S O U TH EA S T A S IA N J TR O P M ED P U B LIC H EA LTH

378 Vol 36 N o. 2 M arch 2005

C orrespondence: S iripen K alayanarooj, Q ueen S irikit
N ational Institute of C hild H ealth, M inistry of H ealth,
B angkok 10400, Thailand.
Tel: +66 (0) 2246-1260-8 ext 3902, 3904; Fax: +66
(0) 2246 6750
E-m ail: sirip@ health.m oph.go.th
IN TR O D U C TIO N
D engue infection is currently one of the
m ost im portant m osquito-borne viral diseases
in the tropical parts of the w orld (W H O , 1999).
The disease is characterized by high fever and
hem orrhagic m anifestations. D engue fever (D F)
and dengue hem orrhagic fever (D H F) are tw o
w ell recognized clinical presentations of den-
gue infections. D H F (N im m annitya, 1997, 2002)
is m ore severe than D F because it has the dis-
tinct pathophysiologic hallm ark of selective
plasm a leakage into the pleural and abdom inal
cavities, w hich can result in shock if the plasm a
leakage is extensive (H alstead, 1997). D elay or
untreated shock can lead to com plications of
hepatic, renal or m ultiple organ failure and death.
M assive bleeding due to dissem inated intravas-
cular coagulation (D IC ) and hepatic failure af-
ter prolonged shock is another characteristic
of severe and com plicated D H F patients be-
fore death ((N im m annitya, 1997, 2002).
The annual reported cases of D F/D H F in
Thailand from 1999 to 2003 w ere quite high,
ranging from 30,000-120,000 cases, w ith case
fatality rates (C FR ) of 0.12-0.21% (B ureau of
Epidem iology, 1999-2003). O bese children are
considered to be higher risk because obese
patients are m ore likely to have com plications
or death (N im m annitya et al 1999; K alayanarooj
et al, 2003a). There are no reported studies re-
garding the nutritional status of D H F patients
and the severity of the illness. M any studies
confirm that m ost D H F patients have good nu-
tritional status and m alnourished children are
less co m m o nly o b served to have D H F
(Thisyakorn and N im m annitya, 1992; N im m an-
IS D EN G U E S EVER ITY R ELATED TO N U TR ITIO N A L S TATU S ?
S iripen K alayanarooj
1
and S uchitra N im m annitya
2
1
Q ueen S irikit N ational Institute of C hild H ealth,
2
D epartm ent of M edical S ervices, M inistry of
P ublic H ealth, B angkok, Thailand
Abstract. A retrospective review of dengue patients adm itted to Q ueen S irikit N ational Institute of
C hild H ealth (previously know n as C hildrens H ospital) from 1995 to 1999 revealed 4,532 con-
firm ed cases of dengue infection; 80.9% w ere dengue hem orrhagic fever (D H F) and 19.1% w ere
dengue fever cases (D F). A m ong the D H F patients; 30.6% had shock. The m ajority of them ,
66.6% , had a norm al nutritional status, w hile 9.3% w ere m alnourished and 24.2% had obesity as
classified by w eight for age. C om pared w ith control patients w ith other diagnoses (excluding
H IV/A ID S patients), m alnourished children had a low er risk of contracting dengue infection (odds
ratio=0.48, 95% C I=0.39-0.60, p=0.000) w hile obese children had a greater risk of infection w ith
dengue viruses (odds ratio=1.96, 95% C I=1.55-2.5, p=0.000). The clinical signs, sym ptom s and
laboratory findings of dengue w ere alm ost the sam e am ong the 3 groups of m alnourished, nor-
m al, and obese patients. The m inor differences observed w ere that in obese children liver en-
largem ent w as found less often; m aculopapular/convalescence rash and elevations of alanine
am inotransferase w ere found m ore often. M alnourished patients had a higher risk of developing
shock (37.8% ) than norm al (29.9% ) and obese patients (30.2% ) (p= 0.000). O bese patients had
m ore unusual presentations: encephalopathy (1.3% ) and associated infections (4.8% ), than nor-
m al (0.5% and 2.7% ) and m alnourished patients (1.2% and 3.1% ). C om plications of fluid over-
load w ere found m ore in obese patients (6.5% ) com pared to norm al (3.2% ) and m alnourished
patients (2.1% ) (p=0.000). The case-fatality rates (C FR ) in m alnourished patients and obese pa-
tients w ere 0.5% and 0.4% , respectively, w hile in norm al patients the C FR w as 0.07% . U nder
and over nutrition D H F patients had either a greater risk of shock or unusual presentations and
com plications, w hich can lead to severe disease or com plications and probably a higher C FR .
D EN G U E S EVER ITY A N D N U TR ITIO N A L S TATU S
Vol 36 N o. 2 M arch 2005 379
nitya, 2002). This study w as planned to dis-
cover w hether nutritional status has any effect
on the severity of the dengue illness.
M ATER IA LS A N D M ETH O D S
R etrospective review of hospital charts of
all adm itted dengue patients at Q ueen S irikit
N ational Institute of C hild H ealth (Q S N IC H , pre-
viously know n as C hildrens H ospital) w as done
for the period of 1995-1999. O nly dengue con-
firm ed cases w ith recorded body w eights (B W )
on adm ission w ere included for analysis. Labo-
ratory confirm ation w as done by the A rm ed
Forces R esearch Institute of M edical S ciences
(A FR IM S ) using the antibody test, enzym e linked
im m unosorbent assay (ELIS A ) and/or hem ag-
glutination inhibition test (H I). The serotype w as
identified by polym erase chain reaction (P C R )
and or virus isolation (m osquito inoculation tech-
nique).
D engue w as classified as dengue fever (D F),
dengue hem orrhagic fever (D H F) or dengue
shock syndrom e (D S S ) according to W H O cri-
teria (W H O , 1997).
D engue patients w ere divided into 3 groups
according to their nutritional status. N utritional
status w as classified by percent ideal body
w eight (IB W ), using the w eight for age stan-
dard grow th curve for Thai children from the
D epartm ent of H ealth (2000). G roup 1 (norm al)-
norm al/m ild protein energy m alnutrition (P EM )
w as a B W =75-110% of IB W . G roup 2 (m alnour-
ished)-m oderate to severe P EM w as a B W =
<75% of IB W . G roup 3 (obesity)-overw eight to
obesity w as a B W >110% of IB W .
S even hundred thirty-four records of pa-
tients adm itted to the Q S N IC H dengue w ard
during the sam e period w ith other diagnoses
(excluding H IV/AID S patients) w ere used to com -
pare nutritional status w ith the dengue patients.
D em ographic data, history, physical exam i-
nation, and laboratory investigations w hich w ere
relevant w ere recorded.
S tatistical analysis of the data w as done
using S P S S for W indow version 10.0.
R ES U LTS
There w ere 4,532 confirm ed dengue pa-
tients adm itted to Q S N IC H from 1995 to 1999.
O f the dengue patients; 3,667 had D H F (80.9% )
and 865 had D F (19.1% ). The m ajority of D F/
D H F patients (65-67% ) had a norm al nutritional
status, w hile 9-11% had m oderate to severe
PEM and 23-24% w ere overw eight/obese. There
w as no statistical difference betw een the nu-
tritional status of the D F and D H F patients (Table
1). M alnourished patients had D S S m ore than
patients w ith a norm al nutritional status. P a-
tients w ith norm al nutritional status and obe-
sity had no difference in the num ber of cases
of D S S (Table 2). C ontrol patients w ith other
diagnoses had significantly m ore m alnourished
patients (19.6% ) and few er obese patients
(12.5% ) (Table 3).
The m ale to fem ale ratios of m alnourished,
norm al, and obese dengue patients w ere 1.09:1,
1.1:1, and 1.1:1, respectively (p = 0.881).
The m ean ages of the dengue and the con-
trol patients w ere 7.93.8 and 5.8 3.5 years,
respectively (p=0.000). The m ean ages of the
Table 1
N utritional status of D F/D H F patients.
D H F D F
O dd ratio !
2
p-value
N % N % (95% C I)
M alnourished 323 8.8 96 11.1 0.78 4.09 0.43
(0.6-1.0)
N orm al 2,452 66.9 566 65.4
O bese 892 24.3 203 23.5 1.01 0.02 0.876
(0.85-1.22)
Total 3,667 100.0 865 100.0
S O U TH EA S T A S IA N J TR O P M ED P U B LIC H EA LTH
380 Vol 36 N o. 2 M arch 2005
Table 2
N utritional status of D H F/D S S patients.
D S S D H F
O dd ratio !
2
p-value
N % N % (95% C I)
M alnourished 122 10.9 201 7.9 1.43 8.4 0.004
(1.11-1.83)
N orm al 732 65.2 1,720 67.6
O bese 269 24.0 623 24.5 1.01 0.03 0.865
(0.86-1.2)
Total 1,123 100.0 2,544 100.0
Table 3
N utritional status of D F/D H F and control patients.
D F/D H F C ontrol
O dd ratio !
2
p-value
N % N % (95% C I)
M alnourished 419 9.2 144 19.6 0.48 47.67 0.000
(0.39-0.60)
N orm al 3,018 66.6 498 67.9
O bese 1,095 24.2 92 12.5 1.96 33.27 0.000
(1.55-2.5)
Total 4,532 100.0 734 100.00
Table 4
A S T range.
A S T (U )
M alnourished N orm al O bese
Total
N % N % N % (% )
0-40 21 5.1 150 5.1 56 5.2 227
(5.1)
>40-200 278 67.6 2,058 69.9 719 67.3 3,055
(69.1)
>200-1,000 112 27.3 736 25.0 293 27.4 1,141
(25.8)
Total 411 100.0 2,944 100.0 1,068 100.0 4,423
p = 0.205
dengue patients w ho w ere m alnourished, nor-
m al, and obese patients w ere 9.73.1, 7.83.7,
and 7.64.0 years, respectively (p=0.000).
The dengue serotypes w ere not different
betw een the patients w ith different nutritional
status (p=0.394). D engue 3 w as the m ost com -
m on serotype found (50.6% ), and dengue 1,
2, and 4 w ere found in 25.8% , 21% , and 2.6% ,
respectively.
The percentage of prim ary and second-
ary dengue infections w ere not different betw een
the patients w ith different nutritional status
(p=0.066). There w ere 22.9% prim ary and 77.1%
secondary dengue infections.
M ost of the presenting signs and sym p-
tom s of D F/D H F (fever, tourniquet test, bleed-
ing m anifestations, nausea/vom iting, abdom i-
nal pain) w ere not different betw een the pa-
D EN G U E S EVER ITY A N D N U TR ITIO N A L S TATU S
Vol 36 N o. 2 M arch 2005 381
tients w ith different nutritional status, except
for liver enlargem ent, w hich w as found less often
in obese patients (86.8% ) com pared to 92%
and 90.1% in norm al and m alnourished patients,
respectively (p=0.000). M aculopapular rash w as
found m ore often in obese patients (10.2% )
com pared to 6.5% and 5% in norm al and m al-
nourished patients, respectively (p= 0.000).
C onvalescence rash w as found m ore often in
obese patients (25% ) com pared to 17.5% and
17.7% in norm al and m alnourished patients,
respectively (p=0.000).
The laboratory findings: W B C , platelet
counts, percent rising hem atocrit (H ct), total
protein, album in and aspartate am inotransferase
(A S T) levels w ere not different betw een the pa-
tients w ith different nutritional status. The per-
centage of patients w ith abnorm al A S T eleva-
tion >200 U w ere not different (p=0.205) (Table
4), but abnorm al ALT elevation >200 U w as found
m ore in obese patients (12.2% ) com pared to
8.7% and 8.8% in norm al and m alnourished
patients, respectivley (Table 5). The m ean ala-
nine am inotransferase (A LT) level w as higher
in obese patients (131 U ) com pared to norm al
(101 U ) and m alnourished (114 U ) patients
(p=0.024).
O bese patients (85.2% ) received less in-
travenous (IV) fluid than norm al (88% ) and m al-
nourished patients (89.8% ) (p=0.040). The du-
ration of IV fluid w as not different betw een each
group of patients and the m ean durations w ere
37.2, 37.5, and 38.4 hours for obese, norm al,
and m alno urished p atients, resp ectively
(p=0.757). O bese patients received m ore IV fluid
(86.6 m l/kg of IB W ) than norm al (70.8 m l/kg)
and m alnourished (63.5 m l/kg of IB W ) patients
(p=0.000). If the calculation w as based on ac-
tual B W , obese patients received less IV fluid
(64 m l/kg) com pared to norm al (76.6 m l/kg) and
m alnourished (85.9 m l/kg) patients (p=0.000).
B lood w as transfused in 4% , 3.7% , and 5.6%
of obese, norm al, and m alnourished patients,
respectively (p=0.269).
O bese patients had m ore unusual presen-
tations and com plications com pared to norm al
and m alnourished patients, such as encephal-
opathy (1.3% vs 0.5% and 1.2% ), associated
infections (4.8% vs 2.7% and 3.1% ), and fluid
overload (6.5% vs 3.2% and 2.1% ) (Table 6).
M ost D H F patients recovered com pletely
except for 8 patients: 2 m alnourished, 5 nor-
m al and 1 obese patient, w ho had som e de-
gree of liver im pairm ent; all of them recovered
w ell w ithin 1 m onth. There w as one obese pa-
tient w ho had hepatic encephalopathy, intra-
cerebral bleeding and stayed for 3 m onths in
the hospital. H e recovered w ith perm anent neu-
rological sequalae, left hem iparesis and a de-
crease in intellectual quotient (IQ ). C ase fatal-
ity rates for D H F/D S S w ere the highest in m al-
nourished patients (0.5% ), follow ed by obese
patients (0.4% ), and norm al patients (0.07% )
(Table 6).
D IS C U S S IO N
In our study, m ost of the D F/D H F cases
Table 5
A LT range.
A LT (U )
M alnourished N orm al O bese
Total
N % N % N % (% )
0-40 210 51.1 1,386 47.1 387 36.3 1,983
(44.8)
>40-200 165 40.1 1,303 44.2 550 51.5 2,018
(45.6)
>200-1,000 36 8.8 256 8.7 130 12.2 420
(9.5)
Total 411 100.0 2,945 100.0 1,067 100.0 4,423
p = 0.000
S O U TH EA S T A S IA N J TR O P M ED P U B LIC H EA LTH
382 Vol 36 N o. 2 M arch 2005
had a better nutritional status than the control
patients w ith other diagnoses, as previously re-
ported (Thisyakorn et al 1992; N im m annitya,
2002). In addition, our study suggests that obese
children are at higher risk (odds ratio=1.96,
95% C I=1.55-2.5, p=0.000), w hile m alnourished
children are at low er risk of contracting den-
gue viruses (odds ratio=0.48, 95% C I=0.39-0.60,
p=0.000). The developm ent of D F/D H F depends
on the host im m une response. M alnourished
children are spared from severe D H F/D S S be-
cause they have a suppressed cellular im m une
response (H alstead, 1997). In contrast to m al-
nourished children, obese children are expected
to have a stronger im m une response than nor-
m al children, so they are at higher risk of de-
veloping D F/D H F. This hypothesis w as sup-
ported by our data.
The m ean age of the controls (5.83.5
years) w as low er than in the dengue patients
(7.9 3.8 years). The control patients w ere
m ostly patients w ith other com m on infectious
diseases, such as diarrhea, pneum onia, and
other viral infections. Their m ean age w as low er
than the dengue because adm ission to den-
gue w ard w as restricted to patients " 2 years
old. The m ean age of the dengue patients in
our study w as the sam e as in previously re-
ports (H alstead, 1997; N im m annitya, 1997,
2002).
The ratio of adm itted cases of D F:D H F w as
Table 6
C om plications and results.
M alnourished N orm al O bese O dd ratio !
2
p-value
(M ) (N ) (O ) (95% C I)
U pper-M :N
N % N % N % Low er- O :N
Encephalo-pathy 5 1.2 14 0.5 14 1.3 2.74 4.01 0.045
(0.86-8.19)
2.78 7.85 0.005
(1.25-6.19)
Fluid overload 9 2.1 98 3.2 71 6.5 0.69 1.13 0.288
(0.32-1.42)
2.08 21.41 0.000
(1.50-2.88) 1
A ssociated infections
a
13 3.1 81 2.7 53 4.8 1.23 0.45 0.500
(0.64-2.30)
1.85 11.84 0.000
(1.28-2.68) 4
A ssociated conditions
b
19 4.5 76 2.5 23 2.1 1.95 6.69 0.009
(1.13-3.36)
0.83 0.62 0.431
(0.50-1.36)
H epatic
c
dysfunction 2 1.6 5 0.7 1 0.1 2.42 1.18 0.278
(D H F) (0.23-14.99)
0.54
(0.01-4.88) 0.32 0.571
D eath
c
(C FR for D H F) 2 0.5 2 0.07 4 0.4 7.63 5.73 0.016
(0.55-105.49)
5.52 4.92 0.026
(0.79-61.06)
a
eg pneum onia, diarrhea, phlebitis, U TI, etc.
b
eg thalassem ia, G -6-P D deficiency, congenital heart disease, etc.
c
Fisher exact test
D EN G U E S EVER ITY A N D N U TR ITIO N A L S TATU S
Vol 36 N o. 2 M arch 2005 383
not different am ong different nutritional status
groups, but m alnourished patients w ere ob-
served to have a greater risk of D S S com pared
to no rm al child ren (o d d s ratio = 1 .4 3 ,
95% C I=1.11-1.83, p=0.004). This is explained
by the sm aller volum e of extra-cellular fluid and
plasm a volum e in the m alnourished patients
(G reenbaum , 2003) so that they developed
shock m ore rapidly w hen a lesser degree of
plasm a leakage occurred. O ther factors m ay
play a role in this and need further study.
The classical clinical pictures of dengue;
fever, positive tourniquet test, bleeding m ani-
festations, leukopenia, throm bocytopenia, ris-
ing H ct, pleural effusion and ascites (W H O , 1997)
w ere alm ost the sam e for different groups of
patients, except for liver enlargem ent, w hich
w as less often palpated in obese patients, pos-
sibly due to the thick abdom inal w all. R ashes,
including petechii, m aculopapular and conva-
lescence rashes, w ere m ore com m only observed
in obese patients. This suggests that obese
patients have a stronger im m une response, since
rashes are usually the result of interactions
b etw een ho st cells and infected viruses
(B oonpucknavig et al, 1979). S om e petechii m ay
be the result of increased m echanical traum a
to skin in obese patients, w ho have a larger
body surface area.
O bese patients had higher m ean A LT (131
U ) levels and larger num bers of patients w ith
A LT>40 U (63.7% ), w hich is higher than has
previously been reported at Q S N IC H (53.5% )
(K alayanarooj et al, 2003). It needs to be studied
further w hether this is related to the m etabolic,
endocrine, or other factors, and w hether this
resulted in the larger num ber of encephalopa-
thy cases observed in obese patients. M ost of
the encephalopathy cases in D H F had a he-
patic cause (N im m annitya et al, 1987). A sso-
ciated infections, pneum onia, diarrhea, U TI, and
phlebitis w ere m ore com m only observed in
obese patients. This m ay be due to m ore com -
plications of fluid overload seen am ong them
and the need for m ore invasive interventions,
m aking them bed-ridden and m ore prone to
nosocom ial infections.
D uring adm ission, 85.2% of obese D H F
patients received IV fluid, w hich w as less than
in the norm al (88% ) and m alnourished (89.8% )
patients. This is likely due to their natural habit
of eating and drinking m ore. The total am ount
of IV fluid needed for these obese individuals
(86.6 m l/kg IB W ) w as m ore than in the norm al
(70.8 m l/kg IB W ) and in the m alnourished (63.5
m l/kg IB W ) patients. Fluid estim ation in obese
patients is m ore difficult and IV fluid based on
B W m ay be too m uch for obese patients and
m ay relate to the higher com plication rate for
fluid overload seen in obese patients. The thick
thoracic w all m ay add to the observed signs
and sym ptom s of fluid overload in these obese
individuals.
The overall C FR w as 0.2% ; C FR seem ed
to be higher in m alnourished (0.5% ) and obese
(0.4% ) patients w hile it w as very low , 0.07% in
norm al nutritional status patients.
In conclusion, m alnourished children have
a low er risk of dengue infection, but if they con-
tract dengue they are at higher risk of devel-
oping D S S . O bese children have a higher risk
of contracting dengue w ith m ore unusual pre-
sentations; encephalopathy, associated infec-
tions and com plications of fluid overload. Fur-
ther study of the im m une, m etabolic, endocrine
and other factors in m alnourished and obese
patients should be done in order to have im -
proved m anagem ent in these high risk children.
A C K N O W LED G EM EN TS
The authors w ould like to thank the A rm ed
Forces R esearch Institute of M edical S ciences
(A FR IM S ) for all the dengue laboratory confir-
m ations.
R EFER EN C ES
B oonpucknavig S , B oonpaknavig V, B ham arapravati
N , N im m annitya S . Im m unofluorescence study of
skin rash in patients w ith dengue hem orrhagic
fever. Arch Pathol Lab Med 1979; 103: 463-6.
B ureau of Epidem iology. D epartm ent of D isease C on-
trol, M inistry of P ublic H ealth, Thailand. A nnual
epidem iological surveillance report 1999-2003.
D epartm ent of H ealth. M inistry of P ublic H ealth, Thai-
land. The references of w eight, height and indi-
cators for nutritional status of Thai children age
1 day to 19 years old. 2000.
S O U TH EA S T A S IA N J TR O P M ED P U B LIC H EA LTH
384 Vol 36 N o. 2 M arch 2005
G reenb aum LA . C o m p o sitio n o f b o d y fluid . In:
B ehrm an R E, K leigm an R M , Jenson H B , eds.
N elson textbook of pediatrics. 17
th
ed. P hiladel-
phia: S aunders, 2003: 191-3.
H alstead S B . Epidem iology of dengue and dengue
hem orrhagic fever. In: G ubler D J, K uno G . eds.
D eng ue and d eng ue hem o rrhag ic fever.
W allingford, N ew York : C A B International, 1997:
23-44.
K alayanarooj S , N im m annitya S , eds. G uidelines for
diagnosis and m anagem ent of dengue infection
2003. N onthaburi: M inistry of Public H ealth, Thai-
land, 2003a.
K alayanarooj S , N im m annitya S . C linical presentations
of dengue hem orrhagic fever in infants com pared
to children. J Med Assoc Thai 2003b; 86 (suppl
3): s673-80.
K urane I, Ennis FA . Im m unopathogenesis of dengue
virus infections. In: G ubler D J, K uno G . eds.
D engue and dengue hem orrhagic fever. W alling-
ford, N ew York: C AB International, 1997: 273-90.
N im m annitya S , Thisyakorn U , H em srichart V. D engue
hem orrhagic fever w ith unusual m anifestations.
Southeast Asian J Trop Med Public Health 1987;
18: 398-406.
N im m annitya S . D engue hem orrhagic fever: diagno-
sis and m anagem ent. In: G ubler D J, K uno G ,
eds. D engue and dengue hem orrhagic fever.
W allingford : N ew York: C A B International,1997:
89-113.
N im m annitya S , K alayanarooj S , Vidhyasuporn A , eds.
G uidelines for diagnosis and m anagem ent of
dengue infection 1999. N onthaburi: M inistry of
P ublic H ealth, Thailand, 1999.
N im m annitya S . D engue hem orrhagic fever: current
issues and future research. Asian-Oceanian J
Pediar Child Health 2002; 1: 1-20.
Thisyakorn U , N im m annitya S . N utritional status of
children w ith dengue hem orrhagic fever. Clin In-
fect Dis 1992; 16: 295-7.
W orld H ealth O rganization. D engue haem orrhagic fe-
ver. D iagnosis, treatm ent, prevention and control.
2
nd
ed. G eneva : W H O , 1997.
W H O . P revention and control of dengue and dengue
hem orrhagic fever. C om prehensive guidelines.
N ew D elhi : W H O S EA R O , 1999.