Evidence of altered gut microbial composition in BD patients suggests that gut microbial dysbiosis contributes to disease progression and pathophysiology in BD patients (Evans et al., 2017; Lai et al., 2020; McIntyre et al., 2021). Although there are few studies discussing the relationship between gut microbiota and cognitive impairment in BD patients, some have found that specific gut microorganisms are associated with reduced cognitive function (Severance et al., 2016). The relationship between the gut microbiota and specific cognitive domains, including learning, memory, attention, processing speed, and executive function—all of which are also impaired in BD—has been demonstrated in animal and human studies on hepatic encephalopathy (Bajaj et al., 2012), diabetes mellitus (Zheng et al., 2021), and aging (Li et al., 2021). Gut microbial dysbiosis also contributes to cognitive impairment in various neurodegenerative and psychiatric disorders, including Parkinson’s disease (Mulak and Bonaz, 2015; Sampson et al., 2016), AD (Jiang et al., 2017), and schizophrenia (Bioque et al., 2021), suggesting potential overlapping mechanisms between these conditions and BD.

Disruption of intestinal homeostasis affects the host’s metabolism and immune responses, leading to systemic disorders ranging from metabolic syndrome to chronic inflammation, which is closely related to the development of cognitive impairment (Kesika et al., 2021; Liu et al., 2020; McGrattan et al., 2019; Zeng et al., 2021). Therefore, a quantitative analysis of the gut microbiota in BD patients may help to improve an understanding of the mechanisms behind the development of cognitive impairment in this population. The most dominant gut microbial phyla in healthy adults include Firmicutes, Bacteroidetes, Proteobacteria, Actinobacteria, and Verrucomicrobia, with Firmicutes and Bacteroidetes accounting for nearly 80% of the total amount (Eckburg et al., 2005). In contrast, multiple studies have demonstrated alterations in gut microbial composition in BD patients compared to healthy controls (Figure 2). Lai et al. (2020) found that, in BD patients, counts of Faecalibacterium prausnitzii, Bacteroides, Prevotella, Atopobium cluster, Enterobacter spp., and Clostridium cluster organisms were significantly increased, and the log10 (ratio of Bifidobacteria to Enterobacteriaceae) was decreased. Another two studies have shown an increase in organisms of the Firmicutes and Bacteroides phyla but a decrease in those of the Bacteriodetes phylum in BD patients, with the Firmicutes-to-Bacteriodetes ratio being accordingly increased (Rong et al., 2019; Lai WT. et al., 2021). An increased Firmicutes-to-Bacteriodetes ratio is a recognized indicator of obesity (Crovesy et al., 2020; Magne et al., 2020). However, bacteria of the Ruminococcaceae family and Faecalibacterium genus, which are butyrate-producing bacteria, were reduced in BD patients relative to healthy controls in other studies (Hu et al., 2019; Lai WT. et al., 2021; Sublette et al., 2021). In 4 of 5 studies, lower α-diversity was observed in BD patients compared to healthy controls (Sublette et al., 2021).

In conclusion, the most convergent taxonomic finding in BD patients is the reduction in butyrate-producing bacteria, gut microbiota known to impact cognitive function by producing SCFAs, including acetate and butyrate (Tanca et al., 2017). A decrease in butyrate-producing bacteria, such as those of the genus Faecalibacterium, was also observed in patients with AD and accompanied by neuroinflammation and impaired cognitive function (Ling et al., 2020; Marizzoni et al., 2020). Associations between cognitive impairment and decreased butyrate-producing bacteria have also been observed in patients with Parkinson’s disease (Nuzum et al., 2020; Tan et al., 2021) and alcohol use disorders (Leclercq et al., 2020) and are likely mediated by butyrate. It has been demonstrated that butyrate levels in the central nervous system can be influenced by gut microbial composition; for example, Clostridium butyricum, as a probiotic bacteria, restored butyrate content in the brain and significantly alleviated cognitive impairment and histopathological changes in a mouse model of vascular dementia (Liu et al., 2015). Clostridium butyricum treatment also attenuated cognitive impairment and prevented microglia-mediated neuroinflammation in a manner mediated by butyrate in an AD mouse model (Sun et al., 2020). Prolonged treatment with sodium butyrate stimulates neurogenesis and improves memory and associative learning (Kim et al., 2009; Govindarajan et al., 2011). On the one hand, butyrate inhibits histone deacetylase, which plays an important role in intestinal barrier regulation and intestinal energy metabolism, thereby affecting cognitive function (Leonel and Alvarez-Leite, 2012; Tanca et al., 2017). Butyrate can also increase hippocampal neurogenesis and the expression of the neurotrophic factor BDNF, improving learning behavior and long-term memory (Levenson et al., 2004; Sleiman et al., 2016). Considering the neuroprotective and cognitive improvement effects of butyrate, inadequate concentrations of butyrate-producing bacteria may be involved in the pathophysiology of cognitive impairment in BD patients.

The interaction between butyric acid-producing bacteria and cognitive impairment suggests that microbiota-produced small-molecule metabolites mediate host-microbiome interactions (Donia and Fischbach, 2015). An analysis of brain transcriptomic data from BD patients has also revealed disturbances in gut microbial metabolites, such as tryptophan, SCFAs, and BAs (Moolamalla and Vinod, 2020). These overlapping metabolite profiles suggest potential routes for the involvement of the gut microbiota in cognitive impairment in BD.

Gut microbial dysbiosis increases the intestinal permeability and penetration of pro-inflammatory factors, such as LPS, leading to intestinal inflammation, which, in turn, promotes energy absorption and reduces satiety via gastrointestinal hormones (Gomes et al., 2018; Nagpal et al., 2018). These gastrointestinal hormones have receptors expressed in regions of the brain that regulate not only hunger and energy metabolism but also stress, behaviors, and cognitive function (Chan et al., 2003; Bucinskaite et al., 2009).

Glucagon-like peptide-1 (GLP-1), an insulinotropic hormone secreted by entero-endocrine L-cells, can stimulate glucose-dependent insulin secretion in response to carbohydrate uptake (Holt & Trapp, 2016). In rodent models, butyrate and propionate stimulate GLP-1 secretion from entero-endocrine L-cells via FFAR2 (Tolhurst et al., 2012). Therefore, reduced concentrations of butyrate-producing bacteria in BD patients may downregulate GLP-1 secretion. GLP-1 receptors are expressed in the cerebral cortex, hypothalamus (ventral medial nucleus and arcuate nucleus), and limbic system (amygdala and hippocampus), all of which are regions responsible for the regulation of emotion and cognition (Heppner et al., 2015). GLP-1 and GLP-1 receptor (GLP-1R) signaling have a neuroprotective role in the control of insulin resistance, synaptic plasticity, and neuroinflammation and improve cognitive function in learning, memory, executive function, and attention (Müller et al., 2019; Flintoff et al., 2021). Therefore, reduced GLP-1R signaling caused by reduced butyrate-producing bacteria in BD patients may impair synaptic plasticity and cognitive function, as demonstrated by several studies. Liraglutide, a GLP-1 agonist, reversed manic-like symptoms and impairment in working and recognition memory by improving hippocampal oxidation and BDNF levels in a D-amphetamine-induced BD model (Chaves Filho et al., 2020). Serum GLP-1 levels are significantly lower in BD patients compared to healthy controls and negatively correlate with previous mood episodes (Rosso et al., 2015). In non-diabetic BD patients, liraglutide had beneficial effects on several cognitive domains, including auditory verbal learning, working memory, and attention (Mansur et al., 2017). Therefore, it is likely that gut microbial dysbiosis affects cognitive function in BD patients via GLP-1R signaling.

Peptide YY (PYY) is another intestinal satiety hormone produced by entero-endocrine L-cells, which affects the central nervous system by inhibiting orexigenic neurons expressing neuropeptide Y (Wynne and Bloom, 2006). Although there is limited evidence directly linking PYY to BD, a reduction in peripheral PYY concentrations might underlie the lack of GABA inhibition associated with impaired cognitive function in BD, considering that both neuropeptide Y and γ-aminobutyric acid (GABA) are released by arcuate neuropeptide Y neurons (Acuna-Goycolea et al., 2005; Huber et al., 2018).

Adipokines, represented by adiponectin and leptin, are cytokines or hormones secreted by white adipocytes in response to increased circulating inflammatory factors (Aguilar-Valles et al., 2015). Gut microbial dysbiosis elicits the penetration of microbial components (e.g., LPS), compromising leptin signaling and leading to leptin resistance and sustained high leptin levels (Faggioni et al., 1997). Intracerebroventricular administration of LPS induces an elevation in leptin receptors in the hippocampus, accompanied by impaired learning and memory, suggesting that leptin signaling disturbances in the hippocampus are involved in the regulation of cognitive responses (Da Ré et al., 2020). Studies suggest that leptin levels may affect cognitive function in BD patients. The TNF-α antagonist, infliximab, reduces plasma leptin levels in BD patients by modulating soluble tumor necrosis factor receptor 2, contributing to better performance on non-literal memory tasks and increased overall cortical volume, both of which are negatively correlated with the leptin level (Mansur et al., 2020). Adiponectin is the most abundant plasma adipokine that regulates energy expenditure, and it improves insulin sensitivity and fatty acid oxidation as an anti-inflammatory factor (Berg et al., 2002; Kubota et al., 2007). In current studies, adiponectin has been shown to improve cognitive function mainly by regulating insulin sensitivity and inhibiting inflammation (Rizzo et al., 2020). Additionally, it alleviates isoflurane-induced cognitive impairment in aging models by activating the p38-mitogen-activated protein kinase pathway and promoting the proliferation of hippocampal precursor cells (Zhang et al., 2019). Studies on adiponectin levels in BD patients remain controversial since body mass index seems to be more strongly correlated with adiponectin levels than mood status in BD patients (Platzer et al., 2019). Therefore, it is difficult to identify the role of the gut microbiota in the interaction between adiponectin and cognitive function in BD patients, which may be more directly correlated with comorbidities in BD patients.

Cholecystokinin (CCK) is a satiety hormone secreted from I-cells of the proximal small intestines during digestion (Raybould, 2007). Vagal afferent neurons, primary sensory neurons to regulate meal size, exhibit decreased CCK sensitivity when the gastrointestinal tract is colonized with high-fat-type microbiome, leading to increased food intake and body weight (Kim et al., 2020). LPS leakage due to increased inflammatory microbiome including Bacteroides and Prevotella also impairs CCK-induced satiety, and promotes food intake and excessive weight gain in aging mice (Rubio et al., 2021). Therefore, decreased CCK signaling caused by gut microbial dysbiosis may exacerbate emotional and cognitive impairment associated with obesity and metabolic syndrome. Except for binding to local vagally expressed receptors, periphery CCK can also act on CCKB receptors, which is widely expressed in limbic system (including the hippocampus and the prefrontal cortex), and thus directly affect emotion and cognition (Ballaz et al., 2020). Despite multiple studies showing consistent cognitive enhancing effects by CCKB receptor activation (Taghzouti et al., 1999; Hebb et al., 2005; Plagman et al., 2019), its role is virtually unknown in BD. Only one study by Sears et al. (2013) found associations between suicide attempt and 12 SNPs of CCKB receptors in BD patients. Therefore, further investigation into the role of CCK in cognitive function in BD is warranted.

Gut microbiota can produce most of the neurotransmitters or neuromodulators in the human brain, including dopamine (which can be produced by Bacillus and Serratia), norepinephrine (produced by Escherichia, Bacillus, and Saccharomyces), serotonin (produced by Candida, Streptococcus, Escherichia, and Enterococcus), or GABA (produced by Lactobacillus and Bifidobacterium) (Donia and Fischbach, 2015; Strandwitz, 2018). These neurotransmitters mediate gut-brain signals, communicating with various regions of the brain, including the frontal cortex, limbic system, and autonomic and neuroendocrine centers, regulating not only appetite-related sensations but also emotion, cognition, and behavior (Holzer, 2016).

Dysregulation of the dopaminergic system contributes to both pathophysiology and impaired executive function in BD (Bercik et al., 2011; Kao et al., 2018). Various gut microbiota can produce dopamine, leading to dopaminergic aberrations when gut microbial dysbiosis occurs in BD (Tsavkelova et al., 2000). Mice treated with the dopamine transporter protein inhibitor GBR12909 (15 mg/kg) constitute a validated animal model for BD, whereas germ-free mice are less susceptible to GBR12909 with less mania-like behavior, suggesting that the gut microbiota contributes to the disease progression of BD via the dopaminergic system (de Miranda et al., 2020). Additionally, para-Cresol-treated mice with social-behavioral deficits have a similar gut microbial profile to that of BD patients (e.g., depleted Clostridiales), with a reduction in excitability and number of evoked action potentials of dopamine neurons in the ventral tegmental area (Bermudez-Martin et al., 2021). The increase in Bacteroides and Prevotella, which is also observed in BD patients, is negatively associated with dopamine transporter expression in the brain (Hartstra et al., 2020). The application of probiotics has been demonstrated to prevent cognitive impairment by regulating the gut microbiota, thereby increasing serotonin, dopamine, and GABA levels and restoring neuronal impairment (Song et al., 2021). The evidence presented above strongly supports the interaction between the dopaminergic system and gut microbiota, underlying the cognitive impairment in BD.

Serotonin, also known as 5-HT, is significantly correlated with cognitive function in BD patients (Chou et al., 2012). In clinical studies, tryptophan levels and the kynurenine-dependent tryptophan index were reduced in patients with bipolar mania and were positively correlated with the Young Mania Rating Scale and Brief Psychiatric Rating Scale scores (Myint et al., 2007). Reduced 5-HT levels in the brain lead to impaired cognitive flexibility, which is the characteristic type of cognitive impairment in BD (Evers et al., 2007). Notably, the gut microbiota can divert tryptophan metabolism to the production of kynurenine instead of 5-HT by activating indoleamine 2, 3-dioxygenase and tryptophan 2,3-dioxygenase (Badawy, 2017). A lack of aryl hydrocarbon receptor ligands in the intestinal contents of germ-free mice has been consistently observed (Lamas et al., 2016). In germ-free mice, there are also relatively low plasma kynurenine levels accompanied by increased indoleamine 2,3-dioxygenase activity, and kynurenine levels increase in these animals after recolonization of the gut microbiota (Clarke et al., 2013; Van der Leek et al., 2017). The altered dynamic of the tryptophan metabolic pathway has also been implicated in a study of postmortem anterior cingulate gyrus in BD patients, showing increased tryptophan 2,3-dioxygenase activity and kynurenine levels (Miller et al., 2006). Thus, the activation of the kynurenine pathway by gut microbiota dysbiosis consequently increases tryptophan consumption, which potentially contributes to the reduced 5-HT neurotransmission in BD patients, a statement being supported by studies focusing on comorbid irritable bowel syndrome (IBS) in BD patients (Tseng et al., 2016) (Figure 3). IBS, characterized by abdominal pain, bowel movement disorders, and gut microbial dysbiosis, is considered to be a valid contributor to BD (Castellini et al., 2016; Rosenblat et al., 2020). A follow-up study showed that IBS patients in clinical remission still experience persistent attention impairment, with a continuous increase in plasma IL-6 levels, the kynurenine-to-tryptophan ratio, and inactivation of the cortisol awakening response (Clarke et al., 2020). Acute tryptophan depletion significantly reduces plasma tryptophan and 5-hydroxyindoleacetic acid in IBS patients, inducing a negative shift in affective memory but without significant changes in mood (Kilkens et al., 2004). The evidence above suggests that 5-HT regulation by the gut microbiota may be involved in cognitive impairment in BD patients.

The gut microbiota profoundly affects peripheral GABA levels. Multiple organisms of gut microbiota are also involved in GABA synthesis, including a range of Bifidobacterium, Lactobacillus, and 16 intestinal Bacteroides strains (Barrett et al., 2012; Luck et al., 2021; Otaru et al., 2021). In germ-free animals, there is a significant reduction in GABA levels in the gut, serum, and brain (Matsumoto et al., 2013). Intestinal and peripheral GABA levels may influence the central nervous system through the gastrointestinal vagal nervous system. Bravo et al. (2011) found that treatment with Lactobacillus rhamnosus increases GABA(B1b) messenger RNA expression in cortical (cingulate and prefrontal) regions and decreases expression in the hippocampus and amygdala, consequently reducing corticosterone levels and manic-like behaviors. Moreover, the neurochemical and behavioral effects disappeared in vagotomized mice. Lactobacillus casei also stimulates gastrointestinal afferent vagal activity and inhibits stress-induced activation of cells producing pro-adrenocorticotropic hormone in the hypothalamic paraventricular nucleus in a dose-dependent manner, ameliorating somatic symptoms induced by learning stress (Takada et al., 2016). As reviewed by Wagner-Skacel et al. (2020), Lactobacillus has been significantly associated with the circadian rhythm in BD patients, linking a GABA disorder caused by the gut microbiota to cognitive function in BD. Thus, the gut microbiota may affect the GABA system mostly through the intestinal vagus nerve and modulate stress-related behavioral and cognitive functions in BD patients.

Multiple studies have demonstrated the versatility of glutamate, the agonist of N-methyl-d-aspartate receptor (NMDAR), as the foremost excitatory neurotransmitter in the central nervous system and modulator of gastrointestinal metabolism (Baj et al., 2019). A systematic review conducted by Reddy-Thootkur et al. (2020) shows that hippocampal glutamate levels are increased in BD patients, but no associations between glutamate metabolite levels and memory performance are detected. Considering that many of those included studies suffered from small sample sizes, the relationship between glutamate and cognitive impairment in BD require further exploration (Reddy-Thootkur et al., 2020). A postmortem study in 10 BD patients revealed significant lower protein and mRNA levels of NMDAR, indicating the presence of excitotoxicity induced by abnormal glutamatergic signaling in BD frontal cortex (Rao et al., 2010). Although there is evidence in support of the impact of the glutamatergic system in cognitive decline and disease progression in BD patients, information regarding the possible ways glutamate (either from dietary sources or microbial activities) may influence cognitive function in BD are still scarce. A Pilot study noted that the plasma and fecal glutamate levels, influenced by relative abundance of certain bacterial families, are negatively associated with cognitive function including processing speed, mental flexibility and executive function (Palomo-Buitrago et al., 2019). Dietary or luminal glutamate may also activate vagal afferents which directly or indirectly influence brain areas including the cerebral cortex, limbic system, hypothalamus and basal ganglia (Kondoh et al., 2009). From the foregoing, further study is needed to investigate the impact of a dietary glutamate and gut-endogenous glutamate in cognitive function in BD.

As an important part of monoamine metabolism, the role of D-amino acids in the gut-brain axis is gaining wider attention. Gut microbiota contributes to the host pool of D-amino acids via intrinsic amino acid racemases within certain gram-negative microbiome (Radkov and Moe, 2014). Kawase et al. (2017) found that gut microbiota could modulate the metabolism of D-amino acids in the brain. They noted that D-aspartic acid, D-serine were higher in some brain regions of GF mice than in those of SPF mice, indicating that gut microbiota may regulate the activity of aspartic acid racemase and serine racemase in the host brain (Kawase et al., 2017). D-serine is an endogenous ligand for NMDAR and thus play a key role in synaptic plasticity (Schell, 2004). Long-term potentiation, which underlies learning and memory, depends on calcium dependent release of D-serine from astrocytes in adult rat CA1 pyramidal hippocampus cells (Henneberger et al., 2010). Genetic association studies convince the role of D-serine in the pathology of BD, showing an association between BD with the gene G72, whose product activates the D-serine degrading enzyme (Chen et al., 2004; Müller et al., 2011). Furthermore, ketamine metabolites (rac)-dehydronorketamine and (2S,6S)-hydroxynorketamine decrease intracellular D-serine concentrations in a concentration dependent manner in PC-12 cells (Singh et al., 2016). Previous study has found potential pro-cognitive effects with intravenous subanesthetic ketamine in BD patients (Zhou et al., 2021). Based on these findings, we suggest that gut microbiota may influence synaptic D-serine availability and thus modulate cognitive function in BD patients.

BDNF can also promote the growth and development of neurons as a neurotrophic factor. Alterations in a 5-HT receptor, BDNF, and NMDA receptor subunit expression in the hippocampal region have been demonstrated in germ-free mice (Bercik et al., 2011). However, studies of BDNF levels and specific functional and behavioral alterations in germ-free or antibiotic-treated animal models remain controversial (Holzer, 2016).

Changing one’s dietary habits is the easiest method to modulate gut microbiota. Gut microbial dysbiosis is a key factor of cognitive impairment in diet-induced obesity (Desbonnet et al., 2014). A dietary survey of 97 BD patients showed that BD patients had greater intake levels of processed meat and sugar, fat, and salt (Davison and Kaplan, 2012). Long-term consumption of a high-fat diet increases the Firmicutes-to-Bacteroides ratio, which is associated with obesity (Shi et al., 2015; Ussar et al., 2015). A high-fat diet also led to insulin resistance and hyperglycemia in diet-induced obese mice who exhibited neurotransmitter disorders, including increased GABA and decreased tryptophan levels (Scott et al., 2020). Further, a high-fat diet can cause a significant decrease in tyrosine phosphorylation of insulin receptors, accompanied by an increase in inflammatory response signals (e.g., nuclear factor kappa-light-chain-enhancer of activated B-cells, c-Jun N-terminal kinase) in whole-brain lysate and a decrease in synaptic plasticity, leading to learning and memory impairment (Kothari et al., 2017).

In contrast, healthy diets have shown therapeutic potential for neurocognitive disorders. The Mediterranean diet, characterized by a high intake of fruits, vegetables, nuts, whole grains, and high-protein foods (i.e., fish), can help to reduce intestinal inflammation, cognitive impairment, and the risk of dementia (Pistollato et al., 2018). People at high risk of cardiovascular diseases also showed higher scores on the mean Mini-Mental State Examination and Cognitive Dysfunction Test after long-term consumption of the Mediterranean diet (Martínez-Lapiscina et al., 2013). In terms of dietary composition, a diet high in protein and saturated fat contributed to a greater abundance of Bacteroides (Wu et al., 2011). Fruits and vegetables are rich in dietary fiber. Microbiota-accessible carbohydrates found in dietary fiber increase the richness and α-diversity of the gut microbiota and inhibit the hippocampal glial activation and neuroinflammation induced by a high-fat diet. In turn, these carbohydrates improved the performance of mice in nest-building and temporal order memory tests (Shi et al., 2020). Perez (2018) found that BD patients have lower compliance with the Mediterranean diet and higher biomarkers of insulin resistance compared to the healthy population. Therefore, dietary management of BD patients is crucial for reducing the risk of metabolic disorders and cognitive impairment in clinical practice.

The ketogenic diet is another dietary pattern that has attracted increasing attention in the treatment of neuropsychiatric diseases. Several studies have confirmed the role of the ketogenic diet in improving comorbidities and cognitive function in BD patients. The ketogenic diet reduces hyperinsulinemia in BD patients by alleviating mitochondrial dysfunction mediated by impairment of the phosphatidylinositol-3 kinase/protein kinase B/hypoxia-induced factor-1α signaling pathway (Campbell and Campbell, 2020). The ketogenic diet may affect the function of vesicular glutamate transporters and EAAT, Na+, K + -ATPase, Kir4.1, aquaporin-4, Cx34, and KATP channels by affecting the glutamate-glutamine cycle and glutamate synthase activity in astrocytes, thus reducing mild cognitive impairment in BD patients (Morris et al., 2020). Several studies have confirmed the role of the gut microbiota in the ketogenic diet, although only a few studies have discussed whether the gut microbiota is involved in the cognitive improvement effect of the ketogenic diet in BD patients. The ketogenic diet protects against seizures by regulating gut microbiota colonization, which increases GABA and glutamate levels in the hippocampus (Olson et al., 2018). In AD patients with mild cognitive impairment, a 6-week ketogenic diet increased the abundance of Enterobacteriaceae, Akkermansia, Slackia, Christensenellaceae, and Erysipelotriaceae and the synthesis of butyrate, accompanied by an improvement in cognitive function (Nagpal et al., 2020).

Probiotics are live microorganisms that offer non-specific benefits to the health of the host. Some of them may also modulate functions of the central nervous system, reducing psychiatric symptoms and improving cognitive function in patients, and are therefore known as psychobiotics (Evrensel et al., 2019). Clinical studies have demonstrated the role of probiotics in promoting gut microbial diversity and improving cognitive function in BD. Probiotics, including Lactobacillus rhamnosus strain GG and Bifidobacterium animalis subsp. lactis strain Bb12, reduce the risk of re-hospitalization among BD patients (Dickerson et al., 2018). A cohort study conducted by Reininghaus et al. (2018) showed a significant improvement in attention and psychomotor processing speed on the Digit Symbol Test and executive function on the Trail Making Test B in BD patients after long-term administration of a probiotic supplement, mainly containing organisms of the Lactobacillus genera and Bifidobacterium genera.

Because some prebiotics support the growth of specific gut microbiota with psychophysiological effects, some have been designated as psychobiotics, including fructooligosaccharides, inulin, and galactooligosaccharides (Sarkar et al., 2016). Currently, there are limited studies addressing the effect of prebiotics in BD patients, but studies in animal models confirm their potential for cognitive improvement. Prebiotics might suppress inflammation that affects cognitive function. Chitosan oligosaccharides effectively reduced learning and memory impairment in an AD model by inhibiting oxidative stress and reducing the release of pro-inflammatory factors, such as IL-1 and TNF-α (Jia et al., 2016). Prebiotics also regulate the synthesis of gastrointestinal hormones. Four weeks of supplementation with prebiotics increased the expression of anorexigenic gastrointestinal hormones, such as peptide tyrosine-tyrosine, GLP-1, and leptin, while decreasing levels of ghrelin and other anorexigenic hormones and helping to improve learning and memory function in schizophrenia patients (Kao et al., 2018). Weight gain induced by antipsychotics, including olanzapine and risperidone, can also be reduced by prebiotics. Intake of the prebiotic galactooligosaccharide mixture significantly reduced olanzapine-induced weight gain, possibly in association with an increased and decreased number of organisms belonging to the Bifidobacterium and Firmicutes genera, respectively, together with increased cortical phospho-NMDA receptor 1 levels and decreased plasma TNF-α levels. Therefore, these effects suggest that prebiotics may prevent the metabolic and cognitive side effects of olanzapine.

The possible mechanisms by which the gut microbiota modulates energy metabolism and cognitive function by influencing gastrointestinal hormones have been reviewed above. Gastrointestinal hormone analogs may also improve cognitive function in a mode similar to that of gastrointestinal hormones. Liraglutide, the GLP-1 receptor agonist, improved D-amphetamine-induced mania-like symptoms and working and recognition memory impairment in a BD mouse model (Chaves Filho et al., 2020). In non-diabetic BD patients, liraglutide also showed beneficial effects in several cognitive domains, including auditory verbal learning, working memory, and attention (Mansur et al., 2017). Therefore, the neuroprotective effect of liraglutide illustrates the potential for gastrointestinal hormone analogs to serve as promising adjunctive tools for BD treatment.

FMT treatment involves the injection of filtrate feces from a healthy donor into a patient. FMT not only increases the microbial diversity but also provides long-term implantation of donor strains compared to the short-term impact on the gut microbiota achieved by changes in the diet or the addition of psychobiotics (Weingarden and Vaughn, 2017). In animal research, FMT treatment reversed cognitive impairment in AD model mice by altering the gut microbial composition and SCFA profile and increasing synaptic plasticity. Currently, few clinical studies have focused on the feasibility and efficacy of FMT in BD patients, although Hinton (2020) claimed that depressive symptoms disappeared with weight loss in a BD patient who had experienced 9 sessions of FMT treatment. Therefore, FMT may have a potential role in improving the cognitive function of BD patients, but further clinical experiments are warranted.

The vagus nerve establishes one of the important connections between emotional and cognitive areas of the brain and gut functions. Vagal afferent fibers express receptors for multiple gastrointestinal hormones (GLP-1, CCK, peptide YY, ghrelin, etc.), neurotransmitters (dopamine, GABA, NE, 5-HT, etc.), and gut microbial metabolites (SCFAs, LPS, etc.), to transfer microbiota signals to the central nervous system (Bonaz et al., 2018; Breit et al., 2018). Brain pathway activated by oral administration of Campylobacter jejuni, which has been proved to influence behavior and brain functions at subclinical doses, has been mapped using c-fos expression as a marker of neuronal activation (Gaykema et al., 2004). In this study, brain activation was observed in the nucleus tractus solitarius, the vagal afferent ending, and the projections of the nucleus tractus solitarius including parabrachial nucleus, paraventricular nucleus of the hypothalamus, amygdala and thalamus, indicating vagally-mediated microbiota effect on mood and cognitive function (Gaykema et al., 2004). Consistently, Bravo et al. (2011) found that treatment with Lactobacillus rhamnosus decreases GABA(B1b) messenger RNA expression in the hippocampus and amygdala, consequently reducing corticosterone levels and manic-like behaviors. Moreover, the neurochemical and behavioral effects disappeared in vagotomized mice, indicating therapeutic potential for treatment targeting vagal tone.

Vagus nerve stimulation, a medical treatment that is routinely used in the treatment of epilepsy and other neurological conditions, works by applying electrical impulses to the vagus nerve (Wheless et al., 2018; González et al., 2019). A 5-year prospective research in patients with treatment-resistant bipolar depression showed that treatment with vagus nerve stimulation was associated with better medication response and significantly greater mean reduction in suicidality compared to treatment-as-usual (McAllister-Williams et al., 2020). Chronic vagus nerve stimulation also produces sustained clinical and cognitive improvements in BD patients in a treatment-resistant depressive episode (McAllister-Williams et al., 2020). Therefore, vagus nerve stimulation seems to be a promising adjunctive therapy for cognitive impairment in BD patients.