Mental health and gut-brain crosstalk: implications for depression and Alzheimer's disease

Abstract

Brain health and mental health disorders are increasingly becoming an essential priority for modern societies as they affect different parameters of life, such as brain health, quality of life, and productivity with a heavy societal and economic burden. Among mental health disorders, depression and Alzheimer disease (AD) have the higher impact on mental health globally as they are highly prevalent, cause long-term cognitive and mood deficits while they are deeply interconnected, with chronic stress raising as a risk factor and potential link between both disorders. This review focuses on the gut-brain axis, a bidirectional communication network that links the gut microbiome with the central nervous system, and its role in brain malfunction and pathology related to stress, depression and AD. This intricate gut-brain crosstalk is orchestrated through top-down and bottom-up mechanisms implicating the HPA axis, the enteric nervous system as well as gut microbiota-derived metabolites, neurotransmitters, epigenetic processes and extracellular vesicles/exosomes that can contribute to depression and AD. The current work provides a comprehensive summary of evidence linking gut microbial alterations to these brain pathologies, with particular focus on lifestyle and nutrition (e.g. food, water) as critical regulators. Lastly, we explore the therapeutic potential of microbiome-targeted interventions including pro/pre/post-biotics, and trace elements (e.g. lithium and silicon). Overall, this review highlights the potential of microbiome-centered strategies as novel interventions to support mental health and wellbeing.

Graphical abstract

Highlights

• •Brain health and mental health disorders are increasingly becoming an essential priority for modern societies
• •Gut microbiome appears as a crucial interface between environment and brain via the gut-brain axis
• •Emerging evidence suggests a crucial role for gut microbiome homeostasis in brain health as well mental health disorders such as depression and Alzheimer's disease
• •Lifestyle, dietary interventions affect gut microbiome health and maybe involved in the vulnerability to brain pathology.

1. Introduction

Over the last years, mental and brain health has become a strategic priority for major international organizations, including the World Health Organization (WHO), the United Nations, and the World Bank. In particular, mental health is now embedded within the Sustainable Development Goals, particularly under the goal of ensuring healthy lives and promoting wellbeing. Representing a major public health problem, mental health disorders such as depression, anxiety, and Alzheimer's disease (AD) are among the leading causes of disability worldwide. For instance, one in every eight people - approximately 970 million individuals worldwide - were living with a mental disorder in 2019, with anxiety and depressive disorders being the most prevalent. Anxiety and fear-related disorders - such as generalized anxiety disorder, panic disorder, social anxiety disorder, and specific phobias - exhibit a lifetime incidence of 28.8% in the U.S. (Kessler et al., 2005). These conditions typically emerge during adolescence and often co-occur with depression, suggesting overlapping etiological pathways (Bandelow et al., 2017). According to the WHO, depression imposes a significant global burden of disease and is a leading cause of disability in western societies (Ferrari et al., 2013). A recent systematic review highlighted the impact of the COVID-19 pandemic on mental health, estimating a 27.6% increase in global prevalence of major depressive disorder (MDD) and a 25.6% increase in anxiety disorders in 2020 (COVID-19 Mental Disorders Collaborators, 2021). Importantly, current pharmacotherapies for depression are limited by delayed onset of action, incomplete efficacy, and a lack of response in approximately 35% of patients (O'Leary et al., 2015). On the other hand and based on the strong interconnection between mood and cognition, clinical and preclinical evidence suggest that depression is also associated with elevated risk for AD (Ownby et al., 2006; Vyas et al., 2016). Indeed, dementia and AD (which represents around 60-70% of dementia cases) is a WHO health priority problem as they affect more than 57 and 44 million people worldwide, with their prevalence being predicted to triplicate by 2050 (Alzheimer's Disease International Prince et al., 2015) (Fig. 1). Moreover, numerous studies link depression to the onset of Mild Cognitive Impairment (MCI), a prodromal phase of AD, and the transition from MCI to AD with preclinical studies providing common neurobiological mechanisms among depression and AD (Dioli et al., 2023; Modrego and Ferrández, 2004). Although it remains unknown whether depression can act both as a cause and as a result of AD, emerging evidence points towards chronic stress as a precipitating factor of these brain disorders (Burke et al., 2024).

Fig. 1.

Mental health disorders and their global impact. Key features related to the impact of mental health disorders worldwide. Note that 12.5% of the global population suffers from mental disorders while the percentage of untreated depressive patients remains as high as 75%. However, only 2% of healthcare budget is dedicated to mental health with the vast majority of countries not having a national plan for brain diseases.

Specifically, longitudinal studies suggest that sustained psychological stress exposure increases lifetime risk for depression and anxiety (Y.-Z. Liu et al., 2017) as well as AD (Johansson et al., 2010). Furthermore, early-life adversity is associated with enduring epigenetic changes that predispose to these disorders (Meaney and Szyf, 2005; Vyas et al., 2016). Indeed, chronic stress is known to impact brain plasticity, hormonal regulation, and immune signaling. Hereby, disruptions in monoaminergic neurotransmission - particularly serotonin and dopamine - along with dysregulation of the hypothalamic-pituitary-adrenal (HPA) and sympathetic-adrenomedullary systems are strongly implicated (Kokras et al., 2019). Chronic elevation of cortisol impairs neuronal plasticity as well as neurogenesis, particularly in the hippocampus, a region crucial for emotional regulation and memory. Consistent with this, individuals with MDD exhibit hyperactivity of the HPA axis and elevated cortisol levels (Bertollo et al., 2020) while dysregulation of HPA axis is also reported in AD patients (Canet et al., 2019), where the elevated cortisol levels were related to low memory scores (Dronse et al., 2023). Animal studies corroborate these findings, showing that chronic glucocorticoid exposure leads to depressive-like behaviour (Qin et al., 2015). A recent review (Hodes et al., 2024) highlights how these stress responses differ markedly by sex, with females exhibiting heightened corticotropin releasing hormone (CRF) receptor signaling and increased sensitivity to stress-induced neuroendocrine and immune responses, contributing to their higher prevalence of mood and anxiety disorders. In relation to AD, animal studies using transgenic and non-transgenic models of AD have shown that exposure to chronic stress and/or high levels of glucocorticoids trigger both main AD pathomechanisms (e.g. amyloid-β production, Tau hyperphosphorylation and accumulation) leading to neuronal atrophy and exacerbating memory and mood deficits (Sotiropoulos et al., 2011; Lopes et al., 2016a). Interestingly, emerging evidence suggests common neurobiological mechanisms between depression and AD. Specifically, AD-related mechanisms are also found to be affected in depression. For instance, depressive symptoms in cognitively normal elderly people are correlated to Aβ brain levels (Donovan, N.J. et al., 2018). Also, Tau positron emission tomography (Tau-PET) brain analysis in cognitively normal elderly people suggest a link between depressive symptoms and accumulation of Tau in the entorhinal cortex, the brain region that is initially affected in AD showing Tau pathology (Gatchel et al., 2017]. Moreover, animal studies in wild-type animals under chronic stress suggest the involvement of Tau and its hyperphosphorylation in the establishment of stress-driven brain malfunction and related depressive behavior as animals lacking Tau (Tau knockout) did not exhibit the typical neuronal atrophy, suppressed neurogenesis, mood and cognitive deficits under chronic stress exposure (Lopes et al., 2016a; Dioli et al., 2017). Thus, as clinical and preclincial evidence supports a cumulative impact of lifetime stress and stressful conditions on the precipitation of both AD and depression, chronic stress could serve as a connecting link between both disorders, highlighting the stress-driven deficits in brain plasticity as a key parameter in the continuum between mood and cognitive disorders.

An emerging body of research work highlights the gut-brain axis, a complex bidirectional communication network linking brain to gut microbiome, as a crucial interface between the environment and the brain biology (Fig. 2). For instance, chronic stress is shown to disrupt the gut microbiota, leading to increased intestinal permeability (referred to as “leaky gut”), systemic inflammation, and altered neurotransmitter systems, including serotonergic signaling (Cryan and Dinan, 2012). Disruption of the balance of gut microbiome - often referred to as ‘dysbiosis - has been implicated in the development of anxiety, depression, and neurodevelopmental disorders, as gut microbes can influence the synthesis of neurotransmitters, short-chain fatty acids (SCFAs), and immune mediators that affect brain function. Although the term “dysbiosis” is widely used to describe disease-associated alterations in gut microbial composition/function, its definition remains conceptually broad and debated, and it does not necessarily imply a causal state (Hooks and O’Malley, 2017). In this review, we use the term “dysbiosis” cautiously to refer to disease-associated alterations in microbial composition and function. Beyond compositional alterations, emerging evidence suggests that host biological variables also shape gut–brain interactions. In particular, sex differences extend to the gut microbiome and its interactions with the brain. Recent findings (Hodes et al., 2024) suggest that females exhibit greater blood-brain barrier (BBB) permeability in prefrontal regions and different patterns of microglial activation from peripheral influence compared to males under stress. These differences may underlie distinct susceptibilities to stress-induced neuroinflammation and behavioral outcomes. Importantly, lifestyle interventions that restore microbial balance - such as diets rich in fiber, fermented foods, probiotics, and prebiotics - are gaining recognition as adjunctive treatments for stress-related disorders (Sarkar et al., 2018). Moreover, regular physical activity, sleep hygiene, and mindfulness practices have shown benefits in modulating both brain stress responses and gut health. This review explores how modern lifestyle factors, nutrition, and emerging interventions affect gut-brain axis and, consequently, mental health and wellbeing. It also highlights how microbial diversity and composition are shaped by physiological, environmental, and behavioral parameters, and how the gut-brain axis mediates bidirectional communication through neural and gut routes. The review further examines the contribution of gut microbiome imbalance to the precipitation of brain disorders (e.g. depression and AD), while considering the role of microbial metabolites, neurotransmitters, and inflammatory processes in these brain diseases (Fig. 2). Significant focus is placed on the potential role of lifestyle interventions, including dietary fiber, probiotics, postbiotics, prebiotics and physical activity, to support microbial and mental health, as well as the emerging evidence for the neuroprotective and microbiota-modulating properties of trace elements such as lithium and silicon.

Fig. 2.

Gut microbiome as a crucial interface between environment and brain. Different environmental factors (e.g. lifestyle, diet, psychological stress and exercise) are known to affect brain integrity and function regulating several biological elements that impact on the brain (e.g. HPA axis, inflammatory response, epigenetic regulation, synaptic plasticity). Emerging evidence suggests gut microbiome as a crucial interface between environmental and biological factors that regulate brain health and function.

Altogether, this review calls for a paradigm shift in mental health care, one that integrates neurobiological, environmental, and microbiome-related factors, considers sex as a biological variable, and embraces holistic and preventive approaches to promote mental health and wellbeing. Investment in mental health is not only a health priority but also an economic necessity, with evidence showing that early intervention and preventive strategies yield substantial returns in productivity and societal wellbeing.

2. The gut microbiome: composition, lifespan dynamics, and functional significance

The human body hosts a wide array of microorganisms, including bacteria, viruses, protozoa, fungi, and archaea, collectively referred to as the ‘human microbiota’ (Sasso et al., 2023). The community of these microorganisms colonizing the gastrointestinal (GI) tract is termed as ‘gut microbiota’, while the term ‘gut microbiome’ includes not only these microorganisms, but also their structural elements, metabolites, and the surrounding environment (Hou et al., 2022). Gut microbiota is primarily dominated by the phyla: Firmicutes (with Clostridium being the major genus), Bacteroidetes (mainly Bacteroides and Prevotella), Actinobacteria (especially Bifidobacterium), along with Proteobacteria, Fusobacteria, and Verrucomicrobia. Firmicutes and Bacteroidetes alone make up around 90% of the gut microbial community (Arumugam et al., 2011).

The composition and diversity of the gut microbiota are shaped by various natural, non-pathological factors, with aging being one of the most prominent. Microbial colonization of the GI tract evolves significantly throughout human lifespan. In infancy, the microbiota is relatively simple in diversity, dominated mainly by Bifidobacterium genus, a genus that by the end of the first-year declines as other microbes begin to colonize, leading to a greater diversity (Yatsunenko et al., 2012). Around the age of three, as children start consuming various solid foods, the gut microbiota stabilizes and becomes more diverse (Ragonnaud and Biragyn, 2021). In adulthood, the microbiota is primarily dominated by Firmicutes and Bacteroidetes, with smaller proportions of Actinobacteria, Proteobacteria, and Verrucomicrobia (Eckburg et al., 2005). In the elderly, the gut microbiota becomes less diverse and often enriched with inflammation-linked microbes, potentially contributing to age-related diseases. Studies in aged mice and macaques have shown that upon aging, microbiota undergoes a shift, with a reduction in beneficial bacteria (e.g. Akkermansia muciniphila and SCFAs-producing Clostridium species), and an increase in pro-inflammatory microbes (Bodogai et al., 2018). This shift appears intrinsic to the aging process, as similar changes are observed across humans and other species. However, aging is not the only factor that can influence gut microbiota diversity. As will be discussed later, factors such as antibiotic use (Ae et al., 2013), ethnicity and cultural habits, diet (Arumugam et al., 2011), exercise frequency (Bai et al., 2019), pH (Firrman et al., 2022) and even variations across anatomical regions of the GI tract (Flint et al., 2012), also play significant role. These factors do not act individually but interact in complex ways, often enhancing each other's effects on shaping the gut microbiome. A clear example of this interaction is seen in breast-fed versus formula-fed infants. Breast-fed infants typically develop a microbiota dominated by Bifidobacterium, largely due to the presence of human milk oligosaccharides (HMOs), which these bacteria utilize to produce lactate and SCFAs like acetate. In contrast, formula-fed infants are exposed to a more diverse microbial population due to the absence of HMOs (Ragonnaud and Biragyn, 2021). This highlights how dietary inputs can compound age-related microbial shifts.

The intricate interplay of factors that shape the composition of the gut microbiome ultimately determines its functional capacity, which is vital for numerous physiological processes. The gut microbiome plays a pivotal role in the digestion and nutrient absorption (Oliphant and Allen-Vercoe, 2019), as well as the synthesis of bioactive molecules, including vitamins, amino acids, and lipids. The GI tract is responsible for digesting proteins and sugars, with the metabolism of polysaccharides and specific proteins depending on enzymes produced by various bacterial species. For example, Bacteroides species in the large intestine contribute to sugar breakdown (Martens et al., 2008), while pathogenic Enterobacteriaceae also utilize sugars and amino acids within the gut (Ducarmon et al., 2019). Beyond digestion, the gut microbiome is also crucial for immune system maturation, particularly through its role in mediating neutrophil migration, which in turn influences the differentiation of T cells (Owaga et al., 2015). Additionally, the microbiome plays a key role in regulating inflammation. Importantly, the gut microbiome exists in a state of dynamic equilibrium, constantly influenced not only by the microorganisms residing in the GI tract but also by external factors. This intricate balance has profound implications, extending beyond the gut and particularly in its bidirectional communication with the brain, highlighting the emerging role of the gut-brain axis (Fig. 3).

Fig. 3.

The gut-brain axis: a bidirectional communication between the brain and the gut. This diagram illustrates the reciprocal relationship between the gut microbiome and the brain. Clinical and preclinical evidence suggests that a balanced gut microbiome supports proper brain function and mental health and, in its turn, a healthy brain promotes gut microbiome stability and diversity. Conversely, mental health disorders and brain malfunction can disrupt gut microbial balance, leading to gut dysbiosis. Also, gut microbiome dysbiosis seems to contribute to brain malfunction and mental illness, forming a vicious cycle.

3. The gut-brain-gut loop: mechanisms of communication along the gut-brain axis

Even though gut and brain are two totally anatomically different organs and entities, several mechanisms have been identified through which gut microbes can influence the central nervous system (CNS) and vice versa (see Fig. 4). Brain-gut interactions primarily occur through ‘neuronal pathways’ such as the enteric nervous system (ENS), the vagus nerve (VN), and the HPA axis, through which the CNS exerts top-down control over GI function. Conversely, gut-brain signaling involves bottom-up communication mediated by gut-derived factors, including neurotransmitters, microbial metabolites, and extracellular vesicles (EVs), which may influence brain function and behavior. Both directions of communication are critically integrated through the circulatory system, which serves as a central conduit, completing the gut-brain-gut loop and ensuring dynamic, bidirectional regulation. Thus, gut-brain axis establishes a dynamic connection between both enteric and nervous system, linking the brain's cognitive and emotional centers with gut physiology.

Fig. 4.

The complex “roadmap” of the gut–brain axis. This scheme illustrates the bidirectional communication between the gut and the brain which includes both efferent and afferent ways of communication. The brain-to-gut communication path includes: i) efferent fibers of the vagus nerve which modulate gut physiology and microbial composition, and, ii) the activation of the hypothalamic–pituitary–adrenal (HPA) axis that leads to cortisol release, which reaches the gut via the bloodstream and influences both gut function and microbiota diversity. On the other hand, the gut-to-brain route consists of three main mechanisms: 1) neurotransmitters produced by the gut microbiota and enteroendocrine cells (e.g. enterochromaffin cells) activate afferent fibers of the vagus nerve, which projects to the brain via the nucleus tractus solitarius; 2) short-chain fatty acids (SCFAs) produced by the gut microbiota enter systemic circulation and, crossing the blood–brain barrier (BBB), they reach the brain influencing its function; 3) extracellular vesicles (EVs) secreted by the gut microbiota carry a variety of cargo—including neurotransmitters, SCFAs, nucleic acids, and proteins—which also cross the BBB and act on the brain.

3.1. Brain-to-gut communication

3.1.1. Enteric nervous system

ENS is frequently named as the ‘second brain’ due to its structural and functional parallels with the CNS (e.g. complexity, similar neurotransmitters etc) and it comprises over 400 million enteric glial cells and approximately 100 million neurons (Dowling et al., 2022). These cells interact with various other cell types such as interstitial cells of Cajal, smooth muscle cells, epithelial cells and vascular structures within the GI wall (Fried et al., 2021). Enteric glial cells, which are distributed throughout the intestinal wall, provide structural and metabolic support to enteric neurons and play active roles in neuromodulation, neurotransmission, and neuroprotection (Neunlist et al., 2014). Neurons of ENS are functionally and morphologically classified into afferent neurons, efferent neurons, interneurons, and excitatory or inhibitory motoneurons. Among them, intrinsic primary afferent neurons are responsible for sensing mechanical and chemical stimuli and transmitting this sensory information to other neurons within the ENS. These intrinsic sensory pathways form the foundation of the autonomic brain–gut axis. Additionally, communication with the CNS is facilitated by extrinsic sensory afferent neurons, which complete the gut–brain–gut regulatory circuit and contribute to the somatosensory aspects of gut–brain signaling. Interneurons constitute the linkage between intrinsic primary afferent neurons and motor neurons (Costa et al., 2000); the latter are excitatory or inhibitory and regulate the contraction and relaxation of smooth muscle layers, thereby controlling gut motility. Finally, the interstitial cells of Cajal often referred to as the “pacemaker cells” of the GI tract, are responsible for generating the spontaneous electrical activity that initiates and propagates peristaltic movements (Sanders et al., 2016). Altogether, these form the complex ENS which autonomously regulates GI functions while maintaining communication with the CNS.

3.1.2. Vagus nerve

Besides ENS, the VN represents a key anatomical and functional pathway within the gut-brain axis. The VN regulates critical autonomous functions, including digestion, heart rate, respiratory rhythm, vasomotor control, and several reflexes. However, its primary role is transmitting sensory information from internal organs to the CNS (Breit et al., 2018). Approximately 80% of the vagal fibers are afferent, while the remaining 20% are efferent (Bonaz et al., 2021), thus acting as a crucial bidirectional communication bridge between the CNS and peripheral organs. Vagal afferents are polymodal, responding to mechanical, chemical, and hormonal signals (Egerod et al., 2018), and can detect gut microbiota and their metabolites. This gut-to-brain sensory transmission is facilitated by specialized enteroendocrine cells (known as ‘neuropods’) (Bohórquez et al., 2015), which form synapses with vagal neurons, allowing fast and direct signaling from the gut to the brain (Kaelberer et al., 2020). VN is critically connected with emotional and behavioural centers of the CNS, as signals transmitted via the VN reach the nucleus tractus solitarius (NTS) in the brainstem, which in turn projects to different brain regions, such as hypothalamic nuclei, paraventricular nucleus, amygdala, thalamus, hippocampus and prefrontal cortex (Han et al., 2022). Given its extensive anatomical and functional connections, growing evidence supports the therapeutic application of VN stimulation for mood disorders. For instance, preclinical studies using adult male Sprague–Dawley rats have shown that chronic VN stimulation (20 Hz, 250 μsec pulse width, 250 μA, 30 s/every 5 min for two weeks) produced significant anxiolytic and antidepressant effects, with efficacy comparable to pharmacological treatment with sertraline (7.5 mg/kg/day) and desipramine (10 mg/kg/day) (Furmaga et al., 2011). Similarly, in male Wistar Kyoto rats, repeated VN stimulation (20 Hz, 0.5-ms, 1-mA pulses, 30 min daily for four days) exhibited antidepressant effect, paralleling the effects of desipramine and electroconvulsive shock therapy, both exhibiting antidepressant action (Krahl et al., 2004). Building on these preclinical findings, VN stimulation with implanted devices has been approved by the U.S. Food and Drug Administration for the treatment of epilepsy and treatment-resistant depression. This approval was followed by extensive clinical research demonstrating the safety and efficacy of VN stimulation in patients with depression (Bottomley et al., 2019). As an example, VN stimulation treatment for 10 weeks in 30 patients with treatment-resistant depression resulted in a ≥50% reduction in baseline scores on the 28-item Hamilton Depression Rating Scale (28-HDRS), which was achieved in nearly half of the participants, indicating robust and sustained antidepressant responses (Rush et al., 2000). Thus, the VN serves as a vital conduit linking gut physiology with emotional and cognitive processes (Fig. 4).

3.1.3. Hypothalamic-pituitary-adrenal axis

VN is also critically linked with the HPA axis, the third part of the anatomical connections of the gut-brain axis (Fig. 4). Vagal afferent pathways are involved in the activation/regulation of the HPA axis as emerging evidence supports the interplay between the VN, HPA axis regulation, and the gut microbiome (Howland, 2014). In a recent animal study, non-stressed adult male mice that received gut microbiota of stressed animals exhibited depressive-like behavior similar to stressed animals suggesting that gut microbiota of stressed animals can trigger brain changes in control, non-stressed animals leading to depressive bahavior (Siopi et al., 2023). These effects seemed to be mediated by VN activation, as subdiaphragmatic vagotomy abolished stress-induced neuroinflammation, deficits in hippocampal neurogenesis, and alterations in brain serotonin (5-HT) and dopamine pathways. Clinically, VN stimulation therapy in patients with treatment-resistant depression was associated with reduced corticotropin-releasing hormone (CRH)/adrenocorticotropin hormone (ACTH) responses after three months, suggesting that vagal modulation can influence HPA axis activity and stress hormone regulation (O'Keane et al., 2005). Altogether, these findings highlight a complex, bidirectional relationship in which the gut microbiome modulates HPA axis function via VN, thereby influencing stress-related behavioral deficits. In addition to its anatomical connection with the VN, the HPA axis also interacts with the gut microbiome and GI tract through the circulatory system – see Fig. 4. Specifically, HPA axis stimulation produce corticosteroids that reach the GI tract, in which various gut cells express glucocorticoid receptors (GR) (Shukla et al., 2022); upon binding to GR, corticosteroids induce changes in gut microbiota composition and diversity (Petrullo et al., 2022). This intricate interplay between the VN, HPA axis, and gut microbiome underscores the integrated nature of neuroendocrine and microbial influences on stress physiology and mood regulation.

3.2. Gut-brain pathways

3.2.1. Neurotransmitters

The gut microbiota significantly contributes to the production of neurotransmitters within the GI tract. Although these microbiota-derived neurotransmitters primarily exert local effects, they can also influence gut-brain communication indirectly, primarily through activation of vagal afferent fibers or, in some cases, via entry into the bloodstream (Dicks, 2022). Approximately 90% of the body's 5-HT is synthesized in the gut, predominantly by enterochromaffin cells located in the mucosal layer of the GI tract (Yano et al., 2015). The 5-HT released by the gut activates 5-HT3 receptors (5-HT3R) located on vagal afferent fibers, thereby transmitting signals to the nucleus tractus solitarius, and subsequently to other brain areas e.g. amygdala, locus coeruleus, and other cortical structures (Fig. 4). In addition to enterochromaffin cells, several gut-resident bacterial strains (e.g. Streptococcus, Lactobacillus, Klebsiella, Escherichia coli, Lactiplantibacillus plantarum FI8595, Streptococcus thermophilus NCFB2392, and Enterococcus spp) are known to synthesize 5-HT (Cryan and Dinan, 2012; K. Gao et al., 2020). This microbial production of 5-HT further highlights the integral role of the microbiome in modulating serotonergic signaling. Recent work has further highlighted intestinal epithelial 5-HT as a key node in gut–brain communication relevant to mood. Hung and colleagues showed that selectively increase of intestinal epithelial 5-HT signaling via ablation of epithelial serotonin reuptake transporter produced anxiolytic and antidepressant-like effects, whereas inhibition of epithelial 5-HT synthesis increased anxious and depressive-like behavior. These behavioral effects depended on intact vagal afferent pathways, supporting a model in which epithelial-derived 5-HT modulates mood via vagus-mediated gut–brain signaling (Hung et al., 2025). Moreover, GABA, the principal inhibitory neurotransmitter, is shown to be produced by a great variety of bacteria strains, such as Bifidobacterium adolescentis (Duranti et al., 2020), Blautia (Zhuang et al., 2020), Limosilactobacillus fermentum L18 (Kaur et al., 2023), Lactobacillus brevis and Bifidobacterium dentium (Barrett et al., 2012) and others. GABA synthesized in the GI tract is proposed to influence CNS activity, in part via GABA transporters located at the BBB (Takanaga et al., 2001). Similarly, glutamate, the major excitatory neurotransmitter, arises from both dietary sources and microbial synthesis by strains integral to the human gut microbiome (Nakayama et al., 2018). Under normal physiological conditions, dietary glutamate does not readily cross the BBB (Montanari et al., 2022). However, enteric glutamate substantially contributes to gut-brain signaling via the activation of glutamatergic receptors distributed along vagal afferent pathways. Furthermore, glutamate receptors located in the dorsal motor nucleus of the vagus facilitate efferent signaling back to the gut, supporting a bidirectional communication network. Collectively, these findings emphasize that neurotransmitters produced both by host cells and gut microbiota are critical mediators of gut-brain communication.

3.2.2. Microbial metabolites

The gut microbiota, particularly anaerobic phyla, produces a wide array of microbial metabolites, with short-chain fatty acids (SCFAs) being the most abundant. SCFAs are organic molecules synthesized through the fermentation of dietary fibers by specific bacterial populations residing in the GI tract (Mansuy-Aubert and Ravussin, 2023); the main SCFAs are the acetate, butyrate and propionate (Ee et al., 2020). Following their production, SCFAs are primarily absorbed by colonocytes via monocarboxylate transporters (MCTs; abundantly expressed in endothelial cells), which supports the concept that SCFAs are crossing the BBB making their way to the CNS (Fig. 4). This is further supported by the fact that all three metabolites (acetate, propionate and butyrate) are detectable in the cerebrospinal fluid (CSF) (Human Metabolome Database, Accessed 2025.). Within the CNS, SCFAs can influence various cell types, including neurons and microglia. For example, intraperitoneal administration of acetate (500 mg/kg) has been shown to modulate levels of glutamate, glutamine, and GABA in the hypothalamus of male wild-type (C57BL/6) mice (Frost et al., 2014). Moreover, oral supplementation of three SCFAs (25 mM sodium propionate, 40 mM sodium butyrate, and 67.5 mM sodium acetate, for 4 weeks) in germ-free C57BL/6 mice was sufficient to promote microglial maturation in the absence of a conventional microbiota (Erny et al., 2015). These findings highlight the capacity of gut-derived SCFAs to influence brain function, underscoring their key role in gut–brain communication.

3.2.3. Extracellular vesicles

As discussed above, several mechanisms have been proposed to explain bidirectional communication within the gut-brain axis. However, more recently, extracellular vesicles (EVs) secreted by the gut microbiota have emerged as a putative additional pathway for this communication, complementing previously-established routes of gut-brain communication (Cuesta et al., 2021; Sun et al., 2023). EVs are lipid bilayer sacks secreted from all types of cells (eukarya and bacteria) (Ibrahim and Khan, 2024), with great variance upon release pathway, on their size, cargo and also function (Doyle and Wang, 2019). EVs have the ability to mirror the molecular processes of their source cells, attributable to their intracellular origins, their capacity to protect their cargo, and their extended half-lives (Gomes et al., 2023). EVs can be categorized to microvesicles, exosomes and apoptotic bodies while their cargo consists of nucleic acids, lipids, proteins and metabolites. Emerging evidence highlights the potential roles of gut microbiota EVs in the communication of gut-brain axis, mainly through transportation of a variety of neurotransmitters (e.g. dopamine, 5-HT) and SCFAs as part of their cargo (Haas-Neill and Forsythe, 2020) to the blood stream and thus, to the brain – note that EVs and especially small EVs (e.g. exosomes) may cross the BBB and influence brain function (N. Y. Kim et al., 2024).– see also Fig. 4. Interestingly, a recent study by Kim et al. (2024) used a novel chip mimicking the gut-brain axis communication, where the gut chip contained human epithelial Caco-2 cells to establish an intestinal lumen, and the brain chip was cultured with neurospheres. The brain side of the chip was treated with EVs derived from Lactobacillus casei Hy2782 (Hy2782) and Lactobacillus plantarum Hy7714 (Hy7714) for 5 days. This study showed that EVs derived from microbes promoted the maturation of neural stem cells into neurons with Hy2782-derived EVs treatment promoting synapse maturation. Another study demonstrated that EVs from A. muciniphila that were administered to male mice (10 μg of protein/200 μl of EVs, by oral gavage, daily for 4 weeks) led to increased 5-HT hippocampal levels (Yaghoubfar et al., 2020) whereas EVs from Paenalcaligenes homini (EVs containing 10 μg/kg protein, 5 days) that were administered to specific pathogen free (SPF) male and female C57BL/6 mice (by oral gavage) caused cognitive impairment. Notably, this EVs effect was VN-dependent, as vagotomy inhibited not only the cognitive decline but also hindered the infiltration of EVs into the hippocampus (K.-E. Lee et al., 2020). Extending these observations to neurodegenerative context, recent work in App^NL−G-F Alzheimer's disease mice showed that commensal gut microbiota-derived bacterial EVs can exacerbate amyloid-β pathology and microglial activation, thereby shaping AD progression along the gut–brain axis (Xie et al., 2025). Together, these findings position gut-microbiota EVs as an emerging mediator of gut–brain communication, capable of influencing neural maturation, neurotransmission, and cognitive function. Note that, rather than representing a fully established pathway on their own, EVs are best viewed as vehicles that integrate microbial signals (e.g. metabolites, neurotransmitters etc.) into the broader network of gut–brain interactions.

4. Gut microbiome dysregulation in depression and Alzheimer's disease

4.1. The gut microbiome in depression

Disruptions of the gut microbiome, and related dysfunction of the gut-brain axis, have been emerging over the last years as pivotal factor in the establishment and progression of mental disorders such as depression and anxiety (Xiong et al., 2023). On the other hand, the underlying brain pathology associated with depression and anxiety can also cause and enhance gut microbiota alterations via gut environmental changes (L. Liu et al., 2023). In the following section, we summarize gut microbiome alterations and their association with impaired brain function and neuroinflammation in the context of mental disorders.

Clinical evidence: Alterations in the composition of the gut microbiome, regarding microbial diversity and distinct bacterial taxa abundance, have been widely reported in patients suffering from major depressive disorder (MDD) (Nikolova et al., 2021). Notably, an elevated ratio in the phyla Bacteroidetes/Firmicutes, characterized by an increase in the Bacteroides genus and a marked elimination in beneficial genera such as Faecalibacterium, Blautia, and Coprococcus have been observed in MDD patients (Nikolova et al., 2021; L. Liu et al., 2023). A similar microbial pattern has also been found in patients with anxiety and depression, where increased levels of Bacteroides, Proteobacteria and Prevotella/Prevotellaceae are more prevalent compared to healthy individuals (Simpson et al., 2020). On the other hand, patients with generalized anxiety disorder and depression exhibit reduced Subdoligranulum and Coprococcus levels (Valles-Colomer et al., 2019), as well as higher levels of inflammatory markers, TNF-a (Tumour Necrosis Factor-A), IL-4 and IL-10 (Y. Cheng et al., 2022). Except for bacterial composition alterations, different metabolites derived from the gut microbiota, also appear to have a regulatory role in depression. For example, SCFAs (e.g. butyrate, acetate and propionate) are significantly eliminated in MDD patients (Zheng et al., 2016b), while SCFA treatment exerts antidepressant effects (Van De Wouw et al., 2018; Caspani et al., 2019). Acetate acts as a preventative in case of enteropathogenic infections and it maintains gut-brain axis signaling, regulating the integrity of gut barrier. Moreover, reduced acetate levels in depressive patients induce a decrease of butyric acid, which may promote neuroinflammation and depressive symptoms, due to elevated microglia activation (Capuco et al., 2020). Furthermore, propionate may regulate intestinal permeability and its lower levels in depressive patients could contribute to gut microbiome disruptions and neuroinflammation, promoting symptoms of depression (Capuco et al., 2020).

In line with these findings, patients with depressive symptoms frequently exhibit GI inflammation and bowel diseases, and therefore symptoms such as abdominal discomfort, bloating and nausea, possibly due to disrupted gut-brain axis pathways (Capuco et al., 2020). Importantly, chronic stress has been closely linked to GI disturbances, such as irritable bowel syndrome (Marano et al., 2025), further highlighting the complex interplay between the microbiome, gut-brain axis and mental health. On the other hand, recent studies in humans demonstrate that probiotic interventions (e.g Lactobacillus helveticus and B. longum treatment) may mitigate symptoms in patients with MDD and enhance well-being in those with elevated anxiety and depression scores. Moreover, a positive effect of probiotic B. infantis M-63 on mental health has also been reported (Z. F. Ma et al., 2019; Järbrink-Sehgal and Andreasson, 2020).

Preclinical evidence: Different studies in animals also support the aforementioned, human-based link between chronic stress, gut microbiota changes and related mood and cognitive deficits. For example, different stress models in animals have shown that stress-induced depressive-like behavior and deficits of neuroplasticity are accompanied by gut microbiome changes. 8-week-long, chronic mild stress induced depression-like behavior, cognitive impairment and synaptic loss that were accompanied by marked alterations of gut microbiota composition, characterized by elevated abundance of Proteobacteria and increased levels of microbiota-derived lipopolysaccharides (LPS) (Hao, 2024). Moreover, adolescent mice subjected to a 12-day chronic unpredictable stress paradigm displayed long-lasting behavioral deficits and gut microbiome disruption (Lukic et al., 2025). Similarly, 5-week-long restraint stress is also shown to cause gut microbiota dysbiosis followed by reduced neurotransmitter levels, enhanced neuroinflammation in the brain and mood deficits (Yang et al., 2021). On the other hand, gut microbiota direct manipulation is shown to impact brain plasticity and behavior. For instance, microbiota depletion in adult rodents via antibiotic oral administration led to depressive-like behavior, memory deficits and altered brain neurotransmitter levels (Hoban et al., 2016). Note that the above antibiotic treatment also induced microbiota changes with a marked reduction in the phyla Firmicutes and Bacteroidetes, and a significant enhancement in Proteobacteria and Cyanobacteria (Hoban et al., 2016). Moreover, germ-free mice are shown to display hyperactivity of HPA axis, hippocampal dendritic atrophy and amygdala hypertrophy (Luczynski et al., 2016), as well as hippocampal synaptic deficits such as reduced LTP (Darch et al., 2021). Note that similar brain deficits are evoked by exposure to chronic stress and are related to depressive brain pathology (Lopes et al., 2016a).

Further preclinical support of the implication of gut microbiome in the establishment of stress-driven depressive pathology is based on animal studies using fecal microbiota transplantation (FMT). Interestingly, fecal microbiota from chronically stressed mice administered to germ-free or specific-pathogen-free mice evoked cognitive impairment (Hao, 2024), as well as depressive-like behavior to the recipient, non-stressed, control mice (Chevalier et al., 2020; Hao, 2024; Lukic et al., 2025). These behavioral deficits were accompanied by decreased proliferation and neurogenesis in the hippocampus (Chevalier et al., 2020), as well as by disrupted microglia-induced synaptic pruning in the prefrontal cortex (Hao, 2024). In addition, FMT from depressive patients to microbiota-depleted rats induced depressive-like and anxious behavior (Kelly et al., 2016). On the other hand, stressed mice that received FMT from mice previously treated with a potentially antidepressant herbal medicine, exhibited reduced depressive-like behavior accompanied by restoration of gut microbiota composition, and enhanced synaptic function (Hao, 2024). Collectively, the above human and animal evidence highlights the significant interplay of gut microbiota and brain function in relation to stress-related brain pathologies such as anxiety and depression.

4.2. Gut-brain communication in Alzheimer's disease

Several studies have suggested the involvement of gut microbiome in the onset and development of neurodegenerative diseases, particularly AD (Ma et al., 2024). AD is characterized by intracellular aggregates of hyperphosphorylated Tau, known as neurofibrillary tangles, as well as by the extracellular deposition of amyloid-beta (Aβ) leading to the formation of amyloid plaques. Besides these neuropathological features, neuroinflammation, synaptic dysfunction, and progressive cognitive decline constitute important hallmarks of AD. Interestingly, growing evidence supports a bidirectional connection between the gut microbiome and brain pathology in AD (Loh et al., 2024). In this section, main gut microbiota alterations in AD are discussed in association with key AD neuropathological features.

Clinical evidence: Human studies suggest that the composition and diversity of the gut microbiota are altered in AD patients across different clinical stages (Ticinesi et al., 2018; Chandra et al., 2023). Notably, distinct gut microbiome shifts may constitute important indicators of preclinical stage of AD (Ferreiro et al., 2023). For instance, Aβ-positive, dementia-free individuals show a reduction in the phylum Firmicutes and a marked enhancement of Bacteroidetes's abundance compared to Aβ-negative cognitively normal individuals (Sheng et al., 2022). Similar patterns, including reduced Bifdobacterium and Firmicutes levels and enhanced Bacteroidetes have been reported in microbiome of AD patients (Vogt et al., 2017; P. Liu et al., 2019) and individuals suffering from MCI, the prodromal stage of AD (G. Chen et al., 2023). However, it is noteworthy that contradictory findings in relation to distinct microbiota taxa are found across studies and this may be attributed to factors, such as geography, diet, lifestyle, and the disease stage (Chandra et al., 2023; Y. Liang et al., 2024). Moreover, dysregulation of different metabolites derived from the gut microbiota [e.g. SCFAs, LPS and trimethylamine N-oxide (TMAO)] have been also found in AD patients (S. Liu et al., 2020; Marizzoni et al., 2023). SCFAs is shown to participate in protein-to-protein interactions that lead to Aβ formation (Ho et al., 2018). Also, TMAO may enhance the activity of β-secretase, further promoting Αβ generation (Q. Gao et al., 2019) while it is also found to induce microglial activation and related release of inflammatory factors (Brunt et al., 2021). Importantly, in the CSF of both MCI and AD dementia patients, levels of TMAO are increased and exhibit a positive association with neurofilament light chains (NfLs), a marker of active neuronal degeneration (Vogt et al., 2018). Furthermore, factors produced by gram-negative bacteria, such as rhamnolipids or LPS are elevated in CSF/serum of individuals suffering from MCI or AD (Y. Zhao et al., 2017).

Recently, the link between gut microbiome, Aβ brain levels and neuroinflammation has been receiving attention (Dodiya et al., 2022). In particular, patients with brain amyloidosis and cognitive deficits display decreased levels of anti-inflammatory bacteria, such as Eubacterium rectale, and increased levels of pro-inflammatory bacteria including Escherichia/Shigella, compared to healthy or Aβ-negative, cognitively impaired individuals (Cattaneo et al., 2017). These alterations were associated with elevated expression of pro-inflammatory cytokines (e.g. IL-1β, −6, ΝLRP3, CXCL2) and decreased levels of the anti-inflammatory cytokine, IL-10. Finally, regarding clinical therapeutic investigations, a few studies have been performed in AD patients using transplantation of fecal microbiota. For instance, FMT from an age-matched healthy individual to an AD patient led to improved memory function of the latter (Hazan, 2020). Moreover, an AD patient (92-year-old woman) that received FMT from a young, healthy man (27-year-old) exhibited improved cognitive function as well as amelioration in the diversity of gut microbiome (Park et al., 2021). Although these early studies reveal a possible therapeutic effect of FMT in AD patients, further clinical investigation and trials are necessary to evaluate FMT efficacy against AD brain pathology and cognitive impairment.

Preclinical evidence: Different Tg-AD models (e.g. 5XFAD, APP/PS1 and APPPS1-21 mice) are reported to display alterations in gut microbiota in terms of composition, functionality and diversity (Chandra et al., 2023), which have been associated with behavioral and brain neuropathological features. APP/PS1 Tg mice exhibit gut microbiota composition shifts such as elevated abundance of pro-inflammatory genera (Helicobacter, Odoribacter, Sutterella) and the Erysipelotrichaceae family, along with decreased Prevotella levels; note that some of these alterations are associated with Aβ plaque load in hippocampus and related memory impairment (Shen et al., 2017; Bäuerl et al., 2018). In addition, elevated Bacteroidetes and Tenericutes and reduced Firmicutes and Proteobacteria have been reported in fecal microbiota of these Tg mice compared to wild-type counterparts (Harach et al., 2017). Beyond changes in microbiota taxa, gut microbiota-derived metabolites have also been implicated in the pathogenesis of AD. Notably, levels of TMAO were increased in the plasma of APP/PS1 mice and this elevation was correlated with Αβ load and cognitive deficits (Q. Gao et al., 2019). Furthermore, Zang et al. demonstrated a decrease in the levels of SCFAs (isobutyric and butyric acids) in feces and brain of APP^swe/PS1^ΔE9 Tg mice, with possible negative effects on the deposition of amyloid in both gut and brain (L. Zhang et al., 2017).

Interestingly, besides its brain deposition, Aβ is also accumulated in the gut of AD Tg mouse models; this is related to gut dysbiosis and inflammation (Jin et al., 2023). Previous studies suggest a clear impact of gut microbiome in AD brain pathology. For instance, compared to conventionally grown APP Tg mice, germ-free APP mice exhibited reduced neuroinflammation, decreased Aβ42 levels and amyloid plaques (Harach et al., 2017). Similarly, antibiotic administration altered gut microbiota composition of APP_SWE/PS1_ΔE9 mice and reduced Aβ deposition and glial reactivity, further highlighting the influence of gut microbiota on Aβ brain pathology and neuroinflammation (Minter et al., 2016). On the other hand, colonization of germ-free APP mice with microbiota from conventional APP mice led to increased Αβ pathology, possibly via alteration of neuroinflammation levels (Harach et al., 2017). In line with inflammation, it is important to mention that gut dysbiosis enhances intestinal epithelium permeability and can facilitate the entrance of inflammatory cytokines and gut-derived metabolites into the systemic circulation (Cryan et al., 2019). In parallel, alterations and damage of BBB due to gut dysbiosis have been reported, allowing circulating factors to assess the brain and accelerating AD pathology (Uchida et al., 2023). Further support of the impact gut microbiome in AD brain pathology is based on microbiome-associated therapeutic strategies tested in Tg mice. Kim and colleagues showed that FMT from healthy wild-type mice to AD Tg littermates decreased cognitive deficits, Αβ and phospho-Tau levels accompanied by reduced reactivity of glial cells (M.-S. Kim et al., 2020). Furthermore, FMT from wild-type mice to 6-month-old APPswe/PS1dE9 Tg model, restored gut microbiota and SCFA changes in parallel to reduced levels of Aβ40, Aβ42 and p-Tau and enhanced synaptic markers (J. Sun et al., 2019). However, in another study, FMT in antibiotic-administered APPPS1-21 mice enhanced microglial activation and Aβ deposition (Dodiya et al., 2022) providing further support about the beneficial role of FMT in AD brain pathology. Despite the above evidence suggesting an interplay between gut dysbiosis and AD brain pathology, future preclinical and clinical studies are needed to clarify the etiopathogenic contribution of the gut microbiome in AD brain pathology.

4.3. Confounders and methodological heterogeneity in gut–brain-axis research

The insufficient control of confounding factors that affect gut-brain axis, such as medication use and dietary profiles, as well as methodological microbiome heterogeneity, represent major limitations across clinical studies (Cussotto et al., 2019; McGuinness et al., 2022; Chandra et al., 2023; Michaelis et al., 2024; Hisamatsu et al., 2025). For instance, psychotropic drugs constitute an important regulating factor of gut-brain axis. Despite its main role in CNS, psychotropic medication, particularly antipsychotics, exhibit antimicrobial properties and have been linked to dose-dependent reduction in gut microbiome alpha diversity among patients with anxiety and depression (Tomizawa et al., 2021). They have also been reported to induce alterations in spore-forming taxa involved in serotonin regulation. This drug-mediated antimicrobial activity potentially contributes to both drug therapeutic effects and metabolic side outcomes, such as weight gain (Michaelis et al., 2024). Furthermore, treatment with selective-serotonin reuptake inhibitors (SSRIs) antidepressants provokes specific taxa alterations depended on the studied population, such as a positive association with Eubacterium ramulus abundance and a negative correlation with Turicibacteraceae family levels, in Dutch and TwinsUK cohorts, respectively (Jackson et al., 2018; Vich Vila et al., 2020). Moreover, drugs used in AD such as acetylcholinesterase inhibitors have been linked to pronounced compositional shifts in the salivary microbiome compared to the gut one in AD patients. Notably, significant general alterations in the microbiome have been detected which include Filifactor, Granulicatella, Haemophilus and Lactobacillus, suggesting that drug effects may partially influence microbiome alterations in AD pathogenesis (Hisamatsu et al., 2025). Dietary patterns also represent substantial confounders in the reported research of gut-brain axis under pathological conditions, as they can affect microbiome composition, and may influence behavioral-cognitive status (Ghosh et al., 2020; Clerici et al., 2025). Western dietary habits induce enhanced Firmicutes/Bacteroidetes ratio, decreased microbial diversity and increased inflammation, contributing to a higher depressive and obesity risk (Clerici et al., 2025). Furthermore, mediterranean diet has been associated not only with improved cognitive status in older individuals and reduced AD risk (Andreu-Reinón et al., 2021), but also with gut microbial shifts, such as enhanced appearance of bacteria that produce butyrate, including Roseburia and F. prausnitzii (Ghosh et al., 2020). Moreover, methodological heterogeneity constitutes another important parameter of variability across clinical studies focused on microbiome-related alterations. In a systematic review for mental disorders, McGuinness and colleagues (2022) reported heterogeneity in study design, methods of sample collection and storage, sequencing methodologies (e.g. 16S rRNA sequencing versus metagenomics) and bioinformatic analysis, all of which may contribute to inconsistent findings for microbial diversity across cohorts (McGuinness et al., 2022). In addition, although differences in microbial diversity have been found between patients with depressive disorder and healthy individuals, methodological procedures, including microbiome measurement and analysis with different pipelines and datasets, as well as classification of patients, have been shown to influence research data (M. Gao et al., 2023). In conclusion, given the rapid microbiome research growth and the absence of standard protocols for microbial analysis, methodological confounders remain an important challenge (McGuinness et al., 2022). Moreover, medication use and dietary patterns constitute also important confounding factors that need to be controlled. Consideration of all these limitations is essential to improve the interpretation and reproducibility of findings regarding microbiome composition alterations in depression and AD.

5. The implication of epigenetic processes in the gut-brain axis

In light of their genome-wide effect as well as their trans-generational heritability and reversibility, epigenetic modifications are increasingly recognized as key contributors to disease onset and progression. Most importantly, their sensitivity to environmental factors (e.g. psychological stress) highlights their role as mediators of gene-environment interactions in pathological processes, particularly in mental health disorders (Yuan et al., 2023). Given that gut microbiota is increasingly recognized as an influential environmental factor (Bull and Plummer, 2014), the dynamic interplay between epigenetics and gut microbiomes has lately garnered substantial interest (Gutierrez-Angulo et al., 2023; Reva et al., 2023). Below, we discussed the emerging connection of epigenetic modifications with the gut-brain axis.

DNA methylation: As an epigenetic alteration involving the transfer of a methyl group to the C5 position of cytosine, DNA methylation typically leads to the suppression of gene expression (Moore et al., 2013) and is catalyzed by DNA methyltransferases, which utilize S-adenosine methionine as a universal methyl donor. Emerging evidence highlights the influence of gut microbiome on DNA methylation. For instance, S-adenosine methionine can be produced as a metabolic byproduct of Blautia wexlerae (Hosomi et al., 2022). Furthermore, other metabolites such as folic acid, choline and SCFAs produced by several gut microbial species can directly supplement S-adenosine methionine with the methyl group needed for the methylation process (Wu et al., 2021; Kopczyńska and Kowalczyk, 2024). Given that gut microbiome can influence the availability of these methyl donors, it is possible that DNA methylation patterns can be affected in cases where dysbiosis is found (Begum et al., 2022). A study by Romano and colleagues highlighted the importance of choline in microbiome-driven epigenetic reprogramming (Romano et al., 2017). The use of a gnotobiotic mouse model harboring choline-utilizing gut bacteria showed that the reduced choline bioavailability in these animals triggered a compensatory increase in global DNA methylation (Romano et al., 2017). This was observed in mice subjected to a standard diet. Conversely, under chronic high fat dietary conditions, these mice showcased a decrease in DNA methylation in several tissues including brain and colon (Romano et al., 2017). These findings emphasize the complex crosstalk that exists between epigenetics, metabolism and gut microbiota. Interestingly, this mouse model significantly engaged in maladaptive behaviors such as barbering and infanticide, both of which have been related to anxiety or depressive-like behavior (Romano et al., 2017). Moreover, the transgenerational heritability of these epigenetic changes induced by gut microbiota and their effect on behavior were also confirmed as the offspring of mothers suffering from bacterially-induced methyl-donor deficiency exhibited a high microbial metabolism of choline, reduced DNA methylation within the brain and anxious behavior by 8 weeks of age (Romano et al., 2017). Moreover, administration of SCFAs (e.g. sodium butyrate) in a genetic rat model of depression led to an increase in the levels of ten-eleven translocation methylcytosine dioxygenase 1 (TET1), a demethylation-facilitating enzyme. This upregulation was coupled with a decrease in the methylation of neurotrophic factor BDNF and an antidepressant-like effect (Y. Wei et al., 2015). Furthermore, several studies have also shown that germ-free mice have lower hippocampal and cortical BDNF levels (Sudo et al., 2004; Clarke et al., 2013) accompanied by higher hormonal response to stress (Sudo et al., 2004). Note that methylation of BDNF gene has been reported to be altered, accompanied by lower BDNF levels in depression, anxiety and bipolar disorder (D'Addario et al., 2012; Bakusic et al., 2020).

Histone modification: Another epigenetic mechanism controlling gene expression is based on different histone posttranslational modifications (e.g. acetylation, methylation) which cause condensation or unwinding of heterochromatin and consequently trigger transcriptional activation or repression. SCFA-producing bacteria, particularly those associated with butyrate production, are frequently found to be reduced in individuals with depression (Zheng et al., 2016b; J. Cheng et al., 2024). Sodium butyrate, a well-characterized SFCA that functions as a histone deacetylase (HDAC) inhibitor (Davie, 2003), has demonstrated antidepressant properties in an LPS-induced mouse model of depression. Its administration attenuated microglial activation and exerted antidepressant effects all through upregulating the acetylation of histones H3 and H4 within the hippocampus (Yamawaki et al., 2018). Furthermore, probiotic treatment effectively restored gut microbiota composition in rats exposed to potent neurotoxin lead and consequently reversed lead-induced long-term memory impairment. This outcome was found to be associated with a reduction in hippocampal trimethylation of histone H3 at lysine 27 (H3K27me3) (Xiao et al., 2020).

Non-coding RNAs: Non-coding RNAs (ncRNAs) are a class of RNA molecules that do not encode any proteins but they appear to play essential roles in various biological processes by fulfilling housekeeping and regulatory functions. ncRNAs regulate epigenetic processes by recruiting chromatin-modifying enzymes and complexes which mediate processes like DNA methylation and histone modifications which in its turn, regulate gene expression. Recent advances have been made in elucidating the bidirectional interaction between the gut microbiota and ncRNAs (L. Liang et al., 2015; S. Liu et al., 2016) and specifically, their implication in the gut-brain axis (Moloney et al., 2019). A study conducted in germ-free mice reported that changes of miRNAs (a subclass of scRNAs) have been found in the amygdala and the prefrontal cortex, two regions that are known to be implicated in depression and anxiety (Hoban et al., 2017). Additionally, individuals suffering from MCI exhibited a significant reduction in microbial diversity and a decrease in serum levels of hsa-miR-107, hsa-let-7g-5p and hsa-miR-186-3p (X. Zhang et al., 2021); note that these miRNAs have been previously proven to regulate the gene expression of enzymes implicated in Aβ production (J. Kim et al., 2016; L. Sun et al., 2017; Wang et al., 2019). On the other hand, germ-free mice displayed reduced anxiety levels accompanied by increase of 17 miRNAs and a decrease of 33 miRNAs that are involved in metabolic processes and axon guidance, suggesting that these molecular changes may contribute to altered hippocampal function of the germ-free mice (J.-J. Chen et al., 2017). Nonetheless, microbial colonization of these mice was insufficient to fully reverse the behavioral abnormalities observed, underscoring the intricate nature of gut-brain axis regulation (J.-J. Chen et al., 2017). Collectively, DNA methylation, histone modifications, and ncRNAs represent a set of mechanisms through which the gut microbiome may affect brain function and behavior via the modulation of gene expression. Further studies are necessary to clarify the role of gut microbiome on epigenetic profile of brain and the precipitation of brain pathology.

6. Autophagy as a key regulator of the gut-brain homeostasis and pathology

Autophagy is a conserved lysosome-dependent process that degrades and recycles cytoplasmic components to maintain cellular homeostasis, especially under cellular stress (Ryter et al., 2019). In the intestine, autophagy is essential for preserving epithelial barrier integrity, controlling microbial balance, and modulating immune responses, with defective autophagy contributing to dysbiosis and chronic gut inflammation (Ryter et al., 2019; Larabi et al., 2020). More specifically, autophagy in peripheral tissues, especially the gut, can shape the microbial environment, and in turn, microbial signals can modulate autophagy in the gut. Several studies have shown that deleting key autophagy genes (e.g. Atg5 and Atg7) in intestinal epithelial cells disrupts gut microbial balance. Particularly, intestinal epithelial cell–specific Atg5-KO mice exhibit reduced microbial diversity as seen by depletion of anti-inflammatory taxa (e.g. Lachnospiraceae, Ruminococcaceae, and Akkermansia muciniphila) accompanied by expansion of pro-inflammatory bacteria such as Pasteurellaceae and segmented filamentous bacteria that promote Th1/Th17 differentiation (Yang et al., 2018). Similarly, colonic intestinal epithelial cell-specific Atg7-deficient mice show enrichment of taxa such as Clostridium leptum, Eubacterium cylindroides, and Bacteroides fragilis (Tsuboi et al., 2015). Recent findings have also stated that mice harbouring the ATG16L1T300A variant in ATG16L1, a key component of the autophagy machinery, display significant alterations in gut microbial composition both at normal conditions and during colitis induction (Lavoie et al., 2019). This further corroborates the role of normal autophagic activity in maintaining a healthy microbiome and immune homeostasis. Conversely, gut microbes produce metabolites that directly influence autophagy including SCFAs that can activate the AMPK–mTOR-driven autophagic pathway (Fan et al., 2020). This indicates an active bidirectional communication between the intestinal microbiome and autophagic activity suggesting that disturbances in autophagy might also exacerbate gut-brain axis dysfunction and vice-versa.

Besides its role in gut, autophagy and its dysregulation in the brain is also involved in depression as evidenced by discrepancies in autophagy-related markers in patients with MDD (He et al., 2023; Lu et al., 2023). Consistently, chronic stress has also been repeatedly shown to impair neuronal autophagic flux (Karoussiotis et al., 2022; Zhang et al., 2023; Yang et al., 2025). Changes in the autophagic process of the brain might be modulated by gut microbiota and their metabolites as recent work indicates that the development of depression is closely tied to autophagy mechanisms within the gut-brain axis. A study by Ma and colleagues has reported that supplementation of chronically stressed mice with Lactobacillus plantarum as a probiotic improved gut microbiota balance and enhanced autophagic clearance in hippocampus, reducing anxiety and compulsive-like behavior (Ma et al., 2023). These findings suggest a possible link between gut microbial balance and autophagy in alleviating depressive behavior and related neuronal malfunction (Ma et al., 2023) as autophagy may be responsible for synaptic atrophy/loss under stress (Karoussiotis et al., 2022). Moreover, autophagy dysfunction is suggested to lead to accumulation of Αβ and Tau aggregates in AD brain (Orr and Oddo, 2013; Zhang et al., 2022) while chronic stress in Tau Tg mice is shown to inhibit autophagy leading to neurotoxic tau aggregates and related neurodegeneration and memory deficits (Silva et al., 2019). Furthermore, in a 3xTg-AD mouse model, the administration of SLAB51, a multi-strain probiotic, resulted in the activation of the autophagy regulator, SIRT1, in neurons, leading to increased neuronal autophagy, and a significant reduction in Aβ plaque deposition (Bonfili et al., 2018).

Altogether, these findings highlight autophagy as a central component of the gut-brain communication acting within a dynamic triad to regulate immune signaling, microbial balance, and neuronal (mal)function. As disruption of this tightly coordinated network may significantly contribute to the deterioration of brain health and the development of brain pathology, future studies are warranted to elucidate the cell-specific pathways in both gut and brain and explore their therapeutic potential.

7. Modern lifestyle and nutrition as key modulators of the gut microbiome and its relation to brain health

As modern society and contemporary lifestyle increasingly expose people to fast-paced, demanding rhythms, leading to higher levels of psychological stress in everyday life, there is an increasing interest of people in adopting lifestyle changes to promote mental health and improve their overall well-being. Basic changes such as increasing physical activity, following a healthy diet, consuming pro-, pre- or postbiotics have been found to reduce stress as well as enhance gut health improving mood and cognitive performance (Mee-inta et al., 2019; Donoso et al., 2023) (Fig. 5). Below, we will discuss different lifestyle parameters and the underlying mechanisms through which they modulate the gut microbiome and the brain.

Fig. 5.

Lifestyle and dietary interventions impact on gut-brain axis. A schematic representation of key parameters that modulate the gut–brain axis and thus, impact on both brain and gastrointestinal health: Different types of physical exercise, nutrition such as diverse whole foods, fibers, and micronutrients), pro-, pre-, and post-biotics supplementation as well as trace elements such as lithium and silicon are suggested to promote healthy gut microbiome and influence brain function. These lifestyle and dietary interventions represent promising, non-invasive strategies for promoting brain and gut health.

7.1. The beneficial effects of exercise on gut-brain axis and its implications on mental health

Exercise has been found not only to enhance physical fitness but also impact brain structure and function; however, the exact mechanisms remain unclear. Physical activity is well known to promote brain plasticity and cognitive function by upregulating different neurotrophic factors (e.g. BDNF), particularly in brain regions involved in mood and memory such as hippocampus (De la Rosa et al., 2019). For instance, elderly individuals who engaged in regular exercise for one year showed a notable rise in serum BDNF levels, along with increased hippocampal volume compared to those who remained sedentary. In addition, physical activity is well known for its’ systemic anti-inflammatory effects, specifically for reducing chronic, low-grade inflammation, which is typically linked to mental health disorders such as depression and anxiety (Osimo et al., 2019). Clinical studies have demonstrated that increased physical activity correlates with reduced circulating levels of pro-inflammatory cytokines, such as IL-6 and TNF-α, in both healthy individuals and those suffering by anxiety or depression (Lavebratt et al., 2017; Paolucci et al., 2018; Fernandes et al., 2022). Preclinical studies support a role of microbiota in the beneficial role of physical expertise as transferring microbiota from exercise-trained mice to diabetic mouse model led to a significant reduction of inflammatory cytokines such as IL-1β, IL-6, and TNF-α (Cheng et al., 2025). Indeed, recent studies in humans and rodents also indicate that exercise also promotes microbial diversity increasing the abundance of beneficial microbial taxa such as Akkermansia muciniphila and Faecalibacterium prausnitzii (Allen et al., 2018; Barton et al., 2018; Hintikka et al., 2023; Boytar et al., 2023). Note that these bacterial populations are essential for preserving gut integrity through the beneficial taxa that contribute to the production of SCFAs while it enhances intestinal barrier function and reduces systemic inflammation which may support overall brain health (Effendi et al., 2022). Additionally, aerobic exercise increases synthesis and metabolism of the neurotransmitter 5-HT in the hippocampus, thereby alleviating symptoms of depression and anxiety (Pietrelli et al., 2018). As mentioned above, certain gut bacteria, like Lactococcus lactis and Lactobacillus plantarum, contribute to 5-HT production (Strandwitz, 2018). Although exercise may enhance these microbiome strains through increased gut microbial diversity, further research is needed to confirm whether exercise directly influences 5-HT-producing bacteria and how this promotes mental health (Dalton et al., 2019).

However, the aforementioned beneficial effects of exercise on the gut microbiome depend on the type and intensity of physical activity as well as its combined nutrition. For example, individuals with higher cardiorespiratory fitness levels tend to exhibit a microbial profile that promotes butyrate production, a SCFA known for its neuroprotective effects, in contrast to those with lower or non-aerobic fitness levels (Estaki et al., 2016; Barton et al., 2018). Moreover, differences in microbial activity and fecal SCFAs levels have also been observed among athletes with varying fitness levels. For example, athletes participating in sports with a high dynamic and low static component showed an enrichment of species such as Bifidobacterium, Lactobacillus, Prevotella, and Faecalibacterium in their gut microbiota (O'Donovan et al., 2020). Moreover, the type of physical activity, along with specific nutritional demands of different athlete groups, appears to influence these variations. For instance, a diet high in protein and fat with low carbohydrate intake has been negatively associated with SCFA levels (Jang et al., 2019; Son et al., 2020).

7.2. Potential mechanisms underlying the influence of nutrition on the gut-brain axis

Modifications in dietary intake are increasingly recognized as critical factors influencing gut microbiome composition, which in turn may affect brain and mental health (Schneider et al., 2024). The bidirectional gut-brain connectivity can be modulated by macronutrient and micronutrient uptake. Consequently, changes in dietary habits can be crucial for maintaining or disrupting the gut-brain communication, therefore affecting brain function through microbial metabolites, neuroactive compound production (e.g., neurotransmitters), and immunoregulatory and endocrine pathways (Horn et al., 2022).

More specifically, high-fiber diets, such as the Mediterranean diet characterized by high consumption of fruits, vegetables and whole grain products are associated with decreased anxiety levels and decreased risk for depression (Saghafian et al., 2021). A large population-based cohort study involving 49,261 Swedish women aged 29-49 found that higher adherence to a Mediterranean dietary pattern (MDP) was associated with a lower risk of developing depression over a 20.4-year follow-up period. Specifically, medium adherence to MDP resulted in a 10% risk reduction, while high adherence led to an 18% reduction. This association was even stronger for severe forms of depression, with high adherence linked to a higher risk reduction (Yin et al., 2021). Similarly, a randomized clinical trial in overweight women with type 2 diabetes found that 12 weeks of a calorie-restricted mediterranean-DASH intervention for neurodegenerative delay (MIND) diet, rich in fruits, vegetables, whole grains and antioxidants, significantly improved depression scores, sleep quality, and crucially raised serum BDNF and decreased serum levels of cortisol (the main stress hormone), compared to a standard low-calorie control diet (Golmohammadi et al., 2025). Additionally, a cross-sectional study involving 4957 individuals with obesity found that higher adherence to the Mediterranean diet was associated with a lower prevalence of both anxiety and depression (Menichetti et al., 2025). Preclinical studies using APP/PS1 mice showed that a high-fiber Mediterranean–ketogenic diet significantly reduced pro-inflammatory cytokine expression (IL-1β, IL-6, TNF-α) in the colon as well in the brain areas such as hippocampus and hypothalamus. Furthermore, it promoted the growth of Lactobacillus and enhanced bacteria-derived lactate production, glutamate metabolism, and GABA biosynthesis suggesting microbiota-mediated anti-inflammatory and neuromodulatory effects compared to those fed with a Western-style diet (Park et al., 2024) – see Fig. 6.

Fig. 6.

The impact of mediterranean and western diets on gut microbiome homeostasis. The Mediterranean diet that includes polyphenols, healthy fats and complex carbohydrates and is rich in fibers promote eubiosis of gut microbiome increasing its diversity. On the other hand, western diet which is high in saturated fats, refined sugars and processed food is related to gut microbiome dysbiosis.

This beneficial effect of the Mediterranean diet could be based on its high polyphenol content (Yammine et al., 2021). Numerous in vitro studies depict the regulatory effect of polyphenols in modulating the intestinal micro-organisms. Polyphenols found in grape seeds and grape derivatives such as red wine, exhibited an inhibitory effect on the growth of pathogenic bacteria such as Clostridium histolyticum in gut models of microbiome derived from human fecal samples (Kemperman et al., 2013; Zhou et al., 2016). Conversely, polyphenols such as tannins in pomegranate, gingerol in ginger, and sorghum polyphenols can promote the growth of the beneficial Bifidobacterium, while tannins also support Lactobacillus populations (Wang et al., 2020; Wu et al., 2021; Sost et al., 2021). In line to the above in vitro evidence, in vivo studies showed that the application of a diet of chestnut extract (rich in tannin) to a zebrafish intestinal inflammation model, promoted healthy bacteria in zebrafish (Enterobacteriaceae and Pseudomonas). Also, a diet rich in cranberry polyphenols lowered the increased Firmicutes/Bacteroides ratio in mice subjected to a high-fat/high-sugar diet (Anhê et al., 2015; Orso et al., 2021). While polyphenols help regulate gut bacteria by promoting beneficial strains and inhibiting harmful ones, fiber is the major substrate for fermentation, leading to the production of SCFAs. A high-fiber diet supports the production of SCFAs by promoting a diverse gut microbiome, as it encourages microbes to break down complex polysaccharides into simpler compounds, which are then metabolized into SCFAs (H. Liu et al., 2022). Despite its importance, fiber intake falls short worldwide, averaging 15–26gr/day instead of the recommended 25–35gr/day, due to the rise of western diets (Gill et al., 2021; Fu et al., 2022).

Another important mechanism by which dietary fiber impacts metabolic health is through the regulation of bile acid (BA). Fiber affects BA metabolism both directly, by reducing BA re-absorption in the intestine and indirectly, by modulating gut microbiota, which alters BA synthesis and facilitates the conversion of primary BAs into secondary BAs (Singh et al., 2019; Pezzali et al., 2021). BAs are cholesterol derivatives produced by the liver, primarily in response to dietary intake, and play a crucial role in lipid digestion and absorption, particularly for dietary fats and fat-soluble vitamins. Beyond their digestive function, BAs are also involved in metabolic, endocrine, and inflammatory regulation (Monteiro-Cardoso et al., 2021; Fleishman & Kumar, 2024). It is also worth noting that a high-fiber diet can reduce the re-absorption of BAs back into the bloodstream, leading to an increased excretion in the feces and an upregulation in the BAs production, to restore the balance (Pezzali et al., 2021). Changes in excretion and biliary BAs levels suggest that dietary fiber affects BA metabolism (Naumann et al., 2019). The BAs pool can be manipulated by the gut microbiota as they convert primary synthetized BAs to secondary BAs (Collins et al., 2023; Gao et al., 2024). Emerging evidence suggests that BAs and their receptors are also present in the brain, implicating them in gut-brain communication. Notably, BAs regulate N-methyl-D-aspartate receptor (NMDA) and GABA receptors, which are essential for synaptic plasticity and neurotransmission. This interaction suggests a potential role for BAs in cognitive function and mental health, although further research is needed to elucidate the precise mechanisms (Monteiro-Cardoso et al., 2021).

In contrast, diets high in saturated fats, refined sugars and processed foods such as the Western diet, is usually correlated with an imbalance of microbial composition leading to a reduction of the beneficial bacteria, dysbiosis, gut permeability and leakage of toxic bacterial metabolites and derivates into systemic circulation (referred as “leaky gut syndrome”) (Malesza et al., 2021). This alteration of gut microbiome diversity can also lead to decreased SCFAs production (Agus et al., 2016). Another important parameter is the high concentration of bacterial-derived LPS in the blood circulation. High fat diets are associated with elevated levels of LPS, can trigger systemic inflammation and neuroinflammatory responses, along with disrupting gut-brain communication (Tucureanu et al., 2017). Note that in individuals diagnosed with MDD, LPS-induced inflammatory markers were significantly elevated and associated with specific depressive symptoms and anxiety levels (van Eeden et al., 2020). Indeed, LPS administration in animals causes cognitive deficits, increased anxiety levels and depressive-like behavior (Zhao et al., 2019). This may be due to the ability of LPS, found in the periphery and brain, to promote inflammatory response, acting on leukocyte and microglial TLR4-CD14/TLR2 receptors to produce NFkB- mediated cytokines increase. Note that increased inflammatory response influences Aβ generation exacerbating amyloid brain pathology (Brown and Heneka, 2024).

7.3. Probiotics, postbiotics and prebiotics as modulators of gut microbiota and mental health

Probiotics are beneficial live microorganisms that when administered in adequate amounts, enrich the gut microbiota and provide health benefits to the host organism (Hill et al., 2014). Lactobacillus and Bifidobacterium are the most commercialized probiotic bacteria genera, both of which are classified as gram-positive lactic acid-producing bacteria (Vlasova et al., 2016). This exogenous enrichment of beneficial bacterial populations can help in the restoration of the intestinal microbial balance, if it is disturbed (Xiong et al., 2023; Zhang et al., 2025). Probiotics can also serve as a resource for other beneficial bacteria to grow (e.g. Firmicutes phylum) and finally lead to increased levels of SCFAs, tryptophan, GABA, 5-HT, BDNF, and dopamine (Moens et al., 2019). Additionally, probiotics appear to support gut barrier integrity, preventing the leakage of toxic bacterial metabolites and derivatives into systemic circulation (Candela et al., 2008). This, in turn, helps protect both the gut and brain from inflammatory conditions (Virk et al., 2024) (Fig. 5).

Preclinical evidence suggests that probiotic supplementation during maternal separation stress can reduce anxiety levels in the infant rats (Peng et al., 2019). Supplementation of male mice with Lactobacillus plantarum MTCC 9510 (2 × 10^10 CFU per mouse) under chronic stress or sleep deprivation conditions successfully alleviated stress-related anxiety, depressive-like behavior and memory deficits accompanied by lower oxidative stress markers and inflammatory cytokines in the brain and serum (Dhaliwal et al., 2018). In clinical studies, probiotic administration has been associated with reduced symptoms of depression and anxiety despite some contradictory findings being reported. A meta-analysis reviewing 10 clinical trials with 1349 patients, found no significant overall difference in mood between probiotics and placebo, post-intervention. However, a subgroup analysis revealed significant improvements in individuals with mild to moderate depression, while no effects were observed in healthy individuals (Ng et al., 2018). In contrast, another meta-analysis of randomized controlled trials investigating the effect of probiotics on anxiety found no significant difference between probiotics and placebo in reducing anxiety symptoms (Liu et al., 2018). Furthermore, a random-effects meta-analysis of 34 controlled clinical trials comparing prebiotics and probiotics showed that, unlike prebiotics, probiotics had small but significant beneficial effects against depression and anxiety (Liu et al., 2019).

In parallel with probiotics, postbiotics are defined as non-viable microbial cells and/or their components that exert beneficial impact on the host (Salminen et al., 2021). They may include substances like SCFAs, enzymes, cell wall fragments, and metabolites such as lactic acid and vitamins, with many of those displaying immunomodulatory, anti-inflammatory, antioxidant, and neuroactive properties, in the same manner as probiotics (Chudzik et al., 2021; Salminen et al., 2021). Unlike probiotics, postbiotics do not require microbial viability to exert their physiological effects, offering distinct advantages in terms of safety, stability during processing and storage, and ease of standardization (Salminen et al., 2021). Despite these findings, the evidence for probiotics' efficacy in anxiety relief remains insufficient, highlighting the need for further clinical trials. Overall, the administration of probiotics supports the function of the gut and the brain and the emerging concept of “psychobiotics” underscores their role as modulators of mood, anxiety, and cognition (Kumar et al., 2024).

In the context of dietary supplementation, prebiotics are non-digestible compounds that resist gastric acidity and enzymatic breakdown in the human GI tract. They serve as a selective energy source for beneficial gut bacteria and thereby, they modulate microbiota composition and promote the growth of health-associated taxa (Davani-Davari et al., 2019). Common categories of prebiotics include fructans, like inulin and fructooligosaccharides (FOS) or oligofructose, galactooligosaccharides (GOS), starch (resistant starch) and glucose-derived oligosaccharides but also other types of oligosaccharides (Davani-Davari et al., 2019). Prebiotics are mentioned for their beneficial impact of improving gut barrier function, reducing inflammation, mitigating dysbiosis and supporting the production of SCFAs. Indeed, FOS + GOS administration in chronically stressed mice or high-fat diet fed mice improved behavioral outcomes and reduced peripheral and central inflammation by promoting the growth of acetate-producing bacteria (B. acidifaciens and B. dorei) and improving gut barrier integrity (Burokas et al., 2017; Paiva et al., 2024). However, clinical studies investigating prebiotic supplementation have failed to show significant antidepressant efficacy compared to placebo (Kazemi et al., 2019a; 2019b; Tarutani et al., 2022).

8. The neuroprotective and microbiome-modulating effects of lithium and silicon

Besides diet, drinking water is another source through which our organism receives essential minerals and trace elements to support vital physiological functions. Lithium and silicon, trace elements present in drinking water and diet, are shown to influence both gut microbome and brain (Melenikioti et al., 2026). For example, previous studies suggest that long-term, low-dose exposure to lithium in drinking water may correlate with lower dementia rates, reduced suicide mortality, and improved cognitive health (Brown et al., 2018; Giotakos et al., 2013; Tondo and Baldessarini, 2024) while a very recent study suggested that low lithium levels predisposes to AD (Aron et al., 2025).

Indeed, lithium is commonly prescribed as a mood stabilizer (Szałach et al., 2023) with various clinical applications in acute mania, bipolar disorder and MDD alone or in combination to other drugs (Keramatian et al., 2023; Nuñez et al., 2022). Furthermore, epidemiological studies demonstrate lithium treatment reduces suicidal ideation in patients suffering from affective disorders such as bipolar disorder and MDD (Tondo and Baldessarini, 2024). In clinical practice, Li doses usually range from 600 mg to 1800 mg per day, with the specific dose adjusted according to therapeutic response, side effects, and serum lithium levels (Liaugaudaite et al., 2017). Notably, even its subtherapeutic levels, such as those found in drinking water with high Li concentrations, appear to provide significant benefits, including advantages in the management of affective disorders, suicide rate (particularly in men), violent behaviors and dementia (Brown et al., 2018; Giotakos et al., 2013; Tondo and Baldessarini, 2024). Mechanistically, lithium has been also known for its neurotrophic and neuroprotective properties as it ιs shown to enhance the expression and secretion of neurotrophic factors such as BDNF and vascular endothelial growth factor (VEGF), both crucial for neuronal resilience, hippocampal neurogenesis, and synaptic remodeling (Forlenza et al., 2014). Consistent with these effects, lithium treatment in individuals diagnosed with bipolar disorder, initiated at 600 mg/day and adjusted according to blood concentrations (targeting 0.6–1.2 mmol/L) for 28 days, has been associated with increased plasma levels of BDNF (de Sousa et al., 2011). Moreover, in a clinical pilot study, its use in AD patients led to significant increases in serum BDNF levels and notable cognitive improvement, in contrast to patients in the placebo group (de Leyhe et al., 2009). Another mechanism underlying these effects involves the lithium-evoked inhibition of glycogen synthase kinase-3β (GSK-3β) (Stambolic et al., 1996), a kinase whose activity is increased in the early stages of MCI and AD (Forlenza et al., 2011), and which contributes to both Aβ plaque formation and Tau hyperphosphorylation (Lauretti et al., 2020). Notably, a very recent study shows that lithium is reduced in the brain of individuals with MCI, a precursor of AD, as well as of AD patients (Aron et al., 2025) which is in line with a previous population study in Denmark that found a significant inverse correlation between lithium levels in local drinking water and the incidence of dementia (Kessing et al., 2017). In line to the above clinical evidence, recent experimental evidence in two different AD Tg mouse models demonstrated that lithium-deficient diet led to increased Aβ and Tau brain pathology and related synaptic loss as well as enhanced microglia-activation and increased pro-inflammatory cytokine. Importantly, lithium supplementation back to physiological levels in AD Tg mice under lithium-deficient diet attenuated these brain deficits (Aron et al., 2025).

While less studied than lithium, silicon is also linked to lower risk of dementia, particularly of AD, based on clinical studies (Rondeau et al., 2000, 2009). Moreover, preclinical studies have shown that low silicon concentrations protect human SH-SY5Y neuroblastoma cells by exerting anti-apoptotic effects and reducing inflammation through decreased TNF-α levels (Garcimartín et al., 2015). This neuroprotective effect may be related to silicon ability to interact with aluminum, thereby reducing its bioavailability and toxicity. Supporting this hypothesis, in vivo study in aluminum-administrated male mice demonstrated treatment with silicic acid (50 mg/L) reduced aluminum accumulation by 40%, preserved the expression of antioxidant enzymes, and normalized TNFα mRNA levels, highlighting silicon potential neuroprotective role against aluminum toxicity (Gonzalez-Muñoz et al., 2008). This has also been examined in a controlled clinical study of AD, which showed that drinking up to 1 L of silicon-rich mineral water daily for 12 weeks increased urinary aluminum excretion in both patients and healthy individuals, without affecting the urinary excretion of essential metals such as iron and copper, while also demonstrating cognitive improvement (Davenward et al., 2013).

Although current data remain limited, recent studies also highlight the role of lithium and silicon in maintaining a healthy gut microbiome. Specifically, lithium treatment in rats was associated with a significant increase in microbial species enrichment and diversity. At the phylum level, Li administration led to an increase in Actinobacteria and a decrease in Bacteroidetes. At the family level, elevated relative abundances of Peptostreptococcaceae, Clostridiaceae, and Ruminococcaceae were observed. Notably, lithium also increased the relative abundance of several low-prevalence genera, including Clostridium sensu stricto 1, Ruminiclostridium 5, Intestinibacter, Eubacterium coprostanoligenes, Peptoclostridium followed by reduction in Bacteroides and Ruminococcus 1 (Cussotto et al., 2019). Similarly, an in vivo study using a dextran sodium sulfate-induced colitis mouse model, demonstrates that lithium carbonate alleviates colonic inflammation by modulating gut microbiota composition and enhancing SCFA production, particularly by increasing Akkermansia muciniphila. This enhancement of SCFAs was associated with an anti-inflammatory response by the activation of regulatory T cells in lamina propria, a crucial layer located in the gut (Huang et al., 2022). Moreover, it is recently suggested that lithium's primary site of action may be peripheral, within the GI tract rather than the brain. Lithium appears to initiate behavioural effects via transient receptor potential cation channel (TRPM2)-dependent activation of enterochromaffin cells, engaging vagal and area postrema pathways. These effects were abolished by vagotomy, area postrema lesioning, or ablation of TRPM2-expressing enterochromaffin cells, supporting a causal gut–brain mechanism (Thorpe et al., 2026).

Similarly, silicon is shown to enhance the diversity and stability of gut microflora, altering the gut microbiome toward a healthier profile, while simultaneously affecting the body's inflammatory response. For instance, 4-week-long consumption of silicon-rich water in rats increased beneficial taxa like Ruminococcaceae UCG-005 and Lactobacillus reuteri, while reduced potential pathogens such as Mucispirillum and Rodentibacter. Consumption of silicon-enriched water in both rats and mice is shown to elevate plasma antioxidant activity (glutathione peroxidase) and reduce gastric mucosal oxidative stress (Wu et al., 2021). Another study conducted in diabetic rat models showed that silicon-enriched diet upregulated colonic tight junction proteins and goblet-cell numbers, and activated antioxidant pathways (Nrf2, GPx), while lowered pro-inflammatory cytokines (TNFα, IL-6) (Hernández-Martín et al., 2024). Although further investigation is needed, the above findings suggest that lithium and silicon can modulate both gut and brain integrity and function. Moreover, it remains to be investigated whether disruption of lithium and silicon homeostasis may dysregulate gut-brain communication contributing to the precipitation of brain pathology and if their daily supplementation even in subtherapeutic levels (e.g. lithium/silicon-enriched water) could be beneficial for brain health.

9. Conclusion

This review focuses on the molecular and cellular mechanisms of the gut-brain communication highlighting its influence on brain health as well its emerging role in the precipitation of brain disorders e.g. depression and AD. As the importance of brain health and the prevention of mental health disorders are increasingly recognized by national health systems and modern societies, this review provides a holistic overview of the (dys)regulating parameters (e.g. lifestyle and diet including drinking water trace elements such as lithium and silicon) that may “shape” the gut microbiome and brain integrity towards brain pathology. Although the underlying mechanisms of the essential gut-brain interplay and communication remain quite complex, the understanding of the gut-brain axis role in mental health is vital for the development of novel preventive and/or therapeutic strategies, aiming to protect against mental health disorders and support overall well-being.

Author contribution

KS and FK contributed equally to the literature search, conceptualization, figure preparation (created using BioRender), drafting and revising of the manuscript. PP, MA, and JB contributed to writing and critical revision of the manuscript. NK, CD, and IS contributed to the conceptualization of the study and provided scientific input. AE provided scientific consultation. CD and IS supervised the project and contributed to manuscript refinement and finalization. All authors reviewed and approved the final version of the manuscript.

Conflict of interest

This study was financially supported in part by Souroti Ltd, which had no influence on the literature search, analysis, interpretation and writing of this manuscript. A.E. is an employee of Souroti Ltd. N.K. and C.D. have received speaker's fees, consultancy honoraria and travel support from Janssen, Lundbeck, Medochemie Generics, Elpen, Vian/Vianex and Boehringer Ingelheim, which not relevant to this study. The remaining authors have no conflict of interest to declare.

Glossary

28-HDRS

28-item Hamilton Depression Rating Scale

5-HT

5-hydroxytryptamine or Serotonin

ACTH

Adrenocorticotropic hormone

AD

Alzheimer's disease

Aβ

Amyloid beta

BA

Bile acid

BBB

Blood brain barrier

BDNF

Bone derived neurotrophic factor

CNS

Central nervous system

CRH

Corticotropin releasing hormone

CSF

Cerebrospinal fluid

ENS

Enteric nervous system

EVs

Extracellular vesicles

FMT

Fecal microbiota transplantation

FOS

Fructooligosaccharides

GI

Gastrointestinal tract

GOS

Galactooligosaccharides

GR

Glucocorticoid receptors

GSK-3β

Glycogen Synthase Kinase-3β

H3k27me3

Histone H3 lysine 27 trimethylation

HDAC

Histone deacetylase

HMOs

Human milk oligosaccharides

HPA

Hypothalamic-pituitary-adrenal

LPS

Lipopolysaccharides

MCI

Mild cognitive impairment

MCTs

Monocarboxylate transporters

MDD

Major depressive disorder

ncRNAs

noncoding RNAs

NFLs

Neurofilament light chains

NMDA

N-methyl-D-aspartate receptor

SCFAs

Short chain fatty acids

SLC6A4

Solute carrier family 6 member 4

SSRIs

Selective serotonin reuptake inhibitors

TET1

Ten-eleven translocation 1

TMAO

Trimethylamine N-oxide

TNF-α:

Tumor necrosis factor α

TRPM2

Transient receptor potential cation channel 2

VEGF

Vascular endothelial growth factor

VN

Vagus nerve

WHO

World health organization