1. Introduction

The brain and heart are among the most energy-intensive organs in the body. Both rely predominantly on oxidative metabolism and are therefore particularly sensitive to oxygen deprivation.^1,2 Accordingly, neurons and cardiomyocytes are highly susceptible to ischemic and metabolic stress.³ Reduced perfusion disrupts mitochondrial function, redox homeostasis, and ATP production, creating a shared metabolic vulnerability across disorders of both acute and chronic injury.^1,4 In ischemic stroke, abrupt loss of cerebral blood flow produces an immediate bioenergetic crisis, whereas in Alzheimer’s disease (AD), chronic vascular dysfunction and impaired perfusion impose a more sustained form of metabolic stress.^5,6,7 In this context, ischemic stroke and AD represent acute and chronic manifestations of oxygen and energy deprivation in the nervous system. Throughout this Review, chronic hypoperfusion is considered one important driver of oxygen-energy stress, although metabolic dysfunction, mitochondrial impairment, inflammation, and cardiovascular insufficiency may also contribute to energetic imbalance.^2-4,8 Accumulating clinical evidence indicates that stroke, heart failure, atrial fibrillation, and vascular cognitive impairment frequently coexist and are associated with worse neurological and cardiovascular outcomes, supporting the concept of a bidirectionally coupled brain-heart axis.^8,9

Neural and cardiovascular dysfunction frequently coexist because both systems share common biological constraints of limited regenerative capacity, high metabolic demand, and dependence on continuous oxygen and nutrient delivery. Cardiomyocytes and neurons possess limited regenerative potential under physiological conditions, and both rely heavily on mitochondrial oxidative phosphorylation to sustain function.^10,11 Consequently, injury to either organ often initiates secondary biological responses that extend beyond the primary site of damage, including metabolic reprogramming, mitochondrial dysfunction, inflammatory activation, vascular remodeling, and impaired adaptive capacity.^1,12 These interconnected responses suggest that disturbances in one organ may influence vulnerability in the other through shared metabolic and vascular pathways.^8,9

Growing evidence indicates that communication between the brain and heart extends beyond hemodynamic coupling alone.⁹ Acute cerebral ischemia can trigger autonomic dysregulation, catecholamine excess, myocardial injury, arrhythmias, and ventricular dysfunction, phenomena collectively recognized as stroke-heart syndrome (SHS).^13-15 Conversely, chronic cardiac dysfunction may impair cerebral perfusion, promote blood-brain barrier dysfunction, and accelerate cognitive decline and neurodegenerative processes.^4,16,17 This reciprocal relationship suggests that adaptive and maladaptive responses to oxygen-energy stress may be coordinated across organ systems rather than confined to a single tissue.^9,17

A central candidate for coordinating these responses is hypoxia-inducible factor (HIF) signaling. Across both acute injury and chronic degeneration, HIF functions as a conserved molecular sensor that links oxygen and metabolic stress to adaptive transcriptional responses involving cellular energetics, vascular function, inflammation, and mitochondrial homeostasis.^3,18-21 Because oxygen availability directly constrains mitochondrial ATP production in both the brain and heart, HIF signaling occupies a uniquely positioned role at the interface between oxygen sensing, metabolic adaptation, and tissue resilience.^2,22 However, accumulating evidence suggests that HIF responses are not uniform across tissues and cell types.^23,24 Their biological effects depend on the severity, duration, and cellular context of stress, varying substantially across neurons, astrocytes, endothelial cells, microglia, and cardiomyocytes.^23-25

Emerging evidence further suggests that HIF-1α and HIF-2α may not contribute equally across distinct temporal patterns of hypoxic stress.^26-28 Although both isoforms are regulated through a shared oxygen-sensing machinery, differences in stabilization kinetics, transcriptional selectivity, and cell-type distribution may favor HIF-1α engagement during acute ischemic injury and HIF2α-associated programs during prolonged hypoperfusion.^22,29 HIF-1α is generally associated with rapid adaptation to severe oxygen deprivation and preferentially activates glycolytic and acute stress-response pathways, whereas HIF-2α has more frequently been associated with endothelial homeostasis, angiogenesis, vascular remodeling, and longer-term adaptation to persistent hypoxic stress.^22,30,31 Importantly, much of the current understanding of HIF isoform specialization derives from vascular biology, developmental studies, and cancer research, whereas direct comparative evidence within the adult brain-heart axis remains limited.^27,29 Nevertheless, these distinctions provide a useful conceptual framework for understanding how acute ischemic stroke and chronic AD-associated hypoperfusion may engage partially distinct hypoxia-responsive programs (Fig.1).^27,29

In this Review, we examine how HIF signaling integrates oxygen-energy stress responses across the brain-heart axis. The overall conceptual framework developed throughout this Review is summarized in Fig.1. We then discuss how acute ischemic stroke engages HIF-dependent pathways that influence both neural and cardiac outcomes, followed by an examination of chronic hypoperfusion and neurodegenerative processes in AD. Finally, we consider emerging therapeutic strategies targeting HIF-regulated pathways and discuss how temporal, cell-type-specific, and organ-specific regulation of HIF signaling may inform future interventions for cerebrovascular and cardiometabolic disease. 

2. Molecular Regulation of HIF Signaling Under Oxygen-Energy Stress
Hypoxia-inducible factor (HIF) signaling is the principal molecular system by which mammalian cells sense and respond to changes in oxygen availability. Its activity is mediated primarily by the HIF-1α and HIF-2α isoforms, which share a common oxygen-sensing framework but differ in regulatory kinetics and downstream transcriptional outputs.^26,30,31 Through these isoform-sensitive programs, HIF coordinates cellular adaptation to hypoxic and energetic stress by regulating metabolic reprogramming, mitochondrial function, vascular remodeling, and inflammatory signaling.^3,19,20

This regulatory architecture is especially relevant in disease states marked by disrupted oxygen and energy homeostasis. Acute ischemic injury involves abrupt and spatially heterogeneous oxygen deprivation, whereas neurodegenerative and cardiovascular disorders more often develop under conditions of chronic, low-grade energetic strain.^5,32,33 Across these distinct temporal contexts, HIF serves as a conserved stress-responsive integrator that links oxygen and redox cues to adaptive, or when persistently dysregulated, maladaptive cellular responses.^2,3 Understanding how HIF is regulated across these settings provides the mechanistic basis for interpreting its context-dependent roles along the brain-heart axis.
2.1. Oxygen sensing and graded transcriptional control of HIF

HIF activity is governed primarily by oxygen-dependent post-translational regulation of the HIFα subunit. Under normoxic conditions, prolyl hydroxylase domain (PHD) enzymes hydroxylate conserved proline residues on HIF-α, thereby creating a recognition site for the von Hippel-Lindau (VHL) E3 ubiquitin ligase complex and targeting HIF-α for rapid proteasomal degradation.^19,20,34 Because PHD enzymes require molecular oxygen, iron, and α-ketoglutarate as cofactors, their activity closely reflects both oxygen availability and cellular metabolic state.^19,34

As oxygen tension declines, PHD activity is reduced, permitting HIF-α stabilization, nuclear accumulation, heterodimerization with HIF-β (ARNT), and binding to hypoxia response elements (HREs), thereby initiating hypoxia-responsive transcription. HIF signaling is further refined by factor inhibiting HIF (FIH), which hydroxylates an asparagine residue within the C-terminal transactivation domain and restricts recruitment of transcriptional coactivators such as p300/CBP.³⁵ Through these parallel checkpoints, HIF regulation operates not as a binary switch, but as a graded oxygen- and metabolism-sensitive rheostat.

Although HIF-1α and HIF-2α are governed by the same core sensing machinery, increasing evidence supports isoform-specific differences in stability, transcriptional selectivity, and physiological engagement across acute and chronic hypoxic states.^26,30 This layered regulatory architecture is especially relevant in conditions of partial, regionally heterogeneous, or fluctuating oxygen limitation, including peri-ischemic tissue,³⁶ chronically hypoperfused brain regions,⁵ and neurogenically stressed myocardium.³⁷ Its conservation across neurons, astrocytes, endothelial cells, and cardiomyocytes provides a mechanistic basis for coordinated stress adaptation along the brain-heart axis.³ Recent studies suggest that prolonged hypoxia may induce a gradual transition from HIF-1α-dominated to HIF-2α-dominated signaling, a phenomenon often referred to as the “HIF switch”.²⁹

This transition may reflect several layers of biological specialization.^27,29 HIF-1α is generally associated with rapid adaptation to severe oxygen deprivation and preferentially regulates glycolytic metabolism, acute stress responses, and inflammatory signaling.^2,12,38 In contrast, HIF-2α has been proposed to contribute to endothelial homeostasis, angiogenesis, vascular remodeling, and longer-term adaptation to persistent hypoxic stress.^22,29-31 Cell-type-specific expression patterns may further contribute to this distinction, as HIF-2α signaling is enriched within endothelial and perivascular compartments that play central roles in maintaining neurovascular integrity.^24,25,31 Although these observations do not imply a strict temporal separation between HIF isoforms, they provide a mechanistic framework for understanding why acute ischemic injury and chronic hypoperfusion may recruit partially distinct hypoxia-responsive programs.^27,29,38

Within the brain-heart axis, these distinctions may be particularly relevant because acute ischemic stroke is characterized by abrupt and severe oxygen deprivation, whereas chronic cerebral hypoperfusion and cardiovascular insufficiency expose tissues to prolonged, lower-grade oxygen-energy stress.^29,33,38 Such differences in stress duration and intensity may contribute to preferential engagement of distinct HIF-associated adaptive programs across disease contexts. Importantly, direct comparative evidence examining HIF-1α and HIF-2α dynamics within the adult brain-heart axis remains limited, and much of the current framework derives from vascular, developmental, and cancer biology studies.^27,29

2.2. HIF-driven metabolic reprogramming and mitochondrial restraint
A major downstream consequence of HIF signaling, driven predominantly by HIF-1α, is the reprogramming of cellular energy metabolism under conditions of limited oxygen availability.^2,12,22 HIF directly induces genes involved in glucose uptake and glycolysis, thereby supporting ATP production when oxidative phosphorylation becomes constrained.^2,18

Importantly, HIF does not simply enhance glycolytic flux; it also limits mitochondrial oxidative burden. One key mechanism is the induction of pyruvate dehydrogenase kinase 1 (PDK1), which inhibits pyruvate dehydrogenase and reduces the conversion of pyruvate to acetyl-CoA, thereby decreasing tricarboxylic acid (TCA) cycle flux.¹⁸ This coordinated shift lowers mitochondrial oxygen consumption, reduces electron transport chain over-reduction, and suppresses mitochondrial reactive oxygen species (ROS) generation.³⁹ In this way, HIF promotes a metabolic state that favors short-term bioenergetic maintenance while limiting oxidative injury.^2,18,22

In highly oxidative tissues such as the brain and heart, this metabolic switch carries important pathophysiological consequences. During acute cerebral ischemia, HIF-mediated glycolytic support and mitochondrial restraint may help preserve ATP availability and delay bioenergetic collapse in metabolically salvageable regions.^33,40,41 In cardiomyocytes, comparable programs can reduce mitochondrial stress during ischemia-reperfusion or catecholamine-driven increases in workload.³⁷

However, this form of metabolic plasticity is inherently time dependent, and accumulating evidence suggests that prolonged hypoxic exposure may progressively shift cellular responses from HIF-1α-dominated metabolic adaptation toward alternative HIF-dependent programs.^27,29 When sustained chronically, as may occur in persistent hypoperfusion or heart failure, prolonged engagement of HIF-associated metabolic programs may progressively alter redox balance, substrate preference, and inflammatory sensitivity, thereby compromising long-term tissue resilience.^4,38 Clinical and experimental studies further suggest that chronic cardiac dysfunction may promote similar metabolic stress responses through sustained reductions in cerebral perfusion and oxygen delivery.^4,16 This temporal dependence provides an important conceptual basis for understanding why HIF-linked metabolic adaptation may be protective in acute injury yet become maladaptive in chronic neurodegenerative and cardiometabolic settings.^29,38

Importantly, the consequences of this metabolic shift are likely to differ between acute ischemic injury and chronic hypoperfusion.^2,29 In acute stroke, transient enhancement of glycolytic capacity may help sustain ATP production within metabolically salvageable tissue and delay bioenergetic failure.^33,41 In contrast, persistent reliance on glycolytic programs during chronic neurovascular insufficiency may contribute to progressive metabolic inefficiency, redox imbalance, and tissue dysfunction.^4,38,42 These observations support the concept that HIF-mediated metabolic adaptation is highly dependent on the duration and severity of oxygen-energy stress and may transition from protective to maladaptive as hypoxic exposure becomes prolonged.^22,29

2.3. Mitochondrial quality control and inflammatory-vascular coupling
Beyond its effects on metabolic flux, HIF also regulates mitochondrial quality control and coordinates downstream stress signals. Under hypoxic conditions, stabilized HIF-1α can induce BCL2 interacting protein 3 (BNIP3) and related mitophagy-associated pathways, thereby contributing to mitochondrial quality control and limiting ROS accumulation.^22,38,42,43 In metabolically stressed neurons and cardiomyocytes, this HIF-BNIP3 axis may transiently preserve bioenergetic stability by reducing mitochondrial injury burden.³⁷

However, the effects of BNIP3-dependent remodeling are strongly context dependent. When sustained or excessively activated, this pathway can contribute to mitochondrial permeability transition, energetic failure, and cell death, particularly in injured myocardium.⁴⁴ HIF-mediated mitochondrial remodeling therefore spans a continuum from adaptive quality control to maladaptive injury, with outcome shaped largely by the duration and severity of stress exposure.⁴⁴

In parallel, HIF extends mitochondrial stress responses to the tissue level through vascular and immune pathways. HIF-dependent induction of vascular endothelial growth factor (VEGF) can support angiogenic adaptation, but may also increase vascular permeability under pathological conditions.⁴⁵ In myeloid-lineage cells, HIF promotes aerobic glycolysis and pro-inflammatory cytokine production, linking oxygen sensing directly to innate immune activation.^12,46 Through these combined mitochondrial, vascular, and immunometabolic effects, HIF functions as an integrative stress node capable of coordinating multicellular responses across tissues.^3,46 Within the neurovascular unit, these pathways influence interactions among endothelial cells, pericytes, astrocytes, microglia, and neurons, thereby shaping tissue-level responses to oxygen-energy stress.^24,25 Within the neurovascular unit and myocardium, mitochondrial quality-control pathways, endothelial responses, and immunometabolic signaling networks are tightly interconnected and may collectively determine whether HIF activation promotes adaptive resilience or pathological remodeling under sustained stress.^1,12,24,47 This integrated response is particularly relevant to the brain-heart axis, where vascular instability, metabolic dysfunction, and chronic inflammation frequently coexist and amplify one another across organ systems.^9,25

2.4. Context dependence across the brain-heart axis
Taken together, these mechanisms indicate that the biological consequences of HIF signaling are highly dependent on the intensity, duration, and spatial distribution of oxygen-energy stress. Acute ischemic injury is characterized by abrupt and severe reductions in oxygen and nutrient availability, whereas chronic neurovascular and cardiovascular disorders more commonly expose tissues to persistent, lower-grade metabolic stress.^4,33,42 Consequently, identical hypoxia-responsive pathways may exert distinct effects depending on the temporal context in which they are activated.^22,29

Emerging single-cell and spatial transcriptomic studies further suggest that hypoxia-responsive signaling varies substantially across neurons, astrocytes, endothelial cells, pericytes, microglia, and cardiomyocytes.^23-25 Differences in baseline metabolic demand, mitochondrial density, vascular dependence, and inflammatory sensitivity may contribute to marked heterogeneity in HIF-mediated responses.^24,25 These findings challenge the notion of a uniform hypoxic response and instead support a cell-type-specific model of stress adaptation across the brain-heart axis.^23,24

Accordingly, HIF activation should not be viewed as uniformly protective or detrimental. Rather, its effects depend on the biological setting in which it is engaged.^26,29 In acute ischemic stroke, HIF-dependent metabolic and survival programs may help preserve tissue viability within metabolically salvageable regions.^33,41 In contrast, under conditions of chronic cerebral hypoperfusion, heart failure, or sustained neurovascular dysfunction, prolonged engagement of HIF-associated pathways may contribute to vascular remodeling, metabolic imbalance, and progressive tissue injury.^4,38,42 This context-dependent framework provides the conceptual basis for the disease-specific mechanisms discussed in the following sections.⁹ For clarity, the principal differences between acute ischemic stroke and chronic AD-associated hypoperfusion across the brain-heart axis are summarized in Table 1.

3. HIF in Ischemic Stroke and Neurogenic Cardiac Stress Across the Brain-Heart Axis
Ischemic stroke represents a prototypical state of acute oxygen and energy deprivation in the brain, most commonly caused by thromboembolic occlusion of a cerebral artery that abruptly interrupts regional blood flow. The resulting perfusion failure rapidly triggers ATP depletion, loss of ionic homeostasis, excitotoxic glutamate release, mitochondrial depolarization, and oxidative stress.^32,33 This ischemic landscape is spatially heterogeneous, consisting of an irreversibly injured core surrounded by a metabolically compromised but potentially salvageable penumbra.³⁶ As illustrated in Fig.1, acute ischemic stroke is characterized by a rapid, high-amplitude stress response in which HIF-1α-associated programs promote glycolytic adaptation and metabolic compensation while simultaneously contributing to neurovascular and systemic responses that influence cardiac function.

Within this acute setting, HIF-1α emerges as the dominant hypoxia-responsive regulator engaged during the early ischemic phase.^40,48,49 Consistent with its rapid stabilization kinetics, HIF-1α links abrupt oxygen deprivation to transcriptional programs that support metabolic adaptation, vascular responses, and inflammatory signaling.^2,49 These responses may help preserve cellular function under conditions of severe energetic stress, but their impact is strongly shaped by timing, cellular context, and the extent of tissue injury.

Importantly, the consequences of stroke-induced hypoxic stress are not confined to the brain. Acute cerebral ischemia can trigger systemic responses that alter cardiac metabolism and function through autonomic dysregulation, catecholamine excess, and neurogenic myocardial stress.^50-52 Ischemic stroke therefore provides a useful framework for examining how HIF-linked stress adaptation operates not only within the injured brain, but also across the broader brain–heart axis.
3.1. Early cerebral HIF-1 activation within the ischemic metabolic window

HIF-1α is rapidly stabilized in neurons and vascular-associated cells within minutes of cerebral arterial occlusion, reflecting the abrupt decline in tissue oxygen tension that defines the early ischemic metabolic window.^40,48 This acute loss of oxygen availability occurs in parallel with ATP depletion, ionic imbalance, excitotoxic glutamate release, mitochondrial dysfunction, and oxidative stress, thereby creating a state of severe bioenergetic instability.^32,33 In this setting, early HIF-1α accumulation serves as a sensitive molecular response to the sudden interruption of oxidative metabolism.⁴⁸

Once stabilized, HIF-1α initiates a transcriptional program that supports short-term metabolic adaptation under conditions of impaired mitochondrial respiration. Key downstream responses include increased glucose uptake, enhanced glycolytic flux, and induction of PDK1, which limits pyruvate entry into mitochondria and reduces tricarboxylic acid (TCA) cycle activity.^2,18,39 Through this coordinated shift, HIF-1α helps reduce mitochondrial oxidative burden, limit ROS accumulation, and transiently preserve ATP homeostasis.¹⁸

These effects are especially relevant within the ischemic penumbra, a metabolically compromised but potentially salvageable region in which residual perfusion persists despite impaired mitochondrial efficiency.³⁶ In this context, HIF-1α -associated metabolic buffering may delay bioenergetic collapse, extend the window for tissue rescue, and improve neuronal tolerance to acute hypoxic stress. Consistent with this protective model, neuron-specific deletion of HIF-1α worsens infarct injury in middle cerebral artery occlusion (MCAO) models, whereas genetic or pharmacologic stabilization of HIF can improve neuronal survival under acute hypoxic conditions.^40,53

In parallel with its metabolic effects, HIF-1α also induces VEGF and related angiogenic mediators.^45,54 This response is functionally biphasic: although it may support later vascular remodeling and perfusion recovery, early VEGF upregulation can increase vascular permeability and contribute to edema during reperfusion.⁵⁵ Together, these findings support the view that HIF1α exerts its most beneficial effects within a narrow early window, during which metabolic stabilization predominates before broader vascular and inflammatory liabilities become more pronounced.

Importantly, HIF activation may differ substantially between the ischemic core and surrounding penumbra. Within the ischemic core, profound reductions in oxygen and nutrient availability frequently exceed the capacity of adaptive responses, leading to rapid bioenergetic collapse, irreversible cellular injury, and tissue necrosis despite HIF stabilization.^33,56 In contrast, the penumbra retains partial perfusion and residual metabolic activity, creating conditions in which HIF-dependent transcriptional programs may more effectively support cellular survival, angiogenesis, metabolic adaptation, and tissue recovery.^41,56 Consequently, differences in both the magnitude and biological consequences of HIF activation across these regions may contribute to the context-dependent protective and maladaptive effects attributed to HIF signaling following stroke. However, direct comparative studies examining HIF isoform dynamics within the ischemic core and penumbra remain limited, and further investigation is required to determine whether regional differences in HIF-1α and HIF-2α engagement contribute to divergent outcomes after cerebral ischemia.^29,38

3.2. Neurovascular and immunometabolic amplification
The progression of ischemic injury is tightly coupled to dysfunction of the neurovascular unit (NVU), a multicellular interface composed of endothelial cells, pericytes, astrocytes, and neurons that collectively regulate cerebral perfusion, metabolic exchange, and barrier integrity.³³ In endothelial and perivascular compartments, HIF-1-dependent signaling contributes to blood-brain barrier (BBB) instability during ischemia-reperfusion. Although some of these responses may support later vascular adaptation, early activation of permeability-associated programs can disrupt endothelial junctional organization and increase BBB leakiness.^54,55 Cell-type-specific studies further highlight the importance of perivascular regulation, as pericyte-targeted deletion of HIF1α preserves capillary coverage, maintains junctional integrity, and reduces BBB disruption following ischemia-reperfusion.⁵⁷ These findings identify perivascular HIF-1 signaling as a critical determinant of whether NVU remodeling remains adaptive or shifts toward barrier fragility.⁵⁷

Astrocytes and myeloid-lineage cells provide an additional layer of amplification. In astrocytes, HIF-linked glycolytic reprogramming supports lactate production and metabolic coupling within the hypoxic penumbra,⁵⁸ but sustained glycolytic bias may also alter the extracellular milieu and influence inflammatory thresholds.⁵⁹ Recent single-cell analyses further suggest substantial heterogeneity in hypoxia-responsive signaling across endothelial, glial, immune, and neuronal populations, highlighting the importance of cell-specific responses within the ischemic neurovascular unit.^23,24 In microglia and infiltrating myeloid cells, HIF-1α promotes a shift toward aerobic glycolysis that supports production of pro-inflammatory mediators, including IL1β, IL-6, TNF-α, and nitric oxide.^12,46 Recent work further links HIF-1 signaling to oxidative stress amplification and NLRP3 inflammasome activation in post-ischemic microglia,¹ providing a direct mechanistic connection between hypoxia and innate immune escalation.

Once barrier integrity is compromised, ischemic injury is no longer confined to a local metabolic lesion.³³ BBB disruption permits peripheral immune cell entry, plasma protein extravasation, and broader cytokine exchange, thereby coupling focal cerebral hypoxia to systemic inflammatory and autonomic stress responses.⁵⁷ Through this NVU-centered cascade, acute HIF-1-dominant signaling acquires the capacity to amplify tissue injury locally while also initiating broader stress propagation across the brain-heart axis.⁵²

3.3. Stroke-to-heart metabolic propagation
Acute ischemic stroke can disrupt central autonomic regulation, particularly in regions such as the insular cortex and brainstem, thereby triggering a pronounced sympathetic surge.⁶⁰ The resulting catecholamine excess increases heart rate, contractility, and β-adrenergic signaling, sharply elevating cardiomyocyte ATP demand and mitochondrial workload.⁵⁰ At the cellular level, βadrenergic stimulation enhances L-type Ca²⁺ influx and sarcoplasmic reticulum Ca²⁺ cycling, creating a state of sustained calcium turnover and energetic strain.⁵⁹

Excess sympathetic drive promotes both cytosolic and mitochondrial Ca²⁺ overload, accelerates electron transport chain activity, and increases ROS generation, features that resemble catecholamine-mediated myocardial stunning and stress cardiomyopathy.⁵⁹ Rising mitochondrial Ca²⁺ and oxidative stress together lower the threshold for mitochondrial permeability transition pore (mPTP) opening, leading to loss of mitochondrial membrane potential, impaired oxidative phosphorylation, and ATP depletion^[1]. In parallel, stroke-associated systemic inflammation, including circulating mediators such as IL-6 and TNF-α, may further disrupt myocardial substrate utilization, endothelial function, and redox buffering capacity, thereby compounding mitochondrial vulnerability.^51,52,61

Despite their distinct upstream triggers, these neurogenic stress signals converge on canonical mitochondrial failure pathways that closely parallel classical myocardial ischemia-reperfusion injury.⁶² Within this stressed myocardial environment, HIF-1α is well positioned to translate neurogenically driven energetic and redox imbalance into coordinated adaptive programs.^18,37

HIF-1 activation promotes a shift away from oxidative phosphorylation toward glycolysis, thereby reducing mitochondrial oxygen consumption and limiting ROS production.^18,39 A central effector of this response is HIF-1-dependent induction of PDK1, which restricts pyruvate entry into the tricarboxylic acid cycle and lowers oxidative pressure while helping preserve ATP availability.¹⁸

Importantly, stroke does not necessarily produce primary myocardial hypoxia. Rather, sympathetic overdrive creates a functional hypoxia-like state characterized by heightened ATP turnover, redox imbalance, and mitochondrial strain. Under these conditions, secondary engagement of HIF-1 signaling may support cardiomyocyte stress tolerance and mitochondrial resilience, although the durability of this response is likely to depend on the intensity and duration of systemic stress exposure.³⁷

3.4. PKM2 as a Potential Amplifier of HIF-Dependent Plasticity Following Stroke
Building on the neurogenic cardiomyocyte stress outlined above, metabolic rewiring in the poststroke myocardium is characterized by a rapid shift toward energy conservation and adaptive bioenergetic compensation.^63,64 In the acute phase, this transition is accompanied by suppression of oxidative metabolism, reduced mitochondrial oxygen consumption, and increased reliance on glycolytic pathways to support ATP production under conditions of limited oxygen availability.^2,18 Through coordinated regulation of glucose metabolism and mitochondrial activity, HIF-1αmediated signaling may help preserve cellular energetics and delay bioenergetic collapse within metabolically vulnerable tissues.^12,33

Under these conditions, pyruvate kinase M2 (PKM2) has emerged as a potential mediator linking metabolic adaptation to transcriptional regulation. Beyond its canonical role as a glycolytic enzyme, PKM2 can translocate to the nucleus and function as a coactivator of HIF-1α, thereby amplifying hypoxia-responsive gene expression and reinforcing glycolytic reprogramming.^12,65 Through this reciprocal interaction, the HIF-1α-PKM2 axis may contribute to short-term maintenance of cellular energetics and adaptive stress responses during acute oxygen deprivation.

Recent studies have further implicated PKM2 in tissue repair, regeneration, and stress adaptation. In cardiomyocytes, PKM2 activity has been associated with enhanced proliferative potential and regenerative responses following injury, while in the central nervous system PKM2-mediated signaling has been linked to neuronal plasticity, metabolic adaptation, and recovery following ischemic stress.^65,66 These observations raise the possibility that PKM2dependent pathways may participate in adaptive responses that are shared across metabolically vulnerable tissues.

However, direct evidence supporting a role for PKM2 in SHS remains limited. Most mechanistic studies linking PKM2 to cellular plasticity derive from cancer biology, regenerative medicine, or experimental central nervous system models rather than investigations specifically examining secondary cardiac dysfunction following cerebral ischemia.^27,67 Consequently, extrapolation of PKM2-dependent mechanisms to post-stroke myocardial adaptation should be interpreted cautiously.

Although the majority of current evidence relates to acute ischemic injury, PKM2-dependent metabolic signaling may also have relevance in chronic neurovascular disorders. Persistent alterations in glucose metabolism, mitochondrial function, and HIF-responsive pathways have been implicated in chronic cerebral hypoperfusion and AD-related neurodegeneration.^38,42 Whether PKM2 contributes to these long-term processes through mechanisms analogous to those observed during acute ischemic adaptation remains uncertain. Nevertheless, the possibility that PKM2 participates in both acute and chronic oxygen-energy stress responses highlights its potential relevance across the broader brain-heart axis framework.

At present, PKM2 should therefore be viewed as a candidate mediator rather than an established regulator of brain-heart communication. Its potential significance lies in its ability to couple metabolic state to transcriptional adaptation through interactions with HIF signaling and related stress-response pathways.^12,65 Whether PKM2 functions as a true cross-organ integrator of stroke-induced neurocardiac remodeling or instead represents a shared downstream response to hypoxic stress remains unresolved.

Several important questions warrant further investigation. It remains unclear whether PKM2 activation differs quantitatively or qualitatively between neural and cardiac tissues following stroke, whether PKM2 contributes directly to secondary myocardial dysfunction, and whether modulation of PKM2 signaling influences long-term neurological and cardiovascular outcomes. Future studies employing tissue-specific and temporally controlled approaches will be required to determine whether PKM2 represents a viable therapeutic target within the brain-heart axis.

3.5. Integrated model of acute brain–heart coupling
Taken together, ischemic stroke can be understood as a systems-level metabolic insult in which acute cerebral hypoxia initiates a predominantly HIF-1α -associated response that extends beyond the injured brain. Within the ischemic neurovascular unit, early HIF-1 activation supports short-term metabolic adaptation, but concurrent barrier instability and immunometabolic amplification also create conditions for broader systemic stress propagation.^40,57

Through autonomic and inflammatory pathways, this acute cerebral insult imposes a secondary energetic burden on the heart, where sympathetic overactivation, calcium dysregulation, and mitochondrial stress converge on canonical cardiomyocyte injury mechanisms.^51,52 In this setting, HIF-1-linked metabolic plasticity, reinforced in part by PKM2-dependent transcriptional amplification, may transiently support myocardial stress tolerance by limiting oxidative burden and preserving metabolic plasticity.^37,65

Several factors are likely to influence whether adaptive stress responses remain protective or progress toward secondary cardiac dysfunction. These include infarct size, lesion location, involvement of autonomic regulatory regions such as the insular cortex, baseline cardiovascular reserve, pre-existing cardiac disease, and the magnitude and duration of sympathetic activation.^13,15,50,60 Variability across these factors may help explain why some patients develop clinically significant SHS whereas others exhibit only transient physiological adaptation.

The overall impact of this response is therefore strongly time dependent. When transient and well constrained, HIF-1α -associated adaptation may support acute resilience across the brain-heart axis. When excessive or prolonged, the same pathways are more likely to promote barrier fragility, inflammatory escalation, and cardiometabolic vulnerability. This acute brain-heart coupling framework provides an important contrast to the more chronic, low-grade stress landscape considered in AD, where sustained hypoperfusion and progressive neurovascular dysfunction are more consistent with increasing engagement of HIF-2α-associated programs.^26,31 This distinction forms the basis for the acute-versus-chronic oxygen-energy stress framework developed throughout this Review.

4. Chronic Heart-Brain Axis in Alzheimer’s Disease: HIF-Associated Responses as Components of Persistent Energetic Stress
AD is increasingly viewed as a disorder in which neuronal vulnerability reflects not only proteostatic failure but also sustained metabolic and vascular strain. Epidemiological and mechanistic studies have linked cardiovascular dysfunction, cerebral hypoperfusion, blood-brain barrier disruption, and neurovascular dysfunction to increased risk of cognitive decline and AD-related pathology.^4-7,16 Unlike ischemic stroke, where oxygen deprivation is abrupt and focal, AD more commonly develops under conditions of prolonged, low-grade energetic insufficiency driven by impaired neurovascular coupling, endothelial dysfunction, and progressive mitochondrial stress.^5,38,42 In contrast to the acute HIF-1α-associated response observed in stroke, Fig.1 illustrates a chronic low-grade stress state in which prolonged neurovascular insufficiency, vascular remodeling, and sustained hypoxia-responsive signaling may contribute to progressive brain-heart dysfunction.

Such conditions are less compatible with the rapid, high-amplitude HIF-1α response typical of acute ischemia and may instead favor more sustained vascular and homeostatic responses. These responses have been associated with HIF-2α-related signaling in vascular and perivascular compartments in experimental systems.^22,29-31 However, direct evidence demonstrating preferential HIF-2α engagement in human AD remains limited. Therefore, HIF-2α-associated pathways should be viewed as a proposed component of chronic neurovascular adaptation rather than an established driver of AD progression.^27,29,38

This intermediate hypoxia-responsive state may initially support metabolic compensation by enhancing glycolytic capacity and restraining mitochondrial oxygen demand.^2,18,22 Over time, however, sustained engagement is more likely to recalibrate redox balance, inflammatory thresholds, and neurovascular stability in ways that may progressively reduce tissue resilience.^38,42 Within the brain-heart axis framework, cardiovascular insufficiency may act as a persistent upstream contributor to this signaling state, while evolving neurovascular dysfunction further compromises cerebral perfusion.^4,16,17 Together, these interactions may contribute to a self-reinforcing cycle of energetic stress, inflammatory activation, and impaired neurovascular homeostasis.

Collectively, these features position HIF-associated signaling as one component of a broader chronic brain-heart stress network linking cardiovascular decline to progressive neurodegenerative vulnerability in AD.^8,9

4.1. Chronic hypoperfusion as a contributor of brain vulnerability
A central premise linking cardiovascular dysfunction to AD is that sustained reductions in cardiac output may contribute to chronic cerebral hypoperfusion. Numerous experimental and clinical studies have reported associations between reduced cerebral blood flow, vascular dysfunction, and increased risk of cognitive decline, although the extent to which hypoperfusion directly drives AD pathology remains an area of ongoing investigation.^4,16,68 Rather than acting as an isolated pathogenic mechanism, chronic hypoperfusion is increasingly viewed as one component of a broader neurovascular dysfunction framework that may interact with aging, vascular risk factors, and established AD-related processes.^5,6,38,42 In this context, chronic hypoperfusion is best viewed as one contributor to a broader state of oxygen-energy stress rather than its sole determinant.^2,3,9,38 Within this Review, chronic hypoperfusion is considered one important contributor to a broader state of oxygen-energy stress, which may also arise from mitochondrial dysfunction, metabolic insufficiency, inflammatory activation, and cardiovascular impairment.^2,3,9,38

At the cellular level, this oxygen-energy stress may arise when reduced perfusion limits delivery of oxygen and glucose, thereby constraining oxidative ATP production and increasing reliance on compensatory metabolic pathways.^2,18,38,69 Persistent reductions in oxygen availability may contribute to mitochondrial dysfunction, oxidative stress, and impaired neurovascular coupling, changes that have been associated with cognitive decline in both experimental and clinical studies.^38,42 Over time, chronic disturbances in cerebral perfusion may impair communication among neurons, astrocytes, endothelial cells, and perivascular compartments, thereby reducing the capacity of the neurovascular unit to maintain metabolic homeostasis.^24,25

Importantly, support for this framework is not derived solely from experimental models. Longitudinal cohort studies have reported associations between reduced cardiac function, lower cerebral blood flow, and increased risk of cognitive impairment and dementia.^16,68 Interventional evidence from the SPRINT-MIND trial further supports the clinical relevance of vascular risk control for cognitive outcomes. In this randomized clinical trial, intensive systolic blood pressure control did not significantly reduce probable dementia alone, but it reduced the incidence of mild cognitive impairment and the combined outcome of mild cognitive impairment or probable dementia.⁷⁰ Similarly, heart failure patients frequently exhibit reduced cerebral perfusion and accelerated cognitive decline, suggesting that chronic cardiovascular insufficiency may influence long-term brain health through vascular and metabolic mechanisms.^4,17 However, these findings remain largely associative and do not establish a direct causal pathway between cardiovascular dysfunction and AD pathology.

Within this context, chronic hypoperfusion may be viewed as a persistent oxygen-energy stressor capable of engaging adaptive hypoxia-responsive signaling pathways. Unlike acute ischemic stroke, where severe oxygen deprivation often favors rapid HIF-1α activation, chronic reductions in oxygen availability may create conditions more compatible with sustained vascular and homeostatic responses that have been associated with HIF-2α signaling.^22,29 Although direct evidence linking HIF-2α activity to human AD progression remains limited, this conceptual framework provides a useful basis for examining how chronic neurovascular stress may contribute to progressive brain vulnerability.

4.2. Chronic HIF engagement links metabolic adaptation to amyloid and tau pathology
Under conditions of sustained perfusion limitation, hypoxia-responsive signaling may support a compensatory metabolic response aimed at preserving cellular function despite prolonged energetic stress. Compared with the abrupt oxygen deprivation characteristic of acute ischemic stroke, chronic hypoperfusion generally imposes a lower-grade but more persistent oxygen-energy challenge, creating conditions in which prolonged HIF-dependent adaptation may become increasingly relevant.^22,29 Within this setting, HIF signaling has been proposed to influence glucose metabolism, mitochondrial homeostasis, vascular remodeling, and cellular stress responses, although the precise contributions of individual HIF isoforms remain incompletely defined.^24,25

Over time, persistent engagement of hypoxia-responsive pathways may alter the neurovascular environment in ways that extend beyond immediate metabolic compensation. Experimental studies have reported associations between chronic hypoperfusion, sustained HIF activity, glial activation, endothelial dysfunction, and progressive neurovascular remodeling.^38,42 However, the extent to which these changes directly drive AD pathology remains uncertain, as many observations derive from preclinical models and causal relationships have not been fully established in humans.

One proposed mechanism involves interactions between HIF signaling and amyloidogenic processing pathways. Several experimental studies have suggested that hypoxia may increase expression of β-site amyloid precursor protein (APP)-cleaving enzyme 1 (BACE1) and promote amyloidogenic APP processing, potentially contributing to increased amyloid-β production.^55,71 HIF-1α has been implicated in some of these responses, although the relative contributions of HIF-dependent and HIF-independent mechanisms remain debated.³⁸ Importantly, direct evidence demonstrating that HIF activity is a primary driver of amyloid pathology in human AD remains limited.

Beyond amyloid biology, chronic oxygen-energy stress may also influence tau-related pathways through mechanisms involving oxidative stress, mitochondrial dysfunction, neuroinflammation, and impaired proteostatic regulation.^5,42,72 Persistent disturbances in cellular homeostasis may increase susceptibility to protein misfolding and aggregation, thereby contributing to progressive neurodegenerative vulnerability. Nevertheless, current evidence generally supports a model in which chronic hypoperfusion and HIF-associated signaling act as potential modifiers of disease progression rather than as singular causal determinants of AD pathology.

Emerging single-cell and spatial transcriptomic studies further suggest that hypoxia-responsive signaling varies substantially across endothelial cells, pericytes, astrocytes, microglia, and neurons.^23-25 Such cellular heterogeneity may help explain why prolonged hypoxic stress produces diverse outcomes across neurovascular compartments and highlights the importance of considering cell-type-specific responses when interpreting HIF-associated mechanisms in AD.

4.3. Neurovascular-immune amplification and EV-mediated heart-to-brain communication

NVU may serve as a critical interface through which chronic cardiovascular dysfunction influences brain vulnerability. Persistent reductions in cerebral perfusion, endothelial dysfunction, and low-grade inflammation can progressively compromise BBB integrity and impair neurovascular coupling.^5,38 Rather than representing isolated pathological events, these changes may interact to reduce metabolic resilience and increase susceptibility to neurodegenerative processes.

As neurovascular integrity declines, inflammatory amplification may become increasingly important.⁵ Circulating mediators associated with chronic cardiovascular disease, including IL6, TNF-α, and other inflammatory signals, have been linked to endothelial dysfunction, glial activation, and altered neurovascular homeostasis.^4,17 Within the central nervous system, persistent inflammatory signaling may influence microglial activation states, astrocytic responses, and vascular remodeling, thereby creating conditions that favor progressive tissue vulnerability rather than efficient stress adaptation.

Beyond soluble inflammatory mediators, recent experimental studies have raised the possibility that extracellular vesicles (EVs) may participate in communication between the injured heart and brain. In preclinical models, EV cargo derived from stressed cardiomyocytes has been shown to influence endothelial function, inflammatory signaling, and microglial activation following systemic cardiovascular injury.⁴¹ These findings suggest that EVs may represent one mechanism through which cardiovascular stress signals are transmitted to the central nervous system.

However, current evidence supporting EV-mediated heart-to-brain communication remains largely preclinical. Direct evidence demonstrating that cardiomyocyte-derived EVs drive neurodegenerative progression in humans is currently lacking. Accordingly, EV signaling should be viewed as an emerging and biologically plausible mechanism rather than an established pathway linking cardiovascular dysfunction to AD.⁴¹

Taken together, chronic cardiovascular dysfunction may influence brain vulnerability through multiple interacting pathways involving impaired cerebral perfusion, neurovascular dysfunction, inflammation, and potentially EV-mediated signaling. Rather than acting as independent mechanisms, these processes likely converge to shape the broader oxygen-energy stress environment within the aging brain. Future studies integrating human clinical cohorts, biomarker analyses, and mechanistic investigations will be required to determine the relative contribution of each pathway to neurodegenerative progression.

4.4. Integrated model of chronic brain–heart coupling
Collectively, cardiovascular dysfunction may influence brain vulnerability through interconnected effects on cerebral perfusion, neurovascular integrity, metabolic homeostasis, and inflammatory signaling. Rather than representing independent pathological events, chronic reductions in cardiac function, impaired cerebral blood flow, endothelial dysfunction, and low-grade inflammation may interact to shape a persistent state of oxygen-energy stress within the aging brain.^4,16,17 These processes have been associated with increased risk of cognitive decline and dementia, although the precise causal relationships remain incompletely understood.^16,17,73,74 Within this framework, hypoxia-responsive signaling may act as one component of a broader adaptive response to prolonged metabolic stress. Compared with the abrupt oxygen deprivation characteristic of acute ischemic stroke, chronic cardiovascular insufficiency generally produces lower-grade but more sustained disturbances in oxygen and nutrient delivery. Such conditions may favor prolonged vascular and homeostatic adaptation, processes that have been associated with HIF-2α-related signaling in experimental systems.^22,29 However, direct evidence demonstrating preferential HIF2α engagement in human AD remains limited, and the relative contributions of HIF-1α and HIF2α under chronic disease conditions require further investigation.

In the early stages of chronic hypoperfusion, adaptive responses involving metabolic compensation, vascular remodeling, and neurovascular regulation may help preserve tissue function despite reduced energetic reserve.³⁸ Over time, however, persistent exposure to metabolic stress may be associated with progressive impairment of neurovascular coupling, mitochondrial homeostasis, inflammatory regulation, and proteostatic balance.^24,42 These changes may increase susceptibility to neurodegenerative processes rather than acting as direct determinants of disease.

Accordingly, chronic brain–heart coupling may be best viewed as a progressive systems-level interaction in which cardiovascular dysfunction contributes to a biological environment that is permissive for neurodegenerative vulnerability. Within this model, chronic hypoperfusion, inflammation, neurovascular dysfunction, and hypoxia-responsive signaling are likely to interact with aging, genetic risk factors, and established AD-related pathology. Rather than serving as singular causal drivers, these mechanisms may function as modifiers that influence disease trajectory and resilience across the brain-heart axis. To complement the mechanistic framework discussed above, representative human clinical evidence linking cardiovascular dysfunction, cerebral hypoperfusion, cognitive decline, and stroke-associated cardiac injury is summarized in Table 2.

5. Therapeutic Implications: Isoform-Aware Targeting of HIF and Metabolic Control Across the Brain-Heart Axis
The evidence reviewed above indicates that HIF signaling cannot be approached as a uniformly beneficial or uniformly harmful pathway. Rather, its biological effects depend on isoform, timing, cellular context, and disease state, with distinct consequences across acute ischemic injury and chronic neurovascular degeneration.^3,26,30,31 This context dependence has direct therapeutic implications. Strategies aimed at modulating HIF-linked pathways will likely require greater temporal precision, improved distinction between HIF-1α and HIF-2α outputs, and closer attention to cell-type specificity within the neurovascular and cardiometabolic compartments.^{12,26,30,31,57} More broadly, effective intervention may depend on treating hypoxia-responsive signaling not as a brain-restricted or heart-restricted process, but as part of an integrated brain-heart stress network.⁵

5.1. Why broad HIF modulation is insufficient
Because HIF coordinates adaptive responses to oxygen and energetic stress, pharmacological manipulation of this pathway has attracted substantial translational interest. In acute ischemic stroke, transient enhancement of HIF-1α-dominant programs may support ATP preservation,⁴⁹ reduce mitochondrial ROS burden, and improve survival within the ischemic penumbra.^39,74 Comparable protective effects have also been described in settings of myocardial ischemia and neurogenic cardiac stress.³⁷

The clinical feasibility of targeting this pathway is supported by the approval of prolyl hydroxylase (PHD) inhibitors for anemia, demonstrating that pharmacological HIF stabilization is achievable in humans.⁷⁵ However, the therapeutic logic becomes considerably more complex in chronic disease states. In AD-relevant contexts, persistent HIF engagement has been proposed to influence pathways associated with amyloidogenic processing, tau dysregulation, and neurovascular fragility.^38,55,72 These observations suggest that the major translational challenge is not whether HIF should be activated or inhibited in general, but how to modulate its activity with appropriate timing, amplitude, and isoform selectivity.

Despite these encouraging observations, broad HIF activation presents important translational challenges. HIF regulates a diverse array of downstream targets involved in angiogenesis, vascular permeability, inflammation, erythropoiesis, and metabolic adaptation. Consequently, indiscriminate pathway activation may produce unintended effects that vary across tissues and disease states. Excessive VEGF induction has been associated with blood-brain barrier disruption, vasogenic edema, and hemorrhagic complications in experimental stroke models, highlighting the potential risks of global HIF stimulation.^5,56 Furthermore, many beneficial effects observed in preclinical studies have proven difficult to reproduce in human neurological disease,⁷⁵ in part because therapeutic efficacy appears highly dependent on timing, dosage, cell type, and disease context. These limitations suggest that broad HIF stabilization alone is unlikely to provide a universally effective therapeutic strategy.

5.2. Targeting downstream metabolic and mitochondrial effectors
Given the pleiotropic nature of HIF signaling, therapeutic strategies may achieve greater precision by targeting downstream metabolic and mitochondrial effectors rather than modulating the pathway at its upstream core.^3,26 One potential downstream target is the glycolysis-mitochondria balance regulated in part through HIF-1-dependent induction of PDK1, which shifts substrate utilization toward glycolysis and reduces mitochondrial oxidative pressure.¹⁸ Carefully tuning this axis has been proposed as a means of preserving ATP availability while limiting ROS generation, particularly in metabolically salvageable penumbral tissue and neurogenically stressed myocardium.³⁹ This therapeutic rationale is consistent with prior work showing that HIF-1regulated increases in glucose transport and glycolytic flux can support redox homeostasis, tissue viability, and neuroprotection in experimental ischemic stroke models.⁴⁹

A second opportunity lies in preserving mitochondrial stability. Because mitochondrial permeability transition, redox collapse, and bioenergetic failure represent shared endpoints across neural and cardiac injury, interventions that stabilize mitochondrial integrity, enhance quality control mechanisms such as mitophagy, or buffer oxidative stress may complement adaptive HIF-linked programs without requiring broad hypoxia-pathway activation.^37,43,76 In parallel, modulation of lactate-mediated metabolic coupling may offer an additional means of supporting tissue resilience. Astrocyte-neuron lactate shuttling is an important component of metabolic support after injury, and strategies that improve lactate transport or utilization could strengthen neuronal survival while avoiding the broader liabilities associated with persistent HIF activation.^58,77 Collectively, these downstream approaches aim to preserve the beneficial metabolic outputs associated with acute HIF-1-linked adaptation while minimizing the risks of prolonged, systemwide engagement of hypoxia-responsive signaling.^39,55

5.3. Cell-type-specific and brain-heart axis-directed strategies
Evidence from both acute stroke and chronic neurodegeneration indicates that therapeutic outcomes are shaped not only by neuronal survival, but also by non-neuronal compartments and inter-organ coupling.^12,51,52,57 This suggests that precision strategies directed toward specific cellular niches or axis-level drivers may outperform global manipulation of HIF signaling. Within the neurovascular unit, targeting endothelial and pericyte responses may be particularly important, as these compartments are central to perfusion adaptation and barrier stability, and may be especially relevant to HIF-2α-associated vascular homeostasis.^31,57 In parallel, glial immunometabolic control represents another promising therapeutic layer. Metabolic tuning of astrocytes may strengthen neuronal energetic support,^58,77 whereas modulation of microglial glycolytic reprogramming, much of which is linked to HIF-1α-dependent inflammatory activation, may help restrain persistent immune amplification.¹²

A further implication is that effective intervention may need to extend beyond the brain itself. The brain-heart axis is characterized by self-reinforcing stress propagation: acute stroke can provoke catecholamine-mediated myocardial injury,⁵⁰ whereas chronic cardiac dysfunction can sustain cerebral hypoperfusion, inflammatory signaling, and progressive neurovascular instability.⁷⁸ From this perspective, therapeutic strategies may need to operate at the systems level by attenuating sympathetic surges after acute brain injury, preserving cardiac output to maintain cerebral perfusion, reducing systemic inflammatory tone, and potentially interrupting EV mediated heart-to-brain communication.^41,79 Rather than acting directly on HIF alone, such approaches target the upstream oxygen- and energy-stress drivers that chronically engage hypoxia responsive programs across organs.^3,51,52

Importantly, many proposed brain-heart axis interventions remain at an early experimental stage. Whether modulation of autonomic signaling, systemic inflammation, metabolic coupling, or EV mediated communication can produce meaningful clinical benefit in humans remains uncertain. Translation of these approaches will require prospective clinical studies capable of distinguishing mechanistic plausibility from therapeutic efficacy.

5.4. Temporal precision as a therapeutic principle
A central therapeutic implication emerging from this framework is that HIF signaling must be modulated with close attention to timing. In acute ischemic injury, transient HIF-1α-associated activation may support short-term metabolic adaptation and tissue survival,^39,40,49 whereas under chronic AD-related hypoperfusion, sustained signaling with progressively greater HIF-2αassociated involvement is more likely to contribute to maladaptive neurovascular and metabolic remodeling.^26,31,80 This distinction helps explain why broad HIF activation can appear beneficial in acute injury models yet remain problematic in chronic neurodegenerative settings.^38,49

Nevertheless, substantial uncertainties remain regarding the feasibility of isoform-selective HIF targeting, the durability of therapeutic benefit, and the potential consequences of long-term pathway manipulation. Future studies will need to determine whether the conceptual distinction between acute HIF-1α-associated adaptation and chronic HIF-2α-associated remodeling can be translated into safe and clinically actionable interventions.

6. Conclusions
Stroke and AD can be viewed as acute and chronic manifestations of oxygen-energy stress within the nervous system. Although traditionally studied as distinct disorders, growing evidence suggests that both conditions are linked through a bidirectional brain-heart axis in which neurological injury and cardiovascular dysfunction interact through shared metabolic, vascular, inflammatory, and neurohumoral pathways. Acute cerebral ischemia can trigger secondary cardiac dysfunction through autonomic dysregulation and inflammatory signaling. In contrast, chronic cardiovascular insufficiency may impair cerebral perfusion, disrupt neurovascular homeostasis, and increase vulnerability to cognitive decline. Together, these observations support a systems-level view in which brain and heart health are closely interconnected rather than biologically isolated.

Within this framework, hypoxia-responsive signaling represents an important mechanism through which cells adapt to oxygen and energetic stress. Emerging evidence suggests that acute oxygen-energy stress may preferentially engage HIF-1α-associated programs where it supports short-term glycolytic compensation, limits mitochondrial burden, and helps preserve cellular viability. However, these responses are often temporally constrained, and excessive or prolonged activation may contribute to inflammation and maladaptive remodeling. In contrast, chronic low-grade hypoperfusion and vascular insufficiency may be associated with increased engagement of HIF-2α-associated programs, which have been linked to vascular and perivascular adaptation in experimental systems. In these settings, sustained hypoxia-responsive signaling may favor pathways linked to endothelial instability, vascular remodeling, and long-term adaptation to energetic stress. Nevertheless, direct evidence supporting preferential HIF-2α engagement within the adult brain-heart axis remains limited. Thus, while this temporal distinction may help explain why hypoxia-responsive pathways can appear protective in acute injury yet potentially maladaptive during chronic disease progression, further investigation is required to validate this framework.

A major theme emerging from this Review is that oxygen-energy stress should not be viewed solely as a neuronal phenomenon. Instead, it reflects a broader systems-level challenge involving coordinated responses across neural, vascular, immune, and cardiovascular compartments. Recent advances in single-cell and spatial transcriptomic technologies have begun to reveal substantial heterogeneity in hypoxia-responsive programs across endothelial cells, pericytes, astrocytes, microglia, neurons, and cardiomyocytes,^23-25 highlighting the importance of cellular context in determining adaptive versus maladaptive outcomes. At the same time, growing clinical evidence linking SHS, chronic cardiovascular dysfunction, cerebral hypoperfusion, and cognitive decline reinforces the translational relevance of this framework.^13,16,17,70

Although considerable progress has been made, important questions remain unresolved. The mechanisms governing communication between the brain and heart, including the contributions of autonomic dysregulation, systemic inflammatory signaling, and extracellular vesicle-mediated communication, remain incompletely understood. Likewise, the relative contributions of HIF-dependent and HIF-independent pathways across distinct cellular populations remain unclear, particularly within endothelial, perivascular, immune, and cardiomyocyte compartments. Future studies integrating experimental models, single-cell analyses, and human clinical cohorts will be essential for distinguishing association from causation, identifying clinically actionable targets, and determining whether the proposed distinction between acute HIF-1α-associated adaptation and chronic HIF-2α-associated remodeling represents a biologically meaningful framework across the brain-heart axis.

Ultimately, understanding neurological disease through the lens of the brain-heart axis may provide new opportunities to bridge traditionally separate fields of cerebrovascular, neurodegenerative, and cardiovascular research. By recognizing ischemic stroke and AD as interconnected manifestations of disrupted oxygen and energy homeostasis, future therapeutic strategies may therefore move beyond organ-specific interventions toward integrated approaches that preserve resilience across the entire neurocardiovascular system.