Psychoneuroimmunology (PNI) has matured into a mechanistically specified neuroimmune science in which psychological stressors, peripheral immunity, and neural circuits are studied as bidirectionally coupled systems rather than as correlational “mind–body” associations[1]. Contemporary work converges on the principle that psychosocial stress can durably remodel inflammatory set-points, with chronic stress producing maladaptive, long-lasting consequences across nervous, immune, endocrine, and metabolic systems[2] and with inflammation feeding back to increase reactivity of stress- and reward-related brain structures[2]. In parallel, immune-to-brain information transfer has been decomposed into discrete access routes—neural, humoral, blood–brain barrier (BBB) transport, and cellular trafficking—with defined molecular hubs such as NF-κB at the blood–brain interface[3, 4]. Within psychiatry, the inflammatory hypothesis of depression is increasingly operationalized as biomarker-defined endotypes in which cytokine signaling perturbs HPA-axis feedback and glucocorticoid sensitivity[5], modulates monoamine handling via the serotonin transporter[5], and engages tryptophan–kynurenine pathways through IDO induction in inflammatory states[5]. Causal neuroimmune circuit work has progressed from “existence proofs” to cell-type-resolved control logic: pro- versus anti-inflammatory cytokines activate non-overlapping vagal sensory neuron populations, which engage brainstem cNST nodes that act as a rheostat for peripheral inflammatory balance[6, 7], and silencing this circuit converts otherwise regulated inflammation into runaway responses[6]. Across neurodegeneration and autoimmunity, microglia are no longer treated as homogeneous responders but as programmable states that can drive neurotoxic glial cascades (microglial IL-1α/TNF/C1q induction of A1 astrocytes) and disease-specific phenotypes such as “microglia inflamed in MS” (MIMS) in which C1q is a validated therapeutic node[8, 9]. Finally, immune surveillance at CNS borders has been reframed via meningeal conduits, lymphatic drainage, and leptomeningeal “licensing” checkpoints for effector T cells, enabling new therapeutic concepts based on spatial control of antigen presentation and trafficking[10–12]. Together, these advances reposition PNI as a precision discipline in which measurement, circuit specification, and stratified translation are inseparable from theory building[6, 13, 14].
Introduction
Over the last three decades, PNI research has accumulated evidence for “extensive bidirectional communication between the brain and the immune system,” reframing psychological and behavioral states as biologically embedded drivers of immune function and disease risk[1]. This bidirectionality is supported by anatomical and physiological substrates, including “sympathetic and parasympathetic innervation of organs and tissues associated with the immune system” such as lymph nodes, thymus, spleen, and bone marrow[1]. It is also supported by a shared signaling lexicon in which the nervous and immune systems communicate through “a common biochemical language” that includes hormones, neurotransmitters, cytokines, and receptors shared across systems[1]. Modern PNI is increasingly organized around (i) route-resolved immune-to-brain communication (neural, humoral, BBB transport, cellular) and (ii) circuit-resolved brain-to-immune control, both of which enable mechanistic intervention design rather than generic anti-inflammatory or stress-reduction approaches[3, 6].
A defining paradigm shift is that psychological exposures are being quantified with reliability comparable to other biomedical risk factors, enabling cumulative stress phenotyping and mechanistic linking to downstream biology[13]. In parallel, neuroimmune circuit mapping has progressed to single-cell and functional imaging specification of defined nodes whose perturbation causally transforms inflammatory trajectories, implying that “the brain tightly modulates the course of the peripheral immune response” through identifiable circuit elements rather than diffuse “stress effects”[6].
The Stress–Immune Axis
Acute and chronic stressors produce systematically different immune consequences, with “acute stressors generally…associated with enhanced immunity” whereas “long-term or chronic stressors” are associated with “suppressed immune function” in the available evidence base[1]. Chronic stress is specifically emphasized as producing “long-lasting maladaptive effects, with pathologic consequences on nervous, immune, endocrine, and metabolic systems,” motivating models in which stress is a disease-modifying exposure rather than merely a symptom amplifier[2]. Mechanistically, stress engages neuroendocrine and autonomic channels, and stress responses include a shift in autonomic balance in which “the SNS is up-regulated” while “the PSNS is downregulated,” implying loss of parasympathetic anti-inflammatory constraints during stress physiology[15].
At the level of immune transcriptional control, experimental stressors in humans can drive inflammatory signaling pathways, exemplified by findings that peripheral blood mononuclear cells after an acute psychosocial stressor show “significant increases in NF-κB DNA binding”[4]. The stress–inflammation relationship is also framed as a feedback system in which “psychosocial stress is a powerful regulator of central and peripheral inflammation” and systemic inflammatory factors can “retroact on the CNS and increase the reactivity” of stress- and reward-related brain structures, embedding stress biology in recurrent loops rather than one-way causation[2].
Innovations in stress measurement further strengthen the field’s inferential power, as cumulative lifetime stress can be quantified with high test–retest reliability (e.g., STRAIN lifetime stressor count ) and linked to health outcomes with interpretable effect sizes[13]. In the same framework, greater lifetime stressor count is associated with more doctor-diagnosed autoimmune disorders (IRR reported as 1.028 with confidence intervals), supporting an epidemiologically grounded connection between psychological exposures and immune-mediated disease burden[13].
Neuroinflammation and Depression
A core translational claim of PNI is that major depression is often accompanied by inflammatory activation, as depressed patients have been reported to exhibit “all of the cardinal features of inflammation” compared with nondepressed individuals across medically ill and medically healthy contexts[4]. In this framing, depression can be conceptualized as “mediated by inflammatory responses and cytokines,” explicitly linking affective pathology to immune signaling and stress adaptation failure[16]. Mechanistically, cytokines can disrupt HPA-axis negative feedback by stimulating CRH release and “facilitating glucocorticoid resistance,” aligning immune signaling with enduring neuroendocrine dysregulation and treatment response dynamics[5].
Several molecular pathways have become particularly influential because they provide specific, testable “immune-to-synapse” links rather than general inflammation correlates. Pro-inflammatory cytokines can influence the active serotonin fraction by “upregulating the activity and expression” of serotonin transporters (SERT), thereby coupling immune activation to monoamine handling in a pathway with obvious pharmacologic relevance[5]. Inflammatory challenges can also drive the kynurenine pathway, as LPS-induced inflammation can increase IDO and produce depressive-like behavior, while “blockade of IDO-activation prevents” those behavioral outcomes in the cited model systems[5].
Meta-analytic evidence suggests that targeting cytokine pathways can improve depressive symptoms on average in randomized trials, with a meta-analysis reporting significant improvement in depressive symptoms with anti-cytokine drugs compared to placebo and a standardized mean difference of 0.40 with confidence intervals[17]. However, the same translational literature emphasizes subgroup structure, as treatment effects can interact with baseline inflammation; for example, changes in depressive symptom ratings favored infliximab only at baseline hs-CRP concentrations greater than 5 mg/L and favored placebo at lower baseline hs-CRP values in the cited analysis[18]. This aligns with the broader position that “only subgroups of depressed patients have elevated levels of cytokines,” and that cytokine elevations are not specific to depression, motivating endotype-based rather than diagnosis-based immunopsychiatry[19].
Microglia and Behavior
Microglia occupy a privileged mechanistic position in PNI because they are described as “the primary cellular recipients” of inflammatory signals reaching the CNS, enabling immune-state information to be translated into changes in neural excitability, plasticity, and behavior[3]. When microglia are “activated,” their morphology becomes amoeboid and they release pro-inflammatory cytokines such as IL-6, TNF-α, and IL-1β within the CNS, providing a cellular substrate for sustained neuroinflammatory signaling[3]. Stress biology connects to this microglial axis, as “chronic stress activates brain microglia,” which secrete cytokines and can affect neurogenesis, linking psychosocial exposures to structural and functional plasticity changes[16].
The field has also advanced from generic activation models to defined glial state transitions and inter-glial causality. Activated microglia can induce A1 astrocytes via secretion of IL-1α, TNF, and C1q, and these cytokines are described as “necessary and sufficient” to induce A1 astrocytes, positioning microglia→astrocyte signaling as a concrete mechanism for neurotoxic remodeling[8]. A1 astrocytes are functionally consequential because they “lose the ability to promote neuronal survival, outgrowth, synaptogenesis and phagocytosis” and can “induce the death of neurons and oligodendrocytes,” linking glial immune programs to neurodegenerative outcomes[8].
In demyelinating neuroinflammation, microglial state definitions have become disease-programmatic: “microglia inflamed in MS” (MIMS) are defined as neurodegenerative programming states, and complement component C1q is identified as a “critical mediator” of MIMS activation with genetic and therapeutic validation, including microglia-specific ablation and blockade in chronic EAE paradigms[9]. This is important for PNI because it supports the idea that immune-related brain changes are not only reactive but can be sustained, targetable state programs with measurable biomarkers and intervention points (e.g., C1q inhibition as a therapeutic avenue, with longitudinal monitoring via paramagnetic rim lesions using advanced MRI methods)[9].
Microbiome–Gut–Brain–Immune Axis
The microbiome–gut–brain axis is increasingly treated as a multi-route communication system in which gut microbes influence brain function through at least three pathways that jointly produce bidirectional information flow[20]. Established communication routes include the autonomic nervous system, enteric nervous system, neuroendocrine system, and immune system, positioning gut ecology as a systems-level modulator of both immune tone and neural states relevant to affective symptoms[21]. Within this framework, an immunoregulatory pathway is explicitly described in which microbiota interact with immune cells to affect cytokine levels and related mediators (including prostaglandin E2), while a vagus nerve pathway is also described as a route by which enteric neural activity contributes to gut-to-brain signaling[20].
Causal evidence for microbiome-driven behavioral phenotypes spans developmental programming, infection challenges, and transplantation paradigms. Germ-free mice show altered anxiety-like behavior (reduced anxiety-like behavior in the elevated plus maze relative to specific pathogen free mice), supporting the claim that microbial exposure shapes baseline behavioral phenotypes[22]. These effects can reflect early-life programming, as the low anxiety-like phenotype can “persist” after colonization with intestinal microbiota, suggesting that gut–brain interactions influence CNS wiring early in life and may not be trivially reversible by later recolonization[22].
The axis is also route-specific in ways relevant to PNI circuit logic, because vagal sensory neurons can show activation markers in response to subpathogenic infection “in the absence of a systemic immune response,” and the same experimental context can increase anxiety-like behavior despite no peripheral immune response, providing a mechanistic wedge separating neural sensory pathways from classical systemic inflammation[21]. Transplantation studies further strengthen causality: fecal microbiota transplantation from depressed donors is described as “definitive evidence” that depression-associated microbiome alterations are sufficient to disrupt behavioral and physiological homeostasis, and recipient rats show anhedonia-like behavior in sucrose preference testing after receiving microbiota from depressed pools[23]. These behavioral effects are coupled to biochemical signatures consistent with immune–metabolic links (increased plasma kynurenine and kynurenine/tryptophan ratio) and ecological disruptions (decreased richness and diversity in the depressed group), reinforcing a mechanistic bridge between microbial ecology and host neuroimmune metabolism[23].
A major 2020s reframing is the “gateway” model in which dysbiosis and barrier disruption enable systemic immune activity to influence the brain, as changes in gut microbiota and BBB disruptions are described as gateways through which systemic immune activity can affect brain states[24]. This model is mechanistically anchored by observations that loss of butyrate can weaken intestinal barrier integrity while microbial products such as LPS, peptidoglycan, flagellin, and TMAO can translocate into circulation, providing concrete molecules linking gut ecology to systemic inflammatory tone[24]. Consistent with this, systemic inflammation and gut-derived metabolites are described as promoting BBB dysfunction and microglial activation that “ultimately” drives neuroinflammation and contributes to depressive symptomatology[24].
Inflammatory Reflex
The inflammatory reflex frames neuroimmune coupling as a bidirectional neural circuit rather than an endocrine-only stress response, and it is explicitly articulated that a vagus-related nucleus can inhibit immune functions and cytokine production via acetylcholine release from the vagus nerve[16]. Peripheral cytokine signals can also be conveyed to the brain via cytokine receptors on afferent nerve fibers such as the vagus, delivering signals to brain regions including the nucleus of the solitary tract and hypothalamus, supporting a defined sensory arm of immune-state detection[25]. The functional importance of neural routes for immune-to-brain effects is underscored by evidence that vagotomy can inhibit multiple aspects of responses to peripheral inflammatory stimuli, including HPA-axis activation, catecholamine and serotonin metabolism changes, and depressive-like behavior[26].
A recent circuit-resolution advance demonstrates that pro- and anti-inflammatory cytokines communicate with “distinct populations of vagal neurons” to inform the brain of emerging inflammatory responses, implying cytokine-class-specific sensory channels rather than a single generalized inflammatory afferent line[6]. In the same mechanistic framework, cNST neurons are proposed to function as a “biological rheostat” controlling the extent of peripheral inflammatory responses through positive- and negative-feedback modulation on immune cells, articulating a control-theoretic model for neuroimmune homeostasis[6]. Circuit necessity is supported by perturbation experiments in which chemogenetic inhibition of cNST neurons produces a “dramatic increase” in pro-inflammatory response with concomitant decreases in anti-inflammatory response, described as runaway inflammation, and by the claim that removing this body–brain circuit abolishes essential immune regulation and makes an otherwise normal inflammatory response unregulated[6].
Vagus Nerve Stimulation
The translational premise of vagus nerve stimulation (VNS) and related neuromodulation is that selective manipulation of defined neuroimmune circuit components can suppress pro-inflammatory responses while enhancing anti-inflammatory states, as shown in work combining single-cell RNA sequencing and functional imaging to identify circuit components and demonstrate that selective manipulation can “effectively suppress” pro-inflammatory responses while “enhancing an anti-inflammatory state”[6]. Mechanistic specificity is further sharpened by findings that anti-inflammatory and pro-inflammatory cytokines activate “two discrete non-overlapping populations of vagal sensory neurons,” implying that stimulation paradigms could, in principle, target immune-state codes rather than merely increasing vagal tone globally[7]. Within this mapping, IL-10 but not pro-inflammatory cytokines activates TRPA1-expressing vagal neurons, providing a molecularly identified route for anti-inflammatory sensory encoding that can be used to design pathway-specific neuromodulation strategies[7].
Clinical translation discussions increasingly emphasize mechanism-bridging designs that combine immunomodulation (e.g., anti-IL-6, COX-2 inhibitors) with neuromodulation (e.g., taVNS), aiming to assess additive neuro–immuno–cognitive effects rather than treating neuromodulation and immunotherapy as separate domains[27]. At the same time, PNI contains cautionary demonstrations that physiological adaptation can be temporary and may not change underlying inflammatory disease, emphasizing the need for durable endpoints and mechanistic markers rather than transient physiological shifts as primary outcomes[28].
Meningeal Lymphatics
CNS border immunity has been reconceptualized from “immune privilege” toward structured immune access and surveillance, including identification of CNS-derived regulatory self-peptides presented on MHC-II molecules in the CNS and at its borders as molecular cues for direct CNS–immune communication[11]. During homeostasis, these regulatory self-peptides are found bound to MHC-II “throughout the path of lymphatic drainage from the brain to its surrounding meninges and its draining cervical lymph nodes,” providing a spatially organized antigen presentation landscape that links drainage to adaptive immune regulation[11]. This supports a model in which border antigen presentation can be used to “dampen autoreactive T cell responses” to secure CNS immunosurveillance, explicitly connecting spatial antigen sampling to tolerance maintenance[11].
Anatomically, the meningeal interface is not merely a diffusion boundary, as direct connections between the dura and brain have been described as “bona fide conduits” that convey peripheral molecules to the subarachnoid space adjoining the brain, expanding the potential for rapid peripheral-to-CNS signaling beyond classic BBB-centric models[10]. In autoimmune neuroinflammation models, effector T cells are shown to enter the CSF from the leptomeninges during EAE, and the leptomeninges are described as a checkpoint where activated T cells are “licensed” to enter CNS parenchyma while non-activated T cells are preferentially released into the CSF, reframing trafficking as an active decision point rather than passive leakage[12]. Adhesion and detachment at this checkpoint are mechanistically specified, as T-cell detachment is counteracted by integrins VLA-4 and LFA-1 binding to ligands produced by resident macrophages, creating tractable molecular targets for modulating neuroinflammatory seeding[12].
T Cells and Cognition
Adaptive immune mechanisms increasingly appear as context-dependent causal factors rather than bystanders, as T cell numbers can correlate with neuronal loss and T cells can dynamically transform from activated to exhausted states with unique TCR clonal expansion in tauopathy models, linking immune state trajectories to neurodegenerative injury over time[29]. Disease modification via immune pathway perturbation is supported by findings that tauopathy induces a unique innate and adaptive immune response and that depletion of microglia or T cells blocks tau-mediated neurodegeneration, positioning microglia–T cell hubs as causal levers rather than correlates[29]. Therapeutically, inhibition of interferon-γ and PDCD1 signaling can significantly ameliorate brain atrophy in this context, implicating checkpoint and cytokine axes as neurodegeneration-modifying pathways with relevance for neuroimmune precision targeting[29].
In neuroinflammatory autoimmunity, innate–adaptive gating mechanisms have been clarified by discovery of inflammatory ILC3s derived from circulation that localize near infiltrating T cells in the CNS, function as antigen-presenting cells that restimulate myelin-specific T cells, and are increased in individuals with multiple sclerosis, giving a concrete cellular mechanism for sustained adaptive immune activation in the CNS niche[30]. Notably, antigen presentation by inflammatory ILC3s is “required” to promote T cell responses in the CNS and development of MS-like disease in mouse models, strengthening the causal inference from cell presence to disease-driving function[30]. Conversely, tissue-resident and peripheral ILC3s can retain tolerogenic potential and when targeted to present myelin antigen can eliminate self-specific T cells and prevent demyelinating disease, suggesting that antigen-presentation engineering can be used for immune tolerance without broad immunosuppression[30].
Finally, immunometabolic links connect adaptive immunity to brain-relevant outcomes, as short-chain fatty acids (SCFAs) regulate T-cell cytokine secretion and SCFAs can affect the brain by passing through the BBB and regulating production of neurotransmitters such as serotonin and dopamine, providing a plausible mechanistic chain from gut ecology to adaptive immunity to neuromodulatory chemistry[31].
Behavior and Disease
PNI’s disease interface increasingly emphasizes that brain-driven modulation of immunity can transform inflammatory pathology, as the “brain-evoked transformation of the course of an immune response” is described as offering new possibilities for modulating a wide range of immune disorders from autoimmune disease to cytokine storm and shock[7]. This principle is operationalized in circuit experiments where chemogenetic activation of defined neuronal populations in an immunomodulatory circuit can dramatically transform survival after otherwise lethal immune challenge (approximately 90% alive after LPS challenge), demonstrating that neural control can be sufficient to shift systemic immune outcomes in vivo[7]. The same work reports that activation of TRPA1 vagal neurons protected animals from multiple pathological conditions, reinforcing a mechanistically specific mapping from sensory neuron subtype to systemic disease phenotype control[7].
Translation is constrained not only by biology but by evidence standards and health system adoption, as immunomodulatory options in psychiatric contexts are described as not being included in official clinical guidelines, and clinicians in many countries cannot prescribe off-label compounds unless incorporated into national or international guidelines[14]. This has motivated calls for stratified and adaptive trial designs, including multi-arm and multi-stage trials with stratification by baseline immunological profile, to resolve conflicting results and produce actionable matching rules in the next decade[14]. The field’s stated barrier to guideline incorporation is the need to answer specific precision questions—“which disorder, which patients, which phase of the illness, which compound(s)”—which defines a concrete agenda for precision PNI translation[14].
Long COVID
Long COVID has been framed within PNI as a neuroimmune syndrome in which nervous system–immune system coordination must be understood across space and time, including nervous system influence on immune cell development, distribution, and execution of functions[32]. A complementary framing emphasizes that immune–nervous interactions can be mapped via a spatial framework (communication in brain, within peripheral organs, across distance) and a temporal framework tracking influence across the immune system’s operational lifespan, explicitly expanding PNI models beyond acute response snapshots[32]. In this context, immune–nervous systems are described as collaborating to detect and respond to psychological stress, circadian cues, infection, and tissue injury, which provides conceptual scaffolding for post-infectious symptom persistence models without reducing long COVID to either purely viral persistence or purely psychosomatic explanations[32].
The pandemic has also been proposed as a catalyst event for immunopsychiatric hypothesis generation and validation, conditional on readiness to take a “next leap forward,” suggesting a field-level opportunity to integrate large-scale clinical cohorts with mechanistic neuroimmune models[14]. Clinically, reported epidemiology that an estimated 34% of COVID-19 survivors receive a new neurological or psychiatric diagnosis within six months underscores the urgency of mechanisms-informed follow-up and service planning for neuropsychiatric sequelae[14]. The same discourse emphasizes implementation barriers, noting challenges in “explaining and convincing clinicians, service users, and other stakeholders” of the discipline’s significance, which is itself a translational obstacle to scaling neuroimmune-informed care pathways[14].
Emerging Circuits
A broad synthesis across modern neuroimmunology and PNI emphasizes that the nervous system can shape immune cell development, distribution, and execution of functions, framing neuroimmune coupling as pervasive across immune lifecycle stages rather than confined to acute inflammatory reflexes[32]. Within the same synthesis, immune–nervous interactions are explicitly organized into spatial and temporal frameworks, supporting emerging models in which “dedicated” circuit logic may operate differently across anatomical contexts (brain, organ niches, distance communication) and time scales (developmental programming vs acute responses vs chronic remodeling)[32]. This framework naturally integrates peripheral-to-brain signal routes (neural, humoral, BBB transport, cellular processes) as named mechanisms by which peripheral cytokines communicate with the CNS, enabling experimental designs that test route dominance rather than assuming a single pathway explains heterogeneous disorders[3].
At the intracellular level, convergent signaling hubs provide a mechanistic “alphabet” for circuit-to-cell translation, as pattern-recognition receptors can activate NF-κB, JAK/STAT, and MAPK cascades, while NLRP3 inflammasome biology integrates mitochondrial dysfunction and oxidative stress with IL-1β release and pyroptosis, connecting immune sensing to cellular stress programs in brain and periphery[24]. This intracellular architecture is functionally relevant for PNI because NF-κB is described as an “essential mediator” at the blood–brain interface communicating peripheral inflammatory signals to the CNS, and central blockade of NF-κB can inhibit brain activation markers and inflammation-induced behavioral changes in rodent models, providing a direct mechanistic link between peripheral immune state and CNS network activation[4].
Therapeutic Translation
Therapeutic translation in PNI increasingly depends on precision stratification: markers of inflammation are described as potentially relevant for “more personalised planning and prediction” of antidepressant treatment response, and high baseline TNF-α and IL-6 are associated with treatment resistance in the cited synthesis, motivating biomarker-first treatment algorithms rather than trial-and-error prescribing[5]. Direct evidence of treatment matching is supported by findings that baseline CRP levels can differentially predict outcomes with different antidepressants (CRP–drug interaction reported with effect size and confidence intervals) and by observations that responders can show lower baseline TNF-α levels than non-responders, linking immune phenotypes to pharmacologic responsiveness[33].
A complementary translational mechanism is that inflammation can alter drug disposition, as inflammation may reduce drug bioavailability and dispersibility within organs and antidepressants are mainly metabolized by CYP enzymes (2D6, 1A2, 3A4, 2C19), implying that “grade of inflammation” can impact bioavailability and motivate combined inflammation-marker monitoring and drug monitoring to optimize treatment[5]. This is reinforced by the observation that metabolism of many antidepressants relies on CYP enzymes “mainly located” in hepatic tissue and circulating peripheral blood mononuclear cells, providing a cellular substrate for immune–drug interactions relevant to PNI-informed pharmacology[5].
Mind–body therapies are also being treated as mechanistically testable immunomodulators rather than nonspecific wellness interventions, as mind–body therapies are described as potentially having a neuro-immune regulatory effect mediated by downregulation of NF-κB and reduced inflammation in the cited work[2]. A systematic synthesis evaluating evidence for depression reports that 14 of 21 pieces of evidence supported positive impact of mind–body therapies on pro-inflammatory cytokine levels, suggesting a measurable immune signature for behavioral interventions while simultaneously highlighting evidence heterogeneity and design limitations[34]. Importantly, translation requires identifying moderators and safety constraints, as the question is explicitly raised, “For whom does mindfulness work best and for whom it may be contraindicated,” implying that precision PNI is as relevant to behavioral interventions as to biologics and neuromodulation[35].
Methodological Innovations
Methodological innovation in PNI is increasingly characterized by:
- reliable exposure measurement
- causal circuit perturbation
- multi-scale biological readouts.
Stress exposure measurement has improved via tools reporting excellent reliability for lifetime stressor count and severity outcomes, enabling cumulative exposure modeling without the instability that limits many retrospective stress instruments[13]. These advances are important because lifetime stressor counts have been linked to mental and physical health complaints and sleep quality with reported correlations, supporting quantitative integration of psychological exposure metrics with immune phenotypes[13].
Causal circuit methods have pushed PNI toward mechanistic closure, exemplified by work using single-cell RNA sequencing with functional imaging to identify neuroimmune axis components and show that selective manipulation can suppress pro-inflammatory responses while enhancing anti-inflammatory states, and by perturbation experiments demonstrating that silencing specific circuit nodes can convert immune challenges into unregulated inflammation[6]. At the clinical translation boundary, the field increasingly advocates stratified and adaptive trial designs, including multi-arm and multi-stage designs stratified by baseline immunological profiles, as a methodological requirement to resolve heterogeneous treatment effects and move toward guideline-eligible evidence[14].
The table below summarizes major immune-to-brain and brain-to-immune transmission modes highlighted across the synthesized sources, emphasizing how “routes” become experimentally testable hypotheses rather than narrative metaphors.
Open Questions
A recurring challenge is heterogeneity and specificity: cytokine findings vary substantially across studies, and “not all subjects with depression show increases” in inflammatory cytokines while not all individuals with elevated cytokines have depression, motivating endotype definitions and careful causal inference criteria rather than simple association claims[3]. Relatedly, an inflammation- and cytokine-associated subtype of depression has been proposed specifically because not all high-cytokine subjects develop depressive symptoms and not all depressive patients show elevated mediators, highlighting the need for stratified biomarkers that map onto mechanism, not merely diagnosis[5].
Temporal inference is another limiting factor for both behavioral and immunomodulatory interventions, because trials in psychiatric disorders often assess immunomodulatory drugs over short durations despite these drugs being used long term in other medical conditions, limiting insight into long-term efficacy and safety in psychiatric contexts[36]. Similarly, reviews of mind–body therapy effects on inflammatory markers emphasize that many studies are short term and do not dynamically monitor temporal relationships between inflammatory marker changes and clinical symptoms, motivating longitudinal designs to infer causal mechanisms rather than coincident change[34].
Finally, measurement and route dominance remain open, as existing stress instruments can fail to yield consistent levels over time even when assessing the same time period, creating avoidable noise in stress–immune modeling[13]. Mechanistically, multiple immune-to-brain channels exist (neural, humoral, BBB transport, cellular processes), so a central future direction is determining which routes dominate in specific disorders, stages, and individuals, particularly where interventions could be route-selective (e.g., circuit neuromodulation vs BBB hub targeting vs border checkpoint control)[3].
Conclusion
The most innovative and important contemporary PNI research converges on mechanistic specification across scales: reliable quantification of psychosocial exposures with measurable health links[13], molecular hubs translating peripheral inflammation into CNS state changes (NF-κB, NLRP3, cytokine–HPA coupling)[4, 5, 24], programmable glial and adaptive immune states that drive neurodegenerative and neuroinflammatory pathology[9, 29], and circuit-level control systems in which defined vagal and brainstem nodes bidirectionally regulate peripheral inflammation with causal necessity and sufficiency[6]. Border immunology discoveries—conduits, lymphatic drainage antigen presentation, and leptomeningeal licensing checkpoints—further revise the architecture of immune access to the CNS and create spatially precise therapeutic hypotheses[10–12]. Translation is increasingly framed as a precision engineering problem in which interventions must be matched to immune biotypes and validated via stratified adaptive trials to achieve durable clinical impact and guideline adoption[14].
