Glymphatic Dysfunction as a Convergent Mechanism in Neurodegenerative Disease: Links to Inflammaging, Gut-Brain Signaling, and Lifestyle Modulators

1. Introduction

Efficient clearance of metabolic waste and interstitial solutes is essential for maintaining central nervous system (CNS) homeostasis. Although brain lacks a conventional lymphatic vasculature within its parenchyma, the last decade scientific findings suggest the existence of a complicated mechanism, coined as the glymphatic system, which is responsible for the removal of macromolecules, neurotoxic proteins, and excess fluid [1,2]. Since the description of the glymphatic system, our understanding of brain fluid dynamics has been fundamentally reshaped and a mechanistic framework with direct relevance to aging and neurodegeneration is increasingly unveiled.

The glymphatic system is suggested to be a brain-wide perivascular network that facilitates the exchange between cerebrospinal fluid (CSF) and interstitial fluid (ISF), thereby enabling convective transport of solutes through the brain parenchyma [3]. According to the prevailing theory, CSF enters the brain along periarterial spaces, driven by arterial pulsatility and pressure gradients, mixing with ISF within the interstitium [4]. The astrocytic endfeet ensheathing the cerebral vasculature is considered a key structural element of glymphatic efficiency, by forming the outer boundary of perivascular spaces, expressing aquaporin-4 (AQP4) water channels that enable glymphatic flow. Proper polarization of AQP4 channels at the perivascular astrocytic endfeet is of crucial importance for the efficient waste clearance [5]. After traversing the interstitial compartment, fluid and solutes exit the brain along perivenous spaces, ultimately draining toward meningeal lymphatic vessels and cervical lymph nodes [3]. The glymphatic system via this pathway is suggested to contribute to the clearance of neurotoxic proteins, including amyloid-β (Aβ) [1], tau [6] and α-synuclein (SNCA) [7] (Figure 1).

Aging has been proposed to exert a profound impact on glymphatic structure as seen with loss of perivascular AQP4 polarization [8], vascular stiffness [9] and altered CSF dynamics [10]. Glymphatic function is also altered with advancing age [11] possibly due to reductions in arterial pulsatility [12] and astrocytic reactivity [13]. Chronic low-grade immune activation, also known as inflammaging, further contributes to these changes.

Gut dysbiosis leads to increased expression of pro-inflammatory cytokines resulting in systemic inflammation. Inflammatory mediators from the gut may enter the CNS and impair glymphatic function [14]. Moreover, the production of neurotransmitters and microbial metabolites from the gut may interfere with the circadian rhythms thus leading to sleep disturbances [15]. Given the positive impact of sleep on glymphatic activity [4], any sleep disturbance or disorder may lead to the accumulation of neurotoxic waste in the brain, potentially contributing in excessive neuroinflammation that is present in major CNS disorders.

Given its proposed role in clearing aggregation-prone proteins and inflammatory mediators, there is increasing evidence highlighting the potential implication of glymphatic system in neurodegeneration as a convergent mechanism in multiple CNS disorders [16,17,18,19,20,21,22,23]. Although direct monitoring of glymphatic flow in humans remains challenging, several in vivo imaging approaches have been developed to infer glymphatic function [24]. Contrast-enhanced and contrast-free magnetic resonance imaging (MRI) are among the in vivo techniques that can be applied to humans. The former is invasive and requires the injection of a paramagnetic contrast agent, while the latter analyzes CSF movements without the need for a tracer. Consequently, contrast-free MRI techniques have become the primary tools in human studies, with diffusion tensor imaging along the perivascular space (DTI-ALPS) being the most widely used technique. DTI-ALPS quantifies directional water diffusivity along perivascular spaces, particularly in the white matter, providing an indirect index of perivascular fluid transport. Although there are several limitations in the reflection of glymphatic function from the estimation of the DTI-ALPS index, it is widely employed in a plethora of studies tackling neurodegenerative disorders [25]. Additional MRI-based markers monitor enlarged perivascular space (ePVS) burden [26] and choroid plexus volume [27] that are used separately or in combination with the DTI-ALPS index.

The scope of this review is to synthesize existing preclinical and translational evidence for glymphatic system dysfunction in neurodegenerative diseases. The last part is dedicated to the discussion of the physiological regulators of glymphatic activity. We build upon this synthesis while we integrate emerging concepts of inflammaging and gut-brain axis interactions. Our goal is to highlight the possibility of a systemic nature of the glymphatic system that may contribute to CNS disorders onset. We additionally aim to underscore the importance of the clinical assessment of waste clearance mechanisms towards the unraveling of useful insights for future diagnostic and therapeutic strategies.

2. Neurodegenerative Diseases

2.1. Alzheimer's Disease (AD)

Expression of AQP4 has been reported to correlate with advancing age while loss of its perivascular localization is associated with increased amyloid burden and higher Braak stage [28]. In an experimental model of AD (5XFAD mice) where AQP4 was coupled with a-syntrophin (Snta1), deletion of the latter allowed for controlled depolarization of AQP4 without alterations in its expression [29]. Both deficiency and mislocalization of AQP4 resulted in increased parenchymal Aβ plaque formation and depositions in Snta mice compared to their littermates. Furthermore, human post-mortem case series analysis demonstrated that the loss of perivascular localization of AQP4 in cortical gray matter was more evident in patients with AD compared to subjects with earlier stages of cognitive decline [30]. Genetic analysis has further unveiled AQP4 polymorphisms that are associated with higher risk of cognitive decline [31] and modification in Aβ uptake [32]. However, others have reported a protective role of AQP4 deficiency leading to mitigated astrogliosis [33], neuronal apoptosis and uptake of Aβ [34].

Glymphatic flow has been reported to be dysregulated following depletion or pharmacological inhibition of AQP4 in the Aβ precursor protein/presenilin 1 model (APP/PS1) [33], the 5X FAD model [35] and tauopathy models [6,18,36]. In all instances glymphatic system dysregulation was associated with toxic accumulation of Aβ and tau, astroglial atrophy, loss of synaptic proteins [33], memory deficits, behavioral impairment and neuronal dysfunction [35]. Suppression of clearance of toxic waste following deletion of AQP4 is further supported by the slowing down of CSF influx by almost 70% [1]. Particularly for models of tauopathy, impairment of glymphatic flow seems to take place in specific brain regions, such as the rostral cortex, which are affected early during tau deposition [18]. Glymphatic flow stagnation in these areas is associated with increased volume changes that are potentially implicated in neurodegeneration and diminished cognitive performance [6,36].

Experimental results of AQP4 deficiency widely coincide with findings from studies assessing the effects of meningeal lymphatics disruption that are involved in the drainage of CSF and ISF to cervical lymph nodes. Surgically disrupted meningeal lymphatics in AD mouse models increased Aβ deposition both in the meninges and the brain tissue, accompanied with increased inflammation and behavioral impairment [37]. Disrupted meningeal lymphatics were further correlated with a pro-inflammatory microglial molecular signature like what is observed in AD pathogenesis [38]. Given that the dominating microglial phenotype is suggested to be the key mediator for the association between tauopathy and glymphatic dysfunction on the one hand [39,40], and that Aβ accumulation in the CSF may halt the glymphatic influx on the other hand [41], a waste clearance mechanism dysfunction driven by pathological features observed in AD pathogenesis is further reinforced.

Advanced imaging studies have revealed consistent clinical symptoms associated with glymphatic dysfunction across the AD spectrum. These studies aimed to link glymphatic dysfunction to core AD pathology comparing DTI-ALPS values in patients with AD and patients with prodromal stages of AD pathogenesis, such as patients with mild cognitive impairment (MCI) and subjective cognitive decline (SCD), as well as healthy individuals [16,20,42,43,44,45]. Glymphatic dysfunction was reported in symptomatic AD compared to healthy individuals [42]. Abnormal DTI-ALPS is already present in patients with SCD suggesting that glymphatic dysfunction may be a prodromal symptom at a very early preclinical-risk stage [20] supported by PVS enlargement even before perivenous efflux impairment [19]. Lower DTI-ALPS values were strongly associated with greater amyloid burden [19,20,42,43] and reduced cognitive function [19,20,42]. Interestingly, Aβ and tau burden partially or fully mediate the association of low DTI-ALPS index to progression from prodromal stages to an AD phenotype [16,43] and deterioration of cognitive performance [20,43]. Lower DTI-ALPS index is further associated with enlarged PVS [45], while free water in white matter was correlated with reduced scores in cognitive and functional scales [19]. No significant risk associations were observed when patients with MCI were stratified by sex, age or APOE ε4 status [16].

These findings suggest a clinically meaningful association between reduced water diffusivity and neuronal loss potentially linking glymphatic dysfunction to AD pathology. Glymphatic dysfunction in AD is further supported by a two-stage proteomic analysis in patients with AD and MCI that shows that proteins associated with the proper function of glymphatic system are mainly involved in neurodegenerative processes, immunity and inflammation [46].

2.2. Parkinson's Disease (PD)

Deficiency of AQP4 in established PD experimental models has been shown to drive hypersensitive responses in mice following stimulation with neurotoxic agents [47,48]. These reactions were suggested to be due to increased astrogliosis, since an increased number of astrocytes was observed following AQP4 deficiency in basal as well as neurotoxic conditions [49]. Genetic evidence from PD patients further supports the role of AQP4 in the disease pathogenesis, with patients carrying the AQP4 rs68006382 being associated with faster transition to lower cognitive performance [50].

Reduced clearance of SNCA has been observed following deletion or pharmacological inhibition of AQP4 in experimental models overexpressing the human A53T-SNCA [7]. Accumulated SNCA was associated with neuronal loss and exacerbated symptoms. Additionally, in PD models where SNCA preformed fibrils (PFFs) were injected intrastriatally, glymphatic flow stagnation was associated with behavioral impairments [51]. Furthermore, blockage of meningeal lymphatic drainage, via cervical lymph nodes ligation, resulted in impaired clearance of SNCA and depolarization of AQP4, further highlighting the impact of the clearance mechanisms in PD pathogenesis [52].

A direct feedback loop between glymphatic function impairment and SNCA aggregation has been proposed. Overexpression of A53T-SNCA in mice has been shown to suppress glymphatic flow via modulation of the expression and depolarization of AQP4 [7]. Subsequent blockage of PVS by SNCA aggregates may not further permit the glymphatic efflux. Alternatively, targeted expression of mutated human SNCA in astrocytes drives the loss of AQP4 perivascular localization associated with enhanced astrogliosis in the brainstem and spinal cord of symptomatic mice potentially linking neuroinflammation and glymphatic system dysfunction [53].

Glymphatic function has been consistently reported to be deregulated in PD patients, commonly reflected by a lower DTI-ALPS index compared to matched healthy individuals across cross-sectional and longitudinal cohorts [23,54,55,56,57]. Lower DTI-ALPS index is associated with greater motor severity, as measured by higher Unified Parkinson's Disease Rating Scale (UPDRS)-III scores [55,57,58], whereas baseline glymphatic dysfunction may predict motor dysfunction progression, with lower DTI-ALPS associated with faster worsening of UPDRS-III scores [23] and increased risk of clinical milestones, such as wheelchair dependence [59]. Furthermore, lower DTI-ALPS is associated with worse global cognition, as measured by lower scores in the Mini-Mental State Examination (MMSE) and Montreal Cognitive Assessment (MoCA) scales in several PD cohorts [57,58], while baseline DTI-ALPS values can predict longitudinal cognitive decline [57]. However, findings from other PD cohorts have not been able to replicate this association [23].

Lower DTI-ALPS values, further, seem to decline with advancing age, with older PD subgroups demonstrating stronger negative correlations [23]. Increased age at disease onset is also associated with decreased DTI-ALPS values, suggesting that aging-related glymphatic dysfunction may be associated with the disease onset [56]. Moreover, impaired glymphatic function has been linked with sleep disturbances [58,60] possibly associated with prodromal PD phenotypes although this association is emerging more clearly in clinically meaningful PD subgroups [58]. Female patients have shown higher DTI-ALPS values compared to males at baseline and after of one-year of follow-up [23]; however, other studies do not confirm sex-specific differences [58].

A potential direct link between the glymphatic system impairment and the subsequent neurodegeneration in PD patients may be speculated based on indications about a positive association of the DTI-ALPS index with regional cortical volumes [54]. Mediation analysis demonstrated that cortical degeneration partially mediates the association between glymphatic dysfunction and cognitive impairment. Given that accumulation of SNCA is most often detected in the regional cortex, this evidence suggests a bidirectional mechanism through which SNCA aggregates may disturb the AQP4 anchoring in the astrocytic endfeet, resulting in glymphatic clearance abolition, further promoting its own aggregation.

2.3. Non-AD Dementia

Categories of non-AD dementia encompass a variety of conditions that cause cognitive decline, distinct from AD. Among several conditions, such as frontotemporal dementia (FTD) and idiopathic normal pressure hydrocephalus (iNPH), the occurrence of glymphatic system dysfunction is suggested. The FTD group characterizes brain disorders that primarily affect the frontal and temporal lobes with the most common clinical syndrome being the behavioral variant FTD (bvFTD) which is associated with atrophy of the frontal and anterior temporal lobes [61]. The pathology of FTD is characterized by intracytoplasmic inclusions of aggregated neuronal and glial proteins, such as tau and TDP-43, that eventually lead to disruption of neuronal homeostasis and subsequent cell death [62,63]. On the other hand, iNPH pathogenesis is associated with altered CSF dynamics, enlarged brain ventricles and vascular abnormalities [64].

Experimental findings from studies employing vascular/mixed-dementia experimental models including the microscopic microinfarcts model [65], the hyperhomocysteinemia model of vascular cognitive impairment [66] and a mixed dementia rat model, namely SHRSP/FAD rats [67] point towards a systemic glymphatic dysfunction. The polarization of AQP4 was found to be disturbed in all experimental models while associations with cognitive decline [66], reactive gliosis [68], dilated PVS and delayed waste clearance [65] were observed. These initial findings suggest that glymphatic system dysregulation may emerge as a unifying contributor across diverse dementia types.

Significantly higher levels of AQP4 have been reported in FTD patients compared to subjects not affected by neurodegenerative diseases. A significant, positive correlation was further detected between AQP4 and total tau levels pointing towards enhanced astrogliosis [69]. On the contrary, decreased expression [70] and reduced density of AQP4 in astrocytic endfoot membranes has been reported in biopsies of iNPH patients compared to reference individuals [71].

Glymphatic dysfunction has been shown in patients with FTD syndromes and genetic FTD [22,72,73]. Particularly for bvFTD, increased waste accumulation suggestive of glymphatic abnormality has been reported in characteristic areas of FTD pathology, such as the forebrain and midbrain areas [72]. Across FTD variants, DTI-ALPS abnormalities are subtype-specific and are related to whole-brain white matter hyperintensities, cognitive performance and cortical thickness, suggesting that worse glymphatic function is linked to specific patterns of cortical neurodegeneration [73]. Furthermore, assessment of glymphatic function across symptomatic/presymptomatic mutation carriers and non-carriers, demonstrated that the DTI-ALPS index was lower in symptomatic compared to presymptomatic patients and non-carriers [22]. The index was further associated with disease severity and faster worsening suggesting that glymphatic dysfunction may predispose the disease course based on the different genetic background and disease stage [22].

Similar findings have been reported for cohorts of patients with iNPH. Clearance of gadobutrol has been reported to be delayed in iNPH patients compared to healthy individuals showing increased persistence in the brain parenchyma [74,75,76]. Compared with non-cognitively affected patients, iNPH patients show higher medial temporal lobe atrophy scores and ventricular enlargement (Evans’ index) and reduced entorhinal cortex thickness, tying impaired glymphatic clearance with structural vulnerability in memory-related regions [76]. These observations suggest that reduced clearance of toxic metabolites may underly pathology in non-AD dementia.

2.4. Multiple Sclerosis (MS)

Increased expression and loss of perivascular localization of AQP4 has been shown in the brain and spinal cord tissues in the most commonly used experimental model of MS, experimental autoimmune encephalomyelitis (EAE) [77,78,79]. Deficiency of AQP4 in myelin oligodendrocyte glycoprotein (MOG)-induced EAE, as well as in adoptive transfer models, was associated with mitigated neuroinflammation and diminished clinical symptoms compared to wild-type littermates [80]. Moreover, following adenovirus-mediated overexpression of AQP4 in astrocytes, the latter acquired an enhanced pro-inflammatory profile, further reinforcing the role of AQP4 in mediating the MS inflammatory cascade [80].

Widespread parenchymal localization of AQP4 is associated with impairment of parenchymal CSF circulation in the spinal cord of EAE-induced mice [79]. Although deceleration of glymphatic flow could be plausibly based on the heavy inflammatory burden of MS or by the hypersecretion of CSF [81], an increasing body of research points towards glymphatic system dysfunction in patients with MS. Reduced DTI-ALPS values have been consistently reported to be reduced in cohorts of MS patients compared to age- and sex-matched healthy individuals [17,82]. Although glymphatic deregulation is apparent both in the relapsing remitting and the progressive phenotype of MS, an enhanced decline is observed for the latter, suggestive of a parallel trajectory along clinical course [83]. Lower DTI-ALPS index is further associated with worse clinical disability, as estimated with the Expanded Disability Status Scale, and longer disease duration [17,83].

Impaired glymphatic function seems to co-occur with structural brain damage and global brain volume changes. Lower DTI-ALPS index was reported to be associated with higher white matter and cortical lesion volume, deep grey matter atrophy and reduced fractional anisotropy [17]. Additionally, larger choroid plexus volume correlates with lower DTI-ALPS index in patients with MS, while both indices were associated with higher total, periventricular, and juxtacortical white matter lesion volumes and lower brain, thalamic and cortical volumes [82]. Interestingly, the DTI-ALPS index was reported to partially mediate the association between increased choroid plexus volume and structural brain changes, suggesting that glymphatic impairment links altered CSF dynamics to tissue damage [82].

Choroid plexus enlargement may represent a key factor in glymphatic system dysfunction in MS, potentially driven by increased infiltration of inflammatory cells and soluble mediators [84,85]. Given that choroid plexus is suggested to produce the vast majority of CSF, any structural distortion may potentially lead to stagnation of CSF flow [27]. Given that choroid plexus enlargement takes place during the early stages of MS, correlations with increased presence of immune cells [86], clinical disability and cognitive dysfunction [87] point towards a more progressive phenotype [27].

Collectively, MS-specific aberrations in choroid plexus volume may confer to persistent inflammation that may lead to the impediment of glymphatic flow and subsequent impairment of the clearance of immune toxic mediators [82]. Impaired glymphatic function is associated with both demyelinating and neurodegenerative events, therefore the exact impact of glymphatic system deregulation in the MS pathogenesis warrants further investigation.

2.5. Autoimmune Encephalitis

Autoimmune encephalitis is an umbrella term for multiple disorders that are characterized by the immune system activation against CNS epitopes such as N-methyl-D-aspartate receptor [88]. Concrete findings regarding the expression of AQP4 in autoimmune encephalitis are lacking, although recently the rare coexistence of autoantibodies against neuronal or glial epitopes representative of autoimmune encephalitis and against AQP4 has been reported [89]. This clearly distinguishes the disease mechanisms of autoimmune encephalitis from neuromyelitis optica spectrum disorder.

Only one study has recently evaluated the glymphatic function in patients with autoimmune encephalitis [90]. Patients with autoimmune encephalitis had lower DTI-ALPS values compared to age- and sex-matched healthy individuals. Low baseline DTI-ALPS index was associated with a decline in cognitive function, as estimated with MMSE and MOCA scales, while an increased risk of long-term cognitive impairment was linked to lower DTI-ALPS index. These observations were confirmed across different subgroups depending on the target of the autoimmune pathogenesis, while the increased expression of pro-inflammatory cytokines partially mediate the association of glymphatic function deterioration with cognitive decline [90]. Although preliminary, these observations suggest a link between glymphatic system impairment and cognitive decline with inefficient clearance of pro-inflammatory cytokines.

2.6. Stroke

The presence of abundant toxic extracellular fluids in areas of liquefactive necrosis leads to atrophy, cytotoxic edema, and neuronal loss in regions adjacent to the infarct for weeks following stroke [91]. Therefore, potential dysregulation of waste clearance mechanisms during stroke may contribute to the impartial recovery. The glymphatic flow could be plausibly hampered from the rapid inflow of blood and CSF in the PVS resulting in brain edema and tissue swelling [92,93,94]. Evidence points to the increased enlarged PVS burden, particularly in the basal ganglia and hippocampus, as a significant risk factor of incident stroke and intracerebral hemorrhage (ICH) [95,96]. Additionally, patients with spontaneous ICH and increased size of PVS in centrum semiovale demonstrate increased vascular amyloid burden [97] while in patients with subarachnoid hemorrhage (SAH) a reduced serum/CSF ratio of neurofilament light polypeptide has been observed [98]. These findings point towards impaired glymphatic efflux and reduced leakage to the draining lymph nodes.

The expression of AQP4 in SAH, ICH and ischemic stroke experimental models has been reported to increase very early during induction [99], suggesting its crucial role in mediating glymphatic function integrity. However, it is difficult to define the role of AQP4 in stroke and related conditions due to the contradictory results reported following its deficiency in experimental models. Knockout of AQP4 has been shown to cause enhanced neuronal loss and BBB disorganization following SAH induction [93,100], brain edema [101], as well as pronounced microglial activation in the middle cerebral artery occlusion (MCAO) model compared to wild-type mice [102]. On the other hand, cerebral edema is reported to decrease in AQP4 knockout mice early post MCAO induction [103,104] while reduced infarct volumes following APQ4 inhibition have been also reported [105]. Therefore, defining the role of AQP4 and subsequently the regulation of glymphatic flow in stroke warrants further research.

However, experimental evidence suggests that glymphatic dysfunction occurs during stroke and related conditions. Rodent models of SAH, whether induced by endovascular perforation or blood injection into the cisterna magna, are characterized by a significant and prolonged disruption of CSF circulation and glymphatic function [93,106,107,108] as well as reduced drainage to deep cervical lymph nodes associated with AQP4 depolarization [106]. Findings from a non-human primate model of SAH further corroborates the dysregulation found in rodent models [109]. Glymphatic system dysfunction during the acute phase of multiple experimental models seems to predominantly occur in the substantia nigra and the ventral thalamic nucleus (VTN) areas [110]. Deregulation of glymphatic flow has been associated with accumulation of tau protein and immune cells, glial cell activation, neuroinflammation, neuronal apoptosis [106] and cognitive dysfunction [111] representing extended neuropathological damage.

Accumulating neuroimaging evidence indicates that glymphatic function is consistently impaired in stroke patients, most inferred from reductions in the DTI-ALPS index. Across ischemic and hemorrhagic stroke subtypes, patients demonstrate lower DTI-ALPS values compared to healthy individuals [112,113,114,115,116,117] and disease-specific subgroups [21], supporting the presence of glymphatic dysfunction post-stroke.

In patients with subacute ischemic stroke, significantly lower DTI-ALPS values compared to healthy subjects indicate reduced perivascular fluid transport after ischemic injury [114]. During the chronic phase (3–12 months), stroke survivors demonstrated persistent reduced DTI-ALPS values versus healthy subjects at both 3 months and 1 year, suggesting prolonged glymphatic disturbance beyond the acute/subacute period [116]. This observation supports the existence of long-lasting glymphatic dysfunction, particularly in the hemisphere ipsilateral to infarction [113,114,116,118]. Similar, glymphatic function impairment has been demonstrated in the lesion compared to contralateral side in patients with spontaneous ICH [112]. Patients with cerebral infraction classified in a good prognosis group based on modified Rankin Scale had increased DTI-ALPS values compared to those classified in the poor prognosis group [21]. Preserved glymphatic function was associated with better neurological outcomes and better psychological recovery pointing towards a prognostic value of DTI-ALPS [21]. Particularly for patients with cerebral small vessel disease (CSVD), a condition that causes approximately 20% of strokes [119], reduction in DTI-ALPS indices correlated with enlarged PVS in basal ganglia and existence of lacunes [120], cerebral venous integrity [121] and vascular risk factors [117].

Findings from these studies suggest an association between glymphatic dysfunction and motor and cognitive functions. In subacute ischemic stroke, the DTI-ALPS index correlated positively with Fugl-Meyer motor scores, supporting a relationship between greater glymphatic impairment and worse motor function [114]. Furthermore, the DTI-ALPS index was correlated with 6-month post-stroke cognitive impairment [113]. In concordance with this, the DTI-ALPS index has been reported to be useful for prediction of future dementia risk in patients with lacunar strokes and confluent white matter hyperintensities [122]. Weak correlations also emerged at 3 months between the DTI-ALPS index in the lesioned hemisphere and motor and memory scales, although this correlation could not be detected at 1 year of follow-up, suggesting that the correlation may be transient [116].

The time since stroke onset and stroke duration are among the factors that suggest a time-dependent relationship between glymphatic activity and disease course. Values of DTI-ALPS increase along time since disease onset [113,116,118] while in the sICH cohort, the lesion-side DTI-ALPS index was correlated with disease duration [112]. These observations may represent the activation of organism towards partial recovery after early impairment due to potential remodeling or reversibility of glymphatic function during rehabilitation [113,118]. Application of standard reperfusion therapy results also in partial recovery of glymphatic function [123], further reinforcing its importance in the therapeutic landscape of stroke.

2.7. Traumatic Brain Injury (TBI)

Widespread and persistent disruption of glymphatic system function has been demonstrated in several experimental models of TBI, characterized by impaired CSF influx and efflux very shortly post-injury [124]. This disruption could endure up to 10 weeks in adult mild TBI models particularly in areas such as the olfactory bulb [125,126]. Disruption of normal brain efflux kinetics is further supported by the reduced drainage to the cervical lymph nodes [124]. Glymphatic impairment was linked to loss of motor coordination and balance deficits. Mechanistically, increased noradrenaline levels released due to post-traumatic sympathetic overactivation seem to contribute to the stagnation of glymphatic flow by suppressing vessel contractility, leading to edema and impaired debris clearance [127].

Deregulation of clearance has been associated with alterations in AQP4 expression and polarization [128,129]. However, the role of AQP4 seems to be quite complex. Reduced AQP4 expression is reported in posterior cortical regions and the ventral cortex [124] although other reports have noticed the upregulation of the global AQP4 expression even if depolarized [130,131]. Deficiency of AQP4 expression and loss of astrocytic polarization was directly linked to impaired solute clearance leading to phosphorylated tau (P-tau) accumulation in both the contralateral and ipsilateral hemispheres, exacerbated neurodegeneration and cognitive deficits [132]. On the other hand, small interfering RNA-mediated inhibition of AQP4 significantly reduced brain edema in the acute phase post-TBI (6 to 12 hours) [133]. Temporal patterns of AQP4 expression post-TBI provide further insights into this controversy. Levels of AQP4 seem to peak early during the acute phase (12 and 72 hours post-TBI) and decline later (by day 15) [130]. These changes are closely associated with the volume of brain water content and the severity of the edema. Notably, although AQP4 has an important role in edema formation, in the long-term its function is compensatory assisting in its resolution once the BBB stability has been restored [131]. However, others have noticed the emergence of glymphatic clearance dysfunction around 28 days post-injury in a blast mild TBI model [129].

Across studies of mild TBI (mTBI), subacute TBI, and chronic TBI, glymphatic function was assessed using the DTI-ALPS index and the estimation of ePVS burden. A consistent impairment of glymphatic function was reported [134,135,136,137]. Reduced DTI-ALPS index correlated with decline in cognitive performance [135] including verbal memory, attention, and executive function [134]. Of note, the DTI-ALPS index may have a prognostic utilization for identifying chronic cognitive impairment following TBI [26], while the cognitive decline was not related to factors reflecting clinical burden at the baseline [135]. Collectively, these findings provide cues considering a link between the DTI-ALPS index and cognitive decline that relies predominantly in glymphatic function.

Alongside reduced DTI-ALPS values, patients with mTBI had profound global white matter damage [134]. Global white matter damage partially mediated the association between the DTI-ALPS index and cognition. Furthermore, the DTI-ALPS index was negatively correlated with plasma neurofilament light (NfL) expression, linking reduced DTI-ALPS to the severity of axonal damage biology [137]. These observations suggest that impaired glymphatic activity is present in more severe or diffuse injury phenotypes [134,136,137] pointing towards a relationship between glymphatic dysfunction and diffuse white-matter integrity to explain cognitive deficits after mTBI [134].

On the other hand, reactive glymphatic hyperactivity has also been reported to occur post-injury. Higher DTI-ALPS index in mTBI compared to non-injured subjects, including MRI-negative mTBI patients [138] has been reported. A gradual increase of the DTI-ALPS index from the acute (14 days post-injury) to the chronic phase (6–12 months post-injury) [26] was noticed. These observations may reflect a dynamic regulation of glymphatic activity in turn of neuronal damage and the activation of a compensatory clearance pathway working towards partial recovery. This is supported by improvements in sleep quality that are associated with glymphatic hyperactivity [26]. Additionally, these dynamic fluctuations of glymphatic activity in TBI renders the DTI-ALPS index as a very useful marker in interpreting the correlation between time- and context-dependent shifts in perivascular diffusivity and clinical outcomes.

Collectively, these studies reinforce the growing understanding that glymphatic dysfunction and AQP4, particularly, play a significant role in TBI pathophysiology. Chronic impairment of glymphatic function post TBI may render post-traumatic brain prone to the accumulation of neurotoxic proteins that lead to neurodegeneration [132]. This is supported by the restoration of the glymphatic function exerted by some of the therapeutic strategies applied in experimental models such as Omega-3 polyunsaturated fatty acids [139], adrenergic blockers [127] and GLP-1R agonists [140] which are associated with BBB preservation, improvement of neurological outcomes and reduced accumulation of inflammatory components.

2.8. Amyotrophic Lateral Sclerosis (ALS)

Studies in transgenic animals expressing mutated human superoxide dismutase (SOD) 1, that mimics ALS pathogenesis, have revealed broad changes in astrocytes homeostasis; deregulated expression of AQP4 was accompanied by depolarization [141]. These pathological features were suggested to have contributed to earlier disease onset [142] potentially due to altered CSF dynamics [143]. Glymphatic system dysfunction during early stages of ALS is further supported by a study in transgenic mice expressing cytoplasmic human TDP-43 (hTDP-43ΔNLS) under the control of doxycycline [144]. Following the cessation of doxycycline feeding, clearance of an MR agent (Magnevist) was evaluated using dynamic contrast-enhanced MRI. Altered glymphatic function was observed in transgenic mice three weeks after doxycycline withdrawal, indicating disruption at a relatively early stage in the neurodegenerative process, estimated to begin around four weeks. Subsequently, transgenic mice exhibited signs of progressive neurodegeneration in the primary motor cortex, the primary somatosensory cortex and the corticospinal tract [144].

Across recent human ALS studies, glymphatic function has been assessed using non-contrast MRI biomarkers, most commonly DTI-ALPS and estimation of choroid plexus volume fraction. Multiple datasets demonstrate lower DTI-ALPS index in patients with ALS compared to age- and sex-matched healthy individuals [145,146,147,148] and patients with other motor neuron diseases [149], consistent with impaired perivascular water diffusivity. A longitudinal study confirmed that the glymphatic impairment could be a permanent phenomenon, lasting up to 8 months [149]. Interestingly, similar to stroke, unilateral lowered DTI-ALPS index has been reported, with the right-side being significantly reduced in ALS patients compared to healthy individuals, while the left side was not affected [147]. These observations demonstrate that glymphatic impairment may contribute substantially to neuronal damage in ALS in a unique manner.

Clinical and demographic factors affect glymphatic activity in ALS patients. The DTI-ALPS index has been shown to decline with advancing age [147,149] while choroid plexus volume fraction is also age-associated [146] potentially linking aging with vulnerability to ALS pathology due to gradually impaired glymphatic system. Alternatively, increased choroid plexus volumes may represent compensatory CSF production due to reduced waste clearance [146]. Moreover, female ALS patients were shown to have higher mean DTI-ALPS values compared to their male counterparts [147]. Positive association with metrics of clinical disability [148,149] reflecting sensorimotor dysfunction and disease duration [145] has been also reported, suggestive of a link between glymphatic and functional neural impairment with disease severity. However, others found no correlation with King’s stage, cognitive function questionnaires and other clinical parameters [146,147]. Last, glymphatic dysfunction in ALS patients has been associated with sleep disturbances [148].

Collectively, glymphatic impairment in early stages of ALS pathogenesis may substantially contribute to the disease burden. Therefore, clinical assessment of glymphatic function could potentially provide cues for designing a better therapeutic approach earlier in the disease course.

3. Inflammaging

Enhanced depolarization of AQP4 as well as altered CSF dynamics point towards a deregulation of glymphatic system during aging. Increased expression of AQP4 with advancing age has been detected in human and mice [150,151]. Staining patterns of AQP4 in old, non-demented individuals shows a distinct diffuse parenchymal pattern compared to the perivascular polarization observed in younger individuals. Enlarged PVS and increased water pool is also observed in aging brains [152] pointing towards a reduction in the CSF turnover [10] and impediment of CSF flow [153]. Besides impaired glymphatic flow, draining to meningeal lymphatic vessels was found also to be related to aging in patients undergoing 3-dimensional T1-weighted imaging following intrathecal injection of gadolinium [154]. These observations suggest a holistic glymphatic flow stagnation in aged individuals and related experimental models.

Although limited, evidence points towards an association between glymphatic impairment and aging that leads to neuroinflammatory conditions. The presence of excessive immune cells in PVS due to chronic low-grade inflammation [155] as well as the reduced arterial pulsatility [12] could explain mechanistically the reduced CSF flow. Immune-mediated loss of meningeal drainage capacity has been observed in old mice [156]. Moreover, a substantial decline in the fluid exchange between CSF and the brain parenchyma in mice of advanced age (18 months old) compared to younger ones (2-3 months old) was accompanied by reduced clearance of injected Aβ and loss of perivascular polarization of AQP4 [8]. Of interest was that AQP4 was found to be localized in the areas that affect the recirculation of CSF into and through the brain interstitium while its expression was elevated in areas where the penetration of CSF tracer was reduced, such as the pial surface. Depolarization of AQP4 is also associated with age-related cognitive decline and increased Aβ burden [28], suggesting AQP4-dependent cognitive deterioration. Although these are preliminary evidence, a feedback loop could be suggested: chronic inflammation leads to the accumulation of immune mediators within CNS hampering the glymphatic function that subsequently fosters protein aggregation resulting in enhanced neuroinflammation and further obstruction of glymphatic flow (Figure 2). This theoretical mechanism could thus contribute to the onset of neurocognitive and neurodegenerative conditions.

4. Gut-Brain Axis

Recent evidence points to a possible interplay between the gut-brain inflammatory axis and the dysfunction of the glymphatic system within CNS. Indirect effects of gut dysbiosis may have an impact in the glymphatic function via several mechanisms [157]. Particularly, gut dysbiosis may result in the accumulation of pro-inflammatory mediators in the periphery as well as the activation of peripheral immune cells. Following their migration within CNS, signaling effects in local glial cells may enhance the neuroinflammatory burden that has the potential to halt the glymphatic flow [5]. Besides blood circulation, additional routes, such as the vagus nerve, allow for direct communication between the gut and the brain thus enabling the local build-up of inflammation that may result in the aggregation of neurotoxic peptides via a feedback loop [158]. Moreover, gut microbiota regulates the release of neurotransmitters and neuroactive hormones that may perturb astrocytic functions rendering them incapable for eliciting their role in the regulation of glymphatic flow and waste clearance [159]. Gut dysbiosis may also lead to dysregulation of BBB integrity through the release of short-chain fatty acids [160]. Given that BBB is suspected to serve as a complementary mechanism to the glymphatic system [161], its dysregulation can compromise the waste removal machinery. The composition of gut microbiota is further associated with the regulation of circadian rhythms, which are directly related to the activity of the glymphatic system. Indeed, microbiota dysbiosis may generate potent activators of aryl hydrocarbon receptors which are able to disrupt circadian rhythms via sleep fragmentation [162] and subsequently impair glymphatic flow [163].

We can further speculate on the possible role of gut dysbiosis in the aggregation of neurotoxic proteins within the CNS focusing on studies that employ a gut-targeted modification of microflora. Gut dysbiosis induced by administration of periodontal pathogen in mice has been shown to be associated with increased Aβ accumulation alongside systemic and focal CNS inflammation and reactive astrogliosis [164]. Additionally, dextran sodium sulfate-induced colitis in aged mice led to extended gut inflammation that was accompanied by increased astrogliosis, impaired glymphatic clearance, depolarization of AQP4 and increased Aβ accumulation [165]. In this specific experimental model, glymphatic function was restored and neuroinflammation was mitigated following pharmacological inhibition of the NLRP3 inflammasome suggesting an interconnectedness of gut-induced neuroinflammation and glymphatic dysfunction.

Existing evidence points towards a hypothetic mechanistic interplay between secreted metabolites by the gut and CNS homeostasis, with waste clearance within the CNS being one of the domains affected (Figure 2). However, more research is warranted to confirm that this existing interaction may lead to the deregulation of waste clearance from the brain.

5. Regulation of Glymphatic Function

5.1. Sleep

Injected trackers in the cisterna magna or the choroid plexus have yielded initial findings supporting that clearance of solutes from the brain is accelerated in either natural sleeping or anesthetized mice compared to awake mice while sleep deprivation led to reduced CSF influx [166,167]. Increased CSF influx is proposed to occur due to the expansion of ISF volume, thus facilitating the fluid interchange [168]. Furthermore, sleep fragmentation induced via tilting platforms in 5X FAD mice resulted in compromised short-term and spatial memory accompanied by increased accumulation of Aβ and astrogliosis [169]. The expression of AQP4 was deregulated both in wild type and 5X FAD mice following sleep deprivation suggesting that dysfunction in the glymphatic system may be associated with cognitive deficits.

Using the DTI-ALPS index, Saito and colleagues have recently shown that glymphatic activity was significantly lower in a group of patients with sleep disruption compared to individuals with normal sleep [170]. Furthermore, impaired glymphatic function has been demonstrated in patients with obstructive sleep apnea [171]. Altered glymphatic function has been further correlated with chronic sleep disturbance in patients with iNPH compared to non-demented individuals [172].

Altogether, these findings support the existing hypothesis that adequate sleep keeps the glymphatic flow active maintaining brain homeostasis. Of note, sleep disturbances have been associated with behavioral disturbances in PD patients [173], tau accumulation and faster cognitive deterioration in AD patients [174]. Enhanced vasomotion during sleep may be the key mechanism for explaining sleep-mediated glymphatic clearance enhancement [175], although additional research is warranted to unravel the pathways involved.

Anesthesia

Anesthetic state seems to influence glymphatic function through its effects on neural oscillations, vascular pulsatility, and adrenergic tone. Rodents anesthetized with ketamine/xylazine show higher CSF tracer influx into the brain compared to mice under isoflurane alone or other agents [176]. It is not clear though whether elevated CSF influx depends on the increased CSF production, since contradictory results have been published [177,178]. Enhanced influx is coupled to the induction of slow-wave electroencephalogram activity, which promotes low noradrenergic signaling and synchronized neuronal activity that widens perivascular spaces, thereby facilitating CSF-ISF exchange. In contrast, anesthetics such as isoflurane, which generate lighter sedation with higher heart rates and less delta power, do not have a substantial impact on glymphatic transport. Thus, deeper slow-wave oscillatory activity under certain anesthetics appears to facilitate perivascular CSF entry into the parenchyma.

Structural changes following anesthesia also confer to altered glymphatic function. Prolonged or high-dose isoflurane disrupts AQP4 localization, impairing solute efflux and leading to accumulation of inflammatory proteins and Aβ, phenomena associated with postoperative cognitive decline [179]. Restoration of AQP4 polarization could preserve partially waste clearance and neurocognitive performance. Further, co-administration of dexmedetomidine, an α2-adrenergic agonist that reduces locus coeruleus activity, with low-dose isoflurane enhances glymphatic transport compared to isoflurane alone, pointing towards a dependency on the noradrenergic tone and vascular dynamics [180].

These findings underscore a two-sided effect of anesthesia: when high-dose isoflurane is administered or during prolonged exposure, glymphatic function may be stalled, while the usage of anesthetics that promote slow-wave neuronal states leads to maintenance of vascular pulsatility and AQP4 polarization. Reduced adrenergic signaling results in relaxed perivascular spaces and expansion of interstitial spaces [166] that subsequently lowers vascular resistance, thus promoting clearance [180].

Body Posture

Changes in posture may have an impact on CSF circulation [181]. Furthermore, CSF pressure and volume are influenced by posture due to gravity [182]. This is crucial since elevated intracranial pressure has been implicated in several neurodegenerative conditions [183]. Additionally, postoperative cognitive impairment has been associated with the acquired posture during surgery [184]. These findings suggest an interplay between the body posture and altered CSF dynamics that may have an impact on brain functions.

Using MRI and fluorescence tracers in rodents, researchers have shown that Aβ clearance is significantly superior in the lateral position compared to the supine or prone positions suggesting the existence of an evolved mechanism to sustain brain homeostatic glymphatic function [185]. Similar, in human subjects the CSF exchange seems to be enhanced in supine compared to the upright position, providing benefits for brain waste clearance processes [186]. Supine position seems to be ideal for the glymphatic process possibly by increasing the intracranial pressure which subsequently allows for more CSF movements and higher driving pressure [187]. More insights concerning the relationship between body posture and glymphatic function are warranted to elucidate the mechanisms involved.

Alcohol

Alcohol intake has been shown to exert dose-dependent and temporally distinct effects on glymphatic function. Acute and chronic exposure to high levels of ethanol (>1.5 g/kg) significantly suppresses glymphatic transport, leading to impaired CSF influx and reduced clearance of waste metabolites [188,189]. Suppression is associated with astrogliosis, AQP4 depolarization, and reduced cerebrovascular pulsatility thus contributing to compromised Aβ clearance [189]. By contrast, low-dose ethanol exposure (0.5 g/kg) appears to enhance glymphatic activity which is accompanied by reduced GFAP expression and stabilization of AQP4 polarization [188]. Furthermore, in humans, acute moderate alcohol administration induces reversible ventricular enlargement possibly reflecting a shift in CSF dynamics and water volume suggesting a glymphatic system malfunction [190].

Mechanistically, ethanol-induced vasodilation following low exposure to alcohol may enhance vascular reactivity, thereby promoting perivascular clearance of larger solutes and proteins [191]. Higher volumes of alcohol may have a detrimental effect inducing astrocytic amyloid generation and enhanced astrogliosis [192] as well as in vivo depolarization of AQP4 [193]. Increased astrogliosis was also present in human postmortem studies of patients with alcohol use disorder [194].

Collectively, these observations suggest that alcohol modulates glymphatic clearance in a dose-dependent manner. Low-dose ethanol may transiently enhance waste clearance through increased vascular pulsatility, whereas chronic or high-dose exposure results in glymphatic flow stagnation. An overall of lifestyle modulators is depicted in Figure 2.

6. Conclusions

The evidence synthesized in this review provides available data that supports that glymphatic system dysfunction may constitute a common clinically relevant feature across neurodegenerative and neuroinflammatory conditions. However, it is not clear whether glymphatic dysfunction is a mechanism preceding neurodegeneration or is an epiphenomenon of underlying pathogenesis. Dysfunction appears as a vulnerability that links aging, inflammation and neurodegeneration. More importantly, glymphatic dysfunction seems to be detectable in prodromal and early-stage disease phenotypes implying that a waste clearance deterioration may precede initial stages of disease pathogenesis. These findings converge on the notion that glymphatic dysfunction may be among the drivers of disease onset. Therefore, we cannot make concrete conclusions considering the potential role of glymphatic system while we highlight the need for future experiments that will decipher this experimental void in our knowledge.

Altered brain homeostasis due to the impaired glymphatic flow appears potentially to lead to CNS function deficits, including cognitive decline among others. Cognitive deficits dominate findings of human imaging studies indicating a sensitivity of brain regions, such as the hippocampus and temporal cortex, in the accumulated waste [195]. Furthermore, the consistency of glymphatic deregulation among different diseases of the CNS argues for a shared upstream mechanism rather than a disease-specific process. With respect to all domains that need to be validated, waste clearance inability may rise as a core mechanism for neurodegeneration. Variable interplay of glymphatic system with other waste clearance pathways should be explored further to gain insights regarding different pathways of neurodegeneration [196].

Another limitation regarding pre-clinical models is the translation of experimental findings to humans due to species differences; for example humans exhibit higher anatomical complexity compared to rodents [197] while the latter higher AQP4 perivascular polarization compared to the former [198]. However, multiple evidence from human imaging studies across AD, PD, ALS, MS, FTD, stroke, and TBI, consistently demonstrate altered glymphatic function. Therefore, we believe that this issue does not fall within the limitations of this synthesis. In most human imaging studies glymphatic function has been assessed with the DTI-ALPS index which serves as a proxy. Low DTI-ALPS indices are often associated with increased perivascular space burden, and choroid plexus abnormalities. These observations support a model in which impaired clearance of aggregation-prone proteins, inflammatory mediators, and metabolic byproducts contributes to a neuroinflammatory cascade that potentially leads to neuronal dysfunction. It should be noted that the DTI-ALPS is a practical, non-invasive surrogate marker of glymphatic function; however, limitations concerning its application exist. The method assumes that diffusivity along PVS reflects glymphatic activity. Yet, diffusion anisotropy may also be influenced by other factors including white matter microstructure, axonal degeneration, edema, or small vessel disease, potentially confounding interpretation of outcomes [199]. Moreover, DTI-ALPS index is a snapshot of water diffusivity, and not representative of dynamic flow. Therefore, its application can be influenced by diffusion as well as convective flow, which is a lively debate concerning glymphatic biology [200]. Lack of established outputs as well as variability of monitoring methods may also reduce variability until a consensus is reached. Thus, although DTI-ALPS is a promising clinical tool, its outcomes should be interpreted with caution and in combination with other endpoints to properly assess glymphatic function in clinical routine. Future experiments should be focused on the functional improvement of DTI-ALPS as well as the development of robust and reliable diagnostic markers for the assessment of glymphatic capacity in humans.

Glymphatic function seems to be strongly dependent on the function of AQP4. Through our synthesis the role of AQP4 is framed as a central facilitator of glymphatic transport due to its polarized expression at astrocytic perivascular endfeet. Initial experiments using Aqp4 knockout mice have supported this mechanistic link [1]. However, subsequent studies employing models of Aqp4 deletion have questioned its role in the magnitude of parenchymal tracer spread [201]. Additionally, others have supported those technical issues such as the type of anesthesia used, and the timing as well as the tracer delivery route may influence our understanding of the role of AQP4 [202]. Thus, while AQP4 likely enhances perivascular CSF–ISF exchange by increasing astrocytic water permeability, whether it is essential for large-scale convective glymphatic flow or instead modulates local mixing and dispersion remains unresolved.

Age-associated structural changes in the vascular integrity, alterations in astrocytes physiology with prominent AQP4 depolarization and systemic inflammation caused by gut dysbiosis may result in gradually accumulated inflammatory burden and loss of BBB integrity. Glymphatic dysfunction driven by aging, alongside these background conditions may render the CNS environment vulnerable for neurodegeneration onset. This bidirectional interaction suggests that inflammaging and glymphatic failure work in parallel, providing a theoretical mechanism for explanation of the strong age-dependent risk of progression of neurodegeneration. Within this model, glymphatic dysfunction may result in retention of immunogenic components thus amplifying inflammatory signaling, thereby accelerating pathogenic cascades. In the opposite direction, accumulated inflammatory burden may result in glymphatic efflux stagnation further enhancing the presence of toxic metabolic byproducts and aggregation-prone proteins that cause neuronal damage.

7. Future Directions

Collectively, imaging evidence positions monitoring of glymphatic function as a promising and powerful accessory tool for clinical management of patients with neurodegenerative diseases [2]. Development of biomarkers based on non-invasive MRI-based methods, particularly DTI-ALPS, may offer robust and widespread longitudinal options for the assessment of brain clearance function. Further standardization of acquisition and analysis protocols, as well as the establishment of a consensus among physicians considering normative reference ranges, and assessment of treatment response would entitle the implementation of glymphatic system in the armamentarium of neurologists. This way earlier detection of brain vulnerabilities and risk stratification-based prognosis could thus be enabled. From a therapeutic perspective, therapeutic strategies for enhancing glymphatic clearance could include interventions that improve sleep quality, reduce systemic and CNS inflammation burden and preserve the established astrocytes interactions and AQP4 polarization. Eventually, harnessing the modalities of the glymphatic system could open the path for the development and repurposing of cross-disease therapeutic strategies given its involvement in multiple neurodegenerative pathways.