The gastrointestinal microbiota has emerged as a key contributor to numerous physiological processes and displays remarkable adaptability to environmental and metabolic challenges. It can be regarded as a discrete organ of substantial plasticity, capable of influencing distant organs through the release of bioactive metabolites (Marques et al., 2018; Fujisaka et al., 2023). Among its systemic effects, the gut microbiota plays an important role in cardiovascular regulation (Tang et al., 2017; Marques et al., 2018). Fermentation-derived short-chain fatty acids, such as acetate and propionate, can modulate vascular tone via G-protein-coupled receptors (e.g., GPR41, GPR43, Olfr78), thereby affecting blood pressure and peripheral resistance. These same metabolites and other microbially derived compounds have also been suggested to influence cardiovascular control indirectly by altering the sympathovagal balance and autonomic reflexes (Toral et al., 2019; Muller et al., 2020).

In line with this concept, we have recently demonstrated in healthy subjects that acute stimulation of gut microbial fermentation with lactulose enhances the ventilatory response to transient hypoxia, reflecting a transient increase in peripheral chemoreflex sensitivity (Seredyński et al., 2021). This finding suggests that fermentation-related signalling can presumably modulate carotid body function and, ultimately, the autonomic cardiovascular control.

Sympathetic overactivation and the altered cardiopulmonary reflexes (hyperactive peripheral chemoreceptors [PCh] in particular) are hallmarks of the chronic heart failure (HF) syndrome and are directly linked to adverse outcomes (Ponikowski et al., 2001; Giannoni et al., 2023). Given the postulated interaction between gut microbial activity and chemoreflex function in healthy humans, together with well-documented alterations in gut microbiota composition and barrier integrity in HF (Tang et al., 2017), it appears reasonable to explore whether enhancing intestinal fermentation can acutely modify chemoreceptor-driven ventilatory responses in this patient population.

In this paper, we attempted to translate our findings from healthy subjects to HF patients. In brief, chronic HF patients underwent the assessment of PCh function with the transient hypoxia test twice, before and ∼120 min after the ingestion of a gut-fermentation-stimulating meal. The content of hydrogen in the expired air was assessed repeatedly during the experiment and taken as a measure of fermentation intensity.

Sixteen HF patients were recruited to the study. Four participants were excluded from further analyses due to the incomplete data (2 participants; measurements were terminated due to the reported discomfort, e.g., related to the pressure of the oronasal mask) or low quality recordings (2 participants; data could not be reliably computed due to the periodic breathing and fluctuations in the blood oxygen saturation). Baseline characteristics of 12 patients (all males) included are summarized in Table 1.

We have previously demonstrated that the acute increase in gut microbial fermentation augments peripheral chemoreflex sensitivity in young, healthy volunteers (Seredyński et al., 2021). Present study is the first to examine the presence of such interactions among heart failure patients. Individuals exhibiting early (<60 min) rise in gut microbial fermentation after the lactulose ingestion (High Early Fermentation patients, HEF) revealed higher initial ventilatory and vascular responses to hypoxia, compared to the Low Early Fermentation patients (LEF). Moreover, in the HEF group, initial HVR positively correlated with the hydrogen excretion during the lactulose test.

These observations may be interpreted as indicating a link between enhanced bacterial fermentation in the small intestine and heightened peripheral chemosensitivity in HF patients. However, the assessment of microbiota-dependent fermentation using the hydrogen breath test is inherently influenced by inter-individual variability in oro-cecal transit time, which warrants cautious interpretation of these findings. Nevertheless, as discussed in more detail in the Study Limitations section, we deliberately applied stringent criteria to define early fermentation in order to virtually eliminate false-positive classification, albeit at the expense of potentially failing to identify some HEF individuals. Notably, applying less conservative criteria did not change the main results of the study (see Supplementary Figure S1).

Among 12 study participants, 6 were classified as HEF, reflecting the prevalence of SIBO reported for larger HF cohorts (Song et al., 2021). Indeed, a growing body of evidence links the gut dysbiosis and disrupted intestinal barrier to worsened inflammation and poor metabolic profiles in HF, potentially aggravating disease progression and mortality (Mollar et al., 2019; Song et al., 2021; Song et al., 2024). In comparison, little is known about the relation between gut microbial disturbances and autonomic imbalance–a hallmark malfunction in HF (Floras and Ponikowski, 2015; Giannoni et al., 2023). Nevertheless, several animal (O’Connor et al., 2019; Wang et al., 2022) and human (Seredyński et al., 2021) studies feature gut microbiota as important modulators of autonomic functions, plausibly via its fermentation-derived metabolites.

In this pilot HF study, we indeed obtained a relation between the gut fermentation intensity and ventilatory response to hypoxia, consistent with the general pattern previously reported in healthy volunteers (Seredyński et al., 2021). It should be emphasized, however, that the two studies differed in terms of experimental protocols and conditions and were conducted independently; therefore, no direct quantitative or statistical comparisons between datasets were performed or intended. Nevertheless, the earlier healthy-volunteer study provides a relevant physiological context for interpreting the present findings. Primarily, our HF cohort did not exhibit a significant overall change in HVR after the lactulose ingestion, in contrast to the lactulose-induced HVR upregulation observed in healthy subjects. Instead, HF patients with increased upper gut fermentation (HEF) showed an intrinsically elevated baseline HVR compared to the LEF group, a difference that disappeared post-lactulose. Moreover, only in the HEF group did hydrogen excretion correlate positively with baseline (but not post-lactulose) HVR, indicating that individuals with greater early gut microbial fermentation tend to have heightened chemoreflex sensitivity even before any experimental intervention.

It is tempting to speculate that early rise in the lactulose-induced hydrogen production (often interpreted as a proxy for bacterial overgrowth of the upper gastrointestinal tract) might be associated with increased small intestinal permeability (Lewis and Taylor, 2020; Kitai et al., 2022), leading to the enhanced outflow of microbial metabolites to the bloodstream (Snelson et al., 2025), which in turn alters the sensitivity of peripheral chemoreceptors. Such hypothesis remains coherent with the lack of direct effect of lactulose ingestion on HVR in HEF group. HEF patients’ ventilatory drive might be already elevated at baseline due to chronic chemoreceptor microbial stimulation, leaving little room for further enhancement with an acute fermentative stimulus. In fact, post-lactulose HVR did not change (or even decreased) for the most of HEF subjects, but raised modestly in LEF patients. Interestingly, rise in HVR following lactulose meal was reported in our previous study in healthy volunteers (Seredyński et al., 2021). However, given the methodological limitations of hydrogen breath testing, such interpretations should be treated with caution. Early hydrogen excretion does not uniquely reflect fermentation capacity and may also be influenced by altered oro-cecal transit. The present findings should be interpreted as identifying an association between hydrogen excretion kinetics and chemoreflex responsiveness, rather than establishing a direct mechanistic link between gut fermentation and PCh function.

Among hemodynamic variables, significant differences were obtained solely for the systemic vascular resistance hypoxic response (hyp-SVR). HEF group revealed higher pre- (baseline) and post-lactulose hyp-SVR values compared to LEF patients. Opposite to HVR, though, we failed to obtain correlation between hyp-SVR and hydrogen excretion. This suggests the small intestinal fermentation affects ventilatory responses in a more direct way that the hemodynamic ones (Seredyński et al., 2021). It aligns with a growing body of data that the two arms of the peripheral chemoreflex can be differentially modulated under the same stimulus (Keir et al., 2019; Zera et al., 2019). Another level of complexity might arise from the bi-directional nature of the gut microbiota-sympathetic division communication. HF itself is known to alter the gut environment: chronic HF leads to intestinal edema and reduced visceral perfusion, which in turn might readjust the vascular sympathetic tone (Fudim et al., 2021), potentially attenuating gastrointestinal motility (Zheng et al., 2024) and promoting gut dysbioses. In line with that, there is rising clinical interest in targeting gut microbial overgrowth for HF management, with a rising number of randomized trials ongoing.

Although direct experimental evidence showing that gut-microbiota modulation can reduce sympathetic activity in HF through alterations in the PCh function is currently lacking, several mechanistic studies provide a physiologically plausible foundation for this pathway. Animal models consistently demonstrate that targeted manipulation of the gut microbiome can attenuate sympathetic overactivity in certain sympathetically-mediated disorders; mostly in hypertension (Klippel et al., 2016; Toral et al., 2019; Kong et al., 2021; Liu et al., 2022; Yin et al., 2024), with more limited but emerging evidence in HF models (Liu et al., 2024). (Klippel et al., 2016; Toral et al., 2019; Kong et al., 2021; Liu et al., 2024; Yin et al., 2024). Reductions in sympathetic activity in these models are typically accompanied by significant improvements in blood pressure, further highlighting the functional relevance of gut–autonomic interactions. These findings are supported by human data showing that interventions that substantially modify the gut environment, i.e., Roux-en-Y gastric bypass (Zhang et al., 2014) or prebiotic supplementation (Jama et al., 2023) can exert measurable antihypertensive effects. On the other hand, well-established role of the heightened peripheral chemosensitivity in generating sympathetic overactivation in HF (Marcus et al., 2014; Marcus et al., 2018; Niewinski et al., 2017; Toledo et al., 2017), and the presence of receptors for microbiota-derived metabolites on the carotid bodies (Peng et al., 2020) provide a plausible cellular substrate through which gut-microbiota alterations could modulate carotid-body signaling and, consequently, autonomic outflow.

Our work has several limitations. First, this was a preliminary observational experiment involving 12 H F patients, without randomization or a control group. As such, the study was not designed to determine causality, and all findings should be interpreted as exploratory. Second, we relied on hydrogen breath tests during the quantification of gut microbial fermentation, as well as classification of participants as HEF or LEF. According to the literature, about 30%–62% of the human population are ‘methane producers’, and in some people methane is even a dominant gaseous product of gut microbial fermentation (De Lacy Costello et al., 2013). Therefore, hydrogen breath tests we performed could lead to the underestimation of the gut fermentation intensity. Nevertheless, hydrogen production has been reported as a better predictor of long-term adverse event risk in HF patients than methane, which supports our focus on hydrogen in this pilot study (Mollar et al., 2019). From a practical standpoint, integrating the lactulose breath test with cardiovascular and autonomic assessments within a single morning visit required a bedside technique with immediate readout; a handheld hydrogen analyzer allowed on-site measurements without gas sample collection or storage, thereby facilitating tight temporal alignment of procedures and minimizing between-session variability. Future studies should employ multi-gas (H₂/CH₄) breath testing to more comprehensively characterize fermentation profiles in HF. Moreover, we did not quantify short-chain fatty acids (SCFAs) in blood or feces, so our assessment of fermentation was limited to breath hydrogen; future, larger studies should integrate targeted SCFA/metabolomics profiling to better characterize microbiota-derived metabolites potentially linking gut fermentation with cardiovascular and autonomic regulation.

Thirdly, our HEF/LEF classification relied on an early rise in exhaled hydrogen, an approach that is sensitive to inter-individual variability in oro-cecal transit time (Rangan et al., 2022), which might be of even greater importance in HF with frequent gastrointestinal ailments (Mahenthiran et al., 2024).

Whereas >20 ppm hydrogen cutoff is a recommended value for the lactulose HBT-based detection of the augmented upper gut fermentation in SIBO (Rezaie et al., 2017; Dahlgren et al., 2025), early H₂ rise timeframes proposed in the literature differ from 60 to 100 min; however, as demonstrated with concomitant scintigraphy, time window >80 min after the lactulose ingestion may substantially increase the risk of false-positive results (Sunny et al., 2016; Dahlgren et al., 2025). As our pilot study did not aim to diagnose SIBO, we deliberately applied very strict lactulose HBT criteria (≥20 ppm rise within 60 min after 10 g lactulose) to focus on an early fermentation phenotype and to minimize the contribution of later, likely colonic fermentation to HEF classification, accepting that this strategy reduces sensitivity and may have led to misclassification of some patients as LEF.

Nevertheless, HEF/LEF classification in this study should be regarded as a pragmatic, hypothesis-generating stratification based on hydrogen kinetics, rather than a definitive marker of intrinsic fermentation capacity.

Finally, as this was a single-centre pilot study, small sample size inevitably limited the statistical power of the between-subgroup comparisons. This limitation also influenced our statistical strategy: although several physiological variables were analysed, we did not apply formal corrections for multiple comparisons. Such procedures (e.g., Bonferroni or Holm) are highly conservative in small mechanistic studies and substantially increase the risk of type II errors, potentially obscuring physiologically meaningful effects. Instead, in line with contemporary statistical recommendations for hypothesis-driven exploratory studies (Perneger, 1998; Armstrong, 2014), we prioritised the interpretation of effect sizes and confidence intervals around the means. Notably, for the three key findings of the study, effect sizes quantified as Cohen’s d exceeded 1.4, which is conventionally interpreted as a large effect, supporting the physiological relevance of these observations despite limited sample size (Cohen, 1992). On the other hand, the study population was somewhat homogenous (in terms of sex, NYHA class, pharmacological treatment, etc.), and several precautions were taken to minimize the intrasubject variability (all data–H₂ excretion, resting parameters, hypoxic responses–were collected during one morning session after the overnight fasting). Nevertheless, further controlled trials in a larger cohort are warranted.

Present findings demonstrate that increased upper gut fermentation in HF may contribute to the augmented hypoxic responsiveness prior to any acute challenge. This extends the findings from healthy subjects by indicating that intrinsic gut microbial activity might be an additional factor shaping peripheral chemosensitivity in HF. Unravelling this putative gut–peripheral chemoreflex connection may broaden our understanding of HF pathophysiology, but given the exploratory, non-randomized design and small sample size, these observations should be viewed as hypothesis-generating and require validation in larger cohorts before any clinical implications can be drawn.