Nigerian Journal of Paediatrics 2011;38(2):90-94
ORIGINAL
Dedeke I.O.F
Point-of-admission neonatal hypoglycaemia in a
Okeniyi J.A.O
Nigerian
tertiary
hospital:
incidence,
risk
Owa J.A
factors and outcome.
Oyedeji G.A
Received: 7 March 2011
Abstract
Background Neonatal
(p = 0.034), admission weight less
Accepted: 27 June 2011
hypoglycaemia
is
a
major
t h a n
2 5 0 0 g
( p
=
0 . 0 0 9 ) ,
metabolic problem. It may result in
hypothermia (p = 0.001) and
I.O.F. Dedeke ( )
mortality or severe handicap
preterm birth (p = 0.020) were
Department of Paediatrics,
among survivors. Many babies
significantly more common in
Federal Medical Centre,
admitted for neonatal care are at
babies with hypoglycaemia. Poor
Abeokuta, Nigeria
high risk for hypoglycaemia. The
suck (p = 0.010), cyanosis (p =
E-mail:
present study set out to determine
0.020), convulsion (p = 0.040) and
bisiwunmiddk@yahoo.com
its point-of-admission prevalence,
pallor (p = 0.048) were also more
Tel: +234803 4547 231
clinical presentation and outcome.
common among babies with
Methods: Consecutive neonates
hypoglycaemia. The mortality rate
Okeniyi JAO, Owa JA
who met the study criteria had
in babies with hypoglycaemia was
Department of Paediatrics and
plasma glucose determined at
32.7%, higher than 18.8% in babies
Child Health, Obafemi
admission into the special care
without hypoglycaemia but the
Awolowo University, Ile-Ife,
baby unit of Wesley Guild
difference was not statistically
Nigeria
Hospital. Hypoglycaemia was
significant (p = 0.060).
defined as plasma glucose of <
Conclusion: Hypoglycaemia is
Oyedeji GA
2.5mmol/L. Babies with and
common among high-risk neonates
Department of Paediatrics,
without hypoglycaemia were
and is often associated with
Ladoke Akintola University of
compared for risk factors, clinical
morbidity and mortality. Routine
Technology, Osogbo, Nigeria
features and outcome.
monitoring of blood glucose is
Results: A total of 150 neonates
therefore recommended for this
Ogunlesi TA
were studied out of which 49
class of babies
Department of Paediatrics,
(32.7%) had hypoglycaemia. The
Olabisi Onabanjo University,
mean age, 38.3 ± 71.6 in hours was
Sagamu, Nigeria
significantly
lower
among
Key Words:
Prevalence, Point-
neonates with hypoglycaemia than
o f - a d m i s s i o n ,
N e o n a t a l
those without hypoglycaemia [p =
Hypoglycaemia, Morbidity and
0.006]. Low socioeconomic class
Mortality, Nigeria.
Introduction
Outcome of NNH depends on the rapidity of onset;
associated co-morbidities and adequacy of therapy.
2
Neonatal hypoglycaemia (NNH) is one of the most
Long-term sequelae include poor neurological
common metabolic problems in contemporary
development, poor intellectual function, motor
neonatal medicine. It is defined either as whole
1
deficits (especially spasticity and ataxia) and seizure
disorders.
3
blood glucose of less than 2.2 mmol/L or plasma
glucose of less than or equal to 2.5 mmol/L.
1
Risk factors for NNH include maternal diabetes
91
mellitus, corticosteroid administration during
monitoring blood glucose levels.
pregnancy, toxaemia of pregnancy, prematurity,
Data analysis was done using the Computer Package
intrauterine growth restriction (IUGR), septicaemia,
for Epidemiologists (PEPI). Tests of statistical
polycythaemia and erythroblastosis. Other factors
4
significance was done using Chi-square analysis for
include adverse perinatal events like prolonged
discreet and Student t-test for continuous variables
labour, birth asphyxia, cold stress, severe respiratory
respectively. Statistical significance was set at p
distress (due to increased energy utilisation).
4
values less than 0.05.
While several studies have been done on neonatal
hypoglycaemia in the developed world, reports from
Nigeria and perhaps many other developing
countries on NNH are sparse. Specifically, there has
Results
been no previous documentation on the prevalence
of neonatal hypoglycaemia at the Wesley Guild
A total of 150 babies were studied. They consisted of
Hospital (WGH), Ilesa, Nigeria, a tertiary health
88 males and 62 females giving a male: female ratio
facility which had been in existence for close to a
of 1.42: 1. Table 1 compares the socio-demographic
century. Wesley Guild Hospital is a unit of the
data
between
babies
with
and
without
Obafemi Awolowo University Teaching Hospitals
hypoglycaemia. Of the 150 babies studied, 49
Complex, Ile-Ife. The hospital provides general and
(32.7%) had hypoglycaemia at the time of admission.
specialist paediatric care to the adjourning
Of the remaining 101 (67.3%), 93 (92.1%) babies had
communities of at least four states in Nigeria. The
normal glucose levels while eight (7.9%) had
objective of the present study was to determine the
hyperglycaemia (plasma glucose > 8.3 mmol/l). Of
prevalence, risk factors, clinical presentations at
the 49 babies with hypoglycaemia, 29 (59.2%) were
time of admission, and outcome of NNH in a
males and 20 (40.8%) were females giving a male:
population of hospitalized Nigerian newborn babies.
female ratio of 1.45: 1 similar to overall male: female
ratio of 1.42: 1. A significantly higher proportion [43
(37.1%) of 116] of babies from lower social class had
Methodology
hypoglycaemia compared with six (17.6%) of 34
babies from higher social class (p = 0.034).
This was a prospective study, conducted over a six-
Ahigher proportion of referred babies compared with
month period. Institutional Ethical Clearance and
inborn babies experienced hypoglycaemia but the
parental informed consent were obtained. All
difference was not statistically significant [p = 0.16].
consecutive neonates admitted into the unit during
Babies who experienced hypoglycaemia were
the study period were included except those who had
significantly younger than their counterparts who did
intravenous infusion of fluid(s) that contained
not (p = 0.033). The admission weights ranged
dextrose prior to admission into the unit.
between 600g and 4,100g with a mean of 2,100 ±
Information obtained and entered into a proforma
900g. Hypoglycaemia occurred in 30 (43.5%) of 69
included demographic characteristics of the mothers
babies who were of low birthweight (birth weight less
and their babies, past medical history and illnesses
than 2,500g) compared with 19 (23.5%) of 81 who
during pregnancy especially that of diabetes mellitus
weighed 2,500g or more [p = 0.009].
and hypertension, antenatal booking status, the
The gestational age of babies with hypoglycaemia
mode of delivery, duration and details of the
ranged between 26 to 44 weeks. Hypoglycaemia was
management of labour and place of delivery (for
significantly commoner among preterm babies than
referred babies), birth asphyxia as well as history of
term babies (p = 0.000). The number of post term
feeding prior to admission. Anthropometry, rectal
babies was remarkably smaller than those with
temperature and clinical features were documented.
normal or low values. The frequency of
One millilitre of blood was taken into dry fluoride
hypoglycaemia was more among post-term than term
oxalate-containing specimen bottle and immediately
babies but the difference was not statistically
processed in the laboratory for plasma glucose
significant (p = 0.071). Hypoglycaemia was recorded
estimation using the glucose oxidase reaction
more commonly among twins (13 (41.9%) of 31
method. Other investigations were done as
compared with 36 (30.1%) of 119 singletons though
indicated. Hypoglycaemia, for this study was
the difference was not statistically significant (p =
defined as plasma glucose of = 2.5 mmol/l. Babies
1
0.217).
with hypoglycaemia were managed initially with a
Three (6.1%) of the 49 babies were infants of mothers
bolus of 10 percent intravenous Dextrose-in-water at
with toxaemia of pregnancy.
2 ml/kg. This was followed with 10 percent
Dextrose-in-water at a rate of 6 mg/kg/min (86
ml/kg/24hrs). Babies who had no contraindications
to oral feeding were commenced on feeding while
92
Table 1: Comparison of Socio-demographic Data
Features identified in hypoglycaemic as well as non-
of Babies with and without Hypoglycaemia
hypoglycaemic babies. In comparison however,
Babies with
Babies without
hypoglycaemia was significantly more commonly
Parameters
hypoglycaemia
hypoglycaemia
Total
Statistics
(49)
(101)
150
associated with poor suck, cyanosis, convulsions and
pallor.
Age in hrs (mean ±
38.3 (71.6)
106.0 (162.1)
150
t = 2.79,
SD)
p = 0.006
Sex:
Table 2: Comparison of Clinical Features among
29 (33.0)
59 (67.0)
88
χ
2
Male
= 0.01,
Babies with and without Hypoglycaemia.
Female
20 (32.3)
42 (67.7)
62
p = 0.929
Clinical
Babies with
Babies without
χ
2
hypoglycaemia
hypoglycaemia
Total
p value
Social Class:
features
No (%)
No (%)
χ
2
Higher classes I and
6 (17.6)
28 (82.4)
34
= 4.51,
II
p = 0.034
Floppiness
12 (24.5)
16 (15.8)
28 (18.7)
1.63
0.202
Poor suck
12 (24.5)
9 (8.9)
21 (14.0)
6.65
0.010
Lower classes III-V
43 (37.1)
73 (62.9)
116
Cyanosis
11 (22.5)
9 (8.9)
20 (13.3)
5.23
0.022
Respiratory
Gestational age
distress
8 (16.3)
28 (27.7)
36 (24.0)
2.35
0.125
(wk):
Convulsions
8 (16.2)
6 (5.9)
14 (9.3)
4.21
0.040
< 37
30 (44.1)
38 (55.9)
68
χ
2
= 7.42,
Pallor
7 (14.3)
5 (5.0)
12 (8.0)
3.91
0.048
p = 0.006
High pitched
6 (12.2)
4 (4.0)
10 (6.7)
2.43
0.119*
χ
2
37 to 42
15 (20.3)
59 (79.7)
74
= 10.20,
cry
0 .144
*
p = 0.001
Jitteriness
5 (62.5)
3 (3.0)
8 (5.3)
2.14
0.171
*
χ
2
> 42
4 (50.0)
4 (50.0)
8
= 0.47,
Apnoea
4 (8.2)
2 (2.0)
6 (4.0)
1.87
p = 0.492
+
*
Hypothermia
22 (44.9)
20 (19.8)
42 (28.0)
10.31 0.001
Maternal age
(years):
* =Yates' correction applied
3 (25.0)
9 (75.0)
12
χ
2
<
= 0.07,
p = 0.788
+
20
Figures in parenthesis are percentages of total in
χ
2
20–35
41(32.8)
84 (67.2)
125
= 0.01,
each row
p = 0.938
χ
2
>35
5 (38.5)
8 (61.5)
13
= 0.22,
Admission diagnoses
p = 0.641
*
Maternal parity:
14 (23.7)
45 (76.3)
59
χ
2
= 3.53,
Table 3 shows that preterm delivery, birth asphyxia
Primipara
p = 0.060
and septicaemia were the commonest admission
χ
2
Multipara
31 (37.4)
52 (62.6)
83
= 1.85,
p = 0.173
diagnoses in hypoglycaemic and non-hypoglycaemic
χ
2
4 (50.0)
4 (50.0)
8
= 0.47,
neonates. Hypoglycaemia was significantly more
Grandmultipara
p = 0.492
+
frequent among preterm babies (p = 0.006) and those
16 (26.2)
45 (73.8)
61
χ
2
Inborn (WGH)
= 1.94,
who were small-for-gestational age (p = 0.022).
Referred
33 (37.1)
56 (62.9)
89
p = 0.164
Table 3: Comparison of Diagnoses on Admission in
Keys:
Babies with and without Hypoglycaemia.
The 150 babies were delivered by 136 mothers
Babies
Yates correction applied
with
Babies
Wk=weeks
Diagnoses* on
without
χ
2
admission
hypoglycae
Total
mia No.
Hypoglycae
p value
(%)
mia No. (%)
Clinical features
Preterm
30 (61.2)
38 (37.6)
68
7.42
0.006
Birth asphyxia
19 (38.8)
36 (35.6)
55
0.14
0.709
Septicaemia
8 (16.3)
24 (23.8)
32
1.0\9
0.297
The distributions of the clinical features among
Neonatal
babies with and without hypoglycaemia are
jaundice
4 (8.2)
19 (18.8)
23
2.12
0.145**
compared in Table 2. Twenty-nine (59.2%) of the 49
Severe
anaemia
6 (12.2)
10 (9.9)
16
0.19
0.663
babies with hypoglycaemia were symptomatic
Polycythaemia
7 (14.3)
6 (5.9)
13
2.90
0.088
compared with 28 (27.7%) of the 101 babies without
SGA
8 (16.3)
4 (4.0)
12
5.28
0.022**
hypoglycaemia (p = 0.000). The mean plasma
LGA
2 (4.1)
4 (4.0)
6
0.00
1.000**
glucose of asymptomatic babies was slightly less
Tetanus
1 (2.0)
4 (4.0)
5
0.02
0.897**
than that of symptomatic babies (1.47 ± 0.57 versus
Rhesus
Iso-
0 (0.0)
2 (2.0)
2
0.05
0.816**
1.51 ± 0.64) but the difference was not statistically
immunisation
significant (p = 0.823). The rectal temperature of
babies with hypoglycaemia ranged between 33.70
* = Many babies had more than one diagnoses
and 38.70 C with a mean (SD) of 35.70 (1.10) C.
o
o
**=Yates' correction applied
There was a significantly higher prevalence of
Figures in parenthesis are percentages of total babies
hypoglycaemia among babies with hypothermia (χ
2
with and without hypoglycaemia
= 10.31, p = 0.001). Floppiness, poor suck, cyanosis
SGA= Small for gestational age
and respiratory distress were the commonest clinical
LGA= Large for gestational age
93
Outcome
figures of 38 percent was reported from Nepal, an
Asian country and 38.4 percent Lahore in
Pakistan.
10,11
The outcome of study subjects is presented in Table
4. The prevalence of motor deficit among discharged
However, one major reason that may partly account
hypoglycaemic babies (all spastic) was about ten
for the higher prevalent rates of NNH in the present
times that in other babies (p = 0.009). There were a
study is the use of higher plasma glucose level in the
total of 35 (23.3%) deaths among the 150 babies
definitions of NNH. For example, using cut-off blood
glucose levels of 1.1 mmol/L or 1.7 mmol/L, will
5
studied. Mortality rate was higher among
hypoglycaemic babies but the difference was not
give lower prevalent rates than using blood glucose
levels of 2.2 mmol/L
6-8
statistically significant (p = 0.060).
or plasma level of 2.5 mmol/
L. Even in the same population using lower glucose
10
Table 4: Comparison of Outcome among Babies
levels will exclude babies with higher plasma glucose
with and without Hypoglycaemia.
levels making fewer babies being diagnosed as
Babies with
Babies without
Total
having hypoglycaemia. Another possible reason for
Outcome
hypoglycaemia
hypoglycaemia
No.
χ2
No. (%)
No. (%)
(%)
value
the difference between the prevalent rates of
Discharged
hypoglycaemia may be in the selection of the babies
without
106
studied. Whilst in the present study high-risk babies
motor
27 (55.1)
79 (78.2)
(70.7)
8.51
0.004
deficit
who required hospitalisation for major illnesses were
Discharged
5.09
0.024*
studied, the Nepalese study recruited apparently
with motor
5 (10.2)
1 (1.0)
6 (4.0)
healthy babies from a post-natal ward within the first
deficit
Discharged
0.00
1.000
50 hours of life while Hamid et al in Lahore, Pakistan
against
recruited neonates with known risk factors or
suggestive clinical feature.
10,11
medical
1 (2.0)
1 (1.0)
2 (1.3)
advice
Referred
0 (0.0)
1 (1.0)
1 (0.7)
0.00
1.000
The relatively higher blood glucose cut-off value
used in the present study has the advantage of
Died
16 (32.7)
19 (18.8)
35
3.53
0.060
(23.3)
providing management opportunities for many more
Total
49 (100)
101 (100)
150
(100)
babies who were at risk of hypoglycaemia. This is
clinically beneficial in view of the damaging effect of
Figures in parenthesis are percentages of total in
symptomatic hypoglycaemia on the brain and the fact
each column
that the critical blood glucose level at which damage
χ= 13.0, df = 4, p = 0.01
2
can occur has not been clearly defined.
2,12
Koh et al
12
*WithYates's correction
showed reversible disturbances in evoked potentials
at glucose level below 2.6 mmol/L in asymptomatic
term babies, although no similar studies have been
reported from Nigeria.
Discussion
Findings of higher proportion of babies with
hypoglycaemia among referred babies than inborn
The point-of-admission prevalence of neonatal
babies in the present study is in agreement with the
hypoglycaemia in the present study was 32.7
previous reports of Njokanma and Fagbule. This
6
percent. This was high, especially when compared
could partly be explained on the basis of delay in
with the 6.6 percent reported in 1977 by Omene from
arrival of referred high-risk babies to the hospital; for
Benin,
5
Mid-western Nigeria and 9.5 percent
example, the mean age of outborn babies with
reported in 1994 by Njokanma and Fagbule in
hypoglycaemic was significantly higher when
Sagamu. Sagamu, like Ilesa, is located in South-
6
compared with that of the inborn babies with
western Nigerian. The prevalence observed in the
hypoglycaemia. Very ill babies while waiting to be
present study was also higher than the 20.6 percent
referred or during transfer are hardly fed adequately,
reported by Sexson from the USA and 23.0 percent
7
if at all fed. Such babies are usually subjected to cold
reported by Osier et al from Kenya, an East African
8
injury (from lack of transfer incubators or proper
country. The earlier cited Nigerian reports of 1977
wrappings)
and
may
therefore
develop
and 1994 would suggest a rise in prevalence of NNH
hypoglycaemia. Moreover, the high index of
and the later works from Kenya and USA, though
suspicion and better high-risk neonate identification
from different populations give credence to this
and interventions may be contributory to the lower
trend. The difference between the observed
prevalent rate of hypoglycaemia among inborn
prevalence in the present study and the values cited
babies.
from the more recent USA and Kenyan studies may
Although, hypoglycaemia is a common finding
among infants of diabetic mothers (IDM), none of
13
also be a reflection of higher predisposing factors to
NNH in the locality in which the present study was
the two infants of diabetic mothers seen in the present
done as suggested by Koh et al. Indeed, much higher
9
study developed hypoglycaemia, a finding consistent
94
with that of another Nigerian study. However, the
5
The high prevalent rate of NNH and high mortality
number of infants of diabetic mothers in this study is
rate among babies with NNH in the present study is
too small to have a meaningful conclusion.
an indication that blood glucose estimation should be
The proportion of babies with symptomatic
routine in the care of high-risk neonates. Where
hypoglycaemia in the present study was higher than
standard laboratory facilities are sparse, cheap but
35.7 percent previously documented in Benin and
5
sensitive and reliable strip methods of glucose
40.2 percent in India. The comparison of occurrence
14
estimation may be used. Most of the major risk
of symptoms and signs suggestive of NNH between
factors for NNH in the present study like LBW,
studies is rather difficult since these features are not
hypothermia, birth asphyxia, septicaemia and
specific for hypoglycaemia and different blood
polycythaemia are highly preventable and treatable.
glucose levels have been used to define
Effort to reduce these factors must be pursued
hypoglycaemia. However, babies with features
vigorously. There is need to train and re-train health
suggestive of hypoglycaemia such as hypothermia,
workers, particularly those in the lower tiers of health
cyanosis, convulsion, refusal to suck, and non-
care system on the prevention of NNH. Public
anaemic pallor deserve prompt estimation of blood
awareness on the immediate post-delivery care of the
glucose. Empirical treatment for hypoglycaemia is
newborn and provision of suitable transfer medium
advised if estimation of blood glucose is not available
for referred babies should also be intensified to
or likely to be delayed.
reduce the burden of NNH among out-born babies.
About a third of the babies with hypoglycaemia in the
Affordable and easily accessible maternal and child
present study died. This was almost twice the
health services with improved referral facilities for
mortality rate in non-hypoglycaemic babies. The
high-risk pregnancies will prevent most of the co-
implication is that hypoglycaemia to some extent
morbidities and therefore reduce the high morbidity
contributes to neonatal mortality.
and mortality rates reported in the present study.
References
1.
Cornblath M, Hawdon JM,
7.
Sexson WR. Incidence of
11. Hamid MH, Chishti AL,
W i l l i a m s
A F,
e t
a l .
neonatal hypoglycemia: a
Maqbool S. Clinical utility
Controversies
regarding
matter of definition. J
and accuracy of a blood
definition
of
neonatal
Pediatr 1984; 105: 149-50.
hypoglycemia: suggested
glucose
meter
for
the
8.
Osier FH, Berkley JA, RossA,
operational
thresholds.
detection
of
neonatal
Sanderson F, Mohammed S,
Pediatrics 2000; 105: 1141-5.
hypoglycemia.
J
Coll
Newton
CR. Abnormal
2.
Volpe JJ. Hypoglycaemia and
Physicians Surg Pak 2004;
brain injury. In: Volpe (ed).
glucose concentration on
14: 225-8.
Neurology of the Newborn, 4
th
admission to a rural Kenyan
12. Koh TH, Aynsley-Green A,
Edition. Philadelphia, WB
district hospital: prevalence
Tarbit M, Eyre JA. Neural
Saunders Company 2001;
and outcome. Arch Dis Child
d y s f u n c t i o n
d u r i n g
497-520.
2003; 88: 621-5.
hypoglycaemia. Arch
Dis
3.
Lucas A, Morley R, Cole TJ.
9.
Koh TH, Eyre JA, Aynsley-
Child 1988; 63: 1353-8.
Adverse neurodevelopmental
G r e e n
A .
N e o n a t a l
13. Deorari, A.K, Kabra SK, Paul
outcome of moderate neonatal
h y p o g l y c a e m i a - t h e
VK, Singh M. Perinatal
hypoglycaemia. BMJ 1988;
c o n t r o v e r s y
r e g a r d i n g
outcome of infants born to
297: 1304-8.
definition. Arch Dis Child
diabetic mothers. Indian
4.
Williams AF. Hypoglycaemia
1988; 63: 1386-8.
Pediatr 1991; 28: 1271-5.
in the newborn: a review. Bull
10. Anderson S, Shakya KN,
14. Singh M, Singhai PK, Paul
Wld Hlth Org 1997; 75: 261-
Shrestha LN, Costello AM.
VK, Deorari AK, Sundaram
90.
Hypoglycaemia: a common
KR, Ghorpade MD, Agadi A. .
5.
Omene JA. The incidence of
p r o b l e m
a m o n g
N e u r o d e v e l o p m e n t a l
Neonatal hypoglycaemia in
uncomplicated
newborn
outcome of asymptomatic and
Benin. Nig J Paediatr 1 977; 4:
infants in Nepal. J Trop
symptomatic babies with
19-23.
Pediatr 1993; 39: 273-7.
neonatal hypoglycaemia .
6.
Njokanma OF, Fagbule D.
Indian J Med Res 1991; 94: 6-
Incidence, aetiology and
10.
manifestations of neonatal
h y p o g l y c a e m i a .
N i g
J
Paediatr 1994; 21: 26-8.