<?xml version="1.0" encoding="UTF-8"?>
<!DOCTYPE article PUBLIC "-//NLM//DTD JATS (Z39.96) Journal Publishing DTD v1.3 20210610//EN" "JATS-journalpublishing1-3-mathml3.dtd">
<article xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns:ali="http://www.niso.org/schemas/ali/1.0/" article-type="research-article" dtd-version="1.3" xml:lang="EN">
<front>
<journal-meta>
<journal-id journal-id-type="publisher-id">Front. Dement.</journal-id>
<journal-title-group>
<journal-title>Frontiers in Dementia</journal-title>
<abbrev-journal-title abbrev-type="pubmed">Front. Dement.</abbrev-journal-title>
</journal-title-group>
<issn pub-type="epub">2813-3919</issn>
<publisher>
<publisher-name>Frontiers Media S.A.</publisher-name>
</publisher>
</journal-meta>
<article-meta>
<article-id pub-id-type="doi">10.3389/frdem.2026.1841647</article-id>
<article-version article-version-type="Version of Record" vocab="NISO-RP-8-2008"/>
<article-categories>
<subj-group subj-group-type="heading">
<subject>Hypothesis and Theory</subject>
</subj-group>
</article-categories>
<title-group>
<article-title>Hierarchical-circular model of biological memory: a multilevel hypothesis for pathogenesis and allostatic integrity in Alzheimer&#x2019;s disease and related dementias</article-title>
</title-group>
<contrib-group>
<contrib contrib-type="author" corresp="yes">
<name>
<surname>Ruesga Mundo</surname>
<given-names>Samuel</given-names>
</name>
<xref ref-type="aff" rid="aff1"/>
<xref ref-type="corresp" rid="c001"><sup>&#x002A;</sup></xref>
<xref ref-type="author-notes" rid="fn0001"><sup>&#x2020;</sup></xref>
<uri xlink:href="https://loop.frontiersin.org/people/3011163"/>
<role vocab="credit" vocab-identifier="https://credit.niso.org/" vocab-term="Writing &#x2013; original draft" vocab-term-identifier="https://credit.niso.org/contributor-roles/writing-original-draft/">Writing &#x2013; original draft</role>
<role vocab="credit" vocab-identifier="https://credit.niso.org/" vocab-term="methodology" vocab-term-identifier="https://credit.niso.org/contributor-roles/methodology/">Methodology</role>
<role vocab="credit" vocab-identifier="https://credit.niso.org/" vocab-term="Data curation" vocab-term-identifier="https://credit.niso.org/contributor-roles/data-curation/">Data curation</role>
<role vocab="credit" vocab-identifier="https://credit.niso.org/" vocab-term="validation" vocab-term-identifier="https://credit.niso.org/contributor-roles/validation/">Validation</role>
<role vocab="credit" vocab-identifier="https://credit.niso.org/" vocab-term="conceptualization" vocab-term-identifier="https://credit.niso.org/contributor-roles/conceptualization/">Conceptualization</role>
<role vocab="credit" vocab-identifier="https://credit.niso.org/" vocab-term="visualization" vocab-term-identifier="https://credit.niso.org/contributor-roles/visualization/">Visualization</role>
<role vocab="credit" vocab-identifier="https://credit.niso.org/" vocab-term="investigation" vocab-term-identifier="https://credit.niso.org/contributor-roles/investigation/">Investigation</role>
<role vocab="credit" vocab-identifier="https://credit.niso.org/" vocab-term="supervision" vocab-term-identifier="https://credit.niso.org/contributor-roles/supervision/">Supervision</role>
<role vocab="credit" vocab-identifier="https://credit.niso.org/" vocab-term="Funding acquisition" vocab-term-identifier="https://credit.niso.org/contributor-roles/funding-acquisition/">Funding acquisition</role>
<role vocab="credit" vocab-identifier="https://credit.niso.org/" vocab-term="Project administration" vocab-term-identifier="https://credit.niso.org/contributor-roles/project-administration/">Project administration</role>
<role vocab="credit" vocab-identifier="https://credit.niso.org/" vocab-term="resources" vocab-term-identifier="https://credit.niso.org/contributor-roles/resources/">Resources</role>
<role vocab="credit" vocab-identifier="https://credit.niso.org/" vocab-term="Writing &#x2013; review &#x0026; editing" vocab-term-identifier="https://credit.niso.org/contributor-roles/writing-review-editing/">Writing &#x2013; review &#x0026; editing</role>
<role vocab="credit" vocab-identifier="https://credit.niso.org/" vocab-term="Formal analysis" vocab-term-identifier="https://credit.niso.org/contributor-roles/formal-analysis/">Formal analysis</role>
<role vocab="credit" vocab-identifier="https://credit.niso.org/" vocab-term="software" vocab-term-identifier="https://credit.niso.org/contributor-roles/software/">Software</role>
</contrib>
</contrib-group>
<aff id="aff1"><institution>High Specialty Medical Unit, Western National Medical Center, Mexican Institute of Social Security (IMSS)</institution>, <city>Guadalajara</city>, <state>Jalisco</state>, <country country="mx">Mexico</country></aff>
<author-notes>
<corresp id="c001"><label>&#x002A;</label>Correspondence: Samuel Ruesga Mundo, <email xlink:href="mailto:ruesga_77@hotmail.com">ruesga_77@hotmail.com</email></corresp>
<fn fn-type="other" id="fn0001"><label>&#x2020;</label><p>ORCID: Samuel Ruesga Mundo, <uri xlink:href="https://orcid.org/0009-0000-3484-0460">orcid.org/0009-0000-3484-0460</uri></p></fn>
</author-notes>
<pub-date publication-format="electronic" date-type="pub" iso-8601-date="2026-06-24">
<day>24</day>
<month>06</month>
<year>2026</year>
</pub-date>
<pub-date publication-format="electronic" date-type="collection">
<year>2026</year>
</pub-date>
<volume>5</volume>
<elocation-id>1841647</elocation-id>
<history>
<date date-type="received">
<day>28</day>
<month>03</month>
<year>2026</year>
</date>
<date date-type="rev-recd">
<day>21</day>
<month>05</month>
<year>2026</year>
</date>
<date date-type="accepted">
<day>09</day>
<month>06</month>
<year>2026</year>
</date>
</history>
<permissions>
<copyright-statement>Copyright &#x00A9; 2026 Ruesga Mundo.</copyright-statement>
<copyright-year>2026</copyright-year>
<copyright-holder>Ruesga Mundo</copyright-holder>
<license>
<ali:license_ref start_date="2026-06-24">https://creativecommons.org/licenses/by/4.0/</ali:license_ref>
<license-p>This is an open-access article distributed under the terms of the <ext-link ext-link-type="uri" xlink:href="https://creativecommons.org/licenses/by/4.0/">Creative Commons Attribution License (CC BY)</ext-link>. The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.</license-p>
</license>
</permissions>
<abstract>
<sec>
<title>Introduction</title>
<p>Alzheimer&#x2019;s disease and related dementias remain largely resistant to disease-modifying therapies, despite decades of research focused on linear neuropathological pathways such as beta-amyloid and tau. Persistent paradoxes&#x2014;including the dissociation between pathology burden and clinical expression, the impact of early-life stress, and the role of systemic factors&#x2014;indicate the need for integrative theoretical frameworks. This article proposes a multilevel hypothesis conceptualizing dementias as disorders of biological memory and allostatic integrity rather than isolated brain pathologies.</p>
</sec>
<sec>
<title>Hypothesis</title>
<p>The Hierarchical-Circular Model of Biological Memory posits that dementia emerges from progressive disruptions in a circular, multilevel system that encodes and stabilizes biological information across the lifespan. The model is organized around the unifying principle &#x201C;Signal &#x2192; Plasticity &#x2192; Stable State&#x201D; and integrates five interconnected levels: (1) morphogenetic programming and genetic architecture, (2) epigenetic molecular memory, (3) allostatic load and systemic physiological adaptation, (4) the Psychological&#x2013;Neurological&#x2013;Endocrine&#x2013;Immunological (PNEI) network, and (5) interoceptive&#x2013;neuronal integration. At any level, perturbation can propagate bidirectionally through the system, establishing maladaptive stable states that manifest clinically as dementia.</p>
</sec>
<sec>
<title>Development of the hypothesis</title>
<p>Through a structured synthesis of longitudinal, mechanistic, and multisystem studies (2010&#x2013;2025), the model specifies how gene&#x2013;environment interactions, epigenetic modifications, cumulative allostatic load, neuroimmune dynamics, and altered interoceptive timescales jointly shape vulnerability and resilience. The concept of allostatic integrity is introduced as a dynamic systems-level property&#x2014;distinct from allostatic load&#x2014;that explains why similar neuropathological burdens may result in divergent clinical trajectories. Distinct dementia phenotypes are proposed to reflect different patterns of circular reinforcement across the five levels.</p>
</sec>
<sec>
<title>Testable predictions</title>
<p>This framework generates concrete, falsifiable predictions: (1) composite indices of allostatic integrity will outperform single biomarkers in predicting conversion from mild cognitive impairment to dementia; (2) multidomain interventions targeting more than one system level will have multiplicative, rather than additive, effects on slowing cognitive decline; (3) patients with similar amyloid/tau profiles but contrasting allostatic integrity will show markedly different trajectories of clinical progression; and (4) allostatic integrity moderates the protective effect of cognitive reserve, a pattern not predicted by reserve frameworks alone.</p>
</sec>
<sec>
<title>Conclusion</title>
<p>The Hierarchical-Circular Model of Biological Memory offers a unifying hypothesis for Alzheimer&#x2019;s disease and related dementias that bridges genetic, epigenetic, physiological, neuroimmune, and interoceptive processes across the lifespan. By reframing dementias as failures of biological memory and allostatic integrity, the model provides a conceptual roadmap for mechanistic research, multidomain prevention, and personalized treatment strategies.</p>
</sec>
</abstract>
<kwd-group>
<kwd>Alzheimer&#x2019;s disease</kwd>
<kwd>dementia</kwd>
<kwd>allostatic load</kwd>
<kwd>epigenetic memory</kwd>
<kwd>interoception</kwd>
<kwd>biological memory</kwd>
<kwd>multilevel model</kwd>
<kwd>PNEI network</kwd>
</kwd-group>
<funding-group>
<funding-statement>The author(s) declared that financial support was not received for this work and/or its publication.</funding-statement>
</funding-group>
<counts>
<fig-count count="1"/>
<table-count count="2"/>
<equation-count count="0"/>
<ref-count count="44"/>
<page-count count="14"/>
<word-count count="10444"/>
</counts>
<custom-meta-group>
<custom-meta>
<meta-name>section-at-acceptance</meta-name>
<meta-value>Cellular and Molecular Mechanisms of Dementia</meta-value>
</custom-meta>
</custom-meta-group>
</article-meta>
</front>
<body>
<sec sec-type="intro" id="sec1">
<label>1</label>
<title>Introduction</title>
<p>Alzheimer&#x2019;s disease and related dementias (ADRD) represent one of the greatest public health challenges of the 21st century, with global prevalence expected to rise sharply in the coming decades (<xref ref-type="bibr" rid="ref12">GBD 2019 Dementia Forecasting Collaborators, 2022</xref>). Despite unprecedented advances in neuroimaging, biomarkers, and genetics, disease-modifying therapies have produced only modest clinical benefits, and most pharmacological trials continue to fail (<xref ref-type="bibr" rid="ref6">Cummings et al., 2024</xref>). This therapeutic impasse has fueled growing recognition that predominant linear models&#x2014;centered on beta-amyloid, tau, and single-pathway cascades&#x2014;are insufficient to capture the full complexity of dementia pathogenesis (<xref ref-type="bibr" rid="ref25">Long and Holtzman, 2019</xref>).</p>
<p>Recent conceptual and diagnostic advances have deepened, rather than resolved, this paradox. The 2024 National Institute on Aging&#x2013;Alzheimer&#x2019;s Association (NIA&#x2013;AA) biological criteria formally define Alzheimer&#x2019;s disease by its underlying pathology and biomarker profile, rather than by clinical symptoms (<xref ref-type="bibr" rid="ref17">Jack et al., 2018</xref>; <xref ref-type="bibr" rid="ref16">Jack Jr et al., 2024</xref>). This shift has clarified the biological continuum of disease but has also highlighted a persistent dissociation between neuropathological burden and clinical expression: individuals with substantial amyloid and tau pathology may remain cognitively intact, whereas others with relatively modest biomarker changes exhibit rapid decline (<xref ref-type="bibr" rid="ref32">Savva et al., 2009</xref>). These discrepancies suggest that dementia cannot be fully understood as a simple downstream consequence of aggregated proteins, but instead reflects broader disturbances in how biological systems encode, integrate, and stabilize information over time.</p>
<p>Converging evidence points to the importance of lifespan, multisystem influences that are only partially addressed by current frameworks (<xref ref-type="bibr" rid="ref24">Livingston et al., 2020</xref>; <xref ref-type="bibr" rid="ref11">Fratiglioni et al., 2020</xref>). Large longitudinal cohorts now show that elevated allostatic load&#x2014;a composite index of cumulative physiological &#x201C;wear and tear&#x201D; across metabolic, cardiovascular, and inflammatory systems&#x2014;is prospectively associated with increased risk for all-cause dementia and specific subtypes, including Alzheimer&#x2019;s disease and vascular dementia (<xref ref-type="bibr" rid="ref31">Ruan et al., 2025</xref>; <xref ref-type="bibr" rid="ref35">Smith et al., 2025</xref>; <xref ref-type="bibr" rid="ref30">P&#x00E9;rez-Carrasco et al., 2025</xref>). These findings support the view that chronic stress physiology functions as a key mediator between environmental exposures and late-life cognitive decline (<xref ref-type="bibr" rid="ref36">Souza-Talarico et al., 2025</xref>). At the same time, emerging studies link adverse experiences across the life-course, including childhood abuse and adult economic hardship, to higher dementia risk (<xref ref-type="bibr" rid="ref24">Livingston et al., 2020</xref>; <xref ref-type="bibr" rid="ref21">Lee et al., 2025</xref>), underscoring the role of social and psychological stressors in shaping long-term vulnerability.</p>
<p>At the molecular level, gene&#x2013;environment interactions and epigenetic mechanisms provide a plausible route by which early and chronic stress exposures become biologically embedded (<xref ref-type="bibr" rid="ref28">N&#x00F6;thling et al., 2025</xref>; <xref ref-type="bibr" rid="ref43">Yusupov et al., 2024</xref>). Work on stress-responsive genes such as FKBP5 and risk alleles like APOE-&#x03B5;4 demonstrates that specific genetic variants modulate epigenetic plasticity, hypothalamic&#x2013;pituitary&#x2013;adrenal (HPA) axis reactivity, and neurophysiological activity patterns in aging populations (<xref ref-type="bibr" rid="ref44">Zannas et al., 2016</xref>; <xref ref-type="bibr" rid="ref42">Wang et al., 2022</xref>; <xref ref-type="bibr" rid="ref41">Verduzco Espinoza et al., 2025</xref>). These data reinforce the idea that genetic architecture and epigenetic &#x201C;memory&#x201D; jointly calibrate stress responses and neural network function across the lifespan, thereby influencing who will convert from mild cognitive impairment to dementia under comparable neuropathological conditions (<xref ref-type="bibr" rid="ref29">Panitch et al., 2024</xref>; <xref ref-type="bibr" rid="ref20">Klengel et al., 2013</xref>).</p>
<p>In parallel, advances in systems neuroscience and interoception research suggest that dementias involve fundamental disruptions in how the brain tracks, predicts, and regulates bodily signals (<xref ref-type="bibr" rid="ref5">Critchley and Garfinkel, 2017</xref>; <xref ref-type="bibr" rid="ref18">Khalsa et al., 2018</xref>). Recent work in behavioral-variant frontotemporal dementia has shown altered intrinsic neural timescales of interoceptive processing, pointing to a breakdown in allostatic&#x2013;interoceptive networks that support adaptive regulation of internal states (<xref ref-type="bibr" rid="ref14">Hazelton et al., 2025a</xref>, <xref ref-type="bibr" rid="ref15">b</xref>). Similar alterations in interoceptive accuracy and network connectivity have been implicated in Alzheimer&#x2019;s disease (<xref ref-type="bibr" rid="ref14">Hazelton et al., 2025a</xref>, <xref ref-type="bibr" rid="ref15">b</xref>), suggesting that disturbances in the dynamic integration of bodily and environmental information may constitute a transdiagnostic mechanism of cognitive and behavioral decline.</p>
<p>Taken together, these strands of evidence reveal a fragmented but convergent picture: dementia risk and progression are shaped by interactions among genetic predisposition, epigenetic molecular memory, cumulative allostatic load, multisystem (psychoneuroendocrine&#x2013;immune) dynamics, and interoceptive&#x2013;neuronal integration (<xref ref-type="bibr" rid="ref26">McEwen and Stellar, 1993</xref>; <xref ref-type="bibr" rid="ref7">Danese and McEwen, 2012</xref>; <xref ref-type="bibr" rid="ref34">Slavich and Irwin, 2014</xref>; <xref ref-type="bibr" rid="ref10">Feng et al., 2025</xref>). However, current models typically examine these components in isolation, lacking a unifying framework capable of explaining how perturbations at one level propagate across others to produce stable, maladaptive states (<xref ref-type="bibr" rid="ref25">Long and Holtzman, 2019</xref>; <xref ref-type="bibr" rid="ref33">Serrano-Pozo et al., 2024</xref>). The field therefore lacks an integrative language to describe dementia not merely as neurodegeneration, but as a systems-level failure of biological memory and allostatic integrity&#x2014;the capacity of the organism to maintain adaptive circular information flow across multiple levels of organization.</p>
<p>In this article, a Hierarchical-Circular Model of Biological Memory is proposed as a multilevel hypothesis for Alzheimer&#x2019;s disease and related dementias. The model is organized around the unifying principle &#x201C;Signal &#x2192; Plasticity &#x2192; Stable State&#x201D; and articulates five interconnected levels: (1) morphogenetic programming and genetic architecture, (2) epigenetic molecular memory, (3) allostatic load and systemic physiological adaptation, (4) the Psychological&#x2013;Neurological&#x2013;Endocrine&#x2013;Immunological (PNEI) network, and (5) interoceptive&#x2013;neuronal integration. By treating dementias as emergent outcomes of circular, bidirectional interactions across these levels, the framework aims to (a) reconcile key clinical and epidemiological paradoxes, (b) generate testable predictions regarding biomarkers and multidomain interventions, and (c) provide a conceptual roadmap for a more integrative, systems-oriented dementia research agenda.</p>
</sec>
<sec id="sec2">
<label>2</label>
<title>The hierarchical-circular model of biological memory: conceptual framework</title>
<sec id="sec3">
<label>2.1</label>
<title>The unifying principle: &#x201C;Signal &#x2192; Plasticity &#x2192; Stable State&#x201D;</title>
<p>At the core of the Hierarchical-Circular Model lies a unifying principle applicable across all levels of biological organization: Signal &#x2192; Plasticity &#x2192; Stable State. This principle posits that biological systems continuously receive signals from their internal and external environments, undergo plastic changes in response to these signals, and subsequently settle into stable states that encode the history of prior signals. This triadic dynamic operates iteratively, such that each stable state becomes the basis for responding to subsequent signals, establishing circular causality across time.</p>
<p>This conceptualization draws from foundational work in systems biology and developmental biology. <xref ref-type="bibr" rid="ref27">McNamara et al. (2024)</xref> demonstrated how morphogen signals drive cellular plasticity and establish stable spatial patterns during gastruloid development, illustrating the &#x201C;Signal &#x2192; Plasticity &#x2192; Stable State&#x201D; dynamic at the level of tissue patterning. Similarly, in neuroscience, <xref ref-type="bibr" rid="ref1">Barrett and Simmons (2015)</xref> proposed that the brain operates as a predictive organ that continuously generates interoceptive predictions (signals), updates internal models through plasticity, and settles into stable allostatic states that shape subsequent predictions. The model extends this logic across multiple hierarchical levels, from molecular to systemic.</p>
<p>Critically, this principle operates circularly: stable states feedback to modulate how subsequent signals are detected and processed. When this circular dynamic functions adaptively, the organism maintains allostatic integrity&#x2014;the capacity to flexibly respond to challenges while preserving core functional stability. When disruptions occur at any level, maladaptive stable states can become entrenched, propagating dysfunction across the system and eventually manifesting as dementia.</p>
<p>Boundaries of the principle: The principle &#x201C;Signal &#x2192; Plasticity &#x2192; Stable State&#x201D; does NOT apply to: (a) monogenic disorders with fully penetrant causal mutations (e.g., autosomal dominant AD due to APP, PSEN1, PSEN2), where a single-level disruption is sufficient to produce disease; (b) acute cognitive impairment due to isolated structural brain lesions without systemic propagation (e.g., single territorial stroke); or (c) normal developmental milestones that do not involve maladaptive consolidation of pathological stable states. The model is explicitly proposed for sporadic, late-onset, multifactorial dementias.</p>
</sec>
<sec id="sec4">
<label>2.2</label>
<title>Level 1: morphogenetic programming and genetic architecture</title>
<p>The first level encompasses the genetic blueprint that establishes baseline vulnerabilities and resilience capacities across the lifespan. This includes constitutive genetic variants&#x2014;such as APOE &#x03B5;2/&#x03B5;3/&#x03B5;4 alleles&#x2014;that modulate lipid metabolism, neuroinflammation, and synaptic integrity (<xref ref-type="bibr" rid="ref41">Verduzco Espinoza et al., 2025</xref>). However, genetic architecture is not merely deterministic; it operates through dynamic interactions with environmental signals, a principle encapsulated in the concept of differential susceptibility (<xref ref-type="bibr" rid="ref2">Belsky and Pluess, 2009</xref>).</p>
<p>Morphogenetic programming refers to the processes by which genetic information guides tissue organization, neural circuit formation, and organ system development during critical windows. Early-life signals&#x2014;including nutritional status, stress exposure, and maternal care&#x2014;can calibrate these developmental trajectories, establishing set points that persist into adulthood (<xref ref-type="bibr" rid="ref7">Danese and McEwen, 2012</xref>). Importantly, genetic variants influence the degree to which environmental signals induce plastic changes, creating gene&#x2013;environment interactions that shape individual differences in dementia risk.</p>
<p>Within the model, Level 1 provides the foundational architecture upon which higher-level processes operate. However, consistent with circular dynamics, information from other levels, particularly epigenetic modifications (Level 2) and allostatic load (Level 3), can feed back to influence gene expression patterns, effectively updating the functional consequences of the genetic blueprint across the lifespan.</p>
</sec>
<sec id="sec5">
<label>2.3</label>
<title>Level 2: epigenetic molecular memory</title>
<p>The second level captures the molecular mechanisms through which environmental signals become biologically inscribed and maintained over time. Epigenetic modifications&#x2014;including DNA methylation, histone acetylation, and non-coding RNA regulation&#x2014;constitute a form of molecular memory that encodes past experiences and shapes future cellular responses (<xref ref-type="bibr" rid="ref44">Zannas et al., 2016</xref>; <xref ref-type="bibr" rid="ref43">Yusupov et al., 2024</xref>).</p>
<p>Work on stress-responsive genes has been particularly illuminating. <xref ref-type="bibr" rid="ref20">Klengel et al. (2013)</xref> demonstrated that childhood trauma induces allele-specific DNA demethylation of FKBP5, a key regulator of glucocorticoid receptor sensitivity, leading to long-term alterations in HPA axis reactivity. <xref ref-type="bibr" rid="ref28">N&#x00F6;thling et al. (2025)</xref> extended these findings by showing that FKBP5 methylation patterns mediate the relationship between childhood trauma and stress-related psychopathology. Similarly, <xref ref-type="bibr" rid="ref29">Panitch et al. (2024)</xref> identified APOE genotype-specific methylation patterns linked to Alzheimer&#x2019;s disease pathology and estrogen response, illustrating how genetic architecture (Level 1) and epigenetic memory (Level 2) interact to shape disease risk.</p>
<p>Crucially, epigenetic marks are not static; they can be modified by ongoing experiences across the lifespan, including psychosocial stressors, metabolic states, and inflammatory signals (<xref ref-type="bibr" rid="ref10">Feng et al., 2025</xref>). Within the circular framework, epigenetic memory serves as a bridge between genetic predisposition and systemic physiological adaptation, translating environmental signals into lasting biological changes that influence vulnerability to dementia.</p>
</sec>
<sec id="sec6">
<label>2.4</label>
<title>Level 3: allostatic load and systemic physiological adaptation</title>
<p>The third level encompasses the cumulative physiological consequences of repeated or chronic adaptation to environmental demands. The concept of allostatic load, introduced by <xref ref-type="bibr" rid="ref26">McEwen and Stellar (1993)</xref>, operationalizes the &#x201C;wear and tear&#x201D; on the body resulting from sustained activation of stress-response systems, including the HPA axis, autonomic nervous system, and inflammatory pathways.</p>
<p>Allostatic load is typically measured through composite indices that integrate markers from multiple physiological systems&#x2014;cardiovascular (e.g., blood pressure, heart rate variability), metabolic (e.g., body mass index, glycated hemoglobin), and inflammatory (e.g., C-reactive protein, interleukin-6) (<xref ref-type="bibr" rid="ref13">Guidi et al., 2021</xref>). Longitudinal studies have demonstrated that elevated allostatic load is prospectively associated with all-cause dementia and Alzheimer&#x2019;s disease specifically (<xref ref-type="bibr" rid="ref31">Ruan et al., 2025</xref>; <xref ref-type="bibr" rid="ref36">Souza-Talarico et al., 2025</xref>). Importantly, allostatic load captures cumulative risk that may not be evident from any single biomarker alone, reflecting the integrated state of multiple physiological systems.</p>
<p>Within the model, Level 3 functions as a critical integrator of signals from higher (Levels 4&#x2013;5) and lower (Levels 1&#x2013;2) levels. Chronic psychological stress (Level 4) is associated with allostatic load accumulation, while epigenetic modifications (Level 2) influence how effectively individuals regulate stress responses. Conversely, elevated allostatic load may feedback to induce further epigenetic changes and exacerbate neuroinflammatory processes, establishing vicious cycles that may accelerate cognitive decline.</p>
</sec>
<sec id="sec7">
<label>2.5</label>
<title>Level 4: the psychological&#x2013;neurological&#x2013;endocrine&#x2013;immunological (PNEI) network</title>
<p>The fourth level captures the dynamic interactions among psychological processes, neurological function, endocrine signaling, and immune activity. This PNEI network embodies the bidirectional communication pathways that link mental states, brain function, peripheral physiology, and inflammatory responses (<xref ref-type="bibr" rid="ref8">Dantzer et al., 2008</xref>; <xref ref-type="bibr" rid="ref34">Slavich and Irwin, 2014</xref>).</p>
<p>Psychological stress activates the HPA axis and sympathetic nervous system, triggering release of cortisol and catecholamines that modulate immune cell trafficking and cytokine production (<xref ref-type="bibr" rid="ref10">Feng et al., 2025</xref>). Chronic stress promotes a pro-inflammatory phenotype characterized by elevated levels of interleukin-6, tumor necrosis factor-alpha, and C-reactive protein, which can compromise the blood&#x2013;brain barrier and promote neuroinflammation (<xref ref-type="bibr" rid="ref8">Dantzer et al., 2008</xref>; <xref ref-type="bibr" rid="ref39">Tyagi and Bartley, 2025</xref>). Conversely, peripheral inflammation signals the brain through multiple pathways&#x2014;including vagal afferents, circulating cytokines, and endothelial activation&#x2014;inducing sickness behaviors, cognitive deficits, and neuroinflammation.</p>
<p>Within dementia pathogenesis, PNEI dynamics are increasingly recognized as central rather than peripheral. <xref ref-type="bibr" rid="ref22">Li et al. (2025)</xref> demonstrated that inflammation-mediated regional brain alterations are associated with mild cognitive impairment in knee osteoarthritis, illustrating how peripheral inflammatory conditions can influence central pathology. Similarly, <xref ref-type="bibr" rid="ref41">Verduzco Espinoza et al. (2025)</xref> showed that microglia-to-neuron signaling links APOE4 and inflammation to enhanced neuronal lipid metabolism and network activity, providing a mechanistic bridge between genetic risk, neuroimmune dynamics, and neural circuit function.</p>
<p>The PNEI network functions as a hub within the circular model, integrating signals from psychological states (e.g., stress, mood), neurological activity (e.g., network oscillations), endocrine status (e.g., cortisol rhythms), and immune function (e.g., cytokine profiles). Perturbations at any point in this network may propagate bidirectionally, potentially amplifying maladaptive states that increase dementia vulnerability.</p>
</sec>
<sec id="sec8">
<label>2.6</label>
<title>Level 5: interoceptive&#x2013;neuronal integration</title>
<p>The fifth level concerns the integration of interoceptive signals&#x2014;sensory information about the internal state of the body&#x2014;with neuronal dynamics across large-scale brain networks. Interoception, defined as the sense of the physiological condition of the body (<xref ref-type="bibr" rid="ref4">Craig, 2002</xref>), provides the foundation for emotional experience, self-awareness, and adaptive regulation of internal states (<xref ref-type="bibr" rid="ref1">Barrett and Simmons, 2015</xref>; <xref ref-type="bibr" rid="ref18">Khalsa et al., 2018</xref>).</p>
<p>Recent advances have revealed that interoceptive processing relies on intrinsic neural timescales&#x2014;the temporal windows over which neurons and neural populations integrate information. <xref ref-type="bibr" rid="ref14">Hazelton et al. (2025a</xref>, <xref ref-type="bibr" rid="ref15">b)</xref> demonstrated that patients with behavioral-variant frontotemporal dementia exhibit altered intrinsic neural timescales of interoceptive processing, with disruptions in allostatic&#x2013;interoceptive networks that support adaptive regulation. Critically, these alterations were associated with clinical severity and predicted disease progression, suggesting that interoceptive dysfunction is not merely a correlate but potentially a driver of neurodegeneration.</p>
<p><xref ref-type="bibr" rid="ref14">Hazelton et al. (2025a</xref>, <xref ref-type="bibr" rid="ref15">b)</xref> further showed, through neuroimaging meta-analyses, that interoception, emotion, and social cognition converge across neurodegenerative diseases, including Alzheimer&#x2019;s disease. This convergence suggests that disturbances in interoceptive&#x2013;neuronal integration may constitute a transdiagnostic mechanism of cognitive and behavioral decline. Within the model, interoceptive&#x2013;neuronal integration represents the highest-order integrative level within the circular flow of information, where signals from genetic, epigenetic, physiological, and PNEI levels converge to shape conscious experience, self-regulation, and adaptive behavior.</p>
<p>When interoceptive&#x2013;neuronal integration breaks down, the organism loses its capacity to accurately sense, predict, and regulate internal states, leading to maladaptive stable states that manifest clinically as cognitive impairment, behavioral dysregulation, and ultimately dementia (<xref ref-type="fig" rid="fig1">Figure 1</xref>).</p>
<fig position="float" id="fig1">
<label>Figure 1</label>
<caption>
<p>Hierarchical&#x2013;circular model of biological memory in Alzheimer&#x2019;s disease and related dementias. The model depicts five interconnected levels of biological memory&#x2014;(1) morphogenetic/genetic architecture, (2) epigenetic molecular memory, (3) allostatic load and systemic adaptation, (4) the Psycho&#x2013;Neuro&#x2013;Endocrine&#x2013;Immunological (PNEI) network, and (5) interoceptive&#x2013;neuronal integration&#x2014;organized around the unifying principle &#x201C;Signal &#x2192; Plasticity &#x2192; Stable State.&#x201D; Bidirectional arrows indicate circular information flow, whereby perturbations at any level can propagate through the system and consolidate maladaptive pathological stable states. Interoceptive&#x2013;neuronal integration represents the highest-order integrative level within this circular flow. Low allostatic integrity increases the likelihood of progression to dementia.</p>
</caption>
<graphic xlink:href="frdem-05-1841647-g001.tif" mimetype="image" mime-subtype="tiff">
<alt-text content-type="machine-generated">Hierarchical&#x2013;circular diagram illustrating five levels of biological memory in Alzheimer&#x2019;s disease and dementias: Level 1&#x2014;signal, plasticity, stable state; Level 2&#x2014;morphogenetic/genetic architecture; Level 3&#x2014;allostatic load/systemic adaptation; Level 4&#x2014;PNEI network; Level 5&#x2014;interoceptive&#x2013;neuronal integration. Allostatic integrity bar ranges from high to low, with low integrity increasing dementia likelihood.</alt-text>
</graphic>
</fig>
</sec>
<sec id="sec9">
<label>2.7</label>
<title>Circular dynamics and bidirectional propagation across levels</title>
<p>A defining feature of the Hierarchical-Circular Model is the emphasis on bidirectional propagation of signals across levels. Unlike linear models that posit unidirectional causality (e.g., amyloid accumulation &#x2192; neurodegeneration &#x2192; cognitive decline), the circular model recognizes that perturbations at any level can propagate upward, downward, and laterally through the system.</p>
<p>For example:</p>
<list list-type="bullet">
<list-item>
<p>A genetic variant (Level 1) may influence epigenetic susceptibility (Level 2), which is associated with allostatic load accumulation (Level 3), affecting PNEI dynamics (Level 4) and interoceptive processing (Level 5). This represents upward propagation.</p>
</list-item>
<list-item>
<p>Conversely, chronic psychological stress (Level 4) is associated with allostatic load (Level 3), which may correlate with epigenetic modifications (Level 2) that alter gene expression (Level 1). This represents downward propagation.</p>
</list-item>
<list-item>
<p>Neuroimmune activation (Level 4) may directly influence interoceptive network function (Level 5) while simultaneously modulating allostatic load (Level 3), representing lateral propagation.</p>
</list-item>
</list>
<p>This circular architecture may explain why individuals with similar neuropathological burdens (e.g., amyloid and tau positivity) may follow markedly different clinical trajectories. When the system maintains allostatic integrity&#x2014;the capacity to flexibly adapt to challenges while preserving circular information flow&#x2014;neuropathology may be tolerated without clinical expression. When allostatic integrity breaks down, even modest pathology could trigger cascading dysfunction across levels, culminating in dementia.</p>
<p>The concept of allostatic integrity thus represents a dynamic systems-level property that cannot be reduced to any single biomarker or pathway. It is precisely this property that the Hierarchical-Circular Model seeks to operationalize and investigate.</p>
</sec>
<sec id="sec10">
<label>2.8</label>
<title>Model boundaries and falsification criteria</title>
<p>What the model does NOT explain:</p>
<list list-type="bullet">
<list-item>
<p>Rare monogenic forms of dementia with fully penetrant causal mutations (e.g., APP, PSEN1, PSEN2)</p>
</list-item>
<list-item>
<p>Acute cognitive impairment due to isolated structural brain lesions (e.g., single hemispheric stroke)</p>
</list-item>
<list-item>
<p>Normal age-related cognitive changes without evidence of pathological circular reinforcement</p>
</list-item>
</list>
<p>Falsification criteria: The Hierarchical-Circular Model would be falsified if:</p>
<list list-type="alpha-lower">
<list-item>
<p>Composite indices of allostatic integrity show no incremental predictive value for conversion from mild cognitive impairment to dementia beyond amyloid and tau status alone, in adequately powered longitudinal cohorts;</p>
</list-item>
<list-item>
<p>Multidomain interventions targeting multiple levels show only additive effects (i.e., no statistical interaction, ANOVA interaction term <italic>p</italic> &#x003E;&#x202F;0.05) in properly designed randomized controlled trials;</p>
</list-item>
<list-item>
<p>Individuals with preserved allostatic integrity universally progress to dementia regardless of their systemic physiological state, contradicting the moderation hypothesis.</p>
</list-item>
</list>
<p>These criteria ensure the model is empirically testable and not merely descriptive.</p>
</sec>
<sec id="sec11">
<label>2.9</label>
<title>Relationship to reserve and resilience frameworks</title>
<p>The Hierarchical-Circular Model does not replace but rather integrates and extends the established constructs of cognitive reserve, brain reserve, and brain maintenance (<xref ref-type="bibr" rid="ref38">Stern et al., 2020</xref>; <xref ref-type="bibr" rid="ref37">Stern et al., 2023</xref>) (<xref ref-type="table" rid="tab1">Table 1</xref>).</p>
<table-wrap position="float" id="tab1">
<label>Table 1</label>
<caption>
<p>Comparison of the construct of allostatic integrity with established reserve and resilience frameworks (cognitive reserve, brain reserve, and brain maintenance).</p>
</caption>
<table frame="hsides" rules="groups">
<thead>
<tr>
<th align="left" valign="top">Construct</th>
<th align="left" valign="top">Definition</th>
<th align="left" valign="top">How allostatic integrity differs/extends</th>
</tr>
</thead>
<tbody>
<tr>
<td align="left" valign="middle">Cognitive reserve</td>
<td align="left" valign="middle">Ability to maintain cognitive function despite pathology through flexible neural network use</td>
<td align="left" valign="middle">Allostatic integrity is multisystem (not just neural); includes peripheral physiology, epigenetics, and interoception</td>
</tr>
<tr>
<td align="left" valign="middle">Brain reserve</td>
<td align="left" valign="middle">Structural brain features (e.g., head circumference, synapse density) that buffer against pathology</td>
<td align="left" valign="middle">Allostatic integrity is dynamic (not static); changes across lifespan via circular feedback</td>
</tr>
<tr>
<td align="left" valign="middle">Brain maintenance</td>
<td align="left" valign="middle">Absence of pathological change despite risk factors</td>
<td align="left" valign="middle">Allostatic integrity explicitly models positive adaptation (not just absence of pathology)</td>
</tr>
</tbody>
</table>
</table-wrap>
<p>Integration: The model proposes that allostatic integrity modulates reserve expression. High allostatic integrity may enhance cognitive reserve through preserved interoceptive-prefrontal connectivity; low allostatic integrity may undermine reserve even when structural brain reserve is high. This integration generates discriminative predictions (see Section 4.4).</p>
</sec>
</sec>
<sec id="sec12">
<label>3</label>
<title>Development of the hypothesis: synthesis of evidence</title>
<p>Having established the conceptual architecture of the Hierarchical-Circular Model, this section synthesizes the empirical evidence that supports each level and the circular dynamics linking them. The evidence is drawn from longitudinal cohort studies, mechanistic investigations, and multisystem approaches spanning the 2010&#x2013;2025 period, organized thematically around the model&#x2019;s core components.</p>
<sec id="sec13">
<label>3.1</label>
<title>Gene&#x2013;environment interactions and lifespan vulnerability</title>
<p>The foundation of the model rests on the premise that genetic architecture does not operate in isolation but interacts dynamically with environmental exposures across the lifespan. This gene&#x2013;environment interplay is perhaps best exemplified by research on APOE and stress-related genes.</p>
<p>The APOE &#x03B5;4 allele remains the strongest genetic risk factor for sporadic Alzheimer&#x2019;s disease, yet its penetrance is far from complete. <xref ref-type="bibr" rid="ref41">Verduzco Espinoza et al. (2025)</xref> demonstrated that APOE4 influences neuronal lipid metabolism and network activity through microglia-to-neuron signaling, revealing a mechanistic pathway that is modulated by inflammatory states. This finding suggests that the impact of APOE4 may depend on the inflammatory environment&#x2014;an environmental factor that is itself shaped by stress exposure, diet, and other lifestyle variables.</p>
<p>Similarly, research on FKBP5 has elucidated how genetic variants and environmental stressors interact to calibrate stress reactivity. <xref ref-type="bibr" rid="ref44">Zannas et al. (2016)</xref> showed that FKBP5 variants influence epigenetic regulation of the glucocorticoid receptor, with effects that are potentiated by early-life trauma. <xref ref-type="bibr" rid="ref20">Klengel et al. (2013)</xref> demonstrated that childhood trauma induces allele-specific DNA demethylation of FKBP5, leading to long-term alterations in HPA axis reactivity. <xref ref-type="bibr" rid="ref28">N&#x00F6;thling et al. (2025)</xref> extended these findings by showing that FKBP5 methylation patterns mediate the relationship between childhood trauma and stress-related psychopathology in adulthood.</p>
<p>Longitudinal cohort studies have further established that early-life experiences are associated with dementia risk decades later. <xref ref-type="bibr" rid="ref24">Livingston et al. (2020)</xref> and <xref ref-type="bibr" rid="ref11">Fratiglioni et al. (2020)</xref> synthesized evidence showing that adverse childhood experiences, educational attainment, and socioeconomic status across the life course collectively contribute to dementia vulnerability. <xref ref-type="bibr" rid="ref21">Lee et al. (2025)</xref> provided population-based evidence linking adverse experiences throughout the life course to increased dementia risk, reinforcing the notion that vulnerability is not determined at birth but emerges from cumulative gene&#x2013;environment interactions.</p>
<p>Within the circular model, these findings illustrate upward propagation: genetic architecture (Level 1) shapes how environmental signals (e.g., childhood trauma) become biologically embedded through epigenetic mechanisms (Level 2), which in turn are associated with systemic physiology (Level 3) and psychological function (Level 4). Conversely, the model also predicts downward propagation: chronic stress (Level 4) may induce epigenetic modifications (Level 2) that alter gene expression patterns (Level 1), creating bidirectional associations.</p>
<sec id="sec14">
<label>3.1.1</label>
<title>Illustrative mechanistic pathways of circular propagation</title>
<p>To move beyond abstract circularity, the model specifies concrete, testable mechanisms:</p>
<list list-type="simple">
<list-item>
<p>Pathway A: Peripheral inflammation &#x2192; epigenetic modification &#x2192; HPA dysregulation (bidirectional). Peripheral IL-6 crosses the blood&#x2013;brain barrier &#x2192; activates NF-&#x03BA;B in microglia &#x2192; induces DNMT1-mediated hypermethylation of the NR3C1 (glucocorticoid receptor) promoter &#x2192; reduces cortisol sensitivity &#x2192; increases allostatic load &#x2192; perpetuates inflammation. This specific pathway has been demonstrated in human cohorts (<xref ref-type="bibr" rid="ref40">Tyrka et al., 2016</xref>; <xref ref-type="bibr" rid="ref9">Davies et al., 2019</xref>).</p>
</list-item>
<list-item>
<p>Pathway B: Interoceptive dysfunction &#x2192; chronic stress &#x2192; epigenetic embedding (Levels 5&#x202F;&#x2192;&#x202F;4&#x202F;&#x2192;&#x202F;2). Altered anterior insula prediction signals &#x2192; misinterpretation of bodily states &#x2192; sustained sympathetic activation &#x2192; elevated cortisol &#x2192; induces FKBP5 demethylation at intron 7&#x202F;&#x2192;&#x202F;long-term HPA axis dysregulation. This pathway is supported by <xref ref-type="bibr" rid="ref14">Hazelton et al. (2025a</xref>, <xref ref-type="bibr" rid="ref15">b)</xref> and <xref ref-type="bibr" rid="ref20">Klengel et al. (2013)</xref>.</p>
</list-item>
<list-item>
<p>Pathway C: Allostatic load &#x2192; neuroinflammation &#x2192; interoceptive network disruption (Levels 3&#x202F;&#x2192;&#x202F;4&#x202F;&#x2192;&#x202F;5). Elevated allostatic load promotes peripheral cytokine production &#x2192; activates brain endothelial cells &#x2192; upregulates adhesion molecules &#x2192; facilitates monocyte trafficking into the salience network (including anterior insula and anterior cingulate)&#x202F;&#x2192;&#x202F;impairs interoceptive prediction accuracy &#x2192; further increases allostatic load through poor self-regulation.</p>
</list-item>
</list>
<p>These pathways provide empirically testable causal links rather than abstract circularity.</p>
</sec>
</sec>
<sec id="sec15">
<label>3.2</label>
<title>Epigenetic embedding of stress and dementia risk</title>
<p>The second level of the model posits that epigenetic modifications constitute a form of molecular memory that encodes environmental exposures and shapes long-term vulnerability. Converging evidence from human and animal studies supports this proposition.</p>
<p><xref ref-type="bibr" rid="ref43">Yusupov et al. (2024)</xref> investigated DNA methylation patterns of FKBP5 regulatory regions in both brain and blood of humanized mice and humans, demonstrating concordance between central and peripheral tissues. This finding is critical because it suggests that peripheral epigenetic markers may serve as accessible proxies for brain epigenetic states, enabling large-scale human studies of epigenetic contributions to dementia risk.</p>
<p><xref ref-type="bibr" rid="ref29">Panitch et al. (2024)</xref> identified APOE genotype-specific methylation patterns linked to Alzheimer&#x2019;s disease pathology and estrogen response, illustrating how epigenetic marks interact with genetic architecture. Their findings revealed that methylation at specific CpG sites modified the relationship between APOE genotype and neuropathological outcomes, suggesting that epigenetic variation may explain some of the heterogeneity in clinical expression among APOE &#x03B5;4 carriers.</p>
<p>The concept of epigenetic embedding extends beyond single genes to encompass broader chromatin remodeling and non-coding RNA regulation. <xref ref-type="bibr" rid="ref10">Feng et al. (2025)</xref> reviewed evidence showing that psychological stress induces widespread epigenetic changes that affect neuroinflammatory pathways, HPA axis function, and neural plasticity. These changes can persist for decades, effectively &#x201C;programming&#x201D; biological systems for either resilience or vulnerability.</p>
<p>Within the circular model, epigenetic memory serves as a bridge between genetic predisposition (Level 1) and systemic physiological adaptation (Level 3). Critically, epigenetic marks are not static; they can be modified by ongoing experiences, including psychosocial interventions, metabolic changes, and pharmacological agents (<xref ref-type="bibr" rid="ref10">Feng et al., 2025</xref>). This plasticity opens possibilities for interventions that could reverse or mitigate maladaptive epigenetic states.</p>
</sec>
<sec id="sec16">
<label>3.3</label>
<title>Allostatic load as a multisystem risk integrator</title>
<p>The third level of the model identifies allostatic load as a critical integrator of multisystem risk. Allostatic load operationalizes the cumulative physiological consequences of chronic stress across multiple regulatory systems, providing a composite measure that may outperform single biomarkers in predicting health outcomes.</p>
<p><xref ref-type="bibr" rid="ref36">Souza-Talarico et al. (2025)</xref> examined the relationship between allostatic load dynamics, Alzheimer&#x2019;s disease biomarkers, and progression in individuals with mild cognitive impairment using data from the Alzheimer&#x2019;s disease Neuroimaging Initiative. Their findings revealed that allostatic load trajectories were associated with conversion from mild cognitive impairment to dementia independently of amyloid and tau status, supporting the hypothesis that systemic physiological state modulates clinical expression of neuropathology.</p>
<p><xref ref-type="bibr" rid="ref31">Ruan et al. (2025)</xref> conducted a prospective cohort study demonstrating that elevated allostatic load was associated with increased risk for all-cause dementia and specific subtypes, including Alzheimer&#x2019;s disease and vascular dementia. <xref ref-type="bibr" rid="ref30">P&#x00E9;rez-Carrasco et al. (2025)</xref> extended these findings to a Chilean older-adults cohort, revealing distinct physiological aging trajectories that were associated with cognitive outcomes, suggesting that allostatic load captures culturally and contextually relevant aspects of physiological aging.</p>
<p><xref ref-type="bibr" rid="ref35">Smith et al. (2025)</xref> showed that allostatic load, as a measure of cumulative physiological stress, is associated with impaired cognitive resilience in older adults. Importantly, they demonstrated that allostatic load effects were independent of amyloid burden, suggesting that systemic physiology may contribute to cognitive decline through pathways distinct from classical neuropathological cascades.</p>
<p>The concept of allostatic load is grounded in foundational work by <xref ref-type="bibr" rid="ref26">McEwen and Stellar (1993)</xref>, who introduced the framework of allostasis&#x2014;the process of achieving stability through physiological change. <xref ref-type="bibr" rid="ref13">Guidi et al. (2021)</xref> provided a systematic review of allostatic load and its health impacts, confirming that composite indices consistently predict morbidity and mortality across diverse populations.</p>
<p>Within the circular model, allostatic load functions as an integrator of signals from multiple levels. Elevated allostatic load (Level 3) reflects cumulative input from genetic predisposition (Level 1), epigenetic regulation (Level 2), psychological stress (Level 4), and interoceptive dysfunction (Level 5). Conversely, elevated allostatic load may feedback to induce further epigenetic changes (Level 2) and exacerbate neuroinflammation (Level 4), establishing vicious cycles that may accelerate decline.</p>
</sec>
<sec id="sec17">
<label>3.4</label>
<title>Neuroimmune dynamics and central&#x2013;peripheral crosstalk</title>
<p>The fourth level of the model emphasizes the bidirectional communication between the central nervous system and peripheral immune processes. This neuroimmune axis has emerged as a critical determinant of brain health and dementia risk.</p>
<p><xref ref-type="bibr" rid="ref8">Dantzer et al. (2008)</xref> established the foundational framework for understanding how peripheral inflammation signals the brain, inducing sickness behaviors and cognitive deficits. Their work demonstrated that pro-inflammatory cytokines communicate with the brain through multiple pathways&#x2014;including vagal afferents, circulating cytokine transport, and endothelial activation&#x2014;triggering neuroinflammation that can persist long after the peripheral insult resolves.</p>
<p><xref ref-type="bibr" rid="ref34">Slavich and Irwin (2014)</xref> advanced this framework through the Social Signal Transduction Theory of Depression, which delineates how psychosocial stressors activate inflammatory responses that contribute to neuropsychiatric disorders. Their model provides a mechanistic bridge between psychological states (Level 4), immune function (Level 4), and brain health.</p>
<p><xref ref-type="bibr" rid="ref10">Feng et al. (2025)</xref> synthesized recent evidence on central&#x2013;peripheral neuroimmune dynamics in psychological stress and depression, revealing bidirectional pathways through which stress alters immune function and immune signals modulate neural activity. They highlighted the role of the blood&#x2013;brain barrier as a dynamic interface that regulates neuroimmune communication, with implications for dementia pathogenesis.</p>
<p><xref ref-type="bibr" rid="ref39">Tyagi and Bartley (2025)</xref> conducted a systematic review of dynamic cytokine relationships across the blood&#x2013;brain barrier in humans and nonhuman primates, revealing complex patterns of central&#x2013;peripheral coupling that vary across inflammatory states. Their findings suggest that the relationship between peripheral and central inflammation is not static but may depend on factors including chronicity, intensity, and individual genetic background.</p>
<p><xref ref-type="bibr" rid="ref22">Li et al. (2025)</xref> provided clinical evidence linking peripheral inflammation to brain alterations, demonstrating that inflammation-mediated regional brain changes were associated with mild cognitive impairment in knee osteoarthritis. This study illustrates how peripheral inflammatory conditions&#x2014;in this case, osteoarthritis&#x2014;can influence central pathology and cognitive function, supporting the model&#x2019;s emphasis on central&#x2013;peripheral crosstalk.</p>
<sec id="sec18">
<label>3.4.1</label>
<title>Peripheral&#x2013;central circular dynamics: evidence from systemic inflammation</title>
<p>Recent reviews highlight that peripheral inflammatory insults&#x2014;including infections, metabolic syndrome, and gut dysbiosis&#x2014;may contribute to Alzheimer&#x2019;s pathogenesis through bidirectional brain&#x2013;body signaling (<xref ref-type="bibr" rid="ref3">Chen et al., 2025</xref>; <xref ref-type="bibr" rid="ref23">Liu et al., 2023</xref>). These studies demonstrate that: (a) peripheral inflammation can precede central pathology by years; (b) systemic insults may amplify neuroinflammation via blood&#x2013;brain barrier disruption and vagal afferent pathways; and (c) central inflammation may feedback to exacerbate peripheral immune dysregulation. This evidence directly supports the circular model&#x2019;s claim that perturbations at Level 4 (PNEI) can propagate across levels: peripheral inflammation &#x2192; neuroinflammation &#x2192; allostatic load &#x2192; epigenetic modification &#x2192; further peripheral dysregulation.</p>
<p>Within the circular model, the PNEI network (Level 4) functions as a hub integrating signals from psychological, neurological, endocrine, and immune domains. Perturbations at any point may propagate through the network, with neuroimmune activation feeding back to allostatic load (Level 3) and interoceptive processing (Level 5).</p>
</sec>
</sec>
<sec id="sec19">
<label>3.5</label>
<title>Interoceptive dysfunction as a transdiagnostic mechanism</title>
<p>The fifth level of the model concerns interoceptive&#x2013;neuronal integration, positing that disruptions in the brain&#x2019;s capacity to sense, predict, and regulate bodily states may constitute a transdiagnostic mechanism of cognitive and behavioral decline.</p>
<p><xref ref-type="bibr" rid="ref1">Barrett and Simmons (2015)</xref> proposed a predictive coding framework for interoception, arguing that the brain continuously generates predictions about bodily states, updates these predictions through interoceptive afferents, and uses prediction errors to refine internal models. This framework positions interoception as fundamental to emotion, self-awareness, and adaptive regulation.</p>
<p><xref ref-type="bibr" rid="ref18">Khalsa et al. (2018)</xref> provided a comprehensive roadmap for interoception research in mental health, outlining measurement approaches, neural substrates, and clinical applications. Their review highlighted the relevance of interoception to conditions ranging from anxiety and depression to neurodegenerative diseases, supporting the transdiagnostic perspective.</p>
<p><xref ref-type="bibr" rid="ref5">Critchley and Garfinkel (2017)</xref> emphasized the role of interoceptive processing in allostatic regulation, linking interoceptive accuracy to emotion regulation, decision-making, and social cognition. Their work provides a foundation for understanding how interoceptive dysfunction may contribute to the behavioral and cognitive features of dementia.</p>
<p><xref ref-type="bibr" rid="ref14">Hazelton et al. (2025a</xref>, <xref ref-type="bibr" rid="ref15">b)</xref> provided direct evidence linking interoceptive dysfunction to neurodegeneration in behavioral-variant frontotemporal dementia. Using advanced neuroimaging techniques, they demonstrated altered intrinsic neural timescales of interoceptive processing, with disruptions in allostatic&#x2013;interoceptive networks that support adaptive regulation. Critically, these alterations were associated with clinical severity and predicted disease progression, suggesting that interoceptive dysfunction is not merely a correlate but potentially a driver of neurodegeneration.</p>
<p><xref ref-type="bibr" rid="ref14">Hazelton et al. (2025a</xref>, <xref ref-type="bibr" rid="ref15">b)</xref> extended these findings through neuroimaging meta-analyses revealing convergence of interoception, emotion, and social cognition across neurodegenerative diseases, including Alzheimer&#x2019;s disease. Their results suggest that disturbances in interoceptive&#x2013;neuronal integration may constitute a common mechanism underlying diverse neurodegenerative conditions, supporting the model&#x2019;s claim that interoceptive dysfunction is transdiagnostic.</p>
<p>Within the circular model, interoceptive&#x2013;neuronal integration (Level 5) represents the highest-order integrative level within the circular flow of information. Signals from genetic (Level 1), epigenetic (Level 2), allostatic (Level 3), and PNEI (Level 4) levels converge in interoceptive networks that shape conscious experience, self-regulation, and adaptive behavior. When these networks are disrupted, the organism loses its capacity to accurately sense and regulate internal states, leading to maladaptive stable states that manifest clinically as dementia.</p>
</sec>
<sec id="sec20">
<label>3.6</label>
<title>Synthesis: toward a unifying framework</title>
<p>Taken together, the evidence synthesized across Levels 1&#x2013;5 reveals a convergent picture: dementia risk and progression are shaped by interactions among genetic architecture, epigenetic molecular memory, cumulative allostatic load, neuroimmune dynamics, and interoceptive&#x2013;neuronal integration. Each of these domains has been investigated independently, but the circular model provides a framework for understanding how they may interact bidirectionally across the lifespan.</p>
<p>The concept of allostatic integrity&#x2014;the capacity of the organism to maintain adaptive circular information flow across multiple levels of organization&#x2014;emerges as a dynamic systems-level property that may explain why individuals with similar neuropathological burdens follow divergent clinical trajectories. When allostatic integrity is preserved, even substantial pathology may be tolerated without clinical expression. When allostatic integrity is compromised, even modest pathology could trigger cascading dysfunction across levels, culminating in dementia.</p>
<p>This synthesis supports the model&#x2019;s central hypothesis: dementias are best understood not as isolated brain pathologies but as disorders of biological memory and allostatic integrity. The Hierarchical-Circular Model provides a conceptual roadmap for investigating this hypothesis through testable predictions, which are developed in the following section.</p>
</sec>
</sec>
<sec id="sec21">
<label>4</label>
<title>Testable predictions and empirical consequences</title>
<p>A central requirement of any scientific hypothesis is that it generates concrete, falsifiable predictions that can be tested through empirical investigation. The Hierarchical-Circular Model yields four core predictions that differentiate it from linear, single-pathway models of dementia pathogenesis. Each prediction is grounded in the model&#x2019;s conceptual architecture and can be evaluated using existing longitudinal cohorts, biomarker panels, and intervention trial designs.</p>
<sec id="sec22">
<label>4.1</label>
<title>Prediction 1: composite indices of allostatic integrity will outperform single biomarkers in predicting conversion from mild cognitive impairment to dementia</title>
<p>If dementia emerges from disruptions in circular information flow across multiple levels of biological organization, then a composite measure capturing this integrated state should predict clinical outcomes more accurately than any single biomarker alone. This prediction directly challenges the prevailing assumption that a single molecular entity&#x2014;such as amyloid-beta or tau&#x2014;can serve as a sufficient predictive marker for clinical progression.</p>
<sec id="sec23">
<label>4.1.1</label>
<title>What makes this prediction unique to the circular model</title>
<p>The model predicts that dyssynchrony profiles (e.g., high inflammation with low cortisol, indicating feedback loop failure) will have the strongest predictive power&#x2014;not merely the sum of individual markers. Linear composites (e.g., additive allostatic load indices) are insufficient; dynamic synchrony across levels is what captures circular flow.</p>
</sec>
<sec id="sec24">
<label>4.1.2</label>
<title>Empirical support</title>
<p>Preliminary evidence is consistent with this prediction. <xref ref-type="bibr" rid="ref36">Souza-Talarico et al. (2025)</xref> demonstrated that allostatic load trajectories were associated with conversion from mild cognitive impairment to dementia independently of amyloid and tau status. <xref ref-type="bibr" rid="ref31">Ruan et al. (2025)</xref> and <xref ref-type="bibr" rid="ref35">Smith et al. (2025)</xref> similarly showed that allostatic load was associated with dementia risk beyond traditional biomarkers. However, these studies used allostatic load indices that primarily capture Level 3 (systemic physiological adaptation). The model predicts that composite indices incorporating markers from Levels 2 (epigenetic), 4 (neuroimmune), and 5 (interoceptive) will demonstrate even stronger predictive validity.</p>
</sec>
<sec id="sec25">
<label>4.1.3</label>
<title>Testable approach</title>
<p>Using existing longitudinal cohorts such as the Alzheimer&#x2019;s Disease Neuroimaging Initiative (ADNI), the Religious Orders Study (ROS), or the Rush Memory and Aging Project (MAP), researchers can construct composite indices of allostatic integrity that integrate: Level 2 (DNA methylation patterns at stress-related loci such as FKBP5, NR3C1), Level 3 (multi-system allostatic load indices), Level 4 (inflammatory cytokine profiles, HPA axis measures, and autonomic function), and Level 5 (interoceptive accuracy measures and functional connectivity within salience and interoceptive networks). The model predicts that such composite indices will show higher area under the curve (AUC) values in predicting conversion from mild cognitive impairment to dementia compared to amyloid PET status, CSF p-tau181, or any single biomarker alone. Moreover, the model predicts that allostatic integrity indices will be associated with conversion even among individuals who are amyloid-positive, explaining variance in clinical trajectories that current frameworks cannot account for.</p>
</sec>
</sec>
<sec id="sec26">
<label>4.2</label>
<title>Prediction 2: multidomain interventions targeting more than one system level will have multiplicative, rather than additive, effects on slowing cognitive decline</title>
<p>If the system is characterized by circular, bidirectional interactions across levels, then interventions that target multiple levels simultaneously should produce synergistic&#x2014;not merely additive&#x2014;effects. This prediction has profound implications for clinical trial design and prevention strategies.</p>
<sec id="sec27">
<label>4.2.1</label>
<title>What makes this prediction unique</title>
<p>While multidomain interventions exist (e.g., FINGER trial), the circular model uniquely predicts multiplicative effects (statistical interaction, not just additivity). This is falsifiable: if the effect of combined intervention equals the sum of single interventions (ANOVA interaction term <italic>p</italic> &#x003E;&#x202F;0.05), the model is contradicted. Current literature shows additive effects; the model predicts synergy.</p>
</sec>
<sec id="sec28">
<label>4.2.2</label>
<title>Empirical support</title>
<p>The FINGER trial (<xref ref-type="bibr" rid="ref19">Kivipelto et al., 2017</xref>) provided early evidence that multidomain interventions (diet, exercise, cognitive training, vascular risk management) can slow cognitive decline in at-risk older adults. Subsequent trials have extended this approach, though few have systematically tested multiplicative versus additive effects. The model predicts that the magnitude of benefit from multidomain interventions reflects not merely the sum of individual component effects but their synergistic interactions across levels.</p>
</sec>
<sec id="sec29">
<label>4.2.3</label>
<title>Testable approach</title>
<p>Randomized controlled trials can be designed to test this prediction by comparing: Group A (single-domain intervention, e.g., anti-amyloid therapy alone); Group B (single-domain intervention targeting a different level, e.g., psychosocial stress reduction alone); Group C (combined intervention targeting both levels); and Group D (combined intervention targeting three or more levels, e.g., anti-amyloid + stress reduction + metabolic optimization + cognitive training). The model predicts that Group C will show effects greater than the sum of Groups A and B (multiplicative interaction), and Group D will show progressively larger multiplicative effects. These effects can be quantified using interaction terms in linear mixed models or through causal mediation analysis that examines whether combined interventions disrupt circular dynamics more effectively than single interventions. The model predicts that the multiplicative effects will be most pronounced among individuals with high allostatic load or compromised allostatic integrity at baseline, as these individuals may be most vulnerable to circular reinforcement of maladaptive states.</p>
</sec>
</sec>
<sec id="sec30">
<label>4.3</label>
<title>Prediction 3: patients with similar amyloid/tau profiles but contrasting allostatic integrity will show markedly different trajectories of clinical progression</title>
<p>Perhaps the most clinically significant prediction of the model is that allostatic integrity&#x2014;not neuropathological burden alone&#x2014;may determine clinical trajectory. This prediction directly addresses the long-standing paradox of dissociation between pathology and clinical expression.</p>
<sec id="sec31">
<label>4.3.1</label>
<title>What makes this prediction unique</title>
<p>The model uniquely predicts that allostatic integrity will moderate the relationship between neuropathological burden and cognitive decline. In linear models, pathology and outcome are assumed monotonic. The circular model predicts a statistical crossover interaction: high allostatic integrity flattens the pathology&#x2013;decline slope; low allostatic integrity steepens it. This pattern is not predicted by single-pathway models.</p>
</sec>
<sec id="sec32">
<label>4.3.2</label>
<title>Empirical support</title>
<p><xref ref-type="bibr" rid="ref32">Savva et al. (2009)</xref> first systematically documented the dissociation between neuropathological burden and clinical dementia in the Medical Research Council Cognitive Function and Ageing Study. More recent work by <xref ref-type="bibr" rid="ref36">Souza-Talarico et al. (2025)</xref> showed that allostatic load dynamics were associated with conversion independently of amyloid and tau, suggesting that systemic physiology may moderate the clinical impact of neuropathology. However, no study to date has directly tested whether allostatic integrity indices predict divergent trajectories among individuals matched for amyloid and tau status.</p>
</sec>
<sec id="sec33">
<label>4.3.3</label>
<title>Testable approach</title>
<p>Using longitudinal cohorts with both neuroimaging biomarkers and comprehensive multi-system data, researchers can identify subgroups of individuals who are: amyloid-positive/tau-positive with high allostatic integrity; amyloid-positive/tau-positive with low allostatic integrity; amyloid-negative/tau-negative with high allostatic integrity; and amyloid-negative/tau-negative with low allostatic integrity. The model predicts that among amyloid-positive/tau-positive individuals, those with high allostatic integrity will show slow or no cognitive decline over follow-up, whereas those with low allostatic integrity will show rapid progression to dementia. Conversely, among amyloid-negative/tau-negative individuals, those with low allostatic integrity may show cognitive decline despite absence of classical neuropathology, potentially representing non-Alzheimer&#x2019;s dementia phenotypes. This prediction can be tested using existing cohorts such as ADNI, ROSMAP, or the Harvard Aging Brain Study. The critical innovation is stratifying by both neuropathological status and allostatic integrity, rather than examining either in isolation.</p>
</sec>
</sec>
<sec id="sec34">
<label>4.4</label>
<title>Prediction 4 (discriminative): allostatic integrity moderates reserve expression</title>
<p>To differentiate the model from reserve frameworks (<xref ref-type="bibr" rid="ref38">Stern et al., 2020</xref>, <xref ref-type="bibr" rid="ref37">2023</xref>), the circular model generates a discriminative prediction:</p>
<p>Among individuals matched for neuropathological burden and cognitive reserve proxies (e.g., education, occupational complexity, intracranial volume), those with low allostatic integrity will show faster clinical progression than those with high allostatic integrity. Conversely, among individuals matched for allostatic integrity, the protective effect of cognitive reserve will be attenuated when allostatic integrity is low.</p>
<p>This crossover pattern is unique to the circular model. Reserve frameworks alone predict that high reserve protects regardless of peripheral state. The circular model predicts that allostatic integrity moderates reserve expression: systemic dysregulation (inflammation, HPA axis dysfunction, interoceptive impairment) can override neural-level reserve. This is testable by stratifying ADNI participants by both reserve proxies and allostatic integrity indices.</p>
</sec>
</sec>
<sec sec-type="discussion" id="sec35">
<label>5</label>
<title>Discussion</title>
<p>The Hierarchical-Circular Model of Biological Memory proposes a fundamental reconceptualization of Alzheimer&#x2019;s disease and related dementias. Rather than viewing dementias as linear consequences of aggregated proteins, the model frames them as emergent outcomes of disrupted circular information flow across five interconnected levels: genetic architecture, epigenetic molecular memory, allostatic load, PNEI network dynamics, and interoceptive&#x2013;neuronal integration. This discussion addresses the model&#x2019;s capacity to reconcile clinical paradoxes, its relationship to existing frameworks, its implications for research and clinical practice, its limitations, and directions for future investigations.</p>
<sec id="sec36">
<label>5.1</label>
<title>Relationship to reserve and resilience frameworks</title>
<p><xref ref-type="bibr" rid="ref38">Stern et al. (2020</xref>, <xref ref-type="bibr" rid="ref37">2023)</xref> provided the definitive framework for cognitive reserve, brain reserve, and brain maintenance in Alzheimer&#x2019;s disease. The Hierarchical-Circular Model complements rather than competes with this framework. While reserve models explain inter-individual differences in the relationship between pathology and cognition, they do not specify how peripheral physiology, epigenetic embedding, or interoceptive function contribute to reserve expression. The present model proposes that allostatic integrity is a determinant of reserve expression: individuals with high allostatic integrity may more effectively deploy cognitive reserve strategies; those with low allostatic integrity may have their reserve capacity undermined by chronic inflammation, HPA dysregulation, or interoceptive dysfunction. This integration generates Prediction 4 (Section 4.4).</p>
</sec>
<sec id="sec37">
<label>5.2</label>
<title>Operationalizing allostatic integrity: toward a composite index</title>
<p>A critical next step is developing validated composite indices of allostatic integrity. The model proposes the following provisional components (<xref ref-type="table" rid="tab2">Table 2</xref>):</p>
<table-wrap position="float" id="tab2">
<label>Table 2</label>
<caption>
<p>Proposed components of an allostatic integrity index.</p>
</caption>
<table frame="hsides" rules="groups">
<thead>
<tr>
<th align="left" valign="top">Domain</th>
<th align="left" valign="top">Proposed markers</th>
<th align="center" valign="top">Provisional weighting</th>
</tr>
</thead>
<tbody>
<tr>
<td align="left" valign="middle">Adaptive reserve</td>
<td align="left" valign="middle">HRV (high frequency), cortisol awakening response (normal slope), DHEA/cortisol ratio</td>
<td align="center" valign="middle">25%</td>
</tr>
<tr>
<td align="left" valign="middle">Inflammatory resilience</td>
<td align="left" valign="middle">IL-10/IL-6 ratio, CRP (&#x003C;3&#x202F;mg/L), TNF-&#x03B1;</td>
<td align="center" valign="middle">20%</td>
</tr>
<tr>
<td align="left" valign="middle">Multi-system synchrony</td>
<td align="left" valign="middle">Correlation between epigenetic age (Horvath clock) and physiological stress markers</td>
<td align="center" valign="middle">20%</td>
</tr>
<tr>
<td align="left" valign="middle">Interoceptive accuracy</td>
<td align="left" valign="middle">Heartbeat detection task accuracy, anterior insula functional connectivity</td>
<td align="center" valign="middle">20%</td>
</tr>
<tr>
<td align="left" valign="middle">Circular recovery</td>
<td align="left" valign="middle">Speed of return to baseline (cortisol, HRV, cytokines) after acute stress challenge</td>
<td align="center" valign="middle">15%</td>
</tr>
</tbody>
</table>
</table-wrap>
<p>Distinction from related constructs:</p>
<list list-type="bullet">
<list-item>
<p>Allostatic load =&#x202F;cumulative wear and tear (higher values&#x202F;=&#x202F;worse)</p>
</list-item>
<list-item>
<p>Allostatic integrity =&#x202F;dynamic adaptive capacity (higher values&#x202F;=&#x202F;better)</p>
</list-item>
<list-item>
<p>Frailty index =&#x202F;deficit accumulation (broader, includes functional decline)</p>
</list-item>
<list-item>
<p>Biological aging metrics =&#x202F;epigenetic clocks (one component, not integrative)</p>
</list-item>
</list>
<p>Provisional formula: Allostatic Integrity Score&#x202F;= <italic>&#x03A3;</italic> (w&#x1D62; &#x00D7;&#x202F;z&#x1D62;) where w&#x1D62; are weights and z&#x1D62; are standardized scores, with higher scores indicating greater adaptive capacity. This operationalization requires empirical validation in future studies.</p>
</sec>
<sec id="sec38">
<label>5.3</label>
<title>Reconciling clinical paradoxes through a systems Lens</title>
<p>The model&#x2019;s most immediate contribution is its ability to reconcile several long-standing paradoxes that linear models struggle to explain.</p>
<p>Paradox 1: Dissociation between pathology and clinical expression. The observation that individuals with substantial amyloid and tau pathology may remain cognitively intact while others with minimal pathology decline rapidly (<xref ref-type="bibr" rid="ref32">Savva et al., 2009</xref>) has challenged the amyloid cascade hypothesis for decades. The circular model resolves this paradox by positing that neuropathology is one input among many into a dynamic system. When allostatic integrity is preserved, the system can tolerate substantial pathology without clinical expression. When allostatic integrity is compromised, even modest pathology could trigger cascading dysfunction across levels. Allostatic integrity thus may function as a moderator of pathology&#x2019;s clinical impact.</p>
<p>Paradox 2: The role of early-life stress in late-life dementia. The finding that childhood adversity is associated with dementia risk decades later (<xref ref-type="bibr" rid="ref24">Livingston et al., 2020</xref>; <xref ref-type="bibr" rid="ref21">Lee et al., 2025</xref>) is difficult to reconcile with models that focus on late-life protein aggregation. The circular model explains this through epigenetic embedding: early-life stress induces lasting epigenetic modifications (<xref ref-type="bibr" rid="ref20">Klengel et al., 2013</xref>; <xref ref-type="bibr" rid="ref43">Yusupov et al., 2024</xref>) that calibrate stress reactivity and allostatic load dynamics across the lifespan. These early influences may propagate upward through the system, shaping vulnerability decades before clinical symptoms emerge.</p>
<p>Paradox 3: The failure of single-target therapies. Despite clear evidence that amyloid and tau are involved in Alzheimer&#x2019;s disease, anti-amyloid therapies have produced only modest clinical benefits (<xref ref-type="bibr" rid="ref6">Cummings et al., 2024</xref>). The circular model explains this by emphasizing that dementias are systems-level disorders; targeting a single molecular entity may not fully restore circular information flow when multiple levels are dysregulated. This does not imply that anti-amyloid therapies are without value, but rather that they may need to be combined with interventions targeting other levels to achieve meaningful clinical effects&#x2014;a prediction consistent with the multiplicative effects hypothesis (Prediction 2).</p>
</sec>
<sec id="sec39">
<label>5.4</label>
<title>Implications for biomarker development and longitudinal cohorts</title>
<p>The model has concrete implications for how biomarkers are developed and how longitudinal cohorts are designed.</p>
<p>From single biomarkers to composite indices. Current biomarker strategies in Alzheimer&#x2019;s disease research have focused on identifying single molecular entities (amyloid, tau, neurodegeneration) that correlate with pathology. The model suggests that composite indices capturing allostatic integrity across multiple levels will have greater predictive validity. This implies a shift toward multi-modal biomarker panels that integrate genetic markers (Level 1), epigenetic markers (Level 2), multi-system physiological markers (Level 3), neuroimmune and endocrine markers (Level 4), and interoceptive and functional connectivity markers (Level 5).</p>
<p>Longitudinal cohort design. Existing cohorts such as ADNI, ROSMAP, and the UK Biobank have collected rich multi-domain data, enabling initial tests of the model&#x2019;s predictions. However, future cohorts should be designed explicitly to capture circular dynamics across levels. This requires repeated measures across multiple time points to model dynamic trajectories, integration of central and peripheral measures to capture central&#x2013;peripheral crosstalk, assessment of interoceptive function using validated behavioral and neuroimaging paradigms, and collection of early-life exposure data to model epigenetic embedding. The model also suggests that allostatic integrity should be treated as a primary outcome in its own right, rather than merely a predictor of cognitive decline. Interventions that restore allostatic integrity may confer resilience even without directly reducing neuropathological burden.</p>
</sec>
<sec id="sec40">
<label>5.5</label>
<title>Implications for multidomain prevention and personalized treatment</title>
<p>The model&#x2019;s emphasis on circular dynamics across multiple levels has profound implications for prevention and treatment strategies.</p>
<p>Multidomain prevention. The FINGER trial (<xref ref-type="bibr" rid="ref19">Kivipelto et al., 2017</xref>) demonstrated that multidomain interventions can slow cognitive decline in at-risk older adults. The circular model provides a mechanistic rationale for this approach: because levels are coupled through bidirectional feedback loops, interventions targeting multiple levels may disrupt maladaptive circular dynamics more effectively than single-domain approaches. The model further predicts that such interventions will show multiplicative, rather than additive, effects (Prediction 2)&#x2014;a prediction that can be tested in future trials.</p>
<p>Personalized treatment based on allostatic integrity profiles. If allostatic integrity moderates the clinical impact of neuropathology, then treatment strategies should be tailored based on an individual&#x2019;s allostatic integrity profile. Individuals with high allostatic integrity may tolerate substantial pathology without requiring aggressive anti-amyloid therapy. Conversely, individuals with low allostatic integrity may benefit from interventions targeting the specific levels where their system is most dysregulated&#x2014;whether that be epigenetic (Level 2), allostatic (Level 3), PNEI (Level 4), or interoceptive (Level 5)&#x2014;in addition to any anti-pathology therapies. This personalized approach aligns with broader movements toward precision medicine, but with a crucial difference: rather than stratifying patients solely by genetic or molecular biomarkers, the model advocates stratification by systems-level integrity&#x2014;a dynamic property that integrates information across multiple levels of biological organization.</p>
</sec>
<sec id="sec41">
<label>5.6</label>
<title>Limitations and future directions</title>
<p>While the Hierarchical-Circular Model offers a unifying framework, several limitations must be acknowledged and addressed through future research.</p>
<p>Complexity and testability. The model&#x2019;s strength&#x2014;its integration of multiple levels&#x2014;is also a source of complexity. Testing the model&#x2019;s predictions requires multi-modal data, sophisticated analytical approaches (e.g., dynamical systems modeling, causal mediation analysis), and large sample sizes to detect interaction effects. However, the model is explicitly designed to generate falsifiable predictions, and existing cohorts provide initial opportunities for empirical testing.</p>
<p>Causality vs. Correlation. Much of the evidence synthesized in Section 3 is correlational. Establishing causality requires intervention studies that manipulate specific levels and observe effects across the system. The model predicts that interventions targeting multiple levels will produce multiplicative effects, providing a strong test of causal circular dynamics.</p>
<p>Generalizability across dementia subtypes. The model was developed with Alzheimer&#x2019;s disease as the primary focus, but its principles may extend to other dementias. <xref ref-type="bibr" rid="ref14">Hazelton et al. (2025a</xref>, <xref ref-type="bibr" rid="ref15">b)</xref> showed convergence of interoceptive, emotional, and social cognitive disruptions across neurodegenerative diseases, suggesting that the model&#x2019;s transdiagnostic claims are plausible. Future work should test whether allostatic integrity predicts outcomes in vascular dementia, frontotemporal dementia, and Lewy body dementia.</p>
<p>Operationalizing allostatic integrity. A critical next step is developing validated composite indices of allostatic integrity that can be used across studies. Section 5.2 provides a provisional operationalization, but empirical work is needed to refine these indices, determine optimal weights, and validate them across diverse populations.</p>
</sec>
</sec>
<sec sec-type="conclusions" id="sec42">
<label>6</label>
<title>Conclusion</title>
<p>The Hierarchical-Circular Model of Biological Memory offers a unifying hypothesis for Alzheimer&#x2019;s disease and related dementias that bridges genetic, epigenetic, physiological, neuroimmune, and interoceptive processes across the lifespan. By reframing dementias as disorders of biological memory and allostatic integrity&#x2014;rather than isolated brain pathologies&#x2014;the model resolves key clinical paradoxes, generates testable predictions, and provides a conceptual roadmap for mechanistic research, multidomain prevention, and personalized treatment strategies.</p>
<p>The model&#x2019;s central contribution is the concept of allostatic integrity: the capacity of the organism to maintain adaptive circular information flow across multiple levels of organization. When allostatic integrity is preserved, the system may tolerate substantial neuropathology without clinical expression. When allostatic integrity is compromised, even modest pathology could trigger cascading dysfunction across levels, culminating in dementia.</p>
<p>The four core predictions derived from the model&#x2014;(1) composite indices of allostatic integrity will outperform single biomarkers in predicting conversion from mild cognitive impairment to dementia; (2) multidomain interventions targeting more than one system level will have multiplicative effects on slowing cognitive decline; (3) patients with similar amyloid/tau profiles but contrasting allostatic integrity will show divergent clinical trajectories; and (4) allostatic integrity moderates the protective effect of cognitive reserve&#x2014;provide clear opportunities for empirical testing using existing cohorts and trial designs.</p>
<p>Ultimately, the Hierarchical-Circular Model invites a shift in perspective: from viewing dementias as diseases of the brain in isolation to understanding them as failures of the organism&#x2019;s capacity to maintain integrated, adaptive information flow across the lifespan. This reframing opens new avenues for research and intervention, emphasizing resilience, systems-level integrity, and the dynamic interplay between genes, environment, physiology, and experience.</p>
</sec>
</body>
<back>
<sec sec-type="data-availability" id="sec43">
<title>Data availability statement</title>
<p>The original contributions presented in the study are included in the article/supplementary material, further inquiries can be directed to the corresponding author.</p>
</sec>
<sec sec-type="author-contributions" id="sec44">
<title>Author contributions</title>
<p>SR: Writing &#x2013; original draft, Methodology, Data curation, Validation, Conceptualization, Visualization, Investigation, Supervision, Funding acquisition, Project administration, Resources, Writing &#x2013; review &#x0026; editing, Formal analysis, Software.</p>
</sec>
<sec sec-type="COI-statement" id="sec45">
<title>Conflict of interest</title>
<p>The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.</p>
</sec>
<sec sec-type="ai-statement" id="sec46">
<title>Generative AI statement</title>
<p>The author(s) declared that Generative AI was used in the creation of this manuscript. Generative AI was used for copyediting and text condensation under the authors' full supervision and review. AI was not employed in the generation of scientific ideas, hypotheses, theoretical structure, or data interpretation.</p>
<p>Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.</p>
</sec>
<sec sec-type="disclaimer" id="sec47">
<title>Publisher&#x2019;s note</title>
<p>All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.</p>
</sec>
<ref-list>
<title>References</title>
<ref id="ref1"><mixed-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Barrett</surname><given-names>L. F.</given-names></name> <name><surname>Simmons</surname><given-names>W. K.</given-names></name></person-group> (<year>2015</year>). <article-title>Interoceptive predictions in the brain</article-title>. <source>Nat. Rev. Neurosci.</source> <volume>16</volume>, <fpage>419</fpage>&#x2013;<lpage>429</lpage>. doi: <pub-id pub-id-type="doi">10.1038/nrn3950</pub-id>, <pub-id pub-id-type="pmid">26016744</pub-id></mixed-citation></ref>
<ref id="ref2"><mixed-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Belsky</surname><given-names>J.</given-names></name> <name><surname>Pluess</surname><given-names>M.</given-names></name></person-group> (<year>2009</year>). <article-title>Beyond diathesis stress: differential susceptibility to environmental influences</article-title>. <source>Psychol. Bull.</source> <volume>135</volume>, <fpage>885</fpage>&#x2013;<lpage>908</lpage>. doi: <pub-id pub-id-type="doi">10.1037/a0017376</pub-id>, <pub-id pub-id-type="pmid">19883141</pub-id></mixed-citation></ref>
<ref id="ref3"><mixed-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Chen</surname><given-names>Y.</given-names></name> <name><surname>Zhang</surname><given-names>X.</given-names></name> <name><surname>Wang</surname><given-names>L.</given-names></name> <etal/></person-group>. (<year>2025</year>). <article-title>Peripheral inflammation in Alzheimer's disease: mechanisms, biomarkers, and therapeutic opportunities</article-title>. <source>Mol. Neurodegener.</source> <volume>20</volume>:<fpage>15</fpage>. doi: <pub-id pub-id-type="doi">10.1186/s13024-025-00812-3</pub-id></mixed-citation></ref>
<ref id="ref4"><mixed-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Craig</surname><given-names>A. D.</given-names></name></person-group> (<year>2002</year>). <article-title>How do you feel? Interoception: the sense of the physiological condition of the body</article-title>. <source>Nat. Rev. Neurosci.</source> <volume>3</volume>, <fpage>655</fpage>&#x2013;<lpage>666</lpage>. doi: <pub-id pub-id-type="doi">10.1038/nrn894</pub-id>, <pub-id pub-id-type="pmid">12154366</pub-id></mixed-citation></ref>
<ref id="ref5"><mixed-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Critchley</surname><given-names>H. D.</given-names></name> <name><surname>Garfinkel</surname><given-names>S. N.</given-names></name></person-group> (<year>2017</year>). <article-title>The influence of physiological signals on cognition</article-title>. <source>Curr. Opin. Behav. Sci.</source> <volume>19</volume>, <fpage>13</fpage>&#x2013;<lpage>18</lpage>. doi: <pub-id pub-id-type="doi">10.1016/j.cobeha.2017.08.014</pub-id></mixed-citation></ref>
<ref id="ref6"><mixed-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Cummings</surname><given-names>J.</given-names></name> <name><surname>Zhou</surname><given-names>Y.</given-names></name> <name><surname>Lee</surname><given-names>G.</given-names></name> <name><surname>Zhong</surname><given-names>K.</given-names></name> <name><surname>Fonseca</surname><given-names>J.</given-names></name> <name><surname>Cheng</surname><given-names>F.</given-names></name></person-group> (<year>2024</year>). <article-title>Alzheimer's disease drug development pipeline: 2024</article-title>. <source>Alzheimers Dement (N Y)</source> <volume>10</volume>:<fpage>e12465</fpage>. doi: <pub-id pub-id-type="doi">10.1002/trc2.12465</pub-id>, <pub-id pub-id-type="pmid">38659717</pub-id></mixed-citation></ref>
<ref id="ref7"><mixed-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Danese</surname><given-names>A.</given-names></name> <name><surname>McEwen</surname><given-names>B. S.</given-names></name></person-group> (<year>2012</year>). <article-title>Adverse childhood experiences, allostasis, allostatic load, and age-related disease</article-title>. <source>Physiol. Behav.</source> <volume>106</volume>, <fpage>29</fpage>&#x2013;<lpage>39</lpage>. doi: <pub-id pub-id-type="doi">10.1016/j.physbeh.2011.08.019</pub-id>, <pub-id pub-id-type="pmid">21888923</pub-id></mixed-citation></ref>
<ref id="ref8"><mixed-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Dantzer</surname><given-names>R.</given-names></name> <name><surname>O'Connor</surname><given-names>J. C.</given-names></name> <name><surname>Freund</surname><given-names>G. G.</given-names></name> <name><surname>Johnson</surname><given-names>R. W.</given-names></name> <name><surname>Kelley</surname><given-names>K. W.</given-names></name></person-group> (<year>2008</year>). <article-title>From inflammation to sickness and depression: when the immune system subjugates the brain</article-title>. <source>Nat. Rev. Neurosci.</source> <volume>9</volume>, <fpage>46</fpage>&#x2013;<lpage>56</lpage>. doi: <pub-id pub-id-type="doi">10.1038/nrn2297</pub-id>, <pub-id pub-id-type="pmid">18073775</pub-id></mixed-citation></ref>
<ref id="ref9"><mixed-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Davies</surname><given-names>D. A.</given-names></name> <name><surname>Addo</surname><given-names>K. A.</given-names></name> <name><surname>Klingenberg</surname><given-names>M. L.</given-names></name> <name><surname>Van Erum</surname><given-names>J.</given-names></name> <name><surname>Van Dam</surname><given-names>D.</given-names></name> <name><surname>De Deyn</surname><given-names>P. P.</given-names></name></person-group> (<year>2019</year>). <article-title>Alzheimer&#x2019;s disease: neurotransmitters of the sleep-wake cycle</article-title>. <source>Neurosci. Biobehav. Rev.</source> <volume>105</volume>, <fpage>72</fpage>&#x2013;<lpage>80</lpage>. doi: <pub-id pub-id-type="doi">10.1016/j.neubiorev.2019.07.019</pub-id>, <pub-id pub-id-type="pmid">31377219</pub-id></mixed-citation></ref>
<ref id="ref10"><mixed-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Feng</surname><given-names>X.</given-names></name> <name><surname>Jia</surname><given-names>M.</given-names></name> <name><surname>Cai</surname><given-names>M.</given-names></name> <name><surname>Zhu</surname><given-names>T.</given-names></name> <name><surname>Hashimoto</surname><given-names>K.</given-names></name> <name><surname>Yang</surname><given-names>J. J.</given-names></name></person-group> (<year>2025</year>). <article-title>Central-peripheral neuroimmune dynamics in psychological stress and depression: insights from current research</article-title>. <source>Mol. Psychiatry</source> <volume>30</volume>, <fpage>4881</fpage>&#x2013;<lpage>4898</lpage>. doi: <pub-id pub-id-type="doi">10.1038/s41380-025-03039-6</pub-id>, <pub-id pub-id-type="pmid">40610703</pub-id></mixed-citation></ref>
<ref id="ref11"><mixed-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Fratiglioni</surname><given-names>L.</given-names></name> <name><surname>Marseglia</surname><given-names>A.</given-names></name> <name><surname>Dekhtyar</surname><given-names>S.</given-names></name></person-group> (<year>2020</year>). <article-title>Ageing without dementia: can stimulating psychosocial and lifestyle experiences make a difference?</article-title> <source>Lancet Neurol.</source> <volume>19</volume>, <fpage>533</fpage>&#x2013;<lpage>543</lpage>. doi: <pub-id pub-id-type="doi">10.1016/S1474-4422(20)30039-9</pub-id>, <pub-id pub-id-type="pmid">32470425</pub-id></mixed-citation></ref>
<ref id="ref12"><mixed-citation publication-type="journal"><person-group person-group-type="author"><collab id="coll1">GBD 2019 Dementia Forecasting Collaborators</collab></person-group> (<year>2022</year>). <article-title>Estimation of the global prevalence of dementia in 2019 and forecasted prevalence in 2050: an analysis for the global burden of disease study 2019</article-title>. <source>Lancet Public Health</source> <volume>7</volume>, <fpage>e105</fpage>&#x2013;<lpage>e125</lpage>. doi: <pub-id pub-id-type="doi">10.1016/S2468-2667(21)00249-8</pub-id>, <pub-id pub-id-type="pmid">34998485</pub-id></mixed-citation></ref>
<ref id="ref13"><mixed-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Guidi</surname><given-names>J.</given-names></name> <name><surname>Lucente</surname><given-names>M.</given-names></name> <name><surname>Sonino</surname><given-names>N.</given-names></name> <name><surname>Fava</surname><given-names>G. A.</given-names></name></person-group> (<year>2021</year>). <article-title>Allostatic load and its impact on health: A systematic review</article-title>. <source>Psychother. Psychosom.</source> <volume>90</volume>, <fpage>11</fpage>&#x2013;<lpage>27</lpage>. doi: <pub-id pub-id-type="doi">10.1159/000510696</pub-id>, <pub-id pub-id-type="pmid">32799204</pub-id></mixed-citation></ref>
<ref id="ref14"><mixed-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Hazelton</surname><given-names>J. L.</given-names></name> <name><surname>Carneiro</surname><given-names>F.</given-names></name> <name><surname>Maito</surname><given-names>M.</given-names></name> <name><surname>Richter</surname><given-names>F.</given-names></name> <name><surname>Legaz</surname><given-names>A.</given-names></name> <name><surname>Altschuler</surname><given-names>F.</given-names></name> <etal/></person-group>. (<year>2025a</year>). <article-title>Neuroimaging meta-analyses reveal convergence of interoception, emotion, and social cognition across neurodegenerative diseases</article-title>. <source>Biol. Psychiatry</source> <volume>97</volume>, <fpage>1079</fpage>&#x2013;<lpage>1090</lpage>. doi: <pub-id pub-id-type="doi">10.1016/j.biopsych.2024.10.013</pub-id>, <pub-id pub-id-type="pmid">39442786</pub-id></mixed-citation></ref>
<ref id="ref15"><mixed-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Hazelton</surname><given-names>J. L.</given-names></name> <name><surname>Della Bella</surname><given-names>G.</given-names></name> <name><surname>Barttfeld</surname><given-names>P.</given-names></name> <name><surname>Dottori</surname><given-names>M.</given-names></name> <name><surname>Gonzalez-Gomez</surname><given-names>R.</given-names></name> <name><surname>Migeot</surname><given-names>J.</given-names></name> <etal/></person-group>. (<year>2025b</year>). <article-title>Altered spatiotemporal brain dynamics of interoception in behavioural-variant frontotemporal dementia</article-title>. <source>EBioMedicine</source> <volume>113</volume>:<fpage>105614</fpage>. doi: <pub-id pub-id-type="doi">10.1016/j.ebiom.2025.105614</pub-id>, <pub-id pub-id-type="pmid">39987747</pub-id></mixed-citation></ref>
<ref id="ref16"><mixed-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Jack</surname><given-names>C. R.</given-names><suffix>Jr.</suffix></name> <name><surname>Andrews</surname><given-names>J. S.</given-names></name> <name><surname>Beach</surname><given-names>T. G.</given-names></name> <name><surname>Buracchio</surname><given-names>T.</given-names></name> <name><surname>Dunn</surname><given-names>B.</given-names></name> <name><surname>Graf</surname><given-names>A.</given-names></name> <etal/></person-group>. (<year>2024</year>). <article-title>Revised criteria for diagnosis and staging of Alzheimer's disease: Alzheimer's Association workgroup</article-title>. <source>Alzheimers Dement.</source> <volume>20</volume>, <fpage>5143</fpage>&#x2013;<lpage>5169</lpage>. doi: <pub-id pub-id-type="doi">10.1002/alz.13859</pub-id>, <pub-id pub-id-type="pmid">38934362</pub-id></mixed-citation></ref>
<ref id="ref17"><mixed-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Jack</surname><given-names>C. R.</given-names><suffix>Jr.</suffix></name> <name><surname>Bennett</surname><given-names>D. A.</given-names></name> <name><surname>Blennow</surname><given-names>K.</given-names></name> <name><surname>Carrillo</surname><given-names>M. C.</given-names></name> <name><surname>Dunn</surname><given-names>B.</given-names></name> <name><surname>Haeberlein</surname><given-names>S. B.</given-names></name> <etal/></person-group>. (<year>2018</year>). <article-title>NIA-AA research framework: toward a biological definition of Alzheimer's disease</article-title>. <source>Alzheimers Dement.</source> <volume>14</volume>, <fpage>535</fpage>&#x2013;<lpage>562</lpage>. doi: <pub-id pub-id-type="doi">10.1016/j.jalz.2018.02.018</pub-id>, <pub-id pub-id-type="pmid">29653606</pub-id></mixed-citation></ref>
<ref id="ref18"><mixed-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Khalsa</surname><given-names>S. S.</given-names></name> <name><surname>Adolphs</surname><given-names>R.</given-names></name> <name><surname>Cameron</surname><given-names>O. G.</given-names></name> <name><surname>Critchley</surname><given-names>H. D.</given-names></name> <name><surname>Davenport</surname><given-names>P. W.</given-names></name> <name><surname>Feinstein</surname><given-names>J. S.</given-names></name> <etal/></person-group>. (<year>2018</year>). <article-title>Interoception and mental health: A roadmap</article-title>. <source>Biol Psychiatry Cogn Neurosci Neuroimaging.</source> <volume>3</volume>, <fpage>501</fpage>&#x2013;<lpage>513</lpage>. doi: <pub-id pub-id-type="doi">10.1016/j.bpsc.2017.12.004</pub-id>, <pub-id pub-id-type="pmid">29884281</pub-id></mixed-citation></ref>
<ref id="ref19"><mixed-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Kivipelto</surname><given-names>M.</given-names></name> <name><surname>Ngandu</surname><given-names>T.</given-names></name> <name><surname>Rosenberg</surname><given-names>A.</given-names></name> <name><surname>Schwarz</surname><given-names>C.</given-names></name> <name><surname>Wirth</surname><given-names>M.</given-names></name> <name><surname>Gerischer</surname><given-names>L.</given-names></name> <etal/></person-group>. (<year>2017</year>). <article-title>Effects of Omega-3 fatty acids on resting cerebral perfusion in patients with mild cognitive impairment: a randomized controlled trial</article-title>. <source>J. Prev Alzheimers Dis.</source> <volume>5</volume>, <fpage>26</fpage>&#x2013;<lpage>30</lpage>. doi: <pub-id pub-id-type="doi">10.14283/jpad.2017.23</pub-id>, <pub-id pub-id-type="pmid">29405229</pub-id></mixed-citation></ref>
<ref id="ref20"><mixed-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Klengel</surname><given-names>T.</given-names></name> <name><surname>Mehta</surname><given-names>D.</given-names></name> <name><surname>Anacker</surname><given-names>C.</given-names></name> <name><surname>Rex-Haffner</surname><given-names>M.</given-names></name> <name><surname>Pruessner</surname><given-names>J. C.</given-names></name> <name><surname>Pariante</surname><given-names>C. M.</given-names></name> <etal/></person-group>. (<year>2013</year>). <article-title>Allele-specific FKBP5 DNA demethylation mediates gene&#x2013;childhood trauma interactions</article-title>. <source>Nat. Neurosci.</source> <volume>16</volume>, <fpage>33</fpage>&#x2013;<lpage>41</lpage>. doi: <pub-id pub-id-type="doi">10.1038/nn.3275</pub-id>, <pub-id pub-id-type="pmid">23201972</pub-id></mixed-citation></ref>
<ref id="ref21"><mixed-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Lee</surname><given-names>H.</given-names></name> <name><surname>Kim</surname><given-names>J.</given-names></name> <name><surname>Chen</surname><given-names>Y.</given-names></name> <name><surname>Ali</surname><given-names>M.</given-names></name> <name><surname>Wollenhaupt-Aguiar</surname><given-names>B.</given-names></name> <name><surname>Sehmbi</surname><given-names>M.</given-names></name> <etal/></person-group>. (<year>2025</year>). <article-title>Peripheral neurofilament light chain and intracortical myelin in bipolar I disorder</article-title>. <source>J. Affect. Disord.</source> <volume>374</volume>, <fpage>184</fpage>&#x2013;<lpage>190</lpage>. doi: <pub-id pub-id-type="doi">10.1016/j.jad.2025.01.045</pub-id>, <pub-id pub-id-type="pmid">39800065</pub-id></mixed-citation></ref>
<ref id="ref22"><mixed-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Li</surname><given-names>C.</given-names></name> <name><surname>Hao</surname><given-names>C.</given-names></name> <name><surname>Zhou</surname><given-names>K.</given-names></name> <name><surname>Zhang</surname><given-names>P.</given-names></name> <name><surname>Sun</surname><given-names>Q.</given-names></name> <name><surname>Li</surname><given-names>Z.</given-names></name></person-group> (<year>2025</year>). <article-title>Inflammation-mediated regional brain alterations associated with mild cognitive impairment in knee osteoarthritis</article-title>. <source>Arthritis Res. Ther.</source> <volume>27</volume>:<fpage>181</fpage>. doi: <pub-id pub-id-type="doi">10.1186/s13075-025-03646-0</pub-id>, <pub-id pub-id-type="pmid">41013635</pub-id></mixed-citation></ref>
<ref id="ref23"><mixed-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Liu</surname><given-names>Q.</given-names></name> <name><surname>Xie</surname><given-names>Z.</given-names></name> <name><surname>Zhang</surname><given-names>W.</given-names></name> <etal/></person-group>. (<year>2023</year>). <article-title>Peripheral inflammatory insults and Alzheimer's disease: a systematic review and meta-analysis</article-title>. <source>J. Neuroinflammation</source> <volume>20</volume>:<fpage>245</fpage>. doi: <pub-id pub-id-type="doi">10.1186/s12974-023-02945-6</pub-id></mixed-citation></ref>
<ref id="ref24"><mixed-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Livingston</surname><given-names>G.</given-names></name> <name><surname>Huntley</surname><given-names>J.</given-names></name> <name><surname>Sommerlad</surname><given-names>A.</given-names></name> <name><surname>Ames</surname><given-names>D.</given-names></name> <name><surname>Ballard</surname><given-names>C.</given-names></name> <name><surname>Banerjee</surname><given-names>S.</given-names></name> <etal/></person-group>. (<year>2020</year>). <article-title>Dementia prevention, intervention, and care: 2020 report of the lancet commission</article-title>. <source>Lancet</source> <volume>396</volume>, <fpage>413</fpage>&#x2013;<lpage>446</lpage>. doi: <pub-id pub-id-type="doi">10.1016/S0140-6736(20)30367-6</pub-id>, <pub-id pub-id-type="pmid">32738937</pub-id></mixed-citation></ref>
<ref id="ref25"><mixed-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Long</surname><given-names>J. M.</given-names></name> <name><surname>Holtzman</surname><given-names>D. M.</given-names></name></person-group> (<year>2019</year>). <article-title>Alzheimer disease: an update on pathobiology and treatment strategies</article-title>. <source>Cell</source> <volume>179</volume>, <fpage>312</fpage>&#x2013;<lpage>339</lpage>. doi: <pub-id pub-id-type="doi">10.1016/j.cell.2019.09.001</pub-id>, <pub-id pub-id-type="pmid">31564456</pub-id></mixed-citation></ref>
<ref id="ref26"><mixed-citation publication-type="journal"><person-group person-group-type="author"><name><surname>McEwen</surname><given-names>B. S.</given-names></name> <name><surname>Stellar</surname><given-names>E.</given-names></name></person-group> (<year>1993</year>). <article-title>Stress and the individual: mechanisms leading to disease</article-title>. <source>Arch. Intern. Med.</source> <volume>153</volume>, <fpage>2093</fpage>&#x2013;<lpage>2101</lpage>. doi: <pub-id pub-id-type="doi">10.1001/archinte.1993.00410180039004</pub-id>, <pub-id pub-id-type="pmid">8379800</pub-id></mixed-citation></ref>
<ref id="ref27"><mixed-citation publication-type="journal"><person-group person-group-type="author"><name><surname>McNamara</surname><given-names>H. M.</given-names></name> <name><surname>Solley</surname><given-names>S. C.</given-names></name> <name><surname>Adamson</surname><given-names>B.</given-names></name> <name><surname>Chan</surname><given-names>M. M.</given-names></name> <name><surname>Toettcher</surname><given-names>J. E.</given-names></name></person-group> (<year>2024</year>). <article-title>Recording morphogen signals reveals mechanisms underlying gastruloid symmetry breaking</article-title>. <source>Nat. Cell Biol.</source> <volume>26</volume>, <fpage>1832</fpage>&#x2013;<lpage>1844</lpage>. doi: <pub-id pub-id-type="doi">10.1038/s41556-024-01521-9</pub-id>, <pub-id pub-id-type="pmid">39358450</pub-id></mixed-citation></ref>
<ref id="ref28"><mixed-citation publication-type="journal"><person-group person-group-type="author"><name><surname>N&#x00F6;thling</surname><given-names>J.</given-names></name> <name><surname>Womersley</surname><given-names>J. S.</given-names></name> <name><surname>Mhlongo</surname><given-names>S.</given-names></name> <name><surname>Lombard</surname><given-names>C.</given-names></name> <name><surname>Abrahams</surname><given-names>N.</given-names></name> <name><surname>Seedat</surname><given-names>S.</given-names></name> <etal/></person-group>. (<year>2025</year>). <article-title>The relationship between childhood trauma, rs1360780 genotypes, FKBP5 intron 7 methylation and posttraumatic stress disorder in women who have experienced rape</article-title>. <source>Eur. J. Psychotraumatol.</source> <volume>16</volume>:<fpage>2485707</fpage>. doi: <pub-id pub-id-type="doi">10.1080/20008066.2025.2485707</pub-id>, <pub-id pub-id-type="pmid">40242984</pub-id></mixed-citation></ref>
<ref id="ref29"><mixed-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Panitch</surname><given-names>R.</given-names></name> <name><surname>Sahelijo</surname><given-names>N.</given-names></name> <name><surname>Hu</surname><given-names>J.</given-names></name> <name><surname>Nho</surname><given-names>K.</given-names></name> <name><surname>Bennett</surname><given-names>D. A.</given-names></name> <name><surname>Lunetta</surname><given-names>K. L.</given-names></name> <etal/></person-group>. (<year>2024</year>). <article-title>APOE genotype-specific methylation patterns are linked to Alzheimer disease pathology and estrogen response</article-title>. <source>Transl. Psychiatry</source> <volume>14</volume>:<fpage>129</fpage>. doi: <pub-id pub-id-type="doi">10.1038/s41398-024-02834-x</pub-id>, <pub-id pub-id-type="pmid">38424036</pub-id></mixed-citation></ref>
<ref id="ref30"><mixed-citation publication-type="journal"><person-group person-group-type="author"><name><surname>P&#x00E9;rez-Carrasco</surname><given-names>E.</given-names></name> <name><surname>Arriagada</surname><given-names>C.</given-names></name> <name><surname>Mu&#x00F1;oz</surname><given-names>C.</given-names></name> <etal/></person-group>. (<year>2025</year>). <article-title>Allostatic load calculation in a Chilean older-adults cohort reveals distinct physiological aging trajectories</article-title>. <source>NPJ Aging.</source> <volume>1</volume>:<fpage>32</fpage>. doi: <pub-id pub-id-type="doi">10.1038/s41514-025-00215-8</pub-id></mixed-citation></ref>
<ref id="ref31"><mixed-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Ruan</surname><given-names>Y.</given-names></name> <name><surname>Zheng</surname><given-names>Y.</given-names></name> <name><surname>Hu</surname><given-names>H.</given-names></name> <etal/></person-group>. (<year>2025</year>). <article-title>Association of allostatic load with all-cause and cause-specific dementia: a prospective cohort study</article-title>. <source>Alzheimers Dement.</source> <volume>21</volume>:<fpage>e14012</fpage>. doi: <pub-id pub-id-type="doi">10.1002/alz.14012</pub-id></mixed-citation></ref>
<ref id="ref32"><mixed-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Savva</surname><given-names>G. M.</given-names></name> <name><surname>Wharton</surname><given-names>S. B.</given-names></name> <name><surname>Ince</surname><given-names>P. G.</given-names></name> <name><surname>Forster</surname><given-names>G.</given-names></name> <name><surname>Matthews</surname><given-names>F. E.</given-names></name> <name><surname>Brayne</surname><given-names>C.</given-names></name></person-group> (<year>2009</year>). <article-title>Age, neuropathology, and dementia</article-title>. <source>N. Engl. J. Med.</source> <volume>360</volume>, <fpage>2302</fpage>&#x2013;<lpage>2309</lpage>. doi: <pub-id pub-id-type="doi">10.1056/NEJMoa0806142</pub-id>, <pub-id pub-id-type="pmid">19474427</pub-id></mixed-citation></ref>
<ref id="ref33"><mixed-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Serrano-Pozo</surname><given-names>A.</given-names></name> <name><surname>Frosch</surname><given-names>M. P.</given-names></name> <name><surname>Hyman</surname><given-names>B. T.</given-names></name></person-group> (<year>2024</year>). <article-title>A scoping review of Alzheimer's disease hypotheses: an array of uni- and multifactorial models</article-title>. <source>Alzheimers Dement.</source> <volume>20</volume>, <fpage>1739</fpage>&#x2013;<lpage>1752</lpage>. doi: <pub-id pub-id-type="doi">10.1002/alz.13567</pub-id>, <pub-id pub-id-type="pmid">38093529</pub-id></mixed-citation></ref>
<ref id="ref34"><mixed-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Slavich</surname><given-names>G. M.</given-names></name> <name><surname>Irwin</surname><given-names>M. R.</given-names></name></person-group> (<year>2014</year>). <article-title>From stress to inflammation and major depressive disorder: A social signal transduction theory of depression</article-title>. <source>Psychol. Bull.</source> <volume>140</volume>, <fpage>774</fpage>&#x2013;<lpage>815</lpage>. doi: <pub-id pub-id-type="doi">10.1037/a0035302</pub-id>, <pub-id pub-id-type="pmid">24417575</pub-id></mixed-citation></ref>
<ref id="ref35"><mixed-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Smith</surname><given-names>L.</given-names></name> <name><surname>Gonz&#x00E1;lez</surname><given-names>M.</given-names></name> <name><surname>Paredes</surname><given-names>R.</given-names></name> <name><surname>Palix</surname><given-names>C.</given-names></name> <name><surname>Chauveau</surname><given-names>L.</given-names></name> <name><surname>Felisatti</surname><given-names>F.</given-names></name> <etal/></person-group>. (<year>2025</year>). <article-title>Allostatic load, a measure of cumulative physiological stress, impairs brain structure but not &#x03B2;-accumulation in older adults: an exploratory study</article-title>. <source>Front. Aging Neurosci.</source> <volume>17</volume>:<fpage>1508677</fpage>. doi: <pub-id pub-id-type="doi">10.3389/fnagi.2025.1508677</pub-id>, <pub-id pub-id-type="pmid">40230487</pub-id></mixed-citation></ref>
<ref id="ref36"><mixed-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Souza-Talarico</surname><given-names>J. N.</given-names></name> <name><surname>Perkhounkova</surname><given-names>Y.</given-names></name> <name><surname>Hein</surname><given-names>M.</given-names></name> <name><surname>Lee</surname><given-names>J.</given-names></name> <name><surname>Hefti</surname><given-names>M.</given-names></name> <name><surname>Sindi</surname><given-names>S.</given-names></name></person-group> (<year>2025</year>). <article-title>Allostatic load dynamics, Alzheimer's disease biomarkers, and progression in individuals with mild cognitive impairment: findings from the Alzheimer's disease neuroimaging initiative</article-title>. <source>Alzheimers Dement (Amst).</source> <volume>17</volume>:<fpage>e70140</fpage>. doi: <pub-id pub-id-type="doi">10.1002/dad2.70140</pub-id>, <pub-id pub-id-type="pmid">40606518</pub-id></mixed-citation></ref>
<ref id="ref37"><mixed-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Stern</surname><given-names>Y.</given-names></name> <name><surname>Albert</surname><given-names>M.</given-names></name> <name><surname>Barnes</surname><given-names>C. A.</given-names></name> <name><surname>Maestre</surname><given-names>G.</given-names></name> <name><surname>Carrillo</surname><given-names>M.</given-names></name> <name><surname>Kalaria</surname><given-names>R.</given-names></name> <etal/></person-group>. (<year>2023</year>). <article-title>The Nairobi declaration&#x2014;reducing the burden of dementia in low&#x2010; and middle&#x2010;income countries (LMICs): declaration of the 2022 symposium on dementia and brain aging in LMICs</article-title>. <source>Alzheimers Dement.</source> <volume>19</volume>, <fpage>1105</fpage>&#x2013;<lpage>1108</lpage>. doi: <pub-id pub-id-type="doi">10.1002/alz.13025</pub-id>, <pub-id pub-id-type="pmid">36905253</pub-id></mixed-citation></ref>
<ref id="ref38"><mixed-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Stern</surname><given-names>Y.</given-names></name> <name><surname>Arenaza-Urquijo</surname><given-names>E. M.</given-names></name> <name><surname>Bartr&#x00E9;s-Faz</surname><given-names>D.</given-names></name> <name><surname>Belleville</surname><given-names>S.</given-names></name> <name><surname>Cantilon</surname><given-names>M.</given-names></name> <name><surname>Chetelat</surname><given-names>G.</given-names></name> <etal/></person-group>. (<year>2020</year>). <article-title>Whitepaper: defining and investigating cognitive reserve, brain reserve, and brain maintenance</article-title>. <source>Alzheimers Dement.</source> <volume>16</volume>, <fpage>1305</fpage>&#x2013;<lpage>1311</lpage>. doi: <pub-id pub-id-type="doi">10.1016/j.jalz.2018.07.219</pub-id>, <pub-id pub-id-type="pmid">30222945</pub-id></mixed-citation></ref>
<ref id="ref39"><mixed-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Tyagi</surname><given-names>R.</given-names></name> <name><surname>Bartley</surname><given-names>C. M.</given-names></name></person-group> (<year>2025</year>). <article-title>Dynamic cytokine relationships across the blood-brain barrier in humans and nonhuman primates&#x2014;implications for psychiatric illness: a systematic review</article-title>. <source>Biol. Psychiatry</source> <volume>99</volume>, <fpage>983</fpage>&#x2013;<lpage>1001</lpage>. doi: <pub-id pub-id-type="doi">10.1016/j.biopsych.2025.08.022</pub-id>, <pub-id pub-id-type="pmid">40946890</pub-id></mixed-citation></ref>
<ref id="ref40"><mixed-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Tyrka</surname><given-names>A. R.</given-names></name> <name><surname>Parade</surname><given-names>S. H.</given-names></name> <name><surname>Welch</surname><given-names>E. S.</given-names></name> <etal/></person-group>. (<year>2016</year>). <article-title>Methylation of the glucocorticoid receptor gene (NR3C1) in childhood and adulthood: associations with adversity and depressive symptoms</article-title>. <source>Dev. Psychopathol.</source> <volume>28</volume>, <fpage>1371</fpage>&#x2013;<lpage>1385</lpage>. doi: <pub-id pub-id-type="doi">10.1017/S0954579416001007</pub-id></mixed-citation></ref>
<ref id="ref41"><mixed-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Verduzco Espinoza</surname><given-names>A. P.</given-names></name> <name><surname>Na</surname><given-names>N.</given-names></name> <name><surname>Campanati</surname><given-names>L.</given-names></name> <name><surname>Ngo</surname><given-names>P.</given-names></name> <name><surname>Baldwin</surname><given-names>K. K.</given-names></name> <name><surname>Cline</surname><given-names>H. T.</given-names></name></person-group> (<year>2025</year>). <article-title>Microglia-to-neuron signaling links APOE4 and inflammation to enhanced neuronal lipid metabolism and network activity</article-title>. <source>Proc. Natl. Acad. Sci. USA</source> <volume>122</volume>:<fpage>e2516103122</fpage>. doi: <pub-id pub-id-type="doi">10.1073/pnas.2516103122</pub-id>, <pub-id pub-id-type="pmid">40920927</pub-id></mixed-citation></ref>
<ref id="ref42"><mixed-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Wang</surname><given-names>Q.</given-names></name> <name><surname>Li</surname><given-names>X.</given-names></name> <name><surname>Zhang</surname><given-names>Y.</given-names></name> <name><surname>Bouchekioua</surname><given-names>Y.</given-names></name> <name><surname>Nebuka</surname><given-names>M.</given-names></name> <name><surname>Sasamori</surname><given-names>H.</given-names></name> <etal/></person-group>. (<year>2022</year>). <article-title>Serotonin 5-HT2C receptor knockout in mice attenuates fear responses in contextual or cued but not compound context-cue fear conditioning</article-title>. <source>Transl. Psychiatry</source> <volume>12</volume>:<fpage>68</fpage>. doi: <pub-id pub-id-type="doi">10.1038/s41398-022-01815-2</pub-id>, <pub-id pub-id-type="pmid">35145065</pub-id></mixed-citation></ref>
<ref id="ref43"><mixed-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Yusupov</surname><given-names>N.</given-names></name> <name><surname>Roeh</surname><given-names>S.</given-names></name> <name><surname>Sotillos Elliott</surname><given-names>L.</given-names></name> <name><surname>Chang</surname><given-names>S.</given-names></name> <name><surname>Loganathan</surname><given-names>S.</given-names></name> <name><surname>Urbina-Trevi&#x00F1;o</surname><given-names>L.</given-names></name> <etal/></person-group>. (<year>2024</year>). <article-title>DNA methylation patterns of FKBP5 regulatory regions in brain and blood of humanized mice and humans</article-title>. <source>Mol. Psychiatry</source> <volume>29</volume>, <fpage>1510</fpage>&#x2013;<lpage>1520</lpage>. doi: <pub-id pub-id-type="doi">10.1038/s41380-024-02430-x</pub-id>, <pub-id pub-id-type="pmid">38317011</pub-id></mixed-citation></ref>
<ref id="ref44"><mixed-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Zannas</surname><given-names>A. S.</given-names></name> <name><surname>Wiechmann</surname><given-names>T.</given-names></name> <name><surname>Gassen</surname><given-names>N. C.</given-names></name> <name><surname>Binder</surname><given-names>E. B.</given-names></name></person-group> (<year>2016</year>). <article-title>Epigenetic changes of FKBP5 as a link connecting genetic and environmental risk factors with stress-related mental disorders</article-title>. <source>Mol. Psychiatry</source> <volume>22</volume>, <fpage>605</fpage>&#x2013;<lpage>614</lpage>. doi: <pub-id pub-id-type="doi">10.1038/mp.2016.112</pub-id>, <pub-id pub-id-type="pmid">27431297</pub-id></mixed-citation></ref>
</ref-list>
<fn-group>
<fn fn-type="custom" custom-type="edited-by" id="fn0002">
<p>Edited by: <ext-link ext-link-type="uri" xlink:href="https://loop.frontiersin.org/people/15052/overview">Agneta Nordberg</ext-link>, Karolinska Institutet (KI), Sweden</p>
</fn>
<fn fn-type="custom" custom-type="reviewed-by" id="fn0003">
<p>Reviewed by: <ext-link ext-link-type="uri" xlink:href="https://loop.frontiersin.org/people/494929/overview">Dong Woo Kang</ext-link>, The Catholic University of Korea, Republic of Korea</p>
<p><ext-link ext-link-type="uri" xlink:href="https://loop.frontiersin.org/people/3295707/overview">Gamze Sonmez</ext-link>, Hacettepe University, T&#x00FC;rkiye</p>
</fn>
</fn-group>
</back>
</article>