Abstract
Background.
We explored three immunopathogenic biomarkers collected during acute malaria illness as potential moderators of gains from a computerized cognitive rehabilitation training (CCRT) intervention.
Method.
Von Willebrand Factor (vWF), tumor necrosis factor (TNF), and Regulated on Activation, Normal T Expressed and Secreted (RANTES) were assayed from plasma and cerebral spinal fluid (CSF) of children during acute severe malaria anemia or cerebral malaria. Two years after acute malaria illness, 150 surviving children and 150 non-malaria community controls (CC) from their households 6 to 12 years old entered a three-arm randomized controlled trial of titrating and non-titrating CCRT against no CCRT. Tests of cognition (Kaufman Assessment Battery for Children; KABC), Tests of Variables of Attention (TOVA), and Achenbach Child Behavior Checklist (CBCL) were administered before and after 24 CCRT sessions over a 3-month period, and at one-year follow-up. Differences in outcomes by trial arms and biomarker levels were evaluated using linear mixed effects models.
Results.
Severe malaria survivors with lower levels of vWF, lower CSF levels of TNF, and higher levels of plasma and CSF RANTES had better KABC cognitive performance after both titrating and non-titrating CCRT compared to no CCRT. For the CBCL, high plasma RANTES was associated with no benefit from either the titrating and non-titrating CCRT, while high TNF plasma was predictive of the benefit for both interventions. These biomarker moderating effects were not evident for CC children.
Conclusion.
Severe malaria immunopathogenic biomarkers may be related to poorer long-term brain/behavior function as evidenced by diminished benefit from a computerized cognitive rehabilitation intervention.
Keywords: Cerebral malaria, cognitive rehabilitation, computer games, neuropsychology, biomarkers
INTRODUCTION
A computerized cognitive rehabilitation training (CCRT) program called Captain’s Log® was designed to improve neuropsychological performance in children surviving malaria and other diseases that affect the brain.1,2 In past testing the CCRT intervention demonstrated benefits for visual-motor target chasing, learning, and behavioral outcomes, but not for attention, working memory, or academic performance among survivors of cerebral malaria (CM) and severe malaria anemia (SMA).1–3 While this evidence suggests the utility of CCRT for improving neuropsychological outcomes after malaria illness, it is not known which children stand to receive the most benefit from the CCRT, and for which children the CCRT may not be the right approach to cognitive rehabilitation. To address this knowledge gap, the present study explores whether the CCRT benefits differ according to immunopathogenic biomarkers of acute severe malarial illness (CM or SMA). Three biomarkers are evaluated in this study as potential moderators of the CCRT effects: Von Willebrand Factor (vWF), tumor necrosis factor (TNF), and Regulated on Activation, Normal T Expressed and Secreted (RANTES). They were selected based on the literature that supports their relationship to each of the proposed major mechanisms of brain injury from severe malaria.4
Biomarkers of endothelial activation increase significantly during acute CM illness and may also provide sensitive indicators of the likelihood of CM neuropathogenesis.5 The biomarker of vWF reflects endothelial activation during sequestration of parasitized erythrocytes in cerebral vasculature during acute illness.6,7 In CM, endothelial activation has been related to compromise in the blood/brain barrier (BBB) and the likely immunopathogenic pathway by which inflammation in the brain disrupts or damages the function of neural networks.8 In vivo von Willebrand Factor (vWF) was a sensitive biomarker of the activation of vascular endothelial cells in various vascular disorders (sickle cell anemia) and sepsis.9,10 Plasma vWF and its propeptides are significantly increased during acute CM illness and in mild malaria illness to a lesser extent when compared to a non-malaria comparison group of ill Malawian children.11,12 TNF and other cytokines may contribute to the pathogenesis of CM by promoting endothelial activation that accompanies P falciparum sequestration in acute CM.5
Regarding RANTES, our group documented that it is significantly lower in children with CM, and that very low levels of this cytokine are associated with mortality in this disease for severe malaria survivors in general, independent of other cytokine and chemokine levels that were examined.4,13 We propose that sequestration is visible during life in some CM patients as filling defects or mottling of retinal vessels,14 and is predictive of neuropsychological deficits in survivors of both severe malaria anemia (SMA) and CM.4 This cerebral vasculature neuropathogenesis may be a basis for the neurocognitive sequelae of severe malaria evident in poorer long-term brain/behavior function. In this study, we consider neurocognitive gains from CCRT intervention as a reflection of the integrity of brain/behavior function foundation to dynamic learning,15 and explore whether these gains differ according to vWF, TNF, and RANTES measured during acute malaria illness.
MATERIALS AND METHODS
Source of Children Enrolled in the Present Trial.
The CCRT trial dovetailed with an observational study (R01NS055349) of the neuro- and immunopathogenesis of severe malaria (both CM and SMA) in surviving children, along with community control (CC) children from their households.16,17 Children in the CC cohort did not have symptomatic malaria at the time of study enrollment. The severe malaria and CC children were eligible for enrollment in the present study once they completed their two-year follow-up assessments in the observational study. Two separate randomization procedures allocated children from each cohort into the CCRT, non-titrating CCRT, or usual care control arms (see Figure, Supplemental Digital Content 1).15
Study Recruitment, Enrollment, and Retention.
Institutional Review Board (IRB) approval for this study was obtained from Makerere University School of Medicine, University of Michigan, and Michigan State University. Research permission was obtained from the Uganda National Council for Science and Technology. Enrollment, assessment, and training took place from August 2012 to August 2016.
Inclusion Criteria:
1) Aged 6 to 12 years of age and did not meet any of the exclusion criteria listed below; 2) Signed consent from the parent/guardian, assent from children aged 7 years and older; 3). Completion of their 24 months of follow-up testing in the severe malaria pathogenesis source study.
Exclusion Criteria:
Same as the exclusion criteria used in the pathogenesis of severe malaria source study protocol.16–18 Exclusion criteria for all children included (1) known chronic illness requiring medical care; (2) known developmental delay; or (3) history of coma, head trauma, hospitalization for malnutrition, or cerebral palsy. Additional exclusion criteria for children with SMA included (1) impaired consciousness at physical examination, (2) other clinical evidence of central nervous system (CNS) disease, or (3) >1 seizure before admission.
Randomization and Masking
Following enrollment, severe malaria and CC children were randomized to either titrating CCRT, non-titrating CCRT, or passive control using a computer generated list of random numbers. Children assigned to either titrating or non-titrating CCRT completed three training sessions per week for eight weeks (24 training session–lasting about an hour). Training took place in a private setting after school near the child’s home, under the supervision of research assistant. Research assistants supervised the computerized CCRT training program, translating any screen instructions in English by speaking them to the children in the local language (Luganda), with sessions programmed to run for 45 to 60 minutes. The CCRT training program for this study was previously described,15 and is briefly summarized below.
Titrating CCRT.
Children assigned to the titrating CCRT treatment arms started at the simplest program level for each task,15 and Captain’s Log® adjusted the difficulty level of each task to the performance level of the child (titration). In other words, as the child achieved mastery at a given level of difficulty, that task was automatically adjusted at the next level of difficulty.
Non-titrating CCRT.
Children assigned to the non-titrating CCRT arm also started at the simplest program level for each task. However, for this arm, we programmed Captain’s Log® to not progressively advance to the next difficulty level, but to rotate randomly in a set of simple to moderate levels of training for the entire session.
Neuropsychological and Behavioral Outcomes for CCRT Intervention
Neuropsychological assessment took place at the Mulago Hospital pediatric severe malaria clinic before the intervention (baseline), post-intervention (after 24 CCRT sessions, at about 8 weeks), and 1 year after completion of CCRT training. Training and testing was done in the local language (Luganda) by Bachelor’s Degree graduates of Psychology from Makerere University. Testers were masked to intervention arm for the study children, and their malaria cohort status. The principal outcomes for the trial were selected because they represented summary measures of key domains of child behavior (Figure 1).15
Figure 1:
Modifying effect of immunologic biomarkers (CSF and plasma) on the principal neuropsychologic (KABC-II and TOVA) and psychiatric (CBCL) outcomes for computerized cognitive rehabilitation training (CCRT). Severe malaria survivors and CC case/control cohorts are compared in this randomized controlled trial.
Kaufman Assessment Battery for Children – Second Edition (KABC-II).19
The KABC-II evaluates sequential (working memory) and simultaneous processing (visual-spatial analysis and problem solving). This edition also evaluates learning as well as the executive function domain of planning/reasoning. All domains are summarized in the mental processing index (MPI), which was used as a measure of global cognitive performance. This test has been used at our Ugandan study site to document persisting neurocognitive sequelae in severe malaria survivors.18,20–22
Tests of Variables of Attention (TOVA) (www.tovatest.com).23
The TOVA visual test is a computerized, objective test of attention. The test provided us with a principal outcome of vigilance attention (D prime signal detections for the target; higher is better) and impulsivity (percent of non-signal presentations to which the child responded as if it were a signal; higher is worse). These measures have proven sensitive to the persisting neurocognitive effects of cerebral malaria in our previous work.22,24,25
Achenbach Child Behavior Checklist (CBCL) – School Age Version.26,27
This is a psychiatric screening questionnaire for emotional (Internalizing Symptoms), behavior (Externalizing Symptoms), and total psychiatric symptoms and problems for the study child. This questionnaire was read to the principal caregiver out loud in Luganda and did not rely on caregiver literacy. The tool has been validated with severe malaria survivors in Uganda.28,29 The scale measures for this instrument were standardized by age and sex using the available global cross-cultural norms.30
Measures of the following important contextual factors were used as covariates.31,32,33 The middle childhood version of the Caldwell Home Observation for the Measurement of the Environment (HOME) was used to assess the stimulation and learning opportunities offered by the child’s home environment. The Socioeconomic Status (SES) evaluation score was based the checklist of material possessions as opposed to income, characteristics of the home structure and living space, and parental education and occupational levels. Height, weight, and body mass index at the time of enrollment into the trial were standardized on the basis of the WHO 2007 reference norms.
Immunoassay Biomarkers
Immunology biomarkers data were collected as part of the observational study (R01NS055349) of the pathogenesis of severe malaria. As previously described in the publications based on the observational study,34,35 cytokine testing was performed on admission plasma and CSF samples. CSF samples were obtained as part of standard work-up for children with CM in whom a lumbar puncture was not contraindicated. The children with SMA have no clinical indication for a lumbar puncture, so only plasma biomarkers were measured for these children. For children with CM, 0.5–2 mL of the CSF sample was stored at −80°C for later study testing.
Plasma and CSF levels of TNF were measured by magnetic cytometric bead assay (EMD-Millipore, Billerica, MA, USA) according to manufacturer’s instructions with a BioPlex-200 system (Bio-Rad, Hercules, CA, USA). Peripheral blood smears were assessed for Plasmodium species by microscopy with Giemsa staining using standard protocols. PfHRP-2 quantification was performed using the Malaria Ag CELISA (Cellabs, Brookvale, Australia). VWF activity was quantified using the human angiopoietin-2 DuoSet ELISA kit (R&D Systems) and REAADS VWF activity ELISA kit (Corgenix, Broomfield, CO, USA), respectively. RANTES was measured by microbead suspension array technology (SAT) using the Luminex system (Austin, TX) and human-specific bead sets (R&D Systems, Minneapolis, MN).
Statistical Analyses.
The KABC-II manual with American norms was used to obtain standardized and scaled scores to adjust for age and compute the KABC-II Mental Processing Index. The CBCL age- and gender-based norms were used to score this tool. Raw score performance outcomes were used for the TOVA D prime (signal detection of overall performance) and percent commission errors (impulsivity) outcomes.
In the absence of established cut-points for the biomarker levels in this population, the distributions of the biomarkers were assessed, and the biomarker levels were dichotomized as high versus low using the upper tertile of each the biomarker distribution in each cohort (severe malaria or CC) and accounting for the natural breaks in distributions. In addition, the distribution of the biomarker across trial arms was taken into account to ensure that the counts for low and high categories were at least 10 for all arms, and the distributions of high versus low values were similar in the two strata. Because biomarker values for the cerebrospinal fluid were only available for children with CM, they were dichotomized as high versus low at the median.
Since child outcomes followed approximately normal distributions, linear mixed effects (LME) models were fit for two repeated measures of each outcome, with child identifier specified as a random effect. In accordance with randomization, separate LME models were fit the malaria and CC children. Each outcome was analyzed separately in relation to each biomarker as a potential moderator. The least square (LS) means for the biomarker level by trial arm interaction were output from the model. Since the p-values may not reflect the magnitude of the effects,36 we have estimated the standardized mean differences (SMDs) for each CCRT arm versus control by differencing the LS means and dividing by the standard deviation of the control group at baseline. The smallest meaningful SMDs are considered to be near 0.33; 0.5 is often considered a threshold for important differences, while SMDs of 0.8 are greater are considered large.37–39 The patterns of differences in magnitude of the LS means and SMDs were evaluated for the evidence of potential moderating effects. For example, if CCRT effects were different in magnitude for two levels of the biomarker, then generation of moderation hypotheses based on these data would be supported.
RESULTS
The malaria survivors cohort of N=150 children included 93 children with CM (62%), and 57 children with SMA (38%) (see Table, Supplemental Digital Content 2), and the CC cohort included 150 children. There were approximately equal proportions of boys and girls, and age ranged from 5 to 13 years in each cohort. Among malaria survivors, median time from clinical disease to baseline neuropsychological assessment of this trial was 2.5 years. Only two children dropped out before the post-intervention assessment in the malaria cohort. Two more malaria survivors and one CC child dropped out between the 2-month (post CCRT) and 1-year assessments, but they were included in the analysis under the missing at random assumption.
In the malaria cohort, vWF in plasma was available for N=137 children; TNF in plasma was available for N=145, TNF in CSF was available for N=68 (CM only); RANTES in plasma was available for all N=150 children, and RANTES in CSF was available for N=67 CM survivors (Table 1). In the CC cohort, vWF in plasma was available for N=100 children, plasma TNF was available for N=121, and plasma RANTES was available for N=125 children. The correlations among plasma biomarker levels were very weak (0.21 or less in absolute value) in both cohorts. The only moderate correlation of 0.46 was observed between CSF RANTES and CSF TNF among CM survivors.
Table 1.
Distributions of biomarkers by trial arm for severe malaria survivors and community control (CC) comparison children without a history of severe malaria.
| Severe malaria survivors | Community control children | |||||
|---|---|---|---|---|---|---|
| Characteristic# | Titrating CCRT## N=51 | Non-titrating CCRT N=54 | Control N=45 | TitratingCCRT N=50 | Non-titrating CCRT N=55 | Control N=45 |
| Mean (St Dev) | Mean (St Dev) | Mean (St Dev) | Mean (St Dev) | Mean (St Dev) | Mean (St Dev) | |
| vWF (plasma) | 203.42 (160.02) | 174.05 (122.57) | 239.88 (209.12) | 46.06 (37.00) | 71.81 (54.84) | 49.70 (44.63) |
| TNF (plasma) | 116.87 (97.85) | 161.70 (235.24) | 114.36 (112.36) | 41.71 (35.13) | 33.38 (34.25) | 27.73 (15.60) |
| TNF (CSF), CM | 1.95 (2.21) | 1.79 (2.27) | 1.55 (1.75) | N/A | N/A | N/A |
| RANTES (plasma) | 4987.50 (4949.41) | 3371.90 (3113.68) | 2827.92 (2124.20) | 8231.81 (5028.68) | 7960.16 (5022.52) | 7642.85 (5563.06) |
| RANTES (CSF), CM | 12.15 (6.28) | 26.10 (66.50) | 39.44 (116.42) | N/A | N/A | N/A |
| Median (IQR) | Median (IQR) | Median (IQR) | Median (IQR) | Median (IQR) | Median (IQR) | |
| vWF (plasma) | 148.17 (158.86) | 153.81 (173.07) | 175.37 (212.60) | 35.03 (50.17) | 64.40 (51.58) | 42.68 (38.08) |
| TNF (plasma) | 89.46 (107.12) | 91.21 (99.83) | 92.46 (87.13) | 35.03 (50.16) | 64.40 (51.58) | 24.28 (14.56) |
| TNF (CSF), CM | 1.44 (2.04) | 1.10 (2.17) | 0.86 (2.13) | N/A | N/A | N/A |
| RANTES (plasma) | 3442.21 (4797.23) | 2188.31 (3130.87) | 2091.01 (2129.62) | 7540.01 (8284.15) | 7371.59 (8903.45) | 7124.25 (7671.99) |
| RANTES (CSF), CM | 9.85 (5.66) | 8.95 (9.29) | 9.87 (4.24) | N/A | N/A | N/A |
| N (%) | N (%) | N (%) | N (%) | N (%) | N(%) | |
| vWF (plasma) | ||||||
| Low | 28 (63.64) | 34 (66.67) | 24 (57.14) | 21 (67.74) | 16 (44.44) | 24 (72.73) |
| High | 16 (36.36) | 17 (33.33) | 18 (42.86) | 10 (32.26) | 20 (55.56) | 9 (27.27) |
| TNF (plasma) | ||||||
| Low | 30 (62.50) | 35 (66.04) | 32 (72.73) | 21 (55.26) | 31 (67.39) | 25 (67.57) |
| High | 18 (37.50) | 18 (33.96) | 12 (27.27) | 17 (44.74) | 15 (32.61) | 12 (32.43) |
| TNF (CSF), CM | ||||||
| Low | 9 (45.00) | 13 (48.15) | 11 (52.38) | N/A | N/A | N/A |
| High | 11 (55.00) | 14 (51.85) | 10 (47.62) | |||
| RANTES (plasma) | ||||||
| Low | 30 (58.82) | 38 (70.37) | 35 (77.78) | 24 (60.00) | 29 (60.42) | 26 (70.27) |
| High | 21 (41.18) | 16 (29.63) | 10 (22.22) | 16 (40.00) | 19 (39.58) | 11 (29.73) |
| RANTES (CSF), CM | ||||||
| Low | 10 (52.63) | 15 (55.56) | 10 (47.62) | N/A | N/A | N/A |
| High | 9 (47.37) | 12 (44.44) | 11 (52.38) | |||
RANTES: Regulated on Activation, Normal T Expressed and Secreted; vWF: von Willebrand Factor; TNF: Tumor Necrosis Factor; CSF: cerebral spinal fluid;
CCRT: computerized cognitive rehabilitation training.
Among malaria survivors, plasma TNF below 120 pg/mL was categorized as low (67%). Values of TNF in plasma of 120 pg/mL or greater were categorized as high (33%). For CC children, low TNF in plasma was defined as below 33 (64%) versus high TNF in plasma of 60 pg/mL or greater (36%). Among CM children only, TNF in CSF was categorized as high if greater than or equal to 1.1 pg/mL (51%) and low if less than 1.1 pg/mL (49%). Among malaria survivors, RANTES in plasma below 3600 pg/mL was categorized as low (68%). Values of RANTES in plasma of 3600 pg/mL or greater were categorized as high (32%).
Results for the Kaufman Assessment Battery for Children (2nd edition) (KABC)
For the titrating CCRT, greater benefit was seen for children with high TNF plasma (SMD 0.46) or CSF RANTES (SMD 0.63) (Table 2). If TNF plasma level was ignored, then group difference would have been approximately 1.25 points (SMD=0.14); accounting for the TNF plasma level, the difference was nearly zero al low TNF plasma, and approximately 4 points at high TNF plasma, further illustrating the potential moderating effect. Similarly with CSF RANTES, overall differences between the titrating CCRT and control groups were 4.3 points for children with CM and available CSF RANTES levels; these differences were approximately 3.3 points among those with low CSF RANTES, and 5.6 points among those with high CSF RANTES.
Table 2.
Severe malaria survivors: the least square (LS) means and 95% confidence intervals (CIs) for post-intervention neuropsychologic outcomes (mixed model average over time) by trial arm and biomarker level
| Outcome# | Biomarker level## | Titrating CCRT~, LS Mean (95% CI) | Non-titrating CCRT, LS Mean (95% CI) | Control, LS Mean (95% CI) | P-value for treatment arm by biomarker level interaction | SMD for titrating CCRT v. control* | SMD for non-titrating CCRT v. control* |
|---|---|---|---|---|---|---|---|
| KABC mental processing index | vWF plasma low | 65.58 (63.33, 67.85) |
69.16 (67.09, 71.24) |
64.99 (62.59, 67.38) |
.06 | .07 | 0.47 |
| vWF plasma high | 68.30 (65.33, 71.26) |
65.90 (63.03, 68.76) |
65.60 (62.82, 68.38) |
.30 | 0.003 | ||
| TNF plasma low | 66.23 (64.04, 68.43) |
67.59 (65.57, 69.62) |
66.32 (64.22, 68.41) |
.17 | −0.01 | 0.14 | |
| TNF plasma high | 68.14 (65.30, 70.98) |
70.00 (67.14, 72.86) |
64.03 (60.62, 67.44) |
0.46 | 0.67 | ||
| TNF CSF low | 66.48 (62.43, 70.53) |
69.01 (65.61, 72.40) |
61.73 (57.83, 65.58) |
.53 | 0.53 | 0.82 | |
| TNF CSF high | 67.80 (63.91, 71.70) |
66.97 (63.68, 70.27) |
63.65 (59.79, 67.50) |
0.46 | 0.37 | ||
| RANTES plasma low | 66.97 (64.37, 69.20) |
68.94 (66.96, 70.93) |
66.03 (64.01, 68.05) |
.92 | 0.11 | 0.33 | |
| RANTES plasma high | 67.13 (64.47, 69.79) |
68.26 (65.08, 71.44) |
66.44 (62.49, 70.38) |
0.07 | 0.20 | ||
| RANTES CSF low | 64.98 (61.32, 68.63) |
65.43 (62.57, 68.29) |
61.66 (57.85, 65.48) |
.50 | 0.37 | 0.42 | |
| RANTES CSF high | 68.50 (64.72, 72.28) |
70.66 (67.16, 74.17) |
62.86 (59.24, 66.48) |
0.63 | 0.88 | ||
| TOVA D’Prime | vWF plasma low | 2.23 (1.96, 2.51) |
2.31 (2.06, 2.55) |
2.15 (1.86, 2.44) |
.40 | 0.06 | 0.12 |
| vWF plasma high | 2.46 (2.10, 2.82) |
2.13 (1.78, 2.47) |
2.25 (1.91, 2.58) |
0.16 | −0.12 | ||
| TNF plasma low | 2.18 (1.91, 2.44) |
2.26 (2.02, 2.50) |
2.23 (1.98, 2.47) |
.53 | −0.03 | 0.02 | |
| TNF plasma high | 2.45 (2.11, 2.79) |
2.23 (1.89, 2.57) |
2.20 (1.80, 2.60) |
0.19 | 0.02 | ||
| TNF CSF low | 2.71 (2.22, 3.20) |
2.11 (1.70, 2.51) |
2.13 (1.67, 2.58) |
.22 | 0.45 | −0.02 | |
| TNF CSF high | 2.09 (1.64, 2.55) |
2.29 (1.90, 2.67) |
2.07 (1.62, 2.53) |
0.02 | 0.17 | ||
| RANTES plasma low | 2.15 (1.89, 2.40) |
2.34 (2.12, 2.56) |
2.20 (1.98, 2.43) |
.04 | 0.03 | 0.11 | |
| RANTES plasma high | 2.60 (2.29, 2.91) |
2.05 (1.69, 2.42) |
2.41 (1.97, 2.85) |
0.16 | −0.27 | ||
| RANTES CSF low | 2.41 (1.95, 2.88) |
2.15 (1.79, 2.52) |
2.10 (1.64, 2.57) |
.87 | 0.24 | 0.04 | |
| RANTES CSF high | 2.27 (1.78, 2.76) |
2.22 (1.78, 2.66) |
2.02 (1.56, 2.49) |
0.19 | 0.16 | ||
| TOVA % commission errors | vWF plasma low | 13.54 (9.34, 17.73) |
12.02 (8.20, 15.85) |
13.03 (8.65, 17.40) |
.78 | −0.03 | 0.06 |
| vWF plasma high | 9.37 (3.82, 14.92) |
9.24 (3.85, 14.64) |
12.16 (7.04, 17.29) |
0.17 | 0.18 | ||
| TNF plasma low | 11.62 (7.55, 15.68) |
10.27 (6.55, 14.00) |
13.31 (9.53, 17.09) |
.50 | 0.10 | 0.19 | |
| TNF plasma high | 11.56 6.41, 16.72) |
12.77 (7.57, 17.97) |
10.20 (4.01, 16.39)) |
−0.08 | −0.16 | ||
| TNF CSF low | 6.08 (0, 13.23) |
12.87 (7.04, 18.70) |
10.29 (3.72, 16.85) |
.10 | 0.26 | −0.16 | |
| TNF CSF high | 17.55 (10.80, 24.39) |
9.38 (3.67, 15.10) |
14.05 (7.35, 20.75) |
−0.22 | 0.29 | ||
| RANTES plasma low | 14.29 (10.41, 18.17) |
10.33 (6.94, 13.72) |
11.94 (8.45, 15.43) |
.02 | −0.15 | 0.10 | |
| RANTES plasma high | 6.16 (1.40, 10.92) |
13.20 (7.56, 18.85) |
14.63 (7.88, 21.39) |
0.52 | 0.09 | ||
| RANTES CSF low | 8.91 (2.32, 15.49) |
12.54 (7.34, 17.75) |
9.70 (3.00, 16.39) |
.17 | 0.05 | −0.17 | |
| RANTES CSF high | 17.07 (9.77, 24.37) |
9.47 (3.07, 15.86) |
15.43 (8.75, 22.12) |
−0.10 | 0.37 | ||
| CBCL internalizing symptoms | vWF plasma low | 57.49 (54.73, 60.25) |
61.55 (58.71, 64.39) |
57.49 (54.24, 60.73) |
.26 | 0.00 | −0.43 |
| vWF plasma high | 57.92 (54.33, 61.51) |
58.50 (55.00, 62.00) |
60.00 (56.16, 63.83) |
0.01 | 0.16 | ||
| TNF plasma low | 59.05 (56.47, 61.63) |
59.82 (57.37, 62.27) |
58.73 (56.01, 61.46) |
.11 | −0.03 | −0.12 | |
| TNF plasma high | 54.54 (51.08, 58.01) |
61.86 (58.09, 65.63) |
58.75 (53.33, 64.17) |
0.45 | −0.33 | ||
| TNF CSF low | 60.87 (56.60, 65.13) |
62.83 (58.44, 67.22) |
59.34 (54.12, 64.56) |
.50 | −0.16 | −0.37 | |
| TNF CSF high | 55.49 (51.23, 59.76) |
59.83 (56.47, 63.20) |
59.39 (54.85, 63.92) |
0.42 | −0.05 | ||
| RANTES plasma low | 56.89 (54.10, 59.65) |
61.43 (59.03, 63.83) |
59.33 (56.86, 61.81) |
.02 | 0.26 | 0.22 | |
| RANTES plasma high | 59.17 (56.36, 61.98) |
59.23 (55.46, 63.00) |
48.60 (40.40, 56.80) |
−1.13 | −1.14 | ||
| RANTES CSF low | 57.88 (53.55, 62.20) |
59.69 (56.14, 63.24) |
59.59 (55.12, 64.05) |
.70 | 0.18 | −0.001 | |
| RANTES CSF high | 58.65 (53.76, 63.53) |
62.93 (56.40, 69.45) |
58.50 (52.56, 64.45) |
−0.02 | −0.47 | ||
| CBCL externalizing symptoms | vWF plasma low | 56.41 (53.56, 59.26) |
61.35 (58.49, 64.20) |
60.52 (57.17, 63.87) |
.13 | 0.50 | −0.10 |
| vWF plasma high | 58.13 (54.44, 61.82) |
56.73 (53.16, 60.31) |
61.45 (57.53, 65.36) |
0.41 | 0.58 | ||
| TNF plasma low | 57.47 (54.73, 60.21) |
60.34 (57.69, 62.78) |
59.64 (56.75, 62.52) |
.05 | 0.27 | 0.01 | |
| TNF plasma high | 56.00 (52.32, 59.68) |
57.96 (53.98, 61.94) |
66.63 (60.96, 72.31) |
1.30 | 1.06 | ||
| TNF CSF low | 60.06 (54.87, 65.25) |
60.57 (55.33, 65.81) |
65.01 (58.78, 71.24) |
.95 | 0.61 | 0.54 | |
| TNF CSF high | 57.33 (52.23, 62.43) |
59.03 (55.01, 63.05) |
61.78 (56.40, 67.15) |
0.55 | 0.34 | ||
| RANTES plasma low | 57.54 (54.62, 60.45) |
60.05 (57.52, 62.57) |
61.77 (59.15, 64.38) |
.14 | 0.52 | 0.21 | |
| RANTES plasma high | 57.05 (54.09, 60.01) |
59.46 (55.45, 63.47) |
51.67 (43.00, 60.35) |
−0.66 | −0.96 | ||
| RANTES CSF low | 59.23 (54.64, 63.82) |
60.24 (56.74, 63.74) |
63.52 (58.75, 68.30) |
.61 | 0.53 | 0.40 | |
| RANTES CSF high | 55.33 (50.00, 60.66) |
55.00 (48.69, 61.31) |
63.37 (57.11, 69.64) |
0.99 | 1.03 |
Negative sign before SMD absolute value indicates differences favoring the control group. SMDs exceeding 1/3 are bolded.
KABC: Kaufman Assessment Battery for Children, 2nd edition; TOVA: Tests of Variables of Attention; CBCL: Achenbach Child Behavior Checklist.
RANTES: Regulated on Activation, Normal T Expressed and Secreted; vWF: von Willebrand Factor; TNF: Tumor Necrosis Factor; CSF: cerebral spinal fluid; ##CCRT: computerized cognitive rehabilitation training.
CCRT: computerized cognitive rehabilitation training.
For the non-titrating CCRT, greater benefit was for children with low plasma vWF (SMD 0.47), high plasma TNF (SMD 0.67), low CSF TNF (SMD 0.82), or high CSF RANTER (SMD 0.88). Without accounting for the levels of these biomarkers, overall differences between non-titrating CCRT arm and control would have corresponded to SMD of 0.32 among children with available plasma biomarker levels, and SMD of 0.58 among CM children with available CSF biomarker data. The SMDs for the other level of these biomarkers were much lower (Table 2), suggesting potential moderating effects.
Results for the Tests of Variables of Attention (TOVA)
The titrating CCRT had greater benefits for TOVA D prime when CFS TNF was low (SMD 0.45) and for TOVA percent commission errors when RANTES in plasma was high (SMD 0.52) (Table 2). The corresponding overall SMDs across biomarker levels were markedly different, 0.04 and 0.07, respectively, indicating that trial arm differences were much larger than overall average at these biomarker levels. The non-titrating CCRT had greater benefit for TOVA percent commission errors when CSF RANTES was high (SMD 0.37), compared to the biomarker-averaged overall SMD of 0.08.
Results for the Achenbach Child Behavior Checklist (parent report) (CBCL)
The titrating CCRT had greater benefits for children with high plasma TNF, or CSF TNF, or high CSF RANTES (for externalizing symptoms, Table 2). The non-titrating CCRT had greater benefits on externalizing symptoms for children with high vWF (SMD 0.58), high TNF plasma (SMD 1.06), or high CSF RANTES (SMD 1.03). The corresponding overall SMDs were 0.13 among all malaria survivors, and 0.51 among CM malaria survivors who had the values of CSF biomarkers. When RANTES plasma was low, there were no benefits for either internalizing or externalizing symptoms from the titrating or non-titrating CCRT, with large SMDs favoring the control condition.
For the CC cohort, the potential modifying effects of biomarkers were few and smaller in magnitude compared to the malaria survivors. There was no indication of moderating effects for the KABC MPI. Titrating CCRT showed greater benefits when on TOVA commission errors for children when vWF was low (Table 3), while non-titrating CCRT was more beneficial when vWF was high. For the CBCL, non-titrating CCRT was more beneficial for the externalizing problems when TNF was high, while titrating CCRT as associated with greater benefits for the internalizing problems when vWF or TNF were high or RANTES was low.
Table 3.
Community Control (CC) non-severe-malaria comparison children: the least square (LS) means and 95% confidence intervals (CIs) for post-intervention neuropsychological outcomes (mixed model average over time) by trial arm and biomarker level
| Outcome# | Biomarker level## | Titrating CCRT~, LS Mean (95% CI) |
Non-titrating CCRT, LS Mean (95% CI) | Control, LS Mean (95% CI) | P-value for arm*biomarker level interaction | SMD for titrating CCRT v. control* | SMD for non-titrating CCRT v. control* |
|---|---|---|---|---|---|---|---|
| KABC mental processing index | vWF plasma low | 71.85 (68.92, 74.78) |
70.79 (67.33, 74.25) |
71.90 (69.13, 74.68) |
.56 | −0.005 | −0.11 |
| vWF plasma high | 71.14 (66.78, 75.50) |
70.43 (67.29, 73.57) |
67.97 (63.41, 72.53) |
0.30 | 0.23 | ||
| TNF plasma low | 70.76 (67.89, 73.62) |
70.74 (68.37, 75.87) |
69.55 (66.94, 72.16) |
.52 | 0.11 | 0.11 | |
| TNF plasma high | 71.60 (68.33, 74.87) |
68.93 (65.55, 72.31) |
71.09 (67.01, 75.17) |
0.06 | −0.20 | ||
| RANTES plasma low | 70.27 (67.60, 72.94) |
69.28 (66.66, 71.89) |
69.28 (66.66, 71.89) |
.78 | 0.09 | 0.00 | |
| RANTES plasma high | 71.74 (68.47, 75.01) |
71.92 (68.91, 71.89) |
72.71 (68.77, 76.65) |
0.003 | −0.07 | ||
| TOVA D Prime Signal Detection | vWF plasma low | 2.42 (2.10, 2.73) |
2.29 (1.92, 2.66) |
2.39 (2.10, 2.69) |
.73 | 0.03 | −0.09 |
| vWF plasma high | 2.36 (1.90, 2.83) |
2.55 (2.21, 2.88) |
2.46 (1.98, 2.94) |
−0.09 | 0.08 | ||
| TNF plasma low | 2.39 (2.08, 2.70) |
2.29 (2.03, 2.54) |
2.32 (2.04, 2.61) |
.45 | 0.06 | −0.03 | |
| TNF plasma high | 2.26 (1.91, 2.62) |
2.39 (2.01, 2.77) |
2.65 (2.20, 3.09) |
−0.33 | −0.22 | ||
| RANTES plasma low | 2.16 (1.86, 2.46) |
2.27 (2.00, 2.55) |
2.33 (2.04, 2.62) |
.42 | −0.15 | −0.04 | |
| RANTES plasma high | 2.49 (2.13, 2.85) |
2.24 (1.91, 2.57) |
2.26 (1.82, 2.69) |
0.20 | −0.01 | ||
| TOVA commission errors | vWF plasma low | 6.36 (2.38, 10.33) |
10.62 (5.90, 15.34) |
12.98 (9.25, 16.70) |
.20 | 0.41 | 0.15 |
| vWF plasma high | 11.90 (6.00, 17.80) |
7.23 (2.96, 11.51) |
14.27 (8.20, 20.35) |
0.15 | 0.44 | ||
| TNF plasma low | 13.17 (8.55, 17.78) |
10.39 (6.62, 14.16) |
13.31 (9.10, 15.72) |
.80 | 0.01 | 0.18 | |
| TNF plasma high | 8.65 (3.45, 13.85) |
8.32 (2.68, 13.95) |
12.08 (5.53, 18.64) |
0.21 | 0.23 | ||
| RANTES plasma low | 9.95 (5.43, 14.46) |
9.84 (5.70, 13.99) |
13.81 (9.41, 81.21) |
.76 | 0.24 | 0.25 | |
| RANTES plasma high | 13.42 (7.89, 18.94) |
10.00 (5.06, 14.94) |
14.14 (7.51, 20.77) |
0.04 | 0.26 | ||
| CBCL internalizing symptoms | vWF plasma low | 57.56 (53.30, 61.82) |
58.65 (54.54, 62.75) |
57.07 (53.91, 60.24) |
.67 | −0.05 | −0.16 |
| vWF plasma high | 56.33 (50.47, 62.09) |
56.14 (52.56, 59.72) |
58.26 (52.59, 63.93) |
0.19 | 0.21 | ||
| TNF plasma low | 56.08 (52.75, 59.40) |
58.44 (55.78, 61.11) |
55.39 (52.51, 58.27) |
.07 | −0.08 | −0.30 | |
| TNF plasma high | 56.68 (52.29, 60.87) |
53.91 (50.33, 57.49) |
58.77 (54.36, 63.18) |
0.21 | 0.48 | ||
| RANTES plasma low | 57.42 (53.88, 60.95) |
53.91 (51.27, 56.56) |
55.82 (52.98, 58.66) |
.05 | 0.16 | 0.19 | |
| RANTES plasma high | 56.79 (53.29, 60.25) |
60.96 (57.73, 64.20) |
56.90 (52.89, 60.90) |
0.01 | −0.40 | ||
| CBCL externalizing symptoms | vWF plasma low | 57.95 (53.39, 62.51) |
55.79 (51.47, 60.11) |
57.94 (54.58, 61.30) |
.47 | −0.01 | 0.24 |
| vWF plasma high | 54.74 (48.76, 60.71) |
57.83 (54.11, 61.55) |
60.65 (54.71, 66.59) |
0.65 | 0.31 | ||
| TNF plasma low | 56.21 (52.76, 59.67) |
57.64 (54.89, 60.39) |
57.47 (54.45, 60.48) |
.58 | 0.14 | 0.02 | |
| TNF plasma high | 53.18 (48.78, 57.57) |
55.19 (51.47, 58.91) |
58.13 (53.54, 62.72) |
0.55 | 0.32 | ||
| RANTES plasma low | 56.82 (53.00, 60.65) |
55.57 (52.73, 58.41) |
57.04 (53.91,60.16) |
.25 | 0.02 | 0.16 | |
| RANTES plasma high | 54.48 (50.69, 58.28) |
59.11 (55.56, 62.66) |
58.64 (54.29, 63.00) |
0.46 | −0.05 |
Negative sign before SMD absolute value indicates differences favoring the control group. SMDs exceeding 1/3 are bolded.
KABC: Kaufman Assessment Battery for Children, 2nd edition; TOVA: Tests of Variables of Attention; CBCL: Achenbach Child Behavior Checklist.
RANTES: Regulated on Activation, Normal T Expressed and Secreted; vWF: von Willebrand Factor; TNF: Tumor Necrosis Factor; CSF: cerebral spinal fluid; ##CCRT: computerized cognitive rehabilitation training.
CCRT: computerized cognitive rehabilitation training.
DISCUSSION
This exploratory hypothesis-generating study was the first to examine potential moderating effects of three immunopathogenic biomarkers collected during acute malaria illness as potential moderators of gains from a computerized cognitive rehabilitation training (CCRT) intervention. Some of the neurocognitive benefits of CCRT differed on the basis of level of vWF, TNF, and RANTES during acute illness. Severe malaria survivors with lower levels of vWF, lower CSF levels of TNF, and higher levels of plasma and CSF RANTES had better KABC cognitive performance after both titrating and non-titrating CCRT compared to no CCRT. Low CSF TNF or higher RANTES were associated with better TOVA outcomes from titrating or non-titrating CCRT. In contrast to the findings for the KABC and TOVA, for the behavioral outcomes measured with the CBCL, non-titrating CCRT showed greater benefit for externalizing symptoms among survivors with high vWF plasma, and no benefit for internalizing symptoms when vWF was high. High plasma RANTES was associated with no benefit from either the titrating and non-titrating CCRT for the behavioral outcomes, while high TNF plasma was predictive of the behavioral benefit for both interventions.
Higher levels of vWF and TNF and lower levels of RANTES during acute illness tend to be associated with more adverse clinical outcomes. This was consistent with findings by our group with previous cohorts of CM survivors at our study setting.12,13,22,34 Furthermore, other studies have implicated some of these same immunopathogenic biomarkers for both CM and SMA,40,41 including a randomized double-blinded placebo-controlled trial evaluating the neurological and neurocognitive protective benefits of inhaled nitric oxide (NO) versus placebo as an adjunctive therapy for severe malaria.42 In that trial, severe malaria was consistently associated with high rates of neurocognitive impairment for both the CM and SMA children, and treatment with inhaled NO was associated with reduced risk of fine motor impairment for both groups.
In the present study, high plasma levels of TNF and high CSF RANTES during acute illness were associated with greater overall cognitive benefits (as measured by the KABC MPI) for both titrating and non-titrating CCRT, suggesting potential moderating effects of these biomarkers for both interventions. Low level of vWF was associated with greater benefit for the non-titrating CCRT only. The differences in moderating effects for titrating versus non-titrating CCRT may be explained by the differences in two interventions. The non-titrating CCRT arm involved greater repetition and practice at less difficult levels of game training, resulting in improved processing speed gains,15 especially at low levels of vWF, an endothelial activation biomarker of CM during acute illness. We provide preliminary evidence to support the further investigation of vWF as a potential moderator of the non-titrating CCRT intervention. We are cautious in this suggestion, because vWF levels, while considered to be a sensitive marker for endothelial activation for severe malaria in general,11,43 may also simply correspond to inflammatory cascades. Therefore, it may not be causally predictive of a breakdown in the BBB and a corresponding underlying insult to the integrity of brain/behavior function foundational to dynamic learning in children.
As noted in the introductory background to the present study, three biomarkers are evaluated in this study because they are differentially related to each of the three proposed major mechanisms of brain injury from severe malaria (both CM and SMA).13,34,35,44 These mechanisms are, the hypoxic/ischemic brain injury effects of sequestration; the metabolic dysfunction for functioning neural networks compounded by sequestration and possibly seizure; and the immuno-pathogenic effects on brain function resulting from blood brain barrier compromise as mediated by the pro-inflammatory cascade of the immunity system in response to severe malaria infection. In a recent study, Harawa et al. (2018) noted that CM is characterized by elevated circulating pro-inflammatory cytokines TNF and IFN-gamma and anti-inflammatory cytokine IL-10.45 However, using MRI brain imaging during acute illness to document brain swelling in comatose Malawian CM children, Harawa and colleagues were able to demonstrated that whereas levels of IL-1beta, IL-6, IL-8 and IL-10 did not differ between CM patients with and without severe brain swelling, IL-12 and TNF levels were higher in children with severe swelling compared to those with moderate or no swelling. They concluded that severe brain swelling in pediatric CM was independent of serum-based pro-inflammatory and anti-inflammatory cytokines which are markers of systemic inflammation.45 However, CSF biomarkers of TNF could prove very sensitive to brain swelling in CM children during acute illness, and play an important role not only in the swelling, but also in the poorer neurocognitive benefit from CCRT intervention. In this study, CSF-based biomarkers (TNF and RANTES) were available just for CM survivors because of the lumbar puncture as a standard of care diagnostically during coma from acute illness. For both CSF TNF and RANTES, evidence of potential moderating effects was seen for the KABC MPI and CBCL, indicating that these CSF biomarkers may be reflective of compromised neuropsychological function in learning, as reflected by differential responses to the CCRT.
The limitations of the study include relatively small available sample sizes for the biomarkers and absence of a priori stated hypotheses, so that power considerations were not applicable and p-values had limited utility. Instead, this study was hypothesis-generating based on the different magnitude of SMDs for interventions effects at different levels of biomarkers. We used distributional approach to defining cut-offs since no normative values are currently available. Further, it was not possible to use the cut-offs obtained for the community control children among children with malaria since the distributions of biomarkers did not overlap. The potential moderating effects of vWF, TNF, and RANTES identified in this study will need to be formally tested in future trials of CCRT or other rehabilitative interventions outcomes which may reflect the integrity of brain/behavior function in dynamic learning processes in children.
Supplementary Material
Acknowledgements:
Biomarker data was provided by Chandy C. John and Dibyadyuti Datta of the Department of Microbiology and Immunology and the Ryan White Center for Pediatric Infectious Disease and Global Health, University of Indiana, Indianapolis, Indiana, USA. Their work was supported by the National Institutes of Health (NIH) grant R01 NS055349 (MPI: John, Opoka). Professor John also provided extensive comments on earlier drafts of this manuscript, and his efforts are greatly appreciated. Cognitive training, assessment, and on-site study management was conducted by members of the Global Health Uganda study team for this project at Mulago Hospital, Kampala, Uganda. Team members were Lyagoba Monica (study coordinator), Alex Mutebe (data management), Robert Tuke (medical officer), Susan Nalubwama (study nurse), Agatha Kuteesa, Michael Sengendo, Ethel Wandeka Nuwamanya, Maria Kateete, Richard Seviiri, Stella Akayo, Titus Sessanga, and Stella Butala. Their efforts are greatly appreciated. Drs. Esther van der Lugt, Kimberley Walhof, Christopher Adam, Hailey Wouters, and Jennifer Neva made significant contributions to this study as medical student research interns; as did graduate students Jacquelyn Moore, Katherine Finn.
Funding: This work was supported by the National Institute of Neurological Disease and Stroke (NINDS) [grant number R01 NS05534] and Kennedy Shriver National Institute of Child Health and Development [grant number R01 HD064416]. The funders had no role in the design of the study and collection, analysis, and interpretation of data and in writing the manuscript, and in the decision to publish this manuscript.
Footnotes
Disclosures: The authors have no conflicts of interest or funding to disclose.
Clinical Trials Registration: ClinicalTrials.gov Identifier NCT01743417, Submitted September 5, 2012
REFERENCES
- 1.Bangirana P, Allebeck P, Boivin MJ, et al. Cognition, behaviour and academic skills after cognitive rehabilitation in Ugandan children surviving severe malaria: a randomised trial. BMC Neurol. 2011;11:96. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Bangirana P, Boivin MJ, Giordani B. Computerized Cognitive Rehabilitation Therapy (CCRT) for African Children: Evidence for Neuropsychological Benefit and Future Directions In: Boivin MJ, Giordani B, eds. Neuropsychology of Children in Africa: Perspectives on Risk and Resilience. Vol 1 New York: Springer Science+Business Media; 2013:277–298. [Google Scholar]
- 3.Bangirana P, Giordani B, John CC, Page C, Opoka RO, Boivin MJ. Immediate neuropsychological and behavioral benefits of computerized cognitive rehabilitation in Ugandan pediatric cerebral malaria survivors. J Dev Behav Pediatr. 2009;30(4):310–318. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Conroy AL, Datta D, John CC. What causes severe malaria and its complications in children? Lessons learned over the past 15 years. BMC Med. 2019;17(1):52. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Dunst J, Kamena F, Matuschewski K. Cytokines and Chemokines in Cerebral Malaria Pathogenesis. Front Cell Infect Microbiol. 2017;7:324. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Taylor T, Fu W, Carr R, et al. Differentiating the pathologies of cerebral malaria by postmortem parasite counts. Nat Med. 2004;10(2):143–145. [DOI] [PubMed] [Google Scholar]
- 7.MacPherson GG, Warrell MJ, White NJ, Looareesuwan S, Warrell DA. Human cerebral malaria. A quantitative ultrastructural analysis of parasitized erythrocyte sequestration. Am J Pathol. 1985;119(3):385–401. [PMC free article] [PubMed] [Google Scholar]
- 8.Seydel KB, Milner DA Jr., Kamiza SB, Molyneux ME, Taylor TE. The distribution and intensity of parasite sequestration in comatose Malawian children. J Infect Dis. 2006;194(2):208–205. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.de Wit TR, van Mourik JA. Biosynthesis, processing and secretion of von Willebrand factor: biological implications. Best Pract Res Clin Haematol. 2001;14(2):241–255. [DOI] [PubMed] [Google Scholar]
- 10.van Mourik JA, Romani de Wit T. Von Willebrand factor propeptide in vascular disorders. Thromb Haemost. 2001;86(1):164–171. [PubMed] [Google Scholar]
- 11.Hollestelle MJ, Donkor C, Mantey EA, et al. von Willebrand factor propeptide in malaria: evidence of acute endothelial cell activation. Br J Haematol. 2006;133(5):562–569. [DOI] [PubMed] [Google Scholar]
- 12.Phiri HT, Bridges DJ, Glover SJ, et al. Elevated plasma von Willebrand factor and propeptide levels in Malawian children with malaria. PLoS ONE. 2011;6(11):e25626. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.John CC, Opika-Opoka R, Byarugaba J, Idro R, Boivin MJ. Low levels of RANTES are associated with mortality in children with cerebral malaria. J Infect Dis. 2006;194(6):837–845. [DOI] [PubMed] [Google Scholar]
- 14.Beare NA, Harding SP, Taylor TE, Lewallen S, Molyneux ME. Perfusion abnormalities in children with cerebral malaria and malarial retinopathy. J Infect Dis. 2009;199(2):263–271. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Boivin MJ, Nakasujja N, Sikorskii A, et al. Neuropsychological benefits of computerized cognitive rehabilitation training in Ugandan children surviving severe malaria: A randomized controlled trial. Brain Res Bull. 2019;145:117–128. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Bangirana P, Opoka RO, Boivin MJ, Idro R, Hodges JS, John CC. Neurocognitive domains affected by cerebral malaria and severe malarial anemia in children. Learning and Individual Differences. 2015; 10.1016/j.lindif.2015.01.010. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Bangirana P, Opoka RO, Boivin MJ, et al. Severe malarial anemia is associated with longterm neurocognitive impairment. Clinical Infectious Diseases. 2014;59(3):336–344. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Bangirana P, Menk J, John CC, Boivin MJ, Hodges JS. The association between cognition and academic performance in Ugandan children surviving malaria with neurological involvement. PLoS ONE. 2013;8(2):e55653. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Kaufman AS, Kaufman NL. Manual for the Kaufman Assessment Battery for Children, Second Edition Circle Pines, MN: American Guidance Service Publishing/Pearson Products Inc.; 2004. [Google Scholar]
- 20.Bangirana P, Seggane M, Allebeck P, et al. A preliminary examination of the construct validity of the KABC-II in Ugandan children with a history of cerebral malaria. Afr Health Sci. 2009;9(3):186–192. [PMC free article] [PubMed] [Google Scholar]
- 21.Boivin MJ, Bangirana P, Byarugaba J, et al. Cognitive impairment after cerebral malaria in children: a prospective study. Pediatrics. 2007;119(2):e360–366. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.John CC, Bangirana P, Byarugaba J, et al. Cerebral malaria in children is associated with long-term cognitive impairment. Pediatrics. 2008;122(1):e92–99. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.The T.O.V.A. (Version 6.X) (Computer Program) [computer program]. Los Alamitos, CA: 1993. [Google Scholar]
- 24.Bergemann TL, Bangirana P, Boivin MJ, Connett JE, Giordani BJ, John CC. Statistical approaches to assess the effects of disease on neurocognitive function over time J Biomet Biostat. 2012;S7:016. doi: 10.4172/2155-6180.S7-016 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Boivin MJ, Bangirana P, Byarugaba J, et al. Cognitive impairment after cerebral malaria in children: a prospective study. Pediatrics. 2007;119(2):e360–366. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Achenbach TM. Integrative Guide ot the 1991 CBCL/4–18 YSR, and TRF Profiles. University of Vermont, Department of Psychology Pediatrics. 1991;108(1):e14. [Google Scholar]
- 27.Achenbach TM. The Multicultural Supplement to the Manual for the ASEBA Preschool Forms and Profiles Burlington, VT: ASEBA; 2010. [Google Scholar]
- 28.Bangirana P, Nakasujja N, Giordani B, Opoka RO, John CC, Boivin MJ. Reliability of the Luganda version of the Child Behaviour Checklist in measuring behavioural problems after cerebral malaria. Child Adolesc Psychiatry Ment Health. 2009;3:38. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Familiar I, Ruisenor-Escudero H, Giordani B, et al. Use of the Behavior Rating Inventory of Executive Function and Child Behavior Checklist in Ugandan children with HIV or a history of severe malaria. J Dev Behav Pediatr. 2015;36(4):277–284. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Achenbach TM, Rescorla LA. Multicultural Supplement to the Manual for the ASEBA School-Age Forms & Profiles. Burlington, VT: University of Vermont, Research Center for Children, Youth, & Families; 2007. [Google Scholar]
- 31.Bangirana P, Musisi S, Allebeck P, et al. A preliminary investigation of the construct validity of the KABC-II in Ugandan children with prior cerebral insult. African Health Sciences. 2009;9(3):186–192. [PMC free article] [PubMed] [Google Scholar]
- 32.Bangirana P, John CC, Idro R, et al. Socioeconomic predictors of cognition in Ugandan children: Implications for community based interventions. PLoS ONE. 2009;4(11):e7898 doi:7810.1371/journal.pone.0007898. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Boivin MJ, Giordani B, Ndanga K, Maky MM, Manzeki KM, Ngunu N. Economic advantage and the cognitive ability of rural children in Zaire. J Psychol. 1996;130(1):95–107. [DOI] [PubMed] [Google Scholar]
- 34.Park GS, Min M, Opika-Opoka R, Boivin MJ, John CC. von Willebrand Factor, but not sVCAM-1 or sICAM-1, effectively discriminates between cerebral and uncomplicated malaria in Ugandan children. Paper presented at: Keystone Conference on Infectious Disease2008; Vienna, Austria. [Google Scholar]
- 35.Shabani E, Ouma BJ, Idro R, et al. Elevated cerebrospinal fluid tumour necrosis factor is associated with acute and long-term neurocognitive impairment in cerebral malaria. Parasite Immunol. 2017;39(7): doi: 10.1111/pim.12438. Epub 12017 May 12428. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Wasserstein RL, Lazar NA. The ASA’s statement on p-values: context, process, and purpose.. The American Statistician. 2016;70(2):129–133. [Google Scholar]
- 37.Cella D, Eton DT, Fairclough DL, et al. What is a clinically meaningful change on the Functional Assessment of Cancer Therapy-Lung (FACT-L) Questionnaire? Results from Eastern Cooperative Oncology Group (ECOG) Study 5592. J Clin Epidemiol. 2002;55(3):285–295. [DOI] [PubMed] [Google Scholar]
- 38.Cohen J Statistical Power Analysis for the Behavioral Sciences (2nd ed.). Hillsdale, NJ: Erlbaum; 1988. [Google Scholar]
- 39.Norman GR, Sloan JA, Wyrwich KW. Interpretation of changes in health-related quality of life: the remarkable universality of half a standard deviation. Medical Care. 2003;41(5):582–592. [DOI] [PubMed] [Google Scholar]
- 40.Conroy AL, Hawkes M, McDonald CR, et al. Host Biomarkers Are Associated With Response to Therapy and Long-Term Mortality in Pediatric Severe Malaria. Open Forum Infect Dis. 2016;3(3):ofw134. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Graham SM, Chen J, Chung DW, et al. Endothelial activation, haemostasis and thrombosis biomarkers in Ugandan children with severe malaria participating in a clinical trial. Malar J. 2016;15:56. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Bangirana P, Conroy AL, Opoka RO, et al. Inhaled nitric oxide and cognition in pediatric severe malaria: A randomized double-blind placebo controlled trial. PLoS One. 2018;13(1):e0191550. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.Larkin D, de Laat B, Jenkins PV, et al. Severe Plasmodium falciparum malaria is associated with circulating ultra-large von Willebrand multimers and ADAMTS13 inhibition. PLoS Pathog. 2009;5(3):e1000349. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.John CC, Panoskaltsis-Mortari A, Opoka RO, et al. Cerebrospinal fluid cytokine levels and cognitive impairment in cerebral malaria. Am J Trop Med Hyg. 2008;78(2):198–205. [PMC free article] [PubMed] [Google Scholar]
- 45.Harawa V, Njie M, Kessler A, et al. Brain swelling is independent of peripheral plasma cytokine levels in Malawian children with cerebral malaria. Malar J. 2018;17(1):435. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.