J. VAN DEN HOMBERGH ET AL.
Does Iron Therapy Benefit Children with Severe
Malaria-associated Anaemia? A Clinical Trial with
12 Weeks Supplementation of Oral Iron in Young
Children from the Turiani Division, Tanzania
by J. van den Hombergh,* MD MSc, E. Dalderop,** MDam/Y. Smit, MD*
* Diocesan Health Programme Morogoro, PO Box 119, Mikumi, Tanzania
**St Anna Hospital, Oss, and Institute of International Health, University of Nijmegen, The Netherlands
Summary
Oral iron supplementation is often routinely ghen to children with malaria-associated anaemia, bat its
contribution to recovery is controversial. A randomized clinical trial, evaluating such routine, was
carried out among 100 children, who had a haemoglobin of <5 g/dl and a positive blood smear for
malaria parasites. AD children received malaria therapy (chloroquin + fansidar) and were randomly
allocated to two groups, one receiving additional oral iron treatment, the other being the control. In the
12-week follow-up period the haemoglobin level and malaria indices were measured at 2, 4, 8, and 12
weeks. There was a 100 per cent compliance during the follow-up period. In each group 20 children (40
per cent) required a blood transfusion. In the remaining 60 children, after 2 weeks the haemoglobin had
risen 3.7 g/dl in the ferrous-supplemented group compared to 3.5 g/dl in the non-ferrous group.
Thereafter, the increase in haemoglobin in both groups was steady. At follow-up measurements, the
groups did not differ for haemoglobin levels. The mean haemoglobin at 12 weeks was 9.2 and 9.0 g/dl,
respectively. It was concluded that iron supplementation did not have any effect on the rate of
parasltaemia and on parasite density during the 12 weeks. However, the iron-supplemented group had a
significantly increased morbidity from other causes than malaria. It appears that iron does not have an
effect on the recovery of haemoglobin level in children with malaria-associated anaemia. This study
provides no evidence supporting routine iron supplementation to these children.
Correspondence: Jan van den Hombergh, Nonnenplaats 10,
65-11-VM Nijmegen, The Netherlands; Tel. (31)80-606015;
Fax: 3180606015.
much as 50 per cent or more of the all infant
mortality in malaria-endemic areas may be attributable to malaria.3 Chloroquine resistance has become
widespread.4 Recently developed effective drugs are
prone to rapidly developing resistance of the parasite
and their cost may be prohibitive for widespread use
in the most affected countries. Anaemia and cerebral
malaria are the foremost life-threatening complications of clinical malaria in children. The anaemia
associated with malaria is the result of a combination
of different mechanisms affecting erythrocytes, such
as destruction by the parasite, vascular sequestration
of parasitized cells, immune elimination, sequestration by the spleen and defective erythropoiesis.5"7
The clinical presentation of a child with severe
anaemia is variable and directly related to the pace at
which the anaemia developed. If the anaemia
worsened gradually following chronic parasitaemia,
the child may have little or no symptoms and signs,
apart from pallor. If the fall in haemoglobin has
occurred more rapidly, the child will present with
signs such as dyspnoea and intercostal retractions,
rapid pulse, enlarged liver, and a palpable spleen.
This is an essential clinical problem, since it has
220
Journal of Tropical Pediatrics
Introduction
Malaria causes 100 million morbid episodes and 1
million deaths yearly worldwide.1 The disease causes
anaemia in children and pregnant women, and
increases vulnerability to. other diseases. It takes the
life of one out of 20 children before the age of 5 years
in tropical Africa.1 Dramatic though these figures
are, they are likely to be a considerable underestimate
of the true situation.2 It has been estimated that as
Acknowledgements
We would like to express our gratitude to the management
team, nursing, and laboratory staff of the Turiani Hospital
for their commitment to the many anaemic children and
their strong co-operation in this study, to Patricia Dalderop
for her assistance in the field work, to Dr G. T. Heikens for
his advice during the design, analysis and write-up of this
study, to Dr W. Deville for assistance in the statistical
analysis of the nutritional indicators, and to S. Cousins for
his review of earlier drafts.
© Oxford University Press 1996
Vol. 42
August 1996
I. VAN DEN HOMBERGH ET AL.
consequences for the management of the patient: viz.
blood transfusion or not.
If the anaemia is acute and related to severe
malaria infection, a further drop in haemoglobin
level may be expected within 1 or 2 days, necessitating a blood transfusion, whereas a patient with past
or low grade infection may already berecoveringand
thus require drug treatment only. Recent data from
Ifakara, Morogoro Region Tanzania, suggest that in
the majority of admitted children with anaemia
(Hb < 5 g/dl) and malaria; this anaemia results from
an acute drop in haemoglobin, rather than being due
to a gradual decline in haemoglobin.8
Until recently, the WHO recommended a blood
transfusion for any patient with a haemoglobin
below 5 g/dl. In pediatric wards of hospitals in
endemic malaria areas numerous blood transfusions
were routinely given, with a considerable risk of
complications. A recent study of records of 497
blood recipients in one region of Tanzania concluded
that 62 per cent of transfusions to under-fives was
avoidable.9 Introduction of a protocol in Malawi
reduced transfusion rates from 44 to 11 per cent in
under-fives, while mortality rates remained the
same.10 Since the recognition of the HTV transmission risk, most health facilities have tried to reduce
blood transfusions and have reviewed the indications.
The value of haematinic drugs like ferrous and
folk acid is subject to controversy, but these are often
routinely prescribed for variable periods after initial
treatment. The interaction between iron, malaria,
and malaria-associated anaemia has been the subject
of a longstanding debate. In spite of a wealth of
research, a clear understanding of the role of iron in
the pathogenesis of malaria and its sequelae has not
been achieved. As early as 1868 Armand Trousseau
noted the deleterious effect of iron repletion on
pulmonary tuberculosis."
The adverse effect of iron on the course of
infections has been studied extensively.12"13 As
malaria causes predominantly a haemolytic anaemia,
it would not be expected to lead to iron deficiency.
Anaemia caused by Plasmodium falciparum has been
described as either normocytic, megalocytic, or
microcytic, but the anaemia which follows a single
defined episode of malaria is invariably normocytic in
character. Precisely how repeated or persistent
malaria induces changes consistant with iron deficiency is not clear. It may do so by depressing
absorption of iron.M Iron metabolism is disturbed
during acute episodes of malaria and the possible
effects of recurrent asymptomatic malaria infections
on iron metabolism need further research.15
Oral or intra-muscular iron supplementation is
common practice in the prevention and treatment of
iron deficient anaemia, but it was not until 1970 that
it was noticed that large quantities of intravenous
iron dextran given to pregnant women increased the
Journal of Tropical Pediatrics
Vol. 42
August 1996
frequency of malaria attacks.16 In the following
years, inconsistent evidence emerged from a number
of studies.17"24 In the Turiani Hospital and its
surrounding region in Tanzania, oral iron is routinely
prescribed for infants and young children with
moderate and severe anaemia.
Given the conflicting views concerning the adverse
effect of iron treatment on malaria infection, a trial
with oral iron was carried out among children
younger than 30 months with severe malariaassociated anaemia.
Methods
Study area and population
Turiani Hospital is a 200-bed voluntary agency
hospital situated in the village of Bwagala, 3 km
from Turiani, in the Morogoro Region of eastern
Tanzania. It functions as a district hospital in the
direct catchment area. The Morogoro Region can be
generally regarded as consisting of a few 'islands' in
which malaria is slight or absent, set in an area of
holoendemic transmission. The mean annual rainfall
is 880 mm The main rainy season is between April
and June. Short rains usually occur between October
and January. Permanent rivers flowing from mountain ranges support not only the human, but also the
anopheline populations of small towns and villages,
such as Turiani, situated at the junction of the
foothills and the dry plain. All four species of malaria
parasite occur, P. falciparum constituting more than
90 per cent of all infections.23
The main vectors are Anopheles gambiae and
Anopheles funestus. In the humid lowlands, which
make up most of the Morogoro Region, villagers may
receive five or more infective bites per night. It has
been observed in another endemic malaria area in
Tanzania, that sporozoite inoculation rates reached
about six per week per person. After effective parasite
clearance with pyrimethamine/sulphadoxin (Fansidar), almost all children becamere-infectedwithin 24 weeks after disappearance of the protective effect of
this drug.26 In north-east Tanzania, a recent study
documented chloroquine resistant infections (RI 53
per cent, RII 2 per cent, RIII 4 per cent) in underfives.27 In addition, chloroquine resistance is assumed
to have contributed to the increased human infectiousness to blood-feeding mosquitoes and increased
infectivity of the malaria parasite in north-east
Tanzania.
The chloroquine resistance in the Turiani Division
was evaluated.29 RII and RUI resistance in vivo was
found in 74 per cent of children under-five. A recent
survey in the Turiani Division showed that parasitaemia rates in healthy infants and young children
range from 40 to 80%. The mean haemoglobin was
found not to exceed 9 g/dl and was significantly lower
in parasitaemic children as compared with parasitefree children.30
221
J. VAN DEN HOMBERGH ET AL.
Subjects
Children, who live within the Turiani Division, were
recruited from 2 April to 10 June 1993, at the outpatient department (OPD) and paediatric ward of the
hospital. This area was chosen to achieve full
compliance, the maximum distance between residence and the hospital being 30 km. The selection
procedure took place during normal opening times.
Children from within the division, younger than 30
months, with a haemoglobin <5 g/dl and having
malaria parasites in the thick blood smear were
eligible. Children excluded from the study were those
with cerebral malaria, non-falciparum malaria, sickle
cell anaemia, and children fulfilling the criteria, but in
whom the malarial anaemia was not the main
medical problem (e.g. meningitis, measles).
Methods and materials
The blood smear for asexual malaria parasites was
counted against 100 leucocytes. The density per /zl
was calculated by assuming 8000 leucotyes per /J
blood. Haemoglobin was measured in g/dl from
capillary blood, using a Hemocue spectrophotometer. After obtaining verbal informed consent from
the caretaker, the children were allocated to two
treatment groups by simple randomization.
Baseline information was collected and additional
haematological tests were performed. Haematocrit
was measured as a percentage, using heparinized
capillaries (Hawksley) an a micro-Ht centrifuge. The
reticulocytes were stained with 1 per cent Brilliant
Cresyl Blue in a citrate buffer, incubated at 37°C for
15 min, counted against 250 red blood cells (RBCs)
and expressed as proportion of 1000 RBCs. Name,
age, sex, and village of residence were recorded.
Bodyweight (in kg), axillary temperature (in °Q,
respiratory rate (RR), pulse rate, and spleen index
(cm below costal arc) were measured.
A standard set of criteria for pediatric blood
transfusion in the hospital was used to assist in
rational decision making. These criteria are: a
haemoglobin of 3.9 g/dl or less combined with a
respiratory rate >60/min, a pulse rate > 120, and
clinical signs of respiratory distress, e.g. grunting,
nasalflaring,inter(sub)costal recessions, and gasping.
Follow-up examination of all study children was
carried out at the hospital's child health clinic (MCH)
after 2, 4, 8, and 12 weeks. If the mothers did not
return for follow-up, they were visited at home. The
mothers were encouraged to attend the clinic in
between if the child had symptoms of malaria or was
otherwise ill. Attendance and treatment were free of
charge. At each visit a short history was taken and a
physical examination performed. The remaining
iron/folic acid medication was counted and discussed
with the mother in order to ascertain compliance.
Haematological data collected during follow-up
included haemoglobin, haematocrit, reticulocyte
count and a blood smear for malaria parasites. In
case of clinical symptoms, treatment was provided.
After 2 weeks the respiratory rate was counted and at
the end of the follow-up period, at 12 weeks, the
spleen and bodyweight were measured.
Data storage and analysis
All data were entered in Epilnfo 6.0 and analysed,
using ANOVA, Bartlett's test for homogeneity of
variance and Kruskal-Wallis test. Linear regression
was used to analyse the association between continuous variables. Analysis of covariance was performed to analyse individual changes over time.
Treatment protocol
All children were treated with the standard oral
second-line malaria drug regimen in the hospital:
Quinine sulphate, 10 mg per kg, three times daily for
3 days and Fansidar (sulphadoxin 500 mg with
pyrimethamin 25 mg), as a single dose (4-6.5 kg: 1/2
tablet; 6.5-8.5 kg: 3/4 tablet; 8.5-11 kg: 1 tablet).
In addition, all children received haematinic drugs:
Group I (iron group): ferrous sulphate 200 mg daily,
and folic acid, 1 mg daily, for 12 weeks. Group n
(controls): folic acid, 1 mg daily, for 12 weeks.
After enrolment, each patient was managed
according to the clinical condition. This varied from
treatment as an out-patient to being admitted and
receiving additional treatment by the medical officer
in charge of the pediatric ward. Forty children
required a blood transfusion. The decision to give a
blood transfusion was made by the medical officer at
the pediatric ward and based on a clinical assessment
of the child.
Results
A total of 100 children entered the study. The male/
female ratio was 45/55. Baseline data are given in
Table 1. Geometric means are used where data were
skewed. Differences in values for the baseline
variables between the ferrous and non-ferrous group,
are observed for the haemoglobin. Of those receiving
iron, 50 per cent had a haemoglobin of less than 4 g/
dl, as compared with 32 per cent in the control group.
In addition, a difference in the rate of palpable
spleen (94 per cent in the ferrous groups v. 82 per cent
in the control group) was noticed. Geometric mean
parasite densities are higher in the ferrous group, but
the difference is not significant. The age distribution
of the children is given in Fig. 1. Mean haemoglobin
values are presented in Table 2. Four children were
excluded, whoreceiveda blood transfusion (BT) later
than day 1 or 2. There are no significant differences in
the change in haemoglobin values between the iron
and non-iron supplemented children in each category. There is an expected relatively large increase in
haemoglobin level in the first 2 weeks for those who
received a blood transfusion. The benefits of a BT
have largely disappeared by week 4. The overall
222
Journal of Tropical Pediatrics
Vol. 42
August 1996
J. VAN DEN HOMBEROH ET AL.
TABLE 1
Values of base-line variables of the study population
Variable
Ferrous (n - 50)
No ferrous (n-50)
Total (n= 100)
Hb 2-2.9 g/dl
Hb 3-3.9 g/dl
Hb4-5
6 (12%)
19 (38%)
25 (50%)
5 (10%)
11 (22%)
34 (68%)
11(11%)
30 (30%)
59 (59%)
Mean HB g/dl (SD)
4.0 (0.8)
4.2 (0.7)
4.1 (0.8)
8536
(5021-14510)
48%,
(" = 45)
4403
(2610-7426)
53V>
(/i = 49)
6039
(4340-8880)
51%.
(n = 94)
57(17)
37.9 (0.9)
128 (21)
56(16)
37.7 (1)
126 (19)
56(16)
37.8 (0.9)
127 (20)
Geom. mean pst. dens, (pi)
confidence interval
Geometric mean count of
reticulocytes (1000 RBCs)
Mean respiratory rate/min (SD)
Mean Temperature (°C) (SD)
Mean pulse rate/min (SD)
Palpable spleen
Mean Z-score (W/age) (SD)
47 (94%)
41 (82%)
88(91%)
-1.29(1.14)
-1.33(1.08)
-1.31(1.11)
20 (40%)
20 (40%)
40 (40%)
Blood transfusion day 0
TABLE 2
Mean haemoglobin (SD) at each date of follow-up and mean increase (SD) in
haemoglobin level in gr/dl after each period between two dates offollow-up
Blood transfusion YES
fer+ (n = 20)
f e r - (n-20)
Blood transfusion NO
fer+ (n = 218)
f e r - (n = 28)
Day 0 Mean Hb
3.3 (0.6)
3.5 (0.6)
4.4 (0.5)
4.6 (0.3)
Week 2:
Mean
Mean increase
9.4 (1.1)
6.1 (1.0)
9.6 (2.1)
6.1 (2.1)
8.1 (1.4)
3.7 (1.3)
8.1 (1.4)
3.5 (1.7)
Week 4:
Mean
Mean increase
9.7(1.5)
0.3 (1.3)
9.9(1.5)
0.3 (2.1)
8.9 (1.2)
0.8 (1.3)
8.7 (1.8)
0.6(1.5)
Week 8:
Mean
Mean increase
8.6 (2.8)
-1.1(2.9)
8.4 (1.8)
-1.5(2.1)
9.1 (1.8)
0.1 (2.1)
8.1 (1.9)
- 0 . 6 (2.5)
Week 12:
Mean
Mean increase
10.1 (1.5)
1.5 (2.6)
9.4 (2.1)
1.2(1.5)
9.2(1.5)
0.2 (2.4)
9.0 (1.5)
0.9 (2.0)
relative decrease in haemoglobin level between weeks
4 and 8 may be due to increasing re-infections with
malaria, the protective effect of Fansidar having
disappeared. Figure 2 illustrates these changes in
haemoglobin levels.
The geometric meanreticulocytelevels are given in
Table 3. The reticulocyte counts are highest at the
start of the study, when haemoglobin is at its lowest
levels. The reticulocytes gradually drop to their
lowest level at week 4, when the haemoglobin is at
its peak.
The geometric mean reticulocyte counts in both
Journal of Tropical Pediatrics
Vol. 42
August 1996
groups did not differ. Parasitaemia rates and
haemoglobin levels for each measure point are given
in Table 4. The parasite densities are reproduced as
geometric means, as can be seen from Table 4 and
Fig. 3. No significant differences in malaria indices
were observed between children with or without iron
supplementation. There were significant differences
between mean haemoglobin levels of parasitaemic
and non-parasitaemic children at 2, 8, and 12 weeks,
confirming the association between parasitaemia and
level of haemoglobin. No association was found
between haemoglobin levels at any moment and
223
J. VAN DEN HOMBERGH BT AL.
0-2
3-5
6-8
•III..
9-1112-14 15-17 18-20 21-23 24-26 27-29
The changes in nutritional status at enrolment and
at week 12 were expressed in Z-scores (weight for
age, Table 5). The mean Z-score decreased in the
ferrous-group, whereas an increase was seen in the
control group.
This difference, controlled for the baseline Zscores, was not significant. During follow-up visits,
but also in case of attendance in between control
visits, clinical assessment of the child was carried out
by one of the authors and a diagnosis was established
according to a specific set of criteria. The number of
extra attendances for clinical care in the ferrous
supplemented group was hgher than in the control
group (14 v. 6, i>=0.05). The total number of
diagnoses made was also higher in the ferrous
supplemented group (107 v. 65, />=0.004).
A significant difference in cause-specific morbidity
between the groups was observed for pneumonia,
which was more frequent in the ferrous-group (26 v.
5, P=0.004). For all other types of scored morbidity
no differences were observed. In each group one
patient died, an overall mortality risk of 2 per cent.
The first child died 14 days after admission, having
recovered from the malaria and anaemia. He developed a fast-spreading ulcer on the mouth and died of
septicaemia. The other child died at home, 55 days
after enrolment; her condition was good at last
follow-up. Verbal autopsy suggested cerebral malaria
or meningitis.
30
Age (months)
FIG. 1. The age distribution of the study population.
Hbnncrease ferrous group
•
Wean Hb ferrous group
-•- Mean Hb non-ferrous group
day 0
Hb-increase non-ferrous group
week 2
week 8
week 12
Discussion
This study supports observations by others in
Tanzania, and elsewhere,23 that severe anaemia is
a complication of malaria which appears to be most
prevalent between ages of 3-18 months. There is a
peak prevalence in the second half of the first year of
life. At this period the protection by passive
immunity after birth has ceased and the child will
gradually develop an acquired immunity following
repeated infections.
The most striking observation in this study is the
rapid recovery of the haemoglobin levels after
FIG. 2. Mean haemoglobin (g/dl) and the mean
increase after each interval.
parasite densities or reticulocyte counts. Interestingly, the haemoglobin level at week 2 was not
associated with reticulocyte counts at day 0. The
spleen rate at day 0 was 91 per cent. In both groups
the spleen rate dropped to 60 per cent. The spleen size
at day 0 was associated with parasite density
(Rruskal-Wallis test, P=0.0035), no association
was observed at 12 weeks.
TABLE 3
Geometric mean reticulocyte counts (per 1000 RBCs) for each category at times of followup. The geometric mean is presented because the data were skewed
224
Intervention category
Day 0
Week 2
Week 4
Week 8
Week 12
All ferrous + ( n - 5 0 )
All ferrous- (n°50)
47.7
53.5
31.0
29.4
18.1
17.7
22.4
30.3
22.6
24.7
B.T. at start of study
Ferrous + (/i = 20)
Ferrous- (n = 20)
39.6
53.0
37.3
27.0
18.6
21.6
37.4
29.6
18.0
29.0
No B.T. during study
Ferrous + ( n - 2 0 )
Ferrous- (n = 28)
55.3
53.7
26.6
31.3
17.7
15.5
20.4
30.7
26.9
22.1
Journal of Tropical Pediatrics
Vol.42
August 1996
J. VAN DEN HOMBERGH ET AL.
TABLE 4
Parasitaemia rates (percentage of children with positive blood smear in each intervention group), mean
haemoglobins and their difference for each category of parasitaemia and geometric mean parasite
densities (per fil) each date offollow-up
Parasitaemia rate
Ferrous+
Ferrous—
Day 0
(n-100)
Week 2
(n-92)
Week 4
(n-95)
Week 8
(n-93)
Week 12
(»-94)
100%
100%
20%
21%
33%
30%
49%
52%
27%
28%
7.9(n=19)
8.8 (n = 73)
0.9 g/1
0.03
8.8 (n-29)
9.3 (n-66)
0.5 g/1
0.1
7.3 (n = 47)
9.6 (n = 46)
2.3 g/1
< 0.001
8.2 (TI = 26)
9.8 (n = 68)
1.6 g/1
<O.O01
1153
673
2777
3866
5825
6634
5308
9302
Mean haemoglobin
Blood smear pos.
Blood smear neg.
Difference
P-value
Geometric mean density
Ferrous +
Ferrous—
8536
4403
TABLE 5
Mean Zrscores (weight/age), and their difference at
start and after 3 months
Mean Z-score
day 0 (SD)
Mean Z-score
week 12 (SD)
Difference
ferN f<rY (wN farY tw N farY I K N fwY tar N
dayO
«n«k2
«mk4
mate
m«k12
Fio. 3. Geometric mean parasite densities ( +confidence interval).
effective malaria therapy. Excluding the effect of
blood transfusion, an increase of more than 3 g/dl in
haemoglobin levels was already seen after 2 weeks.
Thereafter, only a minor increase is observed which
may be due to frequent re-infections after 3 weeks
and more, when the protective effect of Fansidar has
disappeared. The underlying physiology of this rapid
increase is not yet explained but may be related to the
multifactorial pathogenesis of malarial anaemia. The
increase of haemoglobin level in the children who
received a blood transfusion at start is, not surprisingly, much larger after 2 weeks, but this effect is not
observed after 4 weeks (Table 2).
The mean haemoglobin levels at week 12 were less
than 10 g/dl, and not different from haemoglobin
levels measured in a sample (n = 248) of healthy
children from 11 villages in the same health service
area.30
Journal of Tropical Pediatrics
VoL 42
August 1996
Difference
(P-0.421).
in
Ferrous
(n = 48)
No ferrous
-1.39(1.16)
-1.25(1.04)
-1.51(1.07)
-1.27(1.01)
-0.12(0.65)
+0.02(0.55)
change
of
nutritional
(7!-47)
status:
0.14
A positive effect of iron on the recovery from
severe anaemia could not be demonstrated. Differences in haemoglobin levels and reticulocyte counts
between the supplemented and control group were
not statistically significant overall, nor after stratification for having received a blood transfusion. The
observed rapid recovery of haemoglobin levels, and
the fact that the anaemia was normocytic, supports
the impression that these children were not irondeficient. Until more information becomes available
on the pathophysiology of anaemia in such children
with malaria, this observation would seem to be a
justification for discontinuation of routine supplementation of ferrous sulphate to children with
malarial anaemia in this area.
The range of parasite densities at first attendance
was wide in both groups. The geometric means were
below 10000 parasites per fil, a value often accepted
as indicating severe malaria. No association was
found between parasite density and level of anaemia.
This supports the view that the anaemia may
225
J. VAN DEN HOMBERGH ET AL.
develop gradually, after chronic low grade parasitaemia or, suddenly, in parallel with a severe acute
malaria infection. It is realized, however, that this
observation may be due to the restricted range of
haemoglobin at day 0. Prior use of anti-malarial
drugs will likewise influence parasite densities at
admission and follow-up measurements. In 1992, 20
per cent of children presenting with mild malaria
symptoms at the OPD of Turiani Hospital had a
positive ELISA-test for chloroquine in the urine.29 It
can be assumed that a considerable proportion of the
study population will have had malaria treatment in
between days of follow-up, either by self-medication
or at the hospital. The significant association between
parasitaemia rate and haemoglobin at weeks 8 and 12
supports the relationship between parasitaemia and
anaemia. The survey of healthy children in the same
health service area appeared to be consistent with
these findings.30 Out of 248 healthy children, a mean
haemoglobin of 7.3 g/dl was measured in those who
had a positive blood smear for malaria. Those
without had a mean haemoglobin of 9.0 g/dl. This
difference was significant. At day 0, the size of the
spleen was associated with the parasite density, a
higher density correlated with a larger spleen size.
An overall reduction in spleen size was measured
after 12 weeks. However, at this stage there was no
association between splenic enlargement and parasite
density observed anymore. This suggests mat the
reduction in spleen size is a result of the intermittent
treatment of clinical malaria.
The overall parasitaemia rates were high. The fact
that 2 weeks after Fansidar/quinine treatment, a
parasitaemia rate of 20 per cent was observed is
disappointingly high and suggests intense transmission or drug resistance in the area. The conspicuous
difference in parasitaemia rates among the children
between weeks 8 and 12 cannot be readily explained,
but may be due to seasonal fluctuations. The followup at weeks 8 and 12 coincided with a period of high
prevalence usually observed in Turiani after the main
rainy season. With regard to iron supplementation
no difference in parasitaemia rates and parasite
densities were found between both groups at each
control visit, even after stratifying for having had a
blood transfusion. In this sample of children, iron
supplementation did not increase the prevalence and
severity of malaria infection. This corroborates with
earlier observations.18'22
In the Gambia an increase of parasitaemia rates
was seen after supplementing iron deficient children.23 In contrast, however, the overall number of
recorded episodes of morbidity during the 3-month
follow-up was significantly higher in the group with
iron-supplementation. The number of extra attendances by mothers (apart from the regular follow-up
dates) because of an illness of their child, was also
higher in the iron supplemented group, an observation which is not likely to be observer-biased. The
226
diagnosis of pneumonia was made significantly more
often in the iron-supplemented group. The latter,
however, should be interpreted with caution, since
the diagnosis of pneumonia is notoriously difficult in
a malarious area. Some observation bias may have
occurred due to the fact that the diagnosing physician
was not blinded. Nonetheless, this observation is
consistent with other studies that recorded increased
incidence of infections after iron-supplementation."121 There was no effect on the nutritional status
observed.
Conclusions
Infants and young children recover from severe
anaemia associated with malaria within 2 weeks of
effective malaria treatment, with or without iron
supplementation.
Prompt and effective treatment of malaria infection is the single most important element in the
successful management of severe malarial anaemia in
infants and young children. Children with malaria
infection and moderate anaemia should be treated
with an effective second lineregimenwithout further
delay. The low mortality (2 per cent) in this study,
which had a 100 per cent follow-up rate, did not
occur in the period of severe anaemia and malaria.
This supports the view that in an area with a high
prevalence of HTV, syphilis, and hepatitis, the
decision for a blood transfusion to anaemic children
should be made carefully. This risk of death from
irreversible heart failure due to anaemia, has to be
balanced against the potential occurrence of a fatal
incompatibility reaction or transmission of HTVinfection. Our observations indicate that routine
daily supplementation with oral iron neither affects
the prevalence of malaria infection nor the parasite
densities. There is clear evidence that iron supplementation caused increased morbidity from other
conditions. Routine daily iron supplementation to
young children presenting with malaria or malaria
associated anaemia in this area of Tanzania has no
beneficial effect on the recovery of their haemoglobin
levels and, hence, is not warranted in primary care
practice.
References
1. WHO. Global Malaria Control Strategy. CTD/MAL/
EXP/92.3. Ministerial Caonference on Malaria,
Amsterdam, 1992.
2. Sturchler D. How much malaria is there worldwide.
Parasitol Today 1989; 5(2): 39-40.
3. Bradley DJ. Malaria: old infections, changing epidemiology. Hlth Transit Rev 1992; 2 (suppl): 137-53.
4. Keuter M, el al. Parasitological, clinical and haematclogical response of children with plasmodium faltiparum to 4-aminoquinolines and to pyrimethaminesulfadoxine with quinine in western Kenya. Trop Geogr
Med 1991; 44: 1-8.
Journal of Tropical Pediatrics
Vol. 42
August 1996
J. VAN DEN HOMBERGH ET AL.
5. Abdalla SH, Weatherall DJ, Wickramasinghe SN,
Hughes M. The anaemia of P. falciparum malaria. Br
J Haematol 1980; 46: 171-83.
6. Ho M, White NJ, Looaresuwan S. Splenic Fc receptor
function in host defense and anaemia in acute
Plasmodium falciparum malaria. J Inf Dis 1990; 161:
555-61.
7. Weatherall DJ, Abdalla S, Pippard MJ. The anaemia of
Plasmodium falciparum malaria. In Evered D, Welan J
(eds) Malaria and the red cell. Ciba Foundation
Symposium 94. London, Pitman, 1983: 74-88.
8. Snow RW, et al. Severe childhood malaria in two areas
of markedly different falciparum transmission in East
Africa. Acta Tropica 1994; 57: 289-300.
9. Vos J, et al. Are some blood transfusions avoidable? A
hospital record analysis in Mwanza Region, Tanzania.
Trop Geogr Med 1993; 45: 301-3.
10. Craighead IB, Knowles JK. Prevention of transfusion
associated HTV transmission with the use of a transfusion protocol for under 5s. Trop Doct 1993; 23: 59-61.
11. Trousseau A. Lectures on clinical medicine. London:
New Sydenham Society, 1872.
12. Filteau SM, Tomkins AM. Micronutrients and tropical
infections. Trans Roy Soc Med 1994; 88: 1-3.
13. Bullen JJ, Griffiths E (1987). Iron and infection,
molecular, physiological and clinical aspects. Chichester: J. Wiley & Sons.
14. McGregor IA. Malaria: nutritional implications. Rev
Inf Dis 1982; 4: 798-803.
15. Greenwood BM. Asymptomatic malaria infections, do
they matter? Parasitol Today, 1987; 3: 206-13.
16. Byles AB, Avita D'Sa. Reduction of reaction due to
iron-dextran infusion using chloroquine. Br Med J
1970; iii: 625-7..
17. Masawe AEJ, Muindi JM, Swai GBR. Infections in
iron deficiency and other types of anaemia in the
tropics. Lancet 1974; ii: 314-17.
18. Murray MJ, Murray NJ, Murray AB, Murray MB.
Refeeding-malaria and hyperferraemia. Lancet 1975; i:
653-^4.
19. Murray MJ, Murray AB, Murray MB, Murray CJ. The
adverse effect of iron repletion on the course of certain
infections. Br Med J 1978; 2: 1113-15.
Journal of Tropical Pediatrics
Vol. 42
August 19%
20. Oppenheimer SJ, Cashiri P. Serum and red cell
folatelevels associated with malarial parasitaemia.
Trans Roy Soc Trop Med Hyg 1986; 80: 169.
21. Bates CJ, Powers HJ, Lamb WH, Gehnan W, Webb E.
Effect of supplementary vitamins and iron on malaria
indices in rural Gambian children. Trans Roy Soc Trop
Med Hyg 1987; 81: 286-91.
22. Harvey PWJ, et al. The effect of iron therapy on
malarial infection in Papua New Guinean schoolchildren. Am J Trop Med Hyg 1989; 40: 12-18.
23. Smith AW, Hendrickse RG, Harrison C, Hayes RJ,
Greenwood BM. The effects on malaria of treatment of
iron-deficiency anaemia with oral iron in Gambian
children. Ann Trop Paed 1989; 9: 17-23.
24. Snow RW, Byass P, Shenton FC, Greenwood BM. The
relationship between anthropometic measurements and
measurements of iron status and susceptibility to
malaria in Gambian children. Trans Roy Soc Trop
Med Hyg 1991; 85: 584-9.
25. Clyde DF. Malaria in Tanzania. Oxford: Oxford
University Press, 1967.
26. Msuya FHM, Curtis CF. Trial of pyrethroid bednets in
an area of Tanzania holoendemic for malaria Part 4.
Effects of incidence of malaria infection. Acta Tropica,
1991; 49: 165-71.
27. Premji Z, Minjas JN, Schiff CJ. Chloroquine resistant
Plasmodium falciparum in coastal Tanzania. A challenge to the continued strategy of village based
chemotherapy for malaria control. Trop Med Parasitol
1994; 45: 47-8.
28. lines JD, Wilkes TJ, Lyimo EO. Human malaria
infectiousness measured by age-specific sporozoite rates
in Anopheles gambiae in Tanzania. Parasitology 1991;
102: 167-77.
29. De Boer IP, Dunk CP. Drug resistance of P. falciparum
in children in Turiani. A field study on chloroquine and
amodiaquine in Tanzania. 1992. University of Amsterdam (unpublished).
30. Van den Hombergh J. Malarial anaemia in infants and
young children from the Morogoro Region in Tanzania.
Epidemiology and clinical aspects. Thesis Msc. Communicable Diseases Epidemiology. London: School of
Hygiene and Tropical Medicine, 1994.
227