the few existing studies, with most studies be present. To further examine this, MRI studies on anatomical and
focussing on assessment of motor function and not including functional brain networks would be useful.
assessment of minor neurological signs similar to that used in our While the majority of children in our sample with HIE without
study. However, comparison is possible with regards to occurrence CP scored in the normal range for sleep behaviour, the incidence of
of atypical neuromotor signs generally. Perez et al, 2013 6, described clinically relevant sleep problems was over twice that reported in
for a cohort of school and teenage age children with a history of typically developing children. Furthermore, in our study, signifi-
neonatal HIE (not treated with TH) who had survived without cantly more sleep problems were reported in children with MNS
developing CP, impairment in motor speed and quality of move- compared to those with normal neurology. Other studies have also
ments, and this was related to impairment of general cognitive indicated problems with sleep behaviour in children with HIE. For
abilities. This is in line with our findings. No data on behavioural example, Ding et al (2016) [32] reported that overall sleep times of
outcomes were, however, reported for this cohort. More recently, 3-year-old children with HIE were shorter than those of typically
two papers were published on one small sample size cohort of 6e8 developing, age-matched controls, potentially due to difficulties in
year of children who underwent TH. These studies reported that the sleep initiation and maintenance, as well as breathing problems
total score on the Movement Assessment Battery for Children-2 (M- during sleep. Additionally, there was an indication that different
ABC) was significantly lower (indicating poorer motor performance) types of difficulties may be experienced by children with mild and
in children with HIE compared to controls [11,12], and associations moderate HIE. Sleep-related problems in children with HIE can be
between MABC-2 with Full Scale IQ, working memory and percep- observed shortly after birth. Infants with HIE show a delay in
tual reasoning scores, with those showing poorer performance on development of the sleep-wake cycle measured using EEG [33,34],
motor tests performing poorer on the cognitive tests. One of the a delay which is related to the severity of HIE [33]. Furthermore,
papers noted poor predictive value of Bayley-3 motor composite long-term developmental outcome was related to the onset of the
score at 18 months on MABC-2 scores at age 6e8 years, again indi- sleep-wake cycle; 1 to 5.5 year-old-children with HIE whose
cating that long term follow up is essential [11]. newborn sleep-wake cycle started before 36 h had Griffiths scores
Children with developmental difficulties can experience a 8.5 points higher than those whose newborn sleep-wake cycle
widening gap on starting school, when cognitive and behavioural started after 36 h [33]. These findings should be followed up in a
demands increase, and specific learning difficulties are more systematic manner to evaluate both the presence and type of sleep
readily observed, as documented in the literature on school readi- problems, and the relation to long-term outcome.
ness [30,31]. Thus, it is important to assess the longer term devel- As TH is now standard treatment for moderate to severe HIE,
opmental trajectory of children with HIE. A recent systematic contemporaneous comparisons of TH treated and non-treated
review of the literature that focused on follow-up studies in cohorts children is no longer possible. Moving forwards, it may be of
aged 4 years or older reported that a large proportion of children particular interest and relevance to compare children with HIE
with HIE without CP are at increased risk of cognitive impairment, treated with TH with their age matched peers without HIE, in order
with specific cognitive difficulties in attention, language and ex- to explore how they are performing later in life. With regards to
ecutive functions, and limited evidence for behavioural problems assessment of neurology and neuromotor function, it will be
[13]. A wider range of more focussed developmental assessments important in future research to use specifically designed stand-
exist to assess school aged children and the presence of specific ardised and validated tools for detection of minor neurological
84 C.J. Edmonds et al. / European Journal of Paediatric Neurology 27 (2020) 78e85

dysfunction, combined with standardised and validated tests of [2] P.D. Gluckman, J.S. Wyatt, D. Azzopardi, R. Ballard, A.D. Edwards, D.M. Ferriero,
et al., Selective head cooling with mild systemic hypothermia after neonatal
neuromotor function.
encephalopathy: multicentre randomised trial, Lancet 365 (9640) (2005)
Our findings have implications for clinical practice. Children 633e670, https://doi.org/10.1016/S0140-6736(05)17946-X.
surviving HIE who do not present with CP are often routinely dis- [3] S. Shankaran, A. Pappas, S.A. McDonald, B.R. Vohr, S.R. Hintz, K. Yolton, et al.,
charged from clinical care at age 2 years. Children surviving HIE Childhood outcomes after hypothermia for neonatal encephalopathy, N. Engl.
J. Med. 366 (2012) 2085e2092, https://doi.org/10.1056/NEJMoa1112066.
who present with MNS are rarely followed up clinically beyond 2 [4] R. Guillet, A.D. Edwards, M. Thoresen, D.M. Ferriero, P.D. Gluckman,
years of age. Furthermore, clinicians who perform follow-up as- A. Whitelaw, et al., Seven-to eight-year follow-up of the CoolCap trial of head
sessments are not always trained to detect MNS. Our findings cooling for neonatal encephalopathy, Pediatr. Res. 71 (2) (2012) 205e209,
https://doi.org/10.1038/pr.2011.30.
suggest that children surviving HIE with MNS are at heightened risk [5] D. Azzopardi, B. Strohm, N. Marlow, P. Brocklehurst, A. Deierl, O. Eddama, et
of impairment in multiple domains compared to children with HIE al., Effects of hypothermia for perinatal asphyxia on childhood outcomes,
who exhibit normal neurology. This suggests that these children Obstet. Gynecol. Surv. 69 (11) (2014) 639e641, https://doi.org/10.1097/
01.ogx.0000458787.40317.4a.
may benefit from a longer period of routine clinical follow up to [6] A. Perez, S. Ritter, B. Brotschi, H. Werner, J. Caflisch, E. Martin, et al., Long-term
monitor their development and to provide appropriate support. neurodevelopmental outcome with hypoxic-ischemic encephalopathy,
Important limitations of our work include the relatively small J. Pediatr. 163 (2) (2013) 154e159, https://doi.org/10.1016/j.jpeds.2013.02.003.
[7] B.C. Hayes, E. Doherty, A. Grehan, C. Madigan, C. McGarvey, S. Mulvany, et al.,
sample size of the cohort and that not all children completed the Neurodevelopmental outcome in survivors of hypoxic ischemic encephalop-
follow-up assessment. As there were no significant baseline dif- athy without cerebral palsy, Eur J Paediatr 177 (1) (2018) 19e32, https://
ferences between tested and not tested children, while possible, it doi.org/10.1007/s00431-017-3028-3.
[8] L.S. de Vries, M.J. Jongmans, Long-term outcome after neonatal hypoxic-
seems unlikely that attrition bias could explain our findings.
ischaemic encephalopathy, Arch Dis Child Fetal Neonatal 95 (3) (2010)
220e224, https://doi.org/10.1136/adc.2008.148205.
5. Conclusion [9] J. Armstrong-Wells, T.J. Bernard, R. Boada, M. Manco-Johnson, Neurocognitive
outcomes following neonatal encephalopathy, NeuroRehabilitation 26 (1)
(2010) 27e33, https://doi.org/10.3233/NRE-2010-0533.
In conclusion, in our clinical cohort of children with HIE without [10] F.F. Gonzalez, S.P. Miller, Does perinatal asphyxia impair cognitive function
CP treated with TH, our two-year follow-up found that the majority without cerebral palsy? Arch. Dis. Child. Fetal Neonatal Ed. 91 (6) (2006)
454e459, https://doi.org/10.1136/adc.2005.092445.
of the sample showed typical development for cognition, motor [11] S. Jary, R. Lee-Kelland, J. Tonks, F.M. Cowan, M. Thoresen, E. Chakkarapani,
skills, and language. There was no strong evidence of increased Motor performance and cognitive correlates in children cooled for neonatal
autistic traits, but more children exhibited externalising symptoms encephalopathy without cerebral palsy at school age, Acta Paediatr Int J
Paediatr (2019), https://doi.org/10.1111/apa.14780, 2e9.
in the clinical range than in the general population. The subgroup of
[12] R. Lee-Kelland, S. Jary, J. Tonks, F.M. Cowan, M. Thoresen, E. Chakkarapani,
children with HIE and MNS consistently experienced greater diffi- School-age outcomes of children without cerebral palsy cooled for neonatal
culties across multiple domains than those with normal neurology, hypoxic-ischaemic encephalopathy in 2008-2010, Arch Dis Child Fetal
Neonatal (2019), https://doi.org/10.1136/archdischild-2018-316509, 1e6.
with difficulties in internalising behaviour, sleep, and other prob-
[13] M. Schreglmann, A. Ground, B. Vollmer, M.J. Johnson, Systematic Review:
lems. Therefore, this group may benefit from enhanced clinical Long-Term Cognitive and Behavioural Outcomes of Neonatal Hypoxic-
follow up to monitor their development, including sleep patterns, Ischaemic Encephalopathy in Children without Cerebral Palsy, Acta Paediatr.
and to allow physicians to identify those who would benefit from (2019), https://doi.org/10.1111/apa.14821.
[14] D. Azzopardi, B. Strohm, A.D. Edwards, L. Dye, H. Halliday, E. Juszczak, et al.,
early intervention. In addition, clinicians should be trained to Moderate hypothermia to treat perinatal asphyxial encephalopathy, N. Engl. J.
identify minor neurological signs in these children and an appro- Med. 361 (14) (2009) 1349e1358.
priate care pathway should be designed. [15] H.B. Sarnat, M.S. Sarnat, Neonatal encephalopathy following fetal distress. A
clinical and electroencephalographic study, Arch. Neurol. 33 (10) (1976)
696e705.
Funding [16] C. Cans, Surveillance of cerebral palsy in Europe: a collaboration of cerebral
palsy surveys and registers, Dev. Med. Child Neurol. 42 (2000) 816e824,
https://doi.org/10.1111/j.1469-8749.2000.tb00695.x.
This research did not receive any specific grant from funding [17] N. Bayley, Bayley scales of infant and toddler developmente third edition. San
agencies in the public, commercial, or not-for-profit sectors. Antonio, TX: harcourt assessment, J. Psychoeduc. Assess. (25) (2006)
180e190, https://doi.org/10.1177/0734282906297199.
[18] R. Mackin, N Ben Fadel, J. Feberova, L. Murray, A. Nair, S. Kuehn, et al., ASQ3
Contributors’ statement and/or the bayley-III to support clinicians' decision making, PloS One 12 (2)
(2017) 1e13, https://doi.org/10.1371/journal.pone.0170171.
[19] T. Achenbach, C. Edlebrock, Manual for the Child Behavior Checklist and
CJE: Design, data analysis, drafted the manuscript and reviewed
Revised Child Behavior Profile, Burlingt Univ Vermont, Dep Psychiatry, 1993.
and revised the manuscript. [20] T.M. Achenbach, C. Edelbrock, Manual for the Child Behavior Checklist and
SH: Data analysis, drafted the initial manuscript, and reviewed Revised Child Behavior Profile, Queen City Printers, Burlington, VT, 1983.
[21] I. Magiati, D.A. Goh, S.J. Lim, D.Z.Q. Gan, J.C.L. Leong, C. Allison, et al., The
and revised the manuscript.
psychometric properties of the Quantitative-Checklist for Autism in Toddlers
DH: Data collection, review and revision of manuscript. (Q-CHAT) as a measure of autistic traits in a community sample of Singa-
AZ: Data analysis, review and revision of manuscript. porean infants and toddlers, Mol. Autism. 6 (1) (2015) 1e14, https://doi.org/
NG: Data collection, review and revision of manuscript. 10.1186/s13229-015-0032-1.
[22] C. Allison, S. Baron-Cohen, S. Wheelwright, T. Charman, J. Richler, G. Pasco, et
RC: Data analysis, review and revision of manuscript. al., The Q-CHAT (Quantitative CHecklist for Autism in Toddlers): a normally
BV: Design, data collection, review and revision of manuscript. distributed quantitative measure of autistic traits at 18-24 months of age:
All authors approved the final manuscript as submitted and preliminary report, J. Autism Dev. Disord. 38 (8) (2008) 1414e1425, https://
doi.org/10.1007/s10803-007-0509-7.
agree to be accountable for all aspects of the work. [23] P.E.M. Van Schie, J. Schijns, J.G. Becher, F. Barkhof, M.M. Van Weissenbruch,
R.J. Vermeulen, Long-term motor and behavioral outcome after perinatal
hypoxic-ischemic encephalopathy, Eur. J. Paediatr. Neurol. 19 (3) (2015)
Declaration of competing interest
354e359, https://doi.org/10.1016/j.ejpn.2015.01.005.
[24] A.D. Edwards, P. Brocklehurst, A.J. Gunn, H. Halliday, E. Juszczak, M. Levene, et
None to declare. al., Neurological outcomes at 18 months of age after moderate hypothermia
for perinatal hypoxic ischaemic encephalopathy: synthesis and meta-analysis
of trial data, BMJ 340 (7743) (2010) 409, https://doi.org/10.1136/bmj.c363.
References [25] C.E. Ahearne, G.B. Boylan, D.M. Murrary, Short and long term prognosis in
perinatal asphyxia: an update, World J. Clin. Pediatr. 5 (1) (2016) 67e74,
[1] J.J. Kurinczuk, M. White-Koning, N. Badawi, Epidemiology of neonatal en- https://doi.org/10.5409/wjcp.v5.i1.67.
cephalopathy and hypoxic-ischaemic encephalopathy, Early Hum. Dev. 86 (6) [26] M. Hadders-Algra, The Neurological Examination of the Child with Minor
(2010) 329e338, https://doi.org/10.1016/j.earlhumdev.2010.05.010. Neurological Dysfunction, third ed., vol. 6, Mac Keith Press, London, 2010.
 C.J. Edmonds et al. / European Journal of Paediatric Neurology 27 (2020) 78e85 85

[27] C. Arnaud, L. Daubisse-Marliac, M. White-Koning, V. Pierrat, B. Larroque, [31] C. Hughes, S. Foley, N. White, R.T. Devine, School readiness in children with
H. Grandjean, et al., Prevalence and associated factors of minor neuromotor special educational needs and disabilities: psychometric findings from a new
dysfunctions at age 5 Years in prematurely born children, JAMA Pediatr 161 screening tool, the Brief Early Skills, and Support Index, Br. J. Educ. Psychol. 88
(11) (2007) 1053e1061, https://doi.org/10.1001/archpedi.161.11.1053. (4) (2018) 606e627, https://doi.org/10.1111/bjep.12206.
€m, B. Vollmer, J. Bolk, E. Eklo
[28] L. Brostro € f, U. Åde
n, Minor neurological [32] X. Ding, Z. Cheng, B. Sun, J. Huang, L. Wang, X. Han, et al., Distinctive sleep
dysfunction and associations with motor function, general cognitive abilities, problems in children with perinatal moderate or mild hypoxic-ischemia,
and behaviour in children born extremely preterm, Dev. Med. Child Neurol. 60 Neurosci. Lett. 614 (2016) 60e64, https://doi.org/10.1016/
(8) (2018) 826e832, https://doi.org/10.1111/dmcn.13738. j.neulet.2015.12.061.
[29] N. Marlow, A.S. Rose, C.E. Rands, E.S. Draper, Neuropsychological and educa- [33] D. Osredkar, M.C. Toet, L.G.M. van Rooij, A.C. van Huffelen, F. Groenendaal,
tional problems at school age associated with neonatal encephalopathy, Arch L.S. de Vries, Sleep-wake cycling on amplitude-integrated electroencepha-
Dis Child Fetal Neonatal 90 (5) (2005) 380e387, https://doi.org/10.1136/ lography in term newborns with hypoxic-ischemic encephalopathy, Pediat-
adc.2004.067520. rics 115 (2) (2005) 327e332, https://doi.org/10.1542/peds.2004-0863.
[30] C. Hughes, I. Daly, S. Foley, N. White, R.T. Devine, Measuring the foundations [34] T. Takenouchi, E.O. Rubens, V.L. Yap, G. Ross, M. Engel, J.M. Perlman, Delayed
of school readiness: introducing a new questionnaire for teachers - the brief onset of sleep-wake cycling with favorable outcome in hypothermic-treated
early skills and support index (BESSI), Br. J. Educ. Psychol. 85 (3) (2015) neonates with encephalopathy, J. Pediatr. 159 (2) (2011) 232e237, https://
332e356, https://doi.org/10.1111/bjep.12076. doi.org/10.1016/j.jpeds.2011.01.006.