NICI

https://docs.google.com/document/d/1wbIhrFC2YSVWF4DmsKrSugCxcoUr1SQb58I3WBCTIDc/edit?usp=sharing

The Neuro-Immuno-Connective Instability (NICI) Model: A Testable Etiological Framework for a Systemic Endophenotype of Autism

Abstract

A substantial and growing body of evidence indicates a profound epidemiological and genetic overlap between Autism Spectrum Disorder (ASD) and Heritable Disorders of Connective Tissue (HDCTs), such as the Ehlers-Danlos Syndromes (EDS). The Neuro-Immuno-Connective Instability (NICI) model is proposed as a comprehensive and empirically testable etiological framework to explain this association. The model posits that a significant subtype of autism is not a primary brain disorder, but rather the neurodevelopmental expression of a systemic condition rooted in genetic variants affecting the Extracellular Matrix (ECM). The initial developmental cascade is hypothesized to involve the malformation of ECM-derived Perineuronal Nets (PNNs), leading to compromised GABAergic inhibition and a foundational Excitatory/Inhibitory (E/I) imbalance, a neurobiological feature supported by post-mortem studies of autistic brain tissue.1 However, the NICI model extends beyond this static developmental defect, proposing that the inherent mechanical instability of atypical connective tissue actively drives chronic mast cell activation via mechanotransduction. This proposed mechano-immune loop generates systemic inflammation, which in turn is hypothesized to compromise gut and blood-brain barrier integrity, leading to chronic neuroinflammation and oxidative stress.2 This hostile central nervous system (CNS) microenvironment provides a plausible mechanism for the ongoing, dynamic degradation of PNNs via the upregulation of Matrix Metalloproteinase-9 (MMP-9), creating a self-amplifying cycle of neurological dysfunction and systemic physiological distress.3 This model provides a robust, falsifiable framework that mechanistically unites the diverse somatic and neurological manifestations of this phenotype. It mandates a paradigm shift towards sequenced, “bottom-up” therapeutic interventions that prioritize systemic physiological stabilization and provides a clear roadmap for future clinical research.

1. A Systemic Foundation: The Epidemiological and Genetic Nexus

The foundation of any compelling etiological theory rests upon robust, statistically significant evidence. In the case of Autism Spectrum Disorder (ASD), a condition long defined by its neurobehavioral characteristics, a profound link with systemic connective tissue disorders challenges conventional, neuro-centric understanding. The strength of this association, corroborated by familial inheritance patterns, suggests that for a significant portion of the autistic population, the condition’s neurological features may be the downstream manifestation of a systemic, heritable condition.

1.1 Quantifying the Overlap: Meta-Analytic Evidence for the Association Between Autism and Joint Hypermobility

The co-occurrence of joint hypermobility and ASD is a statistically validated phenomenon. To establish the most accurate and conservative estimate of this overlap, it is essential to prioritize systematic reviews and meta-analyses over single-cohort studies. A 2025 systematic review and meta-analysis by Baeza-Velasco and colleagues found that the prevalence of clinically assessed generalized joint hypermobility (GJH) in autistic individuals is approximately 31%.4 This figure stands in stark contrast to the estimated prevalence in the general adult population, which varies from 10-20% depending on the age, sex, and specific Beighton-based criteria used.4 While meta-analytic data provide the most reliable prevalence estimate, findings from individual studies are illustrative. A 2022 study by Csecs and colleagues of a clinically-ascertained cohort of neurodivergent adults (including ASD, ADHD, and Tourette Syndrome) found a GJH prevalence of 51% and calculated an odds ratio of 4.51 compared to the general population.2 Similarly, a 2021 case-control study found that adults with ASD had an adjusted odds ratio of 3.1 for GJH and 4.9 for symptomatic GJH.5 The link becomes even more pronounced when examining specific HDCT diagnoses. A large, nationwide Swedish registry study found that individuals with an EDS diagnosis are 7.4 times more likely to also have an ASD diagnosis.6 This methodological triangulation elevates the connection from a clinical curiosity to a public health concern requiring a mechanistic explanation.

Condition/SymptomPrevalence in Autistic/Hypermobile PopulationPrevalence in General PopulationOdds Ratio / Relative RiskSource Type
Generalized Joint Hypermobility (GJH)~31% (clinically assessed) 4~10-20% (adults, varies by criteria) 4OR 3.1-4.9 for ASD diagnosis 5Meta-analysis / Case-Control
Ehlers-Danlos Syndromes (EDS) / HSDSignificantly elevatedDiagnosed prevalence ~194 per 100,000 77.4 times more likely to have ASD 6Registry Study
Obstructive Sleep Apnea (OSA)~39% in EDS 8; 58% in a pediatric ASD cohort 92-5% (pediatric) 9~6 times more likely in EDS vs. general population 8Meta-analysis / Cohort Study
Strabismus13.4% in ASD 10~2-3% 10OR ~4.7 in ASD diagnosis 11Meta-analysis

1.2 Familial Recurrence Patterns: A Clue to Heritable Underpinnings

Compelling evidence for a shared etiological pathway comes from studies of familial inheritance. A 2020 preprint by Casanova and colleagues, based on survey data, reported a preliminary but highly suggestive finding: over 20% of mothers diagnosed with EDS or HSD have autistic children.12 While this figure requires confirmation in larger, prospectively-followed cohorts, its potential significance becomes clear when contextualized against the established familial recurrence rates for autism. According to a large, multi-site 2024 study by Ozonoff and the Baby Siblings Research Consortium, the sibling recurrence rate for autism is 20.2%.13 This rate is approximately seven times higher than in the general population, confirming a strong heritable component.13 It must be made explicit that survey-based maternal EDS/HSD offspring rates are not directly comparable to prospective sibling recurrence rates; however, the proximity of the figures is hypothesis-generating. This pattern is a textbook example of pleiotropy, where a shared genetic liability may influence multiple, seemingly unrelated traits. The same underlying genetic architecture could manifest with a predominantly connective tissue-related phenotype (EDS/HSD) in one family member and a predominantly neurodevelopmental phenotype (ASD) in another. This suggests that for this specific subtype, the search for “autism genes” should be focused on pathways governing connective tissue homeostasis.

1.3 The “Connectivome Theory”: A Unifying Conceptual Framework

The epidemiological and genetic data present a puzzle: how can lax joints be mechanistically related to the neurobiology of autism? The “Connectivome Theory,” as proposed by Tufano and colleagues, provides the essential conceptual framework.14 It challenges the traditional view by proposing that, in this context, the autistic phenotype is one of many manifestations of a multisystem disorder where alterations in connective elements play a crucial pathogenetic role.14 This theory provides a unifying framework that can account for the wide array of somatic manifestations frequently observed in this population, including musculoskeletal issues, gastrointestinal dysfunction, and autonomic dysregulation.14 By reframing the problem from a pure “brain disorder” to a “systemic disorder with profound neurological manifestations,” the Connectivome Theory directs research toward the ubiquitous molecular components of connective tissue, namely the Extracellular Matrix (ECM), as the logical place to find the mechanistic link between the body and the brain.

2. The Developmental Cascade: From Atypical Extracellular Matrix to Foundational Excitatory/Inhibitory Imbalance

The NICI model posits an initial developmental pathway where a genetic predisposition for an HDCT translates directly into an atypical neurodevelopmental trajectory. This cascade begins with the faulty construction of the brain’s essential scaffolding, the ECM, leading to a specific and critical deficit in the brain’s primary inhibitory system. This section outlines the core developmental hypothesis of the model, which proposes a static, structural vulnerability established early in life.

2.1 The Molecular Genetics of an Atypical ECM: Knowns and Unknowns in HDCTs

HDCTs are caused by genetic variants affecting core ECM components. For instance, classical EDS (cEDS) is often associated with variants in genes encoding type V collagen (COL5A1, COL5A2).15 However, the genetic basis for hypermobile EDS (hEDS), the most common subtype, remains largely unknown, though it is understood to be a heritable disorder with a likely complex, polygenic architecture.15 Recent research has identified a potential role for variants in the Kallikrein gene family, particularly KLK15, in some families with hEDS, offering a promising avenue for future investigation.17 This genetic uncertainty does not preclude investigation. It is biologically parsimonious to hypothesize that a systemic defect in ECM components, regardless of its specific genetic origin, will lead to the construction of a structurally and functionally atypical ECM within the developing brain.

2.2 Perineuronal Nets (PNNs): The ECM’s Specialized Scaffolding for GABAergic Interneurons

Within the CNS, the ECM is a dynamic network comprising 10-20% of the brain’s volume.14 A particularly crucial role is played by Perineuronal Nets (PNNs), highly organized condensations of the ECM that enwrap specific neuronal populations.18 PNNs have a precise molecular architecture, consisting of a hyaluronan backbone anchored to the neuronal surface, to which various chondroitin sulfate proteoglycans (CSPGs) are attached.18 Critically, the glycoprotein Tenascin-R (TN-R) acts as a cross-linking agent, binding the CSPGs together and giving the PNN its stable, net-like structure.19 The structural importance of TN-R is underscored by evidence from knockout mice, which exhibit abnormally formed PNNs.19 Most notably, PNNs are predominantly found around fast-spiking, Parvalbumin-positive (PV+) GABAergic interneurons, the workhorses of cortical inhibition.18

2.3 The Core Developmental Hypothesis: Malformed PNNs and Atypical Critical Period Plasticity

The NICI model requires a specific, testable link between an “atypical ECM” and the “E/I imbalance” theory of autism. PNNs provide this mechanistic bridge. The core developmental hypothesis of the NICI model is that faulty molecular building blocks from a systemically atypical ECM are unable to properly assemble into stable, functional PNNs around developing PV+ interneurons. This is a foundational leap based on inference that requires direct validation. This hypothesis is, however, strongly supported by direct human evidence. A 2022 post-mortem study of autistic brain tissue by Brandenburg and Blatt revealed significantly reduced PNN expression specifically in the globus pallidus of ASD cases compared to controls.1 The globus pallidus is a central node in basal ganglia circuits implicated in the motor stereotypies and repetitive behaviors that form a core diagnostic domain of autism.1 This allows the NICI model to propose a specific mechanism: that malformed PNNs in the developing globus pallidus lead to impaired maturation of its inhibitory neurons, disrupting the E/I balance within these circuits and tilting them toward hyperexcitability. Furthermore, if PNN formation is compromised, the “critical periods” of neuroplasticity they regulate may not close properly, providing a powerful explanatory framework for both the vulnerabilities and unique strengths observed in autistic development.

3. The Dynamic Amplifier: A Proposed Mechano-Immuno-Inflammatory Loop

While the developmental cascade explains a foundational vulnerability, it does not account for the dynamic and inflammatory nature of the clinical presentation. The NICI model introduces self-amplifying feedback loops proposed to transform a static defect into a dynamic process. This section outlines these hypotheses, which require rigorous empirical validation.

3.1 Mechanotransduction: A Plausible Trigger for Mast Cell Activation

The first pillar of the model’s dynamic component is mechanotransduction. In HDCTs, the atypical ECM is mechanically unstable. Residing within this matrix, mast cells are primary mechanosensors.2 The altered stiffness and excessive strain of an unstable ECM are hypothesized to place chronic mechanical stress on these cells. This hypothesis is anchored by experimental evidence confirming that mechanical loading can induce mast cell degranulation independent of traditional allergic triggers, providing a direct, non-allergic mechanism that could explain the high prevalence of mast cell activation-like symptoms in the hEDS/HSD population.20

3.2 The Role of Mast Cell Activation Syndrome (MCAS): A Contentious but Critical Link

The proposed pathway provides a plausible biological basis for the clinical entity known as Mast Cell Activation Syndrome (MCAS). While a high comorbidity of MCAS-like symptoms is clinically observed in the hEDS/HSD population, the formal diagnosis of MCAS is contentious and its true prevalence is uncertain. To establish scientific rigor, any investigation must adhere to the strict international consensus criteria for MCAS diagnosis.21 A definitive diagnosis requires:

  1. Typical, episodic, multi-system symptoms of mast cell activation.
  2. A documented, transient increase in a specific mast cell mediator (e.g., serum tryptase) during a flare, with the tryptase criterion requiring a rise of at least 20% above baseline plus 2 ng/mL.22
  3. An objective response of symptoms to medications targeting mast cells or their mediators.21

This mechano-immune interface, if validated, would create a powerful feedback loop. Mast cell proteases (e.g., tryptase) actively degrade the surrounding ECM, further weakening the tissue and perpetuating the cycle of mechanical stress and mast cell activation.2

3.3 Barrier Dysfunction: A Proposed Gateway to Systemic and Central Nervous System Inflammation

Chronic systemic inflammation would have profound consequences for the body’s critical barriers. Mast cell mediators can disrupt tight junction proteins, increasing intestinal permeability (“leaky gut”).2 To test this hypothesis, the lactulose-mannitol differential urinary excretion test serves as the functional gold standard, providing a more reliable measure than less specific serum markers like zonulin.25 The NICI model further posits a dual assault on the blood-brain barrier (BBB). First, the BBB’s basement membrane may be intrinsically weaker due to the systemic ECM defect. Second, the BBB is proposed to be actively disrupted by circulating inflammatory mediators. The gold-standard fluid biomarker for assessing blood-CSF barrier integrity is the CSF/serum albumin quotient (QAlb), which must be adjusted for age.27 The imaging gold standard is Dynamic Contrast-Enhanced MRI (DCE-MRI), though its use must consider gadolinium exposure. The breach of the BBB is a critical step, facilitating the infiltration of peripheral inflammatory molecules into the CNS.

4. The Hostile CNS Microenvironment: A Testable Model of Progressive PNN Degradation

Once the BBB is breached, the inflammatory cascade is imported into the CNS, creating a hostile microenvironment hypothesized to dynamically degrade the brain’s vulnerable inhibitory infrastructure. This section details the central novel contribution and most critical testable prediction of the NICI model.

4.1 Neuroinflammation, Glial Activation, and the Upregulation of Matrix Metalloproteinase-9 (MMP-9)

Inflammatory signals crossing the compromised BBB are potent activators of microglia and astrocytes, the brain’s resident immune cells.2 This establishes a self-sustaining state of chronic neuroinflammation and oxidative stress. However, it is important to acknowledge the “chicken or egg” complexity: primary dysfunction of these glial cells, perhaps related to the atypical ECM of the “glyco-astroglial interface,” could be an initiating or parallel factor, rather than purely a downstream consequence. A critical mechanistic consequence of this neuroinflammatory state is the upregulation of Matrix Metalloproteinases (MMPs), particularly MMP-9.2 Pro-inflammatory cytokines and oxidative stress are potent inducers of MMP-9 expression.2 While serum MMP-9 levels can be measured, they are heavily influenced by peripheral neutrophils, making CSF levels a more direct, albeit invasive, measure of CNS activity.

4.2 The Central Mechanistic Prediction: MMP-9-Mediated Degradation of PNNs

The central mechanistic prediction of the NICI model is that chronic neuroinflammation leads to the dynamic, ongoing degradation of PNNs via the enzymatic activity of MMP-9. This hypothesis is plausible, as MMP-9 is known to cleave core PNN components like aggrecan and brevican.28 While not yet demonstrated in ASD, a clear precedent exists in Multiple Sclerosis (MS), where post-mortem studies show that elevated MMP-9 activity is directly associated with PNN degradation.3 This provides a powerful proof-of-concept. This proposed mechanism moves the pathophysiology beyond a purely developmental malformation to include a lifelong, dynamic process of degradation. The degradation of PNNs would strip the highly vulnerable PV+ interneurons of their essential neuroprotection, impairing their function, further destabilizing the E/I balance, and fueling a self-perpetuating cycle of neuronal impairment and neuroinflammation.2

4.3 The Trauma-Stress Feedback Loop: A Clinical Hypothesis Integrating Systemic and Psychological Distress

This synthesis provides a physiological framework for understanding phenomena like autistic burnout. Chronic psychological stress is a known activator of the HPA axis, leading to the release of mediators like Corticotropin-Releasing Hormone (CRH), which are potent mast cell activators.2 This creates a devastating feedback loop, conceptualized as a Trauma Kindling Hypothesis: Psychological Trauma → Stress Response → Mast Cell Activation → Systemic & Neuroinflammation → MMP-9 Upregulation → PNN Degradation. 

The resulting degradation of PNNs further destabilizes the E/I balance, increasing sensory sensitivity and emotional dysregulation. This heightened neurological vulnerability, in turn, lowers the threshold for future events to be perceived as traumatic, creating a vicious cycle.

Mediator ClassSpecific ExamplesPrimary Source(s)Key Pathological Actions & Testable Prediction
Mast Cell ProteasesTryptase, ChymaseMast CellsECM degradation; Disruption of gut/BBB tight junctions. Prediction: Serum tryptase will show transient elevation post-flare (e.g., >20% + 2 ng/mL) in NICI-phenotype individuals meeting clinical criteria for MCAS.
Pro-inflammatory CytokinesTNF-α, IL-1\beta, IL-6Mast Cells, Microglia, AstrocytesMicroglial/astrocyte activation; Upregulation of MMP expression; Direct disruption of BBB tight junctions. Prediction: CSF levels of TNF-α and IL-6 will correlate positively with QAlb and CSF MMP-9 levels.
Matrix MetalloproteinasesMMP-9Neutrophils, Microglia, AstrocytesDegradation of ECM (e.g., collagen IV in BBB); Cleavage of PNN components (aggrecan, brevican). Prediction: CSF MMP-9 levels will correlate inversely with in-vivo functional proxies for PNN integrity (e.g., MEG/EEG gamma power).

5. Clinical Synthesis: The Compounding Dysregulation Cascade

The complex biology of the NICI model translates into a severe and multifaceted clinical reality. The intersection of mechanical instability, immune dysregulation, and neurological hyperexcitability creates a phenotype uniquely vulnerable to the accumulation of trauma and physiological distress.

5.1 The Neurobiology of Embodied Trauma in a Mechanically Unstable System

For the individual with this phenotype, trauma can be experienced as a chronic, embodied state driven by the inherent instability of their own body.2 This arises from a confluence of chaotic bodily signals: impaired proprioception, chronic nociceptive (pain) signals, and chaotic interoceptive signals from autonomic dysfunction that can mimic panic.2 The brain, operating as a prediction machine, receives a constant stream of “threat-prediction errors” from within the body, forcing the nervous system into a perpetual state of hyperarousal—the neurobiological signature of complex post-traumatic stress.29

5.2 Evidence for Compromised Processing Pathways: Ocular-Motor, Sleep, and Language Function

The physical manifestations of the connective tissue disorder may create a debilitating triad of impairments in the very systems the brain uses to process and discharge traumatic stress.

Ocular-Motor Pathway: Atypical visual processing is a common feature of ASD. Meta-analytic evidence shows a significantly increased prevalence of strabismus (a misalignment of the eyes) in the autistic population, with a calculated odds ratio of approximately 4.72.11 The prevalence of strabismus in autistic individuals is estimated to be around 13.4%, 3-10 times higher than in the general population.10 Strabismus and related issues like convergence insufficiency can profoundly impact the ability to maintain stable binocular fixation. This instability is not merely a mechanical issue; it can fundamentally disrupt trauma processing. The brain’s inability to fuse two misaligned images into a single, coherent picture generates constant, conflicting visual signals, which can trigger a state of high alert and anxiety, effectively forcing the nervous system into a “fight or flight” mode.30 This chronic, low-level threat perception, driven by unreliable visual input, can exacerbate the hypervigilance common in trauma survivors. Furthermore, this visual instability impairs the ability to process non-verbal social cues, which are critical for navigating relational safety and co-regulation.32 Deficits in these foundational visual skills are also linked to poorer performance in cognitive domains like visual memory and reasoning, which are essential for integrating traumatic experiences.33 Finally, these ocular-motor challenges can directly interfere with therapeutic modalities like Eye Movement Desensitization and Reprocessing (EMDR), which often rely on smooth bilateral eye movements to facilitate memory reprocessing, necessitating clinical adaptations.34

Sleep Pathway Disruption: Laxity in the connective tissues of the upper airway creates a high predisposition for Obstructive Sleep Apnea (OSA). A meta-analysis found that individuals with EDS are approximately six times more likely to have OSA than the general population, with a pooled prevalence of around 39%.8 This aligns with findings from a small, single-center pediatric autistic cohort where polysomnography confirmed OSA in 58% of participants, though this figure requires replication in larger samples.9 OSA severely disrupts Rapid Eye Movement (REM) sleep, the primary state for emotional memory consolidation and fear extinction.9

Language Pathway Inhibition: Acute traumatic stress is known to have a downregulating effect on higher cortical functions. Neuroimaging studies of individuals with PTSD have shown that trauma recall is associated with decreased activation in Broca’s area, the brain’s primary center for language production. While not a universal response, this can result in transient mutism or expressive aphasia, blocking the ability to verbalize the experience.35

6. A New Therapeutic Paradigm: A Sequenced, “Bottom-Up” Approach

The NICI model, with its emphasis on a dysregulated body generating constant “bottom-up” threat signals, demands a fundamental rethinking of therapeutic approaches. Traditional “top-down,” cognitively-focused therapies are likely to be insufficient if the underlying physiological instability is not addressed.

6.1 The Rationale for Prioritizing Physiological Stabilization

The NICI model demonstrates why an individual cannot be expected to “think” their way into a state of safety when their body is perpetually sending physiological signals of threat. The ongoing neuroinflammation and PNN degradation proposed by the model would logically impair the synaptic plasticity necessary for therapeutic learning to occur. Therefore, a “bottom-up” paradigm is required—one that prioritizes establishing somatic safety and physiological regulation as the non-negotiable foundation upon which all other therapeutic work is built.

6.2 A Proposed Three-Phase Therapeutic Sequence

The architecture of the NICI model dictates a logical, neurobiologically-informed sequence for intervention. This sequence is a core, falsifiable prediction of the model itself.

Phase 1: Systemic Stabilization. The first phase must focus on interrupting the core self-amplifying loops. Pharmacological interventions may include mast cell stabilizers (e.g., Ketotifen) and glial modulators (e.g., Low-Dose Naltrexone).2 Concurrently, managing critical comorbidities is essential, including treating OSA with Continuous Positive Airway Pressure (CPAP) or a mandibular advancement device (MAD), and managing POTS to stabilize autonomic function.

Phase 2: Somatic Regulation. Once systemic inflammation is dampened, interventions can focus on reducing the body’s threat signaling. This includes somatic therapies that cultivate interoceptive awareness, manual therapies like myofascial release, and organized proprioceptive and vestibular input.2

Phase 3: Trauma Reprocessing. Only after a degree of physiological and somatic safety has been established can the individual effectively engage in therapies designed to process stored traumatic memories. This may include adapted forms of Eye Movement Desensitization and Reprocessing (EMDR) using tactile or auditory bilateral stimulation to accommodate ocular-motor issues.2

7. A Framework for Empirical Validation: Proposed Clinical Research

The complexity of the NICI model necessitates a rigorous, multi-phase, and biomarker-driven research program to move from a theoretical framework to an empirically validated pathway. This program is designed to systematically test the model’s core tenets, acknowledging its inferential leaps and areas of biological complexity.

7.1 Preclinical Research to Strengthen Foundational Links

Before launching full-scale human trials, targeted preclinical work could powerfully substantiate the model’s core assumptions.

Validate the ECM-PNN Link in an hEDS Animal Model: The most significant inferential leap is from systemic hypermobility to faulty PNNs. A crucial preemptive study would involve using a knock-in mouse model expressing an hEDS-associated variant (e.g., in the KLK15 gene).17 Detailed immunohistochemistry on the brains of these animals could quantify the density and structural integrity of PNNs around PV+ interneurons, providing direct mechanistic evidence.

Investigate the Primary Role of Glial Cells: To address the “chicken or egg” question of neuroinflammation, the same hEDS animal models can be used to study microglia and astrocytes at various developmental stages. This would help determine if glial cell abnormalities exist before the onset of significant systemic inflammation or barrier dysfunction.

7.2 Phase I: Rigorous Phenotyping and Correlational Analysis

The objective of this initial phase is to rigorously define the NICI phenotype and establish key clinical correlations using objective measures.

Methodology: A large cohort (N=1000+) should be recruited across four distinct groups: (1) Autistic + Hypermobile; (2) Autistic + Non-hypermobile; (3) Neurotypical + Hypermobile; and (4) Neurotypical + Non-hypermobile.

Phenotyping: Objective clinical measures must be used. Hypermobility must be assessed by a trained clinician using the Beighton Score. An objective ophthalmological assessment should screen for strabismus, convergence insufficiency, stereoacuity, and suppression.36 An exploratory genetic component should include targeted sequencing of candidate genes (e.g., Kallikrein family) for later stratification.17 A baseline serum marker for neurological stress, such as Neurofilament Light Chain (NfL), should be added.37

Analysis: The analysis plan, including all primary and secondary outcomes and covariates (e.g., age, sex, BMI, medications, site), must be pre-registered to separate confirmatory from exploratory findings.38

7.3 Phase II: Longitudinal Biomarker Validation and Interventional Trials

The objective of this advanced phase is to mechanistically validate the NICI cascade and test the efficacy of a sequenced intervention.

Part A: Longitudinal Biomarker Study: Smaller, deeply-phenotyped sub-cohorts (N=50-100 per group) would be recruited. Data would be collected at baseline and follow-ups to measure activity at each stage of the NICI cascade using a comprehensive, gold-standard biomarker panel (see Table 3).

Part B: Multi-Arm Randomized Controlled Trial (RCT): Individuals meeting the empirically defined NICI phenotype would be recruited for a 24-week, multi-arm RCT.

Arms:

  • Arm 1 (Full NICI Protocol): 12 weeks of Systemic Stabilization (e.g., mast cell stabilizer + LDN) followed by 12 weeks of Somatic Regulation therapy.
  • Arm 2 (Somatic Only): 12 weeks of an active control (e.g., psychoeducation) followed by 12 weeks of Somatic Regulation therapy.
  • Arm 3 (Systemic Only): 12 weeks of Systemic Stabilization followed by 12 weeks of an active control.
  • Arm 4 (Neuroinflammation First): 12 weeks of a glial modulator (e.g., LDN) alone, followed by 12 weeks of an active control.
  • Arm 5 (Active Control): 24 weeks of an active control (psychoeducation).

Outcomes: Primary outcomes would be changes in composite scores of pain, anxiety, and autonomic symptoms. Secondary outcomes would include changes in key biomarkers. The trial should pre-specify a stratified analysis based on baseline biomarker levels (e.g., QAlb, MMP-9) and comorbidities (e.g., PSG-verified OSA).

NICI Pathway StageProposed Biomarker(s)Sample TypeRationale & Gold Standard Status
Genetic FoundationWhole Genome/Targeted Sequencing (e.g., KLK family)Blood/SalivaEstablish underlying genetic predisposition and perform pathway analysis.17
Mechano-Immune Loop (MCAS)Serum Tryptase (baseline & post-flare); Urinary N-methylhistamine, Leukotriene E4Blood, UrineObjective evidence of systemic mast cell activation, adhering to Valent et al. consensus criteria.21
ECM/PNN TurnoverCSF MMP-9 & TIMP-1; Gelatin ZymographyCSFDirect index of CNS ECM degradation balance. Superior to serum MMP-9, which is influenced by peripheral neutrophils.41
Gut Barrier IntegrityLactulose-Mannitol TestUrineFunctional Gold Standard for intestinal hyperpermeability. Superior to less reliable serum markers like zonulin.25
BBB IntegrityCSF/Serum Albumin Quotient (QAlb); Dynamic Contrast-Enhanced MRI (DCE-MRI)CSF, Blood, ImagingFluid Biomarker Gold Standard (QAlb) and Imaging Gold Standard (DCE-MRI) for BBB disruption. Superior to non-specific serum S100B.27
Neuroinflammation/Glial ActivationCSF Cytokines (TNF-α, IL-1\beta, IL-6); CSF GFAP, Neopterin, sTREM2CSFDirect measure of CNS inflammatory state and specific glial (astrocyte, microglia) activation.42
Neuronal StressSerum Neurofilament Light Chain (NfL)BloodMinimally invasive marker of underlying neurological stress/axonal damage.37
E/I Imbalance (Functional Proxies)^{1}H-MRS (GABA/Glutamate); MEG/EEG Gamma Power; TMS-EEG (SICI/LICI)In-vivo Neuroimaging/ ElectrophysiologyDirect (MRS) and functional (EEG/TMS) measures of the core neurochemical deficit and cortical inhibition.44

Conclusion: The NICI Model as a Unifying Etiological Framework

The Neuro-Immuno-Connective Instability (NICI) Model offers a necessary evolution of the foundational Hypermobility-Autism Hypothesis, providing a deeper, dynamic framework that explains the multisystemic, inflammatory, and often progressive nature of this complex phenotype. NICI integrates mechanical instability, immune activation, barrier dysfunction, and metabolic vulnerability into a coherent, self-amplifying cascade. It bridges the gap between the physical reality of hypermobility and the neurobiological reality of autism, proposing a testable mechanism by which chronic inflammation and oxidative stress can dynamically degrade the brain’s inhibitory infrastructure throughout life.

At its core, the NICI cascade can be conceptualized as a state of progressive allostatic overload occurring in a system with fundamentally compromised resilience at multiple biological levels:

  • The structural resilience of the ECM is low, leading to mechanical instability.
  • The barrier resilience of the gut and BBB is low, predisposing them to compromise.
  • The neuro-regulatory resilience of the CNS is low, due to a foundational E/I imbalance.
  • The metabolic resilience of key inhibitory neurons (PV+) is low, making them vulnerable to oxidative stress.

Into this inherently fragile system, the NICI model introduces multiple powerful, self-amplifying feedback loops. The result is not a static disorder, but a progressive, multisystemic decompensation. This integrated understanding mandates a shift in clinical practice towards a systems-based approach. A successful validation of this model would necessitate a radical change in the standard of care for this large subtype of autism, moving from siloed behavioral and psychiatric management to an integrated approach involving neurologists, immunologists, rheumatologists, and gastroenterologists. Rigorous testing of the NICI model, following the empirical roadmap outlined herein, holds the potential to transform the understanding and clinical management of this significant and widely misunderstood condition.

References

  1. Brandenburg, C., & Blatt, G. J. (2022). Region-Specific Alterations of Perineuronal Net Expression in Postmortem Autism Brain Tissue. Frontiers in Molecular Neuroscience, 15, 838918. 
  2. Csecs, J. L., et al. (2022). Joint Hypermobility Links Neurodivergence to Dysautonomia and Pain. Frontiers in Psychiatry, 12, 786916. 
  3. Gray, E., et al. (2008). Elevated Matrix Metalloproteinase-9 and Degradation of Perineuronal Nets in Cerebrocortical Multiple Sclerosis Plaques. Journal of Neuropathology & Experimental Neurology, 67(9), 888–899. 
  4. Baeza-Velasco, C., et al. (2025). Autism in the context of joint hypermobility, hypermobility spectrum disorders, and Ehlers–Danlos syndromes: A systematic review and prevalence meta-analyses. Autism. 
  5. Baeza-Velasco, C., et al. (2021). A case-control study of prevalence of joint hypermobility in adults with autism spectrum disorder. Frontiers in Psychiatry, 12, 803334.
  6. Cederlöf, M., et al. (2016). Nationwide population-based cohort study of psychiatric disorders in individuals with Ehlers-Danlos syndrome or hypermobility syndrome and their siblings. BMC Psychiatry, 16(1), 207.
  7. Demmler, J. C., et al. (2019). Diagnosed prevalence of Ehlers-Danlos syndrome and hypermobility spectrum disorder in Wales, UK: a national electronic cohort study and case-control comparison. BMJ Open, 9(11), e031365.
  8. Sedky, K., et al. (2019). Prevalence of Obstructive Sleep Apnea in Joint Hypermobility Syndrome: A Systematic Review and Meta-Analysis. Journal of Clinical Sleep Medicine, 15(2), 293–299. 
  9. Tomkies, A., et al. (2019). Obstructive Sleep Apnea in Children With Autism. Journal of Clinical Sleep Medicine, 15(10), 1469–1476.
  10. Black, K., et al. (2021). Prevalence of Strabismus in Individuals on the Autism Spectrum: A Meta-analysis. Journal of American Association for Pediatric Ophthalmology and Strabismus, 25(5), 274.e1-274.e7. 
  11. Bafteh, P. R., et al. (2023). Association between Autism Spectrum Disorder (ASD) and vision problems: A systematic review and meta-analysis. Journal of Optometry, 16(4), 285-295.
  12. Casanova, E. L., et al. (2020). Increased rate of joint hypermobility in autism and related neurodevelopmental conditions is linked to dysautonomia and pain. medRxiv.
  13. Ozonoff, S., et al. (2024). Familial Recurrence of Autism: Updates From the Baby Siblings Research Consortium. Pediatrics, 154(2), e2023065297.
  14. Tufano, M., et al. (2022). The “Connectivome Theory”: A New Model to Understand Autism Spectrum Disorders. Frontiers in Neuroscience, 16, 889237.
  15. Aring, A. M., et al. (2021). Ehlers-Danlos Syndromes: A Review of the Literature and Treatment Recommendations. American Family Physician, 103(8), 481-490.
  16. Tinkle, B., et al. (2017). Hypermobile Ehlers-Danlos syndrome (a.k.a. Ehlers-Danlos syndrome Type III and Ehlers-Danlos syndrome hypermobility type): Clinical description and natural history. American Journal of Medical Genetics Part C: Seminars in Medical Genetics, 175(1), 48-69.
  17. Gensemer, C., et al. (2024). Variants in the Kallikrein Gene Family and Hypermobile Ehlers-Danlos Syndrome. medRxiv.
  18. Mueller-Buehl, C., et al. (2022). Brevican, Neurocan, Tenascin-C, and Tenascin-R Act as Important Regulators of the Interplay Between Perineuronal Nets, Synaptic Integrity, Inhibitory Interneurons, and Otx2. Frontiers in Cell and Developmental Biology, 10, 886527.
  19. Morawski, M., et al. (2014). Tenascin-R promotes assembly of the extracellular matrix of perineuronal nets via clustering of aggrecan. Philosophical Transactions of the Royal Society B: Biological Sciences, 369(1654), 20140046.
  20. Fowlkes, V., et al. (2013). Mechanical loading promotes mast cell degranulation via RGD-integrin dependent pathways. Journal of Biomechanics, 46(4), 768-774.
  21. Valent, P., et al. (2012). Definitions, criteria and global classification of mast cell disorders with special reference to mast cell activation syndromes: a consensus proposal. International Archives of Allergy and Immunology, 157(3), 215-225.
  22. Valent, P., et al. (2019). Why the 20% + 2 Tryptase Formula Is a Diagnostic Gold Standard for Severe Systemic Mast Cell Activation and Mast Cell Activation Syndrome. International Archives of Allergy and Immunology, 180(1), 44-51.
  23. Valent, P., et al. (2019). Proposed Diagnostic Algorithm for Patients with Suspected Mast Cell Activation Syndrome. Journal of Allergy and Clinical Immunology: In Practice, 7(4), 1125-1133.e2.
  24. Valent, P., et al. (2021). Updated Diagnostic Criteria and Classification of Mast Cell Disorders: A Consensus Proposal. HemaSphere, 5(11), e646.
  25. D’Eufemia, P., et al. (1996). Abnormal intestinal permeability in children with autism. Acta Paediatrica, 85(9), 1076-1079.
  26. Ganda, A., et al. (2022). A Meta-Analysis of Lactulose:Mannitol Ratio in Coeliac and Crohn’s Disease. Diagnostics, 12(1), 101.
  27. Skillbäck, T., et al. (2017). CSF/serum albumin ratio in clinical neurochemistry. Clinica Chimica Acta, 471, 16-20.
  28. Vafadari, B., et al. (2022). Matrix Metalloproteinase-9 as an Important Contributor to the Pathophysiology of Depression. Frontiers in Neurology, 13, 861843.
  29. van der Kolk, B. A. (2014). The body keeps the score: Brain, mind, and body in the healing of trauma. Viking.
  30. Padula, W. V., et al. (2009). Consequence of spatial visual processing dysfunction caused by traumatic brain injury (TBI). Brain Injury, 23(6), 512-520.
  31. Threatt, T., et al. (2020). Ocular Autonomic Nervous System: An Update from Anatomy to Physiological Functions. Eye & Contact Lens, 46(Suppl 2), S112-S117.
  32. Goodrich, G. L., et al. (2013). Visual function, TBI, and posttraumatic stress disorder. Journal of Rehabilitation Research and Development, 50(6), 789-798.
  33. Masters, M. C., et al. (2016). Vision diagnoses are common after concussion in adolescents. Clinical pediatrics, 55(8), 776-781.
  34. Shapiro, F. (2001). Eye movement desensitization and reprocessing: Basic principles, protocols, and procedures (2nd ed.). Guilford Press.
  35. Hull, A. M. (2002). Neuroimaging findings in post-traumatic stress disorder: systematic review. The British Journal of Psychiatry, 181(2), 102-110.
  36. Stereo Optical. (n.d.). Original Stereo Fly Stereotest. Retrieved from https://www.stereooptical.com/products/stereotests-color-tests/original-stereo-fly/
  37. Gaetz, W. (2020). Neurofilament light chain as a biomarker in neurological disorders. Journal of Neurology, Neurosurgery & Psychiatry, 91(8), 870-878.
  38. Center for Open Science. (n.d.). Preregister. Retrieved from https://www.cos.io/initiatives/prereg
  39. Srivastava, S. B. (2019). Preregistration in the Social and Behavioral Sciences. Social Psychology, 50(5), 261-272.
  40. Office of Planning, Research, & Evaluation. (2022). Pre-registering studies: What is it, how do you do it, and why is it worthwhile?
  41. Henneveld, L., et al. (2024). Glial fibrillary acidic protein as a biomarker in brain disorders. Molecular Psychiatry.
  42. Hesling, I., et al. (2016). sTREM2 cerebrospinal fluid levels are a potential biomarker for microglia activity in early-stage Alzheimer’s disease and associate with neuronal injury markers. EMBO Molecular Medicine, 8(10), 1154-1168.
  43. Holwerda, K. M., et al. (2024). Resting-state gamma-frequency oscillations in schizophrenia: a systematic review and meta-analysis. Scientific Reports, 14(1), 13351.
  44. Kim, M., et al. (2013). Gamma activity in the prefrontal cortex during motor imagery is a potential indicator of BCI performance. Frontiers in Human Neuroscience, 7, 848.
  45. Premoli, I., et al. (2018). Short-interval and long-interval intracortical inhibition of TMS-evoked EEG potentials. Brain Stimulation, 11(3), 547-556.
  46. Snoek-van Beurden, P. A., & Von den Hoff, J. W. (2005). Zymographic techniques for the analysis of matrix metalloproteinases and their inhibitors. BioTechniques, 38(1), 73-83.