Discovering gut microbiome signatures that predict adverse and reverse cardiac remodeling in preclinical heart failure

There is an unmet need to predict adverse or reverse (favorable) cardiac remodeling in preclinical heart failure (stages A and B). We investigated the role of the gut microbiome in cardiac remodeling during asymptomatic heart failure. In this prospective cohort study, 124 subjects were enrolled at baseline (stage A, n = 85; stage B, n = 39). After one-year follow-up, subjects were classified as having adverse (AA vs. AB) or reverse (BB vs. BA) remodeling. Full-length 16S rRNA sequencing of stool DNA was performed. An independent cohort (n = 94) was included to assess model robustness. Here we show that machine learning and differential abundance analyses yield three key findings. First, the top 20 microbiome features predict adverse remodeling (accuracy 0.76; area under the curve 0.86) and reverse remodeling (accuracy 0.85; area under the curve 0.95). Second, taxa identified by both ML and DESeq2 included four genera (Romboutsia, Alloprevotella, Klebsiella, Enterobacter) and five species (R. ilealis, A. rava, K. variicola, K. pneumoniae, E. roggenkampii). Third, pathway analysis suggests beneficial effects of R. ilealis in stage AA via short-chain fatty acid and ketone body metabolism, whereas Klebsiella spp. and E. roggenkampii may be linked to pro-inflammation in stage BB. Cohort-cumulation modeling demonstrates stable performance (area under the curve 0.75-0.80) of the models in the validation cohort. The gut microbiome predicts cardiac remodeling in preclinical heart failure and represents a potential therapeutic target, with possible implications for risk stratification and personalized management. Heart failure can develop gradually before symptoms appear, but it remains difficult to predict whether heart structure and function will worsen or improve during these early stages. We investigated whether bacteria living in the gut could help predict these changes. We analyzed stool samples from people with early-stage heart failure without symptoms and identified gut bacterial patterns associated with changes in heart structure and function over one year. We identified specific gut bacteria that could predict worsening heart structure and function, while others could predict a return to healthier heart structure and function, with good accuracy. These microbial patterns may help identify individuals with different risks of disease progression. Our findings also provide directions for future research into whether the gut microbiome could contribute to strategies for preventing or treating heart failure. Yang, Liao, Lin et al. investigate gut microbiome signatures associated with adverse and reverse cardiac remodeling in preclinical heart failure. Gut microbiome profiles predict both remodeling trajectories, with specific bacterial taxa associated with favorable or adverse remodeling.