Post-stroke cognitive impairment (PSCI) represents a devastating consequence of cerebral infarction (1), with its severity closely associated with lesion topography while demonstrating substantial heterogeneity in cognitive outcomes based on infarct location. Strategic infarctions in the left frontotemporal lobes, thalamus, and right parietal lobe exhibit particularly strong correlations with PSCI development (2).

Cerebral infarction triggers cytotoxic/vasogenic edema, neuroinflammation, and blood–brain barrier disruption (3), which collectively promote secondary neurodegeneration and cognitive decline. The pathological cascade following stroke, including impaired clearance of metabolic byproducts such as β-amyloid (Aβ) (4), tau protein (5), lactate (6) and iron (7), contributes significantly to PSCI pathogenesis, though the underlying mechanisms remain incompletely understood. In recent years, further explore pathophysiological mechanisms of PSCI based on the rise of neuroimaging research (8, 9), may facilitate the acquisition of imaging markers of PSCI and the development of novel treatments for restoring cognitive decline.

Known as the glymphatic system (GS), a pathway for drainage and exchange between cerebrospinal fluid (CSF) and interstitial fluid (ISF). CSF flows into the brain parenchyma along the perivascular spaces of arteries and exchanges with ISF in the GS, subsequently draining out along the perivascular spaces of veins (10). Available studies have found that the glymphatic system affects the pathological changes and outcomes of stroke (11). Based on studies in animal models of post-stroke dementia, chronic cerebral hypoperfusion impairs glymphatic system function, reduces clearance of tau protein from the brain interstitial fluid, and triggers its neuropathological aggregation in peri-infarct regions (5). This demonstrates that impaired glymphatic clearance serves as a critical link connecting post-stroke vascular injury to pathological tau accumulation. Further studies are needed to confirm the relationship between the glymphatic system and PSCI. Treatment strategies to improve the clearance of waste removal may also be a promising approach to improve cognitive impairment and the prevention of PSCI.

Diffusion tensor imaging analysis along the perivascular space (DTI-ALPS) is a quantitative assessment of the function of lymphatic flow and clearance along the perivascular space (PVS) (12). A successful application in the exploring the pathophysiology of Alzheimer’s disease (AD) (13), Parkinson’s disease (PD) (14), headache (15) has been reported. Some reported that DTI-ALPS showed lower values in ischemic stroke, suggesting impaired lymphatic function. After the initial injury, the DTI-ALPS index increased with time after stroke onset, suggesting a recovery of lymphatic function (16). Another study found that the DTI-ALPS index was independently positively associated with episodic memory in cerebral small vessel disease (CSVD) patients and was an early potential marker for recognition of cerebral infarction (17). PSCI affects 35–47% of survivors, yet its pathophysiological mechanisms remain elusive (18). Emerging evidence implicates glymphatic dysfunction in neurodegenerative protein clearance (19, 20), but its role in acute stroke-related cognitive decline is poorly understood. We hypothesize that DTI-ALPS, a novel imaging biomarker of glymphatic activity, may bridge this gap.

Cortical infarctions (involving the frontal, temporal, or parietal lobes) can directly impair cognitive domains. Additionally, due to their typically larger lesion volumes, cortical infarcts may exert mass effects that compress perivascular spaces, thereby impairing glymphatic clearance (21, 22). This mechanism could exacerbate post-stroke cognitive dysfunction independent of direct neuronal damage. To reduce confounding bia, we selected patients with subcortical infarction. In this study, to quantitatively investigate the pathological characteristics of GS in subcortical infarct patients and the relationship with cognitive impairment by non-invasive DTI analysis method, we consider that glymphatic dysfunction after subcortical infarct may be related to early cognitive dysfunction.

This is a positive result in our study that the DTI-AIPS index of subcortical infarct patients was significantly lower than that of HC group, suggesting that the changes of cerebral glymphatic function in patients with stroke. Meanwhile, the DTI-ALPS index on the lesion side was correlated with the MoCA score of day-7 post-stroke, indicating that the DTI-ALPS index was associated with early cognitive function changes. This study found a correlation between early altered in the GS and cognitive impairment after stroke, suggesting that the GS may play an important role in the early changes in cognition function following a stroke. Active interventions targeting GS function in the early stages may provide effective evidence for the prevention and treatment of PSCI.

The energy depletion of oligodendrocytes caused by cerebral ischemia–reperfusion injury leads to demyelination, axonal degeneration and further apoptosis of neurons on the occurrence of cerebral infarction (25, 26). AQP4 water channels in the foot processes of perivascular astrocytes facilitate water movement from the vascular space across the BBB into cells. The glymphatic system is a highly organized fluid transport system that is capable of removing brain waste, regulating the secretion of neurotransmitters and keeping up neuronal function (27). Damaged GS has been found in AD, hydrocephalus, PD, age-related iron deposits as well as tumor-associated brain edema (16). Iliff et al. (20) found that the AQP-4 gene knockout mice have the significantly decreased influx and clearance efficiency on GS than in controls. Further, Lin et al. (28) hold that the perivascular AQP-4 depolarization and intracranial hypertension after ischemic stroke, leading to increased resistance of ISF outflow pathway, great changes of the clearance efficiency have been made in GS, caused by mainly involving the abnormal protein deposition, may be related to vascular dementia (29).

Studies have shown that age-related cognitive decline is associated with the decreased ability of the GS to clear Aβ (30, 31). At the same time, Eide et al. (32) found that the decrease of clearance efficiency of GS leads to the increase of soluble Aβ and tau protein accumulation in the entorhinal cortex-hippocampal circuit, which promotes the development of dementia in patients with idiopathic hydrocephalus. During the acute phase of stroke, pathological changes such as inflammatory stimulation and oxidative stress lead to a reduction in the clearance of harmful substances by lymphocytes. Our research has found that there is glymphatic dysfunction in acute stroke, which is consistent with other studies (33). It is known that the clearance of Aβ drained through the ISF. Dysfunction of the GS leads to an excess of free Aβ, which accumulates in the brain parenchyma and vascular system, promoting the onset of cognitive impairment (34). We found that alterations in the GS are associated with cognitive decline at 7 days post-stroke. Furthermore, we observed significant correlations between 90 days post-stroke MoCA scores and ALPS index, which were consistently demonstrated in the lesion, non-lesion, and mean bilateral measurements. A potential mechanistic explanation is that the enrolled stroke patients exhibited prevalent features of cerebral small vessel disease with white matter hyperintensities and cerebral microbleeds, and these chronic vascular pathologies may exert more profound effects on cognitive function at 90 days than the acute infarct lesions per se. Our ROC analysis demonstrated that the DTI-ALPS index showed moderate discriminative ability (AUC = 0.868) in differentiating patients with cognitive impairment from those with normal cognition following subcortical infarction, exhibiting excellent sensitivity (96.0%) but suboptimal specificity (65.5%). Although these findings suggest the potential clinical utility of DTI-ALPS as a supportive biomarker, its current specificity limitations preclude its recommendation as a standalone screening tool for early detection of post-infarction cognitive impairment. CSF Aβ42, p-tau181 and several plasma p-tau biomarkers (such as p-tau 231) can be used in a specialized memory clinic as a stand-alone biomarker to detect biologically-defined AD (35, 36). The combination of ALPS index and CSF p-tau levels may serve as a stronger predictor of post-stroke cognitive function.

The DTI-ALPS method assesses glymphatic system function by measuring the direction of water molecule movement within the PVS. The ALPS index is calculated for a ROI to quantify GS function. Impaired GS function is represented by a decrease in the DTI-ALPS index (12). DTI-ALPS has been widely utilized to evaluate alterations in the GS associated with neurological and other systemic disorders, such as traumatic brain injury, stroke, carotid atherosclerosis, CSVD, sleep disorders, AD, PD, acute lymphoblastic leukemia in children, and type 2 diabetes mellitus. The largest DTI-ALPS study to date, which included 2,219 participants, found that GS dysfunction, as reflected by DTI-ALPS, is closely related to the occurrence and severity of CSVD. Moreover, the ALPS index demonstrated superior predictive performance for the presence of CSVD among all eligible residents (37). Additionally, the ALPS index shows a significant negative correlation with Aβ and tau protein deposition and a positive correlation with cognitive scores. The ALPS index mediates cognitive dysfunction associated with Aβ and tau deposition (38). Our previous research also found a decrease in the ALPS index, indicating GS dysfunction, in middle-aged and elderly patients with chronic insomnia, which is associated with cognitive impairment (39).

Our study found that the reduction of ALPS in the side of cerebral lesion was not associated with cognitive dysfunction 3 months after stroke, and may represent the compensatory capacity of the GS. Inadequate clearance of GS plays an important role in the pathophysiological mechanisms of cognitive decline after early stroke. It is a therapeutic target to improve early cognitive impairment by actively improving oxidative stress, inflammatory response, reducing apoptosis and maintaining lymphatic system function after stroke. The relative contribution and causal interaction of the GS in cognitive deficits induced by cerebral infarction warrant further study.

At the same time, this study has several limitations that warrant cautious interpretation of the results. The relatively small sample size and single-center recruitment strategy may have reduced statistical power for correlation analysis. While our model showed acceptable discrimination (AUC = 0.868, 95% CI 0.77–0.96), the wide confidence intervals reflect limited precision due to sample size constraints. Second, although we controlled for vascular risk factors and education level, other premorbid cognitive reserve proxies (e.g., occupational complexity, leisure activities) were not systematically collected. Future studies should incorporate comprehensive cognitive reserve assessments using standardized tools like the Cognitive Reserve Index Questionnaire (CRIq). Third, due to clinical constraints, we did not acquire follow-up MRI scans at 90 days post-stroke. Consequently, our analysis could not evaluate temporal changes in glymphatic function (ALPS index) at the same anatomical locations over time. Future studies should incorporate longitudinal MRI assessments at 90-day timepoints to enable analysis of both glymphatic changes at identical anatomical locations and their relationship with 90-day cognitive outcomes.

In conclusion, lymphatic system dysfunction is present in cerebral infarction, and it is positively correlated with early cognitive decline after stroke. The DTI-ALPS index’s predictive value at 7 days post-stroke highlights a therapeutic window for interventions targeting glymphatic clearance. In the future, it is important to conduct the correlation between cerebral lymphoid dysfunction in patients with cerebral infarction and clinical function based on multimodal MRI technology.