Backgrounds:

The effect of SARS-CoV-2 on altering the gut microbiome have been investigated since the beginning of the pandemic. To our knowledge, this is the first observational cohort study on the correlations between gut microbiota (GM) and COVID-19 infection in children.

Methods

The GM profile was investigated by 16S rRNA targeted-metagenomics to analyze ecology and inferred functions of COVID-19 patients compared to healthy subjects (CTRLs). Multiple machine learning (ML) models were exploited. Correlation between ASVs abundances and inflammatory protein levels was investigated using R software.

Results:

The GM in COVID-19 patients (N=67) was characterized by reduction of alfa-diversity compared to CTRLs (p <0.05). In the GM of COVID-19 children an increase of Faecalibacterium, Fusobacterium, Neisseria and decrease of Bifidobacterium, Blautia, Ruminococcus, Collinsella, Coprococcus, Eggerthella, Akkermansia were reported, compared to CTRLs (FDR<0.05). ML models for GM biomarker prediction in COVID-19 patients, compared to CTRLs, identified: Actinomyces, Alistipes, Faecalibacterium, Anaerostipes, Phascolarctobacterium, Dorea, Clostridium, Prevotella, Oscillospira, Staphylococcus, Rominococcus.

The KO-based prediction of GM functional profile of COVID-19, compared to CTRLs, highlighted lipopolysaccharide and peptidoglycan biosynthesis, amino acids biosynthesis, alanine, aspartate and glutamate metabolism specifically associated to COVID-19 patients; while PPAR signaling pathway; ether lipid metabolism; fatty acid degradation; valine, leucine and isoleucine degradation to CTRLs subjects.

A statistically significant positive correlation between Bacteroidetes and pro-inflammatory cytokines (e.g., ITGA11, CLECA4 and DNER) was observed.

Conclusions/Learning Points:

Our study provides a specific characterization of the GM of pediatric COVID-19. Unlike adult, high levels of Faecalibacterium were reported as a specific treat of the COVID-19 in children. GM profile and intrinsic GM anti-inflammatory and fermentative properties may play a central role in the low severity of COVID-19 in children.