Imbalance in Two-Way Communication Between Gut and Brain in Alzheimer's Disease Beyond Microbiome Changes

The current research permanently updates various types of communication channels for the two-way interactions between the gut and the central nervous system (CNS), known as the GBA. Although the gut and brain are physically separated, they are connected through various pathways, primarily via bloodstream, including the transport of intestine-derived molecules to the brain vasculature, where they may penetrate the brain parenchyma if they cross the blood–brain barrier (BBB; Figure 1). The chemical transmission of signals from the gut to the brain forms the most significant, and probably the most sensitive to pathologies, communication channel.

The human gut microbiota is a complex ecosystem of various microorganisms producing various biologically active metabolites. The essential class of these metabolites generated by the gut microbiota are short-chain fatty acids (SCFAs), such as acetic, propionic, and butyric acid [10]. In healthy conditions, a fraction of the SCFAs produced in the colon from carbohydrates enters the bloodstream (approximately 40% acetic acid, 10% propionic acid, and 5% butyric acid) to be distributed to peripheral tissues [11]. SCFAs are generated by a diverse gut microbiota, including members of the phyla Bacteroidetes and Firmicutes (Figure 1) [10,12]. SCFAs act locally, but, when taken to the bloodstream, they are involved in the regulation of such functions as metabolism, the modulation of the immune response, and appetite control [11], suggesting that SCFAs can reach the brain centers that control eating behavior.

In the context of the GBA, increasing attention was paid recently to the gut-produced primary and secondary BAs. After the BAs release from the liver, the intestinal microbiota can convert these primary BAs into secondary BAs, which may acquire new functions. Thus, being neuroactive lipophilic molecules, BAs can, along with SCFAs, be involved in the communication between the gut and the brain (Figure 1) [13,14].

In addition to SCFAs and BAs, the GBA is complemented by signaling through immune cells, being conditioned in the gut wall, and then reaching the brain via the bloodstream. It has been proposed that some immune cells, such as T cells and pro-inflammatory T helper 17 (Th17) cells, can cause inflammation not only locally in the intestine, but also in the brain [15]. These authors demonstrated that, in inflammatory bowel disease, the pathogenic Th17 CD4+ T cells reached the brain, promoting neuroinflammation [15]. Notably, a diet rich in salt can increase the risk of cerebrovascular diseases and dementia via the expansion of Th17 cells in the small intestine, thus promoting interleukin-17 (IL-17) release [16]. While likely limited in healthy conditions, IL-17 can then enter the brain either through the disrupted BBB or via glial limitans, which are functionally linked to the meningeal lymphatic system, densely populated by various types of lymphocytes and other immune cells (Figure 2) [17,18].

Recently, new players in gut–brain communication have been proposed, such as microRNAs (miRNAs), generated in the crosstalk between intestinal cells and bacteria and released to the bloodstream, especially in the case of a compromised intestinal barrier [19,20]. This signaling pathway as a new candidate for GBA communications (Figure 1) looks attractive, consistent with the common signaling role of miRNAs in various functions, but further studies are required to confirm their involvement in GBA signaling.

Collectively, these findings illustrate that the concept of the GBA continues to evolve, with new molecular and cellular mediators expanding the list of humoral connections between the gut and brain in healthy conditions. However, for many members of this ever-expanding list of humoral messengers, the main limitation is the lack of direct evidence of their penetration across the BBB, as well as knowledge of the molecular and cellular targets of these messengers in brain cells.

The humoral communication within the GBA is complemented by neuronal mechanisms, involving autonomic and somatic nerves, as well as by the enteric nervous system (ENS). Within the complex gut innervation, the vagus nerve, containing both afferent and efferent fibers, plays a central role in the GBA, as a direct and fast communication channel between these organs.

One of the most interesting areas of research related to the role of the vagus nerve in intestinal innervation is new discoveries regarding the profile of membrane receptors on nerve endings that function as sensors of intestinal homeostasis. Thus, recently, Liu et al., suggested an original view, that, through the vagus, the gut reports to the brain a new 'sense', which allows the host to respond in real time to stimuli arising from the resident gut microorganisms [21]. This mechanism appears to involve microbial flagellin, which directly activates vagal nodose neurons to reduce feeding behavior [21]. Such an intriguing type of signaling can form a new ascending connection channel from the gut to the brain, whose role in health and disease deserves further studies.

The descending vagal activity is provided by the neurotransmitter acetylcholine (ACh), which controls motility, secretion, and other local gut functions (Figure 1). ACh can provide the beneficial so-called 'cholinergic anti-inflammatory pathway' (CAP) effect, due to the ability to dampen neuroinflammation in the brain [22], as well as peripheral inflammation in the intestine [23]. CAP has been extensively studied in brain pathologies, including AD, and accumulating evidence supports the notion that the cholinergic dysregulation and inflammatory responses play a key role in the pathogenesis of AD [24]. However, its importance for peripheral cholinergic systems as guardians of intestinal homeostasis requires further study.

In the gut, the local abundant neurotransmitter is serotonin (5-HT), produced by intestinal enterochromaffin cells (ECs) [25]. The release of 5-HT is largely dependent on gut motility, while, in turn, this transmitter coordinates secretion and gut motility, along with immune functions and the state of the intestinal barrier (Figure 1) [26,27]. This regulation of gut functions primarily involves serotonin type 3 receptors (5-HT3), expressed in vagal afferent fibers (Figure 1), sending neuronal spiking activity to the brainstem [28,29]. This means that peripheral serotonergic activity forms an important communication channel between the gut and the brain.

In the brain, 5-HT is well-known as a key signaling molecule, which modulates the mood and cognition, and shaping appetite, digestion, and emotional wellbeing [30,31]. This suggests that central serotonergic signaling may indirectly control gut function through the descending (from brain-to-gut) branch of the GBA. Thus, in both the brain and the gut, 5-HT is an important messenger that ensures the efficient functioning of the GBA under healthy conditions.

The vagal innervation of the gut is supplemented by less-explored somatic afferents, which, apart from sending pain signals, detect various intestinal functional states, in particular, motor activity, through recently discovered highly mechanosensitive Piezo channels [32] (Figure 1).

Thus, the complex neuronal network of the intestine, along with its relative independence from the CNS, is an emerging area of research that is likely to yield, in the near future, new discoveries of gut-to-brain interactions. Future studies are expected to expand our knowledge of the profile of membrane receptors in the vagus and somatic nerves, including mechanosensitive channels, which likely control most gut functions.

In addition to neuronal pathways, the GBA also includes the bidirectional hormonal control, when hormones produced by the gut modulate brain functions, while hormones produced by the hypothalamic–pituitary–adrenal (HPA) axis shape the functional state of the gastrointestinal system. Several studies suggest that the gut microbiota develops in parallel with the HPA axis, and that they are in constant interaction [33,34]. In some conditions, like irritable bowel syndrome, it has been shown that the common stress hormone cortisol directly affects immune cells and primary afferent nerve fibers in the gastrointestinal tract [35]. Notably, the stressful stimulation of the HPA can also increase the permeability of the intestinal wall and contribute to the development of dysbacteriosis [36,37].

As participants in the ascending loop of the GBA, intestinal enterochromaffin cells (EC) can secrete a variety of local hormones, such as gastrin, secretin, cholecystokinin, and peptide YY, which, apart from the control of gastrointestinal motility and food transit, can be taken into the circulation to approach other tissues and organs. The secretion of these intestinal hormones is controlled not only by the vagus and somatic nerve fibers, but also by locally produced BAs and SCFAs (Figure 1). Thus, the studies revealed that the G-protein-coupled BA receptor (GPBAR1) is highly expressed in the specialized L-type of EC cells, while the SCFA receptor free fatty acid receptor 2 (FFAR2) is present both in L-type and other EC cells (Figure 1) [38,39]. It means that the function of these hormone-secreted cells is under the control of these principal bacterial metabolites.

One recently proposed mediator of the interaction between the intestinal microbiota and the CNS is the glucagon-like peptide-1 (GLP-1). GLP-1, secreted from L-type cells in response to the nutrition type and microbiota activity (Figure 1), plays an important role in the regulation of glucose homeostasis and the energy balance, and exhibits pronounced anti-inflammatory properties [40,41]. A healthy diet that includes olive oil components, such as hydroxytyrosol, may promote the release of GLP-1 [42,43]. Through specific blood–brain barrier penetration mechanisms, GLP-1 can directly influence brain cells and promote neuroprotection [44]. This novel direct link could be supplemented by the GLP-1-analog–induced stimulation of pancreatic β-cells, mediating insulin secretion [45], which, via microglia, can promote anti-inflammatory effects in the brain [46]. However, the issue of whether the BBB is easily permeable by insulin under healthy conditions, or in disease, remains not entirely clear.

In summary, there are many protective humoral and neuronal mechanisms of the GBA that support healthy brain functions. The action of potentially harmful stimuli in the healthy conditions is minimized by structural factors, such as the intestinal barrier and BBB and the optimal coordination of secretion and digestion, as well as normal gut motility (evident as the lack of constipation), and supplemented by serotonergic cholinergic anti-inflammatory signaling. Table 1 summarizes the major gut-derived and host-regulated factors implicated in the GBA. Together, these multiple mechanisms form a kind of virtuous circle where the ascending and descending branches of the GBA are in optimal balance.

One commonly discussed mechanism, proposed for an abnormal GBA in AD, is dysbiosis, leading to the production of toxic molecules and the activation of the immune system. Different species of Gram-negative bacteria are the primary candidates for dysbiosis, associated with the onset of AD [68,69,70]. Thus, Helicobacter pylori (H. pylori) infection was associated with neurodegenerative changes in cognitively normal men [71]. A long-term analysis of third National Health and Nutrition Examination Surveys (NHANES III) data has demonstrated an association between periodontal pathogens such as Porphyromonas gingivalis, Prevotella melaninogenica, and Campylobacter rectus, and an increased risk of incidence of AD [72]. Other studies showed that patients with AD have an increased proportion of Bacteroidetes type bacteria, while the content of the beneficial Firmicutes and Bifidobacterium types is significantly reduced [73]. The studies of patients with MCI and AD showed an increase in the Gammaproteobacteria, as well as Enterobacteriales and the Enterobacteriaceae family [74]. The meta-genomic sequencing in the preclinical stage AD demonstrated uprising levels of Dorea formicigenerans, Oscillibacter sp., Faecalibacterium prausnitzii, Coprococcus catus, and Anaerostipes hadrus [68] (Figure 1).

There are some conflicting data regarding the composition of the gut microbiota in AD [75,76,77,78]. For example, one study found that the content of Bacteroidetes was reduced in AD patients, while Ruminococcaceae, Enterococcaceae families, and Lactobacillus were increased [67]. These discrepancies are likely due to the differences in the study methodology, population characteristics, and disease stages, as well as environmental factors and diet of the participants.

An intriguing new aspect of the role of bacteria in the development of AD is based on recent research suggesting that, in addition to the local harmful effects on the gastrointestinal tract, pathogenic Gram-negative bacteria have the potential to penetrate the brain. This research suggested several proposed routes for how bacteria can enter brain. First, they could overcome the intestinal barrier and then destroy the intercellular junctions of endothelial cells in brain vessels, which allows them to cross the BBB through the paracellular pathway [57]. In particular, E. coli can bind to the receptors of endothelial cells to destroy the tight junctions between these cells, thus compromising the BBB [79]. Second, Gram-negative bacteria can cross the BBB by transcytosis, through interactions of bacterial proteins with the outer membrane of endothelial cells [58,80]. Finally, Gram-negative bacteria can enter the CNS along the cranial trigeminal and olfactory nerves [81,82]. Even though it remains not fully clear how common this phenomenon is, there are reports that the brain of patients with AD contains 5–10 times more bacteria than a healthy brain [83]. Although such bacterial enrichment of the brain could be an additional factor potentially contributing to the development of AD, evidence for such mechanisms remains weak, as it is mainly derived from in vitro models of the human BBB.

More strong evidence for the brain delivery is provided for bacterial EVs [84,85,86,87], since, apart from in vitro studies, this possibility was explored in animal models [87]. However, further studies demonstrated the presence of lipopolysaccharide (LPS)-containing EVs in the plasma of AD patients [88]. These authors also showed that such bacterial EVs can activate the excessive pruning of synapses via microglia activation [88], which could be a factor contributing to neurodegeneration (Figure 2).

Thus, although these data indicate a significant role of gut microbiota in the progression of AD, there is a variability in the observations obtained from different sources. The evidence for the link between dysbiosis and AD development remains mainly associative and the data on the movement of bacteria across the BBB in vivo are highly debatable, as they are based mainly on the evidence from in vitro studies. Together, this highlights the need for further systematic studies to more accurately establish the role of the abnormal microbiota and dysbiosis in the pathogenesis of AD, including the careful exploration of the bacterial and EV delivery from the gut to the brain.

Out of other toxic compounds delivered by bacteria, the most attention is traditionally paid to LPS (Figure 2), produced by Gram-negative bacteria [89]. The important observation is that the plasma level of LPS is significantly elevated in patients with AD [59,89], consistent with the pro-inflammatory signaling caused by LPS in AD and amyotrophic lateral sclerosis (ALS) pathology [90]. These data are consistent with the current understanding of the contribution of the gut microbiota to neuroinflammation: LPS from Gram-negative bacteria is thought to enhance oxidative stress and pro-inflammatory responses, which may contribute to the progression of neurodegenerative diseases [91]. In the brain, the primary target for LPS is microglia (Figure 2), which normally quickly respond to changes in the brain environment to clean tissue from the waste, including accumulating Aβ [92]. LPS can contribute to the development of AD pathology by stimulating the phagocytic state of microglia via Toll-like receptor 4 (TLR4) and activated nuclear factor-κB (NF-kB) (Figure 2) [93,94,95], which leads to the secretion of multiple pro-inflammatory cytokines and the promotion of the oxidative stress (Figure 2). Apart from TLR4, LPS increases the expression of the receptor for Advanced Glycation End-products (RAGE), which is critically involved in the pathology of AD, including Aβ deposition, and cognitive impairment [96]. LPS activates the microglia via TLR4, switching them to a pro-inflammatory phenotype via the NF-κB pathway, causing neuroinflammation and cognitive impairment. The blockade of TLR4 with the VIPER peptide prevents these effects (Figure 2) [97]. Notably, the long-held "binary" view of microglial functional states (pro-inflammatory M1 and anti-inflammatory M2) is now being updated, suggesting multiple functional subtypes of these glial cells, associated with a given brain disease [98,99]. The latter suggests a promising new direction for further research concerning LPS targets in the brain.

Even though it is generally accepted as a highly pathogenic and pro-inflammatory compound, it is important to note that multiple species of Gram-negative bacteria produce different types of LPS, with distinct abilities to trigger inflammation. Interestingly, the fecal LPS from a healthy human microbiome, primarily from Bacteroidales, weakly activates TLR4 and can suppress cytokine responses induced by highly pro-inflammatory LPS, such as from E. coli [60]. This currently appears to be a controversial issue, as other studies report that Bacteroides fragilis can also produce a pro-inflammatory form of LPS [61].

Thus, the obtained data indicates several complementary molecular mechanisms for the action of LPS in AD, which, together, contribute to the development of this pathology. The potentially beneficial forms of LPS mentioned above are of considerable translational interest, but this unexpected inhibitory mechanism requires further study, particularly its association with the specific bacterial species.

In a recent study, six types of SCFAs were found to be reduced in patients with MCI and AD, with valeric acid showing a particularly notable reduction in AD patients, while five other SCFAs, such as formic, acetic, propanoic, 2-methylbutyric, and isovaleric acids, were significantly different between the MCI and AD groups [100]. Likewise, in AD patients, fecal SCFAs propionate, isovalerate, and propionate-producing bacteria are inversely associated with amyloid status [101]. In the study employing a mouse model of AD, it has been established that SCFAs butyrate reduces Aβ levels and weakens cognitive dysfunction caused by AD [47], thus providing support for the beneficial role of SCFAs in AD. A Kazakhstan-based project 'Study of Gut Microbiota Alterations in Alzheimer's Dementia Patients from Kazakhstan' revealed that, compared to age-matched healthy controls, patients with AD have significant alterations in microbiota, such as a decrease in SCFA-producing bacteria Bifidobacterium and Roseburia [102].

While the majority of data obtained from animal models and human studies suggest a beneficial effect of SCFAs in AD, it is important to note that certain studies have shown that some SCFAs may have a negative effect. Thus, Vinolo et al. found that the SCFAs acetate, propionate, and butyrate increased the release of the cytokine CINC-2αβ (cytokine-induced neutrophil chemoattractant-2αβ), promoting the migration of neutrophils to the site of inflammation, thereby aggravating the inflammatory response [103].

In summary, the available data, including studies in AD patients, indicate that SCFAs play a predominantly neuroprotective role in the pathogenesis of this disorder. Because much of the data regarding SCFAs has been obtained from animal studies and small groups of patients, further research and confirmation in larger patient cohorts is needed.

Studies show that patients with AD have changes in the profile of primary and secondary BAs. Thus, a decrease in the level of the primary BAs, cholic acid (CA), and an increase in the level of the bacteria-generated secondary BAs, deoxycholic acid (DCA), and its conjugates with glycine and taurine have been shown [48]. Notably, some studies showed that DCA and lithocholic acid (LCA), can cross the BBB and affect brain functions in AD (Figure 2) [48,49,50,104]. An increase in the DCA/CA ratio correlates with cognitive decline, suggesting a possible role of the gut microbiota in AD progression [49]. These data, taken together, suggest that BAs could be significant contributors to neurodegenerative diseases, such as AD [49,50,62].

One of the key questions with a significant translational impact is the origin of secondary BAs, and their link to certain gut bacterial taxa. In this regard, it has been shown that the proliferation of pathogenic Clostridium species, such as C.difficile, increases the level of DCA, which, in turn, elevates the serum level of C-C motif ligand 5 (CCL5) and induces CCL5 receptor 5 (CCR5) overexpression in sensory neurons [105]. Notably, the high expression of CCR5 in pro-inflammatory microglia was found in AD patients [106], and it has been proposed that CCR5 is one of the major hub genes in AD [107]. Based on these data and other studies showing a link between C.difficile and AD [108], we can speculate that such CCL5/CCR5 signaling could be involved also in the exaggeration of neuroinflammation in AD patients, having an ingrowth of C.difficile (Figure 2).

Thus, AD-associated changes in the composition of microbiomes and the progression of dysbiosis lead to an abnormal profile of BAs, increasing the levels of toxic secondary BAs, such as DCA and LCA, which can contribute to neuroinflammation and related damage to neurons (Figure 2). Taken together, it appears that the role of BAs in AD is more complex than that of SCFAs, as some BAs can have neurotoxic effects, requiring more careful consideration of the role of bacterial taxa that converts primary BAs to secondary ones.

These and other pathogenic contributors to the GBA are shown in Table 1.

The multiple neuronal pathways appear to be dysfunctional in AD, leading to a disruption in gut motility and gastrointestinal transit, as shown in a large cohort of AD patients, and additionally confirmed in the 5XFAD AD model mice [55]. Slow chyme transit, due to lowered gut motility, through the gut in AD, leads to constipation, which is a common symptom in this disorder [56]. Constipation in AD may also be due to the altered intestinal profiles of SCFAs and BAs, which interact with the neuronal control of intestinal motility (Figure 1).

The dysfunction of the ascending neuronal mechanisms may lead to a disruption in the flow of information from the digestive system to the brainstem and other brain areas that regulate food intake [109]. We speculate that irregular feeding behavior in AD, in return, results in abnormal motor/secretory activity in the gut, a deficiency in key nutrients, and the further progression of dysbiosis.

The precise mechanisms by which the intestine utilizes complex neuronal networks to coordinate vital gut functions, interacting with the brain, are not fully understood and remain an area of active research. In this context, a recent discovery demonstrated that the hippocampus contains specific centers that collect information transmitted by the vagus nerve and somatic afferents from the gut to control eating behavior [110]. Since the hippocampus is the brain region most prone to AD-related neurodegeneration, we can speculate that such pathological processes in the brain can affect the episodic memory associated with food intake.

Common symptoms of AD include weight loss and, as mentioned above, abnormal food-seeking behavior, functions also controlled by the hypothalamus. This brain region, crucial for the control of the metabolism, appeared to also be involved in the pathogenesis of AD [111]. Thus, Robison et al. (2023) found, in the 3xTg-AD mouse model, an increased expression in the hypothalamus of pro-inflammatory cytokines, suggesting the development of the local neuroinflammation [112]. Consistent with this, circadian rhythms, known to be dependent on hypothalamus activity, were already disrupted in a pre-symptomatic cohort of AD patients, along with other biomarkers of this pathology [113].

Recently, Kim et al. (2025) demonstrated the new potential channel of communication within the GBA, which relied on the detection of D-glucose in the gut by somatic (but not visceral) nerves [114]. This ascending signaling results in the modulation of the activity of corticotropin-releasing factor (CRF)-expressing neurons in the hypothalamic paraventricular nucleus. Notably, this paraventricular nucleus of the hypothalamus is undergoing degeneration in AD pathology [115,116,117], which may follow in deviant food behavior in this pathology, affecting multiple functions of the gastrointestinal system, ultimately further unbalancing the GBA (Figure 3). Notably, the dysfunction of the HPA and elevated cortisol can further promote hippocampal damage in AD [63], providing insight on how the interaction between these brain centers may contribute to the development of this disorder.

Overall, recent discoveries reveal a multitude of neural and humoral mechanisms linking the gut and the brain that acquire new deleterious functions in AD, forming a harmful vicious circle, which supports the pathological process.

Optimizing the microbiota to improve ascending gut-to-brain mechanisms in AD could be potentially achieved via diet, more specifically, with prebiotics, probiotics, and postbiotics, or by the fecal transplantation of beneficial bacteria. The former has been addressed elsewhere, and, in our review, we focused on fecal microbiota transplantation, which is an attractive but still rather controversial area of research.

Thus, it has been shown that the treatment of AD mouse with fecal microbiota from a person with a low risk of AD, due to the protective APOEe2 allele, improved memory but promoted neuroinflammation, probably due to the heterogeneous strain of bacteria in the host and donor microbiota [118]. In the other study, Jiang et al., before transplantation, used a prior depletion of endogenous strains by antibiotics and noted that the positive effect is temporal and reduces with time [119]. Similar results were observed by Grabrucker et al., who found that the microbiota from AD patients disrupted cognition and hippocampal neurogenesis into microbiota-depleted rats [120]. These studies mixing the native recipient with donor bacteria strains revealed several unexpected pitfalls in this procedure. Nevertheless, Sakurai et al., in a randomized placebo-controlled trial, showed beneficial effects of the Lactiplantibacillus plantarum OLL2712 strain in patients with memory problems [121], demonstrating that this procedure remains promising.

In this context of controversial research, an interesting alternative for the correction of the microbiome is the use of synthetic bacteria. Thus, a recent Science 2025 article reported the successful generation of Escherichia coli with a 4 Mb synthetic genome opens a potential perspective for enriching the microbiome by bacteria with designed properties [122]. Despite the originality of this approach and its potential coupling in future with AI technologies, its intervention in translational medicine is still far from practical implementation.

In summary, there are various pitfalls in fecal microbiota transplantation aimed at improving the GBA in AD, most notably a temporal beneficial effect of this procedure. As mentioned above, one of the complications is that the pathological strains of bacteria may be present in the recipient stool even after transplantation. It means that fecal microbiota transplantation requires the consideration of bacterial heterogeneity in both the recipient and donor, as well as the different roles of strains in different parts of the gastrointestinal tract.

SCFAs produced by anti-inflammatory gut microbiota regulate immune responses by enhancing the differentiation of regulatory T cells and limiting the production of pro-inflammatory cytokines [123]. Chandra et al. proposed a propionate-based therapeutic strategy to attenuate Th17 signaling, leading to reduced amyloid deposits [124]. These exciting results are to be confirmed in clinical trials, before the application to AD patients.

BAs, particularly secondary BAs, play a predominantly pro-inflammatory role. This suggests that, unlike SCFAs, the therapeutic approach to these intestinal metabolites should primarily focus on reducing their pathogenic role in AD. While the profile of primary BAs is dependent on liver function, the level of secondary BAs is determined by the activity of gut microbiota, consistent with the data that brain BAs have a peripheral origin [125]. Consistent with the former, it is established that diet-induced alterations to liver function can affect brain Aβ homeostasis [126]. Thus, in AD, the treatment options may require optimizing the peripheral BAs metabolism both via diet and by normalizing microbiota.

Surprisingly, recent data suggest that some BAs could provide a positive effect in AD pathology. For instance, Tauroursodeoxycholic acid appeared to play a beneficial role in the prevention and possibly in the treatment of AD [127]. Furthermore, the BAs action via the Takeda G protein-coupled receptor 5 (TGR5) and the farnesoid X receptor (FXR) is of significant interest, as studies have demonstrated the anti-inflammatory effect of TGR5 activation, leading to the suppression of macrophage function and cytokine production, while FXR activation reduced intestinal inflammation and epithelial permeability. These data support the use of specific TGR5 and FXR modulators as potential therapeutic strategies for AD [128,129].

In summary, even though SCFAs and some BAs have emerged as promising therapeutic tools in AD, their therapeutic administration, due to their involvement in multiple aspects of AD pathology, requires further clinical research, considering the microbiota profile in the individual patient.

The ability of GLP-1 to penetrate the BBB and exert a neuroprotective effect by reducing neuroinflammation, modulating amyloidogenesis, and reducing oxidative stress [51] suggests that this gut-produced hormone can be a potential treatment option for AD.

Thus, the dual GLP-1/Gastric inhibitory polypeptide (GIP) receptor agonists such as DA-JC4, DA5-CH, DA-JC1, and DA-CH3 have shown the ability to reduce inflammation, and reversed memory impairments and enhanced synaptic plasticity in the hippocampus of APP/PS1 mice. In 3Tg AD mice treated with DA-JC4, improvements were observed in object recognition, spatial memory, and hippocampal pathology, including reductions in Aβ and tau accumulation.

Apart from animal models, clinical studies suggested that GLP-1 receptor agonists may help to slow cognitive decline in AD patients. For instance, phase 2B clinical trial studies showed that the GLP-1 agonist liraglutide reduced brain atrophy in regions involved in memory and decision-making by nearly 50% compared to the placebo [130,131]. Moreover, the GLP-1 analogue liraglutide has shown promising effects in improving cognitive function in both animal models and patients with AD by suppressing neuroinflammation through reduced microglial activation [132].

Despite these encouraging preclinical findings and the earlier data from patients, clinical trials evaluating GLP-1 receptor agonists in AD patients remain limited. Notably, the recent systematic review concluded that GLP-1 agonists, while showing some positive metabolic and neuroprotective effects, did not reduce beta amyloid and tau biomarkers and did not significantly improve the patient's cognitive state [52]. Further studies are needed to clarify this still attractive approach, especially among patients with a comorbidity of AD and diabetes.

Within the integrative view of the multiple interconnected functions of the gastrointestinal system that we hypothesize are disrupted in AD, in future therapies, it is necessary to coordinate various processes, such as the integrity of the intestinal barrier, gut motility, and secretion of digestive enzymes, as well as the proper activity of local nerves.

In this regard, a new promising role of immunotherapy in improving gut functions emerged recently, based on the CD4 therapy. Thus, Gómez de Las Heras et al. (2025) found, in the early senescing mouse model with T-cell failure associated with dysbiosis and a disrupted gut barrier integrity, that CD4 T cells or Treg-cell-enriched therapy prevented senescence features by restoring the gut barrier integrity [133]. Consistent with these results, Park et al. showed, in a mouse model, the beneficial results of autologous Treg cells in Parkinson's disease [134].

This exciting new data suggests that similar T-cell-based immunotherapy could potentially also be used for patients with AD.

Within the concept of the formation of a vicious circle in GBA, with disrupted descending neuronal signaling from the brain to the gut (Figure 3), the treatment options should be directed to improve such neuronal control. This should go along with therapeutics reducing amyloidosis and neuroinflammation in the brain, which can also, indirectly, optimize the complex gut activity through an improved descending control of the gut by the brain.

Consistent with this view, the direct activation of brain-to-gut communications could be based on vagus nerve stimulation. This technique is widely used already in various pathologies, including AD, where it shows a promising therapeutic effect [53,54,135,136]. Alternative to this could be the stimulation of the somatic–vagal–gastric reflex by electropuncture [137]. Notably, apart from the descending effect, vagus nerve stimulation can provide a beneficial ascending effect on brain functions, in particular, by activating the hypothalamic–pituitary–adrenal axis, which, as mentioned above, plays an important role in the control of the functional state of the gastrointestinal system.

Focused brain stimulation represents a range of dynamically developing techniques [138,139,140] also applicable to AD. For instance, in aged rats, the high-frequency deep stimulation of lateral hypothalamic areas reduced memory decline [141].

In general, non-invasive techniques are emerging, such as a battery of instrumental methods which demonstrate high potential as efficient tools in AD, alone or in combination with pharmacological approaches.