Gut bacteria linked to brain imaging patterns through genetic analysis

Over the past few decades, there has been a growing interest in the intricate interactions between the gut microbiota and brain structure and function. The gut microbiota consists of trillions of highly diverse and metabolically active microbial communities residing in the gut, which play a crucial role in the bidirectional signaling processes of GBA (Cryan et al. 2019). The tripartite connection consisting of the microbiota, gut, and brain links the gastrointestinal system with the central nervous system (CNS) through various regulatory pathways including neural, endocrine, immune, and metabolic mechanisms (Gershon and Margolis 2021), and influences cognitive function, emotional regulation, and neurodegenerative processes (Socała et al. 2021).

Magnetic resonance imaging (MRI), as a non‐invasive neuroimaging technique, serves as a crucial bridge for translating our increasingly deepening understanding of the GBA into clinical applications (Montoro et al. 2022). Studies show that gut microbiota dysbiosis is associated with altered brain connectivity. For example, in amnestic mild cognitive impairment (aMCI) patients, the abundance of Bacteroidetes is negatively associated with the fractional amplitude of low‐frequency fluctuations (fALFF) value of the cerebellar vermis (Liu et al. 2021); in obese individuals, the increased relative abundance of Actinobacteria is associated with elevated fractional anisotropy (FA) in the amygdala and thalamus (Fernandez‐Real et al. 2015). Furthermore, after transplanting fecal microbiota from patients with Attention Deficit Hyperactivity Disorder (ADHD) into germ‐free (GF) mice, alterations in structural connectivity indices in brain regions such as the hippocampus and internal capsule were observed (Tengeler et al. 2020). The absence of commensal microbiota in GF mice during development also results in deficits in brain structure and function (Lu et al. 2018). However, to date, research on the causal interaction of genetic susceptibility of the brain's seven RSNs including DMN, VAN, DAN, Visual Network (VN), Limbic Network (LN), SMN, and FPN with the gut microbiota remains limited.

The gut microbiota may influence brain connectivity through various pathways. Short‐chain fatty acids (SCFAs) (Pascale et al. 2018), the primary metabolites of gut microbiota, play a significant role in the GBA. Ruminococcus, Eubacterium, Coprobacter, and Terrisporobacter are producers of SCFAs within the gut (Vital et al. 2014). Beyond improving gut health (Peng et al. 2009; Lewis et al. 2010), SCFAs can enter the CNS (Vijay and Morris 2014), and exert neuroactive properties (Stilling et al. 2016). In GF mice, colonization with SCFA‐producing strains can alleviate blood–brain barrier (BBB) permeability deficits (Braniste et al. 2014). Moreover, the gut microbiota can regulate neuroinflammation and exert neuroprotective effects through immune pathways (Patnala et al. 2017; Wang et al. 2018; Marizzoni et al. 2020). For example, in patients with end‐stage renal disease (ESRD), dysbiosis of the gut microbiota is associated with elevated pro‐inflammatory factors and abnormal functional connectivity in the DMN and the amygdala (Wang et al. 2019; Zheng et al. 2020). The gut microbiota can also synthesize or regulate neurotransmitters, such as tryptophan, gamma‐aminobutyric acid (GABA), and glutamate, thereby influencing brain network function (Strandwitz 2018; Baj et al. 2019).

The regulatory potential of the GBA has become a research hotspot in nutritional science. Dietary interventions such as supplementing probiotics, prebiotics, and functional foods are considered strategies to enhance brain function and alleviate neurological diseases (Crocetta et al. 2024; Dalile et al. 2025). Bagga et al. found that after 4 weeks of probiotic intake, healthy volunteers showed reduced activity in regions such as the precuneus, middle cingulate cortex, and parahippocampal gyrus (Bagga et al. 2019), accompanied by improvements in emotional regulation and cognitive performance (Bagga et al. 2018). Intervention with probiotic‐containing fermented dairy products in healthy women led to a significant reduction in activation intensity in the periaqueductal gray (PAG) and its associated networks, with improved emotional regulation and cognitive performance (Tillisch et al. 2013). Prebiotics refer to substrates that are selectively utilized by host microorganisms to confer a health benefit (Gibson et al. 2017), widely found in natural fruits, vegetables, and functional foods like yogurt. They primarily influence brain function through SCFAs produced after fermentation by intestinal bacteria (Torre et al. 2021). These intervention strategies may influence the functional connectivity of the DMN and attention networks by modulating the abundance of SCFA‐producers such as Ruminococcus, Eubacterium, and Coprobacter.

Furthermore, as a bidirectional communication system, the GBA allows brain connectivity to reverse‐regulate the composition and function of the gut microbiota. The autonomic nervous system (ANS) triggers rapid changes in the intestinal physiological environment by regulating the enteric nervous system (ENS) (Mayer and Tillisch 2011), and ENS activity is a crucial factor in shaping gut microbial composition (Rolig et al. 2017). In a rat model of intestinal ischemia, electroacupuncture can protect the intestinal barrier by activating a cholinergic anti‐inflammatory mechanism (Hu et al. 2013), and reducing intestinal permeability, potentially modulating the composition of the gut microbiota (Bonaz et al. 2018). Conversely, high‐pressure stress inhibits vagus nerve activity and disrupts gut microbial balance, participating in the pathophysiology of Irritable Bowel Syndrome (IBS) and inflammatory bowel disease (IBD) (Bonaz et al. 2018). Another important pathway is the hypothalamic–pituitary–adrenal (HPA) axis; animal studies have proven that chronic stress activation of the HPA axis has adverse effects on gut microbiota and is associated with increased IL‐6 levels (Bailey et al. 2011). Clinical evidence also confirms that psychosocial stress reduces gut microbiota diversity and hinders the colonization of beneficial bacteria (Ma et al. 2023). In summary, a tight downward regulation exists between the brain and the gut, but the specific mechanisms are not fully elucidated and require further research.

The GBA is jointly regulated by genetic and environmental factors (Berding et al. 2021), but previous observational studies struggle to exclude confounding factors, leaving the causal direction unclear. This study employs the MR method, using the random allocation of alleles during gamete formation as instrumental variables (IVs) to reduce interference from environmental confounding and reverse causality (Sekula et al. 2016). As shown in Figure 1A, this study strictly follows the three core assumptions of MR (Sekula et al. 2016), aiming to establish significant causal associations for the first time between gut microbiota (such as the Ruminococcus and Eubacterium) and structural/functional connectivity in key brain networks (DMN, VAN, DAN, and FPN). Integrating multi‐dimensional research evidence, this study explores the potential of gut microbiota as a neuromodulatory factor and its role in improving neurohomeostasis and cognitive health, providing a theoretical basis for microbiota‐targeted therapies for neuropsychiatric disorders. Figure 2 illustrates the associations between the gut microbiota and IDPs for seven RSNs in the brain, as well as their potential roles in central nervous system disorders.

Our study utilized MR as a robust method to investigate the causal relationship between specific bacterial genera within the gut microbiome and alterations in brain structural and functional connectivity. By harnessing genetic variations as IVs, this approach enabled us to deduce causality from observational data, mitigating confounding factors commonly present in traditional epidemiological studies.

With the rapid progress in neuroimaging and microbiome research, the perspective that the gut microbiota regulates brain structure and function via the GBA has been widely recognized, but the causal direction remains to be elucidated. This research utilizes bidirectional MR methods to confirm the causal link between gut microbiota and brain connectivity from a genetic standpoint. The results show that in the forward MR analysis, R. torques, E. fissicatena, and the Coprobacter genus significantly enhanced DMN connectivity. Terrisporobacter and Gammaproteobacteria exerted positive and negative effects on the connectivity of the DAN and VAN, respectively. Reverse MR reveals that brain network connectivity reciprocally regulates microbial abundance: increased functional connectivity of the DAN was positively correlated with Alloprevotella abundance, whereas enhanced VAN functional connectivity had negative impacts on Alloprevotella, Catenibacterium, and Methanobacteria. Furthermore, the structural connectivity of the FPN and SMN significantly reduced the abundance of Bacilli and Intestinibacter. These findings support the bidirectional regulatory mechanisms of the GBA.

In positive MR analysis, we identified that R. torques, E. fissicatena, and the Coprobacter genus all have significant positive effects on DMN connectivity, suggesting a synergistic role in emotional regulation and the maintenance of cognitive function. As a core brain network closely related to self‐reference, emotional rumination, and episodic memory, abnormal DMN connectivity has been confirmed in various diseases such as IBS, chronic insomnia (CI), and bipolar disorder (BD), which are generally accompanied by gut dysbiosis, suggesting the potential role of the GBA in their pathogenesis (Osadchiy et al. 2020; Li et al. 2022; Feng et al. 2022). Previous studies have shown that the abundance of the Ruminococcaceae family is significantly reduced in IBS patients and is associated with increased anxiety symptoms (Yang et al. 2023; Yusof et al. 2017). As an important member of this family, R. torques abundance is positively correlated with the functional connectivity variability between multiple networks such as the cognitive control network (CCN)–DMN. Changes in the abundance of the Eubacterium and Coprobacter genera have also been reported to be associated with certain brain network indices. Furthermore, Coprobacter abundance is related to the functional integrity of the left angular gyrus (LAG), a key hub of the DMN, whereas reduced regional homogeneity (ReHo) in the LAG is associated with impaired executive function and attention in CI patients (Feng et al. 2022). These lines of evidence suggest that these microbiota may participate in maintaining DMN network homeostasis, thereby affecting emotional and cognitive regulation.

At the mechanistic level, most of these microbiota are SCFA producers (Vital et al. 2014), and their metabolites can participate in the regulation of brain network homeostasis by modulating blood–brain barrier integrity, neuroinflammation, and neurotransmitter metabolism (Wang et al. 2019; Zheng et al. 2020; Li et al. 2022; O'Riordan et al. 2022). For example, Ruminococcus may participate in the bidirectional communication of the GBA by promoting serotonin release, a mechanism that may be impaired in IBS (Labus et al. 2019). In BD patients, gut dysbiosis is accompanied by a decreased capacity to synthesize neuroactive metabolites and weakened limbic–cortical functional connectivity, further suggesting that microbial metabolic abnormalities may be involved in brain network changes (Li et al. 2022). Therefore, the genetic causality revealed in this study provides a theoretical basis for intervention strategies based on diet and microbiome regulation, suggesting that adjusting relevant microbiota structures through dietary fiber intake, prebiotic supplementation, and fermented foods may help improve brain network‐related emotional and cognitive abnormalities. Existing clinical studies provide indirect support: treatment with the probiotic Bifidobacterium longum NCC3001 (BL) can alleviate depressive symptoms in IBS patients and reduce the response of brain regions such as the fronto‐limbic areas to negative emotional stimuli (Pinto‐Sanchez et al. 2017).

Notably, our MR results show the differential impacts of different gut microbiota on the two attention network systems: the SCFA‐producing genus Terrisporobacter has a positive promoting effect on DAN structural connectivity, whereas the pro‐inflammatory Gammaproteobacteria class, containing lipopolysaccharide (LPS), inhibits the functional connectivity of the VAN. Previous clinical studies provide indirect support; for example, Terrisporobacter is significantly reduced in H‐type hypertensive ischemic stroke patients, and its abundance is related to the degree of neurological recovery (Yu et al. 2023); conversely, the LPS components of Gammaproteobacteria can induce the release of pro‐inflammatory factors and trigger neuroinflammatory responses (Oyama et al. 2004; Rothhammer et al. 2016), and its abundance tends to increase in cognitive impairment‐related diseases such as post‐stroke cognitive impairment and Alzheimer's disease (Ling et al. 2020; Liu et al. 2019). As complementary functional attention systems, the DAN and VAN are responsible for top‐down attention control and bottom‐up attention capture, respectively (Alves et al. 2022). Their coordination imbalance has been confirmed in diseases such as major depressive disorder and ADHD, and is associated with abnormal emotional regulation and attention function deficits (Wu et al. 2023).

Crucially, the genetic findings of this study are supported by direct causal evidence from animal models based on human samples. Guzzardi et al. (2023) found that transplanting fecal microbiota from children with high cognitive performance into germ‐free mice resulted in superior learning and memory abilities in the recipient mice during adulthood, with relevant microbiota including members of the Ruminococcus and Eubacterium genera, suggesting that microbiota may participate in the phenotype transfer related to cognitive development. Meanwhile, Tengeler et al. (2020) found that transplanting the microbiota of ADHD patients led to abnormal brain functional connectivity and anxiety‐like behavior in recipient mice, and observed a negative correlation between Eubacterium abundance and anxiety severity, further corroborating the direct regulatory role of human‐derived microbiota differences on attention‐related brain networks. These cross‐species experiments from human donors to animal recipients provide evidence for microbiota–brain phenotype transfer, consistent with the MR inference direction of this study, supporting the causal chain from genetic association to biological evidence. Moreover, gut intervention strategies based on nutritional regulation (such as promoting SCFA producers and inhibiting pro‐inflammatory microbiota) may have potential application value in brain network development and long‐term cognitive protection (Onofrj et al. 2022). However, the specific neurobiological mechanisms still require further verification through longitudinal and experimental studies.

Reverse MR analysis revealed the causal regulatory effect of specific brain networks on the gut microbiota, a novel finding that highlights the importance of top‐down regulatory pathways in the GBA. We speculate that brain network activity may affect the gut physiological state through the autonomic nervous system and neuroendocrine pathways, including regulating intestinal peristalsis, secretion, and immune signaling (Breit et al. 2018), thereby altering the physicochemical environment of the gut and shaping microbial niches (Bonaz et al. 2016). Enhanced DAN functional connectivity may inhibit the release of pro‐inflammatory factors such as TNF by activating the vagus nerve and its mediated cholinergic anti‐inflammatory pathway, thereby reducing intestinal inflammation and maintaining mucosal homeostasis (Pavlov and Tracey 2005), providing a more favorable environment for the Alloprevotella genus. In contrast, when the VAN, FPN, and SMN are over‐activated, it may be accompanied by enhanced HPA axis stress response and up‐regulated sympathetic nerve activity, promoting glucocorticoid release. This process may negatively impact microorganisms such as Alloprevotella and Methanobacteria by interfering with gut barrier function and local immune balance. Previous studies have shown that psychological stress can significantly affect gut microbial composition (Bailey et al. 2011; Bharwani et al. 2016; De Palma et al. 2015) and induce intestinal inflammatory responses (Sun et al. 2013). Furthermore, the opposite regulatory directions of the DAN and VAN toward the Alloprevotella genus also suggest that different attention networks may play different roles in maintaining gut micro‐ecological homeostasis. Although these hypotheses based on neuroanatomy and physiology have certain biological plausibility, future research can combine technologies such as optogenetics or chemogenetics to systematically explore how specific neural network activity regulates the evolution of gut micro‐ecology through neuro‐endocrine‐immune pathways, which will help develop potential brain‐derived microbial intervention strategies for neuropsychiatric diseases.

The potential causal relationships identified in this study between the gut microbiota and brain network connectivity provide important insights for nutritional intervention strategies. Studies have shown that diet is a key factor in regulating the composition and metabolic activity of the gut microbiota (Li et al. 2025; Borrego‐Ruiz and Borrego 2025). For example, foods rich in dietary fiber, whole grains, fruits, and vegetables can alter the gut micro‐ecological structure by promoting the growth of SCFA‐producing bacteria (Crocetta et al. 2024), whereas fermented foods and probiotic supplements can directly regulate the abundance of specific microbial taxa. These diet‐driven microbial changes may further affect brain functional connectivity patterns through the GBA, thus having potential impacts on cognitive function, emotional regulation, and neurodevelopment. Combining the MR results of this study, certain gut genera related to SCFA production may play a role in regulating the structural or functional connectivity of key brain networks. Therefore, promoting the growth of these beneficial microbiota through nutritional intervention means such as dietary fiber, prebiotics, or fermented foods may provide a feasible dietary strategy for improving neurohomeostasis and supporting cognitive health, especially in populations with diseases related to GBA dysfunction.

This study has several advantages. First, bidirectional MR effectively controls for confounding factors and reverse causality common in observational studies. Second, multiple MR methods and sensitivity analyses improve the robustness of the results, mapping brain network features that may be affected by the microbiome. These findings suggest that the gut microbiome may become a potential target for neuropsychiatric interventions. Meanwhile, brain functional abnormalities may in turn affect gut microbial structure, suggesting that improving brain network function through neuromodulation techniques such as brain stimulation therapies (Arulchelvan and Vanneste 2025) may indirectly restore gut micro‐ecological balance. However, this study still has certain limitations. First, the lowest taxonomic level of exposure data are at the genus level, limiting analysis at the species level. Second, human gut microbial composition is affected by environment and lifestyle, which may still lead to some genetic heterogeneity. Furthermore, potential pleiotropy may still exist among some microorganisms. Finally, although MR provides evidence for causal inference, the specific molecular mechanisms still need further verification through animal experiments and clinical studies. Future research can verify these associations in larger samples and different populations, and combine multi‐omics and longitudinal studies to further clarify the molecular mechanisms of the GBA and promote the translational application of microbiome‐based intervention strategies in neurological diseases.

This study evaluates the potential causal relationship between the gut microbiome and brain network connectivity from a genetic perspective through bidirectional MR analysis. The results show that specific gut microbiota can influence the connectivity of different brain networks, whereas brain network activity may also reciprocally regulate gut microbial composition, supporting the bidirectional regulatory mode of the GBA. These findings not only expand our understanding of the mechanisms underlying gut microbiota–brain network interactions but also provide new research perspectives for explaining emotional disorders, cognitive dysfunction, and neurodevelopmental diseases associated with gut–brain imbalance. Meanwhile, this study suggests that targeting gut micro‐ecology and brain network function through strategies such as dietary intervention, probiotic supplementation, or neuromodulation may become potential pathways for intervening in neuropsychiatric disorders. Future research can combine multimodal neuroimaging and multi‐omics approaches to further elucidate the molecular mechanisms of microbiome–brain network interactions, thereby promoting the application of GBA‐based precision intervention strategies in the prevention and treatment of neurological diseases.

Hongchun Liao: software. Hui Liu: methodology. Longtao Yang: formal analysis. Yule Zeng: writing – original draft. Jun Liu: writing – review and editing. Yingxia Zhou: supervision.

The authors declare no conflicts of interest.