Beyond the Stool: Circulating Short-Chain Fatty Acids Better Predict Human Gut-Related Diseases
Simone Baldi, Alessio Rotelli, Francesco Cei, Sara Bertorello, Azzurra Bongianni, Marta Menicatti, Antonino Salvatore Calabrò, Fabio Cianchi, Barbara Colombini, Francesco Sofi, Elena Niccolai, Gianluca Bartolucci, Ernesto Iadanza, Amedeo AmedeiAbstract
Background
Short-chain fatty acids (SCFAs) mediate host–microbiota interactions and are typically measured in feces as markers of gut microbial activity. However, whether fecal SCFAs adequately reflect disease-associated host–microbiota dysfunction remains unclear. We hypothesized that circulating SCFAs, which integrate microbial production, intestinal absorption, barrier function, and host metabolism, provide a more informative representation of gut-related diseases.
Methods
Seven fecal and serum SCFAs were quantified by gas chromatography–mass spectrometry in 193 individuals, including patients with colorectal cancer (CRC, n = 30), celiac disease (CD, n = 35), obesity (OB, n = 33), and healthy controls (HCs, n = 95). We developed an interpretable machine-learning framework combining stacking ensembles, bootstrap validation, permutation testing, SHAP and LIME analyses. Cross-compartment predictive modeling was used to assess the reciprocal informativeness of fecal and circulating SCFAs.
Results
Circulating SCFAs exhibited markedly stronger disease-associated signatures than fecal SCFAs across all groups. Serum-based models achieved excellent classification performance (95% accuracy; multiclass ROC-AUC 1.00), whereas fecal SCFAs provided substantially lower predictive power (59% accuracy). Combining serum and fecal datasets yielded only marginal improvement (97.4% accuracy), indicating most diagnostic relevant information was already captured in circulation. Valeric, propionic, 2-methylbutyric, and isovaleric acids were consistently identified as the main contributors to disease discrimination. Cross-compartment modeling revealed weak predictive relationships between fecal and serum SCFAs, indicating these pools are largely non-redundant and reflect distinct biological processes.
Conclusions
Circulating SCFAs outperform fecal SCFAs as biomarkers of gut-related diseases, capturing the integrated effects of microbial activity and host physiology. These findings support blood-based SCFA profiling as a promising approach for precision diagnostics in microbiota-associated disorders.