Relationships of the gut microbiome with cognitive development among healthy school-age children

Abstract

Background: The gut microbiome might play a role in neurodevelopment, however, evidence remains elusive. We aimed to examine the relationship between the intestinal microbiome and cognitive development of school-age children.

Methods: This cross-sectional study included healthy Israeli Arab children from different socioeconomic status (SES). The microbiome was characterized in fecal samples by implementing 16S rRNA gene sequencing. Cognitive function was measured using Stanford-Binet test, yielding full-scale Intelligence Quotient (FSIQ) score. Sociodemographics and anthropometric and hemoglobin measurements were obtained. Multivariate models were implemented to assess adjusted associations between the gut microbiome and FSIQ score, while controlling for age, sex, SES, physical growth, and hemoglobin levels.

Results: Overall, 165 children (41.2% females) aged 6-9 years were enrolled. SES score was strongly related to both FSIQ score and the gut microbiome. Measures of α-diversity were significantly associated with FSIQ score, demonstrating a more diverse, even, and rich microbiome with increased FSIQ score. Significant differences in fecal bacterial composition were found; FSIQ score explained the highest variance in bacterial β-diversity, followed by SES score. Several taxonomic differences were significantly associated with FSIQ score, including Prevotella, Dialister, Sutterella, Ruminococcus callidus, and Bacteroides uniformis.

Conclusions: We demonstrated significant independent associations between the gut microbiome and cognitive development in school-age children.

Keywords: children; cognitive development; gut microbiome; healthy; school age; socioeconomic status.

Figure 1

The association between full-scale IQ scores and bacterial α-diversity. (A) Box-violin plots of microbial diversity, measured by Shannon's diversity index, across tertiles of FSIQ scores, showing a significant increase in microbial α-diversity with increased FSIQ (p = 0.014). (B) Results of a multivariate analysis of variance displaying the association between FSIQ score and covariates of interest with bacterial diversity. FSIQ score and sex were significantly associated Shannon's bacterial α-diversity index (F = 6.16, p = 0.014 and F = 4.89 p = 0.029, respectively). (C) Box-violin plots of FSIQ scores across tertiles of Shannon's diversity index, show a significant increase in FSIQ scores with increased bacterial diversity (p = 0.002). (D) Results of a multivariate analysis of variance displaying the association between the individuals’ gut Shannon's diversity and covariates of interest with FSIQ scores. Bacterial α-diversity and SES scores were strongly associated with FSIQ score (F = 9.73, p = 0.014 and F = 97.91, p = 0.029, respectively), while hemoglobin levels had a more delicate albeit significant association (F = 3.94, p = 0.049). (E) The correlation between Shannon's α-diversity index and FSIQ score; Pearson's r = 0.20, p = 0.015. SES, socioeconomic status; FSIQ, full-scale IQ; HAZ, height for age z-score at age 18–30 months; BMIZ, body mass index z-score at age 6–9 years. *The x axis in figures (A,C) represents tertiles, T1 = lowest tertile, and T3 = highest tertile. **The mid line in the box plots [figures (A,C)] represents the median, the lower bound of the box represents the 25th percentile, the upper bound of the box represents the 75th percentile, the lowest point of the lower whisker represents the minimum and the highest point of the upper whisker represents the maximum. The violin plot implements a rotated kernel density plot on each side, adding information regarding the full distribution of the measured data; the width of the violin indicates the frequency.

Figure 2

The association between full-scale IQ scores and bacterial β-diversity. (A) Principal coordinate analysis (PCoA) of the Bray-Curtis dissimilarity index, notably altered with changing FSIQ scores (F = 10.79, p = 0.001). (B) PCoA of the phylogenetic unweighted uniFrac distance matrix, significantly separated with altered FSIQ score (F = 5.04, p = 0.001). (C) PCoA of the Jensen-Shannon divergence (JSD), clearly separated according to FSIQ tertiles (F = 15.90, p = 0.001). (D) Stacked (100%) bar-plots of the explained variance in microbial beta diversity by the multivariate models. The FSIQ score explained most of the variance in all β-diversity measurements, followed by SES score. The JSD method explained the highest percentage of variance in microbial β-diversity (16.6%). FSIQ, full-scale IQ; PCoA, principal coordinate analysis; JSD, Jensen-Shannon divergence. *FSIQ scores in Figures (C,D) are represented as tertiles, T1 being the lowest tertile (FSIQ scores between 59 and 96), T2 the middle tertile (FSIQ scores between 97 and105), and T3 the highest tertile (FSIQ scores between 106 and 127). **The midline in the box plots [figure (D)] represents the median, the lower bound of the box represents the 25th percentile, the upper bound of the box represents the 75th percentile, the lowest point of the lower whisker represents the minimum and the highest point of the upper whisker represents the maximum. The violin plot implements a rotated kernel density plot on each side, adding information regarding the full distribution of the measured data; the width of the violin indicates the frequency.

Figure 3

Differentially abundant taxa associated with full-scale IQ scores. (A) Volcano plot showing differentially abundant s-OTUs associated with FSIQ scores in the whole cohort, as detected by ANCOM. The x-axis represents the difference in mean centered log ratio (clr)-transformed abundance between groups, and the y-axis represents the ANCOM W Statistic. s-OTU points are colored by the level of ANCOM significance, with 0.9 being the highest level; s-OTUs in gray were not significant. (B–I) Boxplots of clr-transformed abundance of s-OTUs significantly associated with FSIQ scores, adjusted for sex, age, SES score, hemoglobin level, HAZ and BMIZ scores. Tertiles of FSIQ were categorized as low, middle and high FSIQ score tertiles, T1 being the lowest tertile (FSIQ scores between 59 and 96), T2 the middle tertile (FSIQ scores between 97 and105), and T3 the highest tertile (FSIQ scores between 106 and 127). The midline in the box plots represents the median, the lower bound of the box represents the 25th percentile, the upper bound of the box represents the 75th percentile, the lowest point of the lower whisker represents the minimum and the highest point of the upper whisker represents the maximum. FSIQ, full-scale IQ; s-OTUs, sub-operational taxonomic units; ANCOM, analysis of composition of microbiomes; SES, socioeconomic status; HAZ, height for age z-score at infancy (18–30 months); BMIZ, body mass index z-score at age 6–9 years.

Figure 4

Differentially abundant taxa by village of residence and socioeconomic status. (A,B) Volcano plots showing differentially abundant s-OTUs as detected by ANCOM, stratified by village; A/B [high/intermediate SES] (A), and C [low SES] (B). The x-axis represents the difference in mean centered log ratio (clr)-transformed abundance between groups, and the y-axis represents the ANCOM W Statistic. s-OTU points are colored by level of ANCOM significance, with 0.9 being the highest level; s-OTUs in gray were not significant. (B,C) Boxplots of clr-transformed abundance of s-OTUs significantly associated with FSIQ scores in villages A/B [high/intermediate SES] (C) and in village C [low SES] (D), adjusted for sex, age, SES score, hemoglobin level (g/dl), HAZ and BMIZ scores. Tertiles of FSIQ were categorized as low, middle and high FSIQ score tertiles, T1 being the lowest tertile (FSIQ scores between 59 and 96), T2 the middle tertile (FSIQ score between 97 and 105), and T3 the highest tertile (FSIQ scores between 106 and 127). The mid line in the box plots represents the median, the lower bound of the box represents the 25th percentile, the upper bound of the box represents the 75th percentile, the lowest point of the lower whisker represents the minimum and the highest point of the upper whisker represents the maximum. FSIQ, full-scale IQ; s-OTUs, sub-operational taxonomic units; ANCOM, analysis of composition of microbiomes; SES- socioeconomic status; HAZ, height for age z-score at infancy (18–30 months); BMIZ, body mass index z-score at age 6–9 years.