Gut microbiome patterns depending on children's psychosocial stress: Reports versus biomarkers

Aim: Chronic stress increases disease vulnerability factors including inflammation, a pathological characteristic potentially regulated by the gut microbiota. We checked the association between the gut microbiome and psychosocial stress in children/adolescents and investigated which stress parameter (negative versus positive emotion, self-report versus parental report, events versus emotions, biomarker cortisol versus parasympathetic activity) is the most relevant indicator herein. Methods: Gut microbiome sequencing was completed in fecal samples from 93 Belgian 8-16y olds. Stress measures included negative events, negative emotions, emotional problems reported by parents, happiness, hair cortisol and heart rate variability (pnn50 parameter reflecting parasympathetic activity). Alpha diversity, beta diversity and linear discriminant analysis were the unadjusted analyses. Age, sex, socio-economic status, diet, physical activity, sleep and weight status were adjusted for via a redundancy analysis and differential abundance via zero-inflated negative binomial regression. Results: High stress as reflected by low pnn50 and more negative events were associated with a lower alpha diversity as indicated by the Simpson index. Happiness and pnn50 showed significant differences between high and low stress groups based on weighted UniFrac distance, and this remained significant after confounder adjustment. Adjusted and unadjusted taxonomic differences were also most pronounced for happiness and pnn50 being associated respectively with 24 OTU (=11.8% of bacterial counts) and 31 OTU (=13.0%). As a general pattern, high stress was associated with lower Firmicutes at the phylum level and higher Bacteroides, Parabacteroides, Rhodococcus, Methanobrevibacter and Roseburia but lower Phascolarctobacterium at genus level. Several genera gave conflicting results between different stress measures e.g. Ruminococcaceae UCG014, Tenericutes, Eubacterium coprostanoligenes, Prevotella 9 and Christensenellaceae R7. Differential results in preadolescents versus adolescents were also evident. Conclusion: Even in this young healthy population, stress parameters were cross-sectionally associated with gut microbial composition but this relationship was instrument specific. Positive emotions and parasympathetic activity appeared the strongest parameters and should be integrated in future microbiota projects amongst other stress measures.