How Diet and Gut Bacteria Are Linked to Mental Health

This review was conducted and reported in accordance with the PRISMA 2020 guidelines for systematic reviews [21]. The study aimed to systematically synthesize clinical evidence published between January 2020 and October 2025 investigating the relationship between dietary patterns, microbiota-targeted interventions, and mental health outcomes-specifically depression and anxiety-in adult populations. Given the expected heterogeneity of interventions, populations, and outcome measures, no quantitative meta-analysis was planned. Instead, a structured qualitative synthesis was conducted, emphasizing consistency, methodological quality, and clinical applicability of the available evidence.

Studies were selected according to predefined inclusion and exclusion criteria formulated using the PICOS framework (Population, Intervention/Exposure, Comparator, Outcomes, Study Design). Population: Adults aged 18 years or older, from general or clinical populations, assessed for depressive or anxiety outcomes using validated diagnostic instruments or standardized rating scales. Interventions/Exposures: MD interventions or adherence-based analyses; UPF exposure or consumption; and psychobiotic strategies including probiotic or prebiotic supplementation. Comparators: usual diet, minimal intervention, lower exposure, or placebo. Outcomes: primary outcomes included change in depressive symptoms, incidence of depression, or remission rates; and secondary outcomes encompassed anxiety symptom changes or incident anxiety disorders when co-reported. Study design: RCTs, quasi-experimental studies, prospective cohort studies, and systematic reviews/meta-analyses (SR/MAs); studies focusing exclusively on pediatric samples, psychosis, cognitive decline, or without validated psychiatric measures were excluded.

Time frame: publications between 1 January 2020 and 22 October 2025. Language: English only.

A comprehensive literature search was conducted in PubMed/MEDLINE, Scopus, and Web of Science (Core Collection). The search strategy combined MeSH and free-text terms relating to diet and microbiota ("Mediterranean diet," "dietary pattern," "ultra-processed food," "probiotic," "prebiotic," "psychobiotic," "gut microbiota," "gut–brain axis") with psychiatric terms ("depression," "depressive disorder," "anxiety," "anxiety disorder"). The final PubMed query was: ("Mediterranean diet"[Mesh] OR "Mediterranean diet"[tiab] OR "dietary pattern"[tiab] OR "ultra-processed food"[tiab] OR "probiotic"[Mesh] OR "prebiotic"[Mesh] OR "psychobiotic"[tiab] OR "gut microbiota"[Mesh] OR "gut-brain axis"[tiab]) AND ("depression"[Mesh] OR "depressive disorder"[tiab] OR "anxiety"[Mesh] OR "anxiety disorder"[tiab]).

Equivalent search strings were adapted for Scopus and Web of Science. Filters were applied for human studies, English language, and publication years 2020–2025. In addition, reference lists of included systematic reviews and key mechanistic papers were manually screened to identify relevant studies not captured by electronic searches.

The screening and selection process was performed in two independent stages by two reviewers (G.M. and M.M.). In the first stage, titles and abstracts were screened for relevance. In the second stage, full texts of potentially eligible articles were retrieved and assessed against inclusion criteria. Discrepancies were resolved by consensus or consultation with a third reviewer (G.T.). Risk of bias for included randomized trials was assessed using the Cochrane risk-of-bias framework, evaluating random sequence generation, allocation concealment, blinding, incomplete outcome data, selective reporting, and funding sources. For observational studies, risk of bias was assessed based on confounding control, exposure/outcome assessment, attrition, and statistical adjustment. Discrepancies between reviewers were resolved through consensus.

The process is illustrated in Figure 1 (PRISMA 2020 Flow Diagram). A total of 2357 records were identified (2315 from databases and 42 from reference lists). After removal of duplicates, 1876 records were screened by title and abstract. Of these, 1642 were excluded for irrelevance or lack of psychiatric outcomes. 234 full-text articles were evaluated for eligibility, and 198 were excluded because they did not meet inclusion criteria (wrong population, design, or outcome). Finally, 36 studies were included in the qualitative synthesis, comprising 14 RCTs, 9 cohort studies, and 13 systematic reviews/meta-analyses. No quantitative synthesis was performed due to heterogeneity of interventions, study designs, and outcome measures.

Data were extracted into a predesigned matrix covering study design, population characteristics, intervention or exposure details, duration, psychiatric outcomes, principal findings, and quality assessment. Extraction was performed independently by two reviewers. Because of substantial methodological and conceptual heterogeneity, particularly regarding dietary definitions, psychobiotic strain variability, and outcome reporting, a narrative synthesis was undertaken. Findings were grouped by thematic domain (MD interventions, UPF exposure, psychobiotic/prebiotic strategies, and mechanistic insights), and consistency and clinical relevance were graded across bodies of evidence.

Over the last five years, multiple RCTs and MAs have consolidated the role of the MD as a promising adjunctive intervention for depressive symptoms. Recent studies demonstrate significant reductions in depressive symptom severity compared with control diets or usual care conditions, underscoring the feasibility and clinical potential of MD-based interventions in both clinical and subclinical depression [10,22].

In a comprehensive meta-analysis including over 2000 participants across 14 RCTs, Bizzozero-Peroni et al. reported that adherence to a Mediterranean dietary pattern significantly improved depressive symptoms compared with control diets, with standardized mean differences ranging from −0.22 to −0.40 depending on intervention duration and adherence level [10]. Importantly, these benefits were observed across diverse populations, including individuals with major depressive disorder (MDD), those with subthreshold symptoms, and adults at metabolic risk. The overall certainty of evidence was rated as moderate, suggesting a robust, clinically meaningful association between MD adherence and mood improvement.

These findings were corroborated by Lassale et al. [23], who identified consistent antidepressant effects of MD interventions across both primary prevention and clinical treatment studies, emphasizing that high-fidelity implementation and sustained adherence are critical determinants of efficacy. Interventions incorporating dietitian-led counseling, structured meal planning, or behavioral adherence tools, such as digital prompts, cooking workshops, or group support, tended to achieve greater reductions in depressive scores than those relying solely on written dietary guidance [24].

From a mechanistic standpoint, the MD exerts its antidepressant effects through several complementary pathways. The diet's high content of omega-3 fatty acids, polyphenols, and dietary fiber reduces systemic inflammation and oxidative stress, thereby modulating neuroinflammatory cascades implicated in depression [25]. Improved gut microbiota diversity associated with MD adherence enhances SCFA production, which strengthens intestinal barrier integrity and influences neurotransmitter metabolism [9,26]. Additionally, MD adherence has been linked to reduced activation of the tryptophan–kynurenine pathway, leading to improved serotonin availability and reduced neurotoxic metabolite production [27].

Despite these encouraging results, several limitations temper the current evidence base. Most available RCTs enrolled small to medium-sized samples (n < 250) and were conducted over relatively short durations (typically 8–12 weeks). Moreover, heterogeneity in dietary adherence monitoring and control conditions, ranging from minimal advice to fully structured interventions, complicates direct comparisons. Few studies employed blinded assessors or standardized psychiatric diagnostic tools, which may introduce bias. Nevertheless, the direction and consistency of effects across diverse settings reinforce the clinical relevance and safety of the MD as an adjunctive strategy for depression management [28].

Implementation science insights suggest that adherence strategies play a decisive role in sustaining benefits over time. Interventions supported by personalized meal plans, cooking skill development, and digital adherence tracking yield larger and more durable improvements in depressive symptoms. Conversely, programs with low engagement or poor adherence monitoring often demonstrate attenuated or non-significant effects.

Overall, current research supports the Mediterranean-style diet as a feasible, low-risk, and biologically plausible adjunct in psychiatric care pathways. It is supported by preliminary but promising evidence as a feasible, low-risk adjunct preventive strategy for mental health, particularly in preventive and therapeutic frameworks for depression. Large-scale, long-term RCTs with rigorous adherence assessment and mechanistic endpoints are warranted to confirm durability and generalizability of these effects.

Growing epidemiological evidence indicates that high consumption of UPFs, industrial formulations characterized by refined carbohydrates, added sugars, saturated fats, and numerous cosmetic additives, is associated with adverse mental health outcomes, including higher prevalence and incidence of depression and anxiety disorders [12,29]. Over the past five years, several large-scale cohort studies and umbrella reviews have strengthened the link between UPF intake and affective disorders, revealing a robust and biologically plausible dose–response relationship [30].

In a landmark umbrella review encompassing 45 meta-analyses and over 10 million participants, Lane et al. confirmed that high UPF consumption is consistently associated with an increased risk of depression and anxiety symptoms across diverse populations [29]. The association remained significant after adjustment for BMI, socioeconomic factors, and overall diet quality, suggesting an effect independent of total caloric intake.

Mechanistically, the deleterious impact of UPFs on mental health appears multifactorial. Diets high in UPFs promote chronic low-grade inflammation, oxidative stress, and gut microbiota dysbiosis, all of which contribute to neuroimmune dysregulation and affective vulnerability [31]. Excessive intake of emulsifiers, artificial sweeteners, and preservatives alters microbial composition, leading to decreased production of beneficial SCFAs and increased intestinal permeability, which in turn may activate systemic inflammatory pathways and cytokine cascades implicated in depression [26,32]. In parallel, high-glycemic-load foods provoke postprandial glucose spikes and insulin resistance, disrupting energy metabolism and neuroplastic signaling within limbic structures involved in mood regulation.

Recent neuroimaging data suggest that chronic exposure to UPFs may affect reward-related brain circuits, potentially through dopaminergic sensitization and impaired satiety signaling. This neurobehavioral dysregulation may partially explain the bidirectional relationship between emotional eating and depressive symptoms, particularly among individuals with stress-related vulnerabilities or obesity [33].

The magnitude of association between UPF intake and depression risk is comparable to that observed for other modifiable risk factors such as physical inactivity or heavy alcohol use, underscoring the potential public-health relevance of dietary composition beyond caloric balance. Nevertheless, residual confounding remains a concern: individuals with higher UPF consumption often exhibit lower socioeconomic status, higher rates of smoking, and reduced access to health-promoting resources [34]. Despite these limitations, the consistency of findings across diverse cohorts, analytic methods, and geographical regions strengthens the causal inference that UPF exposure negatively influences mental health.

Furthermore, emerging data indicate that reductions in UPF consumption may lead to short-term improvements in psychological well-being and perceived stress, independent of weight change or macronutrient intake. Interventional trials, though still scarce, suggest that replacing UPFs with minimally processed alternatives rich in fiber, polyphenols, and unsaturated fats can improve mood and fatigue within 2–4 weeks, possibly via microbiota restoration and decreased systemic inflammation [35].

These findings suggest that UPF consumption is a modifiable, population-level risk factor for depression and anxiety. Addressing UPF exposure through dietary counseling, food-labeling policies, and public-health campaigns could represent a scalable strategy for primary and secondary prevention of common mental disorders. Future research should integrate dietary assessment with biological endpoints, such as inflammatory cytokines, SCFAs, and gut microbiota composition, to further elucidate the causal pathways linking UPFs and affective outcomes.

Over the past decade, and particularly since 2020, the concept of psychobiotics has gained substantial traction in nutritional psychiatry. The term refers to live microorganisms (probiotics) or non-digestible substrates (prebiotics) that confer mental health benefits via modulation of the gut microbiota–brain axis [13,14]. Growing evidence from RCTs and meta-analyses suggests that psychobiotic and prebiotic interventions can attenuate depressive and anxiety symptoms, albeit with modest and variable effect sizes [36,37].

A comprehensive meta-analysis by including 36 RCTs reported that probiotic supplementation led to a statistically significant reduction in depressive symptoms compared with placebo [36]. The greatest improvements were observed in participants with subclinical depression or elevated depressive scores rather than those with established MDD.

Across these studies, probiotic strains belonging to the genera Lactobacillus and Bifidobacterium, particularly L. helveticus R0052, L. rhamnosus GG, and B. longum NCC3001, have demonstrated reproducible antidepressant and anxiolytic effects in both clinical and non-clinical populations. These effects are thought to be mediated through several convergent biological mechanisms. Psychobiotics modulate systemic and neuroinflammatory signaling, reducing levels of pro-inflammatory cytokines such as interleukin-6 (IL-6), tumor necrosis factor-α (TNF-α), and C-reactive protein (CRP), which are consistently elevated in depressive states [38]. They enhance intestinal epithelial integrity and increase the production of SCFAs, which exert anti-inflammatory, epigenetic, and neuroprotective effects. These SCFAs influence blood–brain barrier permeability and modulate the activity of microglial cells, thereby affecting neuroinflammatory tone and synaptic plasticity [39].

No consensus exists regarding optimal dosing, duration, or responder characteristics, and strain heterogeneity limits the generalizability of findings. Standardization of strain selection and reporting remains a major unmet need.

Psychobiotic supplementation has been shown to normalize hypothalamic–pituitary–adrenal (HPA) axis hyperactivity, leading to lower cortisol levels and improved stress reactivity. In both human and animal studies, probiotics attenuate stress-induced increases in adrenocorticotropic hormone (ACTH) and corticosterone, suggesting a regulatory influence on central stress pathways [40,41]. Finally, modulation of tryptophan metabolism via the kynurenine pathway may underpin psychobiotics' antidepressant potential. By dampening inflammation-driven indoleamine-2,3-dioxygenase (IDO) activity, these interventions may preserve tryptophan availability for serotonin synthesis, reducing neurotoxic kynurenine metabolites that have been implicated in depression and fatigue [42].

Despite encouraging results, several limitations persist. Considerable heterogeneity exists across studies in probiotic species, dosages, intervention duration, and outcome measures, complicating cross-trial comparison. Many studies enrolled small samples and used self-reported questionnaires rather than clinician-administered scales, potentially inflating effect sizes. Additionally, few trials have performed strain-specific analyses or mechanistic biomarker assessments, such as microbiota sequencing, SCFA quantification, or cytokine profiling, which are essential to delineate causal pathways. A major limitation across dietary and psychobiotic research is the lack of replication, compounded by substantial heterogeneity in strains, dosages, dietary protocols, and outcome measures.

Emerging evidence also supports prebiotic interventions, particularly fructooligosaccharides (FOS) and galactooligosaccharides (GOS), which selectively stimulate the growth of beneficial Bifidobacteria and promote SCFA production [43]. Early trials suggest that prebiotic supplementation can improve emotional regulation and reduce cortisol awakening response in healthy adults, though data in clinical populations remain limited.

Compared with depressive symptoms, findings for anxiety outcomes are more heterogeneous. Some RCTs report modest reductions in state or trait anxiety, whereas others show no significant effect, particularly in populations with established anxiety disorders. Variability in assessment tools and sample characteristics contributes to inconsistent results [15]. Overall, psychobiotic effects on anxiety remain less robust and require targeted trials.

Research indicates that psychobiotic and prebiotic interventions exert small-to-moderate but clinically relevant effects on depressive and anxiety symptoms. They appear to complement traditional psychiatric treatments through anti-inflammatory, neuroendocrine, and microbiota-mediated mechanisms. While not a substitute for established pharmacotherapy or psychotherapy, psychobiotics represent a safe, accessible, and biologically grounded adjunct in the management of mood disorders.

Across the reviewed studies, converging evidence highlights a shared set of biological pathways linking diet quality, gut microbiota composition, and affective regulation. Mediterranean-style dietary patterns, UPF exposure, and psychobiotic or prebiotic interventions appear to act on common molecular and neuroimmune mechanisms, albeit in opposite directions. Collectively, these findings reinforce the concept of an immune–metabolic continuum underlying mood disorders, wherein nutritional patterns shape inflammatory tone, microbial metabolites, and neurotransmitter homeostasis.

Dietary modulation of systemic inflammation emerges as a central node connecting all three domains. The MD exerts robust anti-inflammatory and antioxidant effects through high intake of omega-3 polyunsaturated fatty acids, polyphenols, and vitamins C and E, which suppress pro-inflammatory cytokines and reduce oxidative stress. Conversely, UPF-rich diets are associated with increased circulating concentrations of IL-6, tumor necrosis factor-alpha (TNF-α), and CRP, as well as higher oxidative stress biomarkers. Chronic low-grade inflammation can cross the blood–brain barrier, activating microglia and perturbing monoaminergic signaling, ultimately leading to anhedonia, fatigue, and depressive mood. Psychobiotic supplementation, on the other hand, appears to attenuate inflammation by restoring intestinal barrier integrity and inhibiting pro-inflammatory cytokine cascades. Certain Lactobacillus and Bifidobacterium strains downregulate NF-κB activation and decrease plasma IL-6 and TNF-α concentrations. The resulting reduction in peripheral inflammation may restore serotonergic transmission and improve mood [44].

The gut microbiota represents a critical interface through which diet influences brain function. Adherence to a Mediterranean-style diet increases microbial richness and promotes the growth of Bifidobacterium and Faecalibacterium species, both major producers of SCFAs such as acetate, propionate, and butyrate. These metabolites enhance epithelial tight-junction integrity, reduce intestinal permeability, and modulate immune responses through inhibition of histone deacetylases (HDACs). SCFAs can cross the blood–brain barrier, where they influence neuroinflammation, synaptic plasticity, and neurogenesis. By contrast, UPFs are poor in fermentable fiber and rich in emulsifiers and artificial sweeteners that disrupt microbial balance and decrease SCFA production. This dysbiotic shift may trigger metabolic endotoxemia and systemic immune activation, further reinforcing depressive pathophysiology. Probiotics directly increase SCFA-producing taxa, while prebiotics (e.g., galacto- and fructooligosaccharides) provide the substrates required for their proliferation. Psychobiotics have been proposed as potential modulators of microbial function, although human microbiome responses to probiotic supplementation are highly variable, and some authors report minimal or no significant compositional changes [45].

Another key mechanism linking diet and mood involves the tryptophan–kynurenine pathway, a metabolic route whereby inflammatory activation diverts tryptophan from serotonin synthesis toward neuroactive kynurenine metabolites. Diets rich in anti-inflammatory nutrients (e.g., omega-3 fatty acids, polyphenols) and gut-derived SCFAs may influence IDO activation, preserving tryptophan availability for serotonin production; however, most supporting evidence derives from preclinical studies, and human data remain preliminary. Conversely, UPF-driven inflammation enhances kynurenine formation, resulting in neurotoxic intermediates such as quinolinic acid that promote excitotoxicity and neurodegeneration [44].

Emerging evidence indicates that probiotic supplementation can modulate tryptophan metabolism by reducing IDO activity and increasing circulating serotonin precursors, thereby counteracting inflammatory diversion within this pathway. Such findings underscore the biochemical convergence between anti-inflammatory dietary patterns and microbiota-targeted therapies in normalizing neurotransmitter homeostasis.

Diet also influences mood through regulation of the HPA axis, the primary neuroendocrine stress pathway. Chronic UPF consumption and metabolic inflammation are associated with hypercortisolemia and impaired negative feedback of cortisol signaling, whereas MD adherence and psychobiotic interventions reduce stress reactivity and normalize cortisol awakening response [46]. These effects are hypothesized to be partially mediated by vagal pathways linking gut microbial metabolites to central autonomic circuits, based primarily on preclinical evidence; direct confirmation in humans remains limited.

Cumulative evidence suggests an emerging immune–metabolic model in which diet may act as a potential trigger and modulator of affective vulnerability, although causal pathways remain insufficiently established. Protective dietary patterns such as the MD promote microbiota diversity, SCFA production, and anti-inflammatory tone, stabilizing neurochemical and neuroendocrine function. Conversely, deleterious dietary exposures, notably UPFs, disrupt these regulatory networks, promoting inflammation, dysbiosis, and HPA-axis overactivation. Psychobiotics and prebiotics, positioned mechanistically between these two poles, provide a biologically plausible but still investigational approach of restoring microbial–immune homeostasis and complementing nutritional and pharmacological interventions in depression and anxiety.

The interplay among dietary quality, gut microbiota, and immune regulation provides a unifying framework for understanding how nutrition influences the "immune mind." These mechanistic insights seem to indicate the clinical potential of nutritional psychiatry and justify integrative approaches that combine dietary modification, microbiota-targeted therapy, and lifestyle optimization to contribute to prevent and treat mood disorders. These relationships are summarized in Figure 2, which depicts the bidirectional interactions between diet quality, gut microbiota, immune signaling, and mood regulation, the conceptual core of the "immune mind" model.

Newer prospective syntheses assessing overall diet quality report inverse associations with subsequent depression risk, complementing the interventional MD literature. While residual confounding cannot be excluded, alignment between cohort and RCT/SR evidence strengthens the case for diet-first adjunctive strategies in mood disorders [47].

A concise overview of the principal clinical findings across MD, UPF exposure, and psychobiotic/prebiotic interventions is presented in Table 1.