CORPUS CALLOSUM MORPHOLOGY IN RETINOPATHY-POSITIVE AND
RETINOPATHY-NEGATIVE MALAWIAN CHILDREN DIAGNOSED WITH CEREBRAL
MALARIA
By
Lisa Vroman

A THESIS
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
MASTERS OF ARTS
Psychology
2011

ABSTRACT
CORPUS CALLOSUM MORPHOLOGY IN RETINOPATHY-POSITIVE AND
RETINOPATHY-NEGATIVE MALAWIAN CHILDREN DIAGNOSED WITH CEREBRAL
MALARIA
By
Lisa Vroman
The current study investigated corpus callosum (CC) morphology in retinopathy-positive
and retinopathy-negative pediatric survivors of cerebral malaria (CM). Participants were 20
children who were admitted to the Queen Elizabeth Central Hospital in Blantyre, Malawi, and
satisfied the clinical case definition of CM. There were two groups: retinopathy-positive CM (n
= 16) and retinopathy-negative CM (n = 4). Analysis of covariance and multivariate analysis of
covariance were conducted and effect sizes were calculated on the area measurements to
determine whether overall brain area, CC area, and five segments of the CC (i.e., genu, body,
midbody, isthmus, and splenium) differed as a function of retinopathy status. No significant
group differences emerged in any of the analyses. However, a large effect size was noted in
overall brain area and a moderate effect size was detected in total CC area. For the five CC
segments, a moderate effect size was detected for the genu, body, and midbody, and a large
effect size was found for the splenium. The effect sizes indicate that these regions may be
smaller in the retinopathy-positive group. The results suggest the retinopathy-positive group
may have lost white matter in the CC, particularly in the splenium, and perhaps elsewhere in the
brain. Research indicates attention deficits frequently occur in pediatric survivors of CM. This
study is consistent with previous research Attention-Deficit/Hyperactivity Disorder, suggesting
the CC, particularly the splenium, may be especially important in attentional processes.

ACKNOWLEDGEMENTS

This project would not have been possible without the help of many individuals. I would
first like to express my sincerest thanks to my advisor, Margaret Semrud-Clikeman, for her
continuous guidance on this project. I would also like to thank my committee members, Terrie
Taylor, Karl Seydel, and Robert Caldwell for their support. Also, many thanks to all those on
the Blantyre, Malawi Project who not only generously provided their data, but offered their time
and technical assistance on this project as well. I would also like to thank Jodene Fine for
providing her expertise. And last, but certainly, not least, I would like to express my thanks to
my classmates and family. Without your continued support and encouragement, none of this
would have been possible.

iii

TABLE OF CONTENTS

LIST OF TABLES ......................................................................................................................... vi
LIST OF FIGURES ...................................................................................................................... vii
INTRODUCTION .......................................................................................................................... 1
Infection .............................................................................................................................. 2
Pathogenesis of CM ............................................................................................................ 3
Tumor necrosis factor ............................................................................................. 3
Blood-brain barrier.................................................................................................. 4
Sequestration ........................................................................................................... 5
Diagnosis............................................................................................................................. 5
New diagnostic technique ....................................................................................... 6
Medical Treatment .............................................................................................................. 9
Neuropsychology of CM................................................................................................... 11
Neuroimaging in CM ........................................................................................................ 14
White matter and edema ....................................................................................... 16
Corpus callosum.................................................................................................... 17
Conclusions ....................................................................................................................... 18
The Present Study ............................................................................................................. 19
METHOD ..................................................................................................................................... 20
Participants ........................................................................................................................ 20
Participant criteria ................................................................................................. 21
Image Acquisition ............................................................................................................. 22
Loss of scans ......................................................................................................... 22
Image Analysis.................................................................................................................. 22
Alignment ............................................................................................................. 22
Midsagittal slice identification.............................................................................. 23
Corpus callosum area ............................................................................................ 23
Supratentorial area ................................................................................................ 24
RESULTS ..................................................................................................................................... 24
Power and Statistical Analyses ......................................................................................... 24
Power .................................................................................................................... 25
Statistical analyses ................................................................................................ 25
Gender ................................................................................................................... 26
DISCUSSION ............................................................................................................................... 27
Major Findings .................................................................................................................. 28
Splenium ............................................................................................................... 30
Exploratory Analyses on Gender ...................................................................................... 31
Limitations ........................................................................................................................ 32
Clinical Implications ......................................................................................................... 33
iv

Future Directions .............................................................................................................. 35
REFERENCES ............................................................................................................................. 45

v

LIST OF TABLES

Table 1. Description of Coma Scales ............................................................................................ 36
Table 2. Participant Demographics ............................................................................................... 37
Table 3. Correlations for the Measured Variables ........................................................................ 38
Table 4. Means, Standard Deviations, and Effect Sizes for the Retinopathy-Positive and Negative
Groups ........................................................................................................................................... 39
Table 5. Means, Standard Deviations, and Effect Sizes for Retinopathy-Positive Groups .......... 40

vi

LIST OF FIGURES
Figure 1. Retinal findings ............................................................................................................. 41
Figure 2. Corpus callosum area measurement .............................................................................. 42
Figure 3. Segmentation of the corpus callosum ............................................................................ 43
Figure 4. Supratentorial area measurement .................................................................................. 44

vii

Corpus Callosum Morphology in Retinopathy-Positive and Retinopathy-Negative Malawian
Children Diagnosed with Cerebral Malaria
Cerebral malaria (CM) is a prevailing cause of illness and often death. It is estimated that
CM affects more than 200 million people in Africa yearly and results in one million deaths
annually (Snow, Craig, & Marsh, 1999). Cerebral malaria is particularly devastating in children,
with an estimated case fatality rate of 19% (Murphy & Breman, 2001). A CM diagnosis is rare
in adults living in malarial endemic regions, as they tend to acquire immunity; non-immune
adults, however, are quite susceptible (World Health Organization, 2000).
While the investigation of pediatric CM is one of medical interest, pediatric CM is also
an illness of clinical interest and concern. Neurological deficits following CM, such as paresis
and ataxia, typically resolve within six months post-discharge from the hospital (van Hensbroek,
Palmer, Jaffar, Schneider, & Kwiatkowski, 1997). Neurocognitive deficits, however, are
common among pediatric survivors of CM and have been documented to be more persistent in
nature (e.g., Carter et al., 2005b). Neurocognitive deficits in this population commonly occur in
the areas of working memory and attention (Boivin et al., 2007). Birbeck and colleagues (2010)
commented on the similarities in the neurocognitive profiles of pediatric survivors of
retinopathy-positive CM and disruptive behavior disorders in children, specifically AttentionDeficit/Hyperactivity Disorder (ADHD). These researchers noted that 11% of the pediatric
survivors of retinopathy-positive CM displayed symptoms of ADHD. Additionally,
neuroimaging research in both pediatric populations reveals similarities in brain structure. For
example, white matter abnormalities including corpus callosum abnormalities appear in both
populations. Given the apparent link between pediatric survivors of CM and ADHD
symptomatology, further investigation of the brain the structure of pediatric survivors of CM,

1

particularly the corpus callosum, is necessary. Continued investigation of brain structure is
essential to further elucidate the pathogenesis of the disease, as well as to inform future
intervention and treatment options. The following sections will begin by outlining the infection
phase of the illness and theories on the pathogenesis of CM, diagnostic techniques, and treatment
of CM. The extant neuropsychological and neuroimaging studies on pediatric CM will then be
discussed.
Infection
Humans can be infected with five species of malaria: Plasmodium vivax, Plasmodium
ovale, Plasmodium malariae, Plasmodium knowlesi and Plasmodium falciparum (Cox-Singh et
al., 2008). Plasmodium falciparum, or P falciparum, is the species most commonly associated
with severe and complicated malaria. The life cycle of P falciparum begins with sporozoites, the
transmissible form of the parasite. The sporozoites develop within the mosquito and are later
transferred to humans when the mosquito bites the human host. Within 30 minutes the
sporozoites travel to the liver. Inside the liver the parasite matures into a schizont, the next
developmental phase of the asexual reproductive life cycle. A merozoite is formed after the
division of the schizont. The merozoites are later released to invade the erythrocytes (red blood
cells) of the host. The parasitized erythrocytes bind to receptors found on the endothelial cells of
blood vessels. This process occurs when the parasite produces a ligand, a protein that binds to a
receptor on the erythrocyte. The ligand that has been the most studied in P falciparum infections
is the P falciparum erythrocyte membrane protein (PfEMP-1). The parasite sends the PfEMP-1
to the erythrocyte membrane to adhesive nodules called knobs (Adams, Brown, & Turner, 2002).
The process of merozoites infecting erythrocytes is referred to as invasion and marks the
beginning of the erythrocytic phase of the cycle. This process occurs many times. The

2

merozoites residing in the erythrocytes are initially in ring form and later develop into
trophozoites, which develop into schizonts. Merozoites are released at the time of schizont
rupture, when the erythrocytes are destroyed. The merozoites then invade other red blood cells.
When the schizonts rupture, other toxins are also released. This results in an initial symptom of
fever noted in all forms of malarial illness. Common symptoms in P falciparum infections are
fever, headache, nausea, diarrhea, and vomiting. If the symptoms go untreated, severe malaria
may result. Severe malaria can take many forms; cerebral malaria, hypoglycemia, acidosis, and
severe anemia are all potential complications (WHO, 2000). The earliest symptom is often fever
in children diagnosed with CM. Following fever, frequent symptoms include refusal to eat or
drink, vomiting, and cough. However, while the infection phase of CM has been welldocumented, the pathogenesis of CM is still largely uncertain.
Pathogenesis of CM
In reviewing the literature, it appears that three hypotheses on the pathogenesis of CM are
currently under primary investigation. The first hypothesis implicates tumor necrosis factor
(TNF) in the development of the disease. The second hypothesis involves the investigation of
the potential role of the blood brain barrier (BBB). Lastly, the third hypothesis suggests
sequestration of parasitized erythrocytes in the microvasculature of the brain may be involved in
the pathogenesis of CM. Each of these hypotheses will be examined in turn, in the sections that
follow.
Tumor necrosis factor. Tumor necrosis factor (TNF) has been implicated in a
breakdown of the BBB. Tumor necrosis factor and interleukin are both subsumed under the
category of cytokines, or small proteins that mediate the body’s immune responses. In high
quantities, cytokines can be toxic to the central nervous system (CNS) of a child diagnosed with

3

CM. Increased levels of TNF have been noted in children diagnosed with CM relative to
children diagnosed with a “mild disease” (Karunaweera, Grau, Gamage, Carter, & Mendis,
1992). One study reported that during paroxysm, tumor necrosis factor levels in CM patients
are lower than those in humans infected with Plasmodium vivax, in which there are no
neurological symptoms (Karunaweera et al., 1992).
Blood-brain barrier. A breakdown of the blood-brain barrier (BBB) has also been
implicated in the pathogenesis of CM. The BBB is composed of endothelial cells, astrocyte
projections known as astrocytic feet, and a basement membrane. The BBB restricts the exchange
of substances from the blood into brain tissue. Studies using murine models have demonstrated a
breakdown of the BBB (Thumwood, Hunt, Clark, & Cowden, 1988). In a post-mortem
examination completed by Brown and colleagues (2001) on Malawian children diagnosed with
CM, the researchers discovered an activation of cerebral endothelial cells as well as a disruption
of intercellular junctions following CM. A small increase in the albumin index was discovered.
This suggests a minor breakdown of the BBB occurred, although the authors report this
breakdown was not sufficient to fully account for the cerebral edema noted in the children.
Animal model research also suggests a breakdown of the BBB. Murine models of fatal
CM demonstrate two stages of a BBB breakdown. The first occurs three days following
infection, in which a small but focal breakdown of the BBB occurs. Seven days after infection, a
widespread breakdown of the BBB takes place. The plasma first escapes into the perivascular
space, the area around the blood vessels, while an up-regulation of endothelial adhesion
molecules, proteins found on the cell surface, is triggered by an immune response. An increase
in the BBB permeability soon follows (Adams et al., 2002).

4

Sequestration. Cerebral edema has been observed to frequently occur in pediatric cases
of CM (Newton et al., 1994). Sequestration, or the binding of the parasitized erythrocytes to the
vessels, has been proposed as one potential cause of cerebral edema and cerebral edema may
play a role in the pathogenesis of coma. Other cited possible causes of cerebral edema are
increased intracranial pressure, changes in blood flow, cellular hypoxia, and some evidence
suggests increased permeability of the blood brain barrier (Adams et al., 2002). Additionally,
brain swelling has been documented in African children and at least some of this swelling may
be attributed to intravascular sequestration of parasitized erythrocytes (Newton et al., 1994).
Nonetheless, there is currently an incomplete understanding of the pathogenesis of CM. Despite
an imperfect understanding of the development of CM, diagnosis is a necessary step of the
treatment of CM. The diagnosis of pediatric CM is largely determined by the presence of P
falciparum on a blood smear and coma.
To summarize, pediatric CM is a particularly devastating cause of illness in children. It
is known that CM is a potential outcome following the bite of a mosquito infected with the
parasite P falciparum, but the pathogenesis of CM is not yet completely understood. Cytokines,
the blood brain barrier, and the sequestration of parasitized erythrocytes have been implicated in
the pathogenesis of CM but further investigation is required (Adams et al., 2002).
Diagnosis
Cerebral malaria is defined by the World Health Organization (WHO) as the presence of
P falciparum on a blood smear, coma as exemplified by a Glasgow Coma Scale score ≤ 8 (or
Blantyre Coma Scale score ≤ 2), with no other known cause for coma. Table 1 provides a
description of the Glasgow and Blantyre Coma Scales. The GCS was developed as a measure of
consciousness and is generally utilized during the diagnostic process for traumatic brain injury.

5

The Blantyre Coma Scale was developed specifically for the assessment of children and is
commonly utilized as a measure of severity of CM (Newton, et al., 1997).
The diagnosis of CM is made difficult by the knowledge that parasitemia is not specific
to malaria disease. It is important to note that the presence of parasitemia does imply infection,
but does not necessarily imply malarial illness. Thus, parasitemia is necessary but not sufficient
to warrant a diagnosis of malaria illness. Consequently, at this point in time, it is not possible to
causally link the presence of parasites to the patient’s symptoms. In malaria-endemic regions,
asymptomatic parasitemia is quite common and has been reported at approximately 40-70%
(Mwangi, Ross, Snow, & Marsh, 2005). The diagnostic criteria to date for CM are not specific
to severe malaria. Recently, however, a new technique has been implemented in the diagnosis of
pediatric CM, which evaluates the appearance of the retina.
New diagnostic technique. P falciparum infections in children may result in any of the
following four complications: cerebral malaria, respiratory distress, severe anemia, and
hypoglycemia (World Health Organization, 2000). These four complications represent the most
common complications and each may present with similar symptoms, but it is possible to
develop any of these four syndromes without being infected with malaria. Diagnostically,
however, the difficulty lies not with identifying parasitemic children, but distinguishing which
syndromes are due to malaria and which are not. Because a number of syndromes present with
similar symptomatology, it is challenging to make a correct diagnosis. The WHO definition does
not in and of itself exclude other possible diagnoses. As a result, misdiagnosis is a common
occurrence. It is therefore important that additional diagnostic tools with empirically
demonstrated sensitivity and specificity be utilized. One such tool is direct and indirect
ophthalmoscopy. One study detailed abnormal retinal findings referred to as malaria retinopathy

6

in children diagnosed with CM (Beare, Taylor, Harding, Lewallen, & Molyneux, 2006). The
retina is developmentally an element of the central nervous system and allows for an in vivo
investigation of infected microvasculature, the portion of the circulatory system referring to the
small vessels (Beare et al., 2006).
Four retinal findings have been characterized: retinal whitening, vessel changes,
hemorrhages, and papilledema. Only three of these retinal findings, however, constitute the
features of malarial retinopathy: retinal whitening, vessel changes, and hemorrhages, as
papilledema alone is not indicative of malaria (Lewallen et al., 2008). Retinal whitening results
in scattered whitening of the retina. Vessel changes result in a color change in the retinal vessels
to a pale orange or white color. Hemorrhages generally appear white-centered and are intraretinal. Papilledema is the swelling of the optic disk and is often a sign of raised intracranial
pressure. Research suggests papilledema and retinal hemorrhages are not noted only in CM;
although, white-centered hemorrhages are not noted frequently in other conditions. Vessel
changes and the specific pattern of retinal whitening, however, have yet to be characterized in a
disease outside of malaria (Lewallen et al., 2008). Thus, if papilledema is noted with no signs of
retinal whitening, vessel changes, or hemorrhages, an alternative diagnosis should be considered.
These features have been observed in children in Malawi, Kenya, and Gambia (Beare et al.,
2004a). Malarial retinopathy is now considered to be one of the best diagnostic features of CM,
helping physicians distinguish between malarial and non-malarial coma (Beare et al., 2006;
Taylor et al., 2004). See Figure 1 for pictures of each of the four retinal findings.
Beare and colleagues (2004a) discovered that children with retinopathy had greater than
average parasite density, were more likely to have respiratory distress, and were less likely to
have repeated or prolonged convulsions than those without retinopathy. There was no

7

statistically significant relationship between those with retinopathy who later developed
neurological sequelae and those who did not have any neurological sequelae. Retinopathy was
observed in a greater percentage of children who later died. Of the children with retinopathy
who survived, the resolution of the coma was significantly longer as compared to children who
did not have retinopathy. Additionally, the severity of the retinal signs was found to be
correlated with death and coma resolution time.
In addition, papilledema and retinal hemorrhages were differentially related to death.
Beare and colleagues (2004a) suggest that these retinal abnormalities are the result of two
distinct processes. Papilledema and retinal hemorrhages were found to be independent
predictors of death (Beare et al., 2004a). Cortical blindness does not appear to be correlated to
raised intracranial pressure as indicated by papilledema. Additionally, the retinal changes
associated with malarial retinopathy do not seem to impact visual acuity after one month (Beare,
Sourthern, Kayira, Taylor, & Harding, 2004b).
To date, the majority of studies have not included retinopathy as a diagnostic criterion.
Instead the clinical definition of CM, or the WHO definition, requires the presence of P
falciparum on a blood smear, GCS or BCS core of ≤8 or ≤2, respectively, and no other cause for
coma. Taylor et al. (2004) demonstrated malarial retinopathy is the only clinical feature that
distinguished between CM and non-cerebral malaria cases. This study conducted autopsies on
patients diagnosed with clinically defined CM (n = 31) and patients with non-malarial causes of
coma (n = 11). It was determined that 23% (n = 7) of cases were misdiagnosed with CM, despite
the fact that each of these cases fulfilled the WHO definition of CM. Examination of these
patients failed to reveal parasite sequestration. Further histological examination revealed that
three of the patients died from Reye’s syndrome, arteriovenous malformation, and hepatic

8

necrosis, respectively. The remaining four patients died from severe anemia, pneumonia, and
meningitis. However, when these children were diagnosed with CM, malarial retinopathy was
not a recognized indicator of CM. Histopathological study revealed malarial retinopathy to be a
key diagnostic feature of CM and the authors suggest the absence of malarial retinopathy is a
strong indicator that CM is not the most appropriate diagnosis.
In one post-mortem investigation, ten ophthalmologists over a period of 10 years
performed funduscopic examinations on the same patient as a technique for determining fatal
CM. The ten ophthalmologists collectively achieved a sensitivity of 95% and specificity of 90%
(Beare et al., 2006). When the WHO criteria were applied to this group of children, a specificity
rate of 61% was achieved.
In summary, the WHO diagnostic criteria are not highly specific to CM; but, a new
diagnostic sign, malarial retinopathy, has demonstrated high specificity and sensitivity to CM
(Beare et al., 2006). This technique has improved the diagnosis of CM and medical treatments
are available once a diagnosis has been made.
Medical Treatment
When treatment can be provided for CM, the standard procedure is intravenous or
intramuscular quinine (Aceng, Byarugaba, & Tumwine, 2005; Whitty, 2005). Artemisinin
treatments are also available and can be administered orally or via rectal or intramuscular
administration. A rectal artemether has been suggested to be just as effective as intravenous
quinine, appears to be well-tolerated, and was found to be a safe alternative to intramuscular
quinine (Aceng et al., 2005; Barennes, Balima-Koussobue, Nargot, Charpentier, & Pussard,
2006).

9

A recent study examining parenteral artesunate and parenteral quinine treatment for
severe falciparum malaria in African children under the age of 15 indicated that parenteral
artesunate should be considered the preferred treatment for severe malaria (Dondorp et al.,
2010). The authors discovered that while the incidence of neurological sequelae was not
significantly different between the two treatment groups, those treated with artesunate were
significantly less likely to develop a coma or demonstrate a deterioration of the coma score and
convulsions and post-treatment hypoglycemia were also observed significantly less frequently in
the artesunate group. The artesunate treatment was reported to be well-tolerated and did not
have serious negative effects. Overall, this study demonstrated that artesunate treatment is
superior to parenteral quinine.
Another study determined that the time to parasite clearance was significantly shorter in
Malawian children diagnosed with CM treated by intramuscular artemether compared to those
treated by intravenous quinine (Taylor et al., 1993). A more recent randomized clinical trial
examined the efficacy of chloroquine versus sulfadoxine-pyrimethamine in the treatment of
uncomplicated P falciparum malaria in children in Malawi (Laufer et al., 2006). Despite having
been replaced by sulfadoxine-pyrimethamine in Malawi in 1993, this study discovered that
choloroquine is again an efficacious treatment for malaria. Even when children receive expert
care, however, approximately 15-20% of children die (Taylor et al., 2004; White, 1998).
Neurological and neuropsychological sequelae are frequent in children with CM even
when appropriate treatment has been administered. Common neurological sequelae include
ataxia, convulsions, paresis, and visual and hearing difficulties (van Hensbroek et al., 1997).
Neuropsychological deficits in the areas of working memory and attention are most frequently
documented (Boivin et al., 2007). These impairments are highly akin to those defined in

10

children diagnosed with ADHD. The following sections will summarize the relevant
neuropsychological and ADHD literature.
Neuropsychology of CM
While most children diagnosed with CM no longer have frank neurological deficits (e.g.,
paresis and ataxia) six months following discharge from the hospital (van Hensbroek et al.,
1997), the same is not true of neurocognitive deficits. Cognitive impairments have been reported
in 11-28% of pediatric survivors of CM (Boivin et al., 2007) and are typically in the areas of
attention, working memory, and language (Boivin et al., 2007, Carter et al., 2005b). Thus,
cognitive impairment in children following CM is more common and generally more widespread
than neurologic sequelae.
Boivin and colleagues (2007) investigated the neurocognitive sequelae in Ugandan
children aged five – twelve years. Forty-four children diagnosed with CM were tested along
with 54 children diagnosed with uncomplicated malaria and 89 healthy community children.
Three assessments were administered: The Kaufman Assessment Battery for Children (K-ABC),
the visual form of the Test of Variables of Attention (TOVA), and the tactual performance test
(TPT). These tasks assess cognitive and memory abilities, attention, and tactile perception,
respectively. It was discovered that 21.4% of the children had cognitive deficits in the areas of
working memory, attention, and learning three months following their discharge from the
hospital. At a six month follow-up, these children demonstrated even greater impairment. The
children diagnosed with CM were 3.7 times more likely to develop a cognitive deficit in one or
more domains of functioning, relative to the healthy community children. This pattern of
continued cognitive decline suggests a significant number of children are likely to lose
previously learned skills, as well as have difficulties learning new cognitive skills. A follow-up

11

study completed by John et al. (2008) discovered that 26.3% of children diagnosed with CM had
at least one cognitive deficit two years following the child’s discharge from the hospital.
Additionally, the pediatric survivors of CM were at a 3.67-fold increased risk for developing a
cognitive deficit in working memory, attention, or learning as compared to healthy community
controls. Boivin (2002) evaluated 29 Senegalese children diagnosed with CM and 29 agedmatched mild malaria group. The K-ABC and the TOVA were administered to assess cognition,
memory, and attention. It was determined that the CM group performed significantly lower on
the K-ABC than the ‘mild’ malaria group. On the TOVA, the CM group committed significantly
more omission errors than the mild malaria group. It was also determined that length of coma
positively correlated with the number of omission errors committed.
In another neuropsychological study of pediatric survivors of CM, Carter and colleagues
(2005a) assessed Kenyan children six – nine years of age diagnosed with CM and malaria with
complicated seizures (M/S) in the areas of cognition, motor, speech and language, and hearing
and vision. The authors determined the pediatric survivors of CM performed significantly lower
than the M/S group on tasks language ability (i.e., higher level language, vocabulary, pragmatics,
and non-verbal functioning). These studies provide further evidence for the long-term and
extensive nature of the cognitive sequelae of CM. Carter and colleagues (2005b) discovered that
cognitive impairments can persist up to nine years following recovery from CM. The results
indicated that 24% of the CM and M/S children were impaired in at least one domain. While
these two studies demonstrated that developmental impairments were noted in pediatric
survivors of CM particularly in language skills, attention deficits were not assessed.
Nonetheless, given that attention deficits are commonly reported in children diagnosed with CM,
it is critical that such outcomes are assessed.

12

In contrast, one study failed to find evidence of neuropsychological sequelae following
CM when the children did not suffer from neurological sequelae (Muntendam, Jaffar, Bleichrodt,
& van Hendsbroek, 1996). This study evaluated the children using five cognitive tasks that
assessed perception, reasoning, learning, and memory and seven sensory motor tasks that
assessed visual motor coordination, balance, manual speed, and finger/hand/arm dexterity. A
potential explanation for this incongruent finding is the absence of pure measures of attention
from the assessment battery. A common neurocognitive impairment in pediatric survivors of
CM is inattention. Had an attention measure been administered, the results may have been more
congruous with the literature.
A number of risk factors associated with the development of cognitive impairments have
been corroborated. A risk factor for later motor impairment was linked to multiple seizures.
History of seizures, hypoglycemia, being under three years of age, increased intracranial pressure
at admission and prolonged coma were found to be risk factors for speech and language deficits.
Moreover, this study found that prolonged coma and prior seizure episodes were risk factors for
the development of cognitive impairment (Idro et al., 2006). The later risk factors regarding
extended duration of coma and multiple seizure episodes were supported by Boivin et al. (2007)
as well. Holding et al. (1999) found that coma, hypoglycemia, seizures, and the absence of
hyperpyrexia are four key signs in predicting cognitive impairment. Overall, developmental
delays, particularly for language skills were discovered in children with positive retinopathy
cerebral malaria (Boivin et al., 2011).
A common neurocognitive deficit among children who survived CM is inattention (e.g.,
Boivin et al., 2007). As these children look akin to children diagnosed with ADHD, the
investigation of the potential correlates of these two disorders has caught the attention of

13

researchers. Boivin et al. (2011) evaluated 83 Malawian children diagnosed with retinopathypositive CM and 95 age-matched control children. The Malawi Developmental Assessment Test
(MDAT), an assessment of gross and fine motor control, language, and social skills, was
administered to determine if a developmental delay was present. Also, the Achenbach Child
Behavior Checklist (CBCL), a child psychiatric screening test, was administered to evaluate
psychiatric symptoms of the child. Internalizing and externalizing symptoms, including
depression, anxiety, and aggression were evaluated and no significant group differences were
found. However, on the MDAT language domain, the retinopathy-positive group demonstrated a
significantly greater possibility of a developmental delay. It is important to note that this study
was among the first to assess retinopathy-positive cases of pediatric CM. As such, it is likely
that the literature to date has evaluated both true and false cases of CM, suggesting the outcomes
of these studies have been slightly confounded. This study represents an important first step in
rectifying that confound.
To summarize, pediatric survivors of CM often exhibit neurological sequelae; however,
these deficits typically resolve within six months following discharge from the hospital (van
Hensbroek et al., 1997). Neurocognitive deficits, on the other hand, appear to occur in the areas
of working memory and attention and have been documented to persist and even worsen over
time (Boivin et al., 2002; Boivin et al., 2007; Carter et al., 2005a; Carter et al., 2005b; John et al.,
2008). Research has also been conducted on the brain structure of patients diagnosed with CM.
Neuroimaging in CM
While studies linking the neuropsychological assessment of children diagnosed with CM
to ADHD symptomatology have been completed, few neuroimaging studies have been
completed with this population. Numerous neuroimaging studies have been performed with

14

children diagnosed with ADHD, however. Of those, many neuroimaging studies have
investigated corpus callosum morphology because this structure allows information to be
transferred between the hemispheres. Children diagnosed with ADHD demonstrate a slower
response rate to stimuli, suggesting structures involved in the transference of information may be
implicated in the disorder (Semrud-Clikeman et al., 1994). Hynd and colleagues (1991)
revealed, after morphometric analysis of the corpus callosum, that children with ADHD had
smaller corpus callosi overall, but particularly in the genu, isthmus, and splenium. A second
study examined the area of the corpus callosum in children with ADHD who exhibited
overactivity symptomatology. It was discovered that these children had significantly smaller
splenial regions as compared to typically developing children (Semrud-Clikeman et al., 1994).
The authors suggest smaller splenial regions may be correlated with the commonly observed
attention difficulties observed in this population. Neuroimaging findings in children diagnosed
with CM, however, are scarce and inconsistent.
Beyond case studies of adults, there are six neuroimaging studies on pediatric CM. Of
these pediatric neuroimaging studies, two are case studies and a third study includes only one
child patient; the remaining two patients in the study were adults (Cordoliani et al., 1998;
Gamanagatti & Kandpal, 2006; Saavedra-Lozano, Booth, Weprin, & Ramilo, 2001). The other
three neuroimaging studies completed computed tomography (CT) scans on children diagnosed
with CM; one study examined eight children and the other two studies examined 14 children
(Newton et al., 1994; Potchen et al., 2010; Ngoungo et al., 2006). Although the base rate is high,
CM occurs largely in developing regions where appropriate medical care is largely unavailable
(Birbeck, 2004). The neuroimaging technology necessary to conduct clinical and research scans
is largely nonexistent as well. Consequently, little is known about the structural and functional

15

composition of the brain of children with CM. Of the available research, a few similarities
among findings are noted, but most results have yet to be corroborated. The main findings of
these studies implicated white matter, edema, and corpus callosum abnormalities.
White matter and edema. White matter abnormalities have been noted in a number of
studies (Cascio et al., 2006; Cordoliani et al., 1998; Lecours et al., 2001; Patankar et al., 2002).
Hyperintense areas of white matter were found in two adult patients and cortical infarcts were
noted in one child diagnosed with CM (Cordoliani et al., 1998). Hyperintensities in the occipital
cortex, and bilateral periventricular white matter and thalami were discovered in adult patients
(Yadav, Sharma, Kumar, & Kumar, 2008). Patankar et al. (2002) linked neuroimaging and
postmortem findings with clinical data in patients diagnosed with CM. The autopsies revealed
cerebellar white matter hypoattenuation in five of the 21 patients. In addition, diffuse cerebral
edema was noted in eight of the 21 patients. Newton et al. (1994) also noted diffuse cerebral
edema in six of the 14 children who underwent CT scans. The authors also reported that CT
scans of four of the 14 children demonstrated a decrease of grey and white matter, which
appeared consistent with diffuse ischemic or hypoglycemic insult. Additionally, in a case study
of a seven month-old male, CT results showed cerebellar edema and obstructive hydrocephalus
(Saavedra-Lozano, Booth, Weprin, & Ramilo, 2001). Potchen and colleagues (2010) completed
CT scans on children diagnosed with CM with prolonged coma. The authors found diffuse
cerebral edema, with and without brainstem involvement and the infarction of large vessels. In a
neuroimaging study completed on adults, the authors report that the MRI images for most of the
patients appeared normal unless viewed in relation to other scans. Here, swelling of the brain
during the acute phase of CM became apparent. Overall, only two patients showed evidence of

16

cerebral edema. Additionally, swelling of the brainstem and hydrocephalus were not noted in the
patients (Looareesuwan, et al., 1995).
Ngoungou and colleagues (2006) report disparate CT results from children diagnosed
with CM. Of the eight children that were scanned, five exhibited abnormalities. Diffuse atrophy
was noted in one child. Asymmetrical enlargement of the ventricles was noted in another. A
third child demonstrated atrophy in one hemisphere following bacterial meningitis. The
remaining two children both showed necrosis of the pallidi.
Corpus callosum. A thinning of the corpus callosum was noted in children with proteinenergy malnutrition during infancy (Lecours et al., 1998) and in children with developmental
delay (Cascio et al., 2006). Bilateral symmetric hyperintensities in the centrum semiovale were
also noted in a magnetic resonance (MR) study of a 13 year-old girl (Gamanagatti & Kandpal,
2006). These abnormalities noted in the centrum semiovale suggest possible cytotoxic edema,
which is consistent with the cerebral edema noted in pediatric CM cases. Cordoliani et al. (1998)
discovered white matter hyperintensities in the centrum semiovale and a lesion in the splenium in
one of the three patients examined. In the MR study conducted by Cordoliani and colleagues
(1998), neuropsychological testing determined deficits in memorization tasks. One week
following the initial MR examination, the hyperintensities in the centrum semiovale had
disappeared and the lesion in the splenium of the corpus callosum had decreased slightly. The
noted deficits in memorization resolved five months later; however, the lesion of the splenium
remained the same. A second study also noted a transient lesion in the splenium of the corpus
callosum in three patients with epilepsy (Polster et al., 2001). Lesions in the splenium of the
corpus callosum were noted in adult patients with cerebral malaria who presented with altered
sensorium (Yadav et al., 2008).

17

To summarize, the neuroimaging research to date is sparse, particularly concerning
pediatric CM. The common themes at present include white matter lesions, including
abnormalities in the corpus callosum.
Conclusions
Cerebral malaria commonly afflicts children, particularly in the region with the most
intense transmission of the causative agent, P falciparum. Despite the fact that CM affects
millions of children and adults, little is currently known about the pathogenesis of CM.
Research, particularly with animal models, has implicated cytokines, NO, and a slight
breakdown in the BBB. Diagnosis of CM is often complicated as ‘incidental parasitemias’ are
common in this population. The presence of parasitemia is itself insufficient to infer a causal
relationship between symptom presentation and the presence of parasites. Consequently,
misdiagnoses can occur. Recently however, a new diagnostic indicator of CM, malarial
retinopathy, was identified and has demonstrated an increase in diagnostic specificity (Beare et
al., 2006). Research studies concentrating on the neuropsychological deficits following pediatric
CM are beginning to be completed. Deficits in the areas of working memory, attention, and
language are among those most frequently documented (Boivin, et al., 2007; Carter et al.,
2005a). Currently, few neuropsychological long-term follow-up studies have been completed.
Of those that have, neuropsychological deficits have been noted as long as nine years following
recovery from CM and have been demonstrated to increase over time (Boivin et al., 2007; Carter
et al., 2005b). Additional long-term follow-up studies should be completed to determine the
extent of the noted neuropsychological deficits.
The inattention and working memory deficits found in survivors of CM parallel those
noted in ADHD. Numerous neuroimaging studies have been completed on children diagnosed

18

with ADHD. Neuroimaging studies that have investigated corpus callosum morphology have
discovered that children diagnosed with ADHD have smaller overall corpus callosi, but
specifically in the genu and splenium (e.g., Hynd et al., 1991; Semrud-Clikeman et al., 1994).
Few neuroimaging studies, outside of case studies, have been completed on children diagnosed
with CM. Of the available research, few findings appear to be consistent across studies.
However, white matter abnormalities, cerebral edema, and structural changes in the corpus
callosum have been noted. Additionally, only one neuroimaging study has been completed with
retinopathy-positive CM. As previously discussed, misdiagnoses can occur in pediatric CM
cases, but malarial retinopathy has demonstrated improved diagnostic sensitivity and specificity
(Beare et al., 2006). Moreover, the inclusion of retinopathy-positive cases enhances the
likelihood that true CM cases are being evaluated. Thus, future research should address pediatric
cases of retinopathy-positive CM.
The Present Study
The purpose of this study was to investigate corpus callosum morphology in children
diagnosed with retinopathy-positive and retinopathy-negative CM from Malawi, Africa. Little is
known about the structure of the brain of children diagnosed with retinopathy-positive CM and
the distinction between retinopathy-positive and retinopathy-negative pediatric CM has yet to be
evaluated. To date, only one pediatric neuroimaging study has been completed with children
diagnosed with retinopathy-positive CM. The current study was the second pediatric CM
neuroimaging study to investigate retinopathy-positive. As discussed, pediatric survivors of CM
demonstrate neurocognitive deficits similar to those noted in children diagnosed with ADHD
(e.g., inattention, working memory). Neuroimaging studies completed with children diagnosed
with ADHD have examined the corpus callosum because this structure is implicated in the

19

transference of information from one hemisphere to the other. Children diagnosed with ADHD
demonstrate difficulty responding to stimuli quickly, which suggests the children’s ability to
transfer incoming information may be compromised. As pediatric survivors of CM exhibit signs
of ADHD, an examination of the corpus callosum in these children was warranted. Regions of
the corpus callosum of particular interest were the genu and the splenium, as these two regions
have been implicated in sustained attention in children diagnosed with ADHD (e.g., Hynd et al.,
1991; Semrud-Clikeman et al., 1994).
Corpus callosum morphology in retinopathy-positive and retinopathy-negative children
was investigated. It was hypothesized that retinopathy-positive children would have smaller
overall corpus callosum area as compared to the retinopathy-negative children. It was also
hypothesized that the genu and the splenium would be significantly smaller in the retinopathypositive children, as compared to the retinopathy-negative children. Additionally, gender
differences were explored within the retinopathy-positive group, as there are no documented
studies examining potential gender differences in the structure of the brain in children in Africa
and none with survivors of CM.
Methods
Participants
Participants were 20 children, 11 males and nine females. The retinopathy-positive group
included 16 children and the retinopathy-negative group included four children. Of the 20
patients, the average age was 48.75 months (SD = 24.93), with a range of 16 months to 9 years.
The mean coma score at admission was 1.50 (SD = 0. 51). The mean length of coma was 44.14
hours (SD = 24.28). Six of the children, though they did not die, never fully regained
consciousness (i.e., not seeing, hearing, or talking), thus length of coma could not be determined
20

for these children. The average parasitemia count upon admission was 199971.40 (SD =
291856). Of the children admitted to the Paediatric Research Ward (see below), the geometric
means of parasitemia in the retinopathy-positive group was (M = 42273) and retinopathynegative group was (M = 21128). See Table 2 for a complete account of patient demographics
for the retinopathy-positive and negative groups. The participants were admitted to the Paediatric
Research Ward, Department of Pediatrics, Queen Elizabeth Central Hospital, Blantyre, Malawi
and treated for cerebral malaria. Queen Elizabeth Central Hospital cares for a population of
533,000 and serves as a teaching center for the only medical school in Malawi. Cases of
complicated malaria are referred for care at no cost. This study is part of a larger study of the
pathogenesis of fatal malaria.
Participant criteria. Cerebral malaria was defined in patients as the presence of P
falciparum parasitemia, an acute onset of coma, with a score of two or less on the Blantyre coma
scale (see Table 1) for a minimum of four hours following admission. No clinical or laboratory
evidence was present for an alternative cause of coma, including hypoglycemia, meningitis, or
postictal state following a seizure. All children were treated with intravenous quinine and fluid;
anticonvulsants, antipyretics, and glucose were administered as needed. To determine
retinopathy status, one physician examined each child by direct and indirect ophthalmoscopy
following pupil dilation using tropicamide 1% and phenylephrine 2.5%. Upon admission and
every day afterward, ophthalmoscopy was performed in accordance with patient cooperation and
depth of coma. Retinal findings were recorded on standardized charts and scored for severity.
The most advanced eye findings for each child were used and that information was relayed to the
investigator. Children with chronic malnutrition were excluded from the study.

21

Image Acquisition
A certified technician located at the Magnetic Resonance Imaging Centre, Queen
Elizabeth Central Hospital in Blantyre, Malawi, ran all scans. The images were read by an onsite, licensed radiologist trained by the Radiology Department at Michigan State University
(MSU). All scans selected for this study were free from gross abnormalities. A sagittal twodimensional T1-weighted spoiled gradient echo pulse sequence (3D-SPGR) was performed on
each subject on a .35 permanent magnet GE Signa Ovation J, with the following parameters:
Repetition time (TR) = 24.0 ms, echo time (TE) = 4.9 ms – 5.1 ms, flip angle = 90°, FOV = 26
cm, matrix = 160 x 128, slice thickness = 1.5 mm/0.0sp. MRI files were received from Malawi
via the MSU satellite downlink. The imaging data was transferred to the Department of
Radiology at MSU and were subsequently transferred to a DVD for morphometric analysis at the
Center for Neurodevelopmental Study.
Loss of scans. A total of 70 scans were acquired; however, fifty scans could not be used.
This was a retrospective study, as the scans were acquired prior to the design of the current
study. Thirty-six were not suitable for the specific analyses, as an SPGR series was not acquired;
eleven scans were uninterpretable due to image artifacts; one scan was from a child who was not
a contemporary study patient and two participants were not CM patients. As a result, 20 scans
were analyzed.
Image Analysis
Alignment. The T1-weighted images were transferred to the local Cognitive Imaging
Research Center (CIRC) server and were later processed using Analysis of Functional
NeuroImages (AFNI; Cox, 1996). To account for variance in head orientation during scanning, a
procedure termed anterior callosal-posterior callosal (AC-PC) alignment was used to align the

22

brains along the anterior and posterior commissures. First, both the superior and posterior edges
of the anterior commissure were identified and marked. Next, the inferior edge of the posterior
commissure was marked. Lastly, two additional midsagittal points were marked. Using these
five marks, AFNI then rotated the brain along the x-y-z axis in order to align the brain to a
consistent orientation. All subsequent area measurements were completed using a personal
computer (PC) version of AFNI (Cox, 1996).
Midsagittal slice identification. Before the corpus callosal (CC) and supratentorial area
measurements were completed, the midsagittal plane was first carefully identified for each
participant by a single investigator blind to patient diagnosis. The midsagittal plane was defined
as the plane in which the fornix, thalamus, and cerebellum were most clearly visible. The
midsagittal plane was then secondarily positive through careful visual inspection of the coronal
and axial plane to ensure that the selected midsagittal plane was centered between the left and
right hemispheres. Next, one-third of the brains were randomly selected to establish reliability
by a second independent rater also blind to patient diagnosis. Interrater agreement was 1.00. All
subsequent area measurements were completed on the identified midsagittal plane.
Corpus callosum area. The midsagittal area analysis of the CC was completed using the
AFNI plug-in module “Draw Dataset” to identify the boundaries of the CC. The CC was defined
as the higher-density band of fibers above the lateral ventricle and thalamus and below the
cingulate gyrus in the midsagittal plane, similar to the definition used by Fine, SemrudClikeman, Keith, Stapleton, and Hynd (2007). Using the DrawDataset module, the area defined
as the CC was “drawn” onto the overlay using a stylus and a drawing tablet. All voxels within
the identified CC were set to a value of “1”. Once the area of the CC was drawn for each patient,
the AFNI plug-in module “3Ddump98” was used to determine the number of voxels assigned to

23

a value of one. One-third of the brains were randomly selected and re-measured by the author
and intra-rater reliability was .989. See Figure 2 for a depiction of the CC area measurement.
Next, an automated program utilizing the x-y-z coordinates in AFNI was used to divide
the corpus callosum into five equal regions. These regions, in order from most anterior to most
posterior are the genu, body, midbody, isthmus, and splenium. When the anterior-posterior
lengths were not evenly divisible by five, 100 segmentation iterations with random placement of
remainders were performed. The calculated average number of voxels for each segment was
then imported into SPSS and used in later analyses. This method has also been used in a study
investigating the CC (Fine et al., 2007). See Figure 3 for an illustration of the CC segmentation.
Supratentorial area. Due to the wide age range, a supratentorial area measurement was
completed for each participant in order to control for brain size. The supratentorial area, defined
as the cerebrum without the cerebellum, was measured to determine intracranial area. It does not
include the skull or meninges. The AFNI DrawDataset module was again used to draw onto the
overlay using a stylus and a drawing tablet. All voxels identified as the intracranial area were
assigned the value “1”. Once the supratentorial area was drawn for each patient, the AFNI
module 3Ddump98 was used to determine the number of voxels assigned the value of “1”, thus
producing the supratentorial area measurement. See Figure 4 for an illustration of the
supratentorial area measurement. After completing the intracranial area measurements for each
patient, one-third of the brains were again randomly selected and re-measured by the author,
resulting in an intrarater reliability of .993.
Results
Power and Statistical Analyses

24

Power. The power to detect significant results was significantly reduced due to a limited
sample size (N = 20). Nonetheless, the following statistical analyses will provide pilot data on
this population and effect sizes rather than formal statistical analyses will be emphasized (Cohen,
1988).
Statistical analyses. First, a bivariate correlation was conducted examining the
relationship between the five segments of the CC (i.e., genu, body, midbody, isthmus, and
splenium), total CC area, supratentorial area, and age. The results indicated that each of the five
CC segments was significantly correlated with one another and total CC area (p < .001) as
expected. The five CC segments were not significantly correlated with the supratentorial area,
which was not expected. Additionally, age was not significantly correlated with the 5 CC
segments, overall CC area, or supratentorial area. Since, the correlation between age and
supratentorial area approached significance (p = .090), a conservative approach was used and it
was decided that age would be used as a covariate in the following analyses. It is particularly
important to control for age, as the brain size increases until around the age of 12. See Table 3
for correlation values. Next, to examine whether overall brain size differed between the
retinopathy-positive and negative groups, an exploratory one-way ANCOVA was conducted.
No significant group differences were found on brain size, F(1,17) = 3.47, n.s. The effect size
was large (d = 1.11), indicating a considerable difference in brain size due to retinopathy status,
such that the retinopathy-positive group (M = 9096.13, SD = 527.36) was found to have smaller
brain size than the retinopathy-negative group (M = 9699.50, SD = 557.25).
A one-way ANCOVA was conducted to determine whether total CC area differed
between the retinopathy-positive and negative groups and no significant group differences were
found on overall CC area, F(1,17) = 0.92, n.s. A moderate effect size was detected (d = .63),

25

which indicates a modest difference in CC area exists, such that the retinopathy-negative group
had larger overall CC (M = 642.75, SD = 98.89) than the retinopathy-positive group (M =
580.06, SD = 99.95). As a large effect size was detected for supratentorial area measurement,
suggesting the retinopathy-positive group may have a smaller overall brain area, a CC-to-brain
size ratio was calculated for each participant. An ANOVA was then conducted with the
calculated CC-to-brain size ratio and similar findings were found, F(1,17) = .082, n.s., d = .23.
A 2 (group) x 5 (CC measure) Wilks Lambda MANCOVA with age as a covariate was
conducted to determine whether the CC segments differed as a function of retinopathy status.
Likely due to the small sample size, the results revealed no significant group effect, F(1,17) =
.31, n.s. for CC segments. The effect of the covariate was also not significant, F(1,17) = .19, n.s.
To more fully understand possible differences due to retinopathy status, effect sizes were
calculated for each of the five segments of the corpus callosum to determine the magnitude of the
relationship between the two groups. A small effect size was found for the isthmus (d = .34),
moderate effect sizes were found for the genu (d = .45), body (d = .56), and midbody (d = .55),
and a large effect size was found for the splenium (d = .89). Moreover, these effect sizes
indicate that each of the five segments of the CC were smaller in the retinopathy-positive group.
See Table 4.
Gender. Exploratory analyses were also performed to investigate whether total brain
area and total CC area differ as a function of gender in the retinopathy-positive group. Such
analyses were not conducted with the retinopathy-negative group due to small numbers. A oneway ANCOVA, with age as a covariate was first performed to determine if overall brain area
differed between the retinopathy-positive males (n = 9) and females (n = 7) and a significant
group difference was found indicating that males had a larger brain (M = 9320.00, SD = 523.79)

26

size than females (M = 8808.29, SD = 397.00), F(1,13) = 4.70, p = .049. The effect size was
large (d = 1.40), indicating a considerable difference in brain size between brain males and
females. A second one-way ANCOVA was conducted to determine if overall CC area differed
between the retinopathy-positive males and females, but no significant group difference was
found, F(1,13) = .298, n.s., d = .28. See Table 5. Despite the previous insignificant result, a 2
(group) x 5 (CC segment) MANCOVA, with age as a covariate, was performed to determine if
the five CC segments differed as a function of gender. As predicted, no significant group effect
was detected, F(1,13) = .431, n.s. Effect sizes were calculated: genu (d = .11), body (d = .52),
midbody (d = .36), isthmus (d = .32), splenium (d = .10).
Discussion
Cerebral malaria is a common cause of illness and often death in children in sub-Saharan
Africa, resulting in an estimated case fatality rate of 19% (Snow et al., 1999). A new diagnostic
method, malarial retinopathy, has been documented to improve the diagnostic specificity and
sensitivity of the WHO (Beare et al., 2006; Taylor et al., 2004). As this is a new diagnostic
technique, few studies, including neuroimaging and neurocognitive studies, have used this
diagnostic method in conjunction with the WHO criteria to determine patient eligibility. Thus,
much of the already limited literature on pediatric CM may have inadvertently included non-CM
cases, which confounds the results in both the neurocognitive and neuroimaging literature.
Few neuroimaging studies have been completed with pediatric survivors of CM. Review
of the available literature on pediatric survivors of CM and children diagnosed with ADHD,
however, reveals that white matter abnormalities have been documented in both populations
(e.g., Hynd et al., 1991; Patankar et al., 2002; Semrud-Clikeman et al., 1994; Yadav et al., 2008).
A study investigating adult patients diagnosed with CM noted lesions in the splenium of the CC

27

(Yadav et al., 2008). The neuroimaging research has yet to investigate whether structural
differences exist between retinopathy-positive and retinopathy-negative cases of pediatric CM.
Consequently, the current study sought to investigate CC morphology in pediatric cases of
retinopathy-positive and retinopathy-negative CM.
Findings from the study provide some support the first hypothesis that children in the
retinopathy-positive group would have smaller overall CC area relative to the retinopathynegative children. Partial support was also found for the hypothesis that the genu and the
splenium would be significantly smaller in the retinopathy-positive children, as compared to the
retinopathy-negative children. The following sections will discuss the results of the proposed
hypotheses, as well as exploratory analyses conducted on gender in the retinopathy-positive
group. Next, the results of the completed analyses will be discussed in light of the available
research. Finally, the clinical implications and limitations of the current study will be discussed
and suggestions for future research will be proposed.
Major Findings
Prior to examining the proposed hypotheses, an examination of whether total brain area
differed between the retinopathy-positive and negative groups was completed. The results
yielded no significant group differences. Nonetheless, the large effect size detected indicates a
sizeable difference in brain size and suggests that the retinopathy-positive group may have a
smaller brain size compared to the retinopathy-negative participants. This finding was not
predicted, as diffuse cerebral edema is frequently noted in CM (e.g., Potchen et al., 2008). As a
result, it was expected that the retinopathy-positive group would have larger overall brain area,
due to the commonly documented widespread swelling of the brain. Age did not significantly
differ between the two groups, suggesting that variance in brain development did not account for

28

the large effect size. It is possible that the white matter is compromised in retinopathy-positive
children which resulted in smaller brain size. The current literature on CM indicates that white
matter abnormalities, including abnormalities in the CC, frequently occur (e.g., Patankar et al.,
2002; Yadav et al., 2008).
Some support was found for the first hypothesis that the children in the retinopathypositive group would have smaller overall CC area compared to the retinopathy-negative group.
Significant results were not detected from the main analyses; however, this finding may be
attributed to a lack of power to detect significant main effects as a result of the small sample size.
Therefore, an effect size was also calculated and a moderate effect size was detected, suggesting
that the retinopathy-positive children may have smaller overall CC area. Thus, consistent with
the literature previously mentioned, the results suggest that white abnormalities, particularly a
reduction in white matter in the CC may occur in the retinopathy-positive children.
The second hypothesis that two segments of the CC, the genu and the splenium, would be
significantly smaller in the retinopathy-positive group was partially supported. The results of the
main analyses did not yield significant results; however, this finding may, again, be attributed to
a lack of power. Consequently, effect sizes were calculated for each of the five segments and a
trend providing some support for the hypotheses was found. A moderate effect size was detected
for the genu, as well as the body and midbody, indicating that these regions were likely smaller
in the retinopathy-positive group. While it was hypothesized that the genu would be
significantly smaller in the retinopathy-positive group as compared to the retinopathy-negative
group, this finding suggests a trend in the data providing some support for the hypothesis. A
small effect size was observed for the isthmus of the CC; but, specific hypotheses were not
generated for this region of the CC, so this result was not surprising. Finally, a large effect size

29

was detected for the splenium, indicating that this region of the CC was smaller in the
retinopathy-positive group. This finding will be discussed in more detail in light of the available
research in the section below.
Splenium. Regarding previous findings on the splenium in the neuroimaging literature,
three studies examining adult patients diagnosed with CM and a fourth examining adult patients
diagnosed with epilepsy have discovered transient lesions in the splenium (Hantson et al., 2010;
Polster et al., 2001; Yadav et al., 2008). Additionally, the neuroimaging literature on children
diagnosed with ADHD has found abnormalities in the splenium of the CC (Hynd et al., 1991;
Semrud-Clikeman et al., 1994). It is possible that the splenium is important in many aspects of
neurocognitive functioning, particularly in attention. As a result, a reduction in white matter
may be associated with attentional processes and may therefore be linked to the deficits in
attention noted in pediatric survivors of CM.
Posner and colleagues (1990) have proposed that the attentional networks in brain may be
divided into three networks. The researchers argue that the attention system in the brain is
structurally distinct from data processing systems in the brain; a network of anatomical areas of
the brain are responsible for carrying out attention; and, the specific regions involved in attention
have different functions. Thus, using this model and previous research implicating the CC in
ADHD, it is possible that the posterior regions of the CC may play a critical role in attention
networks in the brain. Research has demonstrated that pediatric survivors of CM frequently
exhibit attention deficits (e.g., Boivin et al., 2007). Further research has demonstrated
similarities between retinopathy-positive pediatric CM and ADHD, such that 11% of children
diagnosed with CM in one study exhibited symptoms of ADHD (Birbeck et al., 2010).
Additionally, the literature on children diagnosed with ADHD in developed countries suggests

30

that the splenium of the CC may be associated with the observed attention deficits (e.g., Hynd et
al., 1991; Semrud-Clikeman et al., 1994). Thus, the results of this study, consistent with
previous research on individuals in developed countries, suggest it is likely that the posterior
regions of the CC, specifically the splenium, are implicated in attentional processes and are
therefore related to the neurocognitive outcomes of attention deficits in pediatric survivors of
CM.
Exploratory Analyses on Gender
Due to the limited sample size, particularly within the retinopathy-negative group,
exploratory analyses were conducted on the retinopathy-positive group to examine potential
gender differences. Previous work has found that men generally have larger brains compared to
women (e.g., Ankney, 1992). There are no documented studies looking at this finding in
children in Africa and none with survivors of CM. To that end, it was decided to explore
possible gender differences. It was found that the retinopathy-positive CM males had larger
brains than females. This finding is consistent with previous research which established that,
even when controlling for body size, males have larger brains than females (e.g., Ankney, 1992).
The current study was one of the first to possibly document this finding in African children. It is
important to note, however, that the children included in this study were all diagnosed with CM;
thus, this finding may not be readily extrapolated to the population at large.
An investigation of overall CC area was then conducted to determine if gender
differences were present, but no significant group difference was detected. While neuroimaging
studies regarding specific gender differences in the CC have not been conducted on individuals
diagnosed with CM, a large body of literature exists on this topic in control participants residing
in developed countries. The extant research on gender differences in the CC are highly

31

inconsistent, with some researchers finding that females tend to have a larger and more bulbous
splenium than males and others failing to replicate these findings (De Lacoste-Utamsing &
Holloway, 1982; Oppenheim, Benjamin, Lee, Nass & Gazzaniga, 1987). More recent research
indicates that brain volume is the main contributor to overall CC size (Jancke & Steinmetz,
2003). Jancke and Steinmetz (2003) suggest that because the CC-to-brain size relationship
follows a geometric rule, larger brains will be correlated with smaller CC-to-brain size ratios. As
females typically have smaller brains, it follows that they will have larger CC ratios.
Consequently, it was predicted that females would have larger overall CC areas; however, the
data in the current study did not support this prediction. As previously discussed, it is possible
that the white matter is compromised in pediatric survivors of CM and may be associated with
the neurocognitive deficits noted in these children. Future research should continue to examine
this possibility, as well as whether gender differences exist in specific regions of the CC, with a
larger sample size.
Limitations
The main limitation of the current study was the limited sample size, particularly with
regard to the retinopathy-negative group. The small sample likely reduced the power in the
completed analyses, thus decreasing the likelihood that group differences could be detected.
Thus, the results of the current study might best be understood as pilot data for future studies,
such that the current study may provide information on what is to be expected with a larger
sample. In addition, the sample of children included in the study varied widely in age. The ages
of the study patients ranged from sixteen months – nine years. Thus, the results may have been
influenced by the wide variance in age in the sample. Also, it is unclear why the brains of the
retinopathy-positive children have smaller brains than the retinopathy-negative children, as grey

32

and white matter analyses were not completed to determine if either tissue is specifically
affected. It may be the case that the brains of the retinopathy-positive children have always been
smaller than the retinopathy-negative children; but this hypothesis should be explored in future
studies (personal communication with Dr. Michael Potchen, 2011).
Clinical Implications
Current research on the neurological outcomes of pediatric CM indicates that deficits in
this domain typically resolve within six months post-discharge from the hospital (van Hensbroek
et al., 1997). Neurocognitive deficits, on the other hand, appear to be more persistent (e.g.,
Carter et al., 2005b; John et al., 2008). The available literature indicates that neurocognitive
deficits frequently occur in the domains of working memory and attention (Boivin et al., 2007).
Research has also suggested that pediatric survivors of CM and children diagnosed with ADHD
have similar neurocognitive profiles, such that 11% of pediatric survivors in one study also
exhibited symptoms of ADHD (Birbeck et al., 2010).
Also, as both CM and traumatic brain injury (TBI) are forms of acquired brain injury, an
examination of the pediatric TBI literature may be an appropriate means of elucidating the
neurocognitive outcomes one may expect to find following pediatric CM. While CM is the
result of a disease process and TBI is the result of an injury, little is known about acquired brain
damage in Africa and the closest approximation is the TBI literature. Deficits in attention and
executive functioning have also been documented to frequently occur in pediatric TBI cases
(e.g., Yeates, et al., 2005). In one investigation, Yeates and colleagues (2005) determined that
20% of children classified with a severe TBI (i.e., GCS ≤ 8 and brain abnormalities or sustained
loss of consciousness) manifested symptoms of inattention and impulsivity-hyperactivity in
accordance with a diagnosis of ADHD-Combined Type. Similarities between children with a

33

TBI and ADHD have also been demonstrated in investigations of brain structure, particularly
with regard to abnormalities in white matter, including the CC. While the results from the
current study were not statistically significant, a moderate effect size for CC area was found,
indicating that children diagnosed as retinopathy-positive may have smaller overall CC. Thus,
the results of this study indicate that the abnormalities in brain structure noted in pediatric
survivors of CM are likely consistent with the ADHD and TBI literature.
Recent research also suggests that these neurocognitive deficits may not be present
immediately following discharge from the hospital, but instead appear over time. A prospective
study of pediatric survivors of CM discovered that a variety of neurological and neurocognitive
deficits arose approximately three – ten months following recovery (Birbeck et al., 2010).
Specifically, motor, sensory, and language deficits were noted approximately three months
following discharge from the hospital. Disruptive behavioral disorders (i.e., ADHD) were then
evident in some of the survivors five months later. Lastly, seizures were noted approximately 10
months post-discharge from the hospital. Thus, many deficits, not only neurological, but
neurocognitive, may not be present immediately following recovery.
This finding has several clinical implications. First, most children are not followed after
discharge from the hospital; however, this article in conjunction with the current study suggests
that pediatric survivors of CM should be followed over time, as deficits in functioning may not
be evident immediately. Additionally, in light of this finding, it is suggested that at follow-up
visits, if children demonstrate poorer progress, additional testing and a new scan should be
performed. However, completing follow-up scans regardless of the child’s progress would be a
beneficial means of elucidating the possible mechanisms by which CM may cause structural
changes in the brain. Completing scans at the acute phase of the illness may not allow time for

34

changes in brain structure to develop. Furthermore, as recent research suggests that many
neurocognitive deficits (i.e., disruptive behaviors) are not evident until approximately five
months post-discharge from the hospital, follow-up scans at this time period may further
elucidate possible connections between anatomical changes in the brain and neurocognitive
functioning (Birbeck et al., 2010). Further thorough examinations would allow for a more
comprehensive understanding of the child’s level of functioning, which would permit
interventions to be tailored to the child accordingly.
Future Directions
As the current study is among the first to concurrently investigate retinopathy-positive
and negative cases of pediatric CM, future research should continue to investigate whether
structural differences exist in the brain of children diagnosed with retinopathy-positive and
negative CM. Further research would benefit from a more comprehensive understanding of
whether or not these two groups differ in terms of structural changes in the brain. Future
research should examine whether grey or white matter is affected, particularly in the retinopathypositive children to determine why the brains of these children are smaller than the retinopathynegative children. Also, it is critical to examine the relationship between attention measures and
brain structure, particularly concerning the CC. This examination would allow for a more direct
assessment of possible associations between neurocognitive functioning and structural findings
in the brain. Finally, it is important to begin linking neurocognitive and behavioral data with
structural changes in the brain to more fully understand the effect of acquired brain damage on
later functioning.

35

Table 1
Description of Coma Scales
Glasgow Coma Scale
Behavior
Eye Opening
Spontaneous
In response to speech
In response to pinprick
(pain)
No response
Motor (M)
Follows commands
Can localize pain
Withdraws from painful
stimulus
Abnormal flexion to pain
Extensor response to pain
No response
Verbal (V)
Oriented
Confused conversation
Inappropriate words
Incomprehensible sounds
No response

Blantyre Coma Scale
Points Behavior
Eye Movement
4
Watches or follows
3
Fails to watch or follow

Points
1
0

2
1
6
5
4

Best Motor Response
Localizes painful stimulus
Withdraws limb from painful stimulus
No response or inappropriate response

2
1
0

3
2
1
5
4
3
2
1

Best Verbal Response
Cries appropriately with painful stimulus
or if verbal speaks
2
Moan or abnormal cry with painful
stimulus
1
No vocal response to painful stimulus
0

Note. Glasgow Coma Scale source is Semrud-Clikeman & Teeter Ellison (2009). The Blantyre
Coma Scale source is Waller et al. (1995); Molyneux, Taylor, Wirima, & Borgsteinj, (1989).

36

Table 2
Participant Demographics
Retinopathy-positive
Retinopathy-negative
M
SD
M
SD
p
Age
47.00
20.98
55.80
40.68
0.04
Coma Score
1.50
0.52
1.50
0.58
1.00
Coma Duration
53.60
21.82
20.50
9.15
0.26
Seizures
0.75
0.50
0.75
0.44
1.00
Parasitemia
195854.40 295848.60
216439.50 360252.70
0.69
Note. Age in months, Coma Score = admitting Blantyre Coma Scale score, Coma Duration in
hours, Seizures = history of seizures, Parasitemia = parasitemia count on admission per ul.

37

Table 3
Correlations for the Measured Variables
Genu
Body
Midbody
Isthmus
Splenium
CCA
Genu
Body
.76**
Midbody
.76**
.93**
Isthmus
.63**
.75**
.81**
Splenium
.70**
.78**
.74**
.60**
CC Area
-.84**
-.88**
-.87**
-.72**
-.82**
SA
.25
.27
.31
.28
.29
-.02
Age
.26
.26
.27
.23
.23
-.22
Note. **p < 0.001, CC = corpus callosum, CCA = corpus callosum area, SA = supratentorial area.

38

SA

Age

.39

-

Table 4
Means, Standard Deviations, and Effect Sizes for the Retinopathy-Positive and Negative Groups
Retinopathy-Positive
Measure
M
SD
Supratentorial Area
9096.13
527.36
CC Area
580.06
99.95
5 CC Segments
Genu
183.51
32.62
Body
90.63
22.88
Midbody
75.41
15.24
Isthmus
73.67
15.77
Splenium
155.35
27.23
Note. CC = corpus callosum.

39

Retinopathy-Negative
M
SD
9699.50
557.25
642.75
98.89

d
1.11
0.63

198.95
102.30
84.04
79.26
176.44

0.45
0.56
0.55
0.34
0.89

35.98
18.53
16.11
17.01
19.55

Table 5
Means, Standard Deviations, and Effect Sizes for Retinopathy-Positive Group
Males
Measure
M
Supratentorial Area
9320.00
CC Area
566.78
Note. CC = corpus callosum.

Females

SD
523.79
53.18

M
8808.29
597.14

40

SD
397.00
143.53

d
1.40
.28

A) Retinal Whitening

B) Vessel Changes

C) Retinal Hemorrhages

D) Papilledema

Figure 1. Retinal findings. For interpretation of the references to color in this and all other
figures, the reader is referred to the electronic version of this thesis. This figure illustrates the
four retinal findings documented in pediatric patients diagnosed with cerebral malaria. The
pictures are courtesy of Terrie Taylor, DO, Malaria Research Project Ward, Blantyre, Malawi.

41

Figure 2. Corpus callosum area measurement. This figure illustrates the corpus callosum area
measurement completed for participant.

42

Figure 3. Segmentation of the corpus callosum. This figure illustrates a graphical representation
of the corpus callosum following the automatic segmentation process.

43

Figure 4. Supratentorial area measurement. This figure illustrates the supratentorial area
measurement completed for each participant.

44

REFERENCES

45

REFERENCES

Ankney, C. D. (1992). Sex differences in relative brain size: The mismeasure of women,
too? Intelligence, 16, 329-336.
Adams, S., Brown, H., & Turner, G. (2002). Breaking down the blood-brain barrier: Signaling a
path to cerebral malaria? TRENDS in Parasitology, 18, 360-366.
Aceng, J. R., Byarugaba, J. S., & Tumwine, J. K. (2005). Rectal artemether versus intravenous
quinine for the treatment of cerebral malaria in children in Uganda: Randomised clinical
trial. British Medical Journal, 330, 333-336.
Barennes, H., Balima-Koussoube, T., Nagot, N., Charpentier, J., & Pussard, E. (2006). Safety
and efficacy of rectal compared with intramuscular quinine for the early treatment of
moderately severe malaria in children: Randomised clinical trial. British Medical
Journal, 332, 1055-1059.
Beare, N. A., Southern, C., Chalira, C., Taylor, T. E., Molyneux, M. E., & Harding, S. P.
(2004a). Prognostic significance and course of retinopathy in children with severe
malaria. Archives of Ophthalmology, 122, 1141-1147.
Beare, N. A. V., Southern, C., Kayira, K., Taylor, T. E., & Harding, S. P. (2004b). Visual
outcomes in children in Malawi following retinopathy of severe malaria. British Journal
of Ophthalmology, 88, 321-324.
Beare, N. A. V., Taylor, T. E., Harding, S. P., Lewallen, S., & Molyneux, M. E. (2006). Malarial
retinopathy: A newly established diagnostic sign in severe malaria. American Journal of
Tropical Medicine and Hygiene, 75, 790-797.
Birbeck, G. L. (2004). Cerebral malaria. Current Treatment Option in Neurology, 6, 125-137.
Birbeck, G. L., Molyneux, M. E., Kaplan, P. W., Seydel, K. B., Chimalizeni, Y. K., Kawaza, K.,
Taylor, T. E. (2010). Blantyre Malaria Project Epilepsy Study (BMPES) of neurological
outcomes in retinopathy-positive paediatric cerebral malaria survivors: A prospective
cohort study. The Lancet, 9, 1173-1181.
Boivin, M. J. (2002). Cognitive impairment after cerebral malaria on cognitive ability in
Senegalese children. Developmental and Behavioral Pediatrics, 23, 353-364.
Boivin, M. J., Bangirana, P., Byarugaba, J., Opoka, R. O., Idro, R., Jurek, A. M., & John, C. C.
(2007). Cognitive impairment after cerebral malaria in children: A prospective study.
Pediatrics, 119, 360-366. doi:10.1542/peds.2006-2027
Boivin, M. J., Gladstone, M. J., Vokhiwa, M., Birbeck, G. L., Magen, J. G., Page, C., SemrudClikeman, M., Kauye, F., & Taylor, T. E. (2011). Developmental outcomes in
46

Malawian children with retinopathy-positive cerebral malaria. Tropical Medicine and
International Health, 16, 213 – 271.
Brown, H., Rogerson, S., Taylor, T., Tembo, M., Mwenechanya, J., Molyneux, M., & Turner, G.
(2001). Blood-brain barrier function in cerebral malaria in Malawian children. American
Journal of Tropical Medicine and Hygiene, 64, 207-213.
Carter, J. A., Ross, A. J., Neville, B. G. R., Obiero, E., Katana, K., Mung’ala-Odera, V., Lees,
J. A., & Newton, C. R. J. C. (2005a). Developmental impairments following severe
falciparum malaria in children. Tropical Medicine and International Health, 10, 3-10.
Carter, J. A., Mung’ala-Odera, V., Neville, B. G.R., Mturi, N., Musumba, C., & Newton,
C. R. J. C. (2005b). Persistent neurocognitive impairments associated with severe
falciparum malaria in Kenyan children. Journal of Neurology, Neurosurgery, and
Psychiatry, 76, 476-481.
Cascio, C., Styner, M., Smith, R. G., Poe, M. D., Gerig, G., Hazlett, H. C., Jomier, M., Bammer,
R., & Piven, J. (2006). Reduced relationship to cortical white matter volume revealed by
tractography-based segmentation of the corpus callosum in young children with
developmental delay. The American Journal of Psychiatry, 163, 2157-2163.
Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Hillsdale, NJ:
Lawrence Erlbaum Associates.
Cordoliani, Y., Sarrazin, J., Felten, D., Caumes, E., Leveque, C., & Alain, F. (1998). MR of
cerebral malaria. American Journal of Neuroradiology, 19, 871-874.
Cox-Singh, J., Davis, T. M. E., Lee K., Shamsul, S. S. G., Matusop, A., Ratnam, S., Rahman, H.
A., Conway, D. J., Singh, B. (2008). Plasmodium knowlesi malaria in humans is widely
distributed and potentially life threatening. Clinical Infectious Diseases, 46, 165-171.
de Lacoste-Utamsing, C., & Holloway, R. L. (1987). Sexual dimorphism in the human corpus
callosum. Science, 216, 143-1432.
de Souza, J. B., & Riley, E. M. (2002). Cerebral malaria: The contribution of studies in animal
models to our understanding of immunopathogenesis. Microbes and Infection, 4,
291-300.
Dondorp, A. M., Fanello, C. I., Hendriksen, I. C. E., Gomes, E., Seni, A., Kajal, D. C., . . .
White, N. J. (2010). Artesunate versus quinine in the treatment of severe falciparum
malaria in African children (AQUAMAT): An open-label, randomised trial. The Lancet,
376, 1647-1657.
Fine, J. G., Semrud-Clikeman, M., Keith, T. Z., Stapleton, L. M., Hynd, G. W. (2007). Reading
and the corpus callosum: An MRI family study of volume and area. Neuropsychology,
21, 235-241.

47

Gamanagatti, S., & Kandpal, H. (2006). MR imaging of cerebral malaria in a child. European
Journal of Radiology, 60, 46-47.
Holding, P.A., Stevenson, J., Peshu, N., & Marsh, K. (1999). Cognitive sequelae of severe
malaria with impaired consciousness. Transactions of the Royal Society of Tropical
Medicine and Hygiene, 93, 529-534.
Hynd, G. W., Semrud-Clikeman, M., Lorys, A. R., Novey, E. S., Eliopulos, D., & Lyytinen, H.
(1991). Corpus callosum morphology in Attention Deficit-Hyperactivity Disorder:
Morphometric analysis of MR. Journal of Learning Disabilities, 24, 141-146.
Idro, R., Carter, J. A., Fegan, G., Neville, B. G. R., Newton, C. R. J. C. (2006). Risk factors for
persisting neurological and cognitive impairments following cerebral malaria. Archives of
Disease in Childhood, 91, 142-148.
Jancke, L., & Steinmetz, H. (2003). Brain size: A possible source of interindividual variability
in corpus callosum morphology. In E. Zaidal & M. Iacoboni (Eds.), The parallel brain
(pp. 51-63). Cambridge, MA: The MIT Press.
John, C. C., Bangirana, P., Byarugaba, J., Opoka, R. O., Idro, R., Jurek, A. M., . . . Boivin, M. J.
(2008). Cerebral malaria in children is associated with long-term cognitive impairment.
Pediatrics, 122, 92-99. doi:10.1542/peds.2007-3709
Karunaweera, N. D., Grau, G. E., Gamage, P., Carter, R., & Mendis, K. N. (1992). Dynamics of
fever and serum levels of tumor necrosis factor are closely associated during clinical
paroxysms in Plasmodium vivax malaria. The Proceedings of the National Academy of
Sciences U. S. A., 89, 3200-3203.
Laufer, M. K., Thesing, P. C., Eddington, N. D., Masonga, R., Dzinjalamala, F. K., Takala, S. L.,
. . . Plowe, C. V. (2006). Return of chloroquine antimalarial efficacy in Malawi. The New
England Journal of Medicine, 355, 1959-1966.
Lewallen, S., Bronzan, R. N., Beare, N. A., Harding, S. P., Molyneux, M. E., & Taylor, T. E.
(2008). Using malarial retinopathy to improve the classification of children with cerebral
malaria. Transactions of the Royal Society of Tropical Medicine and Hygiene, 102, 10891094.
Looareesuwan, S., Wilairatana, P., Krishna, S., Kendall, B., Vannaphan, S., Viravan, C., White,
N. J. (1995). Magnetic resonance imaging of the brain in patients with cerebral malaria.
Clinical Infectious Diseases, 21, 300-309.
Molyneux, M. E., Taylor, T. E., Wirima, J. J., & Borgsteinj, A. (1989). Clinical features and
prognostic indicators in paediatric cerebral malaria: A study of 131 comatose Malawian
children. QJM: An International Journal of Medicine, 71, 441 – 459.
Muntendam, A. H., Jaffar, S., Bleichrodt, N., & van Hensroek, M. G. (1996). Absence of
48

neuropsychological sequelae following cerebral malaria in Gambian children.
Transactions of the Royal Society of Tropical Medicine and Hygiene, 90, 391-394.
Murphy, S. C., & Breman, J. G. (2001). Gaps in the childhood malaria burden in Africa:
Cerebral malaria, neurological sequelae, anemia, respiratory distress, hypoglycemia,
and complications of pregnancy. American Journal of Tropical Medicine and Hygiene,
64, 57-67.
Mwangi, T. W., Ross, A., Snow, R. W., Marsh, K. (2005). Case definitions of clinical malaria
under different transmission conditions in Kilifi District, Kenya. The Journal of
Infectious Diseases, 191, 1932-1939.
Newton, C. R. J. C., Chokwe, T., Armstrong Schellenberg, J., Winstanley, P. A., Forster, D.,
Peshu, N., . . . Marsh, K. (1997). Coma scales for children with severe falciparum
malaria. Transactions of the Royal Society of Tropical Medicine and Hygiene,
91, 161-165.
Newton, C. R. J. C., Peshu, N., Kendall, B., Kirkham, F. J., Sowunmi, A., Waruiru, C., . . .
Marsh, K. (1994). Brain Swelling and ischaemia in Kenyans with cerebral malaria.
Archives of Disease in Childhood, 70, 281-287.
Ngoungou, E. B., Dulac, O., Poudiougou, B., Druet-Cabanac, M., Dicko, A., Mamadou Traore,
A., . . . Preux, P. (2006). Epilepsy as a consequence of cerebral malaria in area in which
malaria is endemic in Mali, West Africa. Epilepsia, 47, 873-879.
Oppenheim, J. S., Benjamin, A. B., Lee, C. P., Nass, R., & Gazzaniga, M. S. (1987). No sexrelated differences in human corpus callosum on magnetic resonance imagery. Annals of
Neurology, 21, 604-606.
Pantankar, T. F., Karnad, D. R., Shetty, P. G., Desai, A. P., & Prasad, S. R. (2002). Adult
cerebral malaria: Prognostic importance of imaging findings and correlation with
postmortem findings. Radiology, 224, 811-816.
Polster, T., Hoppe, M., & Ebner, A. (2001). Transient lesion in the splenium of the corpus
callosum: Three further cases in epileptic patients and pathophysiological hypothesis.
Journal of Neurology, Neurosurgery, and Psychiatry, 70, 459-463.
Potchen, M. J., Bribeck, G. L., DeMarco, J. K., Kampondeni, S. D., Beare, N., Molyneux, M. E.,
& Taylor, T. E. (2010). Neuroimaging findings in children with retinopathy-confirmed
cerebral malaria. European Journal of Radiology, 74, 262 – 268.
Roch Leours, A., Mandujano, M., Romero, G. (2001). Ontogeny of brain and cognition:
Relevance to nutrition research. Nutrition Reviews, 59, s7-s12.
Saavedra-Lozano, J., Booth, T. N., Weprin, B. E., & Ramilo, O. (2001). Isolated cerebellar
edema and obstructive hydrocephalus in a child with cerebral malaria. The Pediatric
49

Infectious Disease Journal, 20, 908-911.
Semrud-Clikeman, M., Filipek, P. A., Biederman, J., Steingard, R., Kennedy, D., Renshaw, P., &
Bekken, K. (1994). Attention-Deficit Hyperactivity Disorder: Magnetic resonance
imaging morphometric analysis of the corpus callosum. Journal of the American
Academy of Child and Adolescent Psychiatry, 33, 875-881.
Semrud-Clikeman, M., & Ellison, T. (2009). Child neuropsychology: Assessment and
interventions for neurodevelopmental disorders. (2nd ed.). New York, NY: Springer.
Snow, R. W., Craig, M., & Marsh, K. (1999). Estimating mortality, morbidity and disability due
to malaria among Africa’s non-pregnant population. Bulletin of the World Health
Organization, 77, 624-640.
Taylor, T. E., Fu, W. J., Carr, R. A., Whitten, R. O., Mueller, J. G., Fosiko, N. G., . . .
Molyneux, M. E. (2004). Differentiating the pathologies of cerebral malaria by
postmortem parasite counts. Nature Medicine, 10, 143-145. doi:10.1038/nm986
Taylor, T. E., Wills, B. A., Kazembe, P., Chisale, M., Wirima, J. J., Ratsma, E. Y. E. C., &
Molyneux, M. E. (1993). Rapid coma resolution with artemether in Malawian children
with cerebral malaria. The Lancet, 341, 661-662.
Thumwood, C. M., Hunt, N. H., Clark, I. A., & Cowden, W. B. (1988). Breakdown of the bloodbrain barrier in murine cerebral malaria. Parasitology, 96, 579 – 589.
van Hensbroek, M. B., Palmer, A. Jaffar, S., Schneider, G., & Kwiatkowski, D. (1997). Residual
neurologic sequelae after childhood cerebral malaria. The Journal of Pediatrics, 131,
125-129.
Waller, D., Krishna, S., Crawley, J., Miller, K., Nosten, F., Chapman, D., . . . White, N. J.
(1995). Clinical features and outcome of severe malaria in Gambian children. Clinical
Infectious Diseases, 21, 577 – 587.
White, N. J. (1998). Not much progress in treatment of cerebral malaria. The Lancet, 352, 594595.
Whitty, C. J. M., Ansah, E., & Reyburn, H. (2005). Treating severe malaria: Rectal artemether
may be as good as intravenous quinine. British Medical Journal, 330, 317-318.
World Health Organization. (2000). Management of severe malaria (2nd ed.). Retrieved from
http://www.rollbackmalaria.org/docs/hbsm.pdf
Yadav, P., Sharma, R., Kumar, S. & Kumar, U. (2008). Magnetic resonance features of cerebral
malaria. Acta Radiologica, 49, 566-569.
Yeates, K. O., Armstrong, K., Janusz, J., Taylor, H. G., Wade, S., Stancin, T., & Drotar, D.
50

(2005). Long-term attention problems in children with traumatic brain injury. Journal of
the American Academy of Child and Adolescent Psychiatry, 44, 574-584.

51