Persistent Alzheimer-related molecular dysregulation after cerebral ischemia: a hierarchical-circular hypothesis of impaired allostatic recovery.

Introduction: Growing evidence suggests a consistent association between brain ischemia and the subsequent development of Alzheimer's disease (AD)-related molecular changes. Experimental and clinical studies have demonstrated that ischemic episodes trigger the overexpression of genes associated with AD pathology, including amyloid precursor protein (APP), β-secretase (BACE1), presenilin 1 and 2 (PSEN1, PSEN2), and MAPT (tau). However, the mechanisms by which transient or chronic ischemic events convert transient gene dysregulation into sustained neurodegenerative pathology remain poorly understood. We acknowledge that current evidence, largely derived from animal models, does not establish that ischemia alone is sufficient to initiate AD as defined by the 2024 NIA-AA biological criteria (core amyloid and tau biomarkers); rather, ischemia may act as a significant contributor that accelerates or exacerbates AD-related molecular changes in vulnerable individuals. Objective: To propose a multilevel reconceptualization of the ischemic contribution to AD within the hierarchical-circular model of biological memory and the construct of allostatic integrity, integrating recent evidence on post-ischemic gene expression, neuroinflammation, oxidative stress, and epigenetic mechanisms. Hypothesis and synthesis: We propose that cerebral ischemia acts as a high-intensity allostatic challenge that may "freeze" biological memory into a pathological configuration. Within this framework, ischemic episodes are not merely transient vascular events but rather serve as biological "programmers" that dysregulate AD-associated genes across multiple levels of biological organization: genetic-epigenetic (DNA methylation, microRNA dysregulation), cellular-mitochondrial (oxidative stress, mitophagy failure), synaptic-neurotrophic (loss of plasticity), neuroimmune-proteostatic (chronic neuroinflammation, protein misfolding), and behavioral-cognitive (progressive dementia). The persistent overexpression of APP, BACE1, PSEN1, PSEN2, and MAPT represents a failure of the system's capacity to restore adaptive circular information flow after an ischemic insult. Implications: This model reframes the ischemic contribution to AD from a linear cause-effect relationship to a dynamic, multilevel failure of allostatic adaptation. It provides a conceptual roadmap for understanding why ischemic events may predispose to sporadic AD and suggests that therapeutic strategies should target the restoration of allostatic integrity across multiple levels rather than single molecular pathways. Testable predictions: The framework generates falsifiable predictions, including: (1) composite indices of allostatic integrity will predict post-ischemic AD progression better than single biomarkers; (2) interventions that restore allostatic integrity (e.g., anti-inflammatory agents, epigenetic modulators, neuroprotective strategies) will attenuate AD-associated gene dysregulation; and (3) patients with higher allostatic load prior to an ischemic event will show more severe AD-associated gene expression changes and worse clinical outcomes.