Subclinical Inflammation in Asymptomatic Schoolchildren With Plasmodium falciparum Parasitemia Correlates With Impaired Cognition

Abstract

Background: Subclinical inflammation and cognitive deficits have been separately associated with asymptomatic Plasmodium falciparum infections in schoolchildren. However, whether parasite-induced inflammation is associated with worse cognition has not been addressed. We conducted a cross-sectional pilot study to better assess the effect of asymptomatic P. falciparum parasitemia and inflammation on cognition in Kenyan schoolchildren.

Methods: We enrolled 240 children aged 7-14 years residing in high malaria transmission in Western Kenya. Children performed five fluid cognition tests from a culturally adapted NIH toolbox and provided blood samples for blood smears and laboratory testing. Parasite densities and plasma concentrations of 14 cytokines were determined by quantitative PCR and multiplex immunoassay, respectively. Linear regression models were used to determine the effects of parasitemia and plasma cytokine concentrations on each of the cognitive scores as well as a composite cognitive score while controlling for age, gender, maternal education, and an interaction between age and P. falciparum infection status.

Results: Plasma concentrations of TNF, IL-6, IL-8, and IL-10 negatively correlated with the composite score and at least one of the individual cognitive tests. Parasite density in parasitemic children negatively correlated with the composite score and measures of cognitive flexibility and attention. In the adjusted model, parasite density and TNF, but not P. falciparum infection status, independently predicted lower cognitive composite scores. By mediation analysis, TNF significantly mediated ~29% of the negative effect of parasitemia on cognition.

Conclusions: Among schoolchildren with PCR-confirmed asymptomatic P. falciparum infections, the negative effect of parasitemia on cognition could be mediated, in part, by subclinical inflammation. Additional studies are needed to validate our findings in settings of lower malaria transmission and address potential confounders that could affect both inflammation and cognitive performance.

Keywords: Plasmodium falciparum; asymptomatic infection; cognition; inflammation; malaria.

Figure 1.

Study design. Participants, procedures, samples, and data collected for analysis.

Figure 2.

Correlations between cognitive scores, parasitemia, and plasma cytokine concentrations. (A) Spearman correlation matrix for individual cognitive scores, cognitive composite score, plasma cytokine concentrations, parasite density, and age. Only correlations with a false discovery rate < 10% are shown. Numbers within tiles are Spearman ρ. (B) A network plot of correlation data in (A) where more highly correlated variables appear closer together and are joined by more intense edges. Only correlations with |ρ| > 0.20 are shown. Log₁₀(qPCR-estimated parasites/μl + 1) was used for parasite density. Correlations with parasite density were done for all 240 children (no asterisk) or only among the 156 children who had PCR-confirmed P. falciparum parasitemia (asterisk). Abbreviations: DCCS, dimensional change card sort test; PSM, picture sequence; PC, pattern comparison; LS, list sorting.

Figure 3.

Causal mediation analysis between parasite density and plasma TNF. Causal mediation analysis was performed to evaluate whether the effect of parasitemia, as measured by log₁₀(parasite/μl + 1), on compositive cognitive score was mediated through circulating TNF in plasma with adjustments for age, gender, and maternal education while accounting for an interaction between age and presence of any parasitemia by PCR. (A) Shown are the pathways, β coefficients, and p values for the indicated covariate (start of arrow) on the response variable (end of the arrow) for each linear regression model. Parasite density positively associates with TNF, which, in turn, negatively associates with cognitive score. Reduction in both magnitude of β and significance for parasite density with inclusion of TNF in the model relative to without TNF suggests that the effect of parasitemia on cognitive score was mediated through TNF. (B) The results of the mediation analysis with nonparametric bootstrap confidence intervals (500 simulations).