INTRODUCTION
Mental and neurodevelopmental disorders, including autism spectrum disorder (ASD), attention‐deficit/hyperactivity disorder (ADHD), schizophrenia, anxiety disorders, post‐traumatic stress disorder (PTSD) and depression, impose a major global burden (GBD 2019 Mental Disorders Collaborators, 2022). The World Health Organization (WHO) estimates that mental illness is among the leading causes of disability‐adjusted life years (DALYs), affecting more than 970 million people worldwide. Although pharmacological options have expanded, widely used treatments such as atypical antipsychotics, benzodiazepines, selective serotonin reuptake inhibitors and mood stabilizers are still constrained by delayed onset, partial or non‐response, adverse effects and poor adherence, particularly in treatment‐resistant populations (Accettulli et al., 2022). These limitations have increased interest in integrative approaches that target biological systems beyond canonical monoaminergic pathways.
The microbiota–gut–brain axis (MGBA) is a bidirectional network connecting the gastrointestinal (GI) tract with immune and endocrine signalling and the central nervous system (CNS). Growing evidence suggests that the gut microbiome and its metabolites can influence cognition, emotion regulation, synaptic signalling and neurodevelopment (Wu et al., 2025). Dysbiosis, defined as disruption of gut microbial communities, has been linked to multiple psychiatric and neurological conditions, including depression, ASD, schizophrenia, Parkinson's disease and generalized anxiety disorder (Generoso et al., 2020), encouraging investigation of microbiome‐targeted strategies in psychiatric research and care (Moerkl et al., 2020).
Psychobiotics were initially defined by Dinan et al. as live organisms or microbe‐derived substances that, when ingested in adequate amounts, confer mental health benefits via the MGBA (Dinan et al., 2013). In contrast to conventional probiotics that primarily target GI health, psychobiotics are framed as microbiome‐based agents that may act through neuroactive pathways, including neurotransmitter‐related signalling (e.g., GABA, serotonin), immunomodulation, hypothalamic–pituitary–adrenal (HPA) axis regulation, and support of neuroplasticity (Del Toro‐Barbosa et al., 2020). The concept has broadened to include not only live strains but also postbiotics, synbiotics, para‐psychobiotics and prebiotics, expanding the therapeutic and safety spectrum, particularly for immunocompromised or vulnerable populations (Zielińska et al., 2022). An explanation of the terms used is provide in Box 1.
Preclinical studies indicate that selected preparations can modulate synaptic plasticity, neuroinflammation, stress‐hormone responses and behaviour, with some effects depending on vagus nerve integrity (Andreeva et al., 2022; Jia et al., 2024). Early randomized controlled trials (RCTs) in humans report improvements in perceived stress, anxiety, sleep quality, depressive symptoms, and selected outcomes in ASD and ADHD, but effects are typically modest and heterogeneous across strains, doses, populations and follow‐up periods (Magalhães‐Guedes et al., 2020). Translation is further limited by short trial durations (often 4–12 weeks), small and diverse samples, inconsistent strain‐level reporting, and host‐dependent variability related to diet, genetics, concomitant medications (especially antibiotics and psychotropics) and baseline microbiota composition (Mosquera et al., 2024). Regulatory ambiguity and limited standardization in labelling and classification of microbiome‐based therapeutics also slow clinical adoption (Andreeva et al., 2022).
With an emphasis on neurobiological mechanisms, strain‐specific profiles, clinical applications and translational challenges, this narrative review evaluates current evidence and clinical promise in psychobiotic research. We synthesized findings across molecular biology, psychiatry and microbiome science, based on searches of PubMed, Scopus, Web of Science and Google Scholar for studies published between 2010 and 2025 using terms such as ‘psychobiotics’, ‘microbiota–gut–brain axis’, ‘probiotics’, ‘anxiety’, ‘depression’, ‘ASD’, ‘ADHD’, ‘neuroinflammation’, and ‘BDNF’. We prioritized original preclinical, clinical and mechanistic studies that assessed neuroimmune, neuroendocrine, or behavioural outcomes; used review papers selectively for background; and excluded studies with unclear strain identification, limited mechanistic detail, or a primary focus outside psychiatry, to balance depth with translational relevance and maintain transparency in study selection.
THE MICROBIOTA–GUT–BRAIN AXIS: NEUROIMMUNE AND NEUROENDOCRINE INTERFACES
The MGBA is a bidirectional communication system linking the GI microbiota with the CNS. Psychobiotics exert their neuroactive and therapeutic effects through four interrelated routes that together provide the biological background for all subsequent sections (Del Toro‐Barbosa et al., 2020; Garzone et al., 2025) (Figure 1).
Gut microbiota–brain communication occurs through parallel pathways involving microbial metabolites (e.g., neurotransmitters and SCFAs) that can influence the CNS, vagus nerve signalling from the gut to the brain, and circulating mediators (hormones and immune‐derived cytokines) that affect brain function after entering systemic circulation. BBB, blood–brain barrier; SCFAs, short‐chain fatty acids; VN, vagus nerve.
Neural pathway: Vagus nerve and enteric nervous system
Neural signalling is primarily conveyed through the vagus nerve and the enteric nervous system (ENS). Afferent vagal fibres project from the gut to emotion‐ and memory‐related regions such as the amygdala and hippocampus. In a preclinical, vagus‐dependent model, Lactobacillus rhamnosus JB‐1 altered GABA receptor expression and reduced anxiety‐like behaviour. Subdiaphragmatic vagotomy abolished this effect, confirming that an intact vagus is necessary for this psychobiotic action (Bravo et al., 2011; Neska et al., 2024). This strain is a clear example of a neural and vagal route.
Immune pathway: Gut barrier, lipopolysaccharide and cytokines
Gut microbes maintain systemic immune tone by supporting epithelial integrity and modulating cytokine secretion (Mazziotta et al., 2023). Dysbiosis can increase intestinal permeability and facilitate the translocation of pro‐inflammatory molecules such as lipopolysaccharide (LPS). This induces peripheral cytokine release, including interleukin (IL)‐6 and tumour necrosis factor α (TNF‐α), which has been associated with depressive symptoms and cognitive decline (Chen et al., 2022). Preclinical work in an LPS‐induced model of neuroinflammation showed that probiotic intervention ameliorated LPS‐driven injury and was accompanied by favourable changes in Aβ‐related proteins (Aβ1–42, APP, γ‐secretase) and increased brain‐derived neurotrophic factor (BDNF) relevant to fetal neurodevelopment (Kar et al., 2022). Human and animal studies with Lactobacillus helveticus and Bifidobacterium longum also report decreased pro‐inflammatory cytokines and increased IL‐10, which indicates that these strains act mainly through an immune‐modulatory route (Zakharova et al., 2022).
Endocrine pathway: HPA axis regulation
Stress activates the HPA axis and results in altered cortisol levels, changes in GI motility and mood disturbances (Oroojzadeh et al., 2022). Germ‐free rodents display exaggerated corticosterone responses, which are normalized after microbial colonization. This shows that a normal microbiota is required for correct HPA reactivity (Sharma et al., 2021). In both animal and human studies, Lactobacillus casei Shirota and Clostridium butyricum have been reported to lower cortisol and reduce stress reactivity, which indicates a microbiota‐sensitive endocrine route (Aziz et al., 2024; Samtiya et al., 2022). In humans, B. longum 1714 has also been shown to reduce perceived stress and cortisol and to improve cognition, so this clinically studied strain can be placed in the same HPA and stress regulation pathway (Allen et al., 2016).
Metabolic pathways: Microbial short‐chain fatty acids and neuroplasticity
Commensal bacteria ferment dietary substrates to generate short‐chain fatty acids (SCFAs) such as acetate, propionate and butyrate. These metabolites can enter the circulation, cross the blood–brain barrier (BBB) and affect microglial activation, neuroinflammation and the expression of neurotrophic factors, particularly BDNF, in hippocampal and cortical regions (Gupta et al., 2026). Multiple preclinical studies have reported antidepressant‐like or pro‐cognitive effects that are consistent with SCFA‐driven support of synaptic plasticity and neurogenesis (Roy et al., 2024). Psychobiotic strategies that increase endogenous SCFA‐producing taxa, for example by using prebiotics or synbiotics, therefore act mainly through this metabolic arm (Binda et al., 2024).
Integrative view
The four routes outlined above explain how the MGBA can influence behaviour, emotion and cognition (Gupta et al., 2026). Unlike many psychiatric drugs that target a single CNS receptor, psychobiotics can rebalance host physiology at several regulatory levels at the same time, which facilitates their use alongside conventional treatments (Dacaya et al., 2025; Sharma et al., 2021) (Table 1, Figure 2).
Mechanistic routes linking selected psychobiotic strains to the microbiota–gut–brain axis and neurobiological readouts. The left panel lists exemplar strains that act through neural, immune, endocrine and metabolic pathways, including vagal signalling, cytokine modulation, HPA axis regulation and short chain fatty acid (SCFA) production. These routes converge on limbic and prefrontal brain regions and are associated with changes in GABA receptors, BDNF, inflammatory cytokines, cortisol, and behavioural outcomes such as anxiety, mood, cognition and gastrointestinal symptoms. Icons indicate whether supporting data come from rodent models or human studies. BDNF, brain derived neurotrophic factor; ENS, enteric nervous system; HPA, hypothalamic–pituitary–adrenal; IL, interleukin; LPS, lipopolysaccharide; MGBA, microbiota–gut–brain axis; SCFA, short chain fatty acid; TNF alpha, tumour necrosis factor alpha.
| Strain/class | Primary pathway(s) | Model (rodent/human) | Biomarker/readout | Key citation |
|---|---|---|---|---|
| JB‐1Lactobacillus rhamnosus | Neural (vagus), GABA receptor modulation | Rodent | ↓ anxiety‐like behaviour, altered GABA/GABAexpressionAB | Bravo et al. () [2011] |
| 1714Bifidobacterium longum | Endocrine (HPA), stress modulation, cognition | Human | ↓ perceived stress, ↓ cortisol, altered brain activity | Allen et al. () [2016] |
| +(immune active strains)Lactobacillus helveticusB. longum | Immune, gut barrier, cytokine shift | Human/rodent | ↓ IL‐6, ↓ TNF‐α, ↑ IL‐10 | Mohammadi et al. () [2019] |
| /ShirotaClostridium butyricumL. casei | HPA normalization, endocrine–GI link | Human/rodent | ↓ cortisol, ↓ stress reactivity | Takada et al. () [2016] |
| SCFA‐producing species/postbiotics (butyrate) | Metabolic, SCFA → BDNF, microglia | Rodent | ↑ BDNF, ↓ neuroinflammation, antidepressant‐like effects | Yamawaki et al. () [2018] |
| Probiotics in LPS‐induced neuroinflammation | Immune → Aβ processing → BDNF | Rodent, maternal–fetal | ↓ LPS injury, changes in Aβ1‐42, APP, γ‐secretase, ↑ BDNF | Kar et al. () [2022] |
THE MECHANISTIC VARIETY OF PSYCHOBIOTICS: FROM MICROBIAL METABOLITES TO LIVE STRAINS
Building on the MGBA framework outlined earlier, psychobiotics can be grouped into product classes that differ in viability, safety profile and their predominant route of action. While the original concept emphasized live probiotic strains, current usage also includes prebiotics, synbiotics, postbiotics and inactivated para‐psychobiotics. Across classes, effects ultimately converge on the same MGBA signalling arms (neural, immune, endocrine and metabolic), but the relative weighting of these routes varies by formulation and host context.
Probiotic psychobiotics (live strains)
Probiotics are live microorganisms that confer a health benefit when administered in adequate amounts. Well‐characterized Lactobacillus and Bifidobacterium strains have been linked to neuroactive effects through pathways that include metabolite production (e.g., SCFAs, serotonin precursors, GABA), modulation of vagal signalling, HPA‐axis reactivity and cytokine balance (Allen et al., 2016; Bravo et al., 2011). A central translational point is that these effects are strain‐specific; findings for L. rhamnosus JB‐1 or B. longum 1714 cannot be generalized across a genus or species. Accordingly, studies should report the exact strain designation and daily dose/colony‐forming units (CFU) to support reproducibility and interpretation (Wang et al., 2019).
Prebiotics
Prebiotics are non‐digestible substrates (e.g., fructooligosaccharides, galactooligosaccharides) that selectively promote beneficial taxa (Dacaya et al., 2025). Their psychobiotic impact is largely indirect because they enrich endogenous SCFA‐producing communities (e.g., Faecalibacterium prausnitzii), thereby supporting neuroplasticity, microglial homeostasis and anti‐inflammatory signalling via the metabolic arm of the MGBA (Ramadan et al., 2025). Because their effects depend on resident microbiota, baseline composition and diet are expected moderators of response.
Synbiotics
Synbiotics combine a defined probiotic strain with a compatible prebiotic substrate. This pairing can enhance colonization, metabolite output and formulation stability, which may explain stronger stress‐ or mood‐related effects in some trials compared with probiotics alone (Aziz et al., 2024). Rather than introducing a distinct biological pathway, synbiotics primarily amplify and prolong MGBA signalling already attributed to their probiotic and prebiotic components.
Postbiotics
Postbiotics are non‐viable microbial products or metabolites, including SCFAs (e.g., butyrate, acetate), tryptophan derivatives, and structural components such as peptidoglycan (Gupta et al., 2026). Because they do not require live colonization, postbiotics may be preferable in vulnerable or immunocompromised populations (Wadan et al., 2025). Mechanistically, butyrate has been linked to increased hippocampal BDNF, improved synaptic plasticity and reduced anxiety‐like behaviour in rodents, consistent with metabolic and neurotrophic MGBA signalling (Ramadan et al., 2025). Microbe‐derived products may also reduce LPS‐related immune activation by supporting gut barrier function and downstream cytokine regulation, aligning with immune‐mediated routes reported in maternal–fetal neuroinflammation models (Kar et al., 2022).
Para‐psychobiotics (inactivated cells)
Para‐psychobiotics are non‐viable but structurally intact microbial preparations that retain bioactive surface features (Lai & Shen, 2023). They may reduce the risk of translocation while still engaging gut and immune pattern‐recognition pathways. For example, heat‐killed Lactobacillus gasseri has been associated with improved sleep and reduced anxiety in stressed individuals, despite the absence of live colonization (Sharma et al., 2021). Their dominant effects are typically framed as immune‐centric, with secondary neuroendocrine consequences.
Implications
Together, these categories highlight that ‘psychobiotics’ comprise multiple intervention types that converge on MGBA communication channels while differing in feasibility and safety. To reduce categorical drift and improve translational value, future studies should specify the product class, the exact strain (or defined metabolite/constituent), the administered dose and the most plausible dominant pathway, rather than attributing effects broadly to ‘probiotics’ (Aziz et al., 2024).
THE EFFECTS OF STRAIN‐SPECIFIC PSYCHOBIOTICS: FUNCTIONAL AND GENOMIC DESCRIPTION
The specificity of the microbial strain has a substantial impact on the therapeutic efficacy of psychobiotics (Wadan et al., 2025). Different strains of the same species can show distinct metabolic outputs, behavioural effects and genomic profiles, which makes precise strain‐level identification and functional characterization essential in both clinical and experimental work (Gupta et al., 2026). One of the best‐characterized strains is L. rhamnosus JB‐1, which shows robust anxiolytic and antidepressant‐like effects in mice. Bravo et al. demonstrated that JB‐1 reduces corticosterone in chronically stressed animals and modulates GABAA and GABAB receptor expression in the amygdala and hippocampus; these effects disappear after vagotomy, indicating a vagus‐dependent mechanism (Bravo et al., 2011; Hunjan et al., 2025). Importantly, this neuroactive profile is not shared by all L. rhamnosus strains, which highlights the distinctiveness of JB‐1 (Yadav et al., 2022).
Bifidobacterium longum 1714 is another well‐studied psychobiotic with effects demonstrated in both rodents and humans. In clinical trials it improved memory‐related performance, reduced cortisol, and lowered daily perceived stress, and functional magnetic resonance imaging studies confirmed altered activity in brain regions involved in emotional processing (Gupta et al., 2026). Additional strains, including Bifidobacterium breve CCFM1025 and L. helveticus R0052, have shown neurotrophic and anti‐inflammatory actions, such as reduced IL‐6, increased hippocampal BDNF, improved cognition, and attenuated microglial activation in neuroinflammatory models (Gupta et al., 2026). Psychobiotic effects may also be enhanced in multispecies formulations. For example, the combination of L. helveticus R0052 and B. longum R0175 has been reported to alleviate depressive and anxiety symptoms in humans, probably through complementary actions on HPA‐axis activity, neurotransmission and inflammatory cytokines (Kim et al., 2023).
Whole‐genome sequencing, metabolomics, and neurophenotyping are now being used to define composite ‘psychobiotic signatures’ that include gene clusters for short‐chain fatty acid synthesis, GABA production and tryptophan metabolism (Li et al., 2024). These signatures may eventually support predictive selection of strains based on desired immunological or neurochemical outcomes. Overall, strain specificity is a central determinant of psychobiotic efficacy rather than a minor detail. Developing reliable and personalized microbial interventions for mental health will require rigorous validation of functional properties at the strain level in addition to genus and species classification (Yousuf et al., 2025).
CLINICAL AND PRECLINICAL PERSPECTIVES ON PSYCHOBIOTICS IN MOOD AND ANXIETY DISORDERS
Mood and anxiety disorders are the most frequently explored indications for psychobiotic interventions, partly because they overlap mechanistically with low‐grade inflammation, gut dysbiosis and dysregulation of the HPA axis (Tiwari & Paramanik, 2025).
Preclinical evidence
Rodent studies show relatively consistent signals for a small group of strains. L. rhamnosus JB‐1 reduced anxiety‐like behaviour and was associated with vagus‐dependent modulation of GABA receptors in the amygdala and hippocampus (Lachmansingh et al., 2024), which is in line with the neural route outlined earlier. B. breve CCFM1025 improved cognition and reduced microglial activation in neuroinflammatory models (Oroojzadeh et al., 2022). These behavioural effects typically co‐occurred with lower circulating IL‐6 and TNF‐α and higher hippocampal BDNF, which supports an anti‐inflammatory and pro‐neuroplastic profile (Zubareva & Melik‐Kasumov, 2021). Because these outcomes have been demonstrated only for specific strains under defined experimental conditions, they should not be generalized to all Lactobacillus or Bifidobacterium species.
Clinical trials in depression, anxiety and related psychiatric conditions
Randomized trials in major depressive disorder (MDD) and stress‐related states suggest that psychobiotic effects are strain‐ and context‐dependent, and are often more apparent in adjunctive or biologically dysregulated populations. In an 8‐week double‐blind RCT (n = 40), Akkasheh et al. reported greater reductions in Beck Depression Inventory (BDI) scores with a multistrain probiotic (Lactobacillus acidophilus, L. casei, Bifidobacterium bifidum; ∼2 × 109 CFU/strain/day) versus placebo, alongside improvements in metabolic and inflammatory markers (Akkasheh et al., 2016). In a three‐arm 8‐week RCT (n = 110), Kazemi et al. found that L. helveticus plus B. longum reduced depressive symptoms more than both a galactooligosaccharide (GOS) prebiotic and placebo, with modest shifts in tryptophan–kynurenine‐related indices (Kazemi et al., 2019).
Adjunctive studies indicate potential additive benefits in more severe cases, although design limitations remain. An 8‐week open‐label controlled inpatient study (n = 40) suggested that adding C. butyricum MIYAIRI 588 to antidepressants improved depression and anxiety scales more than antidepressants alone, but the absence of blinding limits inference (Miyaoka et al., 2018). In an 8‐week double‐blind adjunctive RCT in selective serotonin reuptake inhibitor (SSRI)‐treated MDD (n = 79), Lactobacillus plantarum 299v improved attention and verbal learning and altered kynurenine markers, while mood outcomes improved similarly in both arms, indicating primarily cognitive/biochemical effects rather than robust incremental antidepressant efficacy (Rudzki et al., 2019).
In healthy volunteers, B. longum 1714 reduced stress‐induced cortisol output and perceived stress and produced small neurocognitive/EEG changes after 4 weeks, consistent with stress‐modulating rather than ‘antidepressant’ effects in low‐symptom samples (Allen et al., 2016). Beyond mood and anxiety, evidence is mixed: in schizophrenia, a 14‐week double‐blind RCT (n = 65) found no improvement in Positive and Negative Syndrome Scale (PANSS) scores but reduced severe constipation risk in antipsychotic‐treated patients (Dickerson et al., 2014). In ASD in a small double‐blind crossover trial (n = 15), L. plantarum WCFS1 altered GI/microbiota measures with unclear behavioural benefit (Parracho et al., 2010), whereas an open‐label microbiota transfer therapy protocol (n = 18) reported sustained GI and autism‐scale improvements, though results remain exploratory due to the non‐randomized, unblinded, multi‐component design (Kang et al., 2017).
Overall, these trials suggest that baseline dysregulation (e.g., clinically significant depression, chronic stress or GI disturbance) and intervention parameters (strain identity, dose, duration) shape whether benefits are primarily emotional, cognitive or somatic. Key study parameters are summarized in Table 2.
| Study (population) | Design//durationn | Intervention | Daily dose | Key outcomes & significance |
|---|---|---|---|---|
| Akkasheh et al. () (MDD) [2016] | DB RCT;= 40; 8 weeksn | Multistrain probiotic (,,)L. acidophilusL. caseiB. bifidum | ∼2 × 10CFU of each strain/day9 | Larger BDI reduction vs. placebo + improved insulin/HOMA‐IR, hs‐CRP, glutathione. |
| Kazemi et al. () (MDD) [2019] | DB 3‐arm RCT;= 110 (81 completers); 8 weeksn | +vs. GOS vs. placeboL. helveticusB. longum | Probiotic ∼10CFU/day (typical)9 | Probiotic reduced BDI more than placebo and GOS (≈ 0.04); modest shift in TRP–KYN markers; GOS no clear benefit.P |
| Allen et al. () (healthy) [2016] | Placebo‐controlled, within‐subject;= 22; 4 weeks placebo → 4 weeks probioticn | 1714B. longum | 1 × 10CFU/day9 | Lower stress‐induced cortisol and perceived stress; small but significant memory/EEG effects. |
| Rudzki et al. () (MDD on SSRI) [2019] | DB adjunctive RCT;= 79 (60 completers); 8 weeksn | 299vL. plantarum | ∼10–10CFU/day910 | Improved cognitive performance (APT/CVLT) + altered KYN profile; incremental antidepressant effect modest. |
| Miyaoka et al. () (TRD inpatients) [2018] | Open‐label controlled;= 40; 8 weeksn | CBM588 + antidepressant vs. antidepressant only | 60 mg/day | Greater HAM‐D/BDI/BAI reductions; response ∼70%, remission ∼35% (blinding limitation). |
| Dickerson et al. () (schizophrenia) [2014] | DB RCT;= 65 (58 completers); 14 weeksn | LGG + Bb12 | ∼2 × 10CFU/day9 | No PANSS benefit; markedly reduced severe constipation risk (hazard ratio ≈ 0.23,≈ 0.003).P |
| Parracho et al. () (ASD) [2010] | DB crossover;= 15 analysed; ∼3 week periodsn | WCFS1L. plantarum | ∼4.5 × 10CFU/day10 | Changed stool consistency/microbiota; behavioural effects unclear (small, high dropout).n |
| Kang et al. () (ASD + GI) [2017] | Open‐label single‐arm;= 18; 10 weeks + 8 weeks FUn | Microbiota Transfer Therapy (multi‐step) | ∼2.5 × 10cells/day initially → ∼2.5 × 10/day129 | ∼80% GI symptom improvement + sustained ASD score improvements and higher diversity; exploratory due to open‐label, multi‐component design. |
Heterogeneity of outcomes
Outcomes across psychobiotic trials in mood and anxiety disorders are heterogeneous and generally modest. Evidence suggests benefits are more likely in mild to moderate depression, while effects in healthy or minimally symptomatic groups are often small or non‐significant, which may reflect ceiling effects and limited baseline HPA axis or immune dysregulation (Zhang, Chen, Zhang, et al., 2023; Zhao et al., 2025). Variability is strongly driven by strain identity, viable dose, intervention duration and sample characteristics, and may also be moderated by sex distribution and host context (Zhao et al., 2025). These patterns support the existence of responder and non‐responder subgroups shaped by baseline microbiota, diet and concomitant medications, emphasizing the need for biomarker informed trial designs (Hsiao et al., 2013).
Implications for clinical translation
Taken together, current evidence supports selected, strain‐specific psychobiotics (notably within Lactobacillus, Bifidobacterium, and butyrate‐related approaches) primarily as adjuncts to standard care for anxiety and depressive disorders, with effects that remain heterogeneous across trials (Gupta et al., 2026). To improve reproducibility and enable responder stratification, future studies should report strain identity and viable dose, standardize intervention duration (e.g., 8–12 weeks), and incorporate mechanistic readouts such as cytokines, cortisol, BDNF and tryptophan–kynurenine markers (Atanasova et al., 2025).
NEURODEVELOPMENTAL AND SEVERE PSYCHIATRIC CONDITIONS ARE TREATED WITH PSYCHOBIOTICS
The treatment of neurodevelopmental disorders, including ADHD and ASD, as well as severe mental illnesses like schizophrenia and bipolar disorder, is intricate. Changes in gut microbiota are increasingly being linked to their pathophysiology, primarily through impaired gut‐brain signalling, disturbed neurotransmitter balance, and early‐life immune activation (Vasiliu, 2023a). Psychobiotics may modulate these underlying mechanisms.
Autism spectrum disorder
Both adults and children with ASD frequently have chronic inflammation, GI symptoms, and a changed gut microbiota composition, usually with more Clostridium spp. And less Bifidobacterium (Dash et al., 2022). According to Kang et al., these alterations are associated with behavioural and social impairments (Kang et al., 2017). In a mouse model of ASD (maternal immune activation), preclinical research has demonstrated that the administration of Bacteroides fragilis corrected gut permeability and restored social behaviour (Harding & Bishop, 2022). Several small‐scale clinical studies have suggested that probiotic supplements may alleviate GI symptoms, hyperactivity, and irritability. For instance, Parracho et al. discovered that L. plantarum WCSF1 improved social behaviour and attention in autistic children, even though the results were inconsistent and contingent upon the strain (Parracho et al., 2010).
Attention‐deficit/hyperactivity disorder
Executive function deficiencies and ADHD impulsivity have also been linked to the gut microbiota. The risk of developing ADHD was significantly reduced by the age of 13 through perinatal supplementation with L. rhamnosus GG, as indicated by a randomized trial conducted by Pärtty et al. This suggests long‐term neurodevelopmental modulation (Dash et al., 2022; Pärtty et al., 2015). Among the suggested mechanisms are modifications to tryptophan pathways, control of systemic inflammation and changes to dopamine metabolism (Novikova et al., 2024; O'Riordan et al., 2025).
Schizophrenia and bipolar disorder
An immunopsychiatric perspective is increasingly being used to explain severe mental illnesses such as bipolar disorder and schizophrenia, where gut dysbiosis is linked to altered blood–brain barrier permeability and systemic inflammation (Bashir & Khan, 2022). Pro‐inflammatory species are more prevalent in schizophrenia, which is frequently characterized by a reduction in microbial diversity (Vasiliu, 2023a). In a double‐blind RCT, Dickerson et al. demonstrated that a 14‐week probiotic intervention (e.g., B. lactis and L. rhamnosus) enhanced GI symptoms and reduced rehospitalization rates but had minimal impact on core psychotic symptoms (Dickerson et al., 2014). Psychobiotic trials are still in their early stages, but patients with bipolar disorder also show increased intestinal permeability and altered microbial taxa linked to mood regulation (Dash et al., 2022).
Limitations and interpretation
The evidence base for psychobiotics in these conditions is not as strong as it is for depression and anxiety, despite encouraging initial findings (Vasiliu, 2023a). The heterogeneity of patient populations, outcome measures and microbial baselines complicates the process of interpretation. In addition, the efficacy of treatment may be influenced by the timing of development, particularly in neurodevelopmental disorders, where early microbiota modulation may have enduring neurobehavioural effects (O'Riordan et al., 2025).
Considering these factors, psychobiotics should currently be viewed as a promising adjunct for neurodevelopmental and severe mental illnesses, especially in patients with prominent immune or GI comorbidity (Novikova et al., 2024) (Table 3, Figure 3).
Schematic representation of the gut–brain axis highlighting the role of psychobiotics in modulating neuropsychiatric and neurodegenerative disorders. Intake of psychobiotics influences gut microbiota composition and preserves intestinal barrier integrity, which in turn communicates with the central nervous system via the vagus nerve. Disruption of the bowel barrier is implicated in the pathogenesis of various neurological and psychiatric conditions, including depression, anxiety, autism, schizophrenia, Parkinson's disease, multiple sclerosis and Alzheimer's disease.
| Disorder | Psychobiotic strains studied | Core symptoms targeted | Proposed mechanisms | Key findings (model/trial) | Limitations | Ref. |
|---|---|---|---|---|---|---|
| ASD | ,WCSF1B. fragilisL. plantarum | Social behaviour, GI symptoms, hyperactivity | Gut barrier integrity, neuroimmune modulation, microbial diversity | restored gut barrier and social behaviourimproved attention in childrenB. fragilisL. plantarum | Small cohorts, inconsistent outcomes, strain‐dependent | Hsiao et al. () [2013] |
| ADHD | GGL. rhamnosus | Impulsivity, executive function | Dopaminergic modulation, tryptophan pathways, inflammation control | Perinatal use reduced ADHD diagnosis by age 13 | No data in treated patients; needs replication | Pärtty et al. () [2015] |
| Schizophrenia | ,B. lactisL. rhamnosus | GI distress, relapse rates | Immune regulation, BBB integrity | ↓ Rehospitalization, improved GI symptoms in RCT | No effect on core psychotic features | Dickerson et al. () [2014] |
| Bipolar disorder | Not strain‐specific | Mood instability, systemic inflammation | Gut permeability, immune markers, microbial taxa shift | Altered microbiota linked to mood regulation | No clinical trials; lacks strain precision | Painold et al. () [2019] |
| Cross‐cutting challenges | — | GI symptoms, immune dysregulation, stress reactivity | Multimodal microbiome‐host signalling | — | Population heterogeneity, timing effects, need for biomarkers | Dinan and Cryan () [2017] |
Fecal microbiota transplantation‐based interventions in neuropsychiatric disorders
Over the past decade, fecal microbiota transplantation (FMT) has been explored as a microbiome‐based therapy for several neuropsychiatric conditions. Experimental work suggests that transferring healthy donor microbiota can reduce anxiety and depression‐like phenotypes, whereas transplantation of dysbiotic microbiota from affected individuals can induce such features in recipients (Chinna Meyyappan et al., 2020). Clinically, the most studied indications include ASD, mood and anxiety disorders, and emerging work in Alzheimer's disease (AD).
Autism spectrum disorder
In ASD, open label studies of microbiota transfer therapy and related FMT protocols have reported marked and durable reductions in GI complaints and core autism symptoms, sometimes maintained or further improved at 2 year follow‐up, together with sustained increases in microbial diversity and enrichment of taxa such as Bifidobacteria and Prevotella (Kang et al., 2019; Zhang, Zhu, Wan, et al., 2023). Case reports and small observational cohorts consistently describe decreases in scores on scales such as ABC, CARS, and SRS after FMT. A 2023 systematic review of five ASD studies concluded that all showed significant post‐FMT symptom improvement, but emphasized that all were open label or observational and that no completed RCT was available at that time (Zhang, Zhu, Wan, et al., 2023). Early data from a larger trial suggest that placebo effects may be substantial (Maniscalco et al., 2025), which underlines the need for rigorously blinded, controlled studies before FMT can be considered beyond an experimental option for children with ASD and comorbid gut symptoms.
Anxiety and related disorders
Anxiety‐related phenotypes have been linked to microbiota composition in both preclinical and clinical work. In animals, microbiota from highly anxious donors can transfer anxiety‐like behaviour, whereas microbiota from resilient donors can confer stress resistance (Chinna Meyyappan et al., 2020). In humans, FMT has shown anxiolytic and antidepressant effects mainly in populations with prominent GI disturbance. For example, in patients with irritable bowel syndrome and comorbid anxiety, FMT from healthy donors improved both digestive symptoms and anxiety and depression scores 1 month after treatment (Yang et al., 2023). A recent meta‐analysis in individuals with chronic insomnia also reported improvements in sleep quality and co‐occurring anxiety symptoms within 4 weeks of FMT (Zhang et al., 2025). These findings suggest that correcting gut dysbiosis may secondarily benefit mood and anxiety, but dedicated RCTs in primary anxiety disorders are still lacking.
Depression
The antidepressant potential of FMT has been evaluated more systematically. A recent meta‐analysis of 12 RCTs including 681 participants (2019–2024) found that FMT significantly alleviated depressive symptoms compared with placebo or standard care, with effects remaining robust in sensitivity analyses (Zhang et al., 2025). One trial reported that the antidepressant benefit of FMT not only persisted but appeared to increase over time, although follow‐up durations differed between studies (Liu et al., 2024; Zhang et al., 2025). Mechanistically, FMT tends to restore microbial diversity and increase anti‐inflammatory, short chain fatty acid producing taxa, which may reduce systemic inflammation and improve neurotransmitter systems relevant to mood regulation (Liu et al., 2024). Case reports of treatment resistant depression describe marked mood improvement within weeks of adjunctive FMT after failure of conventional therapies (Liu et al., 2024; Zhang et al., 2024). Overall, the evidence suggests that microbiota transplantation can improve depression in a subset of patients (Vasiliu, 2023b), and ongoing trials are refining delivery routes and patient selection, with preliminary data hinting that colonoscopic infusion may have somewhat larger effects than oral capsules (Zhang et al., 2025).
Alzheimer's disease and cognitive decline
In Alzheimer's disease and related cognitive disorders, FMT research is still at an early, exploratory stage. Preclinical work in transgenic AD mouse models has shown that FMT from healthy donors can improve cognition and reduce neuropathology, while shifting the recipients’ microbiota profile toward that of healthy controls (Chen et al., 2023). Clinically, case level evidence is cautiously encouraging. In one report, an AD patient who received repeated FMT showed improvement in Mini Mental State Exam score from 20 to 26 at 2 months and to 29 at 6 months post‐treatment (Sun et al., 2019). A small pilot study in patients with mild cognitive impairment found that capsule‐based FMT was safe and well tolerated, stabilized or modestly improved cognitive scores over 6 months, and produced microbial and metabolite shifts consistent with a biological effect, particularly in those with milder baseline impairment (Hazan, 2020; Sun et al., 2019). These studies involve very small samples, yet they highlight the gut–brain axis as a candidate therapeutic target in dementia. A registered feasibility study of oral FMT in AD (Oral Fecal Microbiota Transplant Feasibility Study in Alzheimer's Disease (AMBITION), 2018) was terminated early; therefore, the clinical efficacy of FMT in larger cohorts remains to be determined. For now, FMT in Alzheimer's and related disorders should be regarded as an experimental but promising approach that requires further rigorous evaluation.
PERSONALIZED PSYCHOBIOTIC TREATMENTS: MICROBIOME‐INFORMED MENTAL HEALTH STRATEGIES
Treatment responses to psychobiotics are highly heterogeneous, influenced by host genetics, diet, medication history, environmental exposures and baseline gut microbiota (Fang et al., 2023). This variability has led to the concept of personalized microbial therapeutics tailored to an individual's clinical phenotype and microbiome profile (Rupp & Stengel, 2022). Early data suggest that baseline microbial diversity and specific taxa can predict outcomes. For example, patients with reduced Bifidobacterium abundance or low fecal SCFA levels often respond better to SCFA‐producing strains or prebiotics that enrich butyrate‐producing species, and individuals with MDD who respond to L. plantarum PS128 may show distinct increases in SCFA‐related taxa compared with non‐responders (Fang et al., 2023; Rupp & Stengel, 2022).
Near‐term clinical applications focus on simple stratification schemes. Stool metabolomic profiling could identify patients with low butyrate and high acetate to butyrate ratios who might benefit from butyrogenic formulations or synbiotics designed to restore SCFA balance (Jiang et al., 2024). Similarly, patients with elevated inflammatory markers such as C‐reactive protein, TNF‐α or IL‐6 may be better candidates for strains with documented anti‐inflammatory effects, for example B. longum 35624 or L. casei Shirota, which have reduced systemic cytokines in clinical studies (Mills et al., 2023). Such targeted strategies remain hypothesis driven and experimental, so they should be framed as exploratory, implemented with explicit consent, biomarker documentation and standardized short term follow‐up using instruments such as PHQ‐9, GAD‐7 and sleep quality indices over typical 4–12‐week intervention windows (Butler et al., 2019).
Beyond metabolomic and immune profiling, integrating host genomic information such as variants in solute carrier family 6 member 4 (SLC6A4) or IL6 may further refine strain selection and response prediction, although these approaches still lack standardized validation and regulatory endorsement (Jiang et al., 2024).
CHALLENGES AND FUTURE RESEARCH DIRECTIONS
Despite encouraging preclinical and early clinical findings, several barriers still limit the translation of psychobiotics into routine mental health care. First, weak standardization across trials, including differences in strains, doses, intervention duration, delivery formats and outcome measures, makes studies hard to compare and reduces reproducibility, particularly when strain‐level reporting is incomplete (Mosquera et al., 2024). Second, regulatory ambiguity and uneven product quality remain major obstacles because many commercial products do not reliably disclose validated strain identifiers or guarantee viable dose through shelf life, which prevents direct translation of trial evidence into real‐world recommendations (Binda et al., 2020). Third, responses appear strongly host‐dependent, shaped by factors such as baseline microbiota composition, diet, sex and concomitant medications, suggesting the presence of responder and non‐responder subgroups that current trial designs do not adequately capture (Sniffen et al., 2018). Fourth, mechanistic resolution in humans is still limited because relatively few studies directly link microbial and metabolite changes to neural or synaptic outcomes using tools such as neuroimaging, electrophysiology or cerebrospinal fluid biomarkers (Natale et al., 2021). Finally, most trials remain short, often 4–12 weeks, with limited follow‐up, so durability of benefit and longer‐term safety are not well defined (Ribeiro et al., 2025).
CLINICAL IMPLICATIONS
Current evidence supports psychobiotics as cautious, strain‐specific adjuncts in mental health care rather than stand‐alone treatments. Across trials, selected preparations may improve anxiety, depression, stress‐related outcomes and some ASD‐associated measures, particularly when GI symptoms or low‐grade inflammation are present, but effects are generally modest, heterogeneous and dependent on the exact strain, dose and intervention duration (often 4–12 weeks) (Fereshteh et al., 2020; Garzone et al., 2025; Skowron et al., 2022). In clinical practice, psychobiotics are most defensible as add‐on options for partial responders or subsyndromal presentations, and they should not replace evidence‐based treatments in severe mental illness, acute psychosis or suicidal states (Haiqa & Aslam, 2023).
A near‐term priority is improved patient and product selection. Biomarker‐ and microbiome‐informed stratification, such as targeting individuals with elevated inflammatory markers, impaired gut barrier indices or reduced SCFA‐producing taxa, remains promising but largely experimental and requires standardized monitoring and validation (Amirani et al., 2020). In parallel, the uneven quality of over‐the‐counter products, including inconsistent strain identification and uncertain viable dose across shelf life, limits direct translation from RCTs; when used, clinicians should match strain, dose and duration to published trials and preferentially select products with third‐party verification (Sharma et al., 2021).
CONCLUSION
Psychobiotics represent a promising but still emerging avenue in microbiota‐based mental health research. By acting through the MGBA, selected microbial strains and microbe derived products can modulate stress responses, immune signalling, neuroplasticity and neurotransmitter‐related pathways in ways that complement conventional psychopharmacology. Preclinical data are mechanistically robust, and early clinical trials in anxiety, depression and some neurodevelopmental conditions suggest small to moderate benefits in defined subgroups.
Translational use remains at an early stage and should be explicitly strain specific, biomarker informed where possible, and integrated with existing psychiatric treatments rather than used as a replacement. Progress will depend on rigorous strain level validation, clear reporting of dose and duration, incorporation of mechanistic biomarkers to identify likely responders, and long‐term safety data. If these conditions are met, psychobiotics may evolve from broadly marketed supplements into targeted, mechanism informed adjuncts within precision‐oriented psychiatry, particularly for patients with GI or inflammatory comorbidities.
AUTHOR CONTRIBUTIONS
Amir Arsalan Ghahari, Mehrdad Nourizadeh, Mehrdad SalekShahabi, Shaghayegh Davari, and Saeid Mohammadzadeh Mounesyar: Conceptualization, writing – original draft, writing – review & editing, and supervision. All authors have read and approved the final version of this manuscript and agree to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. All persons designated as authors qualify for authorship, and all those who qualify for authorship are listed.
CONFLICT OF INTEREST
None declared.
FUNDING INFORMATION
No funding was received for this work.
Ghahari, A. A. , Nourizadeh, M. , SalekShahabi, M. , Davari, S. , & Mohammadzadeh Mounesyar, S. (2026). Psychobiotics and the microbiota–gut–brain axis: Emerging paradigms in mental health modulation. Experimental Physiology, 111, 2449–2463. 10.1113/EP093301
Contributor Information
Amir Arsalan Ghahari, Email: amirarsalan.ghahari@iau.ir.
Mehrdad Nourizadeh, Email: mehrdad.nourizadeh@iau.ir, Email: Mehrdadnourizadeh1377@gmail.com.