014 PDF
EEG BASED PREDICTION OF
HYPOGLYCAEMIA IN
CHILDREN WITH T1D
Jesper Johannesen, Pia Foli, Siri Fredheim, Grith
Lrkholm, Martin H. Rose, Jonas Duun-Henriksen,
Claus B. Juhl, Kasper Pilgaard and Birthe Olsen
Conflicts of interest
Fear of Hypoglycaemia
Background for the present study
It is well described that hypoglycaemia is associated with changes
in the electroencephalogram (EEG)
It has recently been demonstrated that hypoglycaemia associated
EEG changes in adults recorded by subcutaneously placed electrodes
precede cognitive failure both during daytime and during sleep
Concept
Aim
The aim of this study is to get a first indication of the possibility to
trace early warning events of hypoglycaemia in children by
continuous EEG monitoring.
(i) To compare quality EEG obtained during normoglycaemia to
hypoglycaemia any difference?
(ii) If any difference could be detected, whether an algorithm
applied to the qEEG would predict an hypoglycamic event
Study Design
Subjects
N: 8 (4 males and 4 females)
Age: 9.62.3 yrs (prepubertal); range 6.4 12.5 yrs
Type 1 diabetes
Diabetes duration: 3.01.4 yrs
Tx: 7/8 on CSII
HbA1c: 553,4 mmol/mol
Frequent (>2 episodes) of symptomatic or biochemical
(<2.5mmol/l) event of Nocturnal hypoglycaemia within the past year
as assessed by parents or patients own reporting
Hypoglycaemia
Hyperinsulinaemic Clamp
Insulin 80mU/m2/min i.v. (2mU/kg/min). e.g 4,8 IE/hr at 40 kg
Glucose 20%: 1 ml/kg/hr (3,3 mg/kg/min)
Euglucaemia for 40 min
Hypoglycaemic period
Restoration of euglycaemia to BG > 8mmol/l
Hypoglycaemia procedure stop
The hypoglycaemia is ceased for one of the following reasons:
Blood glucose < 2.2 mmol/l (two consecutive measurements)
Blood glucose < 3.0 mmol/l and the patient is obvious cognitively
affected by the low blood glucose level
The patient (or parent) want the hypoglycaemia to be ceased for
any reason
The investigator want the hypoglycaemia to be ceased for any
reason
Whichever 1-4 occurs first.
Blood glucose curve
Average nadir: 2.30.5 mmol/l
(range 1.6 2.9 mmol/l)
Amplitude spectrum
Theta: 4-7.75 Hz
Delta: 1-3.75 Hz
Alfa: 8-12.75 Hz
Beta: 13-30 Hz
Amplitude spectrum
Average amplitude spectrum
Results of qEEG analysis
Results from the algorithm
Average time from event to nadir: 18,4 20,3 minutes
(range 0-55 minutes)
BGL at event: 2,50.5 mmol/l
(range 1.6 2.9 mmol/l)
BGL at nadir: 2.30.5 mmol/l
(range 1.6 2.9 mmol/l)
Conclusion
Demonstration of significant differences in qEEG during daytime
comparing normoglycaemia to hypoglycaemia
The algorithm based upon adult data identifies hypoglucaemia prior
to development of severe hypoglucaemia in children
Adrenaline and cortisol levels are decreased under nocturnal versus daytime hypoglycaemia in
pre-pubertal children with T1D
Pia Foli, Siri Fredheim, Grith Lrkholm, Jannet Svensson, Claus B, Juhl, Birthe Olsen, Kasper Pilgaard
and Jesper Johannesen
1Herlev
University Hospital, Department of Paediatrics, Copenhagen, Denmark
Hyposafe, Lyngby, Denmark
e-mail: piafoli@hotmail.com
Background & Objectives
Results
6,3% of Danish children with T1D have severe hypoglycaemia
(SH) and the incidence of SH is 7,6 per 100 person-year in
2013. Nocturnal hypoglycaemia occurs under sleep and is
feared, frequent, often asymptomatic and can be fatal. Counter
regulatory hormone (CR) responses has not jet been well
characterized in children and adolescents during day and night
time hypoglycaemia. Our aim is to investigate the CR profile
(glucagon, cortisol, adrenaline and growth hormone (GH)) in
children during a hyperinsulinaemic hypoglycaemic clamp
procedure in day and night time.
Both blood glucose levels and blood glucose decline were comparable day and night. The levels of adrenalin
(nmol/L) 0.60.5 vs 1.91.3 and cortisol (nmol/L) 4240 vs 319229 were blunted at night compared to
daytime. The levels of glucagon and GH at nadir day and night were comparable. Glucagon response to
hypoglycaemia were insignifikant. The increase in adrenalin and GH from hypoglycaemia to nadir were higher
during daytime vs night (p=0.04 and p=0.01, resp). The increase in hormone concentration from
hypoglycaemia to nadir were independent of blood glucose decline day and night. Results shown in Figur 2 a-d.
Conclusion
The response of adrenalin and
cortisol to hypoglycaemia
were lower at night compared
to daytime. These results
indicates that the
physiological response to
hypoglycaemia is impaired at
night compared to daytime in
children with T1D, with
blunted response of several
counter regulatory hormones.
Materials & Methods
8 children with T1D were recruited from the Department of
Paediatrics in Herlev Hospital. Seven children had an insulin
pump, one child were on insulin pen treatment. The
demographics is shown in Table 1.
Mean SEM
25th, 75th percentiles
Sex (boys/girls)
4/4
Age (years)
9.6 0.8
7.3, 11.3
17.1
15.8, 18.0
Diabetes duration (years)
3.0 0.5
2.2, 3.7
HbA1c (IFCC, mmol/mol)
55 1
52, 58
Table 1
Participant characteristics
BMI (kg
-1 m2)
HbA1c (NGSP, %)
0.1
7.5
Two identical
clamps were performed7.2one
in daytime and6.9,one
at
night during sleep. The protocol set up for the hyperinsulinaemic
hypoglycaemic clamp is shown in Figure 1.
Figure 2a-d: Counter-regulatory (CR) hormone levels expressed as
mean (SEM). X-axis, longitudinal venous blood glucose (VBG) levels.
Normoglycemia: BG at 7-9 mmol/L. Hypoglycemia: BG 3.5 mmol/L.
Nadir: VBG 2.2 mmol/L. *P values < 0.05 (significant) . P values
correspond to the non-parametric paired t-test for the hypothesis that
CR hormone levels are equal during day and night clamp-settings.
Wilcoxon signed rank test for day vs. night comparison was used.
P156
Thanks:
To all participating children and their
parents
