The brain is an energy-intensive organ requiring a constant supply of ATP to maintain the amount of ions across the neuronal membrane, depolarising neurones and support neurotransmission. Because the brain lacks large triglyceride storage and retains only small glycogen reserves, it relies on a constant input of blood-borne metabolites (14). The blood-brain barrier (BBB), a highly selective semipermeable membrane that permits the passive diffusion of small hydrophobic molecules like O_2, CO_2, and ethanol while shielding the neuronal environment from infections and big hydrophilic molecules, controls this influx (15).

Glucose is the brain's principal fuel and a necessary precursor for the synthesis of lipids, proteins, and neurotransmitters like acetylcholine, glutamate, and gamma-aminobutyric acid (GABA). Specific sodium-independent glucose transporters (GLUTs) facilitate transport across the BBB: GLUT1 is largely expressed in endothelial cells and astrocytes, whereas GLUT3 and GLUT4 are localized to neurones (16). Once inside astrocytes and neurones, glucose undergoes glycolysis, converting to pyruvate and producing ATP.

While glucose is the primary metabolic fuel, the brain may actively transport and use alternate substrates through specialized carrier proteins. These include glutamate, pyruvate, lactate, and ketones (17). This metabolic flexibility is critical during times of glucose scarcity or high functional demand.

To satisfy bioenergetic needs, cellular metabolism combines the breakdown of proteins, fats, and carbohydrates. These processes come together to form the tricarboxylic acid (TCA) cycle through glycolysis, fatty acid beta-oxidation, and amino acid transamination (18). Located within the mitochondrial matrix, the TCA cycle produces the reduced cofactors NADH and FADH_2, which serve as the primary electron donors for the electron transport chain (ETC).

The inner mitochondrial membrane (IMM) incorporates five protein complexes that collectively make up the ETC. Complex I (NADH: ubiquinone oxidoreductase) or Complex II (succinate dehydrogenase) carry electrons to the ETC. These electrons are then shuttled to ubiquinone (Q), reducing it to ubiquinol (QH_2). Four protons (H+) are injected into the intermembrane space (IMS) as electrons transit from Complex I to the Q cycle. Interestingly, Complex II has no role in the translocation of protons.

Complex III then oxidizes ubiquinol, facilitating the transport of electrons to cytochrome c. Four more protons are introduced into the IMS for each electron pair passing through this cycle (19). This sequential transfer establishes the electrochemical gradient necessary for ATP synthesis.

The astrocyte-neuron lactate shuttle (ANLS) hypothesis, the foundation of cerebral energy balance is a basic metabolic cooperation between astrocytes and neurones (20, 21). According to this framework, astrocytes serve as the primary site for glucose uptake and its subsequent conversion into lactate. This lactate is then tansported to neurons, where it is oxidized back into pyruvate to fuel the tricarboxylic acid (TCA) cycle for ATP production. It is a pathway that may offer higher energetic efficiency than direct glucose utilization.

Beyond its role as an energy substrate, lactate acts as a crucial signaling molecule in the brain. Recent studies have highlighted its numerous physiological functions, including the modulation of synaptic plasticity, memory consolidation, and neuroprotection (22). Consequently, disruptions in lactate homeostasis are linked to a number of neurological conditions, including as Alzheimer's disease (23), traumatic brain injury (24), and epilepsy (25).

The central nervous system (CNS) is the primary coordinator of physiological function, yet its high metabolic demand renders it uniquely vulnerable. Despite accounting for a small fraction of body mass, the brain consumes approximately 20% of the body's total glucose and oxygen—a rate ten times higher than most other tissues (26). This intense metabolic activity significantly increases the production of free radicals, especially reactive oxygen species (ROS). The brain uses energy from several sources, primarily glucose, to maintain the redox balance in the neuronal and glial cells in order to defend itself against oxidative damage by having an adequate antioxidant system (27).

ROS are continuously generated within brain cells via the mitochondrial electron transport chain and various enzymatic pathways under both physiological and pathological conditions. For instance, as part of normal signaling, monoamine oxidases (MAO-A and MAO-B) metabolize catecholamines, yielding hydrogen peroxide as a byproduct. Furthermore, NADPH oxidases can be pathologically activated by ischemia, seizures, protein misfolding, and overactivity of the glutamatergic system (28, 29).

The metabolic rate and signaling intensity can vary because different regions of the brain are specialized for distinct functions. This heterogeneity extends to the distribution of endogenous antioxidants, such as glutathione (GSH) (30, 31). Consequently, there may be variations in GSH levels between anatomical regions as well as between cell types. Crucially, physical exercise and toxic exposures can produce significant changes in the levels of antioxidants, particularly GSH, in various regions of the brain (32).

Mitochondria are central to the neuropathology of Alzheimer's disease (AD), regulating oxidative stress, intracellular Ca2+ signaling, and glutamate-mediated synaptic transmission (33). Disruptions in mitochondrial function trigger Ca2+ imbalances that drive synaptotoxicity and subsequent cognitive decline (34). Furthermore, impaired mitophagy and other mitochondrial quality control mechanisms are critical to the pathophysiology of AD (35).

In AD, the structural integrity and motility of mitochondria are severely altered. Evidence suggests that neuronal mitochondria in the AD brain exhibit fewer cristae and a condensed, darker matrix, characterized by increased fission and diminished fusion (36). These dysregulated dynamics—a common hallmark across neurodegenerative disorders, including Alzheimer‘s, Parkinson's, and Huntington's diseases, lead to mitochondrial fragmentation, reduced bioenergetic efficiency, and elevated oxidative stress (37).

The association of beta-Amyloid with mitochondrial components induces profound ultrastructural changes and oxidative damage, often mediated by disruptions in AMP-activated protein kinase (AMPK) pathways (38, 39). These strictly regulated processes of fusion and fission, collectively termed mitochondrial dynamics, are essential for maintaining optimal organelle distribution and cellular function (40). When these mechanisms fail, the resulting energy deficits and neuronal injury accelerate the progression of AD (41) (Figure 1).

Transgenic AD mouse models, including 5xFAD, FAD, and APP, have been shown to exhibit mitophagy, Aβ buildup in synaptosomal mitochondria, disruptions in mitochondrial dynamics indicators, and structural alterations (42). Studies focusing on the dysregulation of important mitochondrial fission and fusion proteins, including OPA1, MFN2, and DRP1, have demonstrated notable disturbances in mitochondrial dynamics that may aid in the advancement of disease. However, more research is needed to fully understand the time-dependent evolution of the alterations in mitochondrial dynamics markers in transgenic AD mice that exhibit classic AD pathologies, such as amyloid plaques and cellular degeneration. Furthermore, although AD brain tissue frequently exhibits mitochondrial failure, new research has shown that peripheral mononuclear blood cells' mitochondrial activities have changed (11).

Alzheimer's disease is marked by decreased activity in several enzymatic complexes critical for ATP production, including the pyruvate dehydrogenase complex, α-ketoglutarate dehydrogenase complex, complex I (NADH ubiquinone oxidoreductase), complex IV (cytochrome oxidase), and complex V (ATPase) (43, 44). Additionally, the activities of key glycolytic enzymes—phosphofructokinase (PFK), phosphoglycerate mutase, aldolase, glucose-6-phosphate isomerase, and lactate dehydrogenase—are reduced compared to age-matched non-Alzheimer's brains. Conversely, both the levels and activity of reactive oxygen species (ROS) are elevated. Moreover, mitochondrial translocase facilitates the progressive accumulation of mitochondrial Aβ in the brains of transgenic mice overexpressing human APP, which correlates with diminished activity of electron transport chain complexes III and IV and a decreased rate of oxygen consumption (43).

Maintaining an adequate pool of functional neural mitochondria and restoring those that have been damaged or lost relies on mitochondrial biogenesis. This tightly regulated process requires coordination between the nuclear and mitochondrial genomes. Mitochondrial biogenesis occurs continuously in healthy cells, as mitochondria undergo constant proliferation and fusion, but it can also be induced by oxidative stress, increased energy demands, physical activity, and certain diseases (45). However, the status of mitochondrial biogenesis in neurons affected by Alzheimer's disease (AD) remains unclear. Peroxisome proliferator-activated receptor gamma coactivator-1 alpha (PGC-1α) governs this process by activating multiple transcription factors, including nuclear respiratory factors 1 and 2 (NRF-1 and NRF-2) and mitochondrial transcription factor A (TFAM) (46). NRF-1 and NRF-2 regulate nuclear and mitochondrial genes essential for various critical functions, such as oxidative phosphorylation (OXPHOS), electron transport chain complexes I–V, mitochondrial DNA (mtDNA) transcription and replication, protein import and assembly, ion channel activity, molecular transport, and translation (47). Additionally, in response to oxidative stress, PGC-1α orchestrates a complex reactive oxygen species (ROS) defense system vital for maintaining redox homeostasis. This highlights the crucial role of mitochondrial biogenesis in preserving mitochondrial integrity throughout the organelle's lifespan. Notably, key components of the electron transport chain (ETC) are significantly diminished in AD, contributing to energy hypometabolism and mitochondrial dysfunction.ay indicate either increased mitochondrial clearance or decreased mitochondrial biogenesis (48).

Mitochondria are very important for maintaining stable calcium levels in the cytosol because they facilitate calcium entry and exit through several different pathways. This uptake is most efficient at high-Ca2+ microdomains, particularly within Mitochondria-Associated Membranes (MAMs)—specialized sub-cellular regions where the endoplasmic reticulum (ER) and mitochondria exist in close apposition (49). Various illnesses, such as cancer, neurological conditions, and metabolic syndromes, have been linked to changes in MAM structure or function.

One of the most common and early indicator of brain disorders, such as AD, is altered cellular calcium homeostasis (50). Altered Ca2+ homeostasis is a recognized early hallmark of AD (51):

Extracellular oligomeric Aβ further exacerbates this instability by activating membrane receptors and forming de novo Ca2+-permeable pores in the plasma membrane (52). According to the current paradigm, glutamate-driven excitotoxicity, activation of calcium-dependent proteases, over activation of calcium-regulated kinases and phosphatases, and mitochondrial calcium overload resulting in mPTP opening (53, 54) are some of the mechanisms by which generalized calcium overload in neurons causes neuronal death (54, 55).

Each mitochondrion contains multiple copies of mitochondrial DNA (mtDNA), which is particularly susceptible to oxidative damage and somatic mutations due to its close proximity to the electron transport chain (56). The progressive accumulation of mtDNA mutations, along with increased reactive oxygen species (ROS) and disrupted mitochondrial dynamics, severely compromises mitochondrial membrane potential and exacerbates organelle dysfunction. Loss of mitochondrial integrity is critical in this process. Under severe oxidative stress, oligomerization of voltage-dependent anion channel 1 (VDAC1) facilitates the formation of large pores in the outer mitochondrial membrane. These pores allow mtDNA to move from the matrix into the cytosol, a key event that triggers inflammatory signaling and further cellular deterioration (57) (Table 1).

Mitophagy is a specialized form of autophagy essential for mitochondrial quality control, involving the selective sequestration of damaged mitochondria into autophagosomes for lysosomal degradation. Impaired mitophagy is a hallmark of biological aging and is closely linked to the pathogenesis of neurodegenerative disorders, most notably Alzheimer's and Parkinson's diseases (33). The canonical mitophagy pathway is triggered by depolarization of the inner mitochondrial membrane. Under physiological conditions, the kinase PINK1 is rapidly degraded; however, when mitochondrial membrane potential is lost, PINK1 stabilizes on the outer mitochondrial membrane (OMM). Once stabilized, PINK1 phosphorylates Mitofusin 2 (Mfn2) and ubiquitin, facilitating the recruitment of the E3 ubiquitin ligase Parkin to the OMM (58). The recruitment of Parkin activates the ubiquitin-proteasome system (UPS) and signals the phagophore to engulf the compromised organelle. This process culminates in the formation of a mitophagosome, which then fuses with a lysosome for enzymatic degradation. The pivotal roles of the PINK1-Parkin axis in maintaining mitochondrial integrity have been extensively documented in various experimental models (58).

Fluorodeoxyglucose positron emission tomography (FDG-PET) remains a cornerstone for assessing cerebral glucose metabolism, providing a reliable proxy for synaptic density and neuronal energy demand. Unlike biomarkers that reflect downstream pathological accumulation, glucose hypometabolism can identify early functional deficits within neural networks. These metabolic changes often appear years before irreversible structural atrophy occurs, making them invaluable for early detection (59, 60). Emerging research indicates that metabolic dysfunction is not just a symptom but a critical determinant of an individual's disease trajectory and cognitive decline (61). Clarifying how these metabolic changes affect different progression pathways could fundamentally shift the Alzheimer's disease paradigm from a reactive diagnostic model to a proactive, pathway-guided management strategy. This transition would support the development of precision medicine interventions tailored to the unique physiological trajectory of each patient (62).

A controversial but intriguing concept for understanding Alzheimer's disease (AD) is the “Type 3 Diabetes” (T3D) hypothesis, which suggests that AD may be a form of brain-specific insulin resistance. According to this theory, when neurons become insensitive to insulin, crucial cognitive functions like memory consolidation and synaptic plasticity are significantly impaired. A key player in this hypothesis is the insulin-degrading enzyme (IDE). In conditions of high insulin levels, IDE may be overwhelmed by peripheral insulin, reducing its ability to break down amyloid-beta (Aβ) and leading to the accumulation of toxic proteins (63). While insulin resistance is a shared feature among obesity, Type 2 Diabetes (T2D), and AD, the manifestation in the central nervous system (CNS) is unique. In the brain, resistance is mainly characterized by disrupted signaling pathways, such as the PI3K/Akt pathway, rather than a primary issue with glucose transport, as neuronal glucose uptake is largely independent of insulin.

However, it is important to acknowledge that the terminology for T3D is still a topic of discussion. The relationship between T2D and cognitive decline in AD is well-established, but the evidence is varied. AD pathology is now seen as a complex network where insulin signaling issues, Aβ accumulation, and tau hyperphosphorylation interact and worsen each other within the A/T/N framework (Amyloid/Tau/Neurodegeneration) (64).

Extensive research has highlighted the association between AD and the dysregulation of key glycolytic enzymes, including hexokinase, glyceraldehyde 3-phosphate dehydrogenase (GAPDH), and pyruvate kinase (PK) (65, 66). The pathogenesis of AD is closely linked to the tricarboxylic acid (TCA) cycle; specifically, the accumulation of Aβ and the hyperphosphorylation of Tau inhibit TCA cycle activity. This inhibition leads to critical ATP depletion, which accelerates neuronal death (65, 66). The metabolic decline is further worsened by the accumulation of reactive oxygen species (ROS), which target and impair sensitive TCA cycle enzymes. Enzymes containing iron-sulfur clusters, such as aconitase, are particularly vulnerable; their impairment triggers a feedback loop of redox imbalances and increased oxidative stress (67). These disruptions significantly compromise cerebral function by reducing ATP synthesis, impeding synaptic maintenance, and facilitating the accumulation of neurotoxic metabolites.

Beyond glucose impairment, a growing body of epidemiological and clinical evidence links disrupted lipid metabolism to the pathophysiology and progression of AD. Elevated concentrations of free fatty acids (FFAs) and their metabolic intermediates, such as acyl-carnitines and acyl-CoA, exert neurotoxic effects, leading to mitochondrial uncoupling and severe bioenergetic dysfunction (68). Analysis of cerebrospinal fluid (CSF) in AD patients reveals a significant increase in total FFA levels, although individual lipid species show distinct patterns of alteration (69).

Emerging research indicates that defects in lipid metabolism specifically affect the integrity of lipid rafts and the formation of lipid droplets. In the prodromal and early stages of Alzheimer's disease, researchers have documented significant changes in fatty acid composition within lipid rafts, along with a marked increase in cerebral lipid peroxidation (70). These structural changes likely contribute to the instability of neuronal membranes and signaling pathways.

A characteristic feature of the AD brain is the reduction of unsaturated fatty acids, particularly omega-3 polyunsaturated fatty acids (PUFAs) and monounsaturated fatty acids (MUFAs) such as oleic acid. Specifically, levels of docosahexaenoic acid (DHA)—the most abundant PUFA in the central nervous system—are significantly reduced in the hippocampus and cortex of AD patients (68, 71). This depletion leads to a lower overall unsaturation index, which strongly correlates with cognitive impairment. In contrast, even-chain saturated fatty acids are notably increased in the CSF, further reflecting the systemic metabolic shift associated with neurodegeneration (72).

Synaptic disruption is a primary pathogenic hallmark of Alzheimer's disease, appearing long before widespread neurodegeneration. Epidemiological data show a strong correlation between synapse loss and cognitive decline, indicating that maintaining synaptic integrity is a central to disease progression (73). This dysfunction includes structural and functional changes in the presynaptic terminal, synaptic cleft, and postsynaptic dendrites. The failure of synaptic transmission is caused by multiple factors, including reduced mitochondrial energy supply at the nerve terminal and abnormal synaptic pruning mediated by microglial phagocytosis. Evidence from murine models shows that the accumulation of pathogenic Tau leads to significant synaptic loss and dysregulation even without extensive neuronal death (74).

Current research increasingly highlights that soluble Tau species, particularly oligomers, are more synaptotoxic than insoluble neurofibrillary tangles (NFTs) (75). This shift in understanding underscores the importance of targeting early-stage Tau aggregates rather than late-stage pathological inclusions to preserve cognitive function.

The accumulation of insoluble Aβ in Alzheimer's disease results from a fundamental imbalance between peptide synthesis and clearance. While familial AD, a less prevalent form, is driven by genetic mutations affecting $A\beta$ metabolism, most sporadic AD cases are attributed to inefficient Aβ clearance (76) (Figure 2).

Both insoluble fibrils and soluble Aβ oligomers exert potent neurotoxicity by forming plaques and triggering pathological cascades, including synaptic failure, excitotoxicity, changes in membrane permeability, and chronic inflammation (77, 78). The relationship between Aβ and mitochondria is bidirectional and self-reinforcing. Early Aβ accumulation directly impairs mitochondrial function, while mitochondrial dysfunction increases the production of mitochondria-derived reactive oxygen species (ROS). These ROS further stimulate Aβ production by activating β- and γ-secretases, creating a deleterious feed-forward loop that accelerates neuronal loss (79).

A hallmark of early AD is systemic impairment of cerebral energy metabolism (304). High ROS concentrations cause molecular damage both at the site of generation and in adjacent cellular structures (80). The activities of key metabolic enzymes—including cytochrome c oxidase (Complex IV), the pyruvate dehydrogenase complex, isocitrate dehydrogenase, α-ketoglutarate dehydrogenase, and ATP synthase—are significantly reduced in the AD brain. Notably, while these enzymes decline, the activities of succinate dehydrogenase (Complex II) and malate dehydrogenase often increase. This enzymatic imbalance compromises the inner mitochondrial membrane potential and severely reduces ATP production (80).

The accumulation of mitochondrial Aβ is not simply a passive byproduct of disease; it involves the targeting of specific proteins critical to organelle integrity. Key identified targets include Cyclophilin D (CypD), Aβ-binding alcohol dehydrogenase (ABAD), and human Presequence Protease (hPreP) (81). It has been shown that CypD can form complexes with Aβ within the mitochondria of cortical neurons in APP transgenic mice, increasing the translocation of CypD from the matrix to the inner membrane (82), an essential process in the opening of the mPTP (83). This causes dissipation of the internal membrane potential and generation of ROS, followed by rupture of the external membrane and non-specific release of intermembrane space proteins into the cytosol, which activate various signal transduction pathways such as apoptosis (84). The convergence of Aβ and ROS at the mPTP node serves as a critical checkpoint, shifting the cell from metabolic stress to programmed apoptosis.

Further exacerbating this dysfunction is the interaction between Aβ and ABAD, a mitochondrial matrix enzyme that uses NAD⁺ or NADH as a cofactor for the oxidation and reduction of alcohol groups (85). Aβ-ABAD complexes have been detected in both transgenic mouse models and human AD brains. This pathological interaction prevents NAD⁺ binding, thereby impairing mitochondrial enzymatic activity and directly contributing to the cognitive deficits observed during AD progression (83).

While the exact mechanisms by which aberrant Tau triggers mitochondrial dysfunction are still being elucidated, its impact on synaptic integrity is well documented. Synaptic mitochondria are essential for meeting the high bioenergetic demands of the nerve terminal and maintaining calcium homeostasis. Consequently, disruption of mitochondrial transport to the synapse is a primary driver of synaptic degeneration and eventual neuronal death (67).

Evidence from cellular and murine models indicates that Tau overexpression and hyperphosphorylation severely disrupt the localization and distribution of mitochondria (86). This displacement causes profound axonal dysfunction and synaptic loss (87). Recent studies using induced pluripotent stem cells (iPSCs) from patients with frontotemporal dementia (FTD) carrying the R406W Tau mutation have confirmed these trafficking impairments. In cerebral organoids derived from these iPSCs, axonal mitochondria showed decreased stability and a shift toward retrograde orientation, resulting in a net reduction of mitochondrial density within the axon (88).

Similarly, iPSC-derived neurons with the N279K and P301L Tau mutations show a significant reduction in anterograde axonal transport compared to healthy controls (89). This failure in the forward delivery of mitochondria—caused by Tau interfering with motor proteins such as kinesin—deprives the distal synapse of essential ATP, making it vulnerable to metabolic collapse.

Mitochondrial dysfunction is an early-stage hallmark of AD, characterized by diminished ATP synthesis and a reciprocal surge in reactive oxygen species (ROS) (89). Beyond bioenergetic failure, Aβ peptides directly compromise mitochondrial integrity by destabilizing mitochondrial membranes and inhibiting the respiratory chain, further exacerbating oxidative stress (90).

The impairment of mitochondrial function leads to the initiation of a subsequent inflammatory reaction. Molecules released from dysfunctional mitochondria, including mitochondrial DNA (mtDNA) and cytochrome c, serve as damage-associated molecular patterns (DAMPs). These DAMPs stimulate Pattern Recognition Receptors (PRRs) and the NLRP3 inflammasome, promoting persistent neuroinflammation (91).

Crucially, neuronal homeostasis relies on the structural and functional tethering of the endoplasmic reticulum (ER) and mitochondria via Mitochondria-Associated Membranes (MAMs). These contact sites regulate calcium signaling, lipid synthesis, and protein folding. In Alzheimer's disease (AD), the disruption of MAMs enhances inflammatory signaling and impairs the cellular response to ER stress (92). These structural alterations ultimately dictate cell fate by creating a feedback loop of elevated reactive oxygen species (ROS) production and a diminished antioxidant capacity to neutralize neurotoxic metabolites (93).

A significant shift in Alzheimer's disease (AD) research is represented by the Mitochondrial Cascade Hypothesis, which posits that mitochondrial dysfunction is the primary driver of AD pathophysiology, rather than a secondary consequence of Tau or Aβ accumulation (94). This theory suggests that age-related mitochondrial decline is the fundamental catalyst for the disease. According to this model, an individual's baseline mitochondrial function is determined by genetic inheritance, while the rate of age-associated mitochondrial decline is influenced by environmental factors. The interaction of these two variables determines the age of onset and the rate of disease progression.

Importantly, this theory posits that the production of Aβ is influenced by mitochondrial dysfunction triggered by toxins. Consequently, the buildup of Aβ, which initiates well in advance of the manifestation of symptoms, is viewed as an indicator of aging and metabolic challenges in the brain rather than the root cause (94).

There has been much discussion and modification of a model that describes how AD biomarkers grow in connection to one another as well as to the beginning and progression of clinical symptoms (95). It is important to acknowledge that tau and Aβ pathology may start separately in LOAD. The dual-pathway concept posits that Aβ elevation and tau hyperphosphorylation may be independent pathophysiological processes with harmful synergy. According to the Aβ-tau interaction hypothesis, Aβ and tau pathology, which results in synaptic dysfunction, neuronal death, and cognitive impairment (96), is determined by the intrinsic interaction with Aβ and tau (Aβ42 enhances tau phosphorylation, truncation, and aggregation, while tau can increase Aβ synthesis). The impacts of several risk factors and biomarkers, including as tau hyperphosphorylation and Aβ elevation, may be caused by impaired neuroplasticity and mitochondrial dysfunction (97), which may be a common upstream etiology of AD.

Microglia, the indigenous immune cells of the central nervous system, are part of the myeloid lineage and play a crucial role in protecting the brain. They detect harmful stimuli through Pattern Recognition Receptors (PRRs) by recognizing pathogen-associated molecular patterns (PAMPs) and damage-associated molecular patterns (DAMPs) (98) (Figure 3). Initially, the activation of microglia is beneficial for protecting the brain; Damage-Associated Molecular Patterns (DAMPs) such as Aβ species are cleared by microglia to prevent the formation of plaques (99). However, prolonged activation can lead to chronic inflammation and a harmful change in function. In the presence of ongoing accumulation of Aβ, a subset of microglia becomes abnormally activated, causing a neurotoxic effect that accelerates the progression of neurodegenerative diseases (100, 101).

Recent research suggests that this change in function is driven by changes in metabolism. Studies conducted in laboratory settings show that exposure to Aβ leads to a shift from using oxidative phosphorylation to aerobic glycolysis, similar to the Warburg effect. As the disease progresses, the development of immune tolerance in microglia toward Aβ plaques becomes a key feature (102). Despite the initial shift toward glycolysis, microglia in advanced stages of Alzheimer's disease experience a decline in both glycolytic and oxidative phosphorylation pathways, resulting in a state of “energetic exhaustion” that impairs their ability to clear debris and protect neurons (103).

Neurons primarily rely on mitochondrial oxidative phosphorylation for energy production, while astrocytes have a unique metabolic profile with high glycolytic activity. Astrocytes express high levels of glycogen and key enzymes involved in glycolysis, making them particularly responsive to the changing energy needs of the brain (21).

A crucial aspect of this metabolic collaboration is the Astrocyte-Neuron Lactate Shuttle (ANLS). When astrocytes undergo glycolysis, glucose is quickly converted into L-lactate, which is then transported out of the cells through monocarboxylate transporters (MCTs)—specifically MCT1 and MCT4 on astrocytes and MCT2 on neurons. Once taken up by neurons, L-lactate is converted back to pyruvate to support the Tricarboxylic Acid (TCA) cycle (20). The shuttle system is not just an additional pathway; it is crucial for complex neuronal functions like long-term potentiation (LTP) and memory consolidation. Any interruption in astrocytic glycolysis or the efficiency of MCT transport directly impacts neuronal survival. In Alzheimer's disease, dysfunction in the ANLS is a key metabolic obstacle that plays a significant role in early cognitive decline (20).

Memory loss is a key characteristic of Alzheimer's disease (AD), a neurodegenerative condition caused by a complex pathological process that involves the overproduction of reactive oxygen species (ROS) (104). This oxidative imbalance in the brain is due to a dual mechanism: an increase in pro-oxidant enzymes and a decrease in antioxidant defenses like catalase and superoxide dismutase (SOD) (Table 2).

One of the main enzymes responsible for producing superoxide during long-term neuroinflammation is Nicotinamide adenine dinucleotide phosphate oxidase (NADPH oxidase). This complex, which is located on cell membranes and consists of multiple subunits, plays a significant role in the development of Alzheimer's disease (AD) and Parkinson's disease (PD). In these conditions, the persistent activation of NADPH oxidase in microglia leads to continuous oxidative stress and damage (105).

The activation of phagocytic NADPH oxidase is a tightly controlled process that occurs in specific locations within the cell. When stimulated, the cytosolic components of the enzyme, such as p47phox, p67phox, and p40phox, move to the cell membrane to interact with the membrane-bound subunits, gp91phox and p22phox. The key component of this complex is the NOX2 subunit, which facilitates the transfer of an electron from NADPH to molecular oxygen (O2), resulting in the production of the superoxide radical anion (O2–). In the context of Alzheimer's disease, the excessive and uncontrolled generation of superoxide by microglial NADPH oxidase contributes to significant oxidative damage, directly contributing to neurodegeneration and cognitive impairment (106).

Dysregulated microglial function plays a key role in various neurodegenerative, autoimmune, and viral diseases, such as Alzheimer's and Parkinson's disease. When amyloid beta interacts with microglial receptors, it triggers an inflammatory response characterized by the release of pro-inflammatory cytokines like TNF-alpha, IL-1beta, and IL-6 (107). This inflammatory activation is driven by increased levels of inducible nitric oxide synthase (iNOS) and NADPH oxidase (NOX), leading to the production of reactive oxygen species (ROS). These factors create a persistent inflammatory environment that contributes to neuronal damage and loss of synapses over time (108).

The metabolic and inflammatory status of microglia is closely tied to their mitochondrial health. In the context of Alzheimer's disease (AD), both microglia and neighboring neural cells display a range of mitochondrial abnormalities, including:

The complex interplay of diabetes, dyslipidemia, obesity, and hypertension in clinical settings makes it challenging to pinpoint their individual roles in pathology. Research often concentrates on one factor, neglecting the combined impact of these risk factors on Alzheimer's disease (AD) risk. Therefore, a more inclusive strategy involves examining metabolic syndrome, a collection of conditions such as central obesity, hypertension, hyperlipidemia, and diabetes, as a comprehensive risk profile (110).

Numerous epidemiological studies have investigated the association between diabetes mellitus (DM) and Alzheimer's disease (AD), with a significant body of evidence indicating a direct connection between Type 2 Diabetes (T2D) and an elevated risk of developing AD. Various mechanisms have been proposed to elucidate this relationship, including insulin deficiency, central insulin resistance, impaired insulin receptors, and neurotoxicity induced by high blood sugar levels. Additionally, the harmful effects of advanced glycation end products (AGEs), chronic neuroinflammation, and damage to the brain's blood vessels contribute to the progression of the disease (111).

The rising occurrence of Alzheimer's disease (AD) in individuals with type 2 diabetes (T2D) and obesity suggests common underlying molecular pathways (112). Specifically, impaired insulin signaling plays a crucial role in accelerating AD pathology (113). The insulin-degrading enzyme (IDE) is responsible for clearing both insulin and Aβ, a key protein in AD. However, in cases of hyperinsulinemia due to insulin resistance, IDE can become overwhelmed, prioritizing insulin degradation over Aβ breakdown. This leads to reduced clearance of Aβ and its accumulation in the brain. Additionally, IDE activity, a significant risk factor for sporadic AD, naturally decreases with age, further compromising the brain's ability to break down proteins (114).

The connection between excess body fat and cognitive decline, previously overlooked, is now supported by a strong body of epidemiological evidence. Meta-analytical studies show that individuals with obesity or related metabolic conditions are almost twice as likely to develop Alzheimer's disease (AD) (115). Elevated levels of systemic cholesterol, a common feature of obesity, have a significant impact on the production of Aβ proteins. Although the brain primarily produces its own cholesterol and isolates it from the rest of the body, research suggests that prolonged high cholesterol levels in the body may promote the transfer of cholesterol across the blood-brain barrier (BBB), contributing to the development of amyloid plaques (116).

A key process that connects metabolic issues in the body to brain-related problems is the movement of Aβ from the bloodstream into the central nervous system (CNS). This transfer is mainly facilitated by the receptor for advanced glycation end products (RAGE). When levels of Aβ in the blood remain high for an extended period, it enhances the transport of Aβ into the brain through RAGE. Once in the brain, Aβ interacts with neurons, leading to a series of cellular stress responses and neurotoxic effects (117).

The timing of obesity plays a crucial role in determining the risk of developing dementia. Studies based on population cohorts indicate that reducing midlife obesity by 20% could potentially lower the prevalence of dementia in older individuals by 10% (118). Interestingly, there is a phenomenon known as the “obesity paradox” where having a higher Body Mass Index (BMI) in late life is associated with a lower risk of dementia, despite midlife obesity being a significant risk factor. This paradox may be attributed to the unintended weight loss that occurs during the early stages of dementia (119).

Obesity is essentially a condition characterized by chronic, low-grade inflammation due to the enlargement and increased number of fat cells (120). The excess fat tissue releases inflammatory substances such as TNF-alpha, IL-6, MCP-1, and various adipokines, which attract more immune cells and sustain inflammation. Recent studies suggest that this inflammation in fat tissue is the main cause of cognitive decline in obese individuals (121). This inflammatory environment triggers the formation of inflammasome complexes, which intensify the immune response locally and throughout the body, ultimately affecting the brain's neuro-immune system (122).

The heart-brain axis is a complex system of pathological mechanisms that connect the circulatory and cerebrovascular systems. Cardiovascular disease (CVD) and Alzheimer's disease (AD) are common chronic conditions associated with aging, and their increasing prevalence is thought to be linked to abnormalities within this axis. Although the specific molecular regulators are still being studied, it is now understood that established risk factors for CVD play a significant role in the development of AD-related dementia (123, 124).

Both experimental models and long-term observational cohorts have confirmed the link between Alzheimer's disease (AD) and cardiovascular health. Specifically, levels of amyloid-beta (Aβ 1–40) in the bloodstream are linked to the presence of coronary artery disease, irrespective of cognitive function in individuals (125). This indicates that the dynamics of amyloid in the body may be a shared factor in both heart and brain dysfunction.

The connection between Alzheimer's disease (AD) and atherosclerosis has been a topic of discussion for many years. Recent research indicates that individuals with AD have a notably higher occurrence of cerebrovascular atherosclerosis in comparison to individuals of the same age who do not have AD (126, 127). The strong association between Alzheimer's disease (AD) and atherosclerosis in the Circle of Willis, the main arterial network that supplies the brain, is of great importance. Research based on population studies has shown that structural abnormalities in this crucial vascular junction can lead to reduced cerebral blood flow and damage to the blood-brain barrier. This, in turn, accelerates the buildup of Aβ protein and the advancement of neurodegenerative processes (128).

The gradual buildup of abnormal protein aggregates in the central nervous system (CNS) is a key characteristic of neurodegenerative diseases such as Alzheimer's disease (AD), Parkinson's disease (PD), and amyotrophic lateral sclerosis (ALS) (129). It is noteworthy that the age-old Hippocratic belief that “all disease originates in the gut” remains relevant in modern neurobiology, with growing evidence supporting the importance of the gut-brain axis in maintaining neural well-being (130).

One key characteristic of gut dysbiosis in Alzheimer's disease is an increased ratio of Firmicutes to Bacteroidetes (F/B ratio). This imbalance in gut bacteria is connected to the buildup of amyloid precursor protein (APP) in the intestines at the onset of the disease (131). Studies using APP/PS1 mouse models have shown that changes in gut microbiota composition can lead to higher levels of amyloid beta (Aβ) in the central nervous system, as well as notable declines in memory and spatial learning abilities (132).

The release of bacterial metabolites provides a possible link between the gut and the brain. Short-chain fatty acids (SCFAs) have been found to disrupt the interactions between proteins that are crucial for the assembly of Aβ, potentially offering a protective effect on the brain (133). On the other hand, the microbial metabolite trimethylamine N-oxide (TMAO) has been associated with the pathophysiology of Alzheimer's disease (134). TMAO worsens Aβ pathology by increasing the activity of β-secretase, leading to faster amyloid formation and cognitive decline. Additionally, TMAO affects the release of calcium from cellular stores in response to stimuli, promoting excessive platelet reactivity. This mechanism facilitates the spread of Aβ from platelets into the bloodstream and eventually into the brain tissue, adding to the overall amyloid accumulation (131, 135).

In the clinical evaluation of neurodegenerative disorders, particularly Alzheimer's disease (AD), the use of 18 F-fluorodeoxyglucose positron emission tomography (FDG-PET) plays a crucial role in aiding diagnosis. Traditionally, FDG-PET has been utilized to visualize synaptic loss, neuropil atrophy, and neuronal functional decline. Within the established research paradigm, decreased FDG uptake has been classified as an “N” (neurodegeneration) biomarker, indicating reduced metabolism. However, according to recent research, FDG-PET signals may not only reflect neuronal activity but also astrocytic glucose utilization through glutamate transport (GLT-1). This leads to a reevaluation of metabolic imaging in AD; in particular, by preserving glucose absorption levels, reactive astrogliosis may paradoxically “mask” underlying neuronal loss. Therefore, rather than being a stand-alone indicator of neuronal health, FDG-PET should be viewed as a measurement of the larger neurovascular unit (136).

Recent research suggests that decreased FDG uptake could be indicative of vascular abnormalities, particularly those related to dysfunction in the blood-brain barrier (BBB). Studies conducted within the Alzheimer's Disease Neuroimaging Initiative (ADNI) cohort have examined the need to redefine FDG-PET as a distinct biomarker for metabolic dysfunction, denoted as “F” (representing FDG hypometabolism). This reclassification would separate it from the general “N” category in individuals who are amyloid-positive and tau-positive, allowing for a more detailed understanding of metabolic impairment that is independent of structural atrophy (126).

Magnetic Resonance Spectroscopy (MRS) offers a direct and non-invasive method to examine the bioenergetic and metabolic abnormalities linked to neurodegenerative conditions. In vivo 1H-MRS enables the measurement of various important metabolites, including:

While these metabolites exhibit distinct peaks in the spectrum, others like Glutathione (GSH), GABA, glucose, and lactate display more subtle and overlapping signals, leading to increased measurement variability (137).

Meta-analyses have consistently found a metabolic profile characterized by decreased N-acetylaspartate (NAA) and increased myo-inositol (mI) levels in both Mild Cognitive Impairment (MCI) and Alzheimer's Disease (AD) (138, 139). However, many studies have focused exclusively on proton magnetic resonance spectroscopy (1H-MRS) and have not incorporated markers of oxidative stress (OS) and neuroinflammation. As a result, there is still a gap in diagnostic knowledge, as it remains uncertain whether specific longitudinal metabolic changes can accurately predict the progression from MCI to AD or provide insights into the underlying molecular mechanisms.

Once considered a mere waste product of metabolism, L-lactate is now acknowledged for its various physiological functions in the central nervous system, such as modulating neuronal activity and ensuring energy balance (20). Changes in cerebrospinal fluid (CSF) lactate levels are important clinical markers of metabolic dysfunction; for example, increased CSF lactate is a key feature of energy metabolism impairment in mitochondrial disorders and hepatic encephalopathy (140, 141).

Alzheimer's disease is primarily characterized by impaired glycolytic pathways and mitochondrial dysfunction. Extensive clinical studies have shown that individuals with Alzheimer's disease have notably elevated levels of cerebrospinal fluid lactate (142). Interestingly, higher lactate levels are frequently observed in patients with milder cognitive impairment, indicating a potential compensatory metabolic change or a shift resembling the Warburg effect in the early stages of neurodegeneration. Additionally, there is a negative association between cerebrospinal fluid lactate levels and Tau protein levels, suggesting a possible connection between anaerobic metabolic shifts and cytoskeletal abnormalities.

Mitochondrial dysfunction directly affects the brain's main process for metabolizing glucose, known as the Tricarboxylic Acid (TCA) cycle. Abnormalities in the TCA cycle, along with reduced overall metabolism, result in a failure to produce ATP effectively (143, 144). While the exact sequence of events is still debated, evidence suggests that the buildup of Aβ disrupts mitochondrial function, leading to a breakdown in energy production that hinders synaptic plasticity and cognitive abilities (145).

Mitochondrial DNA copy number (mtDNA-CN) is a valuable indicator of mitochondrial functional capacity and genome stability. Abnormal mtDNA-CN levels can indicate issues with the transcription machinery and replication of mitochondrial DNA (146). In individuals with Alzheimer's disease (AD), pyramidal neurons show a significant decrease in mitochondrial DNA copy number (mtDNA-CN) compared to those without cognitive impairment (147). This reduction is also evident in peripheral blood mononuclear cells of both AD and Mild Cognitive Impairment (MCI) patients (148). These findings suggest that mtDNA-CN could be a valuable preclinical biomarker for mitochondrial dysfunction, potentially offering a target for therapeutic interventions to restore genomic stability in the mitochondria (149) (Figure 4).

Table 3 summarizes the relationship between particular metabolic dysfunctions, their clinical readouts, and the accompanying therapeutic implications in order to give a coherent overview of how these varied markers align with the metabolic-pathophysiological framework of AD.

Metabolomics plays a crucial role in identifying biomarkers for neurodegenerative disorders by allowing for the simultaneous quantification of thousands of metabolites in different organ systems. Recent studies have highlighted specific profiles of bile acids (BA), amino acids (AA), and fatty acids (FA) as potential indicators for Alzheimer's disease (AD) and Mild Cognitive Impairment (MCI) (150, 151). While previous research has focused on low-molecular-weight metabolites, lipid metabolites, which have significant biological activity in the central nervous system, are now gaining recognition in these investigations.

Lipidomics, a specialized field within metabolomics, is dedicated to the thorough identification and characterization of various lipid classes, subclasses, and molecular species. By combining metabolomics with lipidomics, scientists can create a comprehensive overview of the metabolic landscape, offering a valuable tool to investigate the underlying mechanisms of Traditional Chinese Medicine (TCM) and other therapeutic approaches. As the brain is the organ with the highest concentration of lipids in the body, lipids play a crucial role in maintaining the structural integrity of cell membranes and regulating important signaling pathways. Changes in lipid metabolism are closely associated with disruptions in homeostasis and the development of neurodegenerative diseases.

A significant study in lipidomics identified a set of 26 sphingolipids and glycerophospholipids that can effectively differentiate between patients with Alzheimer's disease (AD) and cognitively normal individuals (152). Specifically, elevated concentrations of certain sphingomyelins (SMs) such as SM C16:0, SM C18:1, SM C16:1, and SM (OH) C14:1 were observed in AD patients. These lipid profiles are closely associated with the severity of the disease, indicating their potential utility as precise molecular markers for diagnostic and therapeutic purposes.

While around 75 genetic loci have been associated with an elevated risk of Alzheimer's disease (AD), most cases are sporadic or late-onset variants. Among these, the Apolipoprotein E (APOE) gene, particularly the ε4 allele (APOE4), stands out as the most significant and consistent genetic risk factor. In the last 30 years, close to 10,000 studies have explored the complex connection between APOE and AD pathology (153).

The APOE genotype plays a crucial role in clinical research and trials, impacting prognosis, risk evaluation, and treatment planning. Particularly, the APOE genotype has a notable effect on how individuals respond to anti-amyloid monoclonal antibodies like lecanemab, donanemab, and aducanumab. This genetic interaction has led to the creation of therapies that target the specific disease pathways influenced by the APOE4 protein (154, 155).

Data from long-term studies, like the Alzheimer's Disease Neuroimaging Initiative (ADNI), show that the APOE4 gene variant is a strong indicator of both amyloid buildup and shrinkage of the hippocampus, even in the early stages of Alzheimer's disease. Including APOE status in longitudinal analysis enables more accurate identification of when symptoms may appear and helps pinpoint individuals who are at higher risk before they show any symptoms (156).

In addition to nuclear genetics, variations in mitochondrial DNA (mtDNA) - categorized into established haplogroups - are known to play a significant role in the susceptibility to Alzheimer's disease (AD) (157). A human mitochondrial haplogroup is characterized by a specific set of mutations that have accumulated along a maternal lineage (158). These haplogroups are identified by letters (e.g., H, V, L) and further refined by alphanumeric suffixes (e.g., H51A1). Research indicates that certain haplogroups, such as U, H, and J, may act as significant risk factors for AD, potentially affecting the efficiency of oxidative phosphorylation and production of reactive oxygen species (ROS) (157, 159).

Current evidence supports the hypothesis that ketogenic diets (KDs) can enhance cognitive outcomes in patients with Alzheimer's disease (AD). However, the effectiveness of this therapeutic approach is influenced by the stage of AD, disease progression, and the presence of the ApoE4 genotype, which significantly affects the metabolic response to dietary interventions. Despite maintaining adequate caloric intake, the strict carbohydrate restriction necessary for achieving ketosis may compromise the overall quality of the diet. Low-carbohydrate regimens often lack essential micronutrients such as thiamine, folate, vitamins A, E, B6, and K, as well as minerals like calcium, magnesium, and iron. Additionally, KDs are typically low in dietary fiber.

Long-term adherence to very-low-carbohydrate diets has been associated with increased risks of all-cause mortality, cardiovascular events, and cancer-related mortality (254). As a result, these diets are not recommended for certain patient populations, including those with a history of hypertriglyceridemia-related acute pancreatitis or severe hypertriglyceridemia. Caution is also advised for patients taking sodium-glucose cotransporter 2 (SGLT2) inhibitors or diuretics, as they may be at higher risk of metabolic complications and electrolyte imbalances.

The physiological properties of the blood-brain barrier (BBB) undergo significant changes during the progression of Alzheimer's disease (AD) (Table 6). It is crucial to use various cellular models to study these disease-specific states, as BBB permeability and drug transport kinetics can differ based on the underlying pathology.

In the future, Drug Delivery Systems (DDS) will be increasingly important for targeting the brain, as they can overcome traditional delivery barriers. Currently, multi-functionalized DDS, which incorporate multiple targeting ligands or responsive elements, are more effective than single-functionalized ones. Additionally, combining different DDS platforms shows promise for improving therapeutic outcomes. An example is the drug-in-cyclodextrin-in-liposome system, which combines the stabilization properties of cyclodextrins with the biocompatibility and transport abilities of liposomes to achieve better therapeutic results (255).

Preclinical testing of multi-target therapies poses additional challenges, as current animal models and testing protocols often do not accurately capture the complex interactions of multiple targets (256). Furthermore, the stringent regulatory requirements for new multi-target therapies make the approval process more suitable for single-target drugs, necessitating adjustments that could hinder progress and increase costs. The primary obstacle to widespread adoption of this approach is the substantial research and development expenses, along with the potential high costs involved in bringing it to market (257).

While multi-omic investigations have transformed oncology over the past decade, their application in AD is still in its early stages (265). Despite the emergence of several human cohort studies in the last 5 years, few are truly integrative. Many studies focus on proteomic, transcriptomic, or lipidomic changes in isolation rather than integrating them into a comprehensive biological model. For example, research utilizing microglial extracellular vesicles (EVs) from cryopreserved AD brain tissue analyzed lipidomic, transcriptomic, and proteomic datasets to illustrate how microglial activity shifts in late-stage AD, impacting the clearance of tau-laden neuronal debris (266). This suggests that EVs could serve as a source of multi-omic data and a potential diagnostic indicator of disease progression.

Blood (plasma/serum) remains the most practical biological matrix for large-scale exploratory investigations. Blood-based biomarkers offer significant advantages over neuroimaging or cerebrospinal fluid (CSF) analysis, particularly in terms of cost-effectiveness and accessibility. As a result, they are being positioned as the initial step in a multi-stage diagnostic approach, similar to established protocols in cardiovascular and infectious diseases (267). To successfully integrate these markers into clinical practice, the primary focus should be on establishing Contexts of Use (COUs)—formal statements that define the specific purpose and manner of biomarker application in drug development and clinical diagnosis (268).

Multi-omic studies of AD are still in their early stages, despite the fact that multi-omics has been utilized for more than 10 years to examine changes in other diseases, particularly in the field of oncology (265). However, a number of multi-omics research using human cohorts and data in the context of AD have been conducted over the last 5 years. However, few of these research are genuinely integrative. Despite taking into account several biological levels, they only looked at these changes separately. For example, a study (266) analyzed proteomics, transcriptomics and lipidomics data derived from microglial extracellular vesicles acquired from cryopreserved brain tissue with neuropathologically diagnosed AD. They demonstrated how neuronal debris, including tau in extracellular vesicles, is released when microglial activity is altered in the latter stages of AD. This implies that extracellular vesicles may serve as a possible readout of disease state in addition to being a source of future omics data.

However, a real integrative strategy has been the focus of some recent investigations. Crucially, researchers (269) have analyzed transcriptomics and genomes from AD patients' brain tissue using this method. They have demonstrated that single-omics techniques fail to identify the majority of pathways implicated in AD pathogenesis. In fact, it is more probable to find important biological hubs or pathway changes that are pertinent to disease when several weak signals from several individual molecules are combined. For instance, ATP6V1A, which encodes an ATPase component, was identified as a crucial network regulator implicated in several protein networks linked to AD in a study (304) using post-mortem brain tissue samples with confirmed AD. This shows that this gene is not only a possible risk factor for AD but could also serve as a target for disease alterations, emphasizing the efficacy of integrative techniques for uncovering potential disease modifying medication targets. Furthermore, by finding 20 SNPs linked to amyloid deposition and frontal brain shrinkage, the gap between genomics and clinical imaging was successfully closed (270). When taken as a whole, these studies demonstrate the existence of repeatable metabolic and proteomic signatures; nonetheless, the lack of standardized techniques and the requirement for validation in larger, more varied patient cohorts continue to impede their therapeutic application. In the end, overcoming major obstacles in data standardization, validating findings across various longitudinal cohorts, and creating easily accessible, reasonably priced diagnostic tools are necessary to close the gap between multi-omic discovery and clinical practice.