Embed Size (px)
Citation preview
Glucose Derangements and Brain Function in Neonatal Encephalopathy
by
Elana Pinchefsky
A thesis submitted in conformity with the requirements for the degree of Master of Science
Institute of Medical Science University of Toronto
© Copyright by Elana Pinchefsky 2018
ii
Glucose Derangements and Brain Function in Neonatal
Encephalopathy
Elana Pinchefsky
Master of Science
Institute of Medical Science University of Toronto
2018
Abstract
Hypoglycemia and hyperglycemia are common in neonates with encephalopathy and may be
independently associated with unfavourable outcomes. This thesis describes how glucose
derangements detected with continuous interstitial glucose monitoring commenced soon after
birth are associated with brain function on amplitude-integrated electroencephalography (aEEG)
and continuous EEG (cEEG) in the context of neonatal encephalopathy. We hypothesize that
hypoglycemia, hyperglycemia and glucose variability in neonatal encephalopathy are associated
with abnormal brain background activity and increased electrographic seizures on aEEG and
cEEG. In neonates with encephalopathy, epochs of hyperglycemia, but not hypoglycemia, were
common and temporally associated with worse aEEG background scores, less sleep-wake
cycling, and more electrographic seizures, including after adjusting for clinical markers of
hypoxia-ischemia. Epochs of hyperglycemia also were associated with worse cEEG background
scores, including after adjusting for hypoxia-ischemia severity. Our data support the hypothesis
that proactive avoidance of hyperglycemia may be a neuroprotective strategy in infants with
neonatal encephalopathy.
iii
Acknowledgments
This thesis could not have been produced without the help and support of many amazing people.
First and foremost, I would like to express my sincere gratitude to my supervisors, Drs. Emily
Tam and Cecil Hahn for their mentorship, time and support that have allowed me to develop my
research and clinical skills. I have felt continuously supported, challenged and encouraged
throughout the past 2 years. Your patience, insight and guidance with my research and future
career development have been invaluable.
Secondly, thank you to Dr. Vann Chau for all of your extremely valuable advice,
mentorship and continued friendship that have helped me through my training. And thank you to
Dr. Steven Miller for all your mentorship, words of wisdom, and research and clinical pearls
that you have shared. I have learnt so much during my time at SickKids Hospital and the
University of Toronto and I am very grateful for the kind and supportive work environment.
Importantly, I would also like to express my sincere gratitude to my program advisory
committee members, Drs. Carter Snead and Mark Palmert for their time, expertise and
investment in my research career. Your extremely helpful constructive feedback helped me
shape and improve the work within this thesis.
I am also grateful to Dr. Sampsa Vanhatalo and his research group from Helsinki,
including Nathan Stevenson, Karoliina Tapani and Minna Kauppila for such an exciting
collaboration. Your input and expertise in quantitative EEG analysis have been vital and I look
forward to working further with your group.
Furthermore, thank you to Daphne Kamino, our clinical research project manager, who
keeps the NOGIN study organized and answers my many questions, at all times of the day. I am
so grateful for your support and friendship. I would also like to thank all the members of the
Tam and Miller labs for your support and for providing such a wonderful and inspirational work
environment. I would especially like to thank Diane Wilson, Claire Watt, Ashley LeBlanc,
Torin Glass, Lara Leijser, Julianne Schneider, Amr Al-Shahed, Jojy Varghese, Marie Metrailler,
Isabel Benavente-Fernandez, Dalit Cayam-Rand, and Asma Al- Mazroei. Thank you for your
friendship and support throughout my fellowship. You have all been great team members and
iv
have taught me so much. In addition, this project would not have been possible without Dr.
Ayako Ochi, Dr. Tina Go, Dr. Aideen Moore, the wonderful neurophysiology technologists, and
Matthew Keyzers and the Acute Care Transport Services team.
A special thank you to my family and friends for their continued love, support and
encouragement, and especially my fiancé, Andrei Isac for being unfailingly supportive during
my two years away. You have always pushed me to be the best that I can be.
Lastly, I would like to extend the largest thank you, and my sincerest appreciation to all
the NOGIN study subjects and their families, without whom this study could not have
happened.
v
Table of Contents
Acknowledgments................................................................................................................iii
TableofContents...................................................................................................................v
ListofAbbreviations...........................................................................................................viii
ListofTables.........................................................................................................................xi
ListofFigures.......................................................................................................................xii
ListofAppendices...............................................................................................................xiii
Chapter1...............................................................................................................................1
LiteratureReview.............................................................................................................11
1.1 Introduction..........................................................................................................................1
1.2 NeonatalEncephalopathy.....................................................................................................2
1.2.1 Overview................................................................................................................................2
1.2.2 Prevalenceofglucosederangements....................................................................................5
1.3 Glucosephysiologyinneonates............................................................................................7
1.3.1 Normaltransitionalperiodandchallengesofdefiningneonatalhypoglycemia...................7
1.3.2 Glucosephysiologyinneonatalencephalopathy................................................................11
1.3.3 Glucosemanagementguidelines.........................................................................................14
1.3.3.1 Hypoglycemiamanagement........................................................................................................141.3.3.2 Hyperglycemiamanagement.......................................................................................................17
1.4 Neurodevelopmentaloutcomesfollowingneonatalglucosederangements........................21
1.4.1 Overview..............................................................................................................................21
1.4.2 Shorttermoutcomes...........................................................................................................23
1.4.3 Longtermoutcomes............................................................................................................24
1.4.3.1 Wellbabies..................................................................................................................................241.4.3.2 Prematureinfants........................................................................................................................28
1.4.4 Outcomesfollowingglucosederangementsinneonatalencephalopathy..........................29
1.5 Continuousinterstitialglucosemonitoring..........................................................................34
1.5.1 Continuousinterstitialglucosemonitorsoverview.............................................................34
1.5.2 Continuousinterstitialglucosemonitoruseinnewborns...................................................36
1.5.3 Measuresofglucosevariabilityandcomplexity..................................................................37
vi
1.6 Neurophysiology.................................................................................................................40
1.6.1 Continuouselectroencephalography(cEEG).......................................................................40
1.6.2 Amplitude-integratedelectroencephalography(aEEG).......................................................43
1.7 Neurophysiologyandneurodevelopmentaloutcomes........................................................47
1.7.1 Backgroundandneurodevelopmentaloutcomes...............................................................47
1.7.2 Sleep-wake-cyclingandneurodevelopmentaloutcomes....................................................49
1.7.3 Seizuresandneurodevelopmentaloutcomes.....................................................................51
1.8 GlucosedisturbancesandEEG.............................................................................................64
1.8.1 HypoglycemiaandEEGchanges..........................................................................................64
1.8.2 HyperglycemiaandEEGchanges.........................................................................................66
1.9 ResearchHypothesisandAims............................................................................................69
1.9.1 Aims.....................................................................................................................................69
1.9.2 Hypothesis...........................................................................................................................70
Chapter2.............................................................................................................................71
Methods........................................................................................................................712
2.1 Studypopulation................................................................................................................71
2.2 Amplitude-integratedelectroencephalography(aEEG)monitoringandanalysis..................72
2.3 Continuouselectroencephalography(cEEG)monitoringandanalysis..................................72
2.4 Continuousinterstitialglucosemonitoring..........................................................................74
2.5 Statisticalanalysis...............................................................................................................76
2.5.1 StatisticalanalysisofaEEGmonitoring................................................................................76
2.5.2 StatisticalanalysisofcEEGmonitoring................................................................................77
Chapter3.............................................................................................................................79
Results...........................................................................................................................793
3.1 Prevalenceofglucosederangementsincohortandtreatmentsreceived............................79
3.2 Amplitude-integratedEEG...................................................................................................80
3.2.1 Studypopulation.................................................................................................................80
3.2.2 PrevalenceofglucosederangementsduringaEEGmonitoring...........................................82
3.2.3 Amplitude-integratedEEGfindings.....................................................................................82
3.2.3.1 aEEGmeasuresduringepochsofglucosederangements...........................................................823.2.3.2 aEEGmeasuresandseverityofglucosederangements..............................................................853.2.3.3 aEEGmeasuresandglucosevariability........................................................................................86
vii
3.3 ContinuousEEG...................................................................................................................87
3.3.1 Demographicsandclinicalvariables....................................................................................87
3.3.2 Glucosederangements........................................................................................................87
3.3.3 ContinuousEEGfindings......................................................................................................88
3.3.3.1 cEEGmeasuresduringepochsofglucosederangements...........................................................883.3.3.2 cEEGmeasuresandlevelofglucose............................................................................................903.3.3.3 cEEGmeasuresinsubgroupofpatientswithhyperglycemiaepisodesduringcontinuousglucose
monitoring...................................................................................................................................................913.3.3.4 cEEGmeasuresinsubgroupofpatientswithhypoglycemiaepisodesduringcontinuousglucose
monitoring...................................................................................................................................................93
Chapter4.............................................................................................................................95
Generaldiscussion.........................................................................................................954
4.1 Discussion...........................................................................................................................95
4.2 Limitations.........................................................................................................................100
4.3 Conclusions........................................................................................................................102
Chapter5...........................................................................................................................103
Futuredirections..........................................................................................................1035
5.1 ContinuousEEGmonitoring-Qualitativeanalysis..............................................................103
5.2 ContinuousEEGmonitoring-QuantitativeEEGanalyses....................................................104
5.2.1 PreviousstudiesvalidatinguseofquantitativeEEGanalyses...........................................104
5.2.2 ApplicationofquantitativeEEGanalysesintheNOGINstudy..........................................108
5.3 Shortandlong-termoutcomes...........................................................................................109
5.4 Futuredirections-Summary..............................................................................................110
References.........................................................................................................................112
Appendix1-AppendixTable1...........................................................................................156
Appendix2-Contributions................................................................................................157
CopyrightAcknowledgements...........................................................................................158
viii
List of Abbreviations
AAP American Academy of Pediatrics
ACNS American Clinical Neurophysiology Society
aEEG Amplitude-integrated EEG
AGA Appropriate for gestational age
BS Burst suppression
BSID Bayley Scales of Infant Development
cEEG Continuous EEG
CGM Continuous glucose monitor
CHD Congenital heart disease
CLV Continuous low voltage
CNV Continuous normal voltage
CONGA Continuous overlapping net glycemic action
CP Cerebral palsy
CPS Canadian Pediatric Society
DFA Detrended fluctuation analysis
DNV Discontinuous normal voltage
DWI Diffusion weighted imaging
EEG Electroencephalography
ELBW Extremely low birth weight neonates
ES Electrographic seizures
ESE Electrographic status epilepticus
FSIQ Full-scale Intelligence Quotient
FT Flat tracing
ix
GA Gestational age
GMFCS Gross Motor Function Classification System
GOS-E Peds Glasgow Outcome Scale - Extended, Pediatric
HI Hypoxic-ischemia
HIE Hypoxic-ischemic encephalopathy
IBI Interburst interval
ICU Intensive care unit
IDM Infants of diabetic mother
IQ Intelligence quotient
KOSCHI King’s Outcome Scale for Childhood Head Injury
LGA Large-for-gestational age
MAG Mean absolute glucose change
MAGE Mean amplitude of glucose excursions
MF-DFA Multifractal DFA
MODD Mean of daily differences
MRI Magnetic resonance imaging
NICHD National Institute of Child Health and Human Development
NICU Neonatal intensive care unit
PCPC Paediatric Cerebral Performance Category
PES Pediatric Endocrine Society
PICU Pediatric intensive care unit
PMA Postmenstrual age
pp Peak to peak
RCT Randomized controlled trial
SB Seizure burden
x
SD Standard deviation
SGA Small-for-gestational-age
SWC Sleep-wake cycling
TH Therapeutic hypothermia
VLBW Very low birth weight
xi
List of Tables
Table 1.1 Recent studies of interest on outcomes following electrographic seizures and
electrographic status epilepticus in children and neonates………………………...………....
58
Table 3.1 Clinical and demographic data of the study cohort……………………………..... 81
Table 3.2 Association of glucose derangements with aEEG scores during 6-hour epochs
containing hypoglycemia or hyperglycemia………………………………………...………..
83
Table 3.3 Association of severity of glucose derangements with aEEG scores……………. 85
Table 3.4 Association of glucose variability with aEEG scores……………………………. 86
Table 3.5 Association of glucose derangements with cEEG scores during 5-minute epochs
containing hypoglycemia or hyperglycemia…………………………………………………
89
Table 3.6 Relationship of cEEG scores with mean glucose during 5-minute epochs…….… 90
Table 3.7 Association of hyperglycemia and glucose levels with cEEG background scores
in the subgroup of 8 neonates that had hyperglycemia recorded on CGM…………………...
91
Table 3.8 Association of hypoglycemia and glucose levels with cEEG background scores
in subgroup of the 5 neonates that had hypoglycemia recorded on CGM……………………
93
xii
List of Figures
Figure 1.1 Glucose derangements in neonatal encephalopathy…………………………… 7
Figure 1.2 Metabolic, endocrine and physiological changes which occur at the time of
birth………………………………………………………………………………………….
8
Figure 1.3 Range of plasma glucose concentrations during the first days after birth…….. 9
Figure 1.4 Neuroimaging after neonatal hypoglycemia in a 5-day old term neonate…….. 22
Figure 1.5 Neuroimaging features of neonatal HIE………………………………………. 30
Figure 1.6 The Medtronic iPro2 continuous glucose monitor…………………………….. 36
Figure 1.7 Standard placement for EEG and aEEG electrodes…………………………… 44
Figure 1.8 Examples of aEEG background patterns………………………………………. 46
Figure 1.9 Illustration of potential relationship between seizure burden and outcome…… 57
Figure 2.1 Summary of aEEG analysis methods………………………………………….. 77
Figure 2.2 Summary of cEEG analysis methods………………………………………….. 78
Figure 3.1 Bar graph comparing aEEG scores with glucose derangement category……... 84
Figure 3.2 Bar graph comparing cEEG background scores with glucose derangement
category……………………………………………………………………………………..
89
Figure 3.3 Box plot comparing cEEG background scores with glucose level (mmol/L) in
subgroup of patients with hyperglycemia recorded on CGM………………………………
92
Figure 3.4 Box plot comparing cEEG background scores with glucose level (mmol/L) in
subgroup of patients with hypoglycemia recorded on CGM………………………………..
94
Figure 5.1 Timeline for continuous glucose monitoring, MRI study and follow-up …….. 109
xiii
List of Appendices
Appendix 1 - Appendix Table 1 Demographic data of the study cohort; for participants
included in aEEG analysis and for those excluded for incomplete data……………………
156
Appendix 2 - Contributions………………………………………………………………. 157
1
Chapter 1
Literature Review 1
1.1 Introduction
Neonatal encephalopathy is defined as a clinical syndrome of disturbed neurological function
with many etiologies. Hypoxic–ischemic encephalopathy (HIE) after perinatal asphyxia is the
most common cause of neonatal encephalopathy and is an important cause of neonatal mortality
and morbidity (Volpe, 2012). Although therapeutic hypothermia (TH) has reduced mortality and
major neurodevelopmental disability at 18 months of age (Azzopardi, Strohm, et al., 2014;
Jacobs et al., 2013), mortality rates in moderate to severe HIE remain at approximately 25%.
Similarly, morbidity with major neurodevelopmental disability occurs in approximately 20% of
survivors of HIE (Jacobs et al., 2013). Thus, current neuroprotective strategies require further
optimization.
Both hypoglycemia and hyperglycemia are common in neonates with HIE (Basu et al.,
2016; Tam et al., 2012; Wong et al., 2013), and are potentially modifiable risk factors. In
neonates with HIE, hypoglycemia has been detected in 34% (Basu et al., 2016; Tam et al., 2012;
Wong et al., 2013) and hyperglycemia in almost 50% of neonates using intermittent glucose
testing (Basu et al., 2016). Although studies suggest that hypoglycemia and hyperglycemia may
be independently associated with worse outcomes in these neonates (Basu et al., 2016; Chouthai
et al., 2015; Salhab et al., 2004; Spies et al., 2014; Tam et al., 2012), available evidence is
conflicting and putative mechanisms remain poorly understood (Callahan et al., 1990; Nadeem et
al., 2011; Park et al., 2001; Sheldon et al., 1992; Tottman et al., 2017; Vannucci et al., 1987).
Furthermore, glucose instability and rapid correction of hypoglycemia may also be associated
with worse outcomes (Al Shafouri et al., 2015; McKinlay et al., 2015; McKinlay, Alsweiler, et
al., 2017). Despite the above observations, there is limited evidence available to guide effective
hypoglycemia management to prevent associated brain injury and neurodevelopmental sequelae
(Harding et al., 2017). The Pediatric Endocrine Society advise that a ‘safe target’ during the first
48 hours should be close to the mean for healthy newborns and above the threshold for
neuroglycopenic symptoms (2.8 mmol/L) (Thornton et al., 2015). Moreover, reference ranges in
2
healthy term newborns may not be appropriate in infants at risk for impaired metabolic
adaptation as individual susceptibility to brain injury can vary depending on comorbid conditions
and an infant’s ability to produce and use alternative fuels (Harris et al., 2015).
Electroencephalography (EEG) is an important tool for assessing brain activity since
EEG background abnormalities and recovery are important early prognostic indicators of long-
term neurodevelopmental outcome in neonates with HIE (Azzopardi & Toby study group, 2014;
Chandrasekaran et al., 2017; Dunne et al., 2017; Skranes et al., 2017; Spitzmiller et al., 2007;
Thoresen et al., 2010; van Laerhoven et al., 2013).
We hypothesize that hypoglycemia, hyperglycemia and glucose variability in neonates
with encephalopathy are associated with abnormal brain background activity and increased
electrographic seizures on amplitude-integrated EEG (aEEG) and continuous EEG (cEEG). The
objective of the research presented in this thesis is to inform potential neuroprotective glucose
management strategies. We sought to investigate how abnormalities in glucose homeostasis
correlate with brain function as measured by aEEG and cEEG in the context of neonatal
encephalopathy.
1.2 Neonatal Encephalopathy
1.2.1 Overview
Neonatal encephalopathy is defined as a clinical syndrome of disturbed neurological function,
manifested by abnormal level of consciousness or seizures, and may be associated with other
features such as difficulty initiating and maintaining respiration and depression of tone and
reflexes (D'Alton et al., 2014). There are many etiologies of encephalopathy in newborns (such
as congenital infections, genetic or neurometabolic disorders). The term HIE applies when
peripartum or intrapartum hypoxia-ischemia is identified as the likely pathogenesis. The
incidence varies due to methodological and population differences between studies. Nonetheless,
based on population data, the incidence of neonatal encephalopathy is 3 per 1000 live births
3
(95% CI 2.7 to 3.3) and the incidence of HIE is 1.5 per 1000 live births (95% CI 1.3 to 1.7)
(Kurinczuk et al., 2010).
Guidelines have been developed by the American College of Obstetricians and
Gynecologists’ Task Force on Neonatal Encephalopathy for retrospective definition of an
intrapartum event sufficient to cause cerebral palsy (CP). Neonatal encephalopathy is likely due
to acute hypoxia-ischemia when one or more of the following elements are present:
(1) Neonatal signs consistent with an acute peripartum or intrapartum event:
a. Apgar Score of less than 5 at 5 minutes and 10 minutes,
b. Fetal umbilical artery acidemia (fetal umbilical artery pH <7.0, or base deficit ≥12
mmol/L, or both),
c. Neuroimaging evidence of acute brain injury seen on brain magnetic resonance
imaging (MRI) or Magnetic Resonance Spectroscopy consistent with hypoxia–ischemia,
d. Presence of multisystem organ failure consistent with HIE.
(2) Type and timing of contributing factors that are consistent with an acute peripartum or
intrapartum event:
a. Sentinel hypoxic or ischemic event occurring immediately before or during labor and
delivery (e.g. uterine rupture, abruption placenta, umbilical cord prolapse),
b. Fetal heart rate monitor patterns consistent with an acute peripartum or intrapartum
event,
c. Timing and type of brain injury patterns based on imaging studies consistent with an
etiology of an acute peripartum or intrapartum event (including deep nuclear gray matter
or watershed cortical injury),
d. No evidence of other proximal or distal factors that could be contributing factors.
(3) Developmental outcome is spastic quadriplegia or dyskinetic cerebral palsy:
a. Other subtypes of cerebral palsy are less likely to be associated with acute intrapartum
hypoxic-ischemic events,
b. Other developmental abnormalities may occur, but they are not specific to acute
intrapartum HIE and may arise from a variety of other causes (D'Alton et al., 2014).
4
Eleven randomized controlled trials (RCTs) have shown that therapeutic hypothermia
started within 6 hours after birth for term infants with moderate to severe HIE reduces mortality
or major neurodevelopmental disability at 18 months of age (Azzopardi, Strohm, et al., 2014;
Jacobs et al., 2013). Longer-term neurodevelopmental outcomes up to 6 to 8 years old support
these 18-month findings (Azzopardi, Strohm, et al., 2014; Guillet et al., 2012). Therapeutic
hypothermia has thus become the standard of care for neonates that meet the criteria that were
used in the published trials, which were designed to select neonates with presumed HIE within
the 6-hour time window (Azzopardi et al., 2009; Gluckman et al., 2005; Jacobs et al., 2011;
Shankaran et al., 2005; Simbruner et al., 2010; Zhou et al., 2010). The fundamental
understanding that was key to development of therapeutic hypothermia was that of delayed
secondary injury. Following the initial primary injury during the acute hypoxic-ischemic (HI)
event, there is a “latent” phase of initial partial recovery lasting approximately 6 hours. In
moderate to severe injury, this period is followed by “secondary” injury characterized by
progressive failure of cerebral mitochondrial oxidative metabolism, cytotoxic edema and
seizures (Lorek et al., 1994; Wassink et al., 2014). Still, the number needed to treat with
therapeutic hypothermia for an additional beneficial outcome is seven (Jacobs et al., 2013), thus
there is still a proportion of neonates for whom this therapy is ineffective.
Mortality rates in moderate to severe HIE with therapeutic hypothermia treatment remain
at approximately 25% and with major neurodevelopmental disability occurring in approximately
20% of survivors of HIE (Jacobs et al., 2013). Major disabilities include cerebral palsy,
intellectual disability, sensorineural hearing loss and visual impairment. CP develops in 12.5-
32% of children (Jacobs et al., 2013) and intellectual disability in approximately 25% (Pappas et
al., 2015). In a secondary analysis of the National Institute of Child Health and Human
Development (NICHD) whole-body hypothermia trial, intellectual impairment occurred in 96%
of children with CP at 6 to 7 years old (intelligence quotient (IQ) <70). In children without CP,
9% had intellectual impairment and 31% had IQ scores of 70 to 84 (Pappas et al., 2015). A
variety of other deficits are reported; including specific memory and language problems, deficits
in visual-motor integration, delays in school-related activities (such as reading, spelling, and
arithmetic), lower performance and full-scale IQ scores compared with population norms,
difficulties in executive function, and motor deficits (Lindstrom et al., 2006; Perez et al., 2013;
Robertson et al., 1988; Robertson et al., 1989).
5
There is also growing evidence that mild HIE is not a benign syndrome with
observational studies showing both short- and long-term complications. Short-term consequences
in neonates classified clinically as having mild encephalopathy have included EEG documented
seizures, abnormal aEEG and EEG background (including excessive discontinuity and burst
suppression patterns), MRI abnormalities in up to half of neonates (including deep gray matter
injury, watershed injury, white matter lesions and strokes), feeding difficulties and abnormal
neurological exam on discharge (DuPont et al., 2013; Massaro, Murthy, et al., 2015;
Prempunpong et al., 2018; Walsh et al., 2017). Long-term outcome studies have demonstrated
behaviour problems (van Handel et al., 2010) and lower full-scale IQ, verbal IQ and performance
IQ (Murray et al., 2016).
Several ongoing trials are looking at adjunctive therapies to therapeutic hypothermia.
Promising agents indicated by preclinical and clinical trial data include erythropoietin (Wu et al.,
2012; Y. W. Wu et al., 2016) with a phase III trial underway (NCT03079167), melatonin (Aly et
al., 2015; Robertson et al., 2013), stem cells (Bennet et al., 2012; Cotten et al., 2014), topiramate
(Filippi et al., 2018), argon (Broad et al., 2016; Gunes et al., 2007), and allopurinol (Peeters-
Scholte et al., 2003) with the ALBINO study (NCT03162653) to start recruitment.
Despite advances in neuroprotective therapies, good resuscitation and supportive care are
fundamental for management in HIE to avoid additional brain injury. Supportive management is
directed at possible hypoxic damage to organs other than the brain; such as fluid restriction in the
setting of oliguria, monitoring for coagulopathy, closely monitoring adequacy of ventilation,
volume replacement and inotropic medications as needed to maintain blood pressure and
appropriate cerebral perfusion, and control of seizures. As well, it is important that both
hypoglycemia and hyperglycemia be avoided (Basu et al., 2016; Giesinger et al., 2017;
Lingappan et al., 2016; Martinello et al., 2017; Yager et al., 2009).
1.2.2 Prevalence of glucose derangements
Neonatal hypoglycemia (<2.6 mmol/L) is common and occurs in up to 19% of healthy
newborns (Kaiser et al., 2015). However, neonatal hypoglycemia occurs in approximately half of
at-risk neonates. This category includes infants of diabetic mothers, late-preterm (35 to <37
6
weeks’ gestation), small (<10th percentile or <2500 g) or large (>90th percentile or >4500 g)
weight at birth, or other reasons identified clinically such as poor feeding (Harris et al., 2012). A
first episode of hypoglycemia may occur after 3 normal blood glucose measurements and over
24 hours after birth (Harris et al., 2012). Additional risk factors for neonatal hypoglycemia
include intrauterine growth restriction, intrapartum fever, temperature instability in the newborn,
sepsis, public insurance compared with those privately insured, and perinatal asphyxia (DePuy et
al., 2009; Knobloch et al., 1967; VanHaltren et al., 2013; Wackernagel et al., 2016; Zhou et al.,
2015). There are also several genetic predispositions which place neonates at high risk of
hypoglycemia (Thompson-Branch et al., 2017) such as congenital hyperinsulinism, growth
hormone or cortisol deficiency, and inborn errors of metabolism.
Studies in neonates with encephalopathy have detected hypoglycemia in 13-34% using
intermittent glucose testing (Basu et al., 2016; Tam et al., 2012; Wong et al., 2013). Early
hyperglycemia was found in 48% of a cohort of neonates with HIE (Basu et al., 2016) (Figure
1.1). Hyperglycemia is also frequently encountered in preterm neonates, especially those with
very low birth weight (VLBW; birth weight <1500 g). Hays et al. reported a cohort of extremely
low birth weight neonates (ELBW; birth weight <1000 g) where hyperglycemia (>8.3 mmol/L)
was seen in 57% and glucose values above 13.9 mmol/L were seen in 32% of neonates (Hays et
al., 2006). Furthermore, clinically stable well-developing preterm infants on full enteral feeds
who are beyond their initial period of intensive care can continue to have glucose fluctuations,
with frequent hypoglycemia (<2.6 mmol/L) and hyperglycemia (>8.3 mmol/L) detected by
continuous glucose monitoring (CGM) in approximately 40 and 70% of neonates respectively
(Mola-Schenzle et al., 2015).
7
Figure 1.1 Newborns with encephalopathy have a high incidence of hypoglycemia and hyperglycemia, which may
conspire to worsen brain injury in this population already at high risk of adverse neurodevelopmental outcomes.
Hypoxic-ischemic encephalopathy is a common cause of neonatal encephalopathy.
1.3 Glucose physiology in neonates 1.3.1 Normal transitional period and challenges of defining neonatal
hypoglycemia
Many physiological, metabolic and endocrine changes must occur in order to maintain glucose
homeostasis as a neonate transitions from the fetal environment to the extra-uterine world
(Figure 1.2). In utero, fetal glucose is supplied continuously from the mother (Kalhan et al.,
1979) and fetal insulin functions primarily to regulate intrauterine growth (Hattersley et al.,
1998). Once the umbilical cord is cut, the blood glucose concentration declines during the first 2
to 4 hours of life (Srinivasan et al., 1986). This fall in glucose levels leads to decreased insulin
secretion and increase in levels of counter-regulatory hormones (e.g. glucagon, catecholamines
and glucocorticoids) (Mitanchez, 2007; Sperling et al., 1984). This then initiates endogenous
glucose production via glycogenolysis and gluconeogenesis in order to stabilize the blood
glucose concentrations (Fowden, 1980). Blood glucose concentrations in healthy term neonates
Neonatal Encephalopathy
Hypoxic-Ischemic Encephalopathy
HypoglycemiaHyperglycemia
8
reach levels comparable to fasting levels in children and adults by approximately 72 hours of life
(Hawdon et al., 1992). Brief periods of hypoglycemia commonly seen in normal newborns
during the transition from fetal to extra-uterine life are referred to as transitional neonatal
hypoglycemia (Stanley et al., 2015).
Figure 1.2 The metabolic, endocrine and physiological changes which occur at the time of birth to allow a normal
term newborn to adapt to the change from the fetal environment to the extra-uterine world. Figure originally
published by Guemes et al. (2016) in Arch Dis Child; reproduced with permission from BMJ Publishing Group Ltd.
During the initial transitional phase in the hours after birth, glucose values show marked
variability between neonates and range between 1.4 mmol/L and 6.2 mmol/L in healthy,
appropriate for gestational age (AGA) neonates (Acharya et al., 1965; Diwakar et al., 2002;
Hawdon et al., 1992; Heck et al., 1987; Hoseth et al., 2000; Srinivasan et al., 1986) with a mean
9
initial glucose between 3.1 to 3.6 mmol/L (Stanley et al., 2015) (Figure 1.3). It has also been
shown that blood glucose concentrations in healthy neonates are lower in those being breast fed
than bottle fed, yet ketone body concentrations are higher with breastfeeding (Hawdon et al.,
1992). Glucose concentrations are also relatively stable and unaffected by the timing of the
initial feed (Sweet et al., 1999; Zhou et al., 2017).
A
B
Figure 1.3 Range of plasma glucose concentrations during the first days after birth measured in (A) 223 full-term
appropriate weight for gestational age breastfed neonates (B) 200 term appropriate weight for gestational age and
exclusively breastfed neonates at 3 h, 6 h, 24 h and 72 h of age. A) Reproduced from Hoseth et al. (2000) Arch Dis
Child Fetal Neonatal Ed with permission from BMJ Publishing Group Ltd. B) Figure was adapted from Diwakar et
al. (2000) and originally published by Guemes et al. (2016) in Arch Dis Child; reproduced with permission from
BMJ Publishing Group Ltd.
In older children, ketone synthesis and gluconeogenesis is noted in response to low
glucose levels. However, mechanisms associated with fasting adaptations are not fully developed
at the time of birth. Early studies have shown that during hypoglycemia in term AGA neonates,
ketone levels were low despite elevated levels of fatty acid precursors and higher levels of the
glucose precursors lactate and alanine (Hawdon et al., 1992; Stanley et al., 1979). This
suppression of ketones seen in term neonates is similar to that seen in hyperinsulinemic
hypoglycemia (Stanley et al., 1976). It has also been shown that plasma insulin concentrations
10
are not completely suppressed at the low glucose concentrations seen during transitional neonatal
hypoglycemia in normal newborns (Adam et al., 1968; Hawdon, Aynsley-Green, Alberti, et al.,
1993). Furthermore, transitional neonatal hypoglycemia is also associated with an
inappropriately large glycemic response to glucagon or epinephrine suggestive of inappropriate
retention of liver glycogen stores (Cornblath et al., 1958). This constellation of metabolic and
hormonal features (inefficient ketone production, incomplete suppression of insulin, and
inappropriate retention of glycogen stores) supports the idea that transitional neonatal
hypoglycemia in normal AGA newborns resembles hyperinsulinemic hypoglycemia due to
congenital hyperinsulinism in which the set point for suppression of insulin secretion is reduced
(Stanley et al., 2015).
Controversy remains as to whether the newborn brain is more or less vulnerable to low
glucose levels (Kim et al., 2005; Vannucci et al., 2001). A number of putative physiologic
mechanisms may protect the brain from such low glucose levels. These include a compensatory
increase in cerebral blood flow (Ichord et al., 1994; Mujsce et al., 1989; Pryds et al., 1988; Skov
et al., 1992), enhanced glucose extraction from blood into the brain and low cerebral energy
demands (Mujsce et al., 1989), and the availability and utilization of alternative cerebral fuels
(Hellmann et al., 1982; Vannucci et al., 1981). However, there is no evidence that the glucose
concentration which impairs brain function and leads to neuroglycopenic symptoms is different
in neonates than in older children and adults (Thornton et al., 2015). Neonates also have a
proportionally larger brain relative to their body size and almost all of the daily total body
glucose utilization can be accounted for by the brain (Bier et al., 1977). Although glucose is the
primary substrate metabolized by the brain, there are alternate fuels which can be utilized
including pyruvate, lactate, and ketones (Cremer et al., 1979; Hellmann et al., 1982; Wyss et al.,
2011; Zhang et al., 2013). In neonates, ketone use by the brain is directly proportional to the
plasma concentration of ketones (Bougneres et al., 1986) and lactate may also be used as fuel by
the brain in the neonatal period (Costello et al., 2000; Vannucci et al., 1981; Vannucci et al.,
2000). Yet, in high-risk neonates there may be inefficient ketone production and plasma lactate
levels may not be sufficiently elevated, making them particularly vulnerable to hypoglycemia-
induced brain injury (Harris et al., 2015; Hawdon et al., 1992; Stanley et al., 1979).
Failure of neonatal metabolic adaptations can lead to hyperglycemia as well. Etiologies
for hyperglycemia in neonates may be partially explained by several mechanisms. These include
11
an apparent inefficiency of glucose homeostasis, with studies demonstrating that with
intravenous glucose or insulin infusions, there is variability and lack of suppression of
gluconeogenesis in neonates and in particular preterm neonates (Chacko et al., 2011; Cowett et
al., 1983; Sunehag et al., 1994). Animal studies have also shown newborn pups are less efficient
at handling excess glucose due to decreased hepatic and peripheral sensitivity to insulin (Varma
et al., 1973). Treatments such as epinephrine can increase glucose levels (Cornblath, 1955;
Desmond et al., 1950) and hypothermia may delay the return of glucose concentrations back to
normal levels (Baum et al., 1968; Desmond et al., 1950). A study of 28 neonates undergoing
investigations for hypo- or hyperglycemia showed that blood glucose concentrations are
predominantly influenced by the rate of glucose administration, as there are incomplete internal
control mechanisms. Similar to hypoglycemia, there was a variable insulin response to
hyperglycemia seen, and glucoregulatory hormones (insulin and glucagon) were not adequately
controlling hepatic glucose production (Hawdon, Aynsley-Green, Bartlett, et al., 1993). How
these mechanisms may be altered by conditions such as HIE remains unclear.
1.3.2 Glucose physiology in neonatal encephalopathy
Intrapartum asphyxia disturbs the typical physiological changes described above and increases
the likelihood of hypoglycemia through several mechanisms. Intrapartum asphyxia leads to
increased glucose metabolism in all tissues and is possibly associated with decreased placental
glucose transport to the fetus. Prolonged anaerobic metabolism leads to increased lactate, a fall in
pH and a decrease in high energy phosphates. The increased glucose consumption leads to
depletion of endogenous glucose stores, especially in the liver, and consequent hypoglycemia
once the neonate is delivered (Randall, 1979; Vannucci et al., 2001; Volpe, 2008). Reduced
perfusion and impairment of liver function may also have direct effect on substrate supply.
Furthermore, enteral feeding with breast milk which promotes ketogenesis is often delayed
(Hawdon et al., 1992).
Additionally, neonates with HIE may behave as neonates with hyperinsulinemic
hypoglycemia or transitional neonatal hypoglycemia, during which ketogenesis is suppressed
(Collins et al., 1984; Hoe et al., 2006). Therefore, as mentioned above, it cannot be assumed that
12
ketones are available as an alternative fuel to support brain metabolism during hypoglycemia,
leaving them particularly vulnerable to hypoglycemia-induced brain injury (Harris et al., 2015;
Harris et al., 2011; Stanley et al., 2015). Lactate concentrations may be high enough to
potentially be a source of alternative fuel (Harris et al., 2015), however levels may also not be
sufficiently elevated to compensate for low glucose in theses vulnerable neonates (Harris et al.,
2015; Harris et al., 2011).
Animal studies demonstrate that hypoglycemia during hypoxic-ischemic (HI) injury can
be detrimental (Chang et al., 1999; Kim et al., 1994). Studies by Vannucci et al. in newborn rats
have demonstrated the detrimental effect of hypoglycemia on survival following anoxia with
100% nitrogen. Hypoglycemic animals survived only one-tenth as long as controls with normal
glucose following anoxia and the vulnerability was dependent on the amount of endogenous
cerebral glucose reserves at the time of insult. Increased vulnerability correlated with accelerated
depletion of endogenous high energy phosphate levels. Based on their findings they concluded
that glucose levels that are sufficient to meet energy demands in well neonates may be
inadequate during periods of reduced systemic or cerebral oxygenation (Vannucci et al., 1978).
Similar findings were shown in newborn dogs when insulin-induced hypoglycemia was
combined with asphyxia, there was a more rapid exhaustion of high-energy phosphate reserves.
Furthermore, blood glucose levels were not related to cerebral perfusion after asphyxia, and
rather correlated with systemic blood pressure regardless of blood glucose levels (Vannucci et
al., 1980). Thus, although compensatory increase in cerebral blood flow may be a protective
mechanism during hypoglycemia alone (Ichord et al., 1994; Mujsce et al., 1989; Pryds et al.,
1988; Skov et al., 1992) cerebral vascular autoregulation can be lost during hypoxia or asphyxia
and cerebral blood flow becomes pressure passive (Freeman et al., 1968; Massaro, Govindan, et
al., 2015). Hypoglycemia may also further impair cerebrovascular autoregulation, as
demonstrated in rats and newborn dogs, which may further increase vulnerability of the
watershed regions to ischemia in HIE (Anwar et al., 1988; Siesjo et al., 1983). It is important to
note that animal studies use insulin-induced rather than calorie-restricted hypoglycemia, which is
non-physiological and may lead to distinctly different uses of alternative fuels. Studies have also
demonstrated high levels of ketone bodies available as metabolic fuel to the brain in suckling rats
(Cremer et al., 1979; Hawkins et al., 1971).
13
Hyperglycemia after hypoxia-ischemia may occur due to reduced net metabolism of
severely damaged tissues (Jensen et al., 2006) or prolonged elevation of stress hormones after
asphyxia, which may be further prolonged by TH (Davidson et al., 2008). However, regarding
effects of hyperglycemia in this context, animal studies in neonatal models for HIE have been
contradictory with early studies in neonatal rat models suggesting hyperglycemia prior, during,
or after HI did not lead to accentuated damage and was perhaps neuroprotective (Callahan et al.,
1990; Hattori et al., 1990; Vannucci, Brucklacher, et al., 1996; Voorhies et al., 1986). There were
also beneficial effects seen with pretreatment with glucose in newborn lamb and rhesus monkeys
prior to asphyxia (Dawes, Jacobson, et al., 1963; Dawes, Mott, et al., 1963; Rosenberg et al.,
1990). Plus, hyperglycemia was shown to reduce energy depletion during HI in piglets (Laptook
et al., 1992). Conversely, a study examining hyperglycemia after HI injury in newborn rats
demonstrated more severe neuronal injury (Sheldon et al., 1992). Hyperglycemia during
hypothermic circulatory arrest in newborn dogs increased brain injury in specific regions,
especially the caudate, amygdala and brainstem (Vannucci, Rossini, et al., 1996). Hyperglycemia
prior to ischemia in fetal sheep (Petersson et al., 2004) and during HI injury in newborn piglets
also worsened brain injury (Chang et al., 1998; LeBlanc et al., 1993), whereas a study with
hyperglycemia after HI injury did not (Leblanc et al., 1994). Further studies with newborn piglet
models of HI also suggest that hyperglycemia after HI interferes with recovery of brain cell
membrane function and energy metabolism (Park et al., 2001). There may be methodological and
species differences. It is argued that the newborn piglet is a superior animal model of neonatal
HIE as the piglet brain is comparable in glucose metabolism and maturity to human neonates at
birth (Dobbing et al., 1979), whereas rat pups have limited ability for brain glucose metabolism
(Booth et al., 1980; Fuglsang et al., 1986; Nehlig et al., 1988). Additionally, understanding the
effects of blood glucose after HI is more clinically practical as it may not be possible to modify
glucose levels during the primary asphyxial insult in neonates.
Hyperglycemia is also common in preterm neonates thus animal studies have examined
hyperglycemia in this context and demonstrated a detrimental effect of hyperglycemia. A study
in preterm Sprague-Dawley rats demonstrated that increasing severity of hyperglycemia was
associated with greater severity of brain injury and apoptosis, particularly in the hippocampus.
Severity of hyperglycemia was also associated with increased incidence of mortality and
decreased brain density and weight gain in survivors (Tayman et al., 2014). A study by Lear et
14
al. in preterm sheep (equivalent to 28 to 32 weeks gestational age [GA]) compared fetuses that
received glucose infusions prior to asphyxia to maternal injection of dexamethasone and controls
who received saline infusions, in order to examine the hypothesis that dexamethasone treatment
increased asphyxial brain injury at least in part by producing hyperglycemia. Both maternal
dexamethasone treatment and glucose infusions prior to asphyxia were associated with severe
cystic brain injury in multiple brain regions (compared to diffuse injury in control animals). They
were also both associated with improved neurophysiological adaptation to asphyxia but
worsened neurophysiological recovery, and higher glucose levels were significantly correlated
with greater numbers of seizures (Lear et al., 2017).
1.3.3 Glucose management guidelines
1.3.3.1 Hypoglycemia management
There is controversy and variable definitions of both neonatal hypoglycemia and hyperglycemia
in the various available clinical guidelines for glucose management in neonates (Cornblath et al.,
2000; Harding et al., 2017; Tin, 2014). The National Institutes of Child Health and Human
Development workshop reviewed the evidence and identified many gaps in knowledge regarding
neonatal hypoglycemia and the lack of an evidence based definition in neonates, as well as the
need to determine the effects of comorbidities (e.g. HI, sepsis) and their contribution to adverse
outcomes with low plasma glucose (Hay et al., 2009).
On the basis of neurophysiological and neurodevelopmental outcome studies, a
commonly used definition of neonatal hypoglycemia is less than 2.6 mmol/L (Koh et al., 1988;
Lucas et al., 1988). However, the level and duration of hypoglycemia associated with brain
injury and neurologic sequelae has not been established and the individual neonates’ response to
inadequate delivery of glucose to the target organ (e.g. the brain) depends on their physiological
and hormonal responses, as well as concurrent pathology. Consequently, it is difficult to define a
single blood glucose concentration cutoff for hypoglycemia. The blood glucose concentration
must be interpreted within the clinical context, whether there was recent antecedent
hypoglycemia, and considering the presence of alternative fuels and intermediate metabolites
(Guemes et al., 2016; Thornton et al., 2015). Additionally, reference ranges in healthy term
15
newborns may not be appropriate in infants at risk for impaired metabolic adaptation and at risk
for adverse neurodevelopment, such as with perinatal asphyxia (Harding et al., 2017).
In 2000, Cornblath suggested “operational thresholds” with different cutoffs proposed if
the neonate was ‘symptomatic’ or not (Cornblath et al., 2000). Clinical hypoglycemia refers to
when low blood glucose concentrations are associated with symptoms and signs attributable to
glucose deficiency in the target tissues and which are improved by the administration of glucose
(Vannucci et al., 2001). Neurogenic symptoms occur due to sympathetic nervous system
activation and include tachycardia, vomiting, sweating, pallor, tremulousness, temperature
instability, and irritability. Neurogenic signs and symptoms occur at higher glucose levels than
neuroglycopenic symptoms. Neuroglycopenic symptoms occur when there is impaired brain
energy metabolism. Symptoms are nonspecific and include lethargy, poor feeding, irritability,
hypotonia, apnea, seizures and coma (Mitrakou et al., 1991; Pildes et al., 1967). The glucose
level at which symptoms occur may depend on the occurrence of previous episodes of
hypoglycemia and the presence of alternative fuels, such as ketones and lactate (Guemes et al.,
2016). Although children and adults can communicate when they have symptoms of
hypoglycemia, neonates cannot. Furthermore, these symptoms are nonspecific and may occur
due to many possible etiologies in sick neonates. In the study by Lucas et al., 222 of the 433
neonates with plasma glucose concentrations below 2.6 mmol/L had symptoms of hypoglycemia
(e.g. recurrent apnea, vomiting, jitteriness and seizures), however these symptoms were equally
prevalent in the 228 neonates without hypoglycemia (Lucas et al., 1988). Also, a recent
prospective study be Harris et al. found that almost 80% of neonates that became hypoglycemic
showed no clinical signs (Harris et al., 2012).
The American Academy of Pediatrics (AAP) and Canadian Pediatric Society (CPS)
provide advice for screening and management of transitional hypoglycemia in the first 24 to 36
hours after birth. The AAP guidelines for management of newborns at risk born at ≥ 34 weeks’
GA advises thresholds which are dependent on postnatal age; ranging from 1.4 to 2.2 mmol/L in
the first hours of life, 1.9 to 2.5 mmol/L from 4 to 24 hours, and 2.5 mmol/L for babies over 24
hours old. In babies with clinical signs of hypoglycemia, the advised threshold for intervention is
a blood glucose concentration of ≤ 2.2 mmol/L (Adamkin et al., 2011). The CPS advises that at-
risk babies require intervention if they have a blood glucose concentration <1.8 mmol/L at 2
hours of age or <2.0 mmol/L after subsequent feeding or if blood glucose concentration are
16
repeatedly below 2.6 mmol/L despite subsequent feeding. Though symptomatic neonates should
be treated immediately for a blood glucose concentration <2.6 mmol/L (Aziz K et al., 2004).
Whereas the AAP and CPS guidelines use the lower ranges of glucose concentrations
found in asymptomatic infant, the Pediatric Endocrine Society (PES) approach uses mean
glucose values of healthy newborns and focuses on identification of persistent hypoglycemia
when hypoglycemia persists beyond 48 hours of life. They advise that a ‘safe target’ during the
first 48 hours should be close to the mean for a healthy newborn and above the threshold for
neuroglycopenic symptoms (2.8 mmol/L). This target should be increased after 48 hours to 3.3
mmol/L to be above the threshold for neurogenic symptoms and close to the target for older
infants and children (Thornton et al., 2015).
The PES guidelines take into consideration that neurogenic symptoms are perceived in
older children and adults at plasma glucose concentrations <3.0 mmol/L and cognitive function
is impaired <2.8 mmol/L (Boyle et al., 1988; Schwartz et al., 1987). They also take into account
that although attempts to establish thresholds for neuroglycopenic symptoms in response to
hypoglycemia in neonates have not been successful, there is no reason to assume lower threshold
for intervention in neonates than children and adults, particularly considering they have a
proportionally larger demand for glucose by the brain (Harding et al., 2017; Stanley et al., 2015;
Thornton et al., 2015). The PES also considers that in hypoketotic conditions such as
hyperinsulinism, ketones and lactate are not available in sufficient concentrations to serve as an
alternative fuel for the brain and thus there is greater risk of brain energy failure and
hypoglycemia-induced brain damage (Thornton et al., 2015).
At-risk categories for screening in the CPS and AAP guidelines include small-for-
gestational-age (SGA), large-for-gestational age (LGA), infants of diabetic mothers (IDM) or
preterm neonates (Adamkin et al., 2011; Aziz K et al., 2004). The PES expands the at-risk
categories that should be screened to also include: perinatal stress (birth asphyxia, ischemia,
cesarean section for fetal distress, maternal preeclampsia/eclampsia or hypertension, meconium
aspiration syndrome, erythroblastosis fetalis, polycythemia, hypothermia), premature or
postmature delivery, family history of genetic forms of hypoglycemia, congenital syndromes
(e.g., Beckwith Wiedemann) and neonates with abnormal physical findings (e.g., midline facial
malformations, microphallus).
17
Depending on glucose level and symptoms, management of hypoglycemia includes
feeding or IV dextrose (Aziz K et al., 2004; Lilien et al., 1980). Another option is oral dextrose
gel which has been shown to be effective to improve glucose concentrations in late preterm and
term neonates, and reduce admission to the neonatal intensive care unit (NICU) and formula-
feeding at 2 weeks of age (Harris et al., 2013) and appeared safe at 2-year follow-up (Harris et
al., 2016). For more severe hypoglycemia, treatment options may also include glucagon
(Miralles et al., 2002), diazoxide (Drash et al., 1968) and octreotide (Thornton et al., 1993).
Experts suggest that a balance between the different approaches using the lower threshold
recommended by the AAP and CPS, while maintaining awareness of the need to exclude
persistent hypoglycemia when low glucose concentrations persist beyond 48 hours, may be
considered to reduce concerns of unnecessary screening and potentially unnecessary treatments,
while still identifying persistent hypoglycemia (Harding et al., 2017; Tin, 2014).
1.3.3.2 Hyperglycemia management
Hyperglycemia may be considered as part of the physiological response to stress (Davidson et
al., 2008; Dungan et al., 2009), however there is accumulating evidence that hyperglycemia is
associated with greater mortality and morbidity in preterm neonates (Blanco et al., 2006; Hall et
al., 2004; Hays et al., 2006; Heimann et al., 2007; Kao et al., 2006; Zarif et al., 1976) and with
unfavorable outcomes in term neonates with HIE (Basu et al., 2016; Chouthai et al., 2015; Spies
et al., 2014).
There are no clear guidelines for management of hyperglycemia in neonates, and in
particular there is a lack of studies to guide management decisions of hyperglycemia in neonates
with HIE. As such, there are wide variations in the definition used for hyperglycemia in neonates
as well as large variations in the criteria used in NICUs for starting insulin in neonates with
hyperglycemia.
There are a few studies that have assessed management of hyperglycemia in preterm
neonates, but still, treatment approaches differ widely between centers. A survey in Australasia
found that criteria for commencing insulin in preterm neonates varied in the different neonatal
18
units between glucose levels of 8 to 15 mmol/L (Alsweiler et al., 2007). Similarly, studies
looking at outcomes in preterm neonates with hyperglycemia have used varying threshold;
including >8.3 mmol/L (Blanco et al., 2006; Hays et al., 2006; Heimann et al., 2007), >8.5
mmol/L (Ertl et al., 2006; Slidsborg et al., 2018), >10 mmol/L (Beardsall et al., 2008; Kao et al.,
2006), or >12 mmol/L (Manzoni et al., 2006).
Two small randomized trials of continuous insulin infusion for management of
hyperglycemia in ELBW infants found that it appears safe and helps to maintain normal caloric
intake and weight gain (Collins et al., 1991; Meetze et al., 1998). However, only short-term
outcomes were measured in those trials and a 2011 Cochrane review found insufficient evidence
to determine the effects of insulin treatment on death or major morbidities (Bottino et al., 2011).
A randomized, controlled non-blinded trial compared tight glycemic control to standard practice
in neonates that became hyperglycemic (>8.5 mmol/L) found that tight glycemic control with
insulin was associated with increased risk of hypoglycemia as well as lower rates of linear
growth despite greater head circumference and growth and weight gain (Alsweiler et al., 2012).
The American Society for Parenteral and Enteral Nutrition (A.S.P.E.N.) clinical guidelines for
hyperglycemia in neonates receiving parenteral nutrition suggest that excess energy and dextrose
delivery be avoided and fat emulsion be added to the parenteral nutrition infusion. They make a
strong recommendation against the use of early insulin therapy to prevent hyperglycemia
(Arsenault et al., 2012).
“Early” continuous insulin therapy has been investigated in preterm neonates, in which
insulin is started within 24 hours of birth and supported by 20% dextrose in order to prevent
hyperglycemia. The NIRTURE study was an international randomized, controlled trial in VLBW
infants comparing “early” continuous insulin to standard neonatal care. The efficacy of glucose
control was assessed by continuous glucose monitoring. The authors found that early insulin
therapy may lead to a significant improvement in glucose control and an increase in
energy intake during the first week of life, but with an increased risk of hypoglycemia. There
were no differences between the two study groups with regard to the primary outcome of
mortality at the expected date of delivery or to morbidity outcomes. The study was discontinued
early because of concerns about futility with regard to the primary outcome (Beardsall et al.,
2008). The pilot study had shown that treatment with early insulin therapy improved glucose
19
control and increased weight gain and leg length growth, with no difference in hypoglycemia
between the two groups (Beardsall et al., 2007).
Similarly, potential benefits of insulin therapy in critically ill pediatric and adult patients
also remains unclear. Several studies have looked at intensive insulin therapy in the pediatric
ICU (PICU) as hyperglycemia has also been associated with adverse outcome in critically ill
children (Falcao et al., 2008; Faustino et al., 2005; Srinivasan et al., 2004; Wintergerst et al.,
2006; Y. Wu et al., 2016). A prospective randomized controlled study of 700 critically ill
pediatric patients compared intensive insulin therapy (targeted to age-adjusted normoglycemia)
to conventional therapy (continuous insulin therapy started after 2 blood glucose concentrations
>11.9 mmol/L). The cohort included 317 infants less than 1 year of age (45% of cohort) and
three-quarters of patients were admitted for cardiac surgery for congenital heart defects.
Intensive insulin therapy was shown to improve mortality and morbidity (protected the
cardiovascular system, prevented secondary infections and attenuated the inflammatory
response, and reduced the duration of PICU stay) but increased hypoglycemia (Vlasselaers et al.,
2009). Long-term follow-up was obtained in 569 of the patients and demonstrated that tight
glucose control did not affect the full-scale IQ (FSIQ) score compared to usual care, although
tight glucose controlled improved motor coordination and cognitive flexibility. Episodes of
hypoglycemia secondary to tight glucose control were not associated with worse neurocognitive
outcomes (Mesotten et al., 2012).
The finding in a randomized trial that correction of hyperglycemia improves morbidity
and mortality may be suggestive of a causal relationship between hyperglycemia and worse
outcomes. However, other trials have found contradictory results. Two multicenter trials of
hyperglycemia control after cardiac surgery or admission to the PICU in critically ill neonates
and children (the Control of Hyperglycaemia in Paediatric Intensive Care [CHiP] trial and the
Safe Pediatric Euglycemia in Cardiac Surgery [SPECS]) trial found that tight glycemic control in
critically ill children had no significant effect on major clinical outcomes (Agus et al., 2012;
Macrae et al., 2014) and potentially a higher incidence of hypoglycemia with tight glucose
control (Macrae et al., 2014). Prospective follow-up was performed at 9-18 months in children in
the SPECS trial using the Bayley scale of infant development (BSID), Adaptive Behavior
Assessment System, Ages and Stages Questionnaire, and Brief Infant Toddler Social-Emotional
Assessment. No significant differences were seen in neurodevelopment outcome scores with
20
tight glycemic control compared to standard care. However, moderate to severe hypoglycemia
(<2.8 mmol/L) was associated with worse functioning in the cognitive, language, and motor
domains, including after adjusting for potential confounders, such as complexity of cardiac
physiology and duration of cardiac ICU stay (Sadhwani et al., 2016). Additionally, the recent
Heart and Lung Failure–Pediatric Insulin Titration (HALF-PINT) trial investigated tight
glycemic control in children with cardiovascular or respiratory failure excluding patients who
had undergone cardiac surgery. They also found no significant between-group difference in ICU-
free days or in any secondary outcomes, and the study was stopped early due to a low likelihood
of benefit and evidence of the possibility of harm (nonsignificant higher mortality, health care-
associated infection profile, and risk of severe hypoglycemia) in the lower glucose target group
(Agus et al., 2017).
Equally, in adults with brain injury, hyperglycemia has repeatedly been shown to be
associated with worse clinical outcomes (Bruno et al., 2002; Lam et al., 1991; Passero et al.,
2003; Wartenberg et al., 2006), yet controversy exists regarding glycemic management.
Randomized studies in critically ill adults in intensive care units had demonstrated that tighter
glucose control may improve clinical outcomes (Van den Berghe et al., 2006; van den Berghe et
al., 2001). However, not all studies showed a benefit. An international randomized trial showed
intensive glucose control increased mortality among adults in the ICU (Finfer et al., 2009). A
network meta-analysis did not find any benefit of tight glycemic control for mortality (Yamada
et al., 2017). Furthermore, studies in brain injured adults using cerebral microdialysis (which
allow in vivo detection of neurochemical changes in interstitial tissue of the brain) are
concerning because the data indicate that excessive insulin-induced decrease in glucose in order
to achieve tight glycemic control may compromise brain energy metabolism and worsen
neuronal injury (Oddo et al., 2008; Vespa et al., 2006).
In summary, the literature suggests that while tight glycemic control shows no clear
benefit, the appropriate threshold for treatment of hyperglycemia remains unclear. As well, there
is variability between centers in treatment approaches to hyperglycemia. Further understanding
of the long-term effects of hyperglycemia and different management approaches are crucial to
inform treatment decisions in critically ill neonates at high risk for neurodevelopmental
impairments.
21
1.4 Neurodevelopmental outcomes following neonatal glucose
derangements
1.4.1 Overview
There is agreement in the literature that recurrent and severe hypoglycemia can cause brain
injury. Neuroimaging abnormalities are seen in 18-39% of neonates after severe and recurrent
hypoglycemia (Kinnala et al., 1999; Tam et al., 2008). Severe and recurrent hypoglycemia in
term neonates leads to a specific neuroimaging pattern which includes parieto-occipital white
matter abnormalities, abnormal signal in the thalamus and basal ganglia (Alkalay et al., 2005;
Barkovich et al., 1998; Gataullina et al., 2013; Kinnala et al., 1999; Murakami et al., 1999; Spar
et al., 1994; Traill et al., 1998) and restricted diffusion in the parieto-occipital lobes, corpus
collosum and optic radiations on diffusion-weighted imaging (DWI) (Filan et al., 2006; Kim et
al., 2006; Tam et al., 2008) (Figure 1.4). However, Burns et al. reported a retrospective cohort of
35 infants with neonatal hypoglycemia and more diffuse injury, including cortical watershed
parasagittal lesions, focal infarction and hemorrhage. Still, 29% of the cases had a predominantly
posterior pattern of brain injury. Their cohort may represent a mixture of etiologies; as 40% had
IUGR, and 11% had endocrine disorders. As well, delivery by emergency cesarean section, need
for resuscitation in the delivery room, and/or cord pH 7.0 -7.1 was more frequent in cases than
controls (Burns et al., 2008). Of note, neuroimaging findings following hypoglycemia may not
be persistent and the children may have no apparent neurodevelopmental sequelae at follow-up
(Kinnala et al., 1999; Tam et al., 2008).
22
Figure 1.4 Neuroimaging findings after neonatal hypoglycemia in a 5-day old term neonate demonstrating bilateral
parieto-occipital diffusion restriction which is extending to the left posterior temporal lobe and pulvinar nuclei of the
thalami bilaterally but more marked on the left, and into the posterior limb of the internal capsule. (A) DWI and (B)
ADC maps
Long-term neurodevelopmental outcomes of neonatal hypoglycemia include epilepsy,
microcephaly, global developmental delay, cerebral palsy and intellectual disability (Koivisto et
al., 1972; Per et al., 2008; Udani et al., 2009; Yalnizoglu et al., 2007) and visual outcomes
including cortical visual deficits (e.g. cortical blindness, homonymous hemianopsia)
(Karimzadeh et al., 2011; Tam et al., 2008; Yalnizoglu et al., 2007) have been described.
Symptomatic epilepsy secondary to hypoglycemia is often focal (most frequently occipital lobe
epilepsy) and usually has a good prognosis. However, occasionally neonatal hypoglycemia can
lead to refractory seizures and epileptic encephalopathy (including epileptic spasms, atonic and
tonic seizures). There has not been a clear relationship identified between seizure outcome and
the severity, duration or cause of neonatal hypoglycemia, or the presence of neonatal seizures
(Alrifai et al., 2014; Arhan et al., 2017; Caraballo et al., 2004; Fong et al., 2014; Kumaran et al.,
2010; Montassir et al., 2010; Yang et al., 2016). Occurrence of acute symptomatic seizures in the
A B
23
neonatal period may be associated with worse long-term neurodevelopmental outcomes
(Koivisto et al., 1972; Pildes et al., 1974).
Duration, frequency, and severity of hypoglycemia seem to play a role, although the
specific level and duration of hypoglycemia that leads to brain injury or neurodevelopmental
sequelae is not established. Severe and recurrent episodes of hypoglycemia are more predictable
of long term sequelae than single hypoglycemic episodes (Duvanel et al., 1999; McKinlay,
Alsweiler, et al., 2017). Tam et al. found that hypoglycemia measured on 2 or more days was
associated with cortical visual deficits but none of the neonates with hypoglycemia measured on
only 1 day had cortical visual deficits in long-term follow-up (Tam et al., 2008). Longer duration
of hypoglycemia has also been directly related to neurodevelopmental outcome (Montassir et al.,
2009; Singh et al., 1991).
Regarding hyperglycemia, studies in neonates have focused predominantly on outcomes
in preterm neonates and term neonates with HIE, and will be discussed further below.
1.4.2 Short term outcomes
It was recognized almost 60 years ago that low blood glucose levels in SGA and preterm
neonates were associated with seizures (Cornblath et al., 1959). Further studies have also
reported symptomatic seizures in close temporal relationship to documented hypoglycemia
(Filan et al., 2006; Kim et al., 2006; Tam et al., 2008; Traill et al., 1998).
Short-term effects also include neurophysiologic abnormalities detected using evoked
potentials. Abnormal evoked potentials during hypoglycemia were reported by Koh et al. and
supported the common clinical use of 2.6 mmol/L as a cutoff for hypoglycemia. They performed
evoked potentials (sensory evoked potentials and brainstem auditory evoked potentials) during
spontaneous and induced hypoglycemia in a cohort of 5 neonates and 12 children. Abnormalities
in the evoked potentials occurred over a range of glucose concentrations between 0.7 to 2.5
mmol/L, suggesting individuals have varying levels of susceptibility. No subjects whose blood
glucose concentration was maintained at 2.6 mmol/L or greater had abnormalities noted in their
evoked potentials (Koh et al., 1988). A subsequent study by Cowett et al. of 50 neonates (33-40
weeks GA) with glucose levels between 1.38 to 6.83 mmol/L found no correlation with
24
brainstem auditory evoked potentials. Yet the blood glucose concentrations were predominantly
within the normal range and the single glucose concentration of 1.38 mmol/L was associated
with a prolonged wave V latency in the term neonate (Cowett et al., 1997). In a cohort of term
neonates admitted to NICU with a variety of perinatal conditions (excluding HIE), hypoglycemia
was the main risk factor for reduction in amplitude of wave I on brainstem auditory evoked
responses. The main conditions in the cohort included hypotension, hypoglycemia, meconium
aspiration syndrome, sepsis, metabolic acidosis, hemolytic or non-hemolytic hyperbilirubinemia,
and pneumonia (Jiang et al., 2013). Occipital restricted diffusion after neonatal hypoglycemia
has also been associated with abnormal visual evoked potentials within 1 week after birth (Tam
et al., 2008).
Several studies have shown hyperglycemia to be associated with worse short-term
outcomes in preterm neonates than hypoglycemia. Hyperglycemia in preterm neonates is
associated with an increased risk of mortality (Alexandrou et al., 2010; Hays et al., 2006;
Heimann et al., 2007; Kao et al., 2006), grade 3 or 4 intraventricular hemorrhage (Bermick et al.,
2016; Hays et al., 2006), sepsis (Kao et al., 2006), necrotizing enterocolitis (Hall et al., 2004;
Kao et al., 2006), retinopathy of prematurity (Blanco et al., 2006; Ertl et al., 2006; Mohamed et
al., 2013; Mohsen et al., 2014; Slidsborg et al., 2018), and greater length of hospital stay (Hays et
al., 2006). Data are lacking regarding short-term outcomes of hyperglycemia in term neonates.
1.4.3 Long term outcomes
1.4.3.1 Well babies
As discussed, severe and persistent hypoglycemia can cause acute seizures and brain
injury in neonates, however, the long-term significance of early asymptomatic or transitional
hypoglycemia remains controversial. Neurodevelopmental difficulties may emerge as children
get older, but may not be evident at initial follow-up at younger ages. Therefore, long-term
follow-up studies are essential. A systematic review in 2006 (Boluyt et al., 2006) identified 18
studies spanning over 40 years, of which only 2 studies were of sufficient methodological quality
for quantitative analysis, one study in term (Brand et al., 2005) and the other in preterm neonates
(Lucas et al., 1988).
25
Some but not all studies have found an association between neonatal hypoglycemia with
long-term outcome. Still, presence of hypoglycemia in the neonatal period has been associated
with worse long-term neurodevelopmental outcome in several studies. A study in 28 children of
diabetic mothers assessed long-term outcome at 8 years of age. Thirteen of the children had
hypoglycemia (<1.5 mmol/L) in the neonatal period, and most were asymptomatic. The children
with neonatal hypoglycemia exhibited significantly more difficulties on a screening test for
attention deficit disorder and were more frequently reported to be hyperactive, impulsive, and
easily distracted. On psychological assessment, they had a lower total developmental score than
the normoglycemic children born to diabetic mothers as well as control age-matched healthy
children (Stenninger et al., 1998). In a recent prospective study, 72 hypoglycemic (plasma
glucose <2.8mmol/L) and 70 weight and gestational age matched control neonates (GA >32
weeks) were enrolled from the NICU in the first week of life. There were significantly lower
levels of blood glucose in the symptomatic than asymptomatic hypoglycemia group. The mean
motor and mental developmental quotients were significantly lower in neonates with both
symptomatic and asymptomatic hypoglycemia than euglycemic infants. A cutoff of 2.2 mmol/L
was associated with lower scores. Neurodevelopment was assessed at 6 and 12 months old with
the Developmental Assessment Score for Indian Infants using the DASII scoring system which is
an Indian adaptation of the BSID, and which was administered by the un-blinded principal
investigator. In this study, the proportion of neonates with comorbidities (respiratory distress,
polycythemia and sepsis) was significantly higher in the hypoglycemic group (Mahajan et al.,
2017).
Conversely, the study in term neonates by Brand et al. retrospectively identified a cohort
of 117 LGA healthy neonates with transient hypoglycemia on the first day of life (<2.2 mmol/L
1 hour after birth or <2.5 mmol/L afterwards) and obtained neurodevelopmental testing in 75 of
the children at 4 years old (64% of original cohort). These authors found no significant
difference between neonates with and without hypoglycemia on the Denver Developmental
Scale scores, Child Behavior Checklist scores or total IQ scores. Although one IQ subscale
(reasoning) was significantly lower in hypoglycemic neonates, there was no correlation between
lowest plasma glucose level and the reasoning-IQ. These investigators also noted that the Denver
Developmental Scale score may not have been sensitive enough to detect neurodevelopmental
26
sequelae of hypoglycemia because the scores were found to be normal in almost all children
tested (Brand et al., 2005).
To further investigate the relationship between the duration, frequency, and severity of
low glucose concentrations in the neonatal period and later neurodevelopment, the Children with
Hypoglycaemia and Their Later Development (CHYLD) study followed a large prospective
cohort of term and late-preterm neonates born at risk for hypoglycemia (IDM, preterm birth,
SGA or LGA). The study included 614 well babies born at ≥ 32 weeks GA at risk for
hypoglycemia. Glucose concentrations were tested by intermittent blood glucose measurements
and by masked continuous interstitial glucose monitoring. Hypoglycemia was defined as whole
blood glucose <2.6 mmol/L and interstitial episodes were defined as interstitial glucose
concentrations <2.6 mmol/L for ≥10 minutes. At 2 years of age, they assessed 404 children born
≥35 weeks GA with a comprehensive neuropsychological assessment. Hypoglycemia occurred in
53% of the neonates. The investigators found that hypoglycemia, when treated to maintain blood
glucose concentration ≥2.6 mmol/L, was not associated with increased risk of neurosensory
impairment and processing difficulty compared to at-risk children that did not have
hypoglycemia. Neonates with severe, multiple or prolonged hypoglycemia episodes also did not
have worse outcomes. Further, the risks were not increased among children with clinically
unrecognized hypoglycemia. Interestingly, the factor in the first 48 hours that was most
predictive of an adverse 2-year outcome was glucose instability, which was defined as the time
spent outside the central range of 3 to 4 mmol/L. Both higher glucose levels and less glucose
stability in the first 48 hours were associated with neurosensory impairment, especially cognitive
delay. Of note, the higher glucose concentrations were predominantly within the normal range as
only 3 neonates had clinical hyperglycemia (>8 mmol/L). A more rapid increase in glucose
concentration following dextrose treatment for hypoglycemia in the first 12 hours after birth was
also associated with neurosensory impairments at 2 years of age (McKinlay et al., 2015).
It has been proposed that increased glycemic variability could generate more reactive
oxygen species due to hyperglycemia-induced oxidative stress (Hirsch et al., 2005). Preclinical
studies using cell culture and rodent models have shown that neuronal death was triggered by
glucose reperfusion rather than hypoglycemia itself, and that hyperglycemia following
hypoglycemia can worsen neuronal injury. Superoxide production may be influenced by blood
glucose concentrations achieved during the period immediately after hypoglycemia correction;
27
with correction to higher glucose levels during the first hour after hypoglycemia being associated
with greater superoxide production and neuronal death (Ennis et al., 2015; Suh et al., 2007).
These findings suggest that rate of correction of hypoglycemia as well as prevention of rebound
hyperglycemia may be important and should be investigated further.
The CHYLD study recently published 4.5 year outcomes, when emerging neurocognitive
functions could be further assessed. In this follow-up study, 477 children born at ≥ 32 weeks GA
were assessed. Neonatal hypoglycemia occurred in 59% of children and although it was still not
associated with an increased risk of neurosensory impairments, hypoglycemia was associated
with an increased risk of low executive function and low visual motor function. Risk was highest
in children that had severe, recurrent, or clinically undetected hypoglycemia (interstitial episodes
only). This was the first study to show that clinically undetected low glucose concentrations can
be associated with an adverse outcome. Furthermore, children who developed neurosensory
impairment between 2 and 4.5 years had a steeper rise in interstitial glucose concentrations after
hypoglycemia, whereas children with stable neurosensory status had similar interstitial glucose
concentrations in the first 48 hours (McKinlay, Alsweiler, et al., 2017). Impaired executive
function and visual motor function may be a risk factor for later school difficulties.
In accordance with these findings, early transient newborn hypoglycemia was associated
with lower achievement test scores in literacy and mathematics at 10 years old in a retrospective
population-based cohort where the investigators controlled for multiple perinatal factors,
including maternal educational and socioeconomic status. Newborns underwent universal early
glucose screening and medical data was matched to their 4th grade student achievement test
scores in 1395 newborn-student pairs. Although there are several limitations of the retrospective
design, the associations were large and potentially clinically significant (Kaiser et al., 2015).
In summary, the presence of hypoglycemia in the neonatal period has been associated
with worse long-term neurodevelopmental outcome in several studies. Studies in well babies
suggest that long-term follow-up is essential as the potential long-term neurodevelopmental
consequences of transient hypoglycemia on higher-order cognitive functions may emerge at later
ages. These studies also highlight the potential negative consequences of higher glucose levels as
well as rapid increase in glucose levels with treatment.
28
1.4.3.2 Premature infants
Several studies have assessed the association of glucose derangements in preterm neonates and
long-term outcomes, however the results have been mixed. An influential study by Lucas et al.
examining neonatal hypoglycemia in a cohort of 661 premature neonates demonstrated an
association between repeated episodes of hypoglycemia and reduced mental and motor
developmental scores on the BSID at 18 months corrected age. Bayley scores were regressed on
days of hypoglycemia using varying blood glucose concentration cutoffs between 0.5 to 4
mmol/L and found a significant association with glucose levels below 2.5 mmol/L. They
therefore chose 2.6 mmol/L as a cutoff and showed that the presence of hypoglycemia on 5 or
more days was related to reduced BSID scores (Lucas et al., 1988). Follow-up at 7.5-8 years old
showed persistent impairment in arithmetic and motor function (Lucas et al., 1999).
The findings of an association between hypoglycemia with long term outcome was
supported by Duvanel et al. in a cohort of 85 SGA preterm neonates (27 to 34 weeks GA). They
showed that repeated episodes of hypoglycemia were associated with smaller head
circumferences and lower psychometric scores (perceptive and motricity scales) at 3.5 years,
although a significant difference could not be detected at 5 years (Duvanel et al., 1999). An
association has also been noted in moderately preterm neonates (born 32 to 36 weeks GA) in the
Longitudinal Preterm Outcome Project. Parents of children born moderately preterm completed
the Ages and Stages Questionnaire at 4 years old. The investigators looked at many risk factors
and morbidities from the neonatal period and in multivariate analysis, the only factor associated
with developmental delay at 4 years was hypoglycemia, defined as ≥1 glucose level <1.7
mmol/L in the first 72 h of life (Kerstjens et al., 2012).
However, a study by Tin et al. attempted to reproduce the findings of Lucas et al. in
premature neonates (<32 weeks GA) and they were not able to confirm the association between
recurrent hypoglycemia and neurodevelopmental outcome when measured at 2 and 15 years of
age. Of note, cognitive function at 15 years was relatively low in both the exposed and matched
control group (full scale IQ 80.7 vs 81.2) (Tin et al., 2012; Tin et al., 1998). Moreover, the study
by Lucas et al. adjusted for social class and mother’s educational level whereas the study by Tin
et al. did not. Socioeconomic status or maternal education level have been shown to be
associated with neurodevelopmental outcomes, including in the study by Lucas et al. (Ansell et
29
al., 2017; Dollaghan et al., 1999; Fernald et al., 2013; Lewis et al., 1989; Lucas et al., 1988; Vohr
et al., 2000).
Likewise, another retrospective observational cohort of 443 very preterm neonates (GA
<30 weeks or weight <1500g) examined several measures of neonatal glycemia, including
hypoglycemia, hyperglycemia, unstable glucose (≥1 blood glucose concentration ≤2.5 mmol/L
and ≥1 blood glucose concentration ≥8.6 mmol/L), and glucose variability (defined as the
standard deviation around the mean after log transformation of all blood glucose concentrations).
After correction for gestational age, birth weight z-score and socioeconomic status, measures of
neonatal glycemia were not independent predictors of neonatal illness or outcomes at 2 years old
(Tottman et al., 2017). Although Tottman et al. did not find an association with hyperglycemia
and long-term outcomes, hyperglycemia on the first day of life in extremely preterm neonates
(<27 weeks GA) has been associated with white matter reduction on term equivalent MRI scan
(Alexandrou et al., 2010). And another study of hyperglycemia in very preterm neonates (≤32
weeks GA) showed an association of hyperglycemia with neurological and behavior problems at
2 years of age (van der Lugt et al., 2010).
Hence, several studies have demonstrated that both early life hypo- and hyperglycemia in
premature neonates are associated with worse-long term outcomes, although not all studies have
supported these conclusions.
1.4.4 Outcomes following glucose derangements in neonatal
encephalopathy
As discussed, there is no universally accepted “safe” blood glucose level for all newborns since
individual susceptibility to brain injury varies, such as in the presence of comorbid conditions. In
HIE, the neonates’ ability to produce and use alternative fuels may not be sufficient to
compensate for low glucose levels, making these neonates particularly vulnerable to
hypoglycemia induced brain injury. Several studies in neonates with HIE have shown that
glucose derangements on intermittent glucose testing may be an important factor associated with
brain injury and worse neurodevelopmental outcomes.
30
The posterior predominant MRI pattern of hypoglycemic brain injury can be
distinguished from patterns of brain injury in HIE which include watershed, basal ganglia, total
or focal-multifocal white matter injury (Figure 1.5). In a prospective cohort of 160 term neonates
with HIE, selective posterior white matter and pulvinar edema were most predictive of clinical
hypoglycemia and could be superimposed on the predominant pattern of HIE (Wong et al.,
2013). Hypoglycemia may also be associated with a more severe MRI pattern of perinatal
hypoxia-ischemia. In a cohort of 94 term neonates with encephalopathy, occurrence of
hypoglycemia was associated with increased odds of corticospinal tract injury (Tam et al., 2012).
Figure 1.5 Neuroimaging features of basal ganglia pattern of brain injury in a 3-day old term neonate with HIE. (A)
DWI shows restricted diffusion involving the bilateral basal ganglia and thalami (B) corresponding ADC map
Higher rates of hypoglycemia have been associated with more severe clinical signs of
neonatal encephalopathy (Basu et al., 2009). A retrospective chart review of 185 depressed term
neonates with severe fetal acidemia showed that hypoglycemia was associated with adverse
short-term outcomes (death or evidence of moderate to severe encephalopathy with or without
A B
31
seizures) including in multivariate analysis (Salhab et al., 2004). Subsequently, studies looking at
long-term outcomes have shown neonatal hypoglycemia to also be associated with worse
neurodevelopmental outcomes. In the study discussed above by Tam et al. there were 1-year
outcomes available for 73 children of the cohort of 94 term neonates with encephalopathy. They
demonstrated that hypoglycemia in neonatal encephalopathy was associated with worse motor
outcomes as well as worse cognitive and language score on the BSID (Tam et al., 2012).
Conversely, in a retrospective cohort of 52 term neonates’ with HIE, although early
hypoglycemia (<6 hours of life) was associated with neurodevelopmental outcome at 24 months
old on univariate analysis, it was not a significant predictor after adjusting for severity of clinical
neonatal encephalopathy (Sarnat score at 24 hours of life). Nadeem and colleagues thus
concluded that early hypoglycemia was associated with severe HIE. Occurrence of
hyperglycemia also was not associated with adverse outcome (Nadeem et al., 2011). However,
these authors adjusted for clinical severity of encephalopathy which not only may be related to
severity of HIE, but could be worsened due to hypoglycemia or other comorbidities. Of note,
none of the neonates in the above studies received TH except for 11 of 94 neonates in the study
by Tam et al (2012).
In a post-hoc analysis of the CoolCap study, it was demonstrated that not only
hypoglycemia (≤2.2 mmol/L) but also hyperglycemia (>8.3 mmol/L) were very common in
neonates with HIE and that both were associated with unfavorable neurodevelopmental outcome
(death and/or severe neurodevelopmental disability) at 18 months corrected, independent of
severity of HIE and TH (Basu et al., 2016). The CoolCap Study included 234 neonates ≥36
weeks’ GA with moderate to severe HIE and neonates who were randomized to previous
standard care or head cooling for 72 hours within 6 hours after birth. Plasma glucose
concentrations were collected at specific time points after randomization (0, 4, 8 and 12 hours
after randomization). They further showed that neonates with hyperglycemia within 12 hours
after randomization appeared to benefit most from TH (Basu, Salemi, et al., 2017).
Some noteworthy differences have been reported between neonates with HIE who
experience hypo- or hyperglycemia. Neonates with HIE and hyperglycemia had higher rates of
reported sentinel events (e.g. prolapsed cord, umbilical cord tear, placental abruption and
ruptured uterus) and emergency caesarian delivery compared to normoglycemic and
hypoglycemic neonates (Basu, Salemi, et al., 2017). Consequently Basu et al. hypothesize that
32
hyperglycemic neonates were more likely to have a temporally acute and relatively intense insult
and that the evolution of brain injury in some hyperglycemic neonates may have been more
likely to be within the therapeutic window for TH to be beneficial. These neonates had milder
multiorgan injury and may have intact gluconeogenesis and stress hormone responses with
decreased glucose utilization due to brain injury from perinatal asphyxia (Basu, Salemi, et al.,
2017; Shi et al., 2012; Thorngren-Jerneck et al., 2001). In support of this idea, these authors
presented an abstract at the Pediatric Academic Society Meeting recently describing higher odds
of predominant basal ganglia or global patterns of injury in hyperglycemic neonates (Basu,
Ottolini, et al., 2017). Conversely, neonates with HIE and hypoglycemia have been reported to
more commonly have a watershed pattern of hypoxic-ischemic injury (Basu, Ottolini, et al.,
2017; Wong et al., 2013) which has been associated with prolonged partial asphyxia (Barkovich
et al., 1995). Further, neonates without a history of a sentinel event but rather with a history of
decreased fetal movements, have a watershed predominant injury and less clear response to
therapeutic hypothermia (Bonifacio et al., 2011).
Hyperglycemia has also been shown to adversely affect long-term outcome in neonates
with HIE in two small retrospective studies, in which 24 to 36% of babies underwent TH in each.
In a chart review of 41 neonates with HIE, the authors found that hyperglycemia in the first 12
hours of life was associated with poor gross motor outcomes (Gross Motor Function
Classification System [GMFCS] ≥ 1 or death [in NICU or within 6 months with severe motor
deficits]) in univariate analysis (Spies et al., 2014). The other retrospective study identified 56
neonates with HIE of which outcome data was available for 41 children and they likewise found
an association of hyperglycemia in neonates undergoing TH with poor outcome (death,
microcephaly, and/or moderate to severe CP documented between 18 to 36 months) (Chouthai et
al., 2015).
Lastly, looking at glycemic variability, there was one small retrospective cohort of 23
term neonates with HIE treated with TH which showed that glucose variability was associated
with severe neurodevelopmental disability (GMFCS 3 to 5, Bayley III Motor Standard Score
<70, Bayley III Language Score <70 and Bayley III Cognitive Standard Score <70). As a
measure of glucose variability, they used mean absolute glucose (MAG) change which was
calculated as the sum of differences between successive arterial glucose values divided by 24
hours (Al Shafouri et al., 2015). Although their cohort was small, these findings are consistent
33
with the CHYLD study which showed greater glycemic variability was associated with
neurosensory impairment at 2 years in well babies. Similarly, a relationship between glucose
variability (as measured by standard deviation [SD]) with mortality was demonstrated in VLBW
preterm neonates (Fendler et al., 2012). Also, it has been shown that hypoglycemia,
hyperglycemia and glucose variability (SD or glucose variability index) have been associated
with mortality, nosocomial infections and increased length of hospital stay in term neonates and
children admitted to the PICU (Faustino et al., 2005; Hirshberg et al., 2008; Rake et al., 2010;
Wintergerst et al., 2006). As well, a longer duration of hyperglycemia and higher peak glucose
levels have been associated with increased mortality in the PICU (Srinivasan et al., 2004). In
adults, glucose variability also has been associated with increased mortality in the ICU (Ali et al.,
2008; Bagshaw et al., 2009; Dossett et al., 2008; Egi et al., 2006; Hermanides et al., 2010;
Krinsley, 2008; Pidcoke et al., 2009; Waeschle et al., 2008). Measures of glycemic variability
used in these studies have included SD, coefficient of variability, and presence of both an
episode of hypo- and hyperglycemia within 24 hours of ICU admission.
The data regarding the effects of tight glucose control in critically ill adults on outcome is
conflicting. Van den Berghe et al. showed a reduction in mortality with tight glucose control
(van den Berghe et al., 2001) whereas the NICE-SUGAR study found an increase in mortality
with intensive glucose control (Finfer et al., 2009). The study by van den Berghe et al. showed a
markedly lower SD in the intensively treated group (SD of morning blood glucose, 1.1 mmol/L
in the intense treatment group vs. 1.8 mmol/L in the conventional treatment group) compared to
the NICE-SUGAR study where SD of morning glucose was equal in both groups (Siegelaar et
al., 2010). One hypothesis for these conflicting results is the differential effect of glucose
variability in the studies (Siegelaar et al., 2010).
In summary, in neonates with HIE, not only hypo- and hyperglycemia but also glycemic
variability may be important factors associated with brain injury and poor neurodevelopmental
outcomes. It is essential to identify whether hypoglycemia, hyperglycemia and glucose
variability merely reflect the severity of the underlying illness, or whether these factors
contribute further to brain injury and poorer long-term outcomes.
34
1.5 Continuous interstitial glucose monitoring 1.5.1 Continuous interstitial glucose monitors overview
Continuous interstitial glucose monitors (CGMs) were developed for use in management of
diabetes mellitus and have been found to improve glucose control (Battelino et al., 2012;
Cemeroglu et al., 2010) and were shown to be safe and potentially useful in the ICU setting as
well (Allen et al., 2008; Bridges et al., 2010; Piper et al., 2006). CGM devices and insulin pumps
can also now be combined to form a closed loop system, also known as artificial pancreas
systems. They utilize a complex set of algorithms that interpret CGM glucose levels and adjust
insulin delivery via an insulin pump. Closed-loop systems have been shown to be beneficial to
improve glycemic control without increases in hypoglycemia in children and adults (Anderson et
al., 2016; Chernavvsky et al., 2016; DeBoer et al., 2017; Leelarathna et al., 2014; Tauschmann et
al., 2016).
CGMs contain a disposable glucose oxidase-based platinum electrode sensor inserted
subcutaneously which catalyzes interstitial glucose oxidation generating an electrical current
every 10 seconds. Glucose concentrations are estimated from this signal using proprietary
algorithms which require regular calibration with blood glucose measurements. Although the
sensor measures the concentration of interstitial glucose every 10 seconds, the monitor records
averaged glucose values every 5 minutes. Some systems display the output in real-time, whereas
for others, the data cannot be viewed in real time and are therefore a useful research tool, as it
does not impact on clinical care. The CGM data are downloaded after completion. The software
does not display data when glucose concentrations fall below 2.2 mmol/L or above 22 mmol/L
(Beardsall et al., 2005; Beardsall et al., 2013; McKinlay, Chase, et al., 2017). Retrospective
analysis, filtering and recalibration of the raw signal may be possible (Signal et al., 2012). Of
note, the interstitial glucose sensor can also take as long as 2 hours to stabilize after insertion
(Harris et al., 2010). The sensor is well tolerated, including in small premature infants, with no
evident local edema, inflammation or infection at the sensor site or problems of skin debridement
with removal (Beardsall et al., 2005; Beardsall et al., 2013). However, there is limited evidence
for the benefits of treating newborns based on CGM monitoring.
Errors in CGM measures may occur. There can be random error of the sensor due to
sensor technology and its interstitial location (referred to as zero-mean error). Plus, conversion of
35
the raw signal to a glucose concentrations requires the regular input of calibration measures (i.e.
blood or plasma glucose concentrations) and thus the resulting output will reflect the reliability
of the calibration sample used. Ideally, CGMs should be calibrated to plasma equivalent whole-
blood glucose concentrations rather than point-of-care glucometers (McKinlay, Chase, et al.,
2017; Thomas et al., 2014). There is the potential for sensor drift where there is shifts in sensor
output between calibration points. The CGMs rely on a continuous shifting internal algorithm to
generate glucose concentrations from the raw sensor signal. The regular calibrations use the
‘true’ glucose concentrations which are entered into the device. The drift between calibrations
measurements may impact accuracy, which could result in apparently stable CGM values when
the blood glucose concentrations are falling (McKinlay, Chase, et al., 2017; Signal et al., 2012).
Also, studies have reported that interstitial glucose concentrations can have a variable “lag”
behind the blood glucose concentrations by as long as 20 minutes due to delays in diffusion of
glucose from the vascular into the interstitial space, as well as patient-specific and instrumental
factors. Variability in this time lag has been documented and there may be increasing positive
error as the blood glucose concentrations decrease and increasing negative error when the
glucose concentrations are rising (Boyne et al., 2003; Caplin et al., 2003; D. L. Harris et al.,
2009; Kovatchev et al., 2009; Rebrin et al., 1999; Schmelzeisen-Redeker et al., 2015; Shah et al.,
2018). A study in term and near-term neonates was not able to find a significant time lag but
found a late increase in accuracy, with the highest accuracy for subcutaneous glucose
concentrations observed after a 15 to 19-minute delay following blood glucose sampling
(Wackernagel et al., 2016).
The two main CGM brands that have been used in neonates are Medtronic Minimed
(Northridge, CA, United States) (Figure 1.6) and DexCom (San Diego, CA, United States), both
of which manufacture retrospective and real-time devices. Of note, none of these devices have
been approved for clinical use in neonates. The clinical significance of glucose disturbances
detected on CGM is not clear, and in the absence of well-established clinical guidelines, there is
a risk that CGM could lead to unnecessary or even harmful interventions (McKinlay, Chase, et
al., 2017).
36
Figure 1.6 The Medtronic iPro2 continuous glucose monitor. The sensor is inserted subcutaneously into the lateral
aspect of the neonate’s thigh. The sensor catalyzes interstitial glucose oxidation, which is converted to an average
glucose value every 5 minutes.
1.5.2 Continuous interstitial glucose monitor use in newborns CGMs have been used in studies in term and preterm neonates. They have been validated in
small preterm neonates and have been used over a 7-day period without any deterioration in
accuracy over time. It was shown that the CGM data correlates well with point of care devices,
with minimal bias. Analysis of CGMS System Gold (Medtronic, Minneapolis, Minnesota, USA),
in 188 VLBW preterm neonates showed that overall the CGM marginally under reads but there
was essentially no bias overall. At low glucose levels, there was a tendency to over read and at
higher glucose levels there was a slight bias to under read. Newer CGM systems with modified
sensors and monitors are now thought to have better accuracy, particularly at lower glucose
levels (Beardsall et al., 2013).
The CHYLD study used blinded CGMs and a clinical management protocol aimed at
maintaining blood glucose concentrations ≥2.6 mmol/L. CGM use detected low glucose
concentrations in almost one quarter of neonate which were not detected clinically based on
intermittent blood glucose monitoring (McKinlay et al., 2015). A recent RCT in VLBW preterm
neonates also detected many more episodes of hypoglycemia using CGM than routine
intermittent blood glucose testing (Uettwiller et al., 2015).
A recent randomized trial used CGM to guide glucose infusion rate in a cohort of 50
neonates ≤ 32 weeks’ GA or with birth weight ≤ 1500 g. Neonates were randomly assigned to
receive computer-guided glucose infusion rate adjustments driven by unblinded CGM or by
standard of care on the basis of blood glucose determinations. Neonates in the unblinded CGM
group had a significantly greater percentage of time spent in the euglycemic range and decreased
Sensor Recorder
37
time spent in the mild (2.6 – 3.9 mmol/L) and severe (<2.6mmol/L) hypoglycemia range
compared with the blinded CGM group. Use of CGM also decreased glycemic variability
(measured by SD and coefficient of variation [percent value of SD divided by mean glucose]).
With the ability to rapidly change glucose infusion rates, they were able to prevent both
hypoglycemia and hyperglycemia while maintaining glucose intake and weight gain goals.
Further, they showed that linking a CGM to a control algorithm guiding glucose titration can
successfully achieve glucose control in this population without need for insulin (Galderisi et al.,
2017).
CGMs have the potential to improve the detection of hypoglycemia and hyperglycemia
episodes that would otherwise go undetected with intermittent blood sampling, while also
providing information on duration and severity of the episodes. As well, CGMs have the
potential to be a valuable tool to improve management in these neonates, although further
research is needed.
1.5.3 Measures of glucose variability and complexity
Several methods of measuring glucose variability have been described in the literature, each with
advantages and disadvantages. SD is the most commonly used parameter. It represents an index
of dispersion of the glucose values, however it is important to keep in mind that glucose does not
have a Gaussian or ‘normal’ distribution in the statistical sense (Eslami et al., 2011).
Many of the studies discussed above looking at glycemic variability have used SD, but
various other measures were used as well. For example, the study by Wintergerst et al. in the
PICU population calculated a glucose variability index by dividing the absolute difference of
sequential glucose values by the difference in collection time (in hours + 0.01). The mean of the
ratios for each subject formed was the variability index (Wintergerst et al., 2006). Egi et al.
calculated the coefficient of variability (GluCV= GluSD* 100/GluAve) (Egi et al., 2006) and Ali et
al. calculated a glycemic lability index. (Ali et al., 2008). A glycemic lability index can be
calculated as Σ (Glun - Glun+1 (mmol/L))2*(hn +1 -hn )-1)*(number of readings)-1. A potential
benefit is that the differences are weighted individually, which gives more importance to the
greatest differences, which are likely to be more detrimental (Ali et al., 2008; Le Floch et al.,
38
2016).
CGM provides the opportunity to measure glucose variability more precisely. An analysis
of the blood glucose rate of change can be useful to evaluate the dynamics of glucose
fluctuations on the time scale of minutes. The glucose rate of change at a certain time (ti) is
computed as the ratio [Glucose(ti)—Glucose(ti-1)]/(ti-ti-1), where (ti) and (ti-1) are consecutive
CGM readings taken at times ti and ti-1, respectively. A larger (by absolute value) glucose rate of
change indicates more rapid and pronounced glucose fluctuations. One could also use the
average (over a certain time window, such as 1 h) absolute value of glucose rate of change as a
measure of local glucose variability (Kovatchev et al., 2005).
Many other terminologies and measures have been used to assess variability, some of the
common measures include the mean absolute glucose change per patient per hour (MAG), mean
amplitude of glucose excursions (MAGE; average amplitude of upstrokes or downstrokes with
magnitude greater than 1 SD), mean of daily differences (MODD; mean difference between
glucose values obtained at the same time of day on 2 consecutive days under standardized
conditions, and MODDd which allows calculation of the MODD for a continuous time series
involving many days), and the continuous overlapping net glycemic action (CONGAn; the SD of
the difference between values obtained exactly n hours apart) which was developed specifically
for CGM (Rodbard, 2009a; Siegelaar et al., 2010). Limitations of each these measures exist, for
instance, MAGE arbitrarily ignores excursions of less than 1 SD, yet the importance of these
smaller excursions is not clear (Siegelaar et al., 2010). A systematic review of the literature
identified 12 studies using 13 different indicators to measure glucose variability, with SD and the
presence of both hypo- and hyperglycemia being the most common. In each study, at least one of
the measures appeared to be associated with mortality. They conclude that due to methodological
limitations, heterogeneity of studies, and the possibility of reporting bias, it is still not clear
whether and which measure of glucose variability is independent of other confounders (Eslami et
al., 2011).
Many of these indices have been shown to be highly correlated with each other and SD
and thus convey largely the same information (Rodbard, 2009a; Saisho et al., 2015). As SD
correlates highly with other variability measures, it is often used, even though glucose is not
normally distributed (the mathematical condition for SD) and it doesn’t take into account
39
glycemic swings (Siegelaar et al., 2010). Conversely it has also been shown that the different
indices and ratios may provide some complementary information (Le Floch et al., 2016;
Rodbard, 2009b). One study in diabetic patients showed that the most reliable markers of glucose
variability were the pairs of MAG and MAG/m (MAG to mean ratio) as well as two new
markers; the glucose fluctuation index (GFI), and its ratio to the mean of glucose values, the
glucose coefficient of fluctuation (GCF) (Le Floch et al., 2016). Consequently, no universal
consensus exists on how to measure glycemic variability.
In addition to glucose variability, which looks at the magnitude of glucose fluctuations
over time, one could also consider glucose complexity. Glucose complexity is a dynamic
measure of glucose time series. Glycemic variability depends on both exogenous and
endogenous factors, whereas glucose complexity describes the endogenous glucose regulation
that is independent of exogenous factors (Brunner et al., 2012). It is hypothesized that the
regulatory system should normally make frequent corrections of glucose levels and result in a
‘complex’ profile. Whereas critically ill patients will have a ‘low complexity’ glucose profile due
to inability to correct glucose fluctuations frequently and quickly. In critically ill adult patients,
loss of glucose complexity has been shown to be associated with mortality (Brunner et al., 2012;
Lundelin et al., 2010).
Glucose complexity can be calculated using detrended fluctuation analysis (DFA). DFA
is a unitless measure which estimates the degree of long-range correlations within a signal. It
analyzes how the time series and its linear regression diverge as the ‘time window’ increases.
Lower DFA values represents higher complexities (Brunner et al., 2012; Lundelin et al., 2010).
Loss of complexity has been associated with greater glycemic variability (Garcia Maset et al.,
2016). Two modes of DFA can be used depending on the properties of the time series.
Monofractal DFA is used when the scaling properties of the time series can be quantified by a
single power law exponent (independent of time and space). As real-world physiological signals
often do not exhibit monofractal scaling behavior over the entire time period, multiple scaling
exponent are required to fully characterize the correlation properties of the signal, and
multifractal DFA should be employed. Assessment of sensor glucose data from CGM devices
suggests the scaling properties of the time series are multifractal and that multifractal DFA
appears to be a more appropriate analysis technique (Signal et al., 2013).
40
1.6 Neurophysiology
1.6.1 Continuous electroencephalography (cEEG)
EEG is a fundamental tool for assessing brain activity. Traditionally EEG is interpreted by visual
inspection, using a set of qualitative rules developed through clinical experience by experts in the
field. EEG measures electrical fields generated by neuronal activity by recording the amplified
potential differences between electrodes placed on the scalp. The synchronous electrical activity
of more than 108 neurons in a cortical area is required to record visible scalp EEG potentials. The
major contributors to the EEG signal are postsynaptic potentials produced by the cortical neurons
(both excitatory postsynaptic potentials and inhibitory postsynaptic potentials). The scalp EEG
signal is modified by electrical conductive properties of the tissues between the electrical source
and the recording electrode on the scalp, conductive properties of the electrodes, and the
orientation of the cortical generator (Current Practice of Clinical Electroencephalography, 2014;
Olejniczak, 2006).
Scalp electrodes are traditionally placed according to the International 10-20 system.
Specific measurements are made between bony landmarks, and then from these landmarks,
specific measurements are made and 10% and 20% of a specified distance is used at the inter-
electrode distance. This provides uniform inter-electrode distances in all laboratories and
symmetric placement of electrodes over the left and right hemispheres. The orderly arrangement
of multiple channels is termed a montage. The American Clinical Neurophysiology Society
(ACNS) has recommended using a minimum of 21 electrodes for the International 10-20 system.
Each electrode is named by a letter/number combination. The letter designates the approximate
anatomical location (e.g. “C” for central). Odd-numbers designate electrodes over the left
hemisphere and even-numbered electrodes over the right (Current Practice of Clinical
Electroencephalography, 2014). In neonates, EEG electrodes are also applied using the
international 10-20 system, but often modified with fewer electrodes because of the smaller
neonatal head circumference (Figure 1.7). Simultaneous video recording with EEG is strongly
recommended to provide clinical correlate for seizures and to assist in detecting artifacts.
Recordings should include relevant polygraphic data (electrocardiogram and respiration). Other
extracerebral channels including extraocular movements, oxygenation, and electromyogram are
useful, but not universally required (McCoy et al., 2013; Shellhaas et al., 2011).
41
There is no universally accepted EEG classification for neonatal HIE and published
studies have used different classification schemes (Andre et al., 2010; Birca et al., 2016; Glass et
al., 2014; Holmes et al., 1993; Leijser et al., 2007; Nash et al., 2011; Obeid et al., 2017;
Tsuchida, 2013; Walsh et al., 2011). In 2013 the ACNS published an expert consensus of
standardized neonatal EEG nomenclature with the goal of improving consistency and facilitating
collaborative research. The consensus focuses on normal state changes, background features,
graphoelements (or named neonatal EEG features), seizures, and rhythmic or periodic patterns.
Several normal behavior states of term infants are described. During wakefulness, the EEG
background contains continuous, low to medium amplitude mixed frequency activity
(predominance of theta and delta with overriding beta activity), often referred to as activité
moyenne. Sleep is classified as active, quiet, transitional and indeterminate. The EEG during
active sleep consists of activité moyenne, indistinguishable from normal wakefulness. Quiet
sleep near term consists of tracé alternant, which is an alternating trace of higher voltage bursts
(50–150 mcV peak to peak [pp]) lasting 4 to 10 seconds alternating with briefer, lower voltage
(25–50 mcV pp) interburst intervals (IBI) composed mostly of mixed theta and delta frequencies.
Tracé alternant is first seen at 34 to 36 weeks postmenstrual age (PMA). There is a progressive
decrease in normal IBI durations with increasing PMA. At 34-36 weeks PMA, the maximum
normal IBI is 10 seconds and at 37 to 40 weeks PMA the maximum normal IBI is 6 seconds.
Sleep-wake cycling (SWC) is the normal pattern of alterations among behavioral states. In a term
infant, a complete sleep and waking cycle has a typical duration of 3 to 4 hours. An isolated
sleep-only cycle typically lasts 40 to 70 minutes (Tsuchida et al., 2013).
Excessive discontinuity refers to abnormally discontinuous background activity with
bursts that contain some normal patterns and graphoelements separated by IBIs that are too
prolonged or voltage too depressed for PMA. Tracé discontinu is the normal discontinuous
pattern seen in healthy preterm babies. This EEG pattern is characterized by bursts of high
voltage (50–300 mcV pp) activity that is interrupted by low voltage interburst periods (<25 mV
pp). The duration of the low voltage interburst periods is dependent on PMA, and decreases with
increasing PMA. Burst suppression pattern is characterized by a more severe disruption of the
continuity of the EEG. Burst suppression consists of invariant, abnormally composed EEG bursts
(no normal features within the bursts) separated by prolonged and abnormally low voltage IBIs
periods (<5 mcV pp). The definition does allow for one electrode with sparse activity during the
42
IBI (up to 15 mcV pp) or <2 seconds with transient activity up to 15 mV pp, or >2:1 asymmetry
in voltage in multiple electrodes (Tsuchida et al., 2013). A low voltage suppressed or low voltage
undifferentiated background is also invariant and unreactive. It is composed of persistently low
voltage activity without normal background features. The baseline voltage is <10 mcV, yet can
be interspersed with higher voltage (≥10 mcV pp) transient activity for <2 seconds (Tsuchida et
al., 2013).
Variability (lability) signifies noticeable spontaneous EEG responses to internal stimuli
(e.g. those that occur during typical sleep–wake cycling). Of note, arousals from sleep can result
in transient attenuation of EEG voltages, which should not be considered discontinuity.
Reactivity of the EEG is demonstrated by a noticeable cerebral EEG response to external stimuli.
These EEG responses to stimuli can consist of changes in any electrical domain: frequency,
continuity or voltage (Tsuchida et al., 2013).
The ACNS classifies neonatal seizures as: clinical-only seizure, electroclinical seizure
(clinical signs with simultaneous EEG correlate), or electrographic only (EEG seizure with no
specific visible clinical signs). A clinical-only seizure consists of a sudden paroxysmal clinical
change which does not correlate with a simultaneous EEG seizure. Clinical signs may include
unnatural posturing, obligatory stereotyped movements, sudden behavioral arrest, or autonomic
dysfunction (such as episodic tachycardia or hypertension, flushing, pallor, or salivation). An
electrographic seizure is defined as a sudden, abnormal EEG event defined by a repetitive and
evolving pattern, with a minimum voltage of 2 mcV pp and minimum duration of at least 10
seconds. “Evolving” is defined as an unequivocal evolution in frequency, voltage, morphology,
or location. There is no minimum frequency required in the definition of neonatal seizures, in
contrast to the definition of seizures in older children and adults. If brief rhythmic repetitive
discharges last <10 seconds but with evolution, they would be considered brief rhythmic
discharges and not seizures. There must be at least 10 seconds separating 2 distinct seizure
events. Status epilepticus is present when the summed duration of seizures comprise ≥50% of an
arbitrarily defined 1-hour epoch (Tsuchida et al., 2013). The use of 10 seconds as a minimum
duration for seizures is largely arbitrary, and electrographic events with a clinical correlate are
usually considered electroclinical seizures even if they last less than 10 seconds (Massey et al.,
2018).
43
A study applying the ACNS standardized terminology and categorization found good or
very good interrater agreement for identification of seizures and classification of EEG
background. Interrater agreement was good in classifying EEG backgrounds on a 5-category
scale (normal, excessively discontinuous, burst suppression, status epilepticus, or electrocerebral
inactivity) (kappa = 0.70, p<0.001), with perfect agreement in 72% of records (43 of 60).
Interrater agreement was very good for identification of seizures (kappa = 0.93, p<0.001), with
perfect agreement in 95% of records (57 of 60). Although there was limited interrater agreement
for other background features and abnormal sharp waves (Wusthoff et al., 2017). Another study
had also shown that the inter-observer agreement for neonatal seizure detection on cEEG in a
cohort of term neonates at risk of acute encephalopathy was high with temporal assessment
resulting in a kappa of 0.827 (95% CI: 0.769 - 0.865; n = 70). The median agreement was 83.0%
(IQR 76.6 - 89.5%; n = 33) for seizure and 99.7% (IQR 98.9 - 99.8%; n = 70) for nonseizure
EEGs (Stevenson et al., 2015).
1.6.2 Amplitude-integrated electroencephalography (aEEG)
Continuous EEG is the gold standard for monitoring cerebral activity in neonates however there
are several barriers to its widespread use. Interpretation of the EEG requires specialized training,
as the EEG signals are low amplitude and easily contaminated by artifacts (e.g.
electrocardiographic, respiration, pulsation, movement, electrodes, equipment). A
neurophysiologist may not be available on an emergent basis throughout nights and weekends
for interpretation. Furthermore, it is expensive and access to equipment and technologists to start
a recording can be variable depending on time of day or day of the week.
aEEG is a simplified method for continuously monitoring cerebral activity that can be
applied by bedside caregivers with minimal training and can be interpreted by neonatologists or
other bedside caregivers (such as nurses or respiratory therapists) that do not have specialized
neurophysiology backgrounds (Shah et al., 2015).
Due to the ease of application and interpretation of aEEG compared to cEEG, it has been
used in NICUs in European centers for over 2 decades and increasingly used in North American
centers in the recent years (Glass et al., 2013). EEG processing includes an asymmetric band
44
pass filter (to attenuate activity below 2Hz and above 15 Hz), semi-logarithmic amplitude
compression, rectifying, smoothing, and time compression. The amplitude display is linear
between 0 to 10 mcV and logarithmic from 10 to 100 mcV. Detailed evaluation is lost, however
long-term trends and evolution of the background over time become more evident with the
compressed time scale (Hellström-Westas et al., 2006).
For aEEG, a minimum of 2 electrodes and a reference are placed for a single channel
montage or 4 electrodes and a reference for a dual channel montage (provides one channel from
each hemisphere) (Figure 1.7). Dual-channel aEEG recordings can provide information about the
presence of unilateral brain injury and can improve seizure identification in these neonates (van
Rooij, de Vries, et al., 2010). Originally, single-channel aEEG placed leads over the parietal
region (P3 and P4) because it is over the cerebrovascular watershed, an area at high risk for
acquired injury. However, the adjacent C3 and C4 channels probably provide similar data for
single-channel aEEG. Most contemporary machines allow the display of dual-channel recordings
(usually C3-P3 and C4-P4), along with the raw EEG from which the aEEG signals are derived.
When reduced channel aEEG is obtained simultaneously to complement ongoing cEEG
monitoring, P3 and P4 may be added to the conventional neonatal recording montage (Shellhaas
et al., 2011).
Figure 1.7 Standard placement for EEG electrodes A) according to the 10-20 system modified for neonates. B)
aEEG lead placement for dual channel aEEG recordings
A. B.
45
Generally, there is good correlation between findings on EEG and aEEG (Hellström-
Westas et al., 2006; Shellhaas et al., 2008). aEEG can demonstrate background abnormalities and
sleep wake cycling. Hypothermia does not significantly affect aEEG background activity
(Burnsed et al., 2011; Horan et al., 2007). There are 2 main background classification systems
used. Al Naqeeb et al. created a background classification for term infants which includes 3
categories. The classification is based on aEEG amplitudes. aEEG background is classified as
normal amplitude when the upper margin of aEEG activity is >10 mcV and the lower margin >5
mcV; as moderately abnormal amplitude when the upper margin of aEEG activity is >10 mcV
and the lower margin <5 mcV; and as suppressed when the upper margin of aEEG activity is <10
mcV and lower margin <5 mcV (al Naqeeb et al., 1999).
An alternate classification proposed by Hellstrom-Westas et al. classifies aEEG
background patterns using pattern recognition based on EEG terminology, including
categorization of sleep-wake cycling, which could be used in all newborns. The 5 background
categories were defined as: continuous with minimum (lower) amplitude around (5 to) 7 to 10
mcV and maximum (upper) amplitude of 10 to 25 (to 50) mcV; discontinuous with minimum
amplitude variable, but below 5 mcV, and maximum amplitude above 10 mcV; burst-
suppression as a discontinuous background with minimum amplitude without variability at 0 to 1
(2) mcV and bursts with amplitude >25 mcV; low voltage as a continuous background pattern of
very low voltage (around or below 5 mcV); and inactive, flat describes a primarily inactive
(isoelectric tracing) background below 5 mcV (Figure 1.8). A classification for SWC in the
aEEG was also proposed. SWC is characterized by smooth sinusoidal variations of the
bandwidth, mostly in the minimum amplitude. A wider bandwidth represents the discontinuous
background activity during quiet sleep (tracé alternant EEG in term infants), and the narrower
bandwidth corresponds to the more continuous activity seen during wakefulness and active sleep.
SWC was classified in 3 categories: no SWC (where there is no cyclic variation of the aEEG
background); imminent/immature SWC (when there is some, but not fully developed cyclic
variation of the lower amplitude, and not developed as compared with normative gestational age
representative data); and developed SWC (when there are clearly identifiable sinusoidal
variations between discontinuous and more continuous background activity, with cycle duration
>20 min) (Hellström-Westas et al., 2006).
46
Figure 1.8 Examples of aEEG background patterns.
Reproduced from: Stolwijk LJ, Weeke LC, de Vries LS, van Herwaarden MYA, van der Zee DC, et al. (2017)
Effect of general anesthesia on neonatal aEEG—A cohort study of patients with non-cardiac congenital anomalies.
PLOS ONE 12(8): e0183581.
Classifications similar to the pattern classification system had been previously used and
described in cohorts of neonates with HIE and were shown to be correlated with prognosis.
Neonates with continuous normal voltage (CNV) or discontinuous normal voltage (DNV)
background patterns during the first 6 hours after birth were likely to survive without sequelae,
whereas neonates with burst suppression (BS), continuous low voltage (CLV) or flat tracings
(FT) had a high risk for death or neurodevelopmental disability (Hellstrom-Westas et al., 1995;
ter Horst et al., 2004; Toet et al., 1999). aEEG monitoring can be a useful tool in neonates with
HIE to monitor degree of encephalopathy and to detect seizures, as well as to assess recovery
and predict outcome. Its role in outcome prediction will be discussed further below.
Seizures can be identified on aEEG as an abrupt, transient rise in the lower and upper
margins and narrowing of the bandwidth. aEEG can detect approximately one third of neonatal
seizures. Seizures that are infrequent, short (<30 seconds), and low amplitude are difficult to
detect (Lawrence et al., 2009; Mastrangelo et al., 2013; Rakshasbhuvankar et al., 2015;
Shellhaas, Soaita, et al., 2007). Seizure detection may be dependent on user experience (Frenkel
et al., 2011) and on availability of the raw aEEG for confirmation, which improved the
47
sensitivity to 76% and specificity to 78% in one study (Shah et al., 2008). Still, aEEG is superior
to clinical detection, as clinicians are poor at clinical recognition of seizure and nonseizure
movements (Malone et al., 2009) and up to 85% of seizures in neonates are subclinical
(electrographic only), meaning they would go undetected without EEG (Bye et al., 1995; Glass
et al., 2016; Naim et al., 2015; Nash et al., 2011; Wietstock et al., 2016; Wusthoff et al., 2011). The majority of seizures occur in the central region (56%), occasionally in the temporal (25%)
and occipital regions (14%), and rarely in the frontal regions (5%) (Shellhaas & Clancy, 2007).
Thus, placement of electrodes over the central region is important for seizure detection.
1.7 Neurophysiology and neurodevelopmental outcomes
1.7.1 Background and neurodevelopmental outcomes
aEEG background abnormalities and recovery have been shown to be predictive of long-term
neurodevelopmental outcome in term neonates with HIE (Awal et al., 2016; Azzopardi & Toby
study group, 2014; Chandrasekaran et al., 2017; Dunne et al., 2017; Skranes et al., 2017;
Spitzmiller et al., 2007; Thoresen et al., 2010; van Laerhoven et al., 2013). An early normal or
mildly abnormal EEG is predictive of a favorable neurodevelopmental outcome (Awal et al.,
2016; Murray et al., 2009; Pressler et al., 2001; Toet et al., 1999; Weeke et al., 2016). Whereas,
EEG background abnormalities, in particular burst suppression, low voltage or a flat trace is
associated with poor long-term neurodevelopmental outcomes (at 12 months of age or older)
(Awal et al., 2016). However, even neonates with mild EEG abnormalities have reduced intact
survival rates. A study that performed cognitive and motor assessments at 5 years of age showed
that 25% of neonates with mild EEG abnormalities 6 hours after birth had lower cognitive scores
that the healthy comparison group recruited at birth (Murray et al., 2016). Also, serial recordings
improve the prognostic accuracy of EEGs, as progressive improvements in the background are
associated with more favorable outcomes (Murray et al., 2009; Nash et al., 2011; Pressler et al.,
2001), and the worsening of a background pattern or the development of an abnormal
background is predictive of unfavorable outcomes (Khan et al., 2008). Thus, the evolution of
aEEG background over time, using a classification system that defines specific patterns of
48
change in aEEG over the course of therapeutic hypothermia may also be useful for outcome
prediction, rather than grading only at a discrete time interval (Sewell et al., 2018).
A meta-analysis of 8 studies prior to the introduction of therapeutic hypothermia showed
a pooled sensitivity of 91% and specificity of 88% of early severe aEEG backgrounds to
accurately predict poor neurodevelopmental outcome (Azzopardi & Toby study group, 2014;
Spitzmiller et al., 2007). Suppressed aEEG background patterns within the first 24 hours of life
were predictive of adverse outcome (al Naqeeb et al., 1999; Toet et al., 1999), however with
therapeutic hypothermia, the predictive value of the aEEG is delayed until 24 to 48 hours of life.
Still, a persistently abnormal aEEG background at 48 hours remains an important prognostic
indicator (Chandrasekaran et al., 2017; Cseko et al., 2013; Massaro et al., 2012; Shellhaas et al.,
2015; Thoresen et al., 2010). In a cohort of term neonates with HIE that received therapeutic
hypothermia, the positive predictive value (PPV) of EEG background activity for abnormal
outcome was 100% at 36 hours and 48 hours and the negative predictive value (NPV) was 75%
at 36 hours and 69% at 48 hours (Weeke et al., 2016). Cumulative doses of morphine,
phenobarbital or midazolam do not significantly delay aEEG recovery (Cseko et al., 2013).
Several EEG features have also been associated with unfavorable outcome including,
background amplitude of <30 mcV, interburst interval of >60 seconds and electrographic
seizures (Kontio et al., 2013; Menache et al., 2002; Murray et al., 2009; Obeid et al., 2017;
Pressler et al., 2001; Shah et al., 2014b). Quantitative automated analysis of EEG discontinuity
may improve the objectivity of EEG assessment in newborns. Dunne et al. (2017) calculated
EEG discontinuity using a novel algorithm and showed that higher mean discontinuity at 24
hours and 48 hours were associated with severe cerebral tissue injury on MRI and unfavorable
neurodevelopmental outcome at 24 months. Quantitative analysis of the total suppression length
has also been associated with death or neurodevelopmental disability (Flisberg et al., 2011).
More detailed EEG characteristics of discontinuous tracings relate to outcome and clinical
features (Biagioni et al., 1999). Quantitative measures looking at novel features of bursts such as
measures of burst area and duration and their interrelationships, may improve prognostic
capabilities beyond the dichotomous classification of presence-versus-absence of a burst
suppression pattern or length of IBIs (Iyer et al., 2014). An increase in EEG discontinuity may be
seen with rewarming after therapeutic hypothermia, principally in neonates with severe HIE
(Birca et al., 2016), although another study found that severity of HIE but not core body
49
temperature affected the EEG (Burnsed et al., 2011). Abnormal interhemispheric synchrony
(defined as when the onset of bursts between hemispheres is separated by more than 1.5 seconds
in discontinuous background) was also shown in a study to have good sensitivity and specificity
for predicting survival without disability, especially when combined with EEG background
analysis (Leroy-Terquem et al., 2017).
A support vector machine classifier using a multimodal combination of routine clinical
markers, EEG and heart rate parameters was evaluated for neurodevelopmental outcome
prediction at 24 months in newborn infants with HIE. They identified 12 multimodal features
that provided promising prediction results, 9 of which were EEG features and 3 were clinical
features. EEG achieved the highest performance as a predictor of clinical outcome (Temko et al.,
2015). Accordingly, although therapeutic hypothermia has changed the time-frame in which
tools such as aEEG and EEG are the most predictive of outcome, they are still the best objective
tools to identify the highest-risk patients for poor outcome (Bonifacio et al., 2015).
In summary, EEG is a good marker of long-term neurodevelopmental outcome and is
thus a useful surrogate marker in the acute setting in neonates with encephalopathy.
1.7.2 Sleep-wake-cycling and neurodevelopmental outcomes
Neonatal sleep involves endogenous driven brain activity and is fundamental for neuronal
survival and guidance of brain networks (Dereymaeker et al., 2017; Scher et al., 2009). Long-
term, neonatal sleep cycling has been shown to be a marker of brain function and enhance
prediction of neurodevelopmental outcomes. The failure to develop SWC has been shown to
strongly predict poor neurodevelopmental outcomes in neonates with HIE (Massaro et al., 2012;
Murray et al., 2009; Osredkar et al., 2005; Takenouchi et al., 2011; Thoresen et al., 2010). A
study in term neonates with HIE that received therapeutic hypothermia showed that failure to
develop SWC by 45.5 hours of life was associated with adverse outcomes (Sewell et al., 2018).
Another study in term HIE neonates treated with hypothermia further showed that the presence
of SWC was a better predictor of a good outcome than the recovery of aEEG background
pattern, while the absence of SWC was less likely to predict poor outcome then prolonged
abnormal background pattern (Cseko et al., 2013). The amount of time spent in different awake
50
and sleep states can also be fragmented due to frequent handling and painful procedures in the
NICU. Handling-related sleep deprivation may lead to an increase in quiet sleep (Axelin et al.,
2013). Whereas in healthy term infants, SWC is present immediately after birth and is dominated
by active sleep (Korotchikova et al., 2016).
Furthermore, a recent multimodality neuromonitoring study in a cohort of neonates at risk
for cerebral dysfunction showed that inefficient neonatal sleep patterns were independent
predictors of 18-month neurodevelopmental outcome evaluated with BSID-III. Eleven of 29
children in the cohort had HIE, of which 8 received therapeutic hypothermia. They used
objective measures of sleep–wake physiology including conventional video EEG, cerebral near-
infrared spectroscopy (NIRS), and a 12-hour full bedside polysomnography. Increased time
spent in quiet sleep predicted worse cognitive and motor scores and lower 0.5-2Hz EEG power
during quiet sleep predicted worse language and motor scores at 18-months (Shellhaas et al.,
2017).
Thus, assessment of SWC on EEG enhances prediction of long-term neurodevelopmental
outcomes with aEEG and cEEG.
51
1.7.3 Seizures and neurodevelopmental outcomes This section is modified from work published in Current Opinion in Neurology:
Pinchefsky EF and Hahn CD. Outcomes following electrographic seizures and electrographic
status epilepticus in the pediatric and neonatal ICUs.
Publication Date: 2017/04/01. DOI: 10.1097/WCO.0000000000000425
The work is published here with copyright permission from Wolters Kluwer Health, Inc.
A link to the published paper can be found at:
https://journals.lww.com/co-
neurology/Abstract/2017/04000/Outcomes_following_electrographic_seizures_and.7.aspx
Abstract
Purpose of review:
Increasing recognition of electrographic seizures and electrographic status epilepticus in
critically ill neonates and children has highlighted the importance of identifying their potential
contributions to neurological outcomes in order to guide optimal management.
Recent findings:
Recent studies in children and neonates have found an independent association between
increasing seizure burden and worse short-term and long-term outcomes, even after adjusting for
other important contributors to outcome such as seizure etiology and illness severity. The risk of
worse neurological outcome has been shown to increase above a seizure burden threshold of 12-
13 minutes per hour, which is considerably lower than the conventional definition of status
epilepticus of 30 minutes per hour. Randomized controlled trials in neonates have demonstrated
that EEG-targeted therapy can successfully reduce seizure burden, but due to their small size
these trials have not been able to demonstrate that more aggressive EEG-targeted treatment of
both subclinical and clinical seizures results in improved outcome.
Summary:
Despite mounting evidence for an independent association between increasing seizure burden
and worse outcome, further study is needed to determine whether early seizure identification and
aggressive antiseizure treatment can improve neurodevelopmental outcomes.
52
Key points:
• Evidence has accumulated demonstrating an association between electrographic seizures,
immediate pathophysiological changes and worse short-term and long-term
neurodevelopmental outcomes in neonates and children, even after adjusting for potential
confounders. However, a causal link between seizure burden and outcome remains to be
proven.
• The observed association between seizure burden and outcome is potentially confounded by
the contribution of the underlying etiology of the seizures to acute brain injury.
• Whether more timely electrographic seizure detection and more aggressive electrographic
seizure management improves outcomes remains unknown.
• Further research is needed to delineate disease-specific seizure burden thresholds above
which seizures have clinically relevant adverse effects that warrant aggressive antiseizure
treatment.
INTRODUCTION
With increasing use of continuous electroencephalogram (cEEG) monitoring (Sanchez,
Carpenter, et al., 2013), electrographic seizures and electrographic status epilepticus (ESE) are
being increasingly recognized in critically ill neonates and children. Estimates of seizure
prevalence in the pediatric ICU (PICU) populations have varied widely, with reports of
electrographic seizures in 10-47% and ESE in 9-32% of critically ill children undergoing cEEG
(Abend et al., 2013; Abend et al., 2011; Abend et al., 2009; Arndt et al., 2013; Gwer et al., 2012;
Kirkham et al., 2012; Payne et al., 2014; Piantino et al., 2013; Sanchez, Arndt, et al., 2013;
Schreiber et al., 2012; Shahwan et al., 2010; Topjian et al., 2013; Vaewpanich et al., 2016). This
variability is likely due to variations in the PICU populations and variable indications for cEEG
monitoring (Abend, 2015). In neonates at risk for seizures undergoing cEEG, the reported
prevalence of electrographic seizures is 26-82%, and ESE is 16-25% (Glass et al., 2016; Laroia
et al., 1998; McBride et al., 2000; Wietstock et al., 2016). Among specific neonatal
subpopulations, ES have been observed in 30-65% of neonates with hypoxic ischemic
encephalopathy (HIE) (Bashir et al., 2016; Glass et al., 2011; Glass et al., 2014; Kharoshankaya
et al., 2016; Nash et al., 2011; Srinivasakumar et al., 2015; Wusthoff et al., 2011), 82% of
neonates with stroke (Fox et al., 2016), and 8-13% of neonates and infants with congenital heart
disease (CHD) in the perioperative period (Gaynor et al., 2013; Naim et al., 2015). Importantly,
53
up to 85% of seizures in neonates are subclinical (also known as electrographic only), meaning
they would go undetected without EEG (Bye et al., 1995; Glass et al., 2016; Naim et al., 2015;
Nash et al., 2011; Wietstock et al., 2016; Wusthoff et al., 2011). ‘Uncoupling’, a phenomenon by
which ES persist after clinical seizures have ceased, is common in neonates, especially after
administration of antiseizure medications (Glykys et al., 2009; Scher et al., 2003). Furthermore,
it has been shown that clinical identification of seizures in neonates is unreliable, thus cEEG
monitoring is vital for accurate detection (Malone et al., 2009).
There is growing evidence that electrographic seizures and ESE are associated with
worse neurological outcomes. Greater awareness of the potential deleterious effects of seizures
and the high seizure prevalence among at-risk patients has motivated recommendations for more
widespread cEEG monitoring in the pediatric and neonatal ICU. However, the relationship
between seizures, brain injury and outcomes is likely complex, and remains incompletely
understood. This review will highlight recent evidence that has furthered our understanding of
the potential contribution of electrographic seizures and ESE to brain injury and their
relationship to short and long-term outcomes in neonates and children.
POTENTIAL MECHANISMS FOR SEIZURE-INDUCED BRAIN INJURY
Among critically ill children and neonates, electrographic seizures and ESE are recognized as an
important biomarker of more severe brain injury. However, mounting evidence is revealing that
seizures may independently contribute to brain injury through several plausible mechanisms. In
critically-ill adults, seizures have been linked to glutamate-mediated worsening of cerebral
edema (Vespa et al., 1998), long-term ipsilateral hippocampal atrophy (Vespa et al., 2010),
transient increases in intracranial pressure (Vespa et al., 2007) and cortical spreading depression
(Fabricius et al., 2008; Hartings et al., 2014). Pathological studies in both adults and children
support the hypothesis of excitotoxic neuronal injury (Tsuchida et al., 2007). Recently,
electrographic seizures in adults with severe traumatic brain injury were associated with a state
of “metabolic crisis”, demonstrated by significant elevations in the lactate/pyruvate ratio
measured by cerebral microdialysis (Vespa et al., 2016).
In neonates, there is also evidence that seizures are associated with impairments of
energy metabolism and neuronal integrity. Clinical seizure severity has been associated with
increased lactate/choline ratio and diminished N-acetylaspartate/choline ratio on magnetic
54
resonance spectroscopy (Miller et al., 2002). Neonates who experienced seizures during
phosphorous-31 magnetic resonance spectroscopy demonstrated a 33% decrease in
phosphocreatine during seizures, followed by a rapid normalization following administration of
antiseizure medications (Younkin et al., 1986).
SEIZURES & OUTCOME – ASSOCIATION OR CAUSATION?
Several clinical studies have reported that electrographic seizures or ESE are associated with
worse outcomes, even after adjusting for potential confounders such as diagnosis and illness
severity (Abend et al., 2015; Arndt et al., 2013; Fox et al., 2017; Kirkham et al., 2012; O'Neill et
al., 2015; Payne et al., 2014; Topjian et al., 2013; Wagenman et al., 2014). Although these
observational studies support the hypothesis that seizures contribute to brain injury and worsen
outcome, they cannot prove a causal link between higher seizure burden and worse outcome.
Recently, this hypothesis has been tested by two noteworthy trials in neonates designed to
evaluate whether more aggressive antiseizure treatment can decrease seizure burden, reduce
secondary brain injury and improve neurodevelopmental outcomes. These trials employed
amplitude-integrated EEG (aEEG) or cEEG monitoring to quantify seizure burden, and then
randomized neonates at risk for seizures to either treatment of both clinical and subclinical
(electrographic-only) seizures or treatment of clinical seizures alone.
The first Dutch-Flemish multicenter trial enrolled a total of 33 term neonates who
developed seizures following moderate-to-severe HIE, and employed 1-channel aEEG
monitoring to evaluate seizure burden. The primary outcome measure was an MRI injury score
at 4-10 days after birth. Neonates treated for both clinical and subclinical seizures had a lower
total seizure duration, however this was not significant, perhaps because only 8 of the 19
neonates in this group were appropriately promptly treated for seizures. Longer seizure duration
was associated with worse MRI injury scores when analyzing the entire cohort, however there
was no significant difference in MRI injury scores between the two treatment groups. The lack of
difference may have been due to the small sample size, which resulted from difficulties with
study recruitment (van Rooij, Toet, et al., 2010).
A second American single-center trial enrolled a total of 35 term neonates who developed
seizures following moderate-to-severe HIE, and employed conventional cEEG monitoring to
evaluate seizure burden. Outcome measures included MRI injury scoring and
55
neurodevelopmental outcome assessed at age 18-24 months using the Bayley Scales of Infant
Development (BSID). Neonates treated for both clinical and subclinical seizures had a
significantly lower seizure burden. Seizure treatment was initiated earlier in this group, but there
was no difference in the number of antiseizure medications used or phenobarbital levels between
the two groups. Higher seizure burden was correlated with worse brain injury and worse long-
term neurodevelopmental outcome, but these findings were significant only in the combined
cohort, and were not adjusted for HIE severity. When comparing the two treatment groups, there
was no significant difference in MRI injury scores or neurodevelopmental outcome. Again, due
to difficulties with recruitment, the study was likely underpowered to detect a difference between
the treatment groups (Srinivasakumar et al., 2015).
In summary, there is now substantial evidence for an independent association between
seizure burden and worse outcome, but these findings do not prove causation. Although
randomized controlled trials in neonates have demonstrated that EEG-targeted therapy can
successfully reduce seizure burden, due to their small size, these trials have not been able to
demonstrate that more aggressive EEG-targeted treatment of both subclinical and clinical
seizures results in improved outcome. Therefore, a causal link between electrographic seizures
and outcome remains to be proven.
CRITICALLY ILL CHILDREN
Several observational studies in critically-ill children have examined short-term outcomes after
electrographic seizures and ESE with a few recent studies examining longer-term outcomes. See
Table 1.1 for a summary of the most important and recent studies.
Short-term outcomes
Several studies have found an association between ESE and an increased risk of mortality, even
after adjusting for age, EEG background and neurological diagnosis (Abend et al., 2013;
Lambrechtsen et al., 2008; Topjian et al., 2013). However, no such association has been found
between increased electrographic seizures burden and mortality (Kirkham et al., 2012; Payne et
al., 2014; Piantino et al., 2013; Sanchez Fernandez et al., 2014). ESE has also been associated
with increased length of PICU stay (Abend et al., 2013).
In pediatric traumatic brain injury, the occurrence of seizures and ESE was associated
with increased length of hospital stay (Arndt et al., 2013; O'Neill et al., 2015) and ICU stay
56
(O'Neill et al., 2015), lower King’s Outcome Scale for Childhood Head Injury (KOSCHI) scores
at hospital discharge (Arndt et al., 2013), worse Glasgow Outcome Scale scores (GOS-E Peds),
and worse neurocognitive and functional evaluations (Vaewpanich et al., 2016).
Higher seizure burden has also been associated with worse short-term outcomes in
several studies. A prospective observational study of a combined cohort of 204 comatose
neonates and children using 1-3 channel EEG showed that the presence of electrographic
seizures was associated with an unfavorable neurological outcome (severe handicap or persistent
vegetative state) at 1 month in a multivariate analysis adjusting for etiology, age, pediatric index
of mortality, Adelaide coma score, and EEG background. Moreover, no children had favorable
outcomes if they had greater than 139 seizures, their total duration of electrographic seizures was
greater than 759 minutes, or any individual seizure lasted longer than 360 minutes (Kirkham et
al., 2012). A prospective observational study of 200 children undergoing cEEG for acute
encephalopathy found that ESE, but not electrographic seizures, was associated with higher
mortality and worse short-term outcome, measured as Paediatric Cerebral Performance Category
(PCPC) score at ICU discharge, even after adjusting for age, acute neurologic disorder, prior
neurodevelopmental status and EEG background category (Topjian et al., 2013). Conversely, a
study of 84 children with non-traumatic coma showed that both seizures and status epilepticus
were associated with a similarly increased risk of poor outcome (death or gross motor deficits at
discharge) (Gwer et al., 2012).
To further investigate the relationship between seizure burden and outcome, a prospective
observational single-center study of 259 critically ill neonates and children quantified the
maximum hourly seizure burden on cEEG and correlated this with the PCPC score at hospital
discharge, adjusting for diagnosis and illness severity. On multivariable analysis, for every 1%
increase in maximum hourly seizure burden the odds of neurological decline increased by 1.13.
Furthermore, above a maximum hourly seizure burden threshold of 12 minutes, there was a
marked increase in both the probability and magnitude of neurological decline. These
observations of a seizure burden “dose effect” strengthen the case for an association between
seizures and outcome, and identify an hourly seizure burden threshold of 12 minutes as a
potential therapeutic target (Payne et al., 2014). This study also observed that the magnitude of
the relationship between seizure burden and outcome appears to depend on the underlying cause
of seizures. For example, seizures in the context of anoxic brain injury appear to have less
57
influence on outcome, whereas seizures in the context of stroke or meningitis may have a greater
influence on outcome (Figure 1.9). Further work is required to more precisely define these
varying relationships between seizure burden and outcome, in order to inform patient-specific
decisions on how aggressively to treat electrographic seizures (Hahn et al., 2013).
Figure 1.9 Schematic illustration of potential relationship between seizure burden and outcome. The potential
deleterious effects of seizures in the context of acute brain injury are likely to depend on the underlying cause. The
probability of poor outcome might increase linearly or exponentially with increasing seizure burden, or a threshold
might exist, above which seizures are harmful. Previously published by Hahn and Jette (Hahn et al., 2013).
Long-term outcomes
One study has assessed the impact of seizures on long-term outcomes in critically ill children.
Long-term follow-up was obtained on 60/137 children who were previously
neurodevelopmentally normal, from an original cohort of 300 critically ill children who
underwent clinically-indicated cEEG monitoring (Abend et al., 2015; Wagenman et al., 2014).
Among these 60 children, those with ESE but not electrographic seizures demonstrated worse
long-term functional outcome scores (GOS-E Peds), a lower health-related quality of life, and
were more likely to develop epilepsy, even after controlling for age, acute neurological diagnosis
and EEG background category. A favorable long-term functional outcome was seen in 64% of
critically ill children without seizures, but in only 13% of those with ESE (Wagenman et al.,
2014). Furthermore, both electrographic seizures and ESE were associated with worse adaptive
functioning scores on multivariable analysis (Abend et al., 2015).
NATURE REVIEWS | NEUROLOGY VOLUME 9 | DECEMBER 2013 | 663
NEWS & VIEWS
A recent report by Arndt et al.5 high-lights the importance of subclinical seiz-ures among children with traumatic brain injury (TBI) requiring ICU admission. Among the 87 children studied, a remark-able 42.5% experienced seizures, with subclinical seiz ures (seizures without a clear clinical correlate that would have not been detected without cEEG monitoring) occurring in 16.7% and subclinical status epilepticus (subclinical seizures lasting >15 min, or more than three subclinical seizures per hour) in 13.8% of children. Subclinical seizures and status epilepti-cus were more common among younger children, those with abusive head trauma, and those with intra-axial haemorrhage. Children with subclinical seizures and status epilepticus required increased length of hospital stay and demonstrated worse outcomes, with lower scores on the King’s Outcome Scale for Childhood Head Injury (KOSCHI; scores range from 1–5, with 1 referring to death and 5 indicating good recovery) at hospital discharge.
The study by Arndt et al.5 is important and novel because it is the first to apply cEEG monitoring to a consecutive cohort of children with TBI who were admitted to a paediatric ICU, but a few limit ations must be borne in mind. First, the authors chose to combine two quite hetero geneous cohorts from different institutions, some-what limiting the study’s generalizabil-ity. Second, the authors relied on clinical EEG reporting, which might have been influenced by knowledge of the patient’s clinical status. Although inter-rater reli-ability of EEG reporting was assessed, it was performed using a separate sample of cEEG recordings from another set of pediat ric patients rather from the study parti cipants. It is also unclear what pro-portion of children experienced clinical seizures prior to cEEG monitoring, how soon after injury cEEG monitoring of patients began, and how long it took to record their first seizure. Despite these limitations, this report makes a compelling case that children with moderate or severe TBI—in particular, younger children, and those with abusive head trauma or intra-axial haemorrhage—should undergo cEEG monitoring.
Subclinical seizures present a pressing clinical dilemma: they clearly represent an important biomarker of acute brain injury, but their potential contribution to brain injury remains controversial. Plausible mechanisms for subclinical-seizure-induced
brain injury identified in experimen-tal animal models include excito toxicity, activation of inflammatory cascades, and hypoxia–ischaemia. Evidence for a similar deleterious effect of subclinical seizures in critically ill children and adults is mount-ing. Among adults with acute brain injury, subclinical seizures have been associated with focal metabolic dysfunction, increased intra cranial pressure, and ipsilateral hippo-campal atrophy.6,7 Among both adults and children, electrographic seizures and status epilepticus have been associated with increased mortality and worse short-term outcomes, even after controlling for diag-nosis and injury severity.8,9 Conclusive proof of a causal link between seizures and worse outcomes has, however, remained elusive. To demonstrate that reduction of seizure burden actually improves outcomes, an inter ventional study would probably be required. Fortunately, one such multi centre study is currently underway in critically ill adults.10
The potential for subclinical seizures to cause harm is likely to depend on both the seizure burden and the type and severity of the underlying brain injury (Figure 1). Among patients with anoxic brain injury, the probability of a poor outcome might be so high that seizures have a minimal additional effect on outcome. Among patients with generalized epilepsy of pre-sumed genetic origin (formerly referred to as idiopathic generalized epilepsy), a high seizure burden might be respon-sible for persistent encephalopathy, but once seizures are aborted, patients may recover with few or no long-term sequelae. Among patients with acute focal stroke,
meningo encephalitis, TBI, or septic or meta bolic encepha lopathy, however, out-comes are more variable; in these situ ations, seizures might have a more important modu lating effect on neurological outcome. A clearer understanding of the relation-ships between seizure burden and outcome in the context of a patient’s diagnosis is urgently needed to help guide clinicians about when, and how aggressively, to treat subclinical seizures.
cEEG monitoring remains beyond the reach of many medical centres because of the considerable equipment and staffing costs. Qualified EEG technologists required for electrode application and clinical neuro-physiologists required for interpretation are in short supply, particularly outside normal working hours. Creative strategies can, however, be applied to make cEEG monitoring more feasible. Intensive care nurses can be taught to apply a limited array of EEG electrodes or electrode caps, reducing reliance on EEG technologists. Quantitative EEG trending algorithms that time compress and simplify the EEG display can help to speed up EEG interpretation by clinical neurophysiologists, and enable bedside caregivers to more easily screen for seizures. These and other challenges must be overcome to enable greater access to brain monitoring in both resource-rich and resource-poor settings.
The care of critically ill patients with acute brain injury is challenging. The clinical neurological assessment is often compromised by the use of sedating or paralysing medications. In these situations, the addition of continuous brain monitor-ing has the potential to provide important
Figure 1 | Schematic illustration of potential relationships between seizure burden and outcome. The potential deleterious effects of seizures in the context of acute brain injury are likely to depend on the underlying aetiology. The probability of poor outcome might increase linearly or exponentially with increasing seizure burden, or a threshold might exist, above which seizures are harmful.
Seizure burden (e.g. min per h)
Prob
abili
ty o
f poo
r ou
tcom
e (%
)
0
100
Generalized epilepsy ofpresumed genetic origin
Meningoencephalitis
Stroke
Traumatic brain injury
Anoxic injury
20
80
60
40
© 2013 Macmillan Publishers Limited. All rights reserved
58
Table 1.1 Recent studies of interest on outcomes following electrographic seizures and electrographic status epilepticus in children and neonates
References Population and comorbidities
Type of monitoring
Seizure threshold
When outcome was assessed
Main Findings
Studies in Critically Ill Children Vaewpanich et al. 2016
16 patients with TBI
cEEG SE = ongoing seizure >30 min
Short-term (at discharge and 4-6 wks post- discharge)
- All patients with seizures had poor outcome on the GOS-E peds, and speech pathology neurocognitive/functional evaluations (SPNFE)
Abend et al. 2015
60 patients with normal neurodevelopment prior to PICU admission for AMS
cEEG
SE = seizure >30 min or SB >50% in any 1-h epoch
Long-term (median 2.6 years)
- ES and ESE were associated with worse adaptive functioning scores on multivariate analysis - Non-significant trend to worse scores on behavioral-emotional & executive function scales after ESE
O'Neill et al. 2015
144 patients with TBI
cEEG SE = seizure >30 min or >50% of EEG recording
Short-term (at discharge)
- Presence of seizures did not correlate with discharge disposition - Seizures associated with longer hospital and ICU stay (no difference between ESE compared to ES)
Payne et al. 2014
259 paediatric and neonatal patients admitted to ICU with clinically ordered cEEG
cEEG Maximum hourly SB >20%/h
Short-term (at discharge)
- No association with mortality - ↑ probability and magnitude of neurologic decline with SB > 20% (12 minutes) - On multivariate analysis, every 1% ↑ in maximum hourly SB was associated with a 1.13 odds of neurological decline
Sanchez Fernandez et al. 2014
98 patients with CSE who underwent cEEG monitoring
cEEG SE = seizure >30 min or >50% SB in any 1-h epoch
Short-term (at discharge)
- No association with mortality - Longer PICU stay if ES and significantly longer if ESE
Wagenman et al. 2014
60 patients with normal neurodevelopment prior to PICU admission for AMS
cEEG SE = seizure >30 min or SB >50% in any 1-h epoch
Long-term (median 2.7 years)
- ESE, but not ES, associated with worse long term functional outcome scores and lower health-related quality of life - ESE associated with ↑ risk of subsequent diagnosis of epilepsy
Abend et al. 2013
550 consecutive PICU patients who underwent cEEG monitoring
cEEG SE = seizure >30 min or SB >50% in any 1-h epoch
Short-term (at discharge)
- ESE, but not ES, associated with greater odds of in-hospital death - ↑ PICU LOS if ESE compared to no ES, and compared to ES without ESE
Arndt et al. 2013
87 patients with TBI
cEEG
SE = seizure >15 min or >3 seizures/h
Short-term (at discharge)
- ESE associated with ↑ hospital LOS (no difference in PICU LOS) - ES and ESE associated with ↓ discharge KOSCHI score
59
Piantino et al. 2013
19 patients on ECMO
cEEG
SE = seizure >30 min or SB >50% in any 1-h epoch
Short-term (at discharge)
- ES were associated with structural abnormality on MRI - ES not associated with ↑ mortality -Trend for those with ES having more prolonged ECMO course - No difference in rate of complications
Topjian et al. 2013
200 patients with acute encephalopathy
cEEG SE = seizure >30 min or SB >50% in any 1-h epoch
Short-term (at discharge)
- ESE, but not ES, were associated with ↑ risk mortality - No difference in PICU or hospital LOS - ESE, but not ES, were associated with PCPC worsening
Gwer et al. 2012
82 patients with acute nontraumatic coma
cEEG ES = >6 sec; ESE = seizure >30 min or ≥3 seizures/hr
Short-term (at discharge)
- ES and SE associated with greater risk of poor outcome (death or gross motor deficits at discharge) - SE had even greater risk of poor outcome
Kirkham et al. 2012
204 comatose pediatric and neonatal patients
1-3 channel cEEG/ aEEG
Number, total duration, and longest ES
Short-term and long-term for 84 of the survivors
- ES associated with unfavourable short-term outcome (severe handicap or vegetative state) - No association with mortality
Studies in Critically Ill Neonates Glass et al. 2016
426 neonates with clinically suspected +/or EEG seizures
cEEG
High SB=≥7 ES; ESE=seizures >50% of at least 1-h of recording
Short-term (at discharge)
- High SB was a risk factor for mortality, ↑ hospital LOS, and abnormal neurological exam at discharge
Kharoshankaya et al. 2016
47 neonates with HIE
cEEG Total SB >40 min. Maximum hourly SB > 13 min
Long-term (24-48 months)
-Presence of seizures alone was not associated with abnormal outcome - A total SB of >40 min and a maximum hourly SB of >13 min/hr were associated with ↑ risk of abnormal development
Srinivasakumar et al. 2015
69 neonates with HIE
cEEG SB recorded in seconds
Short-term and long-term (18-24 months)
- ↑ SB associated with higher brain injury score - ↑ SB in combined cohort associated with lower performance on all 3 domains of BSID (cognitive, motor, language)
Shah et al. 2014
85 neonates with HIE
aEEG
High SB = SB >15 min per 1-h epoch ESE = SB >30 min per 1-h epoch
Short-term (MRI brain at median 9 days of life)
- High SB independently associated with greater injury on MRI (adjusted for 10 min Apgar and abnormal aEEG background)
60
Glass et al. 2011
56 neonates with HIE
cEEG
SE = seizure >30 min or SB >50% of 1–3 hrs of recording time
Short-term (MRI brain at median 5 days of life)
- Moderate to severe MRI injury more common in newborns with seizures, and present in all with SE - Moderate to severe MRI injury associated with seizures that were multifocal, later onset, and ongoing seizures following 20 mg/kg phenobarbital
Nash et al. 2011
41 neonates with HIE
cEEG
SE = seizure >30 min or SB >50% of 1–3 h of recording time
Short-term (MRI brain at median 5 days of life)
- Isolated or recurrent seizures were more frequent in patients with moderate to severe MRI injury compared with those with no or mild injury - SE was only seen in newborns with moderate to severe MRI injury
van Rooij et al. 2010
42 neonates with HIE
aEEG
Single seizure, ≥3 seizures per 30 min, or SE = seizure >30 min
Short-term (MRI brain at mean 5.5 days of life)
- Severity of brain injury on MRI associated with a longer duration of seizure patterns in the blinded group and the whole cohort
Studies in Neonates and Infants with CHD Undergoing Cardiac Surgery Naim et al. 2015
161 neonates with CHD
cEEG
SE = seizure >30 min or SB >30 min in any 1-h epoch
Short-term (at discharge)
- ↑ Mortality among neonates with seizures - No difference in cardiac ICU or hospital LOS
Gaynor et al. 2013
132 neonates and infants (<6 mo) with CHD
cEEG
ES >10 sec Long-term (4 years)
- ES were associated with worse executive function and impaired social interactions/restricted behaviour -ES were not associated with worse cognitive, language, or motor skills
Bellinger et al. 2011
139 infants with CHD (TGA)
cEEG
ES >6 sec Long-term (16 years old)
-ES in the postoperative period was the medical variable most consistently related to worse outcomes (academic achievement, memory, executive function, visual-spatial skills, and social cognition)
aEEG, amplitude integrated electroencephalogram; AMS, altered mental status; BSID, Bayley Scales of Infant Development; cEEG, continuous electroencephalogram; CHD, congenital heart disease; CSE, convulsive status epilepticus; ECMO, extracorporeal membrane oxygenation; ES, electrographic seizures; GOS-E Peds, Glasgow Outcome Scale scores; HIE, hypoxic-ischemic encephalopathy; KOSCHI, King’s Outcome Scale for Childhood Head Injury; LOS, length of stay; PCPC, Paediatric Cerebral Performance Category; SB, seizure burden; SE, status epilepticus; TGA, transposition of the great arteries.
61
CRITICALLY ILL NEONATES
Several studies have now shown that neonatal seizures and status epilepticus are associated with
worse outcomes even after adjusting for potential confounding variables, and support the
hypothesis that seizures worsen existing brain injury (see Table 1.1). However, increasing
seizure burden is often accompanied by greater use of antiseizure medications such as
phenobarbital and phenytoin, which have been shown to have neurotoxic effects in rodent
models (Bittigau et al., 2002; Bittigau et al., 2003). Disentangling the potential deleterious
effects of seizures from the potential toxicity due to their treatment remains an important
challenge.
Short-term and surrogate outcomes
Studies in neonates undergoing therapeutic hypothermia for HIE have shown a relationship
between higher seizure burden and worse MRI injury scores in univariate analysis (Glass et al.,
2011; Nash et al., 2011; Srinivasakumar et al., 2015; van Rooij, Toet, et al., 2010). In two of
these studies, all neonates with status epilepticus had moderate to severe injury on imaging
(Glass et al., 2011; Nash et al., 2011). In multivariable analysis, a higher seizure burden on
aEEG (more than 15 minutes in any 1-h epoch) was associated with greater injury on MRI,
independent of aEEG background and Apgar score at 10 min (Shah et al., 2014a).
A multicenter cohort study described 426 term and preterm neonates with various
diagnoses who underwent cEEG to characterize clinically suspected seizures or screen for
subclinical seizures. Electrographic seizures were found in 82% of neonates. In a multivariable
analysis, high seizure burden (defined as ≥7 electrographic seizures in total) was associated with
mortality, longer hospital stay, and abnormal neurological examination at discharge (defined as
abnormalities in consciousness, tone, and/or reflexes) (Glass et al., 2016).
Long-term outcomes
Seizures have also been linked to long-term neonatal outcomes. In a study of 47 neonates with
HIE who underwent cEEG monitoring, outcomes were assessed at 24 to 48 months using either
the Griffiths Mental Development Scales or BSID, as well as clinical diagnoses of cerebral palsy
or epilepsy. The presence of seizures per se was not associated with abnormal long-term
outcomes, but high total seizure burden and maximum hourly seizure burden were associated
with abnormal outcome. For every 1-min increase in total seizure burden the odds of an
62
abnormal outcome increased by 2.2%, and for every 1-min increase in maximum hourly seizure
burden the odds of an abnormal outcome increased by 16%. The seizure burden thresholds used
to predict greater odds of abnormal outcome were a total seizure burden more than 40 min or a
maximum hourly seizure burden more than 13 min, and these findings were independent of HIE
severity or the use of therapeutic hypothermia (Kharoshankaya et al., 2016).
Further support for clinically relevant electrographic seizure burden thresholds include
observations that a total neonatal seizure count more than 75 was associated with microcephaly,
severe CP, and failure to thrive (McBride et al., 2000), an ‘ictal fraction’ more than 10 min was
associated with an unfavorable neurodevelopmental outcome at 18 months (Pisani et al., 2008),
and that status epilepticus was a stronger risk factor than recurrent seizures for poor
neurodevelopmental outcome and postneonatal epilepsy at 24 months of age (Pisani et al., 2007).
NEONATES AND INFANTS WITH CONGENITAL HEART DISEASE
Post-operative seizures in children with CHD have been associated with worse outcome. A study
that systematically utilized cEEG for 48 hours in 161 neonates upon return to the ICU following
neonatal cardiac surgery found electrographic seizures in 8%, of whom 62% had ESE. Seizures
were subclinical in 85% of neonates. The presence of electrographic seizures was associated with
higher mortality (Naim et al., 2015).
Studies examining long-term outcomes after cardiac surgery have demonstrated that later
neurodevelopmental difficulties can emerge that may not have been evident at younger ages. The
greater academic and psychosocial demands placed on children as they grow older can reveal
neurodevelopment abnormalities potentially associated with increased seizure burden in the
developing brain. From an initial cohort of 178 neonates and infants with cardiac surgery
requiring cardiopulmonary bypass who underwent 48 h of cEEG monitoring in the postoperative
period, neurodevelopmental outcomes assessed at age 1 year in 114 children showed that
postoperative electrographic seizures were not associated with worse outcomes on the BSID,
except in the subset of children with frontal-onset seizures (Gaynor et al., 2006). When 132
children of the original cohort were subsequently assessed at age 4 years, postoperative
electrographic seizures were associated with worse executive function and impaired social
interactions or restricted behaviours, but were not associated with cognitive, language, or motor
outcomes (Gaynor et al., 2013).
63
The Boston Circulatory Arrest Study reported on successive long-term follow-up at age
1, 4, 8, and 16 years in a cohort of 171 infants with transposition of the great arteries that
underwent cardiac surgery requiring cardiopulmonary bypass during infancy. cEEG was
performed for at least 2 h pre-operatively, during surgery and for 48 h following surgery.
However, cEEGs were not interpreted clinically; therefore, electrographic seizures remained
unrecognized and untreated. Electrographic seizures were found in 20% and clinical seizures in
6% (Helmers et al., 1997). At age 4 years, increased seizures in the postoperative period were
associated with lower mean intelligent quotient scores and an increased risk of abnormalities on
neurological exam (Bellinger et al., 1999). At age 8 years, postoperative seizures were associated
with social and attention problems (Bellinger et al., 2009). At age 16 years, seizures in the
postoperative period was the medical variable most consistently related to worse outcomes,
specifically academic achievement, memory, executive function, visual-spatial skills, and social
cognition (Bellinger et al., 2011).
CONCLUSIONS
Seizures are a cardinal sign of underlying brain injury. There is mounting evidence that higher
seizure burden during critical illness in neonates and children is associated with worse
neurodevelopmental outcomes, even after adjusting for potential confounding variables such as
age, acute encephalopathy etiology, critical illness severity, MRI injury severity, EEG
background, Apgar scores or use of therapeutic hypothermia. Both animal and human studies
have identified plausible mechanisms by which seizures may cause secondary brain injury. To
date, however, clinical trials have been unable to demonstrate that lowering seizure burden
through more aggressive treatment improves MRI brain injury or short-term outcome. Therefore,
a causal link between increased seizure burden and worse outcomes remains unproven. Larger
randomized controlled trials will be required to distinguish the impact of seizure burden from
other non-modifiable factors such as age, genetic background, seizure etiology and severity of
brain injury, as well as the potential adverse effects of antiseizure medications.
64
1.8 Glucose disturbances and EEG
1.8.1 Hypoglycemia and EEG changes
Few studies have examined acute EEG changes associated with hypoglycemia. In newborns,
studies using aEEG have not been able to detect clear changes during hypoglycemia except for
one case report. A study of 12 neonates of diabetic mothers used subcutateous microdialysis
catheters for glucose measurements and visual interpretation of one-channel aEEG monitoring
(performed calculation of the baseline level, the amplitude and the frequency distribution).
Segments of aEEG during hypoglycemia (<2.2 mmol/L) were compared to recordings during
normoglycemia (>2.5 mmol/L) and no significant changes on aEEG during episodes of
hypoglycemia were found (Stenninger et al., 2001). Harris et al. used continuous interstitial
glucose monitoring and concurrent 3-channel aEEG (C3-P3, C4-P4, O1-O2) recordings in a
cohort of 101 well babies (≥32 weeks GA) at risk for hypoglycaemia. They added an occipital
channel to the standard aEEG because the occipital regions may be most sensitive to
hypoglycemic damage. Twenty-four of the neonates had low interstitial glucose concentrations
(<2.6 mmol/L) during the period with matching aEEG data, and a total of 103 hypoglycemic
episodes were recorded. Total EEG intensity (µV2), spectral edge frequency, amplitude and
continuity measures were assessed, and manual review was performed by a pediatric neurologist
for seizures. No changes on aEEG recordings were identified between 20-minute periods of low
glucose concentration and adjacent 20 minutes of normoglycemia from the same baby (Harris et
al., 2011). The same group had similarly used a modified 3-channel aEEG (parietal and occipital
montages) in newborn lambs during insulin-induced hypoglycemia and had also not detected
changes in any of the measured aEEG parameters during hypoglycemia (D. L. Harris et al.,
2009). Although most studies using aEEG failed to detect any changes during hypoglycemia, a
case report of a term newborn with sepsis and necrotizing enterocolitis detected changes on
single channel aEEG during periods of hypoglycemia. Seizures were suspected initially during
hypoglycemia. This was followed by very little aEEG activity, but the aEEG activity returned to
normal when blood glucose levels improved (Hellstrom-Westas et al., 1989).
In contradistinction to the clinical experience, animal models of insulin-induced
hypoglycemia studies have shown progressive background slowing of the EEG as glucose levels
fall. In newborn dogs, slowing was seen at or below glucose levels of 1.5 mmol/L and
65
progressed to paroxysmal discharges and convulsive seizures as glucose levels below 0.6
mmol/L. A burst suppression pattern was seen in 2 animals at glucose levels of 0.1 mmol/L and
0.3 mmol/L (Vannucci et al., 1981). Another study by Vannucci et al. using insulin-induced
hypoglycemia in newborn dogs suggested that the EEG response to hypoglycemia may depend
not only on the concentration of glucose in the brain but also on the presence of associated
metabolic derangements. No change was seen on EEG in the newborn dogs with hypoglycemia
alone, but the EEG became isoelectric following respiratory arrest (Vannucci et al., 1980).
Studies in adult rats have also shown a slowing of the EEG pattern during insulin-induced
hypoglycemia as well as convulsive polyspike activity and progression to burst suppression and
an isoelectric EEG (Agardh et al., 1983; Auer et al., 1984; Bryan et al., 1994; Lewis, Ljunggren,
Norberg, et al., 1974; Lewis, Ljunggren, Ratcheson, et al., 1974; Norberg et al., 1976). In adult
rats no pathological brain damage was seen unless an isoelectric or burst suppression pattern was
seen on the EEG during hypoglycemia (Auer et al., 1984).
As opposed to neonates, young children may manifest changes on the EEG during
hypoglycemia. An interesting study in diabetic and nondiabetic children performed quantitative
spectral analysis of the EEG during a gradual decline in glucose concentrations using insulin. At
a plasma glucose concentration of 4 mmol/L, an increase in delta and theta amplitudes were seen
in both groups. At 3 mmol/L, there was a further and more widespread increase in low-frequency
EEG activity, with greater slowing and epileptiform discharges seen in the diabetic than
nondiabetic children (Bjorgaas et al., 1998).
There are also several case reports of hypoglycemia in children with concomitant EEG
monitoring. A case series of idiopathic spontaneous hypoglycemia describes a 3-year-old girl
with recurrent hypoglycemia who had an EEG performed during the fasting state (blood glucose
of 2.3 mmol/L) which demonstrated slow wave activity which improved after treatment and
glucose improved to 4.1 mmol/L. In the same series, a 10-month-old girl’s EEG showed high
voltage, widespread slow activity while her blood glucose was 1.9 mmol/L. Repeat EEG several
days later when the blood glucose was 2.3 mmol/L continued to show slow activity, although it
had improved from previous recordings. At that point the child was treated and 30 minutes later
with a blood glucose of 4.2 mmol/L the EEG had normalized (Haworth et al., 1960). Another
case reported in this series was that of an 11-month-old girl with hypoglycemia (0.3 mmol/L)
and coma due to an accidental overdose of a sulfonylurea. The EEG recording commenced
66
during treatment with IV dextrose, and demonstrated asymmetric irregular medium amplitude
slow activity appearing in runs as well as runs of irregular fast activity, periodic suppression and
frequent seizures. The EEG gradually improved but after 2 weeks there were still persistent
abnormalities (excess irregular slow activity and small spikes or sharp waves). Long term the
child developed epilepsy with an EEG that showed persistent abnormalities (Pavone et al., 1980).
Finally, one study in neonates demonstrated persisting EEG changes within one week
following symptomatic neonatal hypoglycemia (plasma glucose levels <2.2 mmol/L).
Conventional 50-minute EEG recordings were performed in a cohort of 20 newborns with
hypoglycemia. The authors observed an increased density of frontal sharp transients in all stages
of sleep and less bilateral synchrony compared to controls. However, the cohort of neonates with
hypoglycemia had various associated pathologies (such as hydrocephalus, asphyxia, sepsis,
toxoplasmosis) whereas controls were normal newborns with no evidence of CNS or other
disorders (Lahorgue Nunes et al., 2000). Another study had examined a cohort of 28 children of
diabetic mothers, of which 13 children had neonatal hypoglycemia (<1.5 mmol/L), for persistent
EEG changes at 8 years of age. Quantitative frequency analysis of the EEGs found no significant
difference between children with or without neonatal hypoglycemia in maximal power frequency
or the relative power frequency distribution quotient. However, a lower relative delta power
frequency distribution in the frontotemporal region and a higher relative alpha power frequency
distribution in frontal and parietal regions was identified in children with neonatal hypoglycemia
vs. those without (Stenninger et al., 1998)
1.8.2 Hyperglycemia and EEG changes
In preterm neonates, hyperglycemia has been shown to be associated with acute EEG and aEEG
changes (Granot et al., 2012; Schumacher et al., 2014; Wikstrom et al., 2011). In extremely
preterm neonates (22-27 weeks GA) Wikström et al. found a U-shaped relationship with plasma
glucose levels, whereby a greater discontinuity was seen on aEEG at plasma glucose
concentrations both above or below 4.0 mmol/L. Moderate hyperglycemia was significantly
associated with greater discontinuity on aEEG. These authors measured IBIs using an automated
algorithm and used the average IBI from 10-minute artifact-free epochs before each blood
sample. In addition, it was observed that increasing PaCO2 was also associated with longer IBIs
67
and aEEG power was attenuated at both higher and lower PaCO2 values (Wikstrom et al., 2011).
Similarly, a prospective cohort study by Granot et al. of premature neonates (<28 weeks’
gestation) during the first 3 days of life found that higher levels of blood glucose were
significantly associated with depressed cerebral activity on aEEG. aEEG patterns were assessed
in the 30 minutes preceding blood sampling. Respiratory acidosis also was correlated with
depressed cerebral activity (Granot et al., 2012). Using cEEG monitoring (8 EEG electrodes)
Schumacher et al. showed that increased blood glucose concentrations were associated with
decreasing total absolute band power during the first 3 days of life in premature infants (<30
weeks GA). They calculated the mean total absolute band power on each day and correlated it
with mean glucose levels and highest glucose levels in each neonate. These authors also
observed an effect of hypercapnia on EEG, with decreased total absolute band power
(Schumacher et al., 2014).
A study in fetal sheep that examined the effects on EEG with glucose infusion prior to
ischemic insult showed that the amplitude and power of the EEG during ischemia was
significantly higher in animals given glucose infusions than saline alone. Further, higher blood
glucose concentrations appeared to help maintain EEG activity during ischemia. In this study,
ischemia caused cerebral oxygen consumption to decrease, glucose uptake to increase and a net
efflux of lactate to occur. The authors hypothesized that elevated blood glucose helped maintain
EEG activity during ischemia perhaps by fueling additional anaerobic energy production (Chao
et al., 1989). However, Lear et al. demonstrated in fetal sheep that although higher glucose levels
improve neurophysiologic adaptation to asphyxia, there is more marked secondary deterioration
and impaired neurophysiologic recovery (Lear et al., 2017). Their findings were consistent with
the so-called ‘‘glucose paradox’’ that is seen in adult animals, whereby hyperglycemia is
associated with improved metabolic function during hypoxia-ischemia, but later development of
greater brain injury (Lear et al., 2017; MacDougall et al., 2011). In preterm fetal sheep
(equivalent to 28-32 weeks human fetuses) either dexamethasone, glucose infusion or saline
were injected before asphyxia. The authors found a higher EEG power and attenuated rate of rise
of cortical impedance in dexamethasone and glucose treatment groups during asphyxia.
However, the dexamethasone and glucose treatment groups showed a marked secondary rise in
cortical impedance and impaired recovery of spectral edge frequency. Dexamethasone and
glucose treatment were also associated with increased numbers of post-asphyxial seizures. At the
68
end of the experiment, there were burst suppression patterns seen in the dexamethasone and
glucose treatment groups which was not observed in the saline-asphyxia group (Lear et al.,
2017).
A study in rats also showed that recovery of EEG activity following cerebral ischemia
was severely impaired in hyperglycemic animals. Following cerebral ischemia, the cerebral
activity reappeared later in the hyperglycemic group with persistent burst suppression pattern
seen at 1 hour, whereas the EEG of normoglycemic group had normalized by that time
(Siemkowicz et al., 1981).
Lastly, we could also consider studies in children with hyperglycemia. A study used
CGM and simultaneous EEG monitoring for 40 hours in children with type 1 diabetes and found
that asymptomatic hyperglycemia (>11 mmol/L and mainly >15.5 mmol/L) was associated with
changes in the power spectra of different frequency bands. During wakefulness, hyperglycemia
was associated with increased EEG power of low frequencies (theta and delta) and decreased
EEG power of high frequencies (beta and gamma). During sleep, hyperglycemia was associated
with increased EEG power of low frequencies in all areas and increased high frequencies in
frontal and central regions (Rachmiel et al., 2016).
In summary, a number of animal and clinical studies suggest that hypoglycemia may be
associated with progressive slowing on the EEG which may not be detected on aEEG monitoring
except perhaps in the presence of critical illness or associated metabolic derangements. However,
hyperglycemia is more consistently associated with increased slow activity, increased
discontinuity and depressed cerebral activity. Changes with hyperglycemia have been detected
on both aEEG and EEG monitoring.
69
1.9 Research Hypothesis and Aims
The primary objective of this study was to understand the relationship between glucose levels
and brain activity in neonates with encephalopathy.
Hypoglycemia and hyperglycemia are common problems in neonates and may be
associated with greater brain injury and worse neurodevelopmental outcomes in neonatal
encephalopathy. However, the level and duration of hypo- and hyperglycemia that is harmful to a
neonate’s developing brain is not known. There is some evidence that glucose derangements can
cause acute neurophysiologic changes, but the specific effects of blood glucose disturbances on
the EEG in neonatal encephalopathy remain poorly understood. EEG abnormalities and recovery
are predictive of long-term neurodevelopmental outcome in neonates with HIE. Thus, these
investigations aim to detect changes in global brain function associated with hypo- and
hyperglycemia using aEEG and cEEG monitoring, and assess whether cEEG monitoring could
be a suitable method to determine the degree and duration of hypo- and hyperglycemia that
correlate with changes in global brain function to aid in determining clinically significant target
glucose ranges in these neonates.
1.9.1 Aims
1. To determine whether abnormal glucose levels and glucose variability in neonates
with encephalopathy are associated with immediate changes in brain background
activity and electrographic seizures, as measured using aEEG.
2. To determine whether abnormal glucose levels in neonates with encephalopathy are
associated with immediate changes in brain background activity and electrographic
seizures, as measured using cEEG.
70
1.9.2 Hypothesis
In keeping with our primary aims, we will test the following hypotheses:
1. Hypoglycemia, hyperglycemia and glucose variability in neonates with
encephalopathy are associated with worse brain background activity and increased
electrographic seizures on aEEG.
2. Hypoglycemia and hyperglycemia in neonates with encephalopathy are associated
with abnormal background activity and increased electrographic seizures on cEEG.
71
Chapter 2
Methods 2
2.1 Study population
In an ongoing prospective cohort study enrolling term neonates with encephalopathy transferred
to The Hospital for Sick Children, subjects enrolled between August 2014 and March 2017 were
considered for this analysis. This comprises a consecutive convenience sample of the
Neurological Outcome of Glucose in Neonatal encephalopathy (NOGIN) study cohort. The
institutional research ethics board approved the study protocol and the parents or guardians of all
participating neonates provided written informed consent. Newborns with encephalopathy were
eligible if they had abnormal consciousness, in addition to either neonatal seizures or
abnormalities in tone or reflexes. Exclusion criteria included suspected or confirmed congenital
malformations, inborn errors of metabolism, congenital infections, gestational age <36 weeks,
weight <1500 grams or if it was expected that a CGM could not be attached within 6 hours of
life.
Clinical data was collected from patient medical records. All point of care testing and
laboratory glucose measurements obtained clinically were collected. Point of care testing
performed with i-STAT® (Abbott Laboratories, Abbott Park, Illinois, USA) uses the gold
standard glucose oxidase reaction, and was thus considered as laboratory values. Study data were
stored and managed using research electronic data capture (REDCap) hosted at The Hospital for
Sick Children (P. A. Harris et al., 2009).
Therapeutic hypothermia (TH) was initiated within six hours after birth if clinically
indicated, with target temperature (33–34 °C core temperature) continued for 72 hours followed
by gradual rewarming by 0.5 °C per hour over 6 hours. Low-dose morphine or fentanyl was
provided during hypothermia as needed.
72
2.2 Amplitude-integrated electroencephalography (aEEG) monitoring and analysis
Amplitude-integrated electroencephalography was commenced as soon as possible after
admission as part of routine clinical care and interpreted by the neonatology team. Subsequent
research analysis of aEEG recordings was performed by a single neurologist (E.F.P.), blinded to
clinical course, glucose levels, and neurological outcome. Only dual channel aEEG recordings
(C3-P3, C4-P4) were included. Over 6-hour epochs, the aEEG background was graded on an
ordinal scale: 0:continuous normal voltage (CNV, aEEG maximum >10 µV and minimum >5
µV); 1:discontinuous normal voltage (DNV, aEEG maximum >10 µV and minimum ≤5 µV);
2:burst suppression pattern (virtual absence of activity [<2 µV] between bursts of high voltage
[>25 µV]); 3:continuous low-voltage (CLV, aEEG maximum ≤10 µV); and 4:flat trace
(isoelectric activity, aEEG maximum ≤5 µV). The predominant, best and worst background score
present in each epoch were graded. Sleep-wake cycling was graded as: 0:developed cycling;
1:immature cycling; and 2:no cycling. Seizures were graded as: 0:no seizures; 1:single seizure;
2:repeated seizures (2 or more); and 3:status epilepticus (‘saw-tooth pattern’), defined as
continuous or recurrent seizures lasting at least 30 minutes, or for more than 50% of the
recording time (Hellström-Westas et al., 2008; Hellström-Westas et al., 2006; Oh et al., 2013; Oh
et al., 2015; Thoresen et al., 2010; Toet et al., 1999) (Figure 2.1).
2.3 Continuous electroencephalography (cEEG) monitoring and analysis
Continuous electroencephalography (cEEG) was arranged as soon as possible after admission,
lasting 48 hours unless clinically indicated. They were acquired using portable Stellate Harmonie
or Xltek Brain Monitor ICU video-EEG systems (Natus Neurology, Oakville, Ontario, Canada).
Electrodes were applied by registered EEG technologists according to the international 10–20
system of electrode placement modified for neonates (Fp1, C3, T3, O1, A1, Fp2, C4, T4, O2,
A2, Fz, Cz, and Pz). With the introduction of the Xltek Brain Monitor ICU video-EEG systems,
P3 and P4 electrodes were added in order to display Persyst aEEG trend on the same device for
the bedside caregivers. All neonates had simultaneous recording of deltoid electromyogram,
73
electrocardiogram and abdominal respiration. As per institutional protocol, a clinical cEEG was
requested if there were definite or suspected clinical seizures, definite or suspected
electrographic seizures on aEEG monitoring, or abnormal background activity on aEEG
monitoring defined as discontinuous normal voltage, burst suppression or suppression. Before
initiating clinical cEEG monitoring, all patients are assessed by our neurology consultation
service. The duration of clinical cEEG monitoring is typically 24 hours, or longer when
electrographic seizures are detected. Thus, the studies were prolonged to complete 48 hours of
recording if no longer clinically indicated prior, or were continued as long as clinically indicated
beyond 48 hours. Clinical and electrographic seizures were treated with medications according to
institutional guidelines and at the discretion of the treating physician. If clinical cEEG
monitoring was not indicated, research cEEGs were arranged and continued for 48 hours. If
seizures were detected during research cEEG monitoring, neurology consultation service was
advised and further treatment and cEEG monitoring was managed by the treating physicians as
described above for clinical cEEG monitoring.
Subsequent research analysis of cEEG recordings was performed by a single neurologist
(E.F.P.), blinded to clinical course, glucose levels and neurological outcome. The EEG
recordings were visually inspected and 5-minute EEG epochs were classified according to the
following categories according to accepted criteria (Tsuchida et al., 2013) as 0:Continuous or
tracé alternant (IBI voltage ≥25µV with IBI duration ≤6 seconds); 1:tracé alternant with
excessive discontinuity (IBI voltage ≥25µV with IBI duration >6 seconds); 2:tracé discontinu:
(IBI voltage <25µV); 3:depressed and undifferentiated (with persistently low-voltage
background activity with amplitude between 5 µV and 15 µV and without normal features) or
burst suppression (abnormally composed EEG bursts separated by prolonged and abnormally
low voltage IBI periods <5 µV pp); 4:very low voltage (amplitude <5 µV or with no discernible
cerebral activity). Epochs were skipped if the background pattern was unable to be assessed due
to seizure or excessive movement artifact (Figure 2.2).
Electrographic seizures were defined as a sudden, abnormal, repetitive and evolving
pattern, with a minimum voltage of 2 mcV pp and minimum duration of at least 10 seconds.
“Evolving” was defined as an unequivocal evolution in frequency, voltage, morphology, or
location. Distinct seizure events were separated by at least 10 seconds. Status epilepticus was
present when the summed duration of seizures comprised ≥50% of an arbitrarily defined 1-hour
74
epoch (Tsuchida et al., 2013). Seizure presence, duration and maximal extent over course of
seizure was marked. Video recordings were reviewed and clinical correlate was classified as
absent or present. Clinical correlates were further defined as 1:clonic; 2:tonic; 3: subtle (e.g. eye
deviation, eye opening); 4:automatisms (e.g. bicycling, lip smacking); 5:autonomic only (e.g.
heart rate changes); or unable to assess (e.g. caregiver in front of patient, patient not on video).
Every 6 hours, the presence of reactivity, variability and state changes was graded as 0:present;
1:some/immature; or 2:none.
2.4 Continuous interstitial glucose monitoring
Medtronic iPro2TM professional continuous glucose monitors with EnliteTM sensors (Medtronic
of Canada Ltd., Brampton, Ontario) were placed as early as possible after study enrolment, either
during transport or upon admission to hospital. The sensors were inserted interstitially into the
lateral aspect of the thigh, secured with clear adhesive dressing. In both term and preterm
neonates, this device has been used safely over a 7-day period and has been shown to be well
tolerated (Beardsall et al., 2013; Harris et al., 2010). The iPro2 monitor is a blinded device that
records average interstitial glucose concentrations every 5 minutes. This data was not made
available to the care team due to limited evidence for treating newborns based on these measures.
Clinicians treated newborns for hypoglycemia or hyperglycemia based on current standard of
care using intermittent glucose testing. As per institutional protocol, blood glucose
concentrations less than 2.7 mmol/L were treated with intravenous dextrose or glucagon.
Hyperglycemia was treated with insulin infusion as clinically indicated.
Continuous glucose monitoring was continued for 72 hours, after which the data was
downloaded onto a Windows-based notebook computer running Medtronic software. Only
laboratory blood glucose values were used to calibrate the CGM. Due to software cut-offs,
interstitial glucose readings below 2.2 and above 22.2 mmol/L were recalculated manually from
the raw data.
75
The definition of hypoglycemia remains controversial and there are variable thresholds
used in different consensus guidelines and research studies. On the basis of neurophysiological
and neurodevelopmental outcome studies, a commonly used definition of neonatal hypoglycemia
is less than 2.6 mmol/L (Koh et al., 1988; Lucas et al., 1988). However, the PES advises that a
‘safe target’ during the first 48 hours should be close to the mean for a healthy newborn and
above the threshold for neuroglycopenic symptoms (2.8 mmol/L) (Thornton et al., 2015).
Additionally, reference ranges in healthy term newborns may not be appropriate in infants at risk
for impaired metabolic adaptation and at risk for adverse neurodevelopment, such as with
perinatal asphyxia (Harding et al., 2017). The PES considers that in hypoketotic conditions,
ketones and lactate may not be available in sufficient concentrations to serve as an alternative
fuel for the brain and thus there is greater risk of brain energy failure and hypoglycemia-induced
brain damage (Thornton et al., 2015).
Accordingly, hypoglycemia was defined as blood or interstitial glucose ≤2.8 mmol/L and
hyperglycemia as glucose >8.0 mmol/L. Episodes of interstitial glucose derangements were
defined as two or more consecutive data points (≥10 minutes) outside the normal range. Episodes
of glucose derangement were treated as contiguous if they were separated by brief periods of
normoglycemia lasting ≤10 minutes.
For each subject, mean, minimum and maximum glucose levels were calculated based on
the interstitial measurements obtained over each 6-hour epoch. The area under the curve was
calculated as the area of interstitial glucose concentration over/under the normal glucose range
(hours*mmol/L), reflecting the extent and duration of the episode of interstitial glucose
derangement. Specifically, for hypoglycemia the area of interstitial glucose concentrations ≤2.8
mmol/L was calculated and for hyperglycemia the area of interstitial glucose concentrations >8.0
mmol/L was calculated. Glucose variability was quantified using the standard deviation and
mean glucose rate of change per hour (mmol/L/hour).
76
2.5 Statistical analysis
All statistical analyses were computed using SPSS, version 20 (SPSS, Inc, Chicago, IL, USA).
Demographic data were summarized by standard descriptive statistics, including
median/interquartile range (IQR) or mean/standard deviation (SD) for continuous variables, or as
number and percentages for categorical variables.
2.5.1 Statistical analysis of aEEG monitoring
The primary analyses compared aEEG scores (background, sleep-wake cycling and seizure
scores) during 6-hour epochs containing episodes of hypoglycemia or hyperglycemia to
normoglycemic epochs using generalized estimating equations for repeated measures with a
linear scale response model. An independent correlation structure and robust variance estimators
were used. Comparisons were made relative to normoglycemia because a prior study reported
greater aEEG background discontinuity at plasma glucose concentrations both above or below
4.0 mmol/L (Wikstrom et al., 2011). Findings were adjusted for clinical markers of hypoxia-
ischemia severity: Apgar scores, umbilical artery pH, and base deficit. Secondary analyses
investigated the association of the aEEG scores with other glucose measures (mean, maximum or
minimum glucose values, duration of glucose disturbance, and area under the curve) and
measures of glucose variability (standard deviation and mean glucose rate of change per hour).
Two-tailed tests with p values of <0.05 were considered statistically significant.
77
Figure 2.1 Summary of aEEG analysis methods. Analysis was performed on 6-h epochs of concomitant CGM and
aEEG. CGM provides average interstitial glucose values every 5-min. Six-hour epochs of CGM were assessed for
the presence or absence hypo- or hyperglycemia, as well as calculation of the mean, minimum, maximum, area
under the curve, standard deviation and mean glucose rate of change in each 6-h epoch. aEEGs were scored every 6-
h for background, sleep wake cycling (SWC) and seizures. Background was graded as: 0:continuous normal voltage
(CNV); 1:discontinuous normal voltage (DNV); 2:burst suppression pattern (BS); 3:continuous low-voltage (CLV);
and 4:flat trace (FT). SWC was graded as: 0:developed cycling; 1:immature cycling; and 2:no cycling. Seizures
were graded as: 0:no seizures; 1:single seizure; 2:repeated seizures; and 3:status epilepticus.
2.5.2 Statistical analysis of cEEG monitoring
The primary analyses compared cEEG scores background scores and presence of seizures during
5-minute epochs with or without episodes of hypoglycemia or hyperglycemia to normoglycemic
epochs using generalized estimating equations for repeated measures with a linear scale response
model. An independent correlation structure and robust variance estimators were used. All
comparisons were again made relative to normoglycemia. Findings were adjusted for clinical
markers of hypoxia-ischemia severity: Apgar scores, umbilical artery pH, and base deficit.
Secondary analyses included comparison of mean interstitial glucose concentration in each 5-
CGM
aEEG background
aEEG SWC
aEEG seizures
aEEG 6hours 6hours 6hours
78
minute epoch with cEEG background scores and presence of seizures. A subgroup analysis was
performed examining the association of hyperglycemia with cEEG background scores only in the
8 patients that had occurrences of hyperglycemia on CGM monitoring as it is hypothesized that
neonates with hyperglycemia may have more severe HIE and thus worse brain activity.
Subgroup analysis was also performed examining the association of hypoglycemia with cEEG
background scores only in the 5 patients that had occurrences of hypoglycemia on CGM. Two-
tailed tests with p values of <0.05 were considered statistically significant.
Figure 2.2 Summary of cEEG analysis methods. Analysis was performed on 5-min epochs of concomitant CGM
and cEEG. CGM provides average interstitial glucose values every 5-min. The presence or absence hypo- or
hyperglycemia in each 5-min epoch as well as the mean interstitial glucose values every 5-min were assessed. cEEG
background was visually graded every 5-min as: 0:Continuous or tracé alternant; 1:tracé alternant with excessive
discontinuity; 2:tracé discontinu; 3:depressed and undifferentiated or burst suppression; 4:very low voltage. Seizures
were marked as present or absent, as well as the duration of seizure in each 5-min epoch.
CGM
cEEG background
cEEG seizure
cEEG5min 5min 5min
79
Chapter 3
Results 3
3.1 Prevalence of glucose derangements in cohort and treatments received
Eighty families were approached to participate in the study and 29 declined (enrollment rate
64%), accordingly 51 neonates were included for analysis. Overall, 41 of 51 neonates (80%) had
abnormal glucose values in the first 3 days of life when including both interstitial values as well
as blood glucose measurements obtained prior to insertion of the CGM. The mean age at the start
of CGM measurements was 8.9 hours of life (SD 3.9 h). There were 13 neonates with
hypoglycemia only (25%), 18 neonates with hyperglycemia only (35%) and 10 neonates with
both hypo- and hyperglycemia (20%). Of the first 51 neonates recruited, 2 had no available
interstitial glucose values. Twenty-five neonates had glucose derangements recorded on CGM
(51%). There were 7 neonates with hypoglycemia only (14%), 15 with hyperglycemia only
(31%) and 3 with both hypo-and hyperglycemia (6%).
During initial resuscitation after birth, 9 neonates required epinephrine, of which 6
developed hyperglycemia within 12 hours after birth (median 2.5 hours of life, IQR 1.94-8.9 h)
and one developed hyperglycemia at 12.67 hours of life. Thirty-three neonates had hypotension
requiring intervention (65%), of which 20 neonates (39%) had hypertension responsive to fluid
boluses. The other 13 neonates (25.5%) required further treatment, including dopamine,
dobutamine, epinephrine, norepinephrine, hydrocortisone or vasopressin. Four of these neonates
developed hyperglycemia after receiving vasopressors.
In the 49 patients with CGM monitoring, there were 22 episodes of hypoglycemia (≤2.8
mmol/L) in 10 neonates, of which there were 16 episodes with a minimum glucose ≤2.6 mmol/L
in 8 neonates. Only three of these episodes on CGM were detected and treated clinically with
D10W IV boluses. Overall, 17 neonates received treatment for hypoglycemia; 14 neonates
required treatment within the first 6 hours of life, 3 only had treatment for hypoglycemia after 6
hours of life, and 4 received both early and later treatment of hypoglycemia. Treatment of
hypoglycemia included IV dextrose boluses or glucagon (IM or infusion). Of the 28 neonates
80
with hyperglycemia episodes since birth, 16 had episodes within the first 6 hours of life. Seven
neonates received treatment for hypoglycemia prior to development of hyperglycemia. Four
neonates required an insulin infusion for treatment of hyperglycemia and all insulin infusions
were started after the first 6 hours of life.
3.2 Amplitude-integrated EEG
3.2.1 Study population
Of the 51 neonates recruited, 6 neonates were excluded from aEEG analysis (2 had no available
interstitial glucose values, 3 had only single channel (C3-C4) recordings and 1 had no aEEG
trace), leaving 45 eligible neonates. Demographic information is summarized in Table 3.1.
Compared to neonates who maintained normoglycemia on CGM, neonates with hypoglycemia
on CGM had higher mean gestational age and higher mean umbilical artery pH. One patient was
not cooled (did not meet institutional criteria for therapeutic hypothermia) and 4 patients died
during the neonatal period. Characteristics of excluded participants are found in Appendix Table
1.
81
Table 3.1 Clinical and demographic data of the full study cohort, and subgroups of neonates with hypoglycemia and hyperglycemia, who were compared to those who maintained normoglycemia on continuous interstitial glucose monitoring
FULL COHORT
n=45
NORMOGLYCEMIA n = 25
HYPOGLYCEMIA n = 7
HYPERGLYCEMIA n = 11
p value p value
Sex 21 ♀: 24 ♂ 12 ♀: 13 ♂ 3 ♀: 4 ♂ 1.000 5 ♀: 6 ♂ 1.000 Gestational age (weeks), mean (SD)
39.54 (1.40) 39.19 (1.51) 40.53 (0.48) 0.001 39.35 (1.16) 0.746
Birth weight (grams), mean (SD)
3414.36 (548.24)
3468.80 (670.03)
3413.29 (274.80)
0.746 3337.09 (399.10)
0.470
Birth length (cm), mean (SD)
50.14 (3.15) 50.63 (3.16) 49.43 (3.77) 0.462 49.59 (3.14) 0.373
Head circumference (cm), mean (SD)
34.22 (1.31) 34.31 (1.47) 34.14 (0.85) 0.703 34.14 (1.33) 0.727
Maternal diabetes, n (%) Type II Diabetes Mellitus Gestational Diabetes
5 (11) 1 (2) 4 (9)
4 (16) 1 (4)
3 (12)
0 (0) 0 (0) 0 (0)
0.552 1 (9) 0 (0) 1 (9)
1.000
Maternal hypertension, n (%) Maternal pre-eclampsia/ eclampsia
4 (9)
2 (4.5)
3 (12)
2 (8)
0 (0)
0 (0)
1.000 1 (9)
0 (0)
1.000
Apgar score at 5 min, mean (SD)
4.18 (2.39) 3.68 (1.82) 5.43 (3.10) 0.196 4.55 (2.51) 0.318
Umbilical arterial cord pH, mean (SD)
6.99 (0.18) 6.96 (0.16) 7.17 (0.17) 0.026 6.95 (0.20) 0.924
Umbilical arterial cord BE, mean (SD)
-15.13 (6.79) -15.83 (6.32) -9.00 (7.43) 0.077 -17.13 (6.83) 0.646
Route of delivery, n (%) Vaginal delivery Cesarean section
20 (44.5) 25 (55.5)
13 (52) 12 (48)
2 (28.5) 5 (71.5)
0.402
5 (45.5) 6 (54.5)
1.000
Sentinel event, n (%)* 17 (38) 9 (36) 2 (28.5) 1.000 5 (45.5) 0.716 Seizures, n (%)** 28 (62) 14 (56) 5 (71) 0.671 7 (63.5) 0.729 Clinical seizures only 16 (35.5) 9 (36) 5 (71) 1 (9) Electrographic and/or electroclinical seizures
12 (26.5) 5 (20) 0 (0) 6 (54.5)
Antiepileptic medications, n (%)
26 (58) 12 (48) 5 (71) 0.402 7 (63) 0.481
Lorazepam only 6 (13) 3 (12) 2 (28.5) 0 (0) Phenobarbital 13 (29) 8 (32) 2 (28.5) 3 (27) ≥2AEDs 7 (15.5) 1 (4) 1 (14) 4 (36) Received opioids, n (%) 44 (98) 24 (96) 7 (100) 1.000 11 (100) 1.000 Minutes of concomitant aEEG and continuous glucose monitoring, mean (SD)
3256.92 (1280.02)
3135.60 (1263.51)
3401.43 (1264.65)
0.634 3563.18 (1204.24)
0.345
2subjectshadbothinterstitialhypoglycemiaandhyperglycemiaeventsduringconcurrentaEEGmonitoringContinuousvariableswereanalyzedwiththestudent’sT-test,whereascategoricalvariableswereanalyzedwithFisher’sexacttestSignificantpvaluesareshowninboldHypoglycemiaisdefinedasatleasttwointerstitialglucosemeasurementslessthanorequalto2.8mmol/Landhyperglycemiaisdefinedasatleasttwointerstitialmeasurementsgreaterthan8mmol/L.
*Sentinelhypoxicevent(e.g.cordprolapse,abruptioplacentae)immediatelybeforeorduringlabor**Electrographicand/orelectroclinicalseizureswereidentifiedbyclinicalteamonaEEGorcEEG.ClinicalseizuresonlyhadnoseizurescapturedonaEEGorcEEGmonitoring
82
3.2.2 Prevalence of glucose derangements during aEEG monitoring
A mean of 55.6 hours (SD 20.4) of concurrent CGM and aEEG monitoring were
available per neonate. Thirty-four episodes of glucose derangements were captured on CGM
concurrent with aEEG monitoring; 16 episodes of hypoglycemia in 9 neonates (20%), and 18
episodes of hyperglycemia in 13 (29%). Hypoglycemia episodes had a median duration of 77.5
minutes (IQR 41.25 - 143.75) lasting up to 220 minutes. Hyperglycemia episodes had a median
duration of 237.5 minutes (IQR 57.5 - 766.25) lasting up to 52.1 hours. During hypoglycemia
episodes, the median interstitial glucose was 2.7 mmol/L (IQR 2.6 - 2.8, minimum 2.2 mmol/L),
and during hyperglycemia episodes the median glucose was 14.4 mmol/L (IQR 10.2 - 19.3,
maximum 22.8 mmol/L).
3.2.3 Amplitude-integrated EEG findings
3.2.3.1 aEEG measures during epochs of glucose derangements
Six-hour epochs of aEEG containing episodes of hypoglycemia did not differ from
normoglycemic aEEG epochs for any of the measures assessed (background score, sleep cycling
score or seizure score). Compared to normoglycemic epochs, aEEG epochs containing
hyperglycemia displayed worse aEEG background scores, less sleep-wake cycling and more
frequent seizures, which remained significant after adjusting for clinical markers of HIE (Table
3.2 and Figure 3.1). No seizures were recorded during epochs with hypoglycemia.
83
Table 3.2 Association of glucose derangements with aEEG scores during 6-hour epochs containing hypoglycemia or hyperglycemia
Glucosederangementsduringepoch
Median(IQR)
UnadjustedRegressionSlope(95%CI)
pvalue AdjustedRegressionSlope(95%CI)a
pvalue
Worstbackgroundscoreduringepochb
Hypoglycemia 2(1–3) 0.032(-0.677–0.741) 0.929 0.211(-0.435–0.857) 0.522None 1(1–3) Hyperglycemia 3(2–4) 0.971(0.324–1.618) 0.003 1.120(0.501–1.738) <0.001 Sleepcyclingscoreduringepochb
Hypoglycemia 1(1–1.75) -0.024(-0.252–0.204) 0.834 -0.020(-0.226–0.186) 0.849None 1(1–2) Hyperglycemia 2(2–2) 0.586(0.445–0.728) <0.001 0.587(0.417–0.757) <0.001 Seizurescoreduringepochb
Hypoglycemia 0(0–0) -0.095(-0.176–-0.013) 0.022 -0.065(-0.150–0.021) 0.139None 0(0–0) Hyperglycemia 0(0–1) 0.394(0.170–0.617) 0.001 0.433(0.185–0.681) 0.001
aAdjustedforApgarscore,cordpHandcordbaseexcessbReferencecategoryisnormoglycemiaSignificantpvaluesareshowninboldHypoglycemiaisdefinedasatleasttwointerstitialglucosemeasurementslessthanorequalto2.8mmol/Landhyperglycemiaisdefinedasatleasttwointerstitialmeasurementsgreaterthan8mmol/L.
84
a)
b)
c)
aScoringsystemforbackground:Continuousnormalvoltage(CNV);Discontinuousnormalvoltage(DNV);Burst-suppression(BS);Continuouslowvoltage(CLV);Flattracing(FT)
bScoringsystemforcycling:Developedsleep-wakecycling;Immaturesleep-wakecycling;Nosleep-wakecycling
cScoringsystemforseizures:Noseizures;Singleseizure;Repeatedseizures;Statusepilepticus
Figure 3.1 Bar graphs comparing aEEG scores with glucose derangement category during epochs of hypoglycemia,
normoglycemia or hyperglycemia. a) worst background scores in epocha; b) sleep-wake cycling scores in epochb; c)
seizure scores in epochc
p<0.001 p<0.001
p=0.001
85
3.2.3.2 aEEG measures and severity of glucose derangements
Other glucose measures were analyzed to quantify the severity of glucose derangements during
each 6-hour epoch (Table 3.3). Longer duration of hypoglycemia and greater area under the
hypoglycemic curve were associated with worse aEEG background score, but no other measures
were significant. Longer duration of hyperglycemia, higher maximum glucose values and higher
mean glucose values were associated with worse aEEG background, less sleep-wake cycling and
more frequent seizures. Greater area under the hyperglycemic curve was associated with worse
aEEG background and less sleep-wake cycling.
Table3.3AssociationofseverityofglucosederangementswithaEEGscores
Glucosemeasuresduringepoch
WorstBGscoreinepocha Cyclingscoreinepocha SeizurescoreinepochaAdjusted
RegressionSlope(95%CI)
pvalue AdjustedRegressionSlope
(95%CI)
pvalue AdjustedRegressionSlope
(95%CI)
pvalue
HYPOGLYCEMIAbTimehypoglycemic(hours)
0.299(0.094–0.503)
0.004 0.065(-0.054–0.184)
0.287 -0.023(-0.059–0.013)
0.207
Minimumglucose(mmol/L)
0.101(-0.131–0.333)
0.392 0.086(-0.009–0.181)
0.076 0.001(-0.034–0.035)
0.963
Meanglucose(mmol/L)
0.120(-0.128–0.369)
0.343 0.098(-0.006–0.201)
0.064 0.016(-0.023–0.056)
0.418
Areaundercurve(hour*mmol/L)
1.235(0.123–2.246)
0.029 0.165(-0.367–0.700)
0.544 -0.147(-0.351–0.056)
0.156
HYPERGLYCEMIAbTimehyperglycemic(hours)
0.240(0.102–0.380)
0.001 0.128(0.091–0.165)
<0.001 0.072(0.016–0.128)
0.012
Maximumglucose(mmol/L)
0.090(0.022–0.157)
0.009 0.054(0.031–0.077)
<0.001 0.023(0.010–0.036)
0.001
Meanglucose(mmol/L)
0.094(0.027–0.161)
0.006 0.055(0.031–0.080)
<0.001 0.018(0.003–0.032)
0.016
Areaundercurve(hour*mmol/L)
0.016(0.002–0.031)
0.028 0.011(0.006–0.015)
<0.001 0.002(-0.001–0.005)
0.270
aAdjustedforApgarscore,cordpHandcordbaseexcessbReferencecategoryisnormoglycemiaSignificantpvaluesareshowninboldHypoglycemiaisdefinedasatleasttwointerstitialglucosemeasurementslessthanorequalto2.8mmol/Landhyperglycemiaisdefinedasatleasttwointerstitialmeasurementsgreaterthan8mmol/L.BG:background
86
3.2.3.3 aEEG measures and glucose variability
The relationship of glucose variability with aEEG scores was also explored (Table 3.4). After
adjusting for clinical markers of HIE, a greater standard deviation in interstitial glucose
concentration was associated with worse sleep-wake cycling and more frequent seizures. Faster
rates of increase in glucose concentration were associated with worse sleep-wake cycling, but
faster rates of decrease were not.
Table3.4AssociationofglucosevariabilitywithaEEGscores
Measureofglucosevariability
UnadjustedRegressionSlope(95%CI)
pvalue AdjustedRegressionSlope(95%CI)a
pvaluea
Worstbackgroundscoreinepoch
Standarddeviation 0.470(-0.054–0.995) 0.079 0.516(-0.055–1.087) 0.077Rateofglucosedecrease(mmol/L/hour) -0.407(-1.411–0.596) 0.426 -0.309(-1.296–0.679) 0.540
Rateofglucoseincrease(mmol/L/hour) 0.209(-0.644–1.063) 0.631 0.375(-0.613–1.364) 0.457
Sleepcyclingscoreinepoch
Standarddeviation 0.348(0.180–0.516) <0.001 0.370(0.181–0.560) <0.001
Rateofglucosedecrease(mmol/L/hour) -0.480(-0.986–0.025) 0.063 -0.460(-0.959–0.038) 0.070
Rateofglucoseincrease(mmol/L/hour) 0.379(0.046–0.712) 0.026 0.359(0.012–0.705) 0.042
Seizurescoreinepoch
Standarddeviation 0.399(0.204–0.594) <0.001 0.413(0.188–0.639) <0.001Rateofglucosedecrease(mmol/L/hour) -0.165(-0.444–0.114) 0.247 -0.165(-0.445–0.115) 0.249
Rateofglucoseincrease(mmol/L/hour) 0.564(-0.005–1.133) 0.052 0.465(-0.126–1.057) 0.123
aAdjustedforApgarscore,cordpHandcordbaseexcessNote:RatesofglucoseincreasehavepositiveslopesforrateofchangeandratesofglucosedecreasehavenegativeslopesforratesofchangeSignificantpvaluesareshowninbold
87
3.3 Continuous EEG
3.3.1 Demographics and clinical variables
The first 30 neonates from the cohort of 51 neonates described above were included for cEEG
analysis. There were 17 males and 13 females. Mean gestational age was 39.6 weeks (SD 1.25).
The mean birth weight was 3351.1 grams (SD 623.63), birth length 50.14 (SD 3.22) and head
circumference 34.0 cm (SD 1.40). Mean Apgar scores at 5 minutes was 4.3 (SD 2.3), arterial
cord pH 6.97 (SD 0.18), and arterial cord base excess -15.46 (SD 7.53). The demographics of the
subgroup are representative of the cohort of neonates included in the aEEG analysis described in
detail above (p>0.05 for all demographic measures described).
All neonates received therapeutic hypothermia as per protocol and 2 patients died during
the neonatal period. Twenty-one neonates had P3 and P4 electrodes added to the standard
international 10–20 system of electrode placement modified for neonates. Clinical seizures only
(not confirmed on aEEG or cEEG monitoring) or electrographic and/or electroclinical seizures
identified by the clinical team on aEEG or cEEG were noted in 17 neonates. Three neonates
received lorazepam only, 7 neonates received phenobarbital and 6 neonates received
phenobarbital and further antiepileptic medications (levetiracetam, phenytoin or midazolam
infusion).
3.3.2 Glucose derangements
There was a mean of 47.6 hours (SD 12.4) of concurrent CGM and cEEG monitoring
available per neonate. Twenty-three episodes of glucose derangements were captured on CGM
concurrent with cEEG monitoring; 9 episodes of hypoglycemia in 5 neonates (17%), and 14
episodes of hyperglycemia in 8 neonates (27%). Hypoglycemia episodes had a median duration
of 45 minutes (IQR 20-120) lasting up to 220 minutes. Hyperglycemia episodes had a median
duration of 212.5 minutes (IQR 57.5-540) lasting up to 20.9 hours. During hypoglycemia
episodes, the median interstitial glucose was 2.7 mmol/L (IQR 2.6-2.8, minimum 2.2 mmol/L),
and during hyperglycemia episodes the median glucose was 11.2 mmol/L (IQR 9-14.8,
maximum 18.1 mmol/L).
88
3.3.3 Continuous EEG findings
Eight neonates had electrographic seizures recorded on continuous EEG monitoring. In these
neonates, there was a median of 27.5 (IQR 6.0 -137.5) seizures per recording, ranging between 2
to 165 seizures. Seven neonates had only subclinical seizures on cEEG (87.5%) and one neonate
had both clinical and subclinical seizures. Median seizure duration was 58 seconds (IQR 35-
79.5), and maximum seizure duration recorded was 1043 seconds. Maximal extent of each
seizures (all lobes involved) was recorded. Of the 469 electrographic seizures, 25% involved the
frontal lobe, 53% involved the central and vertex regions, 54.6% involved the temporal lobes,
and 16% the occipital lobes.
3.3.3.1 cEEG measures during epochs of glucose derangements
Five-minute epochs of cEEG containing episodes of hypoglycemia did not differ from
normoglycemic epochs for background scores or seizures. No seizures were recorded during
epochs with hypoglycemia. Compared to normoglycemia, cEEG epochs containing
hyperglycemia displayed worse cEEG background scores, which remained significant after
adjusting for clinical markers of HIE (Table 3.5 and Figure 3.2). No association was found
between epochs containing hyperglycemia and seizures.
89
Table3.5AssociationofglucosederangementswithcEEGscoresduring5-minuteepochscontaininghypoglycemiaorhyperglycemia
Glucosederangementduringepoch
Median(IQR)
UnadjustedRegressionSlope
(95%CI)
pvalue
AdjustedRegressionSlope(95%CI)a
pvalue
Backgroundscoreb
Hypoglycemia 0(0–2) -0.085(-0.685–0.516) 0.782 0.504(-0.335–1.342) 0.239None 0(0–2) Hyperglycemia 3(2–4) 1.842(0.921–2.763) <0.001 1.621(0.663–2.579) 0.001 Presenceofseizuresb
Hypoglycemia 0(0–0) -0.019(-0.041–0.002) 0.077 0.026(-0.018–0.070) 0.245None 0(0–0) Hyperglycemia 0(0–0) -0.002(-0.037–0.033) 0.908 -0.020(-0.082–0.042) 0.529
aAdjustedforApgarscore,cordpHandcordbaseexcessbReferencecategoryisnormoglycemiaSignificantpvaluesareshowninboldHypoglycemiaisdefinedasatleastoneinterstitialglucosemeasurementlessthanorequalto2.8mmol/Landhyperglycemiaisdefinedasatleastoneinterstitialmeasurementgreaterthan8mmol/L.
Figure 3.2 Bar graph comparing cEEG background scores with glucose derangement category during epochs of
hypoglycemia, normoglycemia or hyperglycemia.
p=0.001
90
Examining reactivity, variability and state changes in 6-hour epochs, no association was
found with the presence of hypoglycemia. However, hyperglycemia was associated with less
cEEG variability and state changes (0.824; 95% CI 0.530-1.119; p<0.001), including after
adjusting for clinical markers of hypoxia-ischemia severity (0.669; 95% CI 0.311-1.027;
p<0.001).
3.3.3.2 cEEG measures and level of glucose
The association of the mean glucose level in each 5-minute epoch with background scores was
also assessed (Table 3.6). Excluding epochs of hypoglycemia, higher glucose levels were also
associated with worse cEEG background scores. Excluding epochs of hyperglycemia, lower
glucose levels were surprisingly associated with better cEEG background scores.
Table3.6RelationshipofcEEGscoreswithmeanglucoseduring5-minuteepochs
Glucoseduringepoch
BGscoreinepoch Seizurescoreinepoch BGscoreinepocha SeizurescoreinepochaUnadjustedRegression
Slope(95%CI)
pvalue
UnadjustedRegression
Slope(95%CI)
pvalue
AdjustedRegression
Slope(95%CI)
pvalue
AdjustedRegression
Slope(95%CI)
pvalue
HYPOGLYCEMIAbMeanglucose(mmol/L)
0.263(0.047–0.479)
0.017 0.012(-0.003–0.026)
0.108 0.232(0.042–0.421)
0.017 0.006(-0.001–0.014)
0.105
HYPERGLYCEMIAbMeanglucose(mmol/L)
0.260(0.176–0.344)
<0.001 0.003(-0.001–0.008)
0.124 0.238(0.169–0.307)
<0.001 0.001(-0.004–0.006)
0.701
aAdjustedforApgarscore,cordpHandcordbaseexcessbReferencecategoryisnormoglycemiaSignificantpvaluesareshowninboldHypoglycemiaisdefinedasatleastoneinterstitialglucosemeasurementlessthanorequalto2.8mmol/Landhyperglycemiaisdefinedasatleastoneinterstitialmeasurementgreaterthan8mmol/L.BG:background
91
3.3.3.3 cEEG measures in subgroup of patients with hyperglycemia episodes during continuous glucose monitoring
A subgroup analysis was performed examining the association of hyperglycemia with cEEG
background scores including only the 8 patients that had episodes of hyperglycemia recorded on
CGM monitoring. Epochs of hyperglycemia were still associated with worse background scores
than epochs of normoglycemia in this subgroup. Higher mean glucose concentrations were also
associated with worse cEEG background scores (Table 3.7, Figure 3.3).
Table3.7AssociationofhyperglycemiaandglucoselevelswithcEEGbackgroundscoresinthesubgroupof8neonatesthathadhyperglycemiarecordedonCGM
Glucosederangementduringepoch
Median(IQR)
UnadjustedRegressionSlope(95%CI)
pvalue AdjustedRegressionSlope(95%CI)a
pvalue
Backgroundscoreb
None 2(0–3) Hyperglycemia 3(2–4) 0.936(0.005–1.866) 0.049 0.739(0.038–1.439) 0.039 Backgroundscore
Meanglucose(mmol/L) NA 0.141(0.008–0.274) 0.038 0.117(0.038–0.195) 0.004
aAdjustedforApgarscore,cordpHandcordbaseexcessbReferencecategoryisnormoglycemiaSignificantpvaluesareshowninboldHyperglycemiaisdefinedasatleastoneinterstitialmeasurementgreaterthan8mmol/LNA;notapplicable
92
Figure 3.3 Box plot comparing cEEG background scores with glucose level (mmol/L) in subgroup of 8 patients that
had hyperglycemia recorded on CGM.
Ends of boxes represent IQR. Middle line represents median. Error bars represent 95% confidence intervals.
93
3.3.3.4 cEEG measures in subgroup of patients with hypoglycemia episodes during continuous glucose monitoring
Subgroup analysis was also performed examining the association of hypoglycemia with cEEG
background scores including only the 5 patients that had episodes of hypoglycemia recorded
during CGM monitoring (Table 3.8, Figure 3.4). Epochs of hypoglycemia were not associated
with the background scores on univariate analysis. However, after adjusting for severity of HIE,
epochs with hypoglycemia had better background scores. Lower mean glucose concentrations
were also associated with better cEEG background scores.
Table3.8:AssociationofhypoglycemiaandglucoselevelswithcEEGbackgroundscoresinsubgroupofthe5neonatesthathadhypoglycemiarecordedonCGM
Glucosederangementduringepoch
Median(IQR)
UnadjustedRegressionSlope(95%CI)
pvalue AdjustedRegressionSlope(95%CI)a
pvalue
Backgroundscoreb
None 0(0–2) Hypoglycemia 0(0–2) -0.272(-1.236–0.692) 0.581 -0.602(-0.814–-0.391) <0.001 Backgroundscore
Meanglucose(mmol/L) NA 0.454(0.015–0.892) 0.043 0.159(0.063–0.255) 0.001
aAdjustedforApgarscore,cordpHandcordbaseexcessbReferencecategoryisnormoglycemiaSignificantpvaluesareshowninboldHyperglycemiaisdefinedasatleastoneinterstitialmeasurementgreaterthan8mmol/LNA;notapplicable
94
Figure 3.4 Box plot comparing cEEG background scores with glucose level (mmol/L) in subgroup of 5 patients that
had hypoglycemia recorded on CGM.
Ends of boxes represent IQR. Middle line represents median. Error bars represent 95% confidence intervals.
95
Chapter 4
General discussion 4
4.1 Discussion
Among neonates with encephalopathy, epochs of hyperglycemia (>8 mmol/L), but not
hypoglycemia (2.2 to 2.8 mmol/L), were temporally associated with worse global brain function
and greater seizure frequency on aEEG monitoring, even after adjusting for clinical markers of
HI severity. Greater variability in glucose concentration was also temporally correlated with
worse sleep-wake cycling and more frequent seizures. Epochs of hyperglycemia were also
temporally associated with worse global brain function on cEEG, even after adjusting for clinical
markers of HI severity. Hyperglycemia was also associated with less variability and state
changes on cEEG as well. These observations provide new insights into the dynamics of glucose
homeostasis and brain function, and highlight the potential importance of hyperglycemia on
impaired brain function.
In this cohort of neonates with encephalopathy, epochs with hyperglycemia were
significantly associated with worse EEG background scores. These findings are consistent with
previous studies in other age cohorts. In extremely preterm neonates, moderate hyperglycemia
was significantly associated with greater discontinuity on aEEG (Wikstrom et al., 2011). Another
prospective study in premature neonates found high blood glucose was associated with depressed
cerebral activity on aEEG (Granot et al., 2012). Schumacher et al. with continuous EEG
monitoring showed increased blood glucose was associated with decreased total absolute band
power during the first 3 days of life in premature infants (Schumacher et al., 2014). A study
using CGM and simultaneous cEEG monitoring in children with type 1 diabetes demonstrated
that asymptomatic hyperglycemia was associated with changes in the power spectra of various
EEG frequency bands (Rachmiel et al., 2016).
EEG background abnormalities and their recovery are predictive of long-term
neurodevelopmental outcome in neonates with HIE (Azzopardi & Toby study group, 2014;
Chandrasekaran et al., 2017; Dunne et al., 2017; Skranes et al., 2017; Spitzmiller et al., 2007;
96
Thoresen et al., 2010; van Laerhoven et al., 2013). A meta-analysis of 8 studies before the advent
of therapeutic hypothermia showed a pooled sensitivity of 91% and specificity of 88% of early
severe aEEG tracings to predict poor neurodevelopmental outcome (Azzopardi & Toby study
group, 2014; Spitzmiller et al., 2007). With therapeutic hypothermia, a persistently abnormal
aEEG background at 48 hours remains an important prognostic indicator (Chandrasekaran et al.,
2017; Thoresen et al., 2010). And, although therapeutic hypothermia has changed the time-
frame, aEEG and EEG are still the best objective predictive tools to identify the highest-risk
patients for poor outcome (Bonifacio et al., 2015). Several other EEG features have been
associated with unfavorable outcome including, background amplitude of less than 30 mcV, IBI
of greater than 60 seconds, and electrographic seizures (Kontio et al., 2013; Menache et al.,
2002; Murray et al., 2009; Obeid et al., 2017; Pressler et al., 2001; Shah et al., 2014b).
Discontinuous patterns with IBI lasting 10-60 seconds may lead to either favorable or
unfavourable outcomes after HIE (Murray et al., 2009; Nevalainen et al., 2017). Quantitative
automated analysis of EEG can improve the objectivity of EEG assessment in newborns and may
improve prognostic capabilities. Quantitative EEG measures that have been associated with
outcomes included EEG discontinuity (Dunne et al., 2017), total suppression time (Flisberg et
al., 2011), and novel features of bursts such as measures of burst area and duration and their
interrelationships (Iyer et al., 2014).
In our cohort, hyperglycemia and greater glucose variability (as measured by SD and rate
of glucose increase), were also associated with worse sleep-wake cycling scores. Neonatal sleep-
wake cycling is a marker of brain function and enhances prediction of long-term
neurodevelopmental outcomes in neonates with HIE (Osredkar et al., 2005; Thoresen et al.,
2010). A recent multimodality neuromonitoring study in neonates at risk for cerebral dysfunction
showed that inefficient neonatal sleep patterns were independent predictors of 18-month
neurodevelopmental outcome (Shellhaas et al., 2017).
We demonstrate that hyperglycemia and greater glucose dispersion (as measured by SD)
are associated with seizure burden on aEEG, although not necessarily time-locked to the 5-
minute epochs with hyperglycemia on cEEG. There is growing evidence that higher seizure
burden is associated with worse short and long-term outcomes, adjusting for confounders
including seizure etiology and illness severity (Glass et al., 2016; Kharoshankaya et al., 2016;
McBride et al., 2000; Pisani et al., 2007; Pisani et al., 2008; Srinivasakumar et al., 2015). Studies
97
in neonates undergoing therapeutic hypothermia for HIE show a relationship between higher
seizure burden and worse MRI injury scores in univariate (Glass et al., 2011; Nash et al., 2011;
Srinivasakumar et al., 2015; van Rooij, Toet, et al., 2010) and multivariable analyses (Dunne et
al., 2017; Shah et al., 2014b).
Electrographic seizures have been observed in 30-65% of neonates with HIE (Bashir et
al., 2016; Glass et al., 2011; Glass et al., 2014; Kharoshankaya et al., 2016; Nash et al., 2011;
Srinivasakumar et al., 2015; Wusthoff et al., 2011) and were identified in 27% of our cohort on
cEEG. Our cohort includes neonates with all severities of encephalopathy (mild to severe)
including neonates with early improvement of aEEG background and no clinical seizures, who
may not be monitored with cEEG during routine clinical practice. While twelve patients had
cEEG requested by the clinical neurology consultation service, 18 cEEGs were research studies
(one of which became a clinical study after electrographic seizures were recorded). In neonates
with seizures, a median of 27.5 (IQR 6.0-137.5) seizures were identified per recording with a
median seizure duration of 58 seconds (IQR 35-79.5). The seizure number and duration in our
cohort is similar to a previously described cohort of 23 neonates with HIE and seizures that
received therapeutic hypothermia. They recorded a median seizure duration of 125 seconds (IQR
82-238) and total median seizure number of 26 (IQR 13-86) per neonate (Lynch et al., 2015).
There is increasing evidence of an association between hyperglycemia and
neurodevelopmental outcomes in term and preterm neonates. Hyperglycemia is common in
preterm neonates and has been demonstrated to be associated with increased mortality, short-
term morbidities (Alexandrou et al., 2010; Bermick et al., 2016; Hays et al., 2006; Kao et al.,
2006) and with long-term outcomes, including white matter reduction on term equivalent MRI
scan (Alexandrou et al., 2010) and neurological and behavior problems at age 2 years (van der
Lugt et al., 2010). In a cohort of well babies at risk for hypoglycemia, higher glucose levels were
associated with neurosensory impairment and cognitive delay at 2 years of age (McKinlay et al.,
2015). And in neonates with HIE, hyperglycemia has also been shown to adversely affect long-
term outcome in two small retrospective studies (Chouthai et al., 2015; Spies et al., 2014), as
well as a post-hoc analysis of the CoolCap therapeutic hypothermia trial in which both hypo- and
hyperglycemia were common and associated with unfavorable neurodevelopmental outcome by
age 18 months (Basu et al., 2016). Longer duration of hyperglycemia and higher peak glucose
levels have also been associated with mortality in the PICU (Srinivasan et al., 2004).
98
The relationship between glucose levels and brain injury is likely complex. A study in
well babies had demonstrated that less glucose stability and steeper increase in glucose
concentrations following hypoglycemia in the first 12 hours after birth were associated with
neurosensory impairments at age 2 years (McKinlay et al., 2015), and development of
neurosensory impairment between ages 2 and 4.5 years (McKinlay, Alsweiler, et al., 2017).
Another retrospective cohort of term HIE found that glucose variability was associated with
severe neurodevelopmental disability (Al Shafouri et al., 2015). An association between glucose
variability and mortality was likewise shown in very low birth weight preterm neonates (Fendler
et al., 2012), term neonates and children admitted to the pediatric ICU (Hirshberg et al., 2008;
Rake et al., 2010; Wintergerst et al., 2006) and adult ICU (Ali et al., 2008; Egi et al., 2006;
Hermanides et al., 2010; Pidcoke et al., 2009). Long-term outcomes are being assessed in our
cohort to identify whether the acute changes seen on aEEG and cEEG associated with
hyperglycemia and glucose variability are predictive of long-term neurodevelopmental
outcomes.
Perhaps surprisingly, in our primary analysis we did not observe any associations
between hypoglycemia (2.2-2.8 mmol/L) and brain function. This may be because episodes of
hypoglycemia were generally shorter in duration (median 77.5 minutes) than hyperglycemic
episodes (median 237.5 minutes). Still, the majority of hypoglycemic episodes on CGM were not
clinically identified and treated. In our secondary analysis, a longer duration of hypoglycemia
and greater area under the hypoglycemic curve were associated with worse background scores,
suggesting that the duration and severity of hypoglycemia may indeed be important.
Neonatal hypoglycemia has been associated with more severe neonatal encephalopathy
(Basu et al., 2009) and studies looking at long-term outcomes have shown hypoglycemia in
neonates with HIE to be associated with worse neurodevelopmental outcomes (Basu et al., 2016;
Salhab et al., 2004; Tam et al., 2012). However, those studies in which aEEG has been used in
newborns to examine acute neurophysiological changes have been unable to detect changes
during hypoglycemia (Harris et al., 2011; Stenninger et al., 2001) except for a single case report
of one critically ill neonate (Hellstrom-Westas et al., 1989). In terms of conventional EEG, there
are reports of children with recurrent hypoglycemia, who showed increased slow activity on
EEG with glucose concentrations ≤2.3 mmol/L which improved with normalization of glucose
concentrations (Haworth et al., 1960; Pavone et al., 1980). In diabetic children, an increase in
99
low-frequency EEG activity was observed when glucose concentrations decreased ≤4.0 mmol/L
(Bjorgaas et al., 1998). Canine models of insulin-induced newborn hypoglycemia demonstrated
progressive EEG slowing with declining glucose concentration (Vannucci et al., 1981) and
suggest that the EEG response to hypoglycemia may depend not only on the concentration of
glucose in the brain but also on the presence of associated metabolic derangements (Vannucci et
al., 1980). In HIE, a neonate’s ability to produce and use alternative fuels may not be sufficient
to compensate for low glucose levels, making them particularly vulnerable to hypoglycemia
induced brain injury (Harris et al., 2015; Harris et al., 2011; Stanley et al., 2015). Nonetheless,
looking in our cohort, we still did not identify EEG changes with hypoglycemia. Our data in this
regard may suggest that either our current management protocols for hypoglycemia are adequate
to maintain brain function, or the impact of hypoglycemia on brain function is not immediately
apparent by visual EEG analysis. Furthermore, it was noted in the subgroup of five neonates with
hypoglycemia captured on CGM, that lower mean glucose concentrations were associated with
better cEEG background scores. Potential explanations for this finding include the possibility
that qualitative EEG grading may not be an adequate method to capture the changes seen on
EEG with hypoglycemia. Previous studies suggest progressive EEG slowing occurs as glucose
levels decrease, with progression to a burst suppression pattern only after a severe decrease in
glucose levels, for example to ≤0.3 mmol/L in newborn dogs (Bjorgaas et al., 1998; Vannucci et
al., 1981). In our cohort, interstitial glucose concentrations during hypoglycemia episodes
recorded on CGM were between 2.2 to 2.8 mmol/L, although several early severe hypoglycemia
episodes (<2.2 mmol/L) were recorded by standard intermittent glucose testing in the first 6
hours of life prior to commencement of CGM, aEEG and cEEG monitoring. Accordingly, the
potential neurophysiological changes that occur at glucose concentrations below 2.2 mmol/L
could not be assessed in this study. Additionally, quantitative frequency and spectral power
analysis may be a more sensitive method to detect changes during hypoglycemia. Furthermore,
differences in distribution of glucose disturbances over time may also contribute. Whereas
hyperglycemia events were more widely distributed over the course of CGM monitoring,
hypoglycemia events occurred later during CGM monitoring. Although several neonates had
early hypoglycemia events prior to CGM insertion, the hypoglycemia events recorded on CGM
started at a mean of 48.2 hours of life (SD 13.8, ranging from 32.5 to 76 hours of life) and
occurred later than hyperglycemia events which started at a mean of 29.5 hours of life (SD 19.4,
ranging from 7.7 to 74 hours, p=0.016). EEG background can improve over the course of
100
therapeutic hypothermia although it also may remain persistently severe or worsen over the
course of monitoring (Nash et al., 2011; Pressler et al., 2001; Sewell et al., 2018). Future
analyses will account for evolution over time.
Our findings are consistent with a growing body of evidence that it is important to
consider the potential harm of allowing permissive hyperglycemia when trying to prevent
hypoglycemia (Lear et al., 2017; McKinlay et al., 2015; McKinlay, Alsweiler, et al., 2017).
Several mechanisms have been proposed to explain the association of hyperglycemia with poor
outcomes during critical illness, including: dyslipidemia, inflammatory cytokine production,
endothelial dysfunction, hypercoagulation (Van den Berghe, 2004), and accelerated glucose
toxicity leading to metabolic disturbances and increased cellular apoptosis (Quagliaro et al.,
2003; Van den Berghe, 2004). The overproduction of superoxide by the mitochondrial electron-
transport chain may underlie the mechanisms implicated in glucose-mediated vascular damage
(Brownlee, 2001). It has also been postulated that increased glucose variability would generate
more reactive oxygen species due to hyperglycemia-induced oxidative stress (Hirsch et al.,
2005). Greater glucose variability has been associated with elevated markers of oxidative stress
in adolescents (Dasari et al., 2016). Preclinical studies using cell culture and rodent models have
shown that neuronal death was triggered by glucose reperfusion rather that hypoglycemia itself,
and that hyperglycemia following hypoglycemia can worsen neuronal injury (Ennis et al., 2015;
Suh et al., 2007). These findings suggest that the rate of correction of hypoglycemia as well as
prevention of rebound hyperglycemia may be important.
4.2 Limitations
Our study has several limitations. Because our study is observational, it does not establish a
causal relationship between episodes of hyperglycemia and abnormal brain function.
Hyperglycemia may be considered as part of the physiological response to stress (Davidson et
al., 2008; Dungan et al., 2009) and whether hyperglycemia is associated with worse brain
function or rather with more severe HIE and therefore worse EEG scores cannot be concluded. A
subgroup analysis demonstrated that epochs with hyperglycemia were still associated with worse
101
cEEG background scores in the subgroup of neonates that had all had episodes of hyperglycemia
on CGM, strengthening our findings.
Further, accounting for clinical severity of HIE is challenging in this cohort because the
clinical signs of hypoglycemia and HIE are overlapping. Signs of encephalopathy may be due to
either hypoxia-ischemia or hypoglycemia. Even attempting to separate these factors on MRI can
be complex, as hypoglycemia has been shown to be associated with worsened HIE-related brain
injury in the corticospinal tracts (Tam et al., 2012). Also, initial clinical assessment of mild
neonatal encephalopathy can still frequently be associated with MRI abnormalities (Walsh et al.,
2017). We thus chose to adjust for Apgar scores, umbilical artery pH and base deficit.
In addition, other potential unmeasured confounders may exist such as carbon dioxide
and pH levels which have been associated with aEEG changes in premature infants (Granot et
al., 2012; Lingappan et al., 2016; Wikstrom et al., 2011). Also, the aEEG tracings were graded in
6-hour epochs due to low temporal resolution of aEEG because of time compression of the EEG
signal. cEEG analysis allows better temporal resolution although there may still be EEG features
altered during glucose derangements that cannot be captured by a visual scoring system.
Finally, although the continuous interstitial glucose monitors have been found to show
good agreement between interstitial and blood glucose measurements (Bailey et al., 2015; Harris
et al., 2010) and allow us to determine the exact duration of glucose derangements, there are still
limitations of the sensors outside of 2.2 to 22.2 mmol/L. Moreover, studies have reported that
interstitial glucose concentrations can have a variable “lag” behind the blood glucose
concentrations by as long as 20 minutes due to delays in diffusion of glucose from the vascular
into the interstitial space, as well as patient-specific and instrumental factors. There is variability
in this time lag documented between studies. Variability in the lag may also be noted depending
on the phase of change in the glucose, in that there may be increasing positive error as the blood
glucose concentrations decrease and increasing negative error when the glucose concentrations
are rising. Although, studies have also documented good agreement between the intermittent and
continuous glucose measurements during the baseline, decline and hypoglycemic periods yet in
the recovery phase the blood glucose levels can be significantly higher than the interstitial
glucose measurements (Boyne et al., 2003; Caplin et al., 2003; D. L. Harris et al., 2009;
Kovatchev et al., 2009; Rebrin et al., 1999; Schmelzeisen-Redeker et al., 2015; Shah et al.,
102
2018). Consequently, an unknown and variable delay may exist between the intermittent and
continuous glucose measurements and shifting the CGM data could be further explored.
Please see future directions section to see how some of the limitations discussed will be
overcome in future analyses.
4.3 Conclusions Our data give credence to our hypotheses that glucose derangements in neonates with
encephalopathy are associated with abnormal brain background activity and increased
electrographic seizures on aEEG and cEEG. More specifically, hyperglycemia (>8 mmol/L), but
not hypoglycemia (2.2-2.8 mmol/L), was temporally associated with worse global brain function
and seizures on aEEG as well as worse brain function on cEEG. Glucose variability was
associated with seizures and impaired sleep-wake cycling. However, it remains to be determined
whether these associations represent a causal link between hyperglycemia, glucose variability
and brain function and whether these immediate perturbations in brain function are associated
with worse long-term outcomes following HIE in and of themselves. Our results are in keeping
with growing evidence in the literature that hyperglycemia and glucose variability are associated
with worse long-term neurodevelopmental outcomes (Al Shafouri et al., 2015; Basu et al., 2016;
Chouthai et al., 2015; McKinlay et al., 2015; McKinlay, Alsweiler, et al., 2017; Salhab et al.,
2004; Spies et al., 2014; Tam et al., 2012). Further analyses with longitudinal modeling to
account for time series data and explore temporal correlations of glucose derangements with
EEG changes are ongoing. Future studies are needed to clarify optimal treatment approaches for
hypo- and hyperglycemia, in particular the optimal rate of glucose correction while maintaining
brain function, in order to prevent subsequent brain injury in this high-risk population.
103
Chapter 5
Future directions 5
5.1 Continuous EEG monitoring - Qualitative analysis
The Neurological Outcome of Glucose in Neonatal encephalopathy (NOGIN) study is an
ongoing prospective cohort study using continuous glucose monitoring, early MRI scans and
long-term follow-up to resolve major knowledge gaps in our understanding of the neurological
consequences of neonatal glucose derangements. Continuous glucose monitors provide average
glucose values every 5 minutes and EEG monitoring provides a method that can detect acute and
ongoing changes that may occur with either hypo- or hyperglycemia or glucose instability. EEG
provides an approach to monitoring the developing brain with excellent temporal resolution, in
the millisecond range, through direct access to brain electrical activity. It can assess dynamic
oscillatory activities and brief transient cortical events. Furthermore, the scalp and skull
impedance in neonates is relatively low, thus decreasing smearing of EEG signals from volume
conduction and enhancing the spatial accuracy of brain activity recorded from scalp electrodes
(Dan et al., 2015). Analysis of 30 cEEGs in our cohort demonstrated the feasibility of detecting
acute changes using this method. Over fifty more neonates have been recruited and recruitment
is ongoing. Further analysis of these cEEGs using the qualitative measures described in this
study for scoring background, reactivity and state changes, and seizures will be performed, as
well as the addition of quantitative EEG analysis as discussed below.
104
5.2 Continuous EEG monitoring - Quantitative EEG analyses
5.2.1 Previous studies validating use of quantitative EEG analyses
Quantitative automated analysis of EEG features may improve the objectivity of EEG
assessment in newborns. Several studies have compared quantitative EEG features with
neuroimaging findings and neurodevelopmental outcomes in neonates with HIE.
Several features of burst suppression have been assessed. A study calculated EEG
discontinuity with a novel algorithm and showed that higher mean discontinuity at 24h and 48 h
were associated with severe cerebral tissue injury on MRI and unfavorable neurodevelopmental
outcome at 24 months old (Dunne et al., 2017). Quantitative analysis of the total suppression
length has also been associated with death or neurodevelopmental disability. The authors argue
that short artefacts can be mistakenly classified as bursts with an automated burst suppression
classifier and can significantly influence the mean IBI but would only slightly reduce the total
suppression length (Flisberg et al., 2011). Iyer et al. examined the distributions of burst area and
duration and their interrelationships, demonstrating their ability to predict neuroimaging and
neurodevelopmental outcomes in a cohort of 20 neonates with HIE. They had extracted 3 novel
burst metrics for each EEG: mean burst duration and its coefficient of variation, scaling exponent
of the cumulative distribution function of the burst areas in each recording, and slope value based
on the relationship between burst duration and burst area (Iyer et al., 2014).
Other quantitative measures have also been assessed in cohorts of neonates with HIE.
The spectral frequency content of bursts occurring during burst suppression can be differentiated
from those during a normal tracé alternant pattern. There is a lower amount of low-frequency
activity in periods of burst suppression, with the relative amount of low- and high-frequency
activity being the parameter that best discriminated between the groups (Thordstein et al., 2010).
A retrospective study showed that total EEG power calculated at a mean age of 8.9 h, was
significantly higher in neonates with no or mild injury on MRI compared to those with
moderate/severe MRI injury (Jain et al., 2017). In a preterm fetal sheep model of HIE, early
hypothermia started within 30 minutes of asphyxia was associated with faster recovery of
spectral edge frequency, reduced seizure burden, and less suppression of EEG power and
amplitude and corresponded with reduced neuronal loss and microglial induction in the striatum
on pathological examination (Wassink et al., 2015). Temko et al. developed a support vector
105
machine classifier for prediction of outcome in neonates with HIE using a multimodal
combination of routine clinical markers, EEG features and heart rate parameters. They evaluated
for neurodevelopmental outcome prediction at 24 months in newborn infants with HIE and
identified 12 multimodal features that provided promising prediction results, 9 of which were
EEG features and 3 were clinical features. EEG achieved the highest performance as a predictor
of clinical outcome (Temko et al., 2015).
Quantitative EEG measures have also been used to assess whether EEG changes occur
with changes in core body temperature during mild therapeutic hypothermia for HIE. A study
using both visual and quantitative EEG measures during rewarming in 15 neonates with HIE
showed that rewarming was associated with a worsening of visual EEG scores which
corresponded to an increase in EEG discontinuity, and was more pronounced in newborns with
severe than moderate HIE. They also examined absolute magnitude and relative spectral
magnitudes and noted an increase in the relative magnitude of slower delta and a decrease in
higher frequency theta and alpha waves with rewarming (Birca et al., 2016). However, another
study that assessed several quantitative features of the aEEG in 10 neonates with HIE found that
severity of HIE but not core body temperature affected the quantitative EEG features (Burnsed et
al., 2011).
As with visual analysis, maturational changes are seen in several quantitative features
with increasing gestational age in preterm neonates with normal neurological follow-up,
including maturational changes in spectral measures with a shift from lower to higher
frequencies, decrease IBI length and length of discontinuous activity (periods with high
interburst-burst ratio) and increase in continuous activity (Niemarkt et al., 2010; Niemarkt et al.,
2011) as well maturational changes in spectral power in EEG bursts which are believed to
correspond to ongoing gyration and postnatal white matter maturation (Jennekens et al., 2012).
Minimal changes in the power spectrum in the first 3 days of life are also noted in healthy term
neonates with normal neurocognitive function at 4 years of age. When comparing the first 6
hours of life to 48 to 72 hours of life, there is overall higher relative power spectrum in the delta
frequency band and lower alpha/delta ratio at 6 hours of life (Castro Conde et al., 2017). Thus,
gestational age and temporal evolution since birth should be considered in analysis of
quantitative measures of discontinuity and spectral measures.
106
Of note, quantitative EEG measures such as amplitude, continuity (skewness, kurtosis
and discontinuity) and frequency content are associated with visual interpretation of the EEG
background. A study had 2 expert electroencephalographers visually grade EEGs of 54 neonates
into 4 categories and showed that a combination of quantitative EEG measures could accurately
predict EEG/HIE grade (Korotchikova et al., 2011). Another study developed a cerebral health
index score using multiple components of the EEG (power spectrum and spectral edge
frequency, entropy, volatility and sub-band spectral entropy) and compared it to clinical
measures of severity of HIE (based on Sarnat scores and visually interpreted EEG). The cerebral
health index score was able to distinguish between normal controls and moderate and severe
encephalopathy and normal versus severe EEG groups (Hathi et al., 2010). Other groups have
also used quantitative EEG features to develop automatic grading scales of the degree of
abnormality of the neonatal EEG (Lofhede et al., 2010; Stevenson et al., 2013) as well as
automated classification of sleep cycling (or referred to with the more general term, brain
activity cycling) (Stevenson et al., 2014).
Furthermore, detrended fluctuation analysis (DFA) has been discussed for CGM as a
method for analyzing scaling behavior in time series and can also be used to investigate the
scale-free amplitude modulation of ongoing neuronal oscillations (Hardstone et al., 2012). In a
complex system, scale-free dynamics can give rise to long-range temporal correlations, which
can be assessed from the EEG using DFA. Multifractal DFA (MF-DFA) can characterize time
series with multiple co-existent dynamic processes that may give rise to the temporal local
fluctuations, including of both small and large magnitudes (Matic et al., 2015). In a cohort of
term neonates with HIE the feasibility of DFA and MF-DFA and its correlation to EEG
background grades was assessed. They demonstrated that DFA could distinguish different grades
of abnormality in the background EEG activity in neonates with HIE, and MF-DFA was superior
to conventional DFA in distinguishing between EEG grades. MF-DFA parameters were also
significantly correlated to IBI intervals. They had visually inspected a large number of DFA
plots and noted that the trends were often multiphasic, with very clear “crossover points” (where
there was a change of slope in the fluctuation function), which are suggestive that the EEG signal
might exhibit multifractal behavior. They thus determined that MF-DFA better represents
neonatal EEG dynamics (Matic et al., 2015).
107
Considering glucose derangements, few previous studies have examined the acute EEG
changes associated with hypo- or hyperglycemia, as previously discussed, some of which used
quantitative EEG measures. A study in young children that was able to detect changes during
hypoglycemia performed quantitative spectral analysis of EEG during a gradual decline in
glucose concentrations induced using insulin. At a plasma glucose of 4 mmol/L, an increase in
delta and theta amplitudes were seen in both groups. At 3 mmol/L, there was a further and more
widespread increase in low-frequency EEG activity, with greater slowing and epileptiform
discharges seen in the diabetic than nondiabetic children (Bjorgaas et al., 1998). Several studies
detected changes with hyperglycemia using quantitative EEG measures. In extremely preterm
neonates, moderate hyperglycemia was significantly associated with greater discontinuity on
aEEG, measured using an automated algorithm to measure average IBI from 10-minute artifact-
free epochs of EEG recordings before each blood sample (Wikstrom et al., 2011). Schumacher et
al. measured on continuous EEG recordings the total absolute band power using spectral EEG
analysis and showed increased blood glucose concentrations were associated with decreased total
absolute band power during the first 3 days of life in premature infants (Schumacher et al.,
2014). A study using CGM and simultaneous cEEG monitoring in children with type 1 diabetes
demonstrated that asymptomatic hyperglycemia was associated with changes in the power
spectra of the different frequency bands during wakefulness and sleep (Rachmiel et al., 2016).
Therefore, several quantitative EEG features have been validated in term neonates with
HIE, are predictive of neuroimaging findings and neurodevelopmental outcomes, and correlate
with visual EEG grading scales. Quantitative EEG analysis may provide a more objective
measure than visual analysis, as well as additional detail and novel measures of EEG complexity.
For example, our visual EEG classification would not capture an increase in slow frequencies or
increase in IBI duration which have been previously reported with hypo- and hyperglycemia.
Thus, the addition of quantitative EEG analysis in our future analyses will provide the
opportunity to assess acute changes in brain activity associated with glucose derangements in
neonates with HIE beyond the features captured in our visual background grading.
108
5.2.2 Application of quantitative EEG analyses in the NOGIN study
Several quantitative EEG features will be assessed in our cohort through a collaborative project
with Dr. Sampsa Vanhatalo, a pediatric neurophysiologist at the University of Helsinki. The
signal will be segmented into 5-minute non-overlapping epochs corresponding to the 5-minute
epochs of mean glucose data available from the continuous interstitial glucose monitor. The
quantitative features are calculated on each epoch for each channel and then averaged across
channels. Quantitative EEG features being assessed will include mean amplitude of the epoch
(mcV), mean frequency of the epoch (Hz) (Stevenson et al., 2013), measures of spectral power
from different frequency ranges (delta 1 [0.5-2Hz], delta 2 [2-4 Hz], theta [4-7 Hz], alpha [7-12
Hz], and beta [12-30 Hz]) (O'Toole et al., 2016), activation synchrony index (measure of
interhemispheric synchrony) (Rasanen et al., 2013), IBI (seconds), duration of bursts (sec), DFA
(Kantelhardt et al., 2001) and MF-DFA (Ihlen, 2012).
The relationship of hypo- or hyperglycemia with the qualitative and quantitative EEG
findings will be analyzed, adjusting for clinical markers of HIE severity (Apgar scores, umbilical
artery pH and base deficit) as well as gestational age and antiseizure medications. Longitudinal
modeling will be used to account for time series data and explore temporal correlations of
glucose levels with EEG changes. The visual and multiple quantitative EEG features will be
assessed to investigate which variables fit better with the changes in glucose over time. We will
incorporate cubic splines in the GEE models to characterize the possibly non-linear relationship
between cEEG features and glucose levels, in particular to investigate the possibility of threshold
effects. Phase shifting of the CGM data will be used to explore the delay that may exist between
the intermittent and continuous glucose measurements.
109
5.3 Short and long-term outcomes
Both short and long-term outcomes will be assessed in our cohort (Figure 5.1) to identify
if the neurophysiological changes associated with glucose derangements are associated with later
outcomes. Previous studies have shown a relationship between EEG background abnormalities
with brain injury on MRI in neonatal encephalopathy (Biagioni et al., 2001; Briatore et al., 2013;
Dunne et al., 2017; Shah et al., 2006), and MRI patterns of injury are predictive of long term
outcomes (Biagioni et al., 2001; Dalmazzo et al., 2011). However, EEG findings have not yet
been linked to outcome in our cohort. Short-term outcomes will include a standardized
neurological examination and MRI at 3-5 days (including 3-dimensional T1 and T2-weighted
sequences, diffusion tensor imaging and MR spectroscopy).
Figure 5.1 Timeline for continuous glucose monitoring, MRI study and follow-up
EEG background abnormalities and recovery have also been shown to be predictive of
long-term neurodevelopmental outcome in neonates with HIE (Azzopardi & Toby study group,
2014; Chandrasekaran et al., 2017; Dunne et al., 2017; Skranes et al., 2017; Spitzmiller et al.,
2007; Thoresen et al., 2010; van Laerhoven et al., 2013). Long-term outcomes are being assessed
in our cohort to identify if the acute changes seen on aEEG and cEEG associated with
hyperglycemia and glucose variability are predictive of long-term neurodevelopmental
outcomes. At 18 and 36 months, children undergo full neurological examination and
standardized tests of visual function by a pediatric neurologist, and cognitive and
neurodevelopmental outcomes will be assessed by a pediatric neuropsychologist using several
measures to assess for language, cognitive, behavioral, visual and motor function.
The optimal timing for assessing children to accurately predict development in childhood
and beyond remains debated. Environmental and educational experiences and supports may alter
developmental trajectories but deficits may also emerge that may not have been evident at
110
younger ages. The greater academic and psychosocial demands placed on children as they grow
older can reveal neurodevelopment abnormalities potentially associated with glucose
derangements that occurred in the neonatal period. The CHYLD study performed initial follow-
up at 2 years and hypoglycemia was not associated with neurodevelopmental impairments
(McKinlay et al., 2015). However, at 4.5 years old, the investigators found that neonatal
hypoglycemia was associated with an increased risk of low executive function and low visual
motor function, with the highest risk in those children that had severe, recurrent, or clinically
undetected hypoglycemia (McKinlay, Alsweiler, et al., 2017). Similarly, in children with
congenital heart disease the investigators showed that postoperative electrographic seizures were
not associated with worse outcomes on the BSID at 1 year of age, except in the subset of
children with frontal-onset seizures (Gaynor et al., 2006). Whereas at 4 years, postoperative
electrographic seizures were associated with worse executive function and impaired social
interactions or restricted behaviours (Gaynor et al., 2013). Furthermore, although the BSID is a
reliable instrument and in particular the subdomain scores can predict later performance (Soysal
et al., 2014), in both term and premature populations, measures of cognitive function during
infancy have been shown to have poor correlation with later childhood IQ and may overestimate
neurocognitive impairments (Aylward, 2004; Hack et al., 2005; Illingworth et al., 1959). Thus,
follow-up at 3 years and beyond will be important in determining whether delays persist into
early childhood.
5.4 Future directions - Summary
This study has demonstrated that acute neurophysiological changes can be detected with glucose
derangements in neonates with encephalopathy. Findings in this study support growing evidence
that it is important to consider the potential harm in allowing permissive hyperglycemia while
trying to prevent hypoglycemia. In several of the previous trials with intensive insulin therapy
for hyperglycemia management there has been an increased incidence of hypoglycemia seen
(Agus et al., 2017; Alsweiler et al., 2012; Macrae et al., 2014; Vlasselaers et al., 2009) and
inconsistant effects on glucose variability (Siegelaar et al., 2010). CGM provide an advantage by
providing continuous glucose trending information that can be acted on rapidly and may protect
111
against sudden unexpected decreases in blood glucose. However, there is currently limited
evidence for treating newborns based on this monitoring. Knowledge about the degree and
duration of hypo- and hyperglycemia that correlate with changes in global brain function will aid
in determining clinically significant target glucose ranges in neonates, leading to the
development of evidence-based algorithms for glucose management and effective
neuroprotection. How management algorithms effect the rate of glucose correction and glycemic
variability and subsequent long-term outcomes will be important to consider. The ultimate goal
is to decrease brain injury and improve long-term outcomes of neonatal encephalopathy.
112
References
Abend, N. S. (2015). Electrographic status epilepticus in children with critical illness: Epidemiology and outcome. Epilepsy Behav, 49, 223-227. doi:10.1016/j.yebeh.2015.03.007
Abend, N. S., Arndt, D. H., Carpenter, J. L., Chapman, K. E., Cornett, K. M., Gallentine, W. B., Giza, C. C., Goldstein, J. L., Hahn, C. D., Lerner, J. T., Loddenkemper, T., Matsumoto, J. H., McBain, K., Nash, K. B., Payne, E., Sanchez, S. M., Fernandez, I. S., Shults, J., Williams, K., Yang, A., & Dlugos, D. J. (2013). Electrographic seizures in pediatric ICU patients: cohort study of risk factors and mortality. Neurology, 81(4), 383-391. doi:10.1212/WNL.0b013e31829c5cfe
Abend, N. S., Gutierrez-Colina, A. M., Topjian, A. A., Zhao, H., Guo, R., Donnelly, M., Clancy, R. R., & Dlugos, D. J. (2011). Nonconvulsive seizures are common in critically ill children. Neurology, 76(12), 1071-1077. doi:10.1212/WNL.0b013e318211c19e
Abend, N. S., Topjian, A., Ichord, R., Herman, S. T., Helfaer, M., Donnelly, M., Nadkarni, V., Dlugos, D. J., & Clancy, R. R. (2009). Electroencephalographic monitoring during hypothermia after pediatric cardiac arrest. Neurology, 72(22), 1931-1940. doi:10.1212/WNL.0b013e3181a82687
Abend, N. S., Wagenman, K. L., Blake, T. P., Schultheis, M. T., Radcliffe, J., Berg, R. A., Topjian, A. A., & Dlugos, D. J. (2015). Electrographic status epilepticus and neurobehavioral outcomes in critically ill children. Epilepsy Behav, 49, 238-244. doi:10.1016/j.yebeh.2015.03.013
Acharya, P. T., & Payne, W. W. (1965). Blood Chemistry of Normal Full-Term Infants in the First 48 Hours of Life. Arch Dis Child, 40, 430-435.
Adam, P. A., King, K., & Schwartz, R. (1968). Model for the investigation of intractable hypoglycemia: insulin-glucose interrelationships during steady state infusions. Pediatrics, 41(1), 91-105.
Adamkin, D. H., & Committee on Fetus and Newborn. (2011). Postnatal glucose homeostasis in late-preterm and term infants. Pediatrics, 127(3), 575-579. doi:10.1542/peds.2010-3851
Agardh, C. D., & Rosen, I. (1983). Neurophysiological recovery after hypoglycemic coma in the rat: correlation with cerebral metabolism. J Cereb Blood Flow Metab, 3(1), 78-85. doi:10.1038/jcbfm.1983.10
Agus, M. S., Steil, G. M., Wypij, D., Costello, J. M., Laussen, P. C., Langer, M., Alexander, J. L., Scoppettuolo, L. A., Pigula, F. A., Charpie, J. R., Ohye, R. G., Gaies, M. G., & Investigators, S. S. (2012). Tight glycemic control versus standard care after pediatric cardiac surgery. New England Journal of Medicine, 367(13), 1208-1219. doi:10.1056/NEJMoa1206044
113
Agus, M. S., Wypij, D., Hirshberg, E. L., Srinivasan, V., Faustino, E. V., Luckett, P. M., Alexander, J. L., Asaro, L. A., Curley, M. A., Steil, G. M., Nadkarni, V. M., Investigators, H.-P. S., & the, P. N. (2017). Tight Glycemic Control in Critically Ill Children. New England Journal of Medicine, 376(8), 729-741. doi:10.1056/NEJMoa1612348
al Naqeeb, N., Edwards, A. D., Cowan, F. M., & Azzopardi, D. (1999). Assessment of neonatal encephalopathy by amplitude-integrated electroencephalography. Pediatrics, 103(6 Pt 1), 1263-1271.
Al Shafouri, N., Narvey, M., Srinivasan, G., Vallance, J., & Hansen, G. (2015). High glucose variability is associated with poor neurodevelopmental outcomes in neonatal hypoxic ischemic encephalopathy. Journal of Neonatal-Perinatal Medicine, 8(2), 119-124. doi:10.3233/NPM-15814107
Alexandrou, G., Skiold, B., Karlen, J., Tessma, M. K., Norman, M., Aden, U., & Vanpee, M. (2010). Early hyperglycemia is a risk factor for death and white matter reduction in preterm infants. Pediatrics, 125(3), e584-591. doi:10.1542/peds.2009-0449
Ali, N. A., O'Brien, J. M., Jr., Dungan, K., Phillips, G., Marsh, C. B., Lemeshow, S., Connors, A. F., Jr., & Preiser, J. C. (2008). Glucose variability and mortality in patients with sepsis. Crit Care Med, 36(8), 2316-2321. doi:10.1097/CCM.0b013e3181810378
Alkalay, A. L., Flores-Sarnat, L., Sarnat, H. B., Moser, F. G., & Simmons, C. F. (2005). Brain imaging findings in neonatal hypoglycemia: case report and review of 23 cases. Clin Pediatr (Phila), 44(9), 783-790. doi:10.1177/000992280504400906
Allen, H. F., Rake, A., Roy, M., Brenner, D., & McKiernan, C. A. (2008). Prospective detection of hyperglycemia in critically ill children using continuous glucose monitoring. Pediatr Crit Care Med, 9(2), 153-158. doi:10.1097/PCC.0b013e3181668b33
Alrifai, M. T., AlShaya, M. A., Abulaban, A., & Alfadhel, M. (2014). Hereditary neurometabolic causes of infantile spasms in 80 children presenting to a tertiary care center. Pediatric Neurology, 51(3), 390-397. doi:10.1016/j.pediatrneurol.2014.05.015
Alsweiler, J. M., Harding, J. E., & Bloomfield, F. H. (2012). Tight glycemic control with insulin in hyperglycemic preterm babies: a randomized controlled trial. Pediatrics, 129(4), 639-647. doi:10.1542/peds.2011-2470
Alsweiler, J. M., Kuschel, C. A., & Bloomfield, F. H. (2007). Survey of the management of neonatal hyperglycaemia in Australasia. Journal of Paediatrics & Child Health, 43(9), 632-635. doi:10.1111/j.1440-1754.2007.01158.x
Aly, H., Elmahdy, H., El-Dib, M., Rowisha, M., Awny, M., El-Gohary, T., Elbatch, M., Hamisa, M., & El-Mashad, A. R. (2015). Melatonin use for neuroprotection in perinatal asphyxia: a randomized controlled pilot study. J Perinatol, 35(3), 186-191. doi:10.1038/jp.2014.186
114
Anderson, S. M., Raghinaru, D., Pinsker, J. E., Boscari, F., Renard, E., Buckingham, B. A., Nimri, R., Doyle, F. J., 3rd, Brown, S. A., Keith-Hynes, P., Breton, M. D., Chernavvsky, D., Bevier, W. C., Bradley, P. K., Bruttomesso, D., Del Favero, S., Calore, R., Cobelli, C., Avogaro, A., Farret, A., Place, J., Ly, T. T., Shanmugham, S., Phillip, M., Dassau, E., Dasanayake, I. S., Kollman, C., Lum, J. W., Beck, R. W., Kovatchev, B., & Control to Range Study, G. (2016). Multinational Home Use of Closed-Loop Control Is Safe and Effective. Diabetes Care, 39(7), 1143-1150. doi:10.2337/dc15-2468
Andre, M., Lamblin, M. D., d'Allest, A. M., Curzi-Dascalova, L., Moussalli-Salefranque, F., T, S. N. T., Vecchierini-Blineau, M. F., Wallois, F., Walls-Esquivel, E., & Plouin, P. (2010). Electroencephalography in premature and full-term infants. Developmental features and glossary. Neurophysiol Clin, 40(2), 59-124. doi:10.1016/j.neucli.2010.02.002
Ansell, J. M., Wouldes, T. A., Harding, J. E., & group, C. S. (2017). Executive function assessment in New Zealand 2-year olds born at risk of neonatal hypoglycemia. PLoS ONE [Electronic Resource], 12(11), e0188158. doi:10.1371/journal.pone.0188158
Anwar, M., & Vannucci, R. C. (1988). Autoradiographic determination of regional cerebral blood flow during hypoglycemia in newborn dogs. Pediatr Res, 24(1), 41-45. doi:10.1203/00006450-198807000-00011
Arhan, E., Ozturk, Z., Serdaroglu, A., Aydin, K., Hirfanoglu, T., & Akbas, Y. (2017). Neonatal hypoglycemia: A wide range of electroclinical manifestations and seizure outcomes. Eur J Paediatr Neurol, 21(5), 738-744. doi:10.1016/j.ejpn.2017.05.009
Arndt, D. H., Lerner, J. T., Matsumoto, J. H., Madikians, A., Yudovin, S., Valino, H., McArthur, D. L., Wu, J. Y., Leung, M., Buxey, F., Szeliga, C., Van Hirtum-Das, M., Sankar, R., Brooks-Kayal, A., & Giza, C. C. (2013). Subclinical early posttraumatic seizures detected by continuous EEG monitoring in a consecutive pediatric cohort. Epilepsia, 54(10), 1780-1788. doi:10.1111/epi.12369
Arsenault, D., Brenn, M., Kim, S., Gura, K., Compher, C., Simpser, E., American Society for, P., Enteral Nutrition Board of, D., & Puder, M. (2012). A.S.P.E.N. Clinical Guidelines: hyperglycemia and hypoglycemia in the neonate receiving parenteral nutrition. JPEN J Parenter Enteral Nutr, 36(1), 81-95. doi:10.1177/0148607111418980
Auer, R. N., Olsson, Y., & Siesjo, B. K. (1984). Hypoglycemic brain injury in the rat. Correlation of density of brain damage with the EEG isoelectric time: a quantitative study. Diabetes, 33(11), 1090-1098.
Awal, M. A., Lai, M. M., Azemi, G., Boashash, B., & Colditz, P. B. (2016). EEG background features that predict outcome in term neonates with hypoxic ischaemic encephalopathy: A structured review. Clin Neurophysiol, 127(1), 285-296. doi:10.1016/j.clinph.2015.05.018
Axelin, A., Cilio, M. R., Asunis, M., Peloquin, S., & Franck, L. S. (2013). Sleep-wake cycling in a neonate admitted to the NICU: a video-EEG case study during hypothermia treatment. J Perinat Neonatal Nurs, 27(3), 263-273. doi:10.1097/JPN.0b013e31829dc2d3
115
Aylward, G. P. (2004). Prediction of function from infancy to early childhood: implications for pediatric psychology. J Pediatr Psychol, 29(7), 555-564. doi:10.1093/jpepsy/jsh057
Aziz K, & Dancey P. (2004). Screening guidelines for newborns at risk for low blood glucose. Paediatr Child Health, 9(10), 723-729.
Azzopardi, D., Strohm, B., Edwards, A. D., Dyet, L., Halliday, H. L., Juszczak, E., Kapellou, O., Levene, M., Marlow, N., Porter, E., Thoresen, M., Whitelaw, A., Brocklehurst, P., & Toby Study Group. (2009). Moderate hypothermia to treat perinatal asphyxial encephalopathy. New England Journal of Medicine, 361(14), 1349-1358. doi:10.1056/NEJMoa0900854
Azzopardi, D., Strohm, B., Marlow, N., Brocklehurst, P., Deierl, A., Eddama, O., Goodwin, J., Halliday, H. L., Juszczak, E., Kapellou, O., Levene, M., Linsell, L., Omar, O., Thoresen, M., Tusor, N., Whitelaw, A., Edwards, A. D., & Toby Study Group. (2014). Effects of hypothermia for perinatal asphyxia on childhood outcomes. New England Journal of Medicine, 371(2), 140-149. doi:10.1056/NEJMoa1315788
Azzopardi, D., & Toby study group. (2014). Predictive value of the amplitude integrated EEG in infants with hypoxic ischaemic encephalopathy: data from a randomised trial of therapeutic hypothermia. Arch Dis Child Fetal Neonatal Ed, 99(1), F80-82. doi:10.1136/archdischild-2013-303710
Bagshaw, S. M., Bellomo, R., Jacka, M. J., Egi, M., Hart, G. K., George, C., & Committee, A. C. M. (2009). The impact of early hypoglycemia and blood glucose variability on outcome in critical illness. Crit Care, 13(3), R91. doi:10.1186/cc7921
Bailey, T. S., Chang, A., & Christiansen, M. (2015). Clinical accuracy of a continuous glucose monitoring system with an advanced algorithm. Journal of Diabetes Science & Technology, 9(2), 209-214. doi:10.1177/1932296814559746
Barkovich, A. J., Ali, F. A., Rowley, H. A., & Bass, N. (1998). Imaging patterns of neonatal hypoglycemia. AJNR Am J Neuroradiol, 19(3), 523-528.
Barkovich, A. J., Westmark, K., Partridge, C., Sola, A., & Ferriero, D. M. (1995). Perinatal asphyxia: MR findings in the first 10 days. AJNR Am J Neuroradiol, 16(3), 427-438. doi:0195-6108/95/1603–0427
Bashir, R. A., Espinoza, L., Vayalthrikkovil, S., Buchhalter, J., Irvine, L., Bello-Espinosa, L., & Mohammad, K. (2016). Implementation of a Neurocritical Care Program: Improved Seizure Detection and Decreased Antiseizure Medication at Discharge in Neonates With Hypoxic-Ischemic Encephalopathy. Pediatric Neurology, 64, 38-43. doi:10.1016/j.pediatrneurol.2016.07.007
Basu, P., Som, S., Choudhuri, N., & Das, H. (2009). Contribution of the blood glucose level in perinatal asphyxia. Eur J Pediatr, 168(7), 833-838. doi:10.1007/s00431-008-0844-5
Basu, S. K., Kaiser, J. R., Guffey, D., Minard, C. G., Guillet, R., Gunn, A. J., & CoolCap Study Group. (2016). Hypoglycaemia and hyperglycaemia are associated with unfavourable
116
outcome in infants with hypoxic ischaemic encephalopathy: a post hoc analysis of the CoolCap Study. Arch Dis Child Fetal Neonatal Ed, 101(2), F149-155. doi:10.1136/archdischild-2015-308733
Basu, S. K., Ottolini, K., Mashat, S., Govindan, V., Ridore, M., Vezina, G., Kaiser, J., & Massaro, A. (2017). Glycemic Profile in Infants with Hypoxic Ischemic Encephalopathy is Associated with MRI Patterns of Brain Injury. E-PAS2017 (Meeting Abstract), 2706.2703.
Basu, S. K., Salemi, J. L., Gunn, A. J., Kaiser, J. R., & CoolCap Study Group. (2017). Hyperglycaemia in infants with hypoxic-ischaemic encephalopathy is associated with improved outcomes after therapeutic hypothermia: a post hoc analysis of the CoolCap Study. Arch Dis Child Fetal Neonatal Ed, 102(4), F299-F306. doi:10.1136/archdischild-2016-311385
Battelino, T., Conget, I., Olsen, B., Schutz-Fuhrmann, I., Hommel, E., Hoogma, R., Schierloh, U., Sulli, N., Bolinder, J., & Group, S. S. (2012). The use and efficacy of continuous glucose monitoring in type 1 diabetes treated with insulin pump therapy: a randomised controlled trial. Diabetologia, 55(12), 3155-3162. doi:10.1007/s00125-012-2708-9
Baum, D., Dillard, D. H., Mohri, H., & Crawford, E. W. (1968). Metabolic aspects of deep surgical hypothermia in infancy. Pediatrics, 42(1), 93-105.
Beardsall, K., Ogilvy-Stuart, A. L., Ahluwalia, J., Thompson, M., & Dunger, D. B. (2005). The continuous glucose monitoring sensor in neonatal intensive care. Arch Dis Child Fetal Neonatal Ed, 90(4), F307-310. doi:10.1136/adc.2004.051979
Beardsall, K., Ogilvy-Stuart, A. L., Frystyk, J., Chen, J. W., Thompson, M., Ahluwalia, J., Ong, K. K., & Dunger, D. B. (2007). Early elective insulin therapy can reduce hyperglycemia and increase insulin-like growth factor-I levels in very low birth weight infants. J Pediatr, 151(6), 611-617, 617 e611. doi:10.1016/j.jpeds.2007.04.068
Beardsall, K., Vanhaesebrouck, S., Ogilvy-Stuart, A. L., Vanhole, C., Palmer, C. R., van Weissenbruch, M., Midgley, P., Thompson, M., Thio, M., Cornette, L., Ossuetta, I., Iglesias, I., Theyskens, C., de Jong, M., Ahluwalia, J. S., de Zegher, F., & Dunger, D. B. (2008). Early insulin therapy in very-low-birth-weight infants. New England Journal of Medicine, 359(18), 1873-1884. doi:10.1056/NEJMoa0803725
Beardsall, K., Vanhaesebrouck, S., Ogilvy-Stuart, A. L., Vanhole, C., VanWeissenbruch, M., Midgley, P., Thio, M., Cornette, L., Ossuetta, I., Palmer, C. R., Iglesias, I., de Jong, M., Gill, B., de Zegher, F., & Dunger, D. B. (2013). Validation of the continuous glucose monitoring sensor in preterm infants. Arch Dis Child Fetal Neonatal Ed, 98(2), F136-140. doi:10.1136/archdischild-2012-301661
Bellinger, D. C., Newburger, J. W., Wypij, D., Kuban, K. C., duPlesssis, A. J., & Rappaport, L. A. (2009). Behaviour at eight years in children with surgically corrected transposition: The Boston Circulatory Arrest Trial. Cardiol Young, 19(1), 86-97. doi:10.1017/S1047951108003454
117
Bellinger, D. C., Wypij, D., Kuban, K. C. K., Rappaport, L. A., Hickey, P. R., Wernovsky, G., Jonas, R. A., & Newburger, J. W. (1999). Developmental and Neurological Status of Children at 4 Years of Age After Heart Surgery With Hypothermic Circulatory Arrest or Low-Flow Cardiopulmonary Bypass. Circulation, 100(5), 526-532. doi:10.1161/01.cir.100.5.526
Bellinger, D. C., Wypij, D., Rivkin, M. J., DeMaso, D. R., Robertson, R. L., Jr., Dunbar-Masterson, C., Rappaport, L. A., Wernovsky, G., Jonas, R. A., & Newburger, J. W. (2011). Adolescents with d-transposition of the great arteries corrected with the arterial switch procedure: neuropsychological assessment and structural brain imaging. Circulation, 124(12), 1361-1369. doi:10.1161/CIRCULATIONAHA.111.026963
Bennet, L., Tan, S., Van den Heuij, L., Derrick, M., Groenendaal, F., van Bel, F., Juul, S., Back, S. A., Northington, F., Robertson, N. J., Mallard, C., & Gunn, A. J. (2012). Cell therapy for neonatal hypoxia-ischemia and cerebral palsy. Ann Neurol, 71(5), 589-600. doi:10.1002/ana.22670
Bermick, J., Dechert, R. E., & Sarkar, S. (2016). Does hyperglycemia in hypernatremic preterm infants increase the risk of intraventricular hemorrhage? J Perinatol, 36(9), 729-732. doi:10.1038/jp.2016.86
Biagioni, E., Bartalena, L., Boldrini, A., Pieri, R., & Cioni, G. (1999). Constantly discontinuous EEG patterns in full-term neonates with hypoxic-ischaemic encephalopathy. Clin Neurophysiol, 110(9), 1510-1515.
Biagioni, E., Mercuri, E., Rutherford, M., Cowan, F., Azzopardi, D., Frisone, M. F., Cioni, G., & Dubowitz, L. (2001). Combined Use of Electroencephalogram and Magnetic Resonance Imaging in Full-Term Neonates With Acute Encephalopathy. Pediatrics, 107(3), 461-468. doi:10.1542/peds.107.3.461
Bier, D. M., Leake, R. D., Haymond, M. W., Arnold, K. J., Gruenke, L. D., Sperling, M. A., & Kipnis, D. M. (1977). Measurement of "true" glucose production rates in infancy and childhood with 6,6-dideuteroglucose. Diabetes, 26(11), 1016-1023.
Birca, A., Lortie, A., Birca, V., Decarie, J. C., Veilleux, A., Gallagher, A., Dehaes, M., Lodygensky, G. A., & Carmant, L. (2016). Rewarming affects EEG background in term newborns with hypoxic-ischemic encephalopathy undergoing therapeutic hypothermia. Clin Neurophysiol, 127(4), 2087-2094. doi:10.1016/j.clinph.2015.12.013
Bittigau, P., Sifringer, M., Genz, K., Reith, E., Pospischil, D., Govindarajalu, S., Dzietko, M., Pesditschek, S., Mai, I., Dikranian, K., Olney, J. W., & Ikonomidou, C. (2002). Antiepileptic drugs and apoptotic neurodegeneration in the developing brain. Proc Natl Acad Sci U S A, 99(23), 15089-15094. doi:10.1073/pnas.222550499
Bittigau, P., Sifringer, M., & Ikonomidou, C. (2003). Antiepileptic drugs and apoptosis in the developing brain. Ann N Y Acad Sci, 993, 103-114; discussion 123-104.
118
Bjorgaas, M., Sand, T., Vik, T., & Jorde, R. (1998). Quantitative EEG during controlled hypoglycaemia in diabetic and non-diabetic children. Diabet Med, 15(1), 30-37. doi:10.1002/(SICI)1096-9136(199801)15:1<30::AID-DIA526>3.0.CO;2-R
Blanco, C. L., Baillargeon, J. G., Morrison, R. L., & Gong, A. K. (2006). Hyperglycemia in extremely low birth weight infants in a predominantly Hispanic population and related morbidities. J Perinatol, 26(12), 737-741. doi:10.1038/sj.jp.7211594
Boluyt, N., van Kempen, A., & Offringa, M. (2006). Neurodevelopment after neonatal hypoglycemia: a systematic review and design of an optimal future study. Pediatrics, 117(6), 2231-2243. doi:10.1542/peds.2005-1919
Bonifacio, S. L., deVries, L. S., & Groenendaal, F. (2015). Impact of hypothermia on predictors of poor outcome: how do we decide to redirect care? Semin Fetal Neonatal Med, 20(2), 122-127. doi:10.1016/j.siny.2014.12.011
Bonifacio, S. L., Glass, H. C., Vanderpluym, J., Agrawal, A. T., Xu, D., Barkovich, A. J., & Ferriero, D. M. (2011). Perinatal events and early magnetic resonance imaging in therapeutic hypothermia. J Pediatr, 158(3), 360-365. doi:10.1016/j.jpeds.2010.09.003
Booth, R. F., Patel, T. B., & Clark, J. B. (1980). The development of enzymes of energy metabolism in the brain of a precocial (guinea pig) and non-precocial (rat) species. J Neurochem, 34(1), 17-25.
Bottino, M., Cowett, R. M., & Sinclair, J. C. (2011). Interventions for treatment of neonatal hyperglycemia in very low birth weight infants. Cochrane Database of Systematic Reviews(10), CD007453. doi:10.1002/14651858.CD007453.pub3
Bougneres, P. F., Lemmel, C., Ferre, P., & Bier, D. M. (1986). Ketone body transport in the human neonate and infant. J Clin Invest, 77(1), 42-48. doi:10.1172/JCI112299
Boyle, P. J., Schwartz, N. S., Shah, S. D., Clutter, W. E., & Cryer, P. E. (1988). Plasma glucose concentrations at the onset of hypoglycemic symptoms in patients with poorly controlled diabetes and in nondiabetics. New England Journal of Medicine, 318(23), 1487-1492. doi:10.1056/NEJM198806093182302
Boyne, M. S., Silver, D. M., Kaplan, J., & Saudek, C. D. (2003). Timing of changes in interstitial and venous blood glucose measured with a continuous subcutaneous glucose sensor. Diabetes, 52(11), 2790-2794. doi:10.2337/diabetes.52.11.2790
Brand, P. L., Molenaar, N. L., Kaaijk, C., & Wierenga, W. S. (2005). Neurodevelopmental outcome of hypoglycaemia in healthy, large for gestational age, term newborns. Arch Dis Child, 90(1), 78-81. doi:10.1136/adc.2003.039412
Briatore, E., Ferrari, F., Pomero, G., Boghi, A., Gozzoli, L., Micciolo, R., Espa, G., Gancia, P., & Calzolari, S. (2013). EEG findings in cooled asphyxiated newborns and correlation with site and severity of brain damage. Brain & Development, 35(5), 420-426. doi:10.1016/j.braindev.2012.07.002
119
Bridges, B. C., Preissig, C. M., Maher, K. O., & Rigby, M. R. (2010). Continuous glucose monitors prove highly accurate in critically ill children. Crit Care, 14(5), R176. doi:10.1186/cc9280
Broad, K. D., Fierens, I., Fleiss, B., Rocha-Ferreira, E., Ezzati, M., Hassell, J., Alonso-Alconada, D., Bainbridge, A., Kawano, G., Ma, D., Tachtsidis, I., Gressens, P., Golay, X., Sanders, R. D., & Robertson, N. J. (2016). Inhaled 45-50% argon augments hypothermic brain protection in a piglet model of perinatal asphyxia. Neurobiol Dis, 87, 29-38. doi:10.1016/j.nbd.2015.12.001
Brownlee, M. (2001). Biochemistry and molecular cell biology of diabetic complications. Nature, 414(6865), 813-820. doi:10.1038/414813a
Brunner, R., Adelsmayr, G., Herkner, H., Madl, C., & Holzinger, U. (2012). Glycemic variability and glucose complexity in critically ill patients: a retrospective analysis of continuous glucose monitoring data. Crit Care, 16(5), R175. doi:10.1186/cc11657
Bruno, A., Levine, S. R., Frankel, M. R., Brott, T. G., Lin, Y., Tilley, B. C., Lyden, P. D., Broderick, J. P., Kwiatkowski, T. G., Fineberg, S. E., & Group, N. r.-P. S. S. (2002). Admission glucose level and clinical outcomes in the NINDS rt-PA Stroke Trial. Neurology, 59(5), 669-674. doi:10.1212/wnl.59.5.669
Bryan, R. M., Jr., Eichler, M. Y., Johnson, T. D., Woodward, W. T., & Williams, J. L. (1994). Cerebral blood flow, plasma catecholamines, and electroencephalogram during hypoglycemia and recovery after glucose infusion. J Neurosurg Anesthesiol, 6(1), 24-34.
Burns, C. M., Rutherford, M. A., Boardman, J. P., & Cowan, F. M. (2008). Patterns of cerebral injury and neurodevelopmental outcomes after symptomatic neonatal hypoglycemia. Pediatrics, 122(1), 65-74. doi:10.1542/peds.2007-2822
Burnsed, J., Quigg, M., Zanelli, S., & Goodkin, H. P. (2011). Clinical severity, rather than body temperature, during the rewarming phase of therapeutic hypothermia affect quantitative EEG in neonates with hypoxic ischemic encephalopathy. J Clin Neurophysiol, 28(1), 10-14. doi:10.1097/WNP.0b013e318205134b
Bye, A. M., & Flanagan, D. (1995). Spatial and temporal characteristics of neonatal seizures. Epilepsia, 36(10), 1009-1016.
Callahan, D. J., Engle, M. J., & Volpe, J. J. (1990). Hypoxic injury to developing glial cells: protective effect of high glucose. Pediatr Res, 27(2), 186-190. doi:10.1203/00006450-199002000-00020
Caplin, N. J., O'Leary, P., Bulsara, M., Davis, E. A., & Jones, T. W. (2003). Subcutaneous glucose sensor values closely parallel blood glucose during insulin-induced hypoglycaemia. Diabet Med, 20(3), 238-241. doi:10.1046/j.1464-5491.2003.00837.x
Caraballo, R. H., Sakr, D., Mozzi, M., Guerrero, A., Adi, J. N., Cersosimo, R. O., & Fejerman, N. (2004). Symptomatic occipital lobe epilepsy following neonatal hypoglycemia. Pediatric Neurology, 31(1), 24-29. doi:10.1016/j.pediatrneurol.2003.12.008
120
Castro Conde, J. R., Gonzalez Barrios, D., Gonzalez Campo, C., Gonzalez Gonzalez, N. L., Reyes Millan, B., & Sosa, A. J. (2017). Visual and Quantitative Electroencephalographic Analysis in Healthy Term Neonates Within the First Six Hours and the Third Day of Life. Pediatric Neurology, 77, 54-60 e51. doi:10.1016/j.pediatrneurol.2017.04.024
Cemeroglu, A. P., Stone, R., Kleis, L., Racine, M. S., Postellon, D. C., & Wood, M. A. (2010). Use of a real-time continuous glucose monitoring system in children and young adults on insulin pump therapy: patients' and caregivers' perception of benefit. Pediatr Diabetes, 11(3), 182-187. doi:10.1111/j.1399-5448.2009.00549.x
Chacko, S. K., Ordonez, J., Sauer, P. J., & Sunehag, A. L. (2011). Gluconeogenesis is not regulated by either glucose or insulin in extremely low birth weight infants receiving total parenteral nutrition. J Pediatr, 158(6), 891-896. doi:10.1016/j.jpeds.2010.12.040
Chandrasekaran, M., Chaban, B., Montaldo, P., & Thayyil, S. (2017). Predictive value of amplitude-integrated EEG (aEEG) after rescue hypothermic neuroprotection for hypoxic ischemic encephalopathy: a meta-analysis. J Perinatol, 37(6), 684-689. doi:10.1038/jp.2017.14
Chang, Y. S., Park, W. S., Ko, S. Y., Kang, M. J., Han, J. M., Lee, M., & Choi, J. (1999). Effects of fasting and insulin-induced hypoglycemia on brain cell membrane function and energy metabolism during hypoxia-ischemia in newborn piglets. Brain Res, 844(1-2), 135-142. doi:10.1016/s0006-8993(99)01940-x
Chang, Y. S., Park, W. S., Lee, M., Kim, K. S., Shin, S. M., & Choi, J. H. (1998). Effect of hyperglycemia on brain cell membrane function and energy metabolism during hypoxia-ischemia in newborn piglets. Brain Res, 798(1-2), 271-280. doi:10.1016/s0006-8993(98)00470-3
Chao, C. R., Hohimer, A. R., & Bissonnette, J. M. (1989). The effect of elevated blood glucose on the electroencephalogram and cerebral metabolism during short-term brain ischemia in fetal sheep. Am J Obstet Gynecol, 161(1), 221-228. doi:10.1016/0002-9378(89)90270-6
Chernavvsky, D. R., DeBoer, M. D., Keith-Hynes, P., Mize, B., McElwee, M., Demartini, S., Dunsmore, S. F., Wakeman, C., Kovatchev, B. P., & Breton, M. D. (2016). Use of an artificial pancreas among adolescents for a missed snack bolus and an underestimated meal bolus. Pediatr Diabetes, 17(1), 28-35. doi:10.1111/pedi.12230
Chouthai, N. S., Sobczak, H., Khan, R., Subramanian, D., Raman, S., & Rao, R. (2015). Hyperglycemia is associated with poor outcome in newborn infants undergoing therapeutic hypothermia for hypoxic ischemic encephalopathy. Journal of Neonatal-Perinatal Medicine, 8(2), 125-131. doi:10.3233/NPM-15814075
Collins, J. E., & Leonard, J. V. (1984). Hyperinsulinism in asphyxiated and small-for-dates infants with hypoglycaemia. Lancet, 2(8398), 311-313. doi:10.1016/s0140-6736(84)92685-0
121
Collins, J. W., Hoppe, M., Brown, K., Edidin, D. V., Padbury, J., & Ogata, E. S. (1991). A controlled trial of insulin infusion and parenteral nutrition in extremely low birth weight infants with glucose intolerance. J Pediatr, 118(6), 921-927.
Cornblath, M. (1955). Reactivation of rabbit liver phosphorylase by epinephrine, glucagon and ephedrine. Am J Physiol, 183(2), 240-244. doi:10.1152/ajplegacy.1955.183.2.240
Cornblath, M., Hawdon, J. M., Williams, A. F., Aynsley-Green, A., Ward-Platt, M. P., Schwartz, R., & Kalhan, S. C. (2000). Controversies Regarding Definition of Neonatal Hypoglycemia: Suggested Operational Thresholds. Pediatrics, 105(5), 1141-1145. doi:10.1542/peds.105.5.1141
Cornblath, M., Levin, E. Y., & Marquetti, E. (1958). Studies of carbohydrate metabolism in the newborn. II. The effect of glucagon on the concentration of sugar in capillary blood of the newborn infant. Pediatrics, 21(6), 885-892.
Cornblath, M., Odell, G. B., & Levin, E. Y. (1959). Symptomatic neonatal hypoglycemia associated with toxemia of pregnancy. J Pediatr, 55(5), 545-562. doi:10.1016/s0022-3476(59)80239-0
Costello, A. M. d. L., Pal, D. K., Manandhar, D. S., Rajbhandari, S., Land, J. M., & Patel, N. (2000). Neonatal hypoglycaemia in Nepal 2. Availability of alternative fuels. Archives of Disease in Childhood - Fetal and Neonatal Edition, 82(1), F52-F58. doi:10.1136/fn.82.1.F52
Cotten, C. M., Murtha, A. P., Goldberg, R. N., Grotegut, C. A., Smith, P. B., Goldstein, R. F., Fisher, K. A., Gustafson, K. E., Waters-Pick, B., Swamy, G. K., Rattray, B., Tan, S., & Kurtzberg, J. (2014). Feasibility of autologous cord blood cells for infants with hypoxic-ischemic encephalopathy. J Pediatr, 164(5), 973-979 e971. doi:10.1016/j.jpeds.2013.11.036
Cowett, R. M., Howard, G. M., Johnson, J., & Vohr, B. (1997). Brain stem auditory-evoked response in relation to neonatal glucose metabolism. Biol Neonate, 71(1), 31-36. doi:10.1159/000244394
Cowett, R. M., Oh, W., & Schwartz, R. (1983). Persistent glucose production during glucose infusion in the neonate. J Clin Invest, 71(3), 467-475. doi:10.1172/jci110791
Cremer, J. E., Cunningham, V. J., Pardridge, W. M., Braun, L. D., & Oldendorf, W. H. (1979). Kinetics of blood-brain barrier transport of pyruvate, lactate and glucose in suckling, weanling and adult rats. J Neurochem, 33(2), 439-445. doi:10.1111/j.1471-4159.1979.tb05173.x
Cseko, A. J., Bango, M., Lakatos, P., Kardasi, J., Pusztai, L., & Szabo, M. (2013). Accuracy of amplitude-integrated electroencephalography in the prediction of neurodevelopmental outcome in asphyxiated infants receiving hypothermia treatment. Acta Paediatr, 102(7), 707-711. doi:10.1111/apa.12226
122
Current Practice of Clinical Electroencephalography. (2014). (J. S. Ebersole, A. M. Husain, & D. R. Nordli Eds. Fourth ed.). Philadelphia: Wolters Kluwer.
D'Alton, M., Hankins, G., Berkowitz, R., Bienstock, J., Ghidini, A., Goldsmith, J., Higgins, R., Moore, T., Natale, R., Nelson, K., Papile, L., Peebles, D., Romero, R., Schendel, D., Spong, C., Waldman, R., Wu, Y., Joseph Jr, G., Hawks, D., Politzer, A., Emig, C., & Thomas, K. (2014). Executive summary: Neonatal encephalopathy and neurologic outcome, second edition. Report of the American College of Obstetricians and Gynecologists' Task Force on Neonatal Encephalopathy. Obstetrics & Gynecology, 123(4), 896-901. doi:10.1097/01.AOG.0000445580.65983.d2
Dalmazzo, C., Pomero, G., Delogu, A., Alpicrovi, C., Borgarello, G., Castellino, N., Simonitti, A., & Gancia, P. (2011). Correlation between MRI and neurological outcome at 1 year of age in asphyxiated cooled infants. Journal of Maternal-Fetal and Neonatal Medicine, 24, 168-169. doi:http://dx.doi.org/10.3109/14767058.2011.607964
Dan, B., Pelc, K., Cebolla, A. M., & Cheron, G. (2015). High-density electroencephalography developmental neurophysiological trajectories. Developmental Medicine & Child Neurology, 57 Suppl 3(S3), 44-47. doi:10.1111/dmcn.12728
Dasari, P. S., Gandomani, B. S., Teague, A. M., Pitale, A., Otto, M., & Short, K. R. (2016). Glycemic Variability Is Associated with Markers of Vascular Stress in Adolescents. J Pediatr, 172, 47-55 e42. doi:10.1016/j.jpeds.2016.01.065
Davidson, J. O., Fraser, M., Naylor, A. S., Roelfsema, V., Gunn, A. J., & Bennet, L. (2008). Effect of cerebral hypothermia on cortisol and adrenocorticotropic hormone responses after umbilical cord occlusion in preterm fetal sheep. Pediatr Res, 63(1), 51-55. doi:10.1203/PDR.0b013e31815b8eb4
Dawes, G. S., Jacobson, H. N., Mott, J. C., Shelley, H. J., & Stafford, A. (1963). The Treatment of Asphyxiated, Mature Foetal Lambs and Rhesus Monkeys with Intravenous Glucose and Sodium Carbonate. J Physiol, 169, 167-184.
Dawes, G. S., Mott, J. C., Shelley, H. J., & Stafford, A. (1963). The Prolongation of Survival Time in Asphyxiated Immature Foetal Lambs. J Physiol, 168, 43-64.
DeBoer, M. D., Breton, M. D., Wakeman, C., Schertz, E. M., Emory, E. G., Robic, J. L., Kollar, L. L., Kovatchev, B. P., & Chernavvsky, D. R. (2017). Performance of an Artificial Pancreas System for Young Children with Type 1 Diabetes. Diabetes Technology & Therapeutics, 19(5), 293-298. doi:10.1089/dia.2016.0424
DePuy, A. M., Coassolo, K. M., Som, D. A., & Smulian, J. C. (2009). Neonatal hypoglycemia in term, nondiabetic pregnancies. Am J Obstet Gynecol, 200(5), e45-51. doi:10.1016/j.ajog.2008.10.015
Dereymaeker, A., Pillay, K., Vervisch, J., De Vos, M., Van Huffel, S., Jansen, K., & Naulaers, G. (2017). Review of sleep-EEG in preterm and term neonates. Early Hum Dev, 113, 87-103. doi:10.1016/j.earlhumdev.2017.07.003
123
Desmond, M. M., Hild, J. R., & Gast, J. H. (1950). The glycemic response of the newborn infant to epinephrine administration: a preliminary report. J Pediatr, 37(3), 341-350. doi:10.1016/s0022-3476(50)80152-x
Diwakar, K. K., & Sasidhar, M. V. (2002). Plasma glucose levels in term infants who are appropriate size for gestation and exclusively breast fed. Arch Dis Child Fetal Neonatal Ed, 87(1), F46-48.
Dobbing, J., & Sands, J. (1979). Comparative aspects of the brain growth spurt. Early Hum Dev, 3(1), 79-83. doi:10.1016/0378-3782(79)90022-7
Dollaghan, C. A., Campbell, T. F., Paradise, J. L., Feldman, H. M., Janosky, J. E., Pitcairn, D. N., & Kurs-Lasky, M. (1999). Maternal education and measures of early speech and language. J Speech Lang Hear Res, 42(6), 1432-1443. doi:10.1044/jslhr.4206.1432
Dossett, L. A., Cao, H., Mowery, N. T., Dortch, M. J., Morris Jr, J. M., & May, A. K. (2008). Blood Glucose Variability Is Associated with Mortality in the Surgical Intensive Care Unit. Am Surg, 74, 679-685.
Drash, A., Kenny, F., Field, J., Blizzard, R., Langs, H., & Wolff, F. (1968). The therapeutic application of diazoxide in pediatric hypoglycemic states. Ann N Y Acad Sci, 150(2), 337-355. doi:10.1111/j.1749-6632.1968.tb19059.x
Dungan, K. M., Braithwaite, S. S., & Preiser, J. C. (2009). Stress hyperglycaemia. Lancet, 373(9677), 1798-1807. doi:10.1016/S0140-6736(09)60553-5
Dunne, J. M., Wertheim, D., Clarke, P., Kapellou, O., Chisholm, P., Boardman, J. P., & Shah, D. K. (2017). Automated electroencephalographic discontinuity in cooled newborns predicts cerebral MRI and neurodevelopmental outcome. Arch Dis Child Fetal Neonatal Ed, 102(1), F58-F64. doi:10.1136/archdischild-2015-309697
DuPont, T. L., Chalak, L. F., Morriss, M. C., Burchfield, P. J., Christie, L., & Sanchez, P. J. (2013). Short-term outcomes of newborns with perinatal acidemia who are not eligible for systemic hypothermia therapy. J Pediatr, 162(1), 35-41. doi:10.1016/j.jpeds.2012.06.042
Duvanel, C. B., Fawer, C. L., Cotting, J., Hohlfeld, P., & Matthieu, J. M. (1999). Long-term effects of neonatal hypoglycemia on brain growth and psychomotor development in small-for-gestational-age preterm infants. J Pediatr, 134(4), 492-498.
Egi, M., Bellomo, R., Stachowski, E., French, C. J., & Hart, G. (2006). Variability of blood glucose concentration and short-term mortality in critically ill patients. Anesthesiology, 105(2), 244-252.
Ennis, K., Dotterman, H., Stein, A., & Rao, R. (2015). Hyperglycemia accentuates and ketonemia attenuates hypoglycemia-induced neuronal injury in the developing rat brain. Pediatr Res, 77(1-1), 84-90. doi:10.1038/pr.2014.146
124
Ertl, T., Gyarmati, J., Gaal, V., & Szabo, I. (2006). Relationship between hyperglycemia and retinopathy of prematurity in very low birth weight infants. Biol Neonate, 89(1), 56-59. doi:10.1159/000088199
Eslami, S., Taherzadeh, Z., Schultz, M. J., & Abu-Hanna, A. (2011). Glucose variability measures and their effect on mortality: a systematic review. Intensive Care Med, 37(4), 583-593. doi:10.1007/s00134-010-2129-5
Fabricius, M., Fuhr, S., Willumsen, L., Dreier, J. P., Bhatia, R., Boutelle, M. G., Hartings, J. A., Bullock, R., Strong, A. J., & Lauritzen, M. (2008). Association of seizures with cortical spreading depression and peri-infarct depolarisations in the acutely injured human brain. Clin Neurophysiol, 119(9), 1973-1984. doi:10.1016/j.clinph.2008.05.025
Falcao, G., Ulate, K., Kouzekanani, K., Bielefeld, M. R., Morales, J. M., & Rotta, A. T. (2008). Impact of postoperative hyperglycemia following surgical repair of congenital cardiac defects. Pediatric Cardiology, 29(3), 628-636. doi:10.1007/s00246-007-9178-8
Faustino, E. V., & Apkon, M. (2005). Persistent hyperglycemia in critically ill children. J Pediatr, 146(1), 30-34. doi:10.1016/j.jpeds.2004.08.076
Fendler, W., Walenciak, J., Mlynarski, W., & Piotrowski, A. (2012). Higher glycemic variability in very low birth weight newborns is associated with greater early neonatal mortality. Journal of Maternal-Fetal & Neonatal Medicine, 25(7), 1122-1126. doi:10.3109/14767058.2011.624220
Fernald, A., Marchman, V. A., & Weisleder, A. (2013). SES differences in language processing skill and vocabulary are evident at 18 months. Dev Sci, 16(2), 234-248. doi:10.1111/desc.12019
Filan, P. M., Inder, T. E., Cameron, F. J., Kean, M. J., & Hunt, R. W. (2006). Neonatal hypoglycemia and occipital cerebral injury. J Pediatr, 148(4), 552-555. doi:10.1016/j.jpeds.2005.11.015
Filippi, L., Fiorini, P., Catarzi, S., Berti, E., Padrini, L., Landucci, E., Donzelli, G., Bartalena, L., Fiorentini, E., Boldrini, A., Giampietri, M., Scaramuzzo, R. T., la Marca, G., Della Bona, M. L., Fiori, S., Tinelli, F., Bancale, A., Guzzetta, A., Cioni, G., Pisano, T., Falchi, M., & Guerrini, R. (2018). Safety and efficacy of topiramate in neonates with hypoxic ischemic encephalopathy treated with hypothermia (NeoNATI): a feasibility study. Journal of Maternal-Fetal & Neonatal Medicine, 31(8), 973-980. doi:10.1080/14767058.2017.1304536
Finfer, S., Chittock, D. R., Su, S. Y., Blair, D., Foster, D., Dhingra, V., Bellomo, R., Cook, D., Dodek, P., Henderson, W. R., Hebert, P. C., Heritier, S., Heyland, D. K., McArthur, C., McDonald, E., Mitchell, I., Myburgh, J. A., Norton, R., Potter, J., Robinson, B. G., Ronco, J. J., & Nice-Sugar Study Investigators. (2009). Intensive versus conventional glucose control in critically ill patients. New England Journal of Medicine, 360(13), 1283-1297. doi:10.1056/NEJMoa0810625
125
Flisberg, A., Kjellmer, I., Lofhede, J., Lindecrantz, K., & Thordstein, M. (2011). Prognostic capacity of automated quantification of suppression time in the EEG of post-asphyctic full-term neonates. Acta Paediatr, 100(10), 1338-1343. doi:10.1111/j.1651-2227.2011.02323.x
Fong, C. Y., & Harvey, A. S. (2014). Variable outcome for epilepsy after neonatal hypoglycaemia. Developmental Medicine & Child Neurology, 56(11), 1093-1099. doi:10.1111/dmcn.12496
Fowden, A. L. (1980). Effects of adrenaline and amino acids on the release of insulin in the sheep fetus. J Endocrinol, 87(1), 113-121.
Fox, C. K., Glass, H. C., Sidney, S., Smith, S. E., & Fullerton, H. J. (2016). Neonatal seizures triple the risk of a remote seizure after perinatal ischemic stroke. Neurology, 86(23), 2179-2186. doi:10.1212/WNL.0000000000002739
Fox, C. K., Mackay, M. T., Dowling, M. M., Pergami, P., Titomanlio, L., Deveber, G., & Investigators, S. (2017). Prolonged or recurrent acute seizures after pediatric arterial ischemic stroke are associated with increasing epilepsy risk. Developmental Medicine & Child Neurology, 59(1), 38-44. doi:10.1111/dmcn.13198
Freeman, J., & Ingvar, D. H. (1968). Elimination by hypoxia cerebral blood flow autoregulation and EEG relationship. Exp Brain Res, 5(1), 61-71. doi:10.1007/bf00239906
Frenkel, N., Friger, M., Meledin, I., Berger, I., Marks, K., Bassan, H., & Shany, E. (2011). Neonatal seizure recognition--comparative study of continuous-amplitude integrated EEG versus short conventional EEG recordings. Clin Neurophysiol, 122(6), 1091-1097. doi:10.1016/j.clinph.2010.09.028
Fuglsang, A., Lomholt, M., & Gjedde, A. (1986). Blood-brain transfer of glucose and glucose analogs in newborn rats. J Neurochem, 46(5), 1417-1428. doi:10.1111/j.1471-4159.1986.tb01757.x
Galderisi, A., Facchinetti, A., Steil, G. M., Ortiz-Rubio, P., Cavallin, F., Tamborlane, W. V., Baraldi, E., Cobelli, C., & Trevisanuto, D. (2017). Continuous Glucose Monitoring in Very Preterm Infants: A Randomized Controlled Trial. Pediatrics, 140(4). doi:10.1542/peds.2017-1162
Garcia Maset, L., Gonzalez, L. B., Furquet, G. L., Suay, F. M., & Marco, R. H. (2016). Study of Glycemic Variability Through Time Series Analyses (Detrended Fluctuation Analysis and Poincare Plot) in Children and Adolescents with Type 1 Diabetes. Diabetes Technology & Therapeutics, 18(11), 719-724. doi:10.1089/dia.2016.0208
Gataullina, S., De Lonlay, P., Dellatolas, G., Valayannapoulos, V., Napuri, S., Damaj, L., Touati, G., Altuzarra, C., Dulac, O., & Boddaert, N. (2013). Topography of brain damage in metabolic hypoglycaemia is determined by age at which hypoglycaemia occurred. Developmental Medicine & Child Neurology, 55(2), 162-166. doi:10.1111/dmcn.12045
126
Gaynor, J. W., Jarvik, G. P., Bernbaum, J., Gerdes, M., Wernovsky, G., Burnham, N. B., D'Agostino, J. A., Zackai, E., McDonald-McGinn, D. M., Nicolson, S. C., Spray, T. L., & Clancy, R. R. (2006). The relationship of postoperative electrographic seizures to neurodevelopmental outcome at 1 year of age after neonatal and infant cardiac surgery. J Thorac Cardiovasc Surg, 131(1), 181-189. doi:10.1016/j.jtcvs.2005.08.062
Gaynor, J. W., Jarvik, G. P., Gerdes, M., Kim, D. S., Rajagopalan, R., Bernbaum, J., Wernovsky, G., Nicolson, S. C., Spray, T. L., & Clancy, R. R. (2013). Postoperative electroencephalographic seizures are associated with deficits in executive function and social behaviors at 4 years of age following cardiac surgery in infancy. J Thorac Cardiovasc Surg, 146(1), 132-137. doi:10.1016/j.jtcvs.2013.04.002
Giesinger, R. E., Bailey, L. J., Deshpande, P., & McNamara, P. J. (2017). Hypoxic-Ischemic Encephalopathy and Therapeutic Hypothermia: The Hemodynamic Perspective. J Pediatr, 180, 22-30 e22. doi:10.1016/j.jpeds.2016.09.009
Glass, H. C., Nash, K. B., Bonifacio, S. L., Barkovich, A. J., Ferriero, D. M., Sullivan, J. E., & Cilio, M. R. (2011). Seizures and magnetic resonance imaging-detected brain injury in newborns cooled for hypoxic-ischemic encephalopathy. J Pediatr, 159(5), 731-735 e731. doi:10.1016/j.jpeds.2011.07.015
Glass, H. C., Shellhaas, R. A., Wusthoff, C. J., Chang, T., Abend, N. S., Chu, C. J., Cilio, M. R., Glidden, D. V., Bonifacio, S. L., Massey, S., Tsuchida, T. N., Silverstein, F. S., Soul, J. S., & Neonatal Seizure Registry Study, G. (2016). Contemporary Profile of Seizures in Neonates: A Prospective Cohort Study. J Pediatr, 174, 98-103 e101. doi:10.1016/j.jpeds.2016.03.035
Glass, H. C., Wusthoff, C. J., & Shellhaas, R. A. (2013). Amplitude-integrated electro-encephalography: the child neurologist's perspective. J Child Neurol, 28(10), 1342-1350. doi:10.1177/0883073813488663
Glass, H. C., Wusthoff, C. J., Shellhaas, R. A., Tsuchida, T. N., Bonifacio, S. L., Cordeiro, M., Sullivan, J., Abend, N. S., & Chang, T. (2014). Risk factors for EEG seizures in neonates treated with hypothermia: a multicenter cohort study. Neurology, 82(14), 1239-1244. doi:10.1212/WNL.0000000000000282
Gluckman, P. D., Wyatt, J. S., Azzopardi, D., Ballard, R., Edwards, A. D., Ferriero, D. M., Polin, R. A., Robertson, C. M., Thoresen, M., Whitelaw, A., & Gunn, A. J. (2005). Selective head cooling with mild systemic hypothermia after neonatal encephalopathy: multicentre randomised trial. Lancet, 365(9460), 663-670. doi:10.1016/S0140-6736(05)17946-X
Glykys, J., Dzhala, V. I., Kuchibhotla, K. V., Feng, G., Kuner, T., Augustine, G., Bacskai, B. J., & Staley, K. J. (2009). Differences in cortical versus subcortical GABAergic signaling: a candidate mechanism of electroclinical uncoupling of neonatal seizures. Neuron, 63(5), 657-672. doi:10.1016/j.neuron.2009.08.022
Granot, S., Meledin, I., Richardson, J., Friger, M., & Shany, E. (2012). Influence of respiratory acidosis and blood glucose on cerebral activity of premature infants. Pediatric Neurology, 47(1), 19-24. doi:10.1016/j.pediatrneurol.2012.03.018
127
Guemes, M., Rahman, S. A., & Hussain, K. (2016). What is a normal blood glucose? Arch Dis Child, 101(6), 569-574. doi:10.1136/archdischild-2015-308336
Guillet, R., Edwards, A. D., Thoresen, M., Ferriero, D. M., Gluckman, P. D., Whitelaw, A., Gunn, A. J., & CoolCap Trial, G. (2012). Seven- to eight-year follow-up of the CoolCap trial of head cooling for neonatal encephalopathy. Pediatr Res, 71(2), 205-209. doi:10.1038/pr.2011.30
Gunes, T., Ozturk, M. A., Koklu, E., Kose, K., & Gunes, I. (2007). Effect of allopurinol supplementation on nitric oxide levels in asphyxiated newborns. Pediatric Neurology, 36(1), 17-24. doi:10.1016/j.pediatrneurol.2006.08.005
Gwer, S., Idro, R., Fegan, G., Chengo, E., Garrashi, H., White, S., Kirkham, F. J., & Newton, C. R. (2012). Continuous EEG monitoring in Kenyan children with non-traumatic coma. Arch Dis Child, 97(4), 343-349. doi:10.1136/archdischild-2011-300935
Hack, M., Taylor, H. G., Drotar, D., Schluchter, M., Cartar, L., Wilson-Costello, D., Klein, N., Friedman, H., Mercuri-Minich, N., & Morrow, M. (2005). Poor predictive validity of the Bayley Scales of Infant Development for cognitive function of extremely low birth weight children at school age. Pediatrics, 116(2), 333-341. doi:10.1542/peds.2005-0173
Hahn, C. D., & Jette, N. (2013). Neurocritical care: Seizures after acute brain injury--more than meets the eye. Nat Rev Neurol, 9(12), 662-664. doi:10.1038/nrneurol.2013.231
Hall, N. J., Peters, M., Eaton, S., & Pierro, A. (2004). Hyperglycemia is associated with increased morbidity and mortality rates in neonates with necrotizing enterocolitis. Journal of Pediatric Surgery, 39(6), 898-901. doi:10.1016/j.jpedsurg.2004.02.005
Harding, J. E., Harris, D. L., Hegarty, J. E., Alsweiler, J. M., & McKinlay, C. J. (2017). An emerging evidence base for the management of neonatal hypoglycaemia. Early Hum Dev, 104, 51-56. doi:10.1016/j.earlhumdev.2016.12.009
Hardstone, R., Poil, S. S., Schiavone, G., Jansen, R., Nikulin, V. V., Mansvelder, H. D., & Linkenkaer-Hansen, K. (2012). Detrended fluctuation analysis: a scale-free view on neuronal oscillations. Front Physiol, 3, 450. doi:10.3389/fphys.2012.00450
Harris, D. L., Alsweiler, J. M., Ansell, J. M., Gamble, G. D., Thompson, B., Wouldes, T. A., Yu, T. Y., Harding, J. E., Children with, H., & their Later Development Study, T. (2016). Outcome at 2 Years after Dextrose Gel Treatment for Neonatal Hypoglycemia: Follow-Up of a Randomized Trial. J Pediatr, 170, 54-59 e51-52. doi:10.1016/j.jpeds.2015.10.066
Harris, D. L., Battin, M. R., Weston, P. J., & Harding, J. E. (2010). Continuous glucose monitoring in newborn babies at risk of hypoglycemia. J Pediatr, 157(2), 198-202 e191. doi:10.1016/j.jpeds.2010.02.003
Harris, D. L., Battin, M. R., Williams, C. E., Weston, P. J., & Harding, J. E. (2009). Cot-side electro-encephalography and interstitial glucose monitoring during insulin-induced hypoglycaemia in newborn lambs. Neonatology, 95(4), 271-278. doi:10.1159/000166847
128
Harris, D. L., Weston, P. J., & Harding, J. E. (2012). Incidence of neonatal hypoglycemia in babies identified as at risk. J Pediatr, 161(5), 787-791. doi:10.1016/j.jpeds.2012.05.022
Harris, D. L., Weston, P. J., & Harding, J. E. (2015). Lactate, rather than ketones, may provide alternative cerebral fuel in hypoglycaemic newborns. Arch Dis Child Fetal Neonatal Ed, 100(2), F161-164. doi:10.1136/archdischild-2014-306435
Harris, D. L., Weston, P. J., Signal, M., Chase, J. G., & Harding, J. E. (2013). Dextrose gel for neonatal hypoglycaemia (the Sugar Babies Study): a randomised, double-blind, placebo-controlled trial. Lancet, 382(9910), 2077-2083. doi:10.1016/S0140-6736(13)61645-1
Harris, D. L., Weston, P. J., Williams, C. E., Pleasants, A. B., Battin, M. R., Spooner, C. G., & Harding, J. E. (2011). Cot-side electroencephalography monitoring is not clinically useful in the detection of mild neonatal hypoglycemia. J Pediatr, 159(5), 755-760 e751. doi:10.1016/j.jpeds.2011.04.026
Harris, P. A., Taylor, R., Thielke, R., Payne, J., Gonzalez, N., & Conde, J. G. (2009). Research electronic data capture (REDCap)--a metadata-driven methodology and workflow process for providing translational research informatics support. J Biomed Inform, 42(2), 377-381. doi:10.1016/j.jbi.2008.08.010
Hartings, J. A., Wilson, J. A., Hinzman, J. M., Pollandt, S., Dreier, J. P., DiNapoli, V., Ficker, D. M., Shutter, L. A., & Andaluz, N. (2014). Spreading depression in continuous electroencephalography of brain trauma. Ann Neurol, 76(5), 681-694. doi:10.1002/ana.24256
Hathi, M., Sherman, D. L., Inder, T., Rothman, N. S., Natarajan, M., Niesen, C., Korst, L. M., Pantano, T., & Natarajan, A. (2010). Quantitative EEG in babies at risk for hypoxic ischemic encephalopathy after perinatal asphyxia. J Perinatol, 30(2), 122-126. doi:10.1038/jp.2009.130
Hattersley, A. T., Beards, F., Ballantyne, E., Appleton, M., Harvey, R., & Ellard, S. (1998). Mutations in the glucokinase gene of the fetus result in reduced birth weight. Nat Genet, 19(3), 268-270. doi:10.1038/953
Hattori, H., & Wasterlain, C. G. (1990). Posthypoxic glucose supplement reduces hypoxic-ischemic brain damage in the neonatal rat. Ann Neurol, 28(2), 122-128. doi:10.1002/ana.410280203
Hawdon, J. M., Aynsley-Green, A., Alberti, K. G., & Ward Platt, M. P. (1993). The role of pancreatic insulin secretion in neonatal glucoregulation. I. Healthy term and preterm infants. Arch Dis Child, 68(3 Spec No), 274-279.
Hawdon, J. M., Aynsley-Green, A., Bartlett, K., & Ward Platt, M. P. (1993). The role of pancreatic insulin secretion in neonatal glucoregulation. II. Infants with disordered blood glucose homoeostasis. Arch Dis Child, 68(3 Spec No), 280-285.
129
Hawdon, J. M., Ward Platt, M. P., & Aynsley-Green, A. (1992). Patterns of metabolic adaptation for preterm and term infants in the first neonatal week. Arch Dis Child, 67(4 Spec No), 357-365.
Hawkins, R. A., Williamson, D. H., & Krebs, H. A. (1971). Ketone-body utilization by adult and suckling rat brainin vivo. Biochemical Journal, 122(1), 13-18. doi:10.1042/bj1220013
Haworth, J. C., & Coodin, F. J. (1960). Idiopathic spontaneous hypoglycemia in children. Report of seven cases and review of the literature. Pediatrics, 25, 748-765.
Hay, W. W., Jr., Raju, T. N., Higgins, R. D., Kalhan, S. C., & Devaskar, S. U. (2009). Knowledge gaps and research needs for understanding and treating neonatal hypoglycemia: workshop report from Eunice Kennedy Shriver National Institute of Child Health and Human Development. J Pediatr, 155(5), 612-617. doi:10.1016/j.jpeds.2009.06.044
Hays, S. P., Smith, E. O., & Sunehag, A. L. (2006). Hyperglycemia is a risk factor for early death and morbidity in extremely low birth-weight infants. Pediatrics, 118(5), 1811-1818. doi:10.1542/peds.2006-0628
Heck, L. J., & Erenberg, A. (1987). Serum glucose levels in term neonates during the first 48 hours of life. J Pediatr, 110(1), 119-122.
Heimann, K., Peschgens, T., Kwiecien, R., Stanzel, S., Hoernchen, H., & Merz, U. (2007). Are recurrent hyperglycemic episodes and median blood glucose level a prognostic factor for increased morbidity and mortality in premature infants </=1500 g? J Perinat Med, 35(3), 245-248. doi:10.1515/JPM.2007.057
Hellmann, J., Vannucci, R., & Nardis, E. (1982). Blood-brain barrier permeability to lactic acid in the newborn dog: lactate as a cerebral metabolic fuel. Pediatr Res, 16(1), 40-44.
Hellström-Westas, L., de Vries, L. S., & Rosen, I. (2008). An Atlas of Amplitude-Integrated EEGs in the Newborn, Second Edition. London: CRC Press.
Hellström-Westas, L., Rosén, I., De Vries, L. S., & Greisen, G. (2006). Amplitude-integrated EEG Classification and Interpretation in Preterm and Term Infants. NeoReviews, 7(2), e76-e87. doi:10.1542/neo.7-2-e76
Hellstrom-Westas, L., Rosen, I., & Svenningsen, N. W. (1989). Cerebral complications detected by EEG-monitoring during neonatal intensive care. Acta Paediatr Scand Suppl, 360, 83-86.
Hellstrom-Westas, L., Rosen, I., & Svenningsen, N. W. (1995). Predictive value of early continuous amplitude integrated EEG recordings on outcome after severe birth asphyxia in full term infants. Arch Dis Child Fetal Neonatal Ed, 72(1), F34-38.
Helmers, S. L., Wypij, D., Constantinou, J. E., Newburger, J. W., Hickey, P. R., Carrazana, E. J., Barlow, J. K., Kuban, K. C., & Holmes, G. L. (1997). Perioperative
130
electroencephalographic seizures in infants undergoing repair of complex congenital cardiac defects. Electroencephalogr Clin Neurophysiol, 102(1), 27-36.
Hermanides, J., Vriesendorp, T. M., Bosman, R. J., Zandstra, D. F., Hoekstra, J. B., & Devries, J. H. (2010). Glucose variability is associated with intensive care unit mortality. Crit Care Med, 38(3), 838-842. doi:10.1097/CCM.0b013e3181cc4be9
Hirsch, I. B., & Brownlee, M. (2005). Should minimal blood glucose variability become the gold standard of glycemic control? J Diabetes Complications, 19(3), 178-181. doi:10.1016/j.jdiacomp.2004.10.001
Hirshberg, E., Larsen, G., & Van Duker, H. (2008). Alterations in glucose homeostasis in the pediatric intensive care unit: Hyperglycemia and glucose variability are associated with increased mortality and morbidity. Pediatr Crit Care Med, 9(4), 361-366. doi:10.1097/PCC.0b013e318172d401
Hoe, F. M., Thornton, P. S., Wanner, L. A., Steinkrauss, L., Simmons, R. A., & Stanley, C. A. (2006). Clinical features and insulin regulation in infants with a syndrome of prolonged neonatal hyperinsulinism. J Pediatr, 148(2), 207-212. doi:10.1016/j.jpeds.2005.10.002
Holmes, G. L., & Lombroso, C. T. (1993). Prognostic value of background patterns in the neonatal EEG. J Clin Neurophysiol, 10(3), 323-352.
Horan, M., Azzopardi, D., Edwards, A. D., Firmin, R. K., & Field, D. (2007). Lack of influence of mild hypothermia on amplitude integrated-electroencephalography in neonates receiving extracorporeal membrane oxygenation. Early Hum Dev, 83(2), 69-75. doi:10.1016/j.earlhumdev.2006.05.004
Hoseth, E., Joergensen, A., Ebbesen, F., & Moeller, M. (2000). Blood glucose levels in a population of healthy, breast fed, term infants of appropriate size for gestational age. Arch Dis Child Fetal Neonatal Ed, 83(2), F117-119.
Ichord, R. N., Helfaer, M. A., Kirsch, J. R., Wilson, D., & Traystman, R. J. (1994). Nitric oxide synthase inhibition attenuates hypoglycemic cerebral hyperemia in piglets. Am J Physiol, 266(3 Pt 2), H1062-1068. doi:10.1152/ajpheart.1994.266.3.H1062
Ihlen, E. A. (2012). Introduction to multifractal detrended fluctuation analysis in matlab. Front Physiol, 3, 141. doi:10.3389/fphys.2012.00141
Illingworth, R. S., & Birch, L. B. (1959). The Diagnosis of Mental Retardation in Infancy: A Follow-Up Study. Arch Dis Child, 34(175), 269-273.
Iyer, K. K., Roberts, J. A., Metsaranta, M., Finnigan, S., Breakspear, M., & Vanhatalo, S. (2014). Novel features of early burst suppression predict outcome after birth asphyxia. Ann Clin Transl Neurol, 1(3), 209-214. doi:10.1002/acn3.32
Jacobs, S. E., Berg, M., Hunt, R., Tarnow-Mordi, W. O., Inder, T. E., & Davis, P. G. (2013). Cooling for newborns with hypoxic ischaemic encephalopathy. Cochrane Database of Systematic Reviews(1), CD003311. doi:10.1002/14651858.CD003311.pub3
131
Jacobs, S. E., Morley, C. J., Inder, T. E., Stewart, M. J., Smith, K. R., McNamara, P. J., Wright, I. M., Kirpalani, H. M., Darlow, B. A., Doyle, L. W., & Infant Cooling Evaluation, C. (2011). Whole-body hypothermia for term and near-term newborns with hypoxic-ischemic encephalopathy: a randomized controlled trial. Arch Pediatr Adolesc Med, 165(8), 692-700. doi:10.1001/archpediatrics.2011.43
Jain, S. V., Zempel, J. M., Srinivasakumar, P., Wallendorf, M., & Mathur, A. (2017). Early EEG power predicts MRI injury in infants with hypoxic-ischemic encephalopathy. J Perinatol, 37(5), 541-546. doi:10.1038/jp.2016.262
Jennekens, W., Niemarkt, H. J., Engels, M., Pasman, J. W., van Pul, C., & Andriessen, P. (2012). Topography of maturational changes in EEG burst spectral power of the preterm infant with a normal follow-up at 2 years of age. Clin Neurophysiol, 123(11), 2130-2138. doi:10.1016/j.clinph.2012.03.018
Jensen, E. C., Bennet, L., Hunter, C. J., Power, G. C., & Gunn, A. J. (2006). Post-hypoxic hypoperfusion is associated with suppression of cerebral metabolism and increased tissue oxygenation in near-term fetal sheep. J Physiol, 572(Pt 1), 131-139. doi:10.1113/jphysiol.2005.100768
Jiang, Z. D., Zhou, Y., Yin, R., & Wilkinson, A. R. (2013). Amplitude reduction in brainstem auditory response in term infants under neonatal intensive care. Clin Neurophysiol, 124(7), 1470-1476. doi:10.1016/j.clinph.2013.02.009
Kaiser, J. R., Bai, S., Gibson, N., Holland, G., Lin, T. M., Swearingen, C. J., Mehl, J. K., & ElHassan, N. O. (2015). Association Between Transient Newborn Hypoglycemia and Fourth-Grade Achievement Test Proficiency: A Population-Based Study. JAMA Pediatr, 169(10), 913-921. doi:10.1001/jamapediatrics.2015.1631
Kalhan, S. C., D'Angelo, L. J., Savin, S. M., & Adam, P. A. (1979). Glucose production in pregnant women at term gestation. Sources of glucose for human fetus. J Clin Invest, 63(3), 388-394. doi:10.1172/JCI109314
Kantelhardt, J. W., Koscielny-Bunde, E., Rego, H. H. A., Havlin, S., & Bunde, A. (2001). Detecting long-range correlations with detrended fluctuation analysis. Physica A: Statistical Mechanics and its Applications, 295(3-4), 441-454. doi:10.1016/s0378-4371(01)00144-3
Kao, L. S., Morris, B. H., Lally, K. P., Stewart, C. D., Huseby, V., & Kennedy, K. A. (2006). Hyperglycemia and morbidity and mortality in extremely low birth weight infants. J Perinatol, 26(12), 730-736. doi:10.1038/sj.jp.7211593
Karimzadeh, P., Tabarestani, S., & Ghofrani, M. (2011). Hypoglycemia-occipital syndrome: a specific neurologic syndrome following neonatal hypoglycemia? J Child Neurol, 26(2), 152-159. doi:10.1177/0883073810376245
Kerstjens, J. M., Bocca-Tjeertes, I. F., de Winter, A. F., Reijneveld, S. A., & Bos, A. F. (2012). Neonatal morbidities and developmental delay in moderately preterm-born children. Pediatrics, 130(2), e265-272. doi:10.1542/peds.2012-0079
132
Khan, R. L., Nunes, M. L., Garcias da Silva, L. F., & da Costa, J. C. (2008). Predictive value of sequential electroencephalogram (EEG) in neonates with seizures and its relation to neurological outcome. J Child Neurol, 23(2), 144-150. doi:10.1177/0883073807308711
Kharoshankaya, L., Stevenson, N. J., Livingstone, V., Murray, D. M., Murphy, B. P., Ahearne, C. E., & Boylan, G. B. (2016). Seizure burden and neurodevelopmental outcome in neonates with hypoxic-ischemic encephalopathy. Developmental Medicine & Child Neurology, 58(12), 1242-1248. doi:10.1111/dmcn.13215
Kim, M., Yu, Z. X., Fredholm, B. B., & Rivkees, S. A. (2005). Susceptibility of the developing brain to acute hypoglycemia involving A1 adenosine receptor activation. Am J Physiol Endocrinol Metab, 289(4), E562-569. doi:10.1152/ajpendo.00112.2005
Kim, S. Y., Goo, H. W., Lim, K. H., Kim, S. T., & Kim, K. S. (2006). Neonatal hypoglycaemic encephalopathy: diffusion-weighted imaging and proton MR spectroscopy. Pediatr Radiol, 36(2), 144-148. doi:10.1007/s00247-005-0020-2
Kim, Y. B., Gidday, J. M., Gonzales, E. R., Shah, A. R., & Park, T. S. (1994). Effect of hypoglycemia on postischemic cortical blood flow, hypercapnic reactivity, and interstitial adenosine concentration. J Neurosurg, 81(6), 877-884. doi:10.3171/jns.1994.81.6.0877
Kinnala, A., Rikalainen, H., Lapinleimu, H., Parkkola, R., Kormano, M., & Kero, P. (1999). Cerebral magnetic resonance imaging and ultrasonography findings after neonatal hypoglycemia. Pediatrics, 103(4 Pt 1), 724-729. doi:10.1542/peds.103.4.724
Kirkham, F. J., Wade, A. M., McElduff, F., Boyd, S. G., Tasker, R. C., Edwards, M., Neville, B. G., Peshu, N., & Newton, C. R. (2012). Seizures in 204 comatose children: incidence and outcome. Intensive Care Med, 38(5), 853-862. doi:10.1007/s00134-012-2529-9
Knobloch, H., Sotos, J. F., Sherard, E. S., Jr., Hodson, W. A., & Wehe, R. A. (1967). Prognostic and etiologic factors in hypoglycemia. J Pediatr, 70(6), 876-884. doi:10.1016/s0022-3476(67)80260-9
Koh, T. H., Aynsley-Green, A., Tarbit, M., & Eyre, J. A. (1988). Neural dysfunction during hypoglycaemia. Arch Dis Child, 63(11), 1353-1358.
Koivisto, M., Blanco-Sequeiros, M., & Krause, U. (1972). Neonatal symptomatic and asymptomatic hypoglycaemia: a follow-up study of 151 children. Developmental Medicine & Child Neurology, 14(5), 603-614.
Kontio, T., Toet, M. C., Hellstrom-Westas, L., van Handel, M., Groenendaal, F., Stjerna, S., Vanhatalo, S., & de Vries, L. S. (2013). Early neurophysiology and MRI in predicting neurological outcome at 9-10 years after birth asphyxia. Clin Neurophysiol, 124(6), 1089-1094. doi:10.1016/j.clinph.2012.12.045
Korotchikova, I., Stevenson, N. J., Livingstone, V., Ryan, C. A., & Boylan, G. B. (2016). Sleep-wake cycle of the healthy term newborn infant in the immediate postnatal period. Clin Neurophysiol, 127(4), 2095-2101. doi:10.1016/j.clinph.2015.12.015
133
Korotchikova, I., Stevenson, N. J., Walsh, B. H., Murray, D. M., & Boylan, G. B. (2011). Quantitative EEG analysis in neonatal hypoxic ischaemic encephalopathy. Clin Neurophysiol, 122(8), 1671-1678. doi:10.1016/j.clinph.2010.12.059
Kovatchev, B. P., Clarke, W. L., Breton, M., Brayman, K., & McCall, A. (2005). Quantifying temporal glucose variability in diabetes via continuous glucose monitoring: mathematical methods and clinical application. Diabetes Technology & Therapeutics, 7(6), 849-862. doi:10.1089/dia.2005.7.849
Kovatchev, B. P., Shields, D., & Breton, M. (2009). Graphical and numerical evaluation of continuous glucose sensing time lag. Diabetes Technology & Therapeutics, 11(3), 139-143. doi:10.1089/dia.2008.0044
Krinsley, J. S. (2008). Glycemic variability: a strong independent predictor of mortality in critically ill patients. Crit Care Med, 36(11), 3008-3013. doi:10.1097/CCM.0b013e31818b38d2
Kumaran, A., Kar, S., Kapoor, R. R., & Hussain, K. (2010). The clinical problem of hyperinsulinemic hypoglycemia and resultant infantile spasms. Pediatrics, 126(5), e1231-1236. doi:10.1542/peds.2009-2775
Kurinczuk, J. J., White-Koning, M., & Badawi, N. (2010). Epidemiology of neonatal encephalopathy and hypoxic-ischaemic encephalopathy. Early Hum Dev, 86(6), 329-338. doi:10.1016/j.earlhumdev.2010.05.010
Lahorgue Nunes, M., Magdalena Penela, M., & Costa Da Costa, J. (2000). Differences in the dynamics of frontal sharp transients in normal and hypoglycemic newborns. Clin Neurophysiol, 111(2), 305-310. doi:http://dx.doi.org/10.1016/S1388-2457%2899%2900234-5
Lam, A. M., Winn, H. R., Cullen, B. F., & Sundling, N. (1991). Hyperglycemia and neurological outcome in patients with head injury. J Neurosurg, 75(4), 545-551. doi:10.3171/jns.1991.75.4.0545
Lambrechtsen, F. A., & Buchhalter, J. R. (2008). Aborted and refractory status epilepticus in children: a comparative analysis. Epilepsia, 49(4), 615-625. doi:10.1111/j.1528-1167.2007.01465.x
Laptook, A. R., Corbett, R. J., Arencibia-Mireles, O., & Ruley, J. (1992). Glucose-associated alterations in ischemic brain metabolism of neonatal piglets. Stroke, 23(10), 1504-1511. doi:10.1161/01.str.23.10.1504
Laroia, N., Guillet, R., Burchfiel, J., & McBride, M. C. (1998). EEG background as predictor of electrographic seizures in high-risk neonates. Epilepsia, 39(5), 545-551. doi:10.1111/j.1528-1157.1998.tb01418.x
Lawrence, R., Mathur, A., Nguyen The Tich, S., Zempel, J., & Inder, T. (2009). A pilot study of continuous limited-channel aEEG in term infants with encephalopathy. J Pediatr, 154(6), 835-841 e831. doi:10.1016/j.jpeds.2009.01.002
134
Le Floch, J. P., & Kessler, L. (2016). Glucose Variability: Comparison of Different Indices During Continuous Glucose Monitoring in Diabetic Patients. Journal of Diabetes Science & Technology, 10(4), 885-891. doi:10.1177/1932296816632003
Lear, C. A., Davidson, J. O., Mackay, G. R., Drury, P. P., Galinsky, R., Quaedackers, J. S., Gunn, A. J., & Bennet, L. (2017). Antenatal dexamethasone before asphyxia promotes cystic neural injury in preterm fetal sheep by inducing hyperglycemia. J Cereb Blood Flow Metab, 271678X17703124. doi:10.1177/0271678X17703124
Leblanc, M. H., Huang, M., Patel, D., Smith, E. E., & Devidas, M. (1994). Glucose given after hypoxic ischemia does not affect brain injury in piglets. Stroke, 25(7), 1443-1447.
LeBlanc, M. H., Huang, M., Vig, V., Patel, D., & Smith, E. E. (1993). Glucose affects the severity of hypoxic-ischemic brain injury in newborn pigs. Stroke, 24(7), 1055-1062. doi:10.1161/01.str.24.7.1055
Leelarathna, L., Dellweg, S., Mader, J. K., Allen, J. M., Benesch, C., Doll, W., Ellmerer, M., Hartnell, S., Heinemann, L., Kojzar, H., Michalewski, L., Nodale, M., Thabit, H., Wilinska, M. E., Pieber, T. R., Arnolds, S., Evans, M. L., Hovorka, R., & Consortium, A. P. h. (2014). Day and night home closed-loop insulin delivery in adults with type 1 diabetes: three-center randomized crossover study. Diabetes Care, 37(7), 1931-1937. doi:10.2337/dc13-2911
Leijser, L. M., Vein, A. A., Liauw, L., Strauss, T., Veen, S., & Wezel-Meijler, G. (2007). Prediction of short-term neurological outcome in full-term neonates with hypoxic-ischaemic encephalopathy based on combined use of electroencephalogram and neuro-imaging. Neuropediatrics, 38(5), 219-227. doi:10.1055/s-2007-992815
Leroy-Terquem, E., Vermersch, A. I., Dean, P., Assaf, Z., Boddaert, N., Lapillonne, A., & Magny, J. F. (2017). Abnormal Interhemispheric Synchrony in Neonatal Hypoxic-Ischemic Encephalopathy: A Retrospective Pilot Study. Neonatology, 112(4), 359-364. doi:10.1159/000478964
Lewis, L. D., Ljunggren, B., Norberg, K., & Siesjo, B. K. (1974). Changes in carbohydrate substrates, amino acids and ammonia in the brain during insulin-induced hypoglycemia. J Neurochem, 23(4), 659-671. doi:10.1111/j.1471-4159.1974.tb04389.x
Lewis, L. D., Ljunggren, B., Ratcheson, R. A., & Siesjo, B. K. (1974). Cerebral energy state in insulin-induced hypoglycemia, related to blood glucose and to EEG. J Neurochem, 23(4), 673-679. doi:10.1111/j.1471-4159.1974.tb04390.x
Lewis, M., & Bendersky, M. (1989). Cognitive and motor differences among low birth weight infants: impact of intraventricular hemorrhage, medical risk, and social class. Pediatrics, 83(2), 187-192.
Lilien, L. D., Pildes, R. S., Srinivasan, G., Voora, S., & Yeh, T. F. (1980). Treatment of neonatal hypoglycemia with minibolus and intravenous glucose infusion. The Journal of Pediatrics, 97(2), 295-298. doi:10.1016/s0022-3476(80)80499-9
135
Lindstrom, K., Lagerroos, P., Gillberg, C., & Fernell, E. (2006). Teenage outcome after being born at term with moderate neonatal encephalopathy. Pediatric Neurology, 35(4), 268-274. doi:10.1016/j.pediatrneurol.2006.05.003
Lingappan, K., Kaiser, J. R., Srinivasan, C., & Gunn, A. J. (2016). Relationship between PCO2 and unfavorable outcome in infants with moderate-to-severe hypoxic ischemic encephalopathy. Pediatr Res, 80(2), 204-208. doi:10.1038/pr.2016.62
Lofhede, J., Thordstein, M., Lofgren, N., Flisberg, A., Rosa-Zurera, M., Kjellmer, I., & Lindecrantz, K. (2010). Automatic classification of background EEG activity in healthy and sick neonates. Journal of Neural Engineering, 7(1), 16007. doi:10.1088/1741-2560/7/1/016007
Lorek, A., Takei, Y., Cady, E. B., Wyatt, J. S., Penrice, J., Edwards, A. D., Peebles, D., Wylezinska, M., Owen-Reece, H., Kirkbride, V., & et al. (1994). Delayed ("secondary") cerebral energy failure after acute hypoxia-ischemia in the newborn piglet: continuous 48-hour studies by phosphorus magnetic resonance spectroscopy. Pediatr Res, 36(6), 699-706.
Lucas, A., & Morley, R. (1999). Authors' reply to Outcome of neonatal hypoglycaemia. Complete data are needed. BMJ, 318(7177), 194-195. doi:10.1136/bmj.318.7177.194a
Lucas, A., Morley, R., & Cole, T. J. (1988). Adverse neurodevelopmental outcome of moderate neonatal hypoglycaemia. BMJ, 297(6659), 1304-1308.
Lundelin, K., Vigil, L., Bua, S., Gomez-Mestre, I., Honrubia, T., & Varela, M. (2010). Differences in complexity of glycemic profile in survivors and nonsurvivors in an intensive care unit: a pilot study. Crit Care Med, 38(3), 849-854. doi:10.1097/CCM.0b013e3181ce49cf
Lynch, N. E., Stevenson, N. J., Livingstone, V., Mathieson, S., Murphy, B. P., Rennie, J. M., & Boylan, G. B. (2015). The temporal characteristics of seizures in neonatal hypoxic ischemic encephalopathy treated with hypothermia. Seizure, 33, 60-65. doi:10.1016/j.seizure.2015.10.007
MacDougall, N. J., & Muir, K. W. (2011). Hyperglycaemia and infarct size in animal models of middle cerebral artery occlusion: systematic review and meta-analysis. J Cereb Blood Flow Metab, 31(3), 807-818. doi:10.1038/jcbfm.2010.210
Macrae, D., Grieve, R., Allen, E., Sadique, Z., Morris, K., Pappachan, J., Parslow, R., Tasker, R. C., Elbourne, D., & Investigators, C. H. (2014). A randomized trial of hyperglycemic control in pediatric intensive care. New England Journal of Medicine, 370(2), 107-118. doi:10.1056/NEJMoa1302564
Mahajan, G., Mukhopadhyay, K., Attri, S., & Kumar, P. (2017). Neurodevelopmental Outcome of Asymptomatic Hypoglycemia Compared With Symptomatic Hypoglycemia and Euglycemia in High-Risk Neonates. Pediatric Neurology, 74, 74-79. doi:10.1016/j.pediatrneurol.2017.05.028
136
Malone, A., Ryan, C. A., Fitzgerald, A., Burgoyne, L., Connolly, S., & Boylan, G. B. (2009). Interobserver agreement in neonatal seizure identification. Epilepsia, 50(9), 2097-2101. doi:10.1111/j.1528-1167.2009.02132.x
Manzoni, P., Castagnola, E., Mostert, M., Sala, U., Galletto, P., & Gomirato, G. (2006). Hyperglycaemia as a possible marker of invasive fungal infection in preterm neonates. Acta Paediatr, 95(4), 486-493. doi:10.1080/08035250500444867
Martinello, K., Hart, A. R., Yap, S., Mitra, S., & Robertson, N. J. (2017). Management and investigation of neonatal encephalopathy: 2017 update. Arch Dis Child Fetal Neonatal Ed, 102(4), F346-F358. doi:10.1136/archdischild-2015-309639
Massaro, A. N., Govindan, R. B., Vezina, G., Chang, T., Andescavage, N. N., Wang, Y., Al-Shargabi, T., Metzler, M., Harris, K., & du Plessis, A. J. (2015). Impaired cerebral autoregulation and brain injury in newborns with hypoxic-ischemic encephalopathy treated with hypothermia. J Neurophysiol, 114(2), 818-824. doi:10.1152/jn.00353.2015
Massaro, A. N., Murthy, K., Zaniletti, I., Cook, N., DiGeronimo, R., Dizon, M., Hamrick, S. E., McKay, V. J., Natarajan, G., Rao, R., Smith, D., Telesco, R., Wadhawan, R., Asselin, J. M., Durand, D. J., Evans, J. R., Dykes, F., Reber, K. M., Padula, M. A., Pallotto, E. K., Short, B. L., & Mathur, A. M. (2015). Short-term outcomes after perinatal hypoxic ischemic encephalopathy: a report from the Children's Hospitals Neonatal Consortium HIE focus group. J Perinatol, 35(4), 290-296. doi:10.1038/jp.2014.190
Massaro, A. N., Tsuchida, T., Kadom, N., El-Dib, M., Glass, P., Baumgart, S., & Chang, T. (2012). aEEG evolution during therapeutic hypothermia and prediction of NICU outcome in encephalopathic neonates. Neonatology, 102(3), 197-202. doi:10.1159/000339570
Massey, S. L., Jensen, F. E., & Abend, N. S. (2018). Electroencephalographic monitoring for seizure identification and prognosis in term neonates. Semin Fetal Neonatal Med. doi:10.1016/j.siny.2018.01.001
Mastrangelo, M., Fiocchi, I., Fontana, P., Gorgone, G., Lista, G., & Belcastro, V. (2013). Acute neonatal encephalopathy and seizures recurrence: a combined aEEG/EEG study. Seizure, 22(9), 703-707. doi:10.1016/j.seizure.2013.05.006
Matic, V., Cherian, P. J., Koolen, N., Ansari, A. H., Naulaers, G., Govaert, P., Van Huffel, S., De Vos, M., & Vanhatalo, S. (2015). Objective differentiation of neonatal EEG background grades using detrended fluctuation analysis. Front Hum Neurosci, 9, 189. doi:10.3389/fnhum.2015.00189
McBride, M. C., Laroia, N., & Guillet, R. (2000). Electrographic seizures in neonates correlate with poor neurodevelopmental outcome. Neurology, 55(4), 506-513.
McCoy, B., & Hahn, C. D. (2013). Continuous EEG monitoring in the neonatal intensive care unit. J Clin Neurophysiol, 30(2), 106-114. doi:10.1097/WNP.0b013e3182872919
McKinlay, C. J., Alsweiler, J. M., Ansell, J. M., Anstice, N. S., Chase, J. G., Gamble, G. D., Harris, D. L., Jacobs, R. J., Jiang, Y., Paudel, N., Signal, M., Thompson, B., Wouldes, T.
137
A., Yu, T. Y., Harding, J. E., & Group, C. S. (2015). Neonatal Glycemia and Neurodevelopmental Outcomes at 2 Years. New England Journal of Medicine, 373(16), 1507-1518. doi:10.1056/NEJMoa1504909
McKinlay, C. J. D., Alsweiler, J. M., Anstice, N. S., Burakevych, N., Chakraborty, A., Chase, J. G., Gamble, G. D., Harris, D. L., Jacobs, R. J., Jiang, Y., Paudel, N., San Diego, R. J., Thompson, B., Wouldes, T. A., Harding, J. E., Children With, H., & Their Later Development Study, T. (2017). Association of Neonatal Glycemia With Neurodevelopmental Outcomes at 4.5 Years. JAMA Pediatr, 171(10), 972-983. doi:10.1001/jamapediatrics.2017.1579
McKinlay, C. J. D., Chase, J. G., Dickson, J., Harris, D. L., Alsweiler, J. M., & Harding, J. E. (2017). Continuous glucose monitoring in neonates: a review. Maternal Health, Neonatology and Perinatology, 3, 18. doi:10.1186/s40748-017-0055-z
Meetze, W., Bowsher, R., Compton, J., & Moorehead, H. (1998). Hyperglycemia in extremely- low-birth-weight infants. Biol Neonate, 74(3), 214-221. doi:10.1159/000014027
Menache, C. C., Bourgeois, B. F., & Volpe, J. J. (2002). Prognostic value of neonatal discontinuous EEG. Pediatric Neurology, 27(2), 93-101.
Mesotten, D., Gielen, M., Sterken, C., Claessens, K., Hermans, G., Vlasselaers, D., Lemiere, J., Lagae, L., Gewillig, M., Eyskens, B., Vanhorebeek, I., Wouters, P. J., & Van den Berghe, G. (2012). Neurocognitive development of children 4 years after critical illness and treatment with tight glucose control: a randomized controlled trial. JAMA, 308(16), 1641-1650. doi:10.1001/jama.2012.12424
Miller, S. P., Weiss, J., Barnwell, A., Ferriero, D. M., Latal-Hajnal, B., Ferrer-Rogers, A., Newton, N., Partridge, J. C., Glidden, D. V., Vigneron, D. B., & Barkovich, A. J. (2002). Seizure-associated brain injury in term newborns with perinatal asphyxia. Neurology, 58(4), 542-548.
Miralles, R. E., Lodha, A., Perlman, M., & Moore, A. M. (2002). Experience With Intravenous Glucagon Infusions as a Treatment for Resistant Neonatal Hypoglycemia. Archives of Pediatrics & Adolescent Medicine, 156(10), 999. doi:10.1001/archpedi.156.10.999
Mitanchez, D. (2007). Glucose regulation in preterm newborn infants. Horm Res, 68(6), 265-271. doi:10.1159/000104174
Mitrakou, A., Ryan, C., Veneman, T., Mokan, M., Jenssen, T., Kiss, I., Durrant, J., Cryer, P., & Gerich, J. (1991). Hierarchy of glycemic thresholds for counterregulatory hormone secretion, symptoms, and cerebral dysfunction. Am J Physiol, 260(1 Pt 1), E67-74. doi:10.1152/ajpendo.1991.260.1.E67
Mohamed, S., Murray, J. C., Dagle, J. M., & Colaizy, T. (2013). Hyperglycemia as a risk factor for the development of retinopathy of prematurity. BMC Pediatr, 13, 78. doi:10.1186/1471-2431-13-78
138
Mohsen, L., Abou-Alam, M., El-Dib, M., Labib, M., Elsada, M., & Aly, H. (2014). A prospective study on hyperglycemia and retinopathy of prematurity. J Perinatol, 34(6), 453-457. doi:10.1038/jp.2014.49
Mola-Schenzle, E., Staffler, A., Klemme, M., Pellegrini, F., Molinaro, G., Parhofer, K. G., Messner, H., Schulze, A., & Flemmer, A. W. (2015). Clinically stable very low birthweight infants are at risk for recurrent tissue glucose fluctuations even after fully established enteral nutrition. Arch Dis Child Fetal Neonatal Ed, 100(2), F126-131. doi:10.1136/archdischild-2014-306168
Montassir, H., Maegaki, Y., Ogura, K., Kurozawa, Y., Nagata, I., Kanzaki, S., & Ohno, K. (2009). Associated factors in neonatal hypoglycemic brain injury. Brain & Development, 31(9), 649-656. doi:10.1016/j.braindev.2008.10.012
Montassir, H., Maegaki, Y., Ohno, K., & Ogura, K. (2010). Long term prognosis of symptomatic occipital lobe epilepsy secondary to neonatal hypoglycemia. Epilepsy Research, 88(2-3), 93-99. doi:10.1016/j.eplepsyres.2009.10.001
Mujsce, D. J., Christensen, M. A., & Vannucci, R. C. (1989). Regional cerebral blood flow and glucose utilization during hypoglycemia in newborn dogs. Am J Physiol, 256(6 Pt 2), H1659-1666. doi:10.1152/ajpheart.1989.256.6.H1659
Murakami, Y., Yamashita, Y., Matsuishi, T., Utsunomiya, H., Okudera, T., & Hashimoto, T. (1999). Cranial MRI of neurologically impaired children suffering from neonatal hypoglycaemia. Pediatr Radiol, 29(1), 23-27. doi:10.1007/s002470050527
Murray, D. M., Boylan, G. B., Ryan, C. A., & Connolly, S. (2009). Early EEG findings in hypoxic-ischemic encephalopathy predict outcomes at 2 years. Pediatrics, 124(3), e459-467. doi:10.1542/peds.2008-2190
Murray, D. M., O'Connor, C. M., Ryan, C. A., Korotchikova, I., & Boylan, G. B. (2016). Early EEG Grade and Outcome at 5 Years After Mild Neonatal Hypoxic Ischemic Encephalopathy. Pediatrics, 138(4). doi:10.1542/peds.2016-0659
Nadeem, M., Murray, D. M., Boylan, G. B., Dempsey, E. M., & Ryan, C. A. (2011). Early blood glucose profile and neurodevelopmental outcome at two years in neonatal hypoxic-ischaemic encephalopathy. BMC Pediatr, 11, 10. doi:10.1186/1471-2431-11-10
Naim, M. Y., Gaynor, J. W., Chen, J., Nicolson, S. C., Fuller, S., Spray, T. L., Dlugos, D. J., Clancy, R. R., Costa, L. V., Licht, D. J., Xiao, R., Meldrum, H., & Abend, N. S. (2015). Subclinical seizures identified by postoperative electroencephalographic monitoring are common after neonatal cardiac surgery. J Thorac Cardiovasc Surg, 150(1), 169-178; discussion 178-180. doi:10.1016/j.jtcvs.2015.03.045
Nash, K. B., Bonifacio, S. L., Glass, H. C., Sullivan, J. E., Barkovich, A. J., Ferriero, D. M., & Cilio, M. R. (2011). Video-EEG monitoring in newborns with hypoxic-ischemic encephalopathy treated with hypothermia. Neurology, 76(6), 556-562. doi:10.1212/WNL.0b013e31820af91a
139
Nehlig, A., de Vasconcelos, A. P., & Boyet, S. (1988). Quantitative autoradiographic measurement of local cerebral glucose utilization in freely moving rats during postnatal development. Journal of Neuroscience, 8(7), 2321-2333.
Nevalainen, P., Marchi, V., Metsaranta, M., Lonnqvist, T., Toiviainen-Salo, S., Vanhatalo, S., & Lauronen, L. (2017). Evoked potentials recorded during routine EEG predict outcome after perinatal asphyxia. Clin Neurophysiol, 128(7), 1337-1343. doi:10.1016/j.clinph.2017.04.025
Niemarkt, H. J., Andriessen, P., Peters, C. H., Pasman, J. W., Zimmermann, L. J., & Bambang Oetomo, S. (2010). Quantitative analysis of maturational changes in EEG background activity in very preterm infants with a normal neurodevelopment at 1 year of age. Early Hum Dev, 86(4), 219-224. doi:10.1016/j.earlhumdev.2010.03.003
Niemarkt, H. J., Jennekens, W., Pasman, J. W., Katgert, T., Van Pul, C., Gavilanes, A. W., Kramer, B. W., Zimmermann, L. J., Bambang Oetomo, S., & Andriessen, P. (2011). Maturational changes in automated EEG spectral power analysis in preterm infants. Pediatr Res, 70(5), 529-534. doi:10.1203/PDR.0b013e31822d748b
Norberg, K., & Siesio, B. K. (1976). Oxidative metabolism of the cerebral cortex of the rat in severe insulin-induced hypoglycaemia. J Neurochem, 26(2), 345-352. doi:10.1111/j.1471-4159.1976.tb04487.x
O'Neill, B. R., Handler, M. H., Tong, S., & Chapman, K. E. (2015). Incidence of seizures on continuous EEG monitoring following traumatic brain injury in children. J Neurosurg Pediatr, 16(2), 167-176. doi:10.3171/2014.12.PEDS14263
O'Toole, J. M., Boylan, G. B., Vanhatalo, S., & Stevenson, N. J. (2016). Estimating functional brain maturity in very and extremely preterm neonates using automated analysis of the electroencephalogram. Clin Neurophysiol, 127(8), 2910-2918. doi:10.1016/j.clinph.2016.02.024
Obeid, R., Sogawa, Y., Gedela, S., Naik, M., Lee, V., Telesco, R., Wisnowski, J., Magill, C., Painter, M. J., & Panigrahy, A. (2017). The Correlation Between a Short-term Conventional Electroencephalography in the First Day of Life and Brain Magnetic Resonance Imaging in Newborns Undergoing Hypothermia for Hypoxic-Ischemic Encephalopathy. Pediatric Neurology, 67, 91-97. doi:10.1016/j.pediatrneurol.2016.10.020
Oddo, M., Schmidt, J. M., Carrera, E., Badjatia, N., Connolly, E. S., Presciutti, M., Ostapkovich, N. D., Levine, J. M., Le Roux, P., & Mayer, S. A. (2008). Impact of tight glycemic control on cerebral glucose metabolism after severe brain injury: a microdialysis study. Crit Care Med, 36(12), 3233-3238. doi:10.1097/CCM.0b013e31818f4026
Oh, S. H., Park, K. N., Kim, Y. M., Kim, H. J., Youn, C. S., Kim, S. H., Choi, S. P., Kim, S. C., & Shon, Y. M. (2013). The prognostic value of continuous amplitude-integrated electroencephalogram applied immediately after return of spontaneous circulation in therapeutic hypothermia-treated cardiac arrest patients. Resuscitation, 84(2), 200-205. doi:10.1016/j.resuscitation.2012.09.031
140
Oh, S. H., Park, K. N., Shon, Y. M., Kim, Y. M., Kim, H. J., Youn, C. S., Kim, S. H., Choi, S. P., & Kim, S. C. (2015). Continuous Amplitude-Integrated Electroencephalographic Monitoring Is a Useful Prognostic Tool for Hypothermia-Treated Cardiac Arrest Patients. Circulation, 132(12), 1094-1103. doi:10.1161/CIRCULATIONAHA.115.015754
Olejniczak, P. (2006). Neurophysiologic basis of EEG. J Clin Neurophysiol, 23(3), 186-189. doi:10.1097/01.wnp.0000220079.61973.6c
Osredkar, D., Toet, M. C., van Rooij, L. G., van Huffelen, A. C., Groenendaal, F., & de Vries, L. S. (2005). Sleep-wake cycling on amplitude-integrated electroencephalography in term newborns with hypoxic-ischemic encephalopathy. Pediatrics, 115(2), 327-332. doi:10.1542/peds.2004-0863
Pappas, A., Shankaran, S., McDonald, S. A., Vohr, B. R., Hintz, S. R., Ehrenkranz, R. A., Tyson, J. E., Yolton, K., Das, A., Bara, R., Hammond, J., Higgins, R. D., & Hypothermia Extended Follow-up Subcommittee of the Eunice Kennedy Shriver, N. N. R. N. (2015). Cognitive outcomes after neonatal encephalopathy. Pediatrics, 135(3), e624-634. doi:10.1542/peds.2014-1566
Park, W. S., Chang, Y. S., & Lee, M. (2001). Effects of hyperglycemia or hypoglycemia on brain cell membrane function and energy metabolism during the immediate reoxygenation-reperfusion period after acute transient global hypoxia-ischemia in the newborn piglet. Brain Res, 901(1-2), 102-108.
Passero, S., Ciacci, G., & Ulivelli, M. (2003). The influence of diabetes and hyperglycemia on clinical course after intracerebral hemorrhage. Neurology, 61(10), 1351-1356. doi:10.1212/01.wnl.0000094326.30791.2d
Pavone, L., Mollica, F., Musumeci, S., Marino, S., & Pampiglione, G. (1980). Accidental glibenclamide ingestion in an infant: clinical and electroencephalographic aspects. Developmental Medicine & Child Neurology, 22(3), 366-371.
Payne, E. T., Zhao, X. Y., Frndova, H., McBain, K., Sharma, R., Hutchison, J. S., & Hahn, C. D. (2014). Seizure burden is independently associated with short term outcome in critically ill children. Brain, 137(Pt 5), 1429-1438. doi:10.1093/brain/awu042
Peeters-Scholte, C., Braun, K., Koster, J., Kops, N., Blomgren, K., Buonocore, G., van Buul-Offers, S., Hagberg, H., Nicolay, K., van Bel, F., & Groenendaal, F. (2003). Effects of allopurinol and deferoxamine on reperfusion injury of the brain in newborn piglets after neonatal hypoxia-ischemia. Pediatr Res, 54(4), 516-522. doi:10.1203/01.PDR.0000081297.53793.C6
Per, H., Kumandas, S., Coskun, A., Gumus, H., & Oztop, D. (2008). Neurologic sequelae of neonatal hypoglycemia in Kayseri, Turkey. J Child Neurol, 23(12), 1406-1412. doi:10.1177/0883073808319075
Perez, A., Ritter, S., Brotschi, B., Werner, H., Caflisch, J., Martin, E., & Latal, B. (2013). Long-term neurodevelopmental outcome with hypoxic-ischemic encephalopathy. J Pediatr, 163(2), 454-459. doi:10.1016/j.jpeds.2013.02.003
141
Petersson, K. H., Pinar, H., Stopa, E. G., Sadowska, G. B., Hanumara, R. C., & Stonestreet, B. S. (2004). Effects of exogenous glucose on brain ischemia in ovine fetuses. Pediatr Res, 56(4), 621-629. doi:10.1203/01.PDR.0000139415.96985.BF
Piantino, J. A., Wainwright, M. S., Grimason, M., Smith, C. M., Hussain, E., Byron, D., Chin, A., Backer, C., Reynolds, M., & Goldstein, J. (2013). Nonconvulsive seizures are common in children treated with extracorporeal cardiac life support. Pediatr Crit Care Med, 14(6), 601-609. doi:10.1097/PCC.0b013e318291755a
Pidcoke, H. F., Wanek, S. M., Rohleder, L. S., Holcomb, J. B., Wolf, S. E., & Wade, C. E. (2009). Glucose variability is associated with high mortality after severe burn. J Trauma, 67(5), 990-995. doi:10.1097/TA.0b013e3181baef4b
Pildes, R., Cornblath, M., Warren, I., Page-El, E., Di Menza, S., Merritt, D. M., & Peeva, A. (1974). A prospective controlled study of neonatal hypoglycemia. Pediatrics, 54(1), 5-14.
Pildes, R., Forbes, A. E., O'Connor, S. M., & Cornblath, M. (1967). The incidence of neonatal hypoglycemia--a completed survey. J Pediatr, 70(1), 76-80. doi:10.1016/s0022-3476(67)80168-9
Piper, H. G., Alexander, J. L., Shukla, A., Pigula, F., Costello, J. M., Laussen, P. C., Jaksic, T., & Agus, M. S. (2006). Real-time continuous glucose monitoring in pediatric patients during and after cardiac surgery. Pediatrics, 118(3), 1176-1184. doi:10.1542/peds.2006-0347
Pisani, F., Cerminara, C., Fusco, C., & Sisti, L. (2007). Neonatal status epilepticus vs recurrent neonatal seizures: clinical findings and outcome. Neurology, 69(23), 2177-2185. doi:10.1212/01.wnl.0000295674.34193.9e
Pisani, F., Copioli, C., Di Gioia, C., Turco, E., & Sisti, L. (2008). Neonatal seizures: relation of ictal video-electroencephalography (EEG) findings with neurodevelopmental outcome. J Child Neurol, 23(4), 394-398. doi:10.1177/0883073807309253
Prempunpong, C., Chalak, L. F., Garfinkle, J., Shah, B., Kalra, V., Rollins, N., Boyle, R., Nguyen, K. A., Mir, I., Pappas, A., Montaldo, P., Thayyil, S., Sanchez, P. J., Shankaran, S., Laptook, A. R., & Sant'Anna, G. (2018). Prospective research on infants with mild encephalopathy: the PRIME study. J Perinatol, 38(1), 80-85. doi:10.1038/jp.2017.164
Pressler, R. M., Boylan, G. B., Morton, M., Binnie, C. D., & Rennie, J. M. (2001). Early serial EEG in hypoxic ischaemic encephalopathy. Clin Neurophysiol, 112(1), 31-37.
Pryds, O., Greisen, G., & Friis-Hansen, B. (1988). Compensatory increase of CBF in preterm infants during hypoglycaemia. Acta Paediatrica Scandinavica, 77(5), 632-637.
Quagliaro, L., Piconi, L., Assaloni, R., Martinelli, L., Motz, E., & Ceriello, A. (2003). Intermittent high glucose enhances apoptosis related to oxidative stress in human umbilical vein endothelial cells: the role of protein kinase C and NAD(P)H-oxidase activation. Diabetes, 52(11), 2795-2804.
142
Rachmiel, M., Cohen, M., Heymen, E., Lezinger, M., Inbar, D., Gilat, S., Bistritzer, T., Leshem, G., Kan-Dror, E., Lahat, E., & Ekstein, D. (2016). Hyperglycemia is associated with simultaneous alterations in electrical brain activity in youths with type 1 diabetes mellitus. Clin Neurophysiol, 127(2), 1188-1195. doi:10.1016/j.clinph.2015.07.011
Rake, A. J., Srinivasan, V., Nadkarni, V., Kaptan, R., & Newth, C. J. (2010). Glucose variability and survival in critically ill children: allostasis or harm? Pediatr Crit Care Med, 11(6), 707-712. doi:10.1097/PCC.0b013e3181e88b1f
Rakshasbhuvankar, A., Paul, S., Nagarajan, L., Ghosh, S., & Rao, S. (2015). Amplitude-integrated EEG for detection of neonatal seizures: a systematic review. Seizure, 33, 90-98. doi:10.1016/j.seizure.2015.09.014
Randall, G. C. (1979). Studies on the effect of acute asphyxia on the fetal pig in utero. Biol Neonate, 36(1-2), 63-69. doi:10.1159/000241208
Rasanen, O., Metsaranta, M., & Vanhatalo, S. (2013). Development of a novel robust measure for interhemispheric synchrony in the neonatal EEG: activation synchrony index (ASI). Neuroimage, 69, 256-266. doi:10.1016/j.neuroimage.2012.12.017
Rebrin, K., Steil, G. M., van Antwerp, W. P., & Mastrototaro, J. J. (1999). Subcutaneous glucose predicts plasma glucose independent of insulin: implications for continuous monitoring. Am J Physiol, 277(3 Pt 1), E561-571.
Robertson, C. M., & Finer, N. N. (1988). Educational readiness of survivors of neonatal encephalopathy associated with birth asphyxia at term. J Dev Behav Pediatr, 9, 298-306.
Robertson, C. M., Finer, N. N., & Grace, M. G. (1989). School performance of survivors of neonatal encephalopathy associated with birth asphyxia at term. J Pediatr, 114(5), 753-760.
Robertson, N. J., Faulkner, S., Fleiss, B., Bainbridge, A., Andorka, C., Price, D., Powell, E., Lecky-Thompson, L., Thei, L., Chandrasekaran, M., Hristova, M., Cady, E. B., Gressens, P., Golay, X., & Raivich, G. (2013). Melatonin augments hypothermic neuroprotection in a perinatal asphyxia model. Brain, 136(Pt 1), 90-105. doi:10.1093/brain/aws285
Rodbard, D. (2009a). Interpretation of continuous glucose monitoring data: glycemic variability and quality of glycemic control. Diabetes Technology & Therapeutics, 11 Suppl 1, S55-67. doi:10.1089/dia.2008.0132
Rodbard, D. (2009b). New and improved methods to characterize glycemic variability using continuous glucose monitoring. Diabetes Technology & Therapeutics, 11(9), 551-565. doi:10.1089/dia.2009.0015
Rosenberg, A. A., & Murdaugh, E. (1990). The effect of blood glucose concentration on postasphyxia cerebral hemodynamics in newborn lambs. Pediatr Res, 27(5), 454-459. doi:10.1203/00006450-199005000-00008
143
Sadhwani, A., Asaro, L. A., Goldberg, C., Ware, J., Butcher, J., Gaies, M., Smith, C., Alexander, J. L., Wypij, D., & Agus, M. S. (2016). Impact of Tight Glycemic Control on Neurodevelopmental Outcomes at 1 Year of Age for Children with Congenital Heart Disease: A Randomized Controlled Trial. J Pediatr, 174, 193-198 e192. doi:10.1016/j.jpeds.2016.03.048
Saisho, Y., Tanaka, C., Tanaka, K., Roberts, R., Abe, T., Tanaka, M., Meguro, S., Irie, J., Kawai, T., & Itoh, H. (2015). Relationships among different glycemic variability indices obtained by continuous glucose monitoring. Prim Care Diabetes, 9(4), 290-296. doi:10.1016/j.pcd.2014.10.001
Salhab, W. A., Wyckoff, M. H., Laptook, A. R., & Perlman, J. M. (2004). Initial hypoglycemia and neonatal brain injury in term infants with severe fetal acidemia. Pediatrics, 114(2), 361-366.
Sanchez Fernandez, I., Abend, N. S., Arndt, D. H., Carpenter, J. L., Chapman, K. E., Cornett, K. M., Dlugos, D. J., Gallentine, W. B., Giza, C. C., Goldstein, J. L., Hahn, C. D., Lerner, J. T., Matsumoto, J. H., McBain, K., Nash, K. B., Payne, E., Sanchez, S. M., Williams, K., & Loddenkemper, T. (2014). Electrographic seizures after convulsive status epilepticus in children and young adults: a retrospective multicenter study. J Pediatr, 164(2), 339-346 e331-332. doi:10.1016/j.jpeds.2013.09.032
Sanchez, S. M., Arndt, D. H., Carpenter, J. L., Chapman, K. E., Cornett, K. M., Dlugos, D. J., Gallentine, W. B., Giza, C. C., Goldstein, J. L., Hahn, C. D., Lerner, J. T., Loddenkemper, T., Matsumoto, J. H., McBain, K., Nash, K. B., Payne, E., Sanchez Fernandez, I., Shults, J., Williams, K., Yang, A., & Abend, N. S. (2013). Electroencephalography monitoring in critically ill children: current practice and implications for future study design. Epilepsia, 54(8), 1419-1427. doi:10.1111/epi.12261
Sanchez, S. M., Carpenter, J., Chapman, K. E., Dlugos, D. J., Gallentine, W. B., Giza, C. C., Goldstein, J. L., Hahn, C. D., Kessler, S. K., Loddenkemper, T., Riviello, J. J., Jr., Abend, N. S., & Pediatric Critical Care, E. E. G. G. (2013). Pediatric ICU EEG monitoring: current resources and practice in the United States and Canada. J Clin Neurophysiol, 30(2), 156-160. doi:10.1097/WNP.0b013e31827eda27
Scher, M. S., Alvin, J., Gaus, L., Minnigh, B., & Painter, M. J. (2003). Uncoupling of EEG - clinical neonatal seizures after antiepileptic drug use. Pediatric Neurology, 28, 277 -280.
Scher, M. S., & Loparo, K. A. (2009). Neonatal EEG/sleep state analyses: a complex phenotype of developmental neural plasticity. Dev Neurosci, 31(4), 259-275. doi:10.1159/000216537
Schmelzeisen-Redeker, G., Schoemaker, M., Kirchsteiger, H., Freckmann, G., Heinemann, L., & Del Re, L. (2015). Time Delay of CGM Sensors: Relevance, Causes, and Countermeasures. Journal of Diabetes Science & Technology, 9(5), 1006-1015. doi:10.1177/1932296815590154
144
Schreiber, J. M., Zelleke, T., Gaillard, W. D., Kaulas, H., Dean, N., & Carpenter, J. L. (2012). Continuous video EEG for patients with acute encephalopathy in a pediatric intensive care unit. Neurocrit Care, 17(1), 31-38. doi:10.1007/s12028-012-9715-z
Schumacher, E. M., Larsson, P. G., Pripp, A. H., & Stiris, T. A. (2014). The effect of blood glucose and pCO2 on spectral EEG of premature infants during the first three days of life. Neonatology, 105(4), 297-305. doi:10.1159/000357291
Schwartz, N. S., Clutter, W. E., Shah, S. D., & Cryer, P. E. (1987). Glycemic thresholds for activation of glucose counterregulatory systems are higher than the threshold for symptoms. J Clin Invest, 79(3), 777-781. doi:10.1172/JCI112884
Sewell, E. K., Vezina, G., Chang, T., Tsuchida, T., Harris, K., Ridore, M., Glass, P., & Massaro, A. N. (2018). Evolution of Amplitude-Integrated Electroencephalogram as a Predictor of Outcome in Term Encephalopathic Neonates Receiving Therapeutic Hypothermia. Am J Perinatol, 35(3), 277-285. doi:10.1055/s-0037-1607212
Shah, D. K., Lavery, S., Doyle, L. W., Wong, C., McDougall, P., & Inder, T. E. (2006). Use of 2-channel bedside electroencephalogram monitoring in term-born encephalopathic infants related to cerebral injury defined by magnetic resonance imaging. Pediatrics, 118(1), 47-55. doi:10.1542/peds.2005-1294
Shah, D. K., Mackay, M. T., Lavery, S., Watson, S., Harvey, A. S., Zempel, J., Mathur, A., & Inder, T. E. (2008). Accuracy of bedside electroencephalographic monitoring in comparison with simultaneous continuous conventional electroencephalography for seizure detection in term infants. Pediatrics, 121(6), 1146-1154. doi:10.1542/peds.2007-1839
Shah, D. K., Wusthoff, C. J., Clarke, P., Wyatt, J. S., Ramaiah, S. M., Dias, R. J., Becher, J. C., Kapellou, O., & Boardman, J. P. (2014a). Electrographic seizures are associated with brain injury in newborns undergoing therapeutic hypothermia. Archives of Disease in Childhood: Fetal and Neonatal Edition, 99(3), F219-F224. doi:http://dx.doi.org/10.1136/archdischild-2013-305206
Shah, D. K., Wusthoff, C. J., Clarke, P., Wyatt, J. S., Ramaiah, S. M., Dias, R. J., Becher, J. C., Kapellou, O., & Boardman, J. P. (2014b). Electrographic seizures are associated with brain injury in newborns undergoing therapeutic hypothermia. Arch Dis Child Fetal Neonatal Ed, 99(3), F219-224. doi:10.1136/archdischild-2013-305206
Shah, N. A., & Wusthoff, C. J. (2015). How to use: amplitude-integrated EEG (aEEG). Arch Dis Child Educ Pract Ed, 100(2), 75-81. doi:10.1136/archdischild-2013-305676
Shah, R., McKinlay, C. J. D., & Harding, J. E. (2018). Neonatal hypoglycemia: continuous glucose monitoring. Current Opinion in Pediatrics, 30(2), 204-208. doi:10.1097/MOP.0000000000000592
Shahwan, A., Bailey, C., Shekerdemian, L., & Harvey, A. S. (2010). The prevalence of seizures in comatose children in the pediatric intensive care unit: a prospective video-EEG study. Epilepsia, 51(7), 1198-1204. doi:10.1111/j.1528-1167.2009.02517.x
145
Shankaran, S., Laptook, A. R., Ehrenkranz, R. A., Tyson, J. E., McDonald, S. A., Donovan, E. F., Fanaroff, A. A., Poole, W. K., Wright, L. L., Higgins, R. D., Finer, N. N., Carlo, W. A., Duara, S., Oh, W., Cotten, C. M., Stevenson, D. K., Stoll, B. J., Lemons, J. A., Guillet, R., Jobe, A. H., National Institute of Child, H., & Human Development Neonatal Research, N. (2005). Whole-body hypothermia for neonates with hypoxic-ischemic encephalopathy. New England Journal of Medicine, 353(15), 1574-1584. doi:10.1056/NEJMcps050929
Sheldon, R. A., Partridge, J. C., & Ferriero, D. M. (1992). Postischemic hyperglycemia is not protective to the neonatal rat brain. Pediatr Res, 32(4), 489-493. doi:10.1203/00006450-199210000-00022
Shellhaas, R. A., Burns, J. W., Hassan, F., Carlson, M. D., Barks, J. D. E., & Chervin, R. D. (2017). Neonatal Sleep-Wake Analyses Predict 18-month Neurodevelopmental Outcomes. Sleep, 40(11). doi:10.1093/sleep/zsx144
Shellhaas, R. A., Chang, T., Tsuchida, T., Scher, M. S., Riviello, J. J., Abend, N. S., Nguyen, S., Wusthoff, C. J., & Clancy, R. R. (2011). The American Clinical Neurophysiology Society's Guideline on Continuous Electroencephalography Monitoring in Neonates. J Clin Neurophysiol, 28(6), 611-617. doi:10.1097/WNP.0b013e31823e96d7
Shellhaas, R. A., & Clancy, R. R. (2007). Characterization of neonatal seizures by conventional EEG and single-channel EEG. Clin Neurophysiol, 118(10), 2156-2161. doi:10.1016/j.clinph.2007.06.061
Shellhaas, R. A., Gallagher, P. R., & Clancy, R. R. (2008). Assessment of neonatal electroencephalography (EEG) background by conventional and two amplitude-integrated EEG classification systems. J Pediatr, 153(3), 369-374. doi:10.1016/j.jpeds.2008.03.004
Shellhaas, R. A., Kushwaha, J. S., Plegue, M. A., Selewski, D. T., & Barks, J. D. (2015). An Evaluation of Cerebral and Systemic Predictors of 18-Month Outcomes for Neonates With Hypoxic Ischemic Encephalopathy. J Child Neurol, 30(11), 1526-1531. doi:10.1177/0883073815573319
Shellhaas, R. A., Soaita, A. I., & Clancy, R. R. (2007). Sensitivity of amplitude-integrated electroencephalography for neonatal seizure detection. Pediatrics, 120(4), 770-777. doi:10.1542/peds.2007-0514
Shi, Y., Zhao, J. N., Liu, L., Hu, Z. X., Tang, S. F., Chen, L., & Jin, R. B. (2012). Changes of positron emission tomography in newborn infants at different gestational ages, and neonatal hypoxic-ischemic encephalopathy. Pediatric Neurology, 46(2), 116-123. doi:10.1016/j.pediatrneurol.2011.11.005
Siegelaar, S. E., Holleman, F., Hoekstra, J. B., & DeVries, J. H. (2010). Glucose variability; does it matter? Endocr Rev, 31(2), 171-182. doi:10.1210/er.2009-0021
Siemkowicz, E., & Hansen, A. J. (1981). Brain extracellular ion composition and EEG activity following 10 minutes ischemia in normo- and hyperglycemic rats. Stroke, 12(2), 236-240.
146
Siesjo, B. K., Ingvar, M., & Pelligrino, D. (1983). Regional differences in vascular autoregulation in the rat brain in severe insulin-induced hypoglycemia. J Cereb Blood Flow Metab, 3(4), 478-485. doi:10.1038/jcbfm.1983.74
Signal, M., Le Compte, A., Harris, D. L., Weston, P. J., Harding, J. E., Chase, J. G., & Chyld Study, G. (2012). Impact of retrospective calibration algorithms on hypoglycemia detection in newborn infants using continuous glucose monitoring. Diabetes Technology & Therapeutics, 14(10), 883-890. doi:10.1089/dia.2012.0111
Signal, M., Thomas, F., Shaw, G. M., & Chase, J. G. (2013). Complexity of continuous glucose monitoring data in critically ill patients: continuous glucose monitoring devices, sensor locations, and detrended fluctuation analysis methods. Journal of Diabetes Science & Technology, 7(6), 1492-1506. doi:10.1177/193229681300700609
Simbruner, G., Mittal, R. A., Rohlmann, F., Muche, R., & neo.n, E. n. T. P. (2010). Systemic hypothermia after neonatal encephalopathy: outcomes of neo.nEURO.network RCT. Pediatrics, 126(4), e771-778. doi:10.1542/peds.2009-2441
Singh, M., Singhal, P. K., Paul, V. K., Deorari, A. K., Sundaram, K. R., Ghorpade, M. D., & Agadi, A. (1991). Neurodevelopmental outcome of asymptomatic & symptomatic babies with neonatal hypoglycaemia. Indian J Med Res, 94, 6-10.
Skov, L., & Pryds, O. (1992). Capillary recruitment for preservation of cerebral glucose influx in hypoglycemic, preterm newborns: evidence for a glucose sensor? Pediatrics, 90(2 Pt 1), 193-195.
Skranes, J. H., Lohaugen, G., Schumacher, E. M., Osredkar, D., Server, A., Cowan, F. M., Stiris, T., Fugelseth, D., & Thoresen, M. (2017). Amplitude-Integrated Electroencephalography Improves the Identification of Infants with Encephalopathy for Therapeutic Hypothermia and Predicts Neurodevelopmental Outcomes at 2 Years of Age. J Pediatr, 187, 34-42. doi:10.1016/j.jpeds.2017.04.041
Slidsborg, C., Jensen, L. B., Rasmussen, S. C., Fledelius, H. C., Greisen, G., & Cour, M. (2018). Early postnatal hyperglycaemia is a risk factor for treatment-demanding retinopathy of prematurity. Br J Ophthalmol, 102(1), 14-18. doi:10.1136/bjophthalmol-2016-309187
Soysal, A. S., Gucuyener, K., Ergenekon, E., Turan, O., Koc, E., Turkyilmaz, C., Onal, E., & Atalay, Y. (2014). The prediction of later neurodevelopmental status of preterm infants at ages 7 to 10 years using the Bayley Infant Neurodevelopmental Screener. J Child Neurol, 29(10), 1349-1355. doi:10.1177/0883073813520495
Spar, J. A., Lewine, J. D., & Orrison, W. W., Jr. (1994). Neonatal hypoglycemia: CT and MR findings. AJNR Am J Neuroradiol, 15(8), 1477-1478.
Sperling, M. A., Ganguli, S., Leslie, N., & Landt, K. (1984). Fetal-perinatal catecholamine secretion: role in perinatal glucose homeostasis. Am J Physiol, 247(1 Pt 1), E69-74. doi:10.1152/ajpendo.1984.247.1.E69
147
Spies, E. E., Lababidi, S. L., & McBride, M. C. (2014). Early hyperglycemia is associated with poor gross motor outcome in asphyxiated term newborns. Pediatric Neurology, 50(6), 586-590. doi:10.1016/j.pediatrneurol.2014.01.043
Spitzmiller, R. E., Phillips, T., Meinzen-Derr, J., & Hoath, S. B. (2007). Amplitude-integrated EEG is useful in predicting neurodevelopmental outcome in full-term infants with hypoxic-ischemic encephalopathy: a meta-analysis. J Child Neurol, 22(9), 1069-1078. doi:10.1177/0883073807306258
Srinivasakumar, P., Zempel, J., Trivedi, S., Wallendorf, M., Rao, R., Smith, B., Inder, T., & Mathur, A. M. (2015). Treating EEG Seizures in Hypoxic Ischemic Encephalopathy: A Randomized Controlled Trial. Pediatrics, 136(5), e1302-1309. doi:10.1542/peds.2014-3777
Srinivasan, G., Pildes, R. S., Cattamanchi, G., Voora, S., & Lilien, L. D. (1986). Plasma glucose values in normal neonates: a new look. J Pediatr, 109(1), 114-117.
Srinivasan, V., Spinella, P. C., Drott, H. R., Roth, C. L., Helfaer, M. A., & Nadkarni, V. (2004). Association of timing, duration, and intensity of hyperglycemia with intensive care unit mortality in critically ill children. Pediatr Crit Care Med, 5(4), 329-336. doi:10.1097/01.pcc.0000128607.68261.7c
Stanley, C. A., Anday, E. K., Baker, L., & Delivoria-Papadopolous, M. (1979). Metabolic fuel and hormone responses to fasting in newborn infants. Pediatrics, 64(5), 613-619.
Stanley, C. A., & Baker, L. (1976). Hyperinsulinism in infancy: diagnosis by demonstration of abnormal response to fasting hypoglycemia. Pediatrics, 57(5), 702-711.
Stanley, C. A., Rozance, P. J., Thornton, P. S., De Leon, D. D., Harris, D., Haymond, M. W., Hussain, K., Levitsky, L. L., Murad, M. H., Simmons, R. A., Sperling, M. A., Weinstein, D. A., White, N. H., & Wolfsdorf, J. I. (2015). Re-evaluating "transitional neonatal hypoglycemia": mechanism and implications for management. J Pediatr, 166(6), 1520-1525 e1521. doi:10.1016/j.jpeds.2015.02.045
Stenninger, E., Eriksson, E., Stigfur, A., Schollin, J., & Aman, J. (2001). Monitoring of early postnatal glucose homeostasis and cerebral function in newborn infants of diabetic mothers. A pilot study. Early Hum Dev, 62(1), 23-32.
Stenninger, E., Flink, R., Eriksson, B., & Sahlen, C. (1998). Long-term neurological dysfunction and neonatal hypoglycaemia after diabetic pregnancy. Arch Dis Child Fetal Neonatal Ed, 79(3), F174-179.
Stevenson, N. J., Clancy, R. R., Vanhatalo, S., Rosen, I., Rennie, J. M., & Boylan, G. B. (2015). Interobserver agreement for neonatal seizure detection using multichannel EEG. Ann Clin Transl Neurol, 2(11), 1002-1011. doi:10.1002/acn3.249
Stevenson, N. J., Korotchikova, I., Temko, A., Lightbody, G., Marnane, W. P., & Boylan, G. B. (2013). An automated system for grading EEG abnormality in term neonates with
148
hypoxic-ischaemic encephalopathy. Annals of Biomedical Engineering, 41(4), 775-785. doi:10.1007/s10439-012-0710-5
Stevenson, N. J., Palmu, K., Wikstrom, S., Hellstrom-Westas, L., & Vanhatalo, S. (2014). Measuring brain activity cycling (BAC) in long term EEG monitoring of preterm babies. Physiol Meas, 35(7), 1493-1508. doi:10.1088/0967-3334/35/7/1493
Suh, S. W., Gum, E. T., Hamby, A. M., Chan, P. H., & Swanson, R. A. (2007). Hypoglycemic neuronal death is triggered by glucose reperfusion and activation of neuronal NADPH oxidase. J Clin Invest, 117(4), 910-918. doi:10.1172/JCI30077
Sunehag, A., Gustafsson, J., & Ewald, U. (1994). Very immature infants (< or = 30 Wk) respond to glucose infusion with incomplete suppression of glucose production. Pediatr Res, 36(4), 550-555. doi:10.1203/00006450-199410000-00024
Sweet, D. G., Hadden, D., & Halliday, H. L. (1999). The effect of early feeding on the neonatal blood glucose level at 1-hour of age. Early Human Development, 55(1), 63-66. doi:10.1016/s0378-3782(99)00004-3
Takenouchi, T., Rubens, E. O., Yap, V. L., Ross, G., Engel, M., & Perlman, J. M. (2011). Delayed onset of sleep-wake cycling with favorable outcome in hypothermic-treated neonates with encephalopathy. J Pediatr, 159(2), 232-237. doi:10.1016/j.jpeds.2011.01.006
Tam, E. W., Haeusslein, L. A., Bonifacio, S. L., Glass, H. C., Rogers, E. E., Jeremy, R. J., Barkovich, A. J., & Ferriero, D. M. (2012). Hypoglycemia is associated with increased risk for brain injury and adverse neurodevelopmental outcome in neonates at risk for encephalopathy. J Pediatr, 161(1), 88-93. doi:10.1016/j.jpeds.2011.12.047
Tam, E. W., Widjaja, E., Blaser, S. I., Macgregor, D. L., Satodia, P., & Moore, A. M. (2008). Occipital lobe injury and cortical visual outcomes after neonatal hypoglycemia. Pediatrics, 122(3), 507-512. doi:10.1542/peds.2007-2002
Tauschmann, M., Allen, J. M., Wilinska, M. E., Thabit, H., Acerini, C. L., Dunger, D. B., & Hovorka, R. (2016). Home Use of Day-and-Night Hybrid Closed-Loop Insulin Delivery in Suboptimally Controlled Adolescents With Type 1 Diabetes: A 3-Week, Free-Living, Randomized Crossover Trial. Diabetes Care, 39(11), 2019-2025. doi:10.2337/dc16-1094
Tayman, C., Yis, U., Hirfanoglu, I., Oztekin, O., Goktas, G., & Bilgin, B. C. (2014). Effects of hyperglycemia on the developing brain in newborns. Pediatric Neurology, 51(2), 239-245. doi:10.1016/j.pediatrneurol.2014.04.015
Temko, A., Doyle, O., Murray, D., Lightbody, G., Boylan, G., & Marnane, W. (2015). Multimodal predictor of neurodevelopmental outcome in newborns with hypoxic-ischaemic encephalopathy. Comput Biol Med, 63, 169-177. doi:10.1016/j.compbiomed.2015.05.017
ter Horst, H. J., Sommer, C., Bergman, K. A., Fock, J. M., van Weerden, T. W., & Bos, A. F. (2004). Prognostic significance of amplitude-integrated EEG during the first 72 hours
149
after birth in severely asphyxiated neonates. Pediatr Res, 55(6), 1026-1033. doi:10.1203/01.pdr.0000127019.52562.8c
Thomas, F., Signal, M., Harris, D. L., Weston, P. J., Harding, J. E., Shaw, G. M., Chase, J. G., & Group, C. S. (2014). Continuous glucose monitoring in newborn infants: how do errors in calibration measurements affect detected hypoglycemia? Journal of Diabetes Science & Technology, 8(3), 543-550. doi:10.1177/1932296814524857
Thompson-Branch, A., & Havranek, T. (2017). Neonatal Hypoglycemia. Pediatr Rev, 38(4), 147-157. doi:10.1542/pir.2016-0063
Thordstein, M., Lofhede, J., Flisberg, A., Lofgren, N., Lindecrantz, K., & Kjellmer, I. (2010). Automated classification of human neonatal EEG. Clin Neurophysiol, 121, S163-S164. doi:http://dx.doi.org/10.1016/S1388-2457%2810%2960671-2
Thoresen, M., Hellstrom-Westas, L., Liu, X., & de Vries, L. S. (2010). Effect of hypothermia on amplitude-integrated electroencephalogram in infants with asphyxia. Pediatrics, 126(1), e131-139. doi:10.1542/peds.2009-2938
Thorngren-Jerneck, K., Ohlsson, T., Sandell, A., Erlandsson, K., Strand, S. E., Ryding, E., & Svenningsen, N. W. (2001). Cerebral glucose metabolism measured by positron emission tomography in term newborn infants with hypoxic ischemic encephalopathy. Pediatr Res, 49(4), 495-501. doi:10.1203/00006450-200104000-00010
Thornton, P. S., Alter, C. A., Katz, L. E., Baker, L., & Stanley, C. A. (1993). Short- and long-term use of octreotide in the treatment of congenital hyperinsulinism. J Pediatr, 123(4), 637-643.
Thornton, P. S., Stanley, C. A., De Leon, D. D., Harris, D., Haymond, M. W., Hussain, K., Levitsky, L. L., Murad, M. H., Rozance, P. J., Simmons, R. A., Sperling, M. A., Weinstein, D. A., White, N. H., Wolfsdorf, J. I., & Pediatric Endocrine Society. (2015). Recommendations from the Pediatric Endocrine Society for Evaluation and Management of Persistent Hypoglycemia in Neonates, Infants, and Children. J Pediatr, 167(2), 238-245. doi:10.1016/j.jpeds.2015.03.057
Tin, W. (2014). Defining neonatal hypoglycaemia: a continuing debate. Semin Fetal Neonatal Med, 19(1), 27-32. doi:10.1016/j.siny.2013.09.003
Tin, W., Brunskill, G., Kelly, T., & Fritz, S. (2012). 15-year follow-up of recurrent "hypoglycemia" in preterm infants. Pediatrics, 130(6), e1497-1503. doi:10.1542/peds.2012-0776
Tin, W., Fritz, S., Wariyar, U., & Hey, E. (1998). Outcome of very preterm birth: children reviewed with ease at 2 years differ from those followed up with difficulty. Arch Dis Child Fetal Neonatal Ed, 79(2), F83-87.
Toet, M. C., Hellstrom-Westas, L., Groenendaal, F., Eken, P., & de Vries, L. S. (1999). Amplitude integrated EEG 3 and 6 hours after birth in full term neonates with hypoxic-ischaemic encephalopathy. Arch Dis Child Fetal Neonatal Ed, 81(1), F19-23.
150
Topjian, A. A., Gutierrez-Colina, A. M., Sanchez, S. M., Berg, R. A., Friess, S. H., Dlugos, D. J., & Abend, N. S. (2013). Electrographic status epilepticus is associated with mortality and worse short-term outcome in critically ill children. Crit Care Med, 41(1), 215-223. doi:10.1097/CCM.0b013e3182668035
Tottman, A. C., Alsweiler, J. M., Bloomfield, F. H., Pan, M., & Harding, J. E. (2017). Relationship between Measures of Neonatal Glycemia, Neonatal Illness, and 2-Year Outcomes in Very Preterm Infants. J Pediatr, 188, 115-121. doi:10.1016/j.jpeds.2017.05.052
Traill, Z., Squier, M., & Anslow, P. (1998). Brain imaging in neonatal hypoglycaemia. Arch Dis Child Fetal Neonatal Ed, 79(2), F145-147. doi:10.1136/fn.79.2.F145
Tsuchida, T. N. (2013). EEG background patterns and prognostication of neonatal encephalopathy in the era of hypothermia. J Clin Neurophysiol, 30(2), 122-125. doi:10.1097/WNP.0b013e3182872ac2
Tsuchida, T. N., Barkovich, A. J., Bollen, A. W., Hart, A. P., & Ferriero, D. M. (2007). Childhood status epilepticus and excitotoxic neuronal injury. Pediatric Neurology, 36(4), 253-257. doi:10.1016/j.pediatrneurol.2006.12.005
Tsuchida, T. N., Wusthoff, C. J., Shellhaas, R. A., Abend, N. S., Hahn, C. D., Sullivan, J. E., Nguyen, S., Weinstein, S., Scher, M. S., Riviello, J. J., Clancy, R. R., & American Clinical Neurophysiology Society Critical Care Monitoring, C. (2013). American clinical neurophysiology society standardized EEG terminology and categorization for the description of continuous EEG monitoring in neonates: report of the American Clinical Neurophysiology Society critical care monitoring committee. J Clin Neurophysiol, 30(2), 161-173. doi:10.1097/WNP.0b013e3182872b24
Udani, V., Munot, P., Ursekar, M., & Gupta, S. (2009). Neonatal hypoglycemic brain - injury a common cause of infantile onset remote symptomatic epilepsy. Indian Pediatrics, 46(2), 127-132.
Uettwiller, F., Chemin, A., Bonnemaison, E., Favrais, G., Saliba, E., & Labarthe, F. (2015). Real-time continuous glucose monitoring reduces the duration of hypoglycemia episodes: a randomized trial in very low birth weight neonates. PLoS ONE [Electronic Resource], 10(1), e0116255. doi:10.1371/journal.pone.0116255
Vaewpanich, J., & Reuter-Rice, K. (2016). Continuous electroencephalography in pediatric traumatic brain injury: Seizure characteristics and outcomes. Epilepsy Behav, 62, 225-230. doi:10.1016/j.yebeh.2016.07.012
Van den Berghe, G. (2004). How does blood glucose control with insulin save lives in intensive care? J Clin Invest, 114(9), 1187-1195. doi:10.1172/JCI23506
Van den Berghe, G., Wilmer, A., Hermans, G., Meersseman, W., Wouters, P. J., Milants, I., Van Wijngaerden, E., Bobbaers, H., & Bouillon, R. (2006). Intensive insulin therapy in the medical ICU. New England Journal of Medicine, 354(5), 449-461. doi:10.1056/NEJMoa052521
151
van den Berghe, G., Wouters, P., Weekers, F., Verwaest, C., Bruyninckx, F., Schetz, M., Vlasselaers, D., Ferdinande, P., Lauwers, P., & Bouillon, R. (2001). Intensive insulin therapy in critically ill patients. New England Journal of Medicine, 345(19), 1359-1367. doi:10.1056/NEJMoa011300
van der Lugt, N. M., Smits-Wintjens, V. E., van Zwieten, P. H., & Walther, F. J. (2010). Short and long term outcome of neonatal hyperglycemia in very preterm infants: a retrospective follow-up study. BMC Pediatr, 10, 52. doi:10.1186/1471-2431-10-52
van Handel, M., Swaab, H., de Vries, L. S., & Jongmans, M. J. (2010). Behavioral outcome in children with a history of neonatal encephalopathy following perinatal asphyxia. J Pediatr Psychol, 35(3), 286-295. doi:10.1093/jpepsy/jsp049
van Laerhoven, H., de Haan, T. R., Offringa, M., Post, B., & van der Lee, J. H. (2013). Prognostic tests in term neonates with hypoxic-ischemic encephalopathy: a systematic review. Pediatrics, 131(1), 88-98. doi:10.1542/peds.2012-1297
van Rooij, L. G., de Vries, L. S., van Huffelen, A. C., & Toet, M. C. (2010). Additional value of two-channel amplitude integrated EEG recording in full-term infants with unilateral brain injury. Arch Dis Child Fetal Neonatal Ed, 95(3), F160-168. doi:10.1136/adc.2008.156711
van Rooij, L. G., Toet, M. C., van Huffelen, A. C., Groenendaal, F., Laan, W., Zecic, A., de Haan, T., van Straaten, I. L., Vrancken, S., van Wezel, G., van der Sluijs, J., Ter Horst, H., Gavilanes, D., Laroche, S., Naulaers, G., & de Vries, L. S. (2010). Effect of treatment of subclinical neonatal seizures detected with aEEG: randomized, controlled trial. Pediatrics, 125(2), e358-366. doi:10.1542/peds.2009-0136
VanHaltren, K., & Malhotra, A. (2013). Characteristics of infants at risk of hypoglycaemia secondary to being 'infant of a diabetic mother'. Journal of Pediatric Endocrinology & Metabolism, 26(9-10), 861-865. doi:10.1515/jpem-2013-0012
Vannucci, R. C., Brucklacher, R. M., & Vannucci, S. J. (1996). The effect of hyperglycemia on cerebral metabolism during hypoxia-ischemia in the immature rat. J Cereb Blood Flow Metab, 16(5), 1026-1033. doi:10.1097/00004647-199609000-00028
Vannucci, R. C., Nardis, E. E., & Vannucci, S. J. (1980). Cerebral metabolism during hypoglycemia dn asphyxia in newborn dogs. Biol Neonate, 38(5-6), 276-286. doi:10.1159/000241377
Vannucci, R. C., Nardis, E. E., Vannucci, S. J., & Campbell, P. A. (1981). Cerebral carbohydrate and energy metabolism during hypoglycemia in newborn dogs. Am J Physiol, 240(3), R192-199. doi:10.1152/ajpregu.1981.240.3.R192
Vannucci, R. C., Rossini, A., & Towfighi, J. (1996). Effect of hyperglycemia on ischemic brain damage during hypothermic circulatory arrest in newborn dogs. Pediatr Res, 40(2), 177-184. doi:10.1203/00006450-199608000-00001
152
Vannucci, R. C., & Vannucci, S. J. (1978). Cerebral carbohydrate metabolism during hypoglycemia and anoxia in newborn rats. Ann Neurol, 4(1), 73-79. doi:10.1002/ana.410040114
Vannucci, R. C., & Vannucci, S. J. (2000). Glucose metabolism in the developing brain. Semin Perinatol, 24(2), 107-115. doi:10.1053/sp.2000.6361
Vannucci, R. C., & Vannucci, S. J. (2001). Hypoglycemic brain injury. Semin Neonatol, 6(2), 147-155. doi:10.1053/siny.2001.0044
Vannucci, R. C., Vasta, F., & Vannucci, S. J. (1987). Cerebral metabolic responses of hyperglycemic immature rats to hypoxia-ischemia. Pediatr Res, 21(6), 524-529. doi:10.1203/00006450-198706000-00002
Varma, S., Nickerson, H., Cowan, J. S., & Hetenyi Jr, G. (1973). Homeostatic responses to glucose loading in newborn and young dogs. Metabolism: Clinical and Experimental, 22(11), 1367-1375. doi:http://dx.doi.org/10.1016/0026-0495%2873%2990252-7
Vespa, P., Boonyaputthikul, R., McArthur, D. L., Miller, C., Etchepare, M., Bergsneider, M., Glenn, T., Martin, N., & Hovda, D. (2006). Intensive insulin therapy reduces microdialysis glucose values without altering glucose utilization or improving the lactate/pyruvate ratio after traumatic brain injury. Crit Care Med, 34(3), 850-856. doi:10.1097/01.CCM.0000201875.12245.6F
Vespa, P., Prins, M., Ronne-Engstrom, E., Caron, M., Shalmon, E., Hovda, D. A., Martin, N. A., & Becker, D. P. (1998). Increase in extracellular glutamate caused by reduced cerebral perfusion pressure and seizures after human traumatic brain injury: a microdialysis study. J Neurosurg, 89(6), 971-982. doi:10.3171/jns.1998.89.6.0971
Vespa, P., Tubi, M., Claassen, J., Blanco, M., McArthur, D., Velazquez, A. G., Tu, B., Prins, M., & Nuwer, M. (2016). Metabolic Crisis occurs with Seizures and Periodic Discharges after Brain Trauma. Ann Neurol. doi:10.1002/ana.24606
Vespa, P. M., McArthur, D. L., Xu, Y., Eliseo, M., Etchepare, M., Dinov, I., Alger, J., Glenn, T. P., & Hovda, D. (2010). Nonconvulsive seizures after traumatic brain injury are associated with hippocampal atrophy. Neurology, 75(9), 792-798. doi:10.1212/WNL.0b013e3181f07334
Vespa, P. M., Miller, C., McArthur, D., Eliseo, M., Etchepare, M., Hirt, D., Glenn, T. C., Martin, N., & Hovda, D. (2007). Nonconvulsive electrographic seizures after traumatic brain injury result in a delayed, prolonged increase in intracranial pressure and metabolic crisis. Crit Care Med, 35(12), 2830-2836.
Vlasselaers, D., Milants, I., Desmet, L., Wouters, P. J., Vanhorebeek, I., van den Heuvel, I., Mesotten, D., Casaer, M. P., Meyfroidt, G., Ingels, C., Muller, J., Van Cromphaut, S., Schetz, M., & Van den Berghe, G. (2009). Intensive insulin therapy for patients in paediatric intensive care: a prospective, randomised controlled study. Lancet, 373(9663), 547-556. doi:10.1016/S0140-6736(09)60044-1
153
Vohr, B. R., Wright, L. L., Dusick, A. M., Mele, L., Verter, J., Steichen, J. J., Simon, N. P., Wilson, D. C., Broyles, S., Bauer, C. R., Delaney-Black, V., Yolton, K. A., Fleisher, B. E., Papile, L. A., & Kaplan, M. D. (2000). Neurodevelopmental and functional outcomes of extremely low birth weight infants in the National Institute of Child Health and Human Development Neonatal Research Network, 1993-1994. Pediatrics, 105(6), 1216-1226.
Volpe, J. (2008). Neurology of the Newborn, 5th ed. Philadelphia, PA: Saunders.
Volpe, J. J. (2012). Neonatal encephalopathy: an inadequate term for hypoxic-ischemic encephalopathy. Ann Neurol, 72(2), 156-166. doi:10.1002/ana.23647
Voorhies, T. M., Rawlinson, D., & Vannucci, R. C. (1986). Glucose and perinatal hypoxic-ischemic brain damage in the rat. Neurology, 36(8), 1115-1118.
Wackernagel, D., Dube, M., Blennow, M., & Tindberg, Y. (2016). Continuous subcutaneous glucose monitoring is accurate in term and near-term infants at risk of hypoglycaemia. Acta Paediatr, 105(8), 917-923. doi:10.1111/apa.13479
Waeschle, R. M., Moerer, O., Hilgers, R., Herrmann, P., Neumann, P., & Quintel, M. (2008). The impact of the severity of sepsis on the risk of hypoglycaemia and glycaemic variability. Crit Care, 12(5), R129. doi:10.1186/cc7097
Wagenman, K. L., Blake, T. P., Sanchez, S. M., Schultheis, M. T., Radcliffe, J., Berg, R. A., Dlugos, D. J., Topjian, A. A., & Abend, N. S. (2014). Electrographic status epilepticus and long-term outcome in critically ill children. Neurology, 82(5), 396-404. doi:10.1212/WNL.0000000000000082
Walsh, B. H., Murray, D. M., & Boylan, G. B. (2011). The use of conventional EEG for the assessment of hypoxic ischaemic encephalopathy in the newborn: a review. Clin Neurophysiol, 122(7), 1284-1294. doi:10.1016/j.clinph.2011.03.032
Walsh, B. H., Neil, J., Morey, J., Yang, E., Silvera, M. V., Inder, T. E., & Ortinau, C. (2017). The Frequency and Severity of Magnetic Resonance Imaging Abnormalities in Infants with Mild Neonatal Encephalopathy. J Pediatr, 187, 26-33 e21. doi:10.1016/j.jpeds.2017.03.065
Wartenberg, K. E., Schmidt, J. M., Claassen, J., Temes, R. E., Frontera, J. A., Ostapkovich, N., Parra, A., Connolly, E. S., & Mayer, S. A. (2006). Impact of medical complications on outcome after subarachnoid hemorrhage. Crit Care Med, 34(3), 617-623. doi:10.1097/01.ccm.0000201903.46435.35
Wassink, G., Barrett, R. D., Davidson, J. O., Bennet, L., Galinsky, R., Dragunow, M., & Gunn, A. J. (2015). Hypothermic neuroprotection is associated with recovery of spectral edge frequency after asphyxia in preterm fetal sheep. Stroke, 46(2), 585-587. doi:10.1161/STROKEAHA.114.008484
Wassink, G., Gunn, E. R., Drury, P. P., Bennet, L., & Gunn, A. J. (2014). The mechanisms and treatment of asphyxial encephalopathy. Front Neurosci, 8, 40. doi:10.3389/fnins.2014.00040
154
Weeke, L. C., Boylan, G. B., Pressler, R. M., Hallberg, B., Blennow, M., Toet, M. C., Groenendaal, F., de Vries, L. S., & consortium, N. E. s. t. w. M. O.-p. (2016). Role of EEG background activity, seizure burden and MRI in predicting neurodevelopmental outcome in full-term infants with hypoxic-ischaemic encephalopathy in the era of therapeutic hypothermia. Eur J Paediatr Neurol, 20(6), 855-864. doi:10.1016/j.ejpn.2016.06.003
Wietstock, S. O., Bonifacio, S. L., Sullivan, J. E., Nash, K. B., & Glass, H. C. (2016). Continuous Video Electroencephalographic (EEG) Monitoring for Electrographic Seizure Diagnosis in Neonates: A Single-Center Study. J Child Neurol, 31(3), 328-332. doi:10.1177/0883073815592224
Wikstrom, S., Lundin, F., Ley, D., Pupp, I. H., Fellman, V., Rosen, I., & Hellstrom-Westas, L. (2011). Carbon dioxide and glucose affect electrocortical background in extremely preterm infants. Pediatrics, 127(4), e1028-1034. doi:10.1542/peds.2010-2755
Wintergerst, K. A., Buckingham, B., Gandrud, L., Wong, B. J., Kache, S., & Wilson, D. M. (2006). Association of hypoglycemia, hyperglycemia, and glucose variability with morbidity and death in the pediatric intensive care unit. Pediatrics, 118(1), 173-179. doi:10.1542/peds.2005-1819
Wong, D. S., Poskitt, K. J., Chau, V., Miller, S. P., Roland, E., Hill, A., & Tam, E. W. (2013). Brain injury patterns in hypoglycemia in neonatal encephalopathy. AJNR Am J Neuroradiol, 34(7), 1456-1461. doi:10.3174/ajnr.A3423
Wu, Y., Lai, W., Pei, J., Zhao, Y., Wang, Q., & Xiang, B. (2016). Hyperglycemia and its association with clinical outcomes in postsurgical neonates and small infants in the intensive care unit. Journal of Pediatric Surgery, 51(7), 1142-1145. doi:10.1016/j.jpedsurg.2016.01.001
Wu, Y. W., Bauer, L. A., Ballard, R. A., Ferriero, D. M., Glidden, D. V., Mayock, D. E., Chang, T., Durand, D. J., Song, D., Bonifacio, S. L., Gonzalez, F. F., Glass, H. C., & Juul, S. E. (2012). Erythropoietin for neuroprotection in neonatal encephalopathy: safety and pharmacokinetics. Pediatrics, 130(4), 683-691. doi:10.1542/peds.2012-0498
Wu, Y. W., Mathur, A. M., Chang, T., McKinstry, R. C., Mulkey, S. B., Mayock, D. E., Van Meurs, K. P., Rogers, E. E., Gonzalez, F. F., Comstock, B. A., Juul, S. E., Msall, M. E., Bonifacio, S. L., Glass, H. C., Massaro, A. N., Dong, L., Tan, K. W., Heagerty, P. J., & Ballard, R. A. (2016). High-Dose Erythropoietin and Hypothermia for Hypoxic-Ischemic Encephalopathy: A Phase II Trial. Pediatrics, 137(6). doi:10.1542/peds.2016-0191
Wusthoff, C. J., Dlugos, D. J., Gutierrez-Colina, A., Wang, A., Cook, N., Donnelly, M., Clancy, R., & Abend, N. S. (2011). Electrographic seizures during therapeutic hypothermia for neonatal hypoxic-ischemic encephalopathy. J Child Neurol, 26(6), 724-728. doi:10.1177/0883073810390036
Wusthoff, C. J., Sullivan, J., Glass, H. C., Shellhaas, R. A., Abend, N. S., Chang, T., & Tsuchida, T. N. (2017). Interrater agreement in the interpretation of neonatal
155
electroencephalography in hypoxic-ischemic encephalopathy. Epilepsia, 58(3), 429-435. doi:10.1111/epi.13661
Wyss, M. T., Jolivet, R., Buck, A., Magistretti, P. J., & Weber, B. (2011). In vivo evidence for lactate as a neuronal energy source. Journal of Neuroscience, 31(20), 7477-7485. doi:10.1523/JNEUROSCI.0415-11.2011
Yager, J. Y., Armstrong, E. A., & Black, A. M. (2009). Treatment of the term newborn with brain injury: simplicity as the mother of invention. Pediatric Neurology, 40(3), 237-243. doi:10.1016/j.pediatrneurol.2008.12.002
Yalnizoglu, D., Haliloglu, G., Turanli, G., Cila, A., & Topcu, M. (2007). Neurologic outcome in patients with MRI pattern of damage typical for neonatal hypoglycemia. Brain & Development, 29(5), 285-292. doi:10.1016/j.braindev.2006.09.011
Yamada, T., Shojima, N., Noma, H., Yamauchi, T., & Kadowaki, T. (2017). Glycemic control, mortality, and hypoglycemia in critically ill patients: a systematic review and network meta-analysis of randomized controlled trials. Intensive Care Med, 43(1), 1-15. doi:10.1007/s00134-016-4523-0
Yang, G., Zou, L. P., Wang, J., Shi, X., Tian, S., Yang, X., Ju, J., Yao, H., & Liu, Y. (2016). Neonatal hypoglycemic brain injury is a cause of infantile spasms. Exp Ther Med, 11(5), 2066-2070. doi:10.3892/etm.2016.3107
Younkin, D. P., Delivoria-Papadopoulos, M., Maris, J., Donlon, E., Clancy, R., & Chance, B. (1986). Cerebral metabolic effects of neonatal seizures measured with in vivo 31P NMR spectroscopy. Ann Neurol, 20(4), 513-519. doi:10.1002/ana.410200412
Zarif, M., Pildes, R. S., & Vidyasagar, D. (1976). Insulin and growth-hormone responses in neonatal hyperglycemia. Diabetes, 25(5), 428-433. doi:10.2337/diab.25.5.428
Zhang, Y., Kuang, Y., Xu, K., Harris, D., Lee, Z., LaManna, J., & Puchowicz, M. A. (2013). Ketosis proportionately spares glucose utilization in brain. J Cereb Blood Flow Metab, 33(8), 1307-1311. doi:10.1038/jcbfm.2013.87
Zhou, W., Yu, J., Wu, Y., & Zhang, H. (2015). Hypoglycemia incidence and risk factors assessment in hospitalized neonates. Journal of Maternal-Fetal & Neonatal Medicine, 28(4), 422-425. doi:10.3109/14767058.2014.918599
Zhou, W. H., Cheng, G. Q., Shao, X. M., Liu, X. Z., Shan, R. B., Zhuang, D. Y., Zhou, C. L., Du, L. Z., Cao, Y., Yang, Q., Wang, L. S., & China Study Group. (2010). Selective head cooling with mild systemic hypothermia after neonatal hypoxic-ischemic encephalopathy: a multicenter randomized controlled trial in China. J Pediatr, 157(3), 367-372, 372 e361-363. doi:10.1016/j.jpeds.2010.03.030
Zhou, Y., Bai, S., Bornhorst, J. A., Elhassan, N. O., & Kaiser, J. R. (2017). The Effect of Early Feeding on Initial Glucose Concentrations in Term Newborns. J Pediatr, 181, 112-115. doi:10.1016/j.jpeds.2016.10.032
156
Appendix 1 - Appendix Table 1
Demographic data of the study cohort; for participants included in aEEG analysis and for participants excluded for incomplete data
COHORT INCLUDED WITH COMPLETE DATA
n=45
EXCLUDED FOR INCOMPLETE DATA
n=6
p value
Sex 21 ♀: 24 ♂ 3♀: 3♂ 1.000 Gestational age (wks), mean (SD)
39.54 (1.40) 39.86 (1.52) 0.645
Birth weight (grams), mean (SD)
3414.36 (548.24) 2745 (487.38) 0.018**
Birth length (cm), mean (SD)
50.14 (3.15) 47.92 (2.11) 0.052
Head circumference (cm), mean (SD)
34.22 (1.31) 33.17 (1.37) 0.123
Maternal diabetes, n (%) Type II Diabetes Mellitus Gestational Diabetes
5 (11) 1 (2) 4 (9)
1 (17) 0(0)
1 (17)
0.548
Maternal hypertension, n (%) Maternal pre-eclampsia/ eclampsia
4 (9)
2 (4.5)
0 (0)
0 (0)
1.000
Apgar score at 5 min, mean (SD)
4.18 (2.39) 4.50 (1.05) 0.572
Umbilical arterial cord pH, mean (SD)
6.99 (0.18) 6.99 (0.23) 0.987
Umbilical arterial cord BE, mean (SD)
-15.13 (6.79) -13.33 (11.37) 0.812
Route of delivery, n (%) Vaginal delivery Cesarean section
20 (44.5) 25 (55.5)
3 (50) 3 (50)
1.000
Sentinel event, n (%) 17 (38) 4 (67) 0.214 Seizures, n (%)* 28 (62) 2(33) 0.214 Clinical seizures only 16 (35.5) 0 (0) Electrographic and/or electroclinical seizures
12 (26.5) 2 (33)
Antiepileptic medications, n (%)
26 (58) 2 (33) 0.390
Lorazepam only 6 (13) 0 (0) Phenobarbital 13 (29) 0 (0) ≥2AEDs 7 (15.5) 2 (33)
Continuous variables were analyzed with the student’s T-test, whereas categorical variables were analyzed with Fisher’s exact test Significant p values are shown in bold *Electrographic and/or electroclinical seizures were identified by clinical team on aEEG or cEEG. Clinical seizures only had no seizures
captured on aEEG or cEEG monitoring **The birth weight was significantly lower in excluded neonates and may be related to greater difficulty with CGM insertion in smaller
neonates
157
Appendix 2 - Contributions
The author was responsible for the writing and preparation of this original thesis. All of the work
presented, including the planning, data analysis, interpretation of results and writing of the
original research, was performed by the author, with the guidance and expertise of the
individuals listed below. The following contributions to the work in this thesis are formally and
inclusively acknowledged:
Dr. Emily Tam conceived and developed study design, data collection, provided guidance with
analytic approach and interpretation of results, and provided important feedback and detailed
revisions of the thesis.
Dr. Cecil Hahn provided guidance with study design, analytic approach, interpretation of
results, provided expertise for development of EEG scoring and review, supervision for clinical
review and reporting of cEEG studies, and provided important feedback and detailed revisions of
the thesis.
Dr. Rollin Brant provided guidance for statistical analyses.
Dr. Steven Miller and Dr. Aideen Moore provided guidance with study design and analytic
approach. Dr. Vann Chau provided guidance with study design, analytic approach and
contributed help with clinical review and reporting of cEEG studies.
Daphne Kamino was responsible for overall study set-up and management, including managing
REB submissions, protocol development, subject recruitment, REDCap database design & data
management of all study data and helped with cEEG coordination. Ashley LeBlanc, Angela
Thompson, and Giselle Da Rocha contributed help with clinical data collection.
Drs. Ayako Ochi, Tina Go, Dragos Nita and Ala Birca contributed help with clinical review
and reporting of cEEG studies. The Neurophysiology technologists including Roy Sharma
applied electrodes and started EEG recordings for all study subjects.
The Acute Care Transport Services team including Marc LePine and Michelle Manning
recruited all study subjects and Matthew Keyzers for study recruitment and storage of aEEG
files.
This work was supported by a CIHR grant awarded to Dr. Emily Tam, as well as the Savoy
Epilepsy Foundation Studentship and Research Training Competition (RESTRACOMP)
Graduate Scholarships and Research Fellowship awarded to Dr. Elana Pinchefsky.
158
Copyright Acknowledgements
Formal copyright permission for use for Figures 1.2, 1.3, and section 1.7.3 follow this page
Please see note below regarding the referenced Figure 1.8
Figure 1.8 was reproduced from: Stolwijk LJ, Weeke LC, de Vries LS, van Herwaarden MYA,
van der Zee DC, et al. (2017) Effect of general anesthesia on neonatal aEEG—A cohort study of
patients with non-cardiac congenital anomalies. PLOS ONE 12(8): e0183581.
159
160
161
162
163
164
165
166
167
168
169
