Exercise Training Transiently Increases Gut Microbiota Diversity and Short-Chain Fatty Acid Production in a Diet-Dependent Manner in Healthy Adults

Abstract

Exercise alters gut microbiome composition, but the temporal dynamics and diet-dependent metabolic interactions remain unclear. We investigated longitudinal changes in gut microbiota, functional pathways, and metabolite profiles during and after an exercise intervention. Twenty-four healthy adults completed a sequential three-phase protocol: an 8-week self-directed exercise intervention, an 8-week washout, and an 8-week no-exercise control period. Fecal samples were collected at T1 (baseline), T2 (post-exercise), T3 (post-washout), and T4 (post-control). Microbiota composition was assessed by 16S rRNA sequencing, functional pathways predicted using PICRUSt2, metabolites predicted using COBRA Toolbox, and fecal SCFAs and bile acids quantified by GC and HPLC. Temporal causal relationships were examined using Tigramite analysis with dietary pattern stratification, and microbiota-environment associations were assessed by redundancy analysis (RDA). Alpha diversity was significantly higher at T2 than at T4 (p < 0.05). Beta diversity differed significantly between T2 and both T3 and T4, with no difference between T1 and T4, indicating reversibility. Ruminococcus gnavus was significantly higher at T2 than at T4 (p < 0.001), with several additional taxa higher at T2 at a less stringent threshold. Propionate and butyrate were elevated at T2, while total bile acids were lower. Bacteroides thetaiotaomicron was positively associated with body weight in the total cohort and specifically under a balanced dietary pattern (BD), with no significant time-lagged associations detected under a Western-style diet (WSD). In RDA, taxa associated with body weight substantially overlapped with taxa found to increase during exercise, whereas physical performance measures showed no direct temporal association with microbiota composition in causal analysis. In conclusion, exercise-induced changes in the gut microbiome were not sustained after structured exercise ended, suggesting that continuous exercise may be required. Diet further shaped whether microbiota–host associations were detectable, underscoring dietary pattern as a factor for future microbiome-targeted exercise interventions.