ORIGINAL

Niger J Paed 2012; 39 (2):63 - 66

Ntia HN

Anah MU

Udo JJ

Ewa AU

Onubi J

Prevalence of hypoglycaemia in

under-five children presenting with

acute diarrhoea in University of

Calabar Teaching Hospital, Calabar

DOI:http://dx.doi.org/10.4314/njp.v39i2.5

hypoglycaemia in under-five chil-

dren presenting with acute diar-

rhoea was 4%. There was no sex

difference. It was commonest

among children of the low socio-

economic class (83.3%). Risk fac-

tors to developing hypoglycaemia

were longer duration of last feeds

greater than five hours and severe

dehydration, both reaching statisti-

cally significant differences

Received: 7th march 2011

Accepted: 27th June 2011

Abstract Background: The clinical

features of hypoglycaemia and se-

vere dehydration are similar, and

these two can occur in a child pre-

senting with acute diarrhoea. Hypo-

glycaemia occurring in a growing

brain is deleterious and must be

detected and treated.

Objective: To determine the preva-

lence of hypoglycaemia among

under-five children presenting with

acute diarrhoea in UCTH, Calabar.

Method: This was a prospective

study of 150 children aged six

weeks to five years presenting with

acute diarrhoea in UCTH, Calabar

from June 1st to October 31st 2008.

Consecutive children who met the

inclusion criteria were recruited

into the study. Blood samples were

collected for random blood sugar

and serum electrotype estimation

using One Touch Ultra Test Strips

Ntia HN (

Ewa AU

) Anah MU, Udo JJ,

Department of Paediatrics

Onubi J

Department of Chemical Pathology,

University of Calabar Teaching

Hospital, PMB 1278, Calabar,

Nigeria.

(

p=0.022 and 0.002; FET respec-

tively). Forty percent of patients

who died had hypoglycaemia con-

stituting 33.3% of patients with

hypoglycaemia.

Conclusion: Children with diar-

rhoea complicated with severe de-

hydration are prone to developing

hypoglycaemia. It causes high mor-

tality and thus this parameter should

be checked for and managed on

time.

Email: mnhappiness237@yahoo.com

2

respectively.

Results: The overall prevalence of

006 model and Flame photometry

Key words: Hypoglycaemia, acute

diarrhoea, under-five children.

Introduction

Acute diarrhoea can be complicated by₁e₀l_-₁e₄ctrolyte im-

balance, dehydration and hypoglycaemia.

Diarrhoea is a common cause of childhood morbidity

and mortality in the developing world accounting fo_-₃r an

Interestingly, the symptoms of severe dehydration and

hypogly₁c₀_-a₁e₂mia are similar, and hence difficult to differ-

1

estimated 1.7 million under-five deaths globally. An

entiate

. Indeed, the two complications may operate

estimated one billion episodes of diarrhoea occur yearly

among u₂ nder-five children in Africa, Asia and Latin

America. The peak age incidence of childhood diar-

in concert with a consequent worsening of morbidity

and mortality, thus an early detection and treatment of

hypoglycaemia in such patients would be expected to

4

12

rhoea is 6-24months. This is attributable to poverty and

improve outcome . The prevalence of hypoglycaemia

ignorance among affected groups who usually constitute

up to 80% of the population.

varies from time to time and from place to place. For

instance, in the same centre in Dhakar different preva-

len₀ce rate were recorded at different times by Bennish et

1

11

It has been observed that in developing countries, in-

cluding Nigeria, childhood diarrhoea is gros_,s₆ly under-

al and Huq et al .

5

reported and the incidence, underestimated. In devel-

The aims of the current study were, therefore, to deter-

mine the prevalence of hypoglycaemia among diarrhoea

admissions in the University of Calabar Teaching Hos-

pital, Calabar as well as identify its risk factors and out-

come in these children.

oping countries, there is an inverse relationship b_-e₉ tween

7

breastfeeding and diarrhoea-related mortality. Diar-

rhoea kills infants directly as a result of shock, and indi-

rectly as a major cause of protein₂_,₅e_,₇nergy malnutrition,

especially marasmic-kwashiorkor.

6

4

Subjects and Methods

trans software was used to import data from Epi-Info.

Descriptive statistics included frequency and percent-

ages were made. Categorical variables were compared

using the chi-square test which was used to identify the

validity of the observed differences in the prevalence.

Fisher's exact test (FET) was used for frequency tables

with cells with small numbers (<5). The level of statisti-

cal significance was P<0.05.

The study was conducted in the Diarrhoea Treatment

and Training Unit (DTU) of the University of Calabar

Teaching Hospital (UCTH), Calabar, where children

with acute diarrhoea are managed. The DTU is a subunit

of the children's emergency room (CHER). The unit was

established in 1995, and provides a 24-hour service. The

medical team in the CHER consists of consultants, sen-

ior and junior residents in₁p₅aediatrics, house officers and

paediatric-trained nurses.

In the DTU, children receive oral rehydration solution

Results

(

ORS), and their mothers are trained on home manage-

ment of diarrhoea. In addition, the DTU offers training

for medical students, nurses, resident doctors and house

officers. It has a median admission of 20-40 patients per

month. Admission peaks are recorded during the dry

season.

One hundred and fifty subjects were recruited for the

study; there were 80(53.3%) males and 70(46.7%) fe-

males, with a male: female ratio of 1.1:1. The mean age

of the subjects was 14.6±10.5months. The peak age

prevalence for diarrhoea was at 2-11months constituting

4

6.7% of the study population. A large proportion

The study was approved by the Ethical Committee of

the Hospital. In addition, a written consent was obtained

from the Care givers.

Consecutive children aged six weeks to 59 months pre-

senting with acute diarrhoea and had not received oral

rehydration solution (ORS) before admission into the

DTU were recruited. Subjects who were less than six

weeks, or those that had received ORS and whose par-

ents did not give consent were excluded.

(78.7%) of the children was of the low socioeconomic

class. Five (3.3%) out of 150 children died during the

study period with hypoglycaemia contributing 40% of

the mortality.

Out of the 150 children that were studied, six children

(4%) had hypoglycaemia, with a male to female ratio of

1:1. Vomiting was present in 81 children among whom

four had hypoglycaemia, while of the 69 children who

did not have vomiting, two had hypoglycaemia. The

difference in vomiting in children with hypoglycaemia

and those without hypoglycaemia was not statistically

significant. P=0.684(FET). Ninety-eight (65.3%) chil-

dren were fed within five hours prior to presentation,

while 52(34.7%) children had not eaten for more than

5hours before presentation. Those who had not fed for

more than 5hours before presentation were more likely

to develop hypoglycaemia than those who ate within

5hours prior to presentation, p=0.022(FET).Table 1

On enrolment, a questionnaire detailing bio-data (age,

sex, parents' educational level and employment), history

of duration of diarrhoea, presence of vomiting, 24hrs

food recall and time of last feed were completed for

each subject. This was done by the principal investigator

who was occasionally assisted by a House officer in

training in the unit.

Information obtained was on a one-on-one basis with

the parent/guardian of the children. Socioeconomic class

(

SEC) was assigned to each subject using Olusanya et al

classification. Physical examination was carried out

Table 1: Blood glucose status and time of last feed prior

to presentation.

1

6

with emphasis on weight, height, pallor, and level of

dehydration.

Blood glucose status Time of last feeds in hours (%) Total

<5hrs

>5hrs

Finger prick was done to estimate random blood sugar

and venous blood taken into a plain bottle for serum

electrolyte estimation. Random blood sugar was tested

using the One Touch Ultra Test Strip 2006 model and

serum electrolyte with Flame photometer model 420

with two channels. Hypoglycaemia was considered at

blood sugar level below 2.2mmol/l, hypokalaemia at

serum potassium level below 3.2mmol/l and hyponatrae-

mia at serum levels of sodium below 132mmol/l as well

as metabolic acidosis at serum level of bicarbonate be-

low 22mmol/l. These are reference values for our labo-

ratory. All the equipment used were standardized on a

daily basis before use.

Hypoglycaemia

Nomoglycaemia

1(16.7)

97(67.4)

5(83.3)

47(32.6)

6(100)

144(100)

Total

98(65.3)

52(34.7)

150(100)

Fisher’s exact = 0.022

Of the six who had hypoglycaemia, two (33.3%) were

moderately dehydrated while four (66.7%) had severe

dehydration. The difference was statistically significant

between children who had hypoglycaemia and dehydra-

tion and those who did not have dehydration, p=0.002

(FET). Table 2.

Data collected were checked for accuracy and entered

into a computer. Epi-Info (version 2002) was used.

Data analysis was done using stata version 10. A Stat-

6

5

1

1,12,

Table 2:Blood glucose and hydration status.

these respective studies

the difference in prevalence

was not due to differences in methodology or sensitivity

of the instrument since they were comparable. However,

this prevalence is comparable to the 4.5% reported by

Blood glucose status

Hydration status

Mild Moderate Severe Total

1

0

Hypoglycaemia

Nomoglycaemia

0

9

2

4

6(4%)

Bennish et al also in Dhakar.

There were slightly

13

122 144(96%)

more males than females admitted into DTU in this

study, though not statistically significant. This was simi-

lar to the 1.4:1 male to female ratio recorded on child-

hood diarrhoea admissions in the same centre four years

Total

9

15

126 150 (100%)

Fisher's exact = 0.002

1

5

earlier. The reasons for this male preponderance are

not quite clear.

Five children (83.3%) out of the six subjects who had

hypoglycaemia were from the lower socioeconomic

class. The relationships between hypoglycaemia and

serum electrolytes did not reach statistically significant

difference. Only one out of ten children who had hypo-

natraemia also had hypoglycaemia, p=0.340(FET).

Nineteen children had hypokalaemia out of whom one

also had hypoglycaemia, p=0.558(FET). Out of the sixty

1

7

The assertion by Mitra et al of de₈lay in presentation to

1

hospital of females nor Madrigal assertion that males

are the pillars of the family cannot stand the inheritance

system in Calabar, an Efik tribe predominant community

where both sexes have equal rights (personal communi-

cation with parents and some indigenes).

-six children who had metabolic acidosis, five children

had hypoglycaemia, p=0.086(FET).

More than half of our subjects were below 24months of

age in this study, which agrees with other reports that

diarrhoe₄_,a₅_,₉i_,s₁₁m_,₁₅ore common in children aged less than 24

Although five out of six children who had hypoglycae-

mia also had metabolic acidosis, but this was not statisti-

cally significant. This is most probably due to the small

number (5) compared to the large number of children

months

. Many reasons have been advanced for

this high incidence in children less than 24months, one

of which is the causative agent that is common in this

age group. Our study did not assess the causative agent.

Weaning food and poor hygiene have also been impli-

cated.

(

61) that had metabolic acidosis alone.

The mean length of stay in the hospital was 3.8days

(SD=4.1).Three (2.1%) children out of the 144(97.9%)

children who did not have Hypoglycaemia died while

two (33.3%) of the six children who had hypoglycaemia

died. The difference in death in children with or without

hypoglycaemia was statistically significant p=0.013

Hypoglycaemia was more in the low social class in this

survey. This group is exposed to the vicious cycle of

poverty, hunger and disease; hence it was not surprising

that hypoglycaemia was commonest in this group.

(

FET). Table 3.

Furthermore, parents tend to withhold feeds on the be-

lief that it would reduce the frequency of diarrhoea espe-

cially when accompanied with vomiting hence the possi-

ble reason while hypoglycaemia was common. Longer

duration of last feeds was found to be a risk factor to the

development of hypoglycaemia. This is consistent with

Table 3: Outcome of the study population. (N=150)

Outcome

Hypoglaecemia

Total FET

Present

Absent

Duration of

admission (days)

1

2

the earlier observations of Reid and Losek . The prob-

able explanation may be due to the fact that these pa-

tients have low levels of gluconeogenetic substrate, thus

when counter regulatory hormones are secreted, the gly-

cogen stores are easily depleted and so enough glucose

<

2 days

Compared

2 days

1(1.9)

5(5.2)

4(2.8)

52(98.1)

92(94.8)

53

0.304

97

>

Discharge

141(97.3) 145

1

is not produced . More so, our patients had not received

any form of oral rehydration solution prior to presenta-

tion which could further explained the reduced blood

sugar level.

Compared

Death

0.013

2(40.0)

3(60.0)

5

Figures in bracket represent percentage.

There was no statistically significant association be-

tween vomiting and the development of hypoglycaemia

in children with acute watery diarrhoea. However, there

is a tendency to withholding feeds from these children,

to reduce vomiting and this could lead to starvation and

hence hypoglycaemia. It is both difficult to know the

quality as well as the quantity of food consumed espe-

cially in our community where food items are consumed

at random and children commonly share food as group

consumption from one plate. Thus no significant asso-

ciation was found with hypoglycaemia. In our study,

most of our patients were below two years, thus the

Discussion

The overall prevalence rate of hypoglycaemia among

children with diarrhoea in the DTU of the University of

Calabar Teaching Hospital was 4%. This prevalence is

comparatively low₁ er than 11% and 9.2% recorded ear-

1

lier by Huq et al in Dhakar, as well as that of Reid and

1

2

Losek in Minnesota USA, respectively. Though our

sample size was smaller as compared to 196 and 782 in

6

1

presence of other co-morbidities like malaria, pneumo-

nia and septicaemia could increase their chances of hav-

ing hypoglycaemia as earlier reported

the 28% reported by Huq et al . This shows that hypo-

glycaemia must be identified and treated in children

with diarrhoea.

1

1,19

.

A significant association was found between severe de-

hydration and hypoglycaemia. Hypoglycaemia observed

in these children may have been due to reduced blood

supply to the liver with subsequent reduction in glucose

metabolism. The clinical importance of this finding is

that children who are admitted with diarrhoea compli-

cated by severe dehydration should have their blood

sugar checked as they are likely to develop hypoglycae-

mia.

Conclusion

The prevalence of hypoglycaemia in diarrhoea admis-

sions was 4%. Longer duration from time of last feeds

and severe dehydration were risk factors to developing

hypoglycaemia. Mortality was high in the children who

had both diarrhoea and hypoglycaemia. Thus there is

need to intensify education on continued oral feeding

and administration of ORS during diarrhoea episodes in

children to reduce severe dehydration and subsequent

hypoglycaemia.

Generally the association between serum electrolytes

and hypoglycaemia was not statistically significant but

most of our subjects with hypoglycaemia had metaboli₂c₀

acidosis. This agrees with the study by Wathen et al

who reported abnormally low levels of bicarbonate as a

significant finding associated with hypoglycaemia in

children who were severely dehydrated, as well as those

requiring intravenous fluid administration.

Conflict of interest: None

Funding: None

The overall case fatality of children with diarrhoea and

hypoglycaemia was 33.3%. This was a little higher than

References

1

. Lancet. Global, Regional and Na-

tional causes of child mortality in

8. Yakubu AM. Diarrhoea in children.

15. Ochigbo SO, Anah MU, Udo JJ,

Uniga JA, Ogon PM. Childhood

diarrhoea in Calabar: An 8-year

review. Global J Med Sci 2007; 6:

85-90.

16. Olusanya O, Okpere E, Ezimokhai

M. The importance of social class

in voluntary fertility control in a

developing country. W Afr J Med

1985; 4:205-12.

17. Mitra AK, Rahim KMM, Fuchs GJ.

Risk factors and gender differen-

tials for death among children

hospitalized with diarrhoea in

Bangladesh. J. Health Pop Nutr

2000; 18: 151-6.

18. Madrigal L. Differential sex mortal-

ity in a rural nineteenth century

population: Escazu, Costa Rica.

Human Biol 1992; 64:199-213.

19. Elusanya JBE, Adejuyibe EA,

Adeodu O. Hypoglycaemia in a

Nigerian Paediatric Emergency

Ward. Obafemi Awolowo Univer-

sity, Ile Ife, Nigeria. J Trop Pedi-

atr 2005; 52: 96-102.

In: Azubuike JC and Nkanginieme

KEO eds. Paediatrics and child

health in a tropical region 2nd Ed.

African Educational Services Ni-

geria. 2007; 268-282.

9. Cardenos de Saude Publica. Risk

factors for diarrhoea disease in

Brazilian infants: A special case

control design application. 2002;

18(3) P464.

10. Bennish ML, Azad AK, Rahman O,

Philips RE. Hypoglycaemia

during diarrhoea in Childhood.

Prevalence, pathophysiology and

outcome. International Centre for

Diarrhoea Disease Research,

Bangladesh, Dhaka. N Engl J Med

1990; 322: 1357-63.

11. Huq S, Hassan MI, Malek MA,

Faruque AS, Salam MA. Hypogly-

caemia in Under Five Children

with Diarrhoea. J Trop Pediatr

2007; 53: 197-201.

12. Reid SR, Losek JD: Hypoglycaemia

complicating dehydration in chil-

dren with acute gastroenteritis. J.

Emerg Med 2005; 29: 141-5.

13. Elliot EJ and Dalby-payne JR.

Acute infectious diarrhoea and

dehydration in children. Med J

Aust 2004; 181(10):565-570.

14. Gastroenteritis in Children. Pub-

lished by Bupals Health Informa-

tion

2

008: a systematic analysis. Lancet

2

010; 375(730): 1969-87.

2

. Margaret K, Caryn B, Richard L.

Global burden of diarrhoeal Dis-

ease. In: Bull Wlto Hlth Org.

8

1:197-204 Geneva 2003.

3

4

.

Grant JP. The State of the World's

Children 2008. UNICEF, New

York.

. Robert C. Diarrhoeal disease in

early childhood. In: Child Health

in the Tropics, Jellife DB, Mill-

Road, Duntron Green, Sevenoaks.

(

eds) 5th Ed Kent TN 132YA

Publishers) 2001; 102-9.

5

6

7

.

Carl ET, William BG. Control of

diarrhoeal diseases. http://

www.annual reviews. Org/

aronline. Assessed May 30th.

2

010.

. Ogbonnaya O, Nebe A, Chigozie U,

Ekperechi SA; Aetiology of acute

infantile diarrhoea and antibiotic

sensitivity profile. Intern J Third

World Med 2008 ; 5: 1539-4646.

. Asindi AA, Meremikwu MM, Antia

20. Wathen JE, Todd M, Joan PB. Use-

fulness of serum electrolyte panel

in the management of paediatric

dehydration treated with intrave-

nous administratered fluids. Pub-

lished online Nov 2004; 114(5):

1227-1234. Available as http://

pediatrics.aappublications.org/cgi/

content/abstract/114/5/1227.

Accessed Jan 30, 2011.

-Obong OE; Influence of breast-

feeding on the occurrence of dys-

entery, persistent diarrhoea and

malnutrition among Nigerian chil-

dren. W Afr J Med. 1997;16: 20-

2

3.

Team, 2008.