Gut Bacteria and Blood Proteins Predict Who Gains Fitness From HIIT in Prediabetes
Multi-omics analysis reveals gut microbiome and proteomic signatures that explain why some men with prediabetes gain far more VO2peak from HIIT than others.
Summary
A 12-week high-intensity interval training (HIIT) study in overweight men with prediabetes found dramatic variation in cardiorespiratory fitness gains — some improved VO2peak by nearly four times the average. Using proteomics, gut metagenomics, and metabolomics, researchers identified biological predictors of this variation. Men with higher baseline levels of short-chain fatty acid-producing gut bacteria — notably Prevotella — showed greater fitness gains. Conversely, higher baseline erythropoietin levels predicted blunted responses. Exercise-induced increases in growth hormone were linked to better outcomes. Lean body mass was the strongest clinical predictor overall. These findings suggest that pre-exercise gut microbiome and blood protein profiles could help personalize HIIT prescriptions to maximize metabolic and cardiovascular benefits in people at risk for type 2 diabetes.
Detailed Summary
Cardiorespiratory fitness, measured as VO2peak, is a powerful predictor of cardiovascular and metabolic health. People with prediabetes tend to have reduced fitness compared to healthy peers, and improving VO2peak through exercise can meaningfully lower disease risk. High-intensity interval training is particularly effective, but responses vary enormously between individuals — making it difficult to predict who will benefit most.
This study enrolled 35 medication-naïve, overweight and obese Chinese men aged 24–62 years with prediabetes. All completed a structured 12-week HIIT program. Researchers collected serum, stool, and metabolite samples before and after the intervention, applying proteomic, fecal metagenomic, and targeted metabolomic profiling to identify biological correlates of VO2peak improvement.
On average, VO2peak increased by 0.47 L/min, but individual gains ranged from 0 to 1.7 L/min — a striking three-fold span. Baseline abundance of short-chain fatty acid-producing gut bacteria, especially Prevotella, Coprococcus, and Hungatella, positively predicted fitness gains. Higher baseline erythropoietin (EPO) levels predicted smaller improvements. Post-exercise increases in growth hormone 1 were positively associated with VO2peak gains, while increases in BACH1 — a repressor of the cytoprotective enzyme heme oxygenase-1 — were negatively associated. Lean mass was the strongest individual clinical predictor, and adding circulating molecular factors to the clinical model improved explanatory power from 27% to 37% of variance.
These findings suggest the gut microbiome may modulate exercise adaptation, potentially through SCFA production influencing energy metabolism and mitochondrial function. Clinically, assessing a patient's baseline gut microbiome or EPO levels before prescribing HIIT could help set realistic expectations or guide adjunct interventions.
Key caveats: the sample was limited to overweight Chinese men, restricting generalizability to women and other ethnic groups. FDR values for several proteomic findings were high, indicating exploratory rather than confirmatory results. Summary is based on the abstract only.
Key Findings
- Higher baseline Prevotella gut bacteria strongly predicted greater VO2peak gains after 12 weeks of HIIT.
- Elevated baseline erythropoietin was associated with blunted cardiorespiratory fitness response to HIIT.
- Exercise-induced increases in growth hormone 1 correlated with larger VO2peak improvements.
- Lean body mass was the single strongest clinical predictor of fitness adaptation in prediabetes.
- Adding multi-omics biomarkers to clinical variables improved variance explained from 27% to 37%.
Methodology
Observational longitudinal study with 35 overweight/obese Chinese men with prediabetes completing a 12-week HIIT program. Pre- and post-intervention serum proteomics, fecal metagenomics, and targeted metabolomics were performed. Statistical analyses included linear regression models with FDR correction.
Study Limitations
The study included only overweight Chinese men, limiting generalizability to women and other populations. Several proteomic associations had high FDR values, indicating these findings are exploratory and require replication in larger cohorts. Summary is based on the abstract only, so methodological details beyond what is reported cannot be assessed.
Enjoyed this summary?
Get the latest longevity research delivered to your inbox every week.
Enter your email to subscribe: