Moderate exercise has limited but distinguishable effects on the mouse microbiome

The gut microbiome is known to have a complex yet vital relationship with host health. While both exercise and the gut microbiome have been shown to impact human health independently, the direct effects of moderate exercise on the intestinal microbiota remain unclear. In this study, we compared gut microbial diversity and changes in inflammatory markers associated with exercise over an 8-week period in mice that performed either voluntary exercise (VE) (n 10) or moderate forced exercise (FE) (n 11) and mice that did not perform any exercise (n 21). VE mice, but not FE mice, had increased food intake and lean body mass compared to sedentary mice. The levels of inflammatory markers associated with exercise were similar for mice in all three groups. Traditional microbial profiles comparing operational taxonomic units (OTUs) in samples (P > 0.1) and multivariate analysis of beta diversity via Adonis testing (P > 0.1) did not identify significantly altered taxonomic profiles in the voluntary or forced exercise group compared to the sedentary controls. However, a random forests machine learning model, which takes into account the relationships between bacteria in a community, classified voluntary exercisers and nonexercisers with 97% accuracy at 8 weeks. The top bacteria used by the model allowed us to identify known taxa (Bacteroides, S24-7, and Lactobacillus) and novel taxa (Rikenellaceae and Lachnospiraceae) associated with exercise. Although aerobic exercise in mice did not result in significant changes of abundance in gut microbes or in host inflammatory response, more sophisticated computational methods could identify some microbial shifts. More study is needed on the effects of various exercise intensities and their impact on the gut microbiome.