Review: Natural products show preclinical promise in vascular dementia

A 2026 review concludes that isolated natural compounds, botanical extracts and herbal formulations can influence several neuroprotective and repair processes in experimental models of vascular dementia, but the evidence for durable benefits in patients remains limited and predominantly preclinical. The analysis covers oxidative stress, neuroinflammation, apoptosis, autophagy, blood–brain barrier integrity, angiogenesis, neurogenesis and synaptic plasticity. Cellular and animal experiments often link changes in molecular targets with better tissue structure or cognitive performance in models of chronic cerebral hypoperfusion. Human trials provide a much weaker and less consistent evidence base, leaving target validation, pharmacokinetics, product standardization and clinical translation as central challenges.

Vascular dementia is a heterogeneous group of disorders in which cerebrovascular disease and persistent vascular dysfunction contribute substantially to progressive cognitive decline. Small-vessel disease, large-vessel atherosclerosis, territorial infarction and mixed forms can produce different combinations of cerebral hypoperfusion, white-matter damage, blood–brain barrier disruption and impaired neurovascular coupling. These vascular changes can trigger oxidative stress, neuroinflammation, neuronal death and synaptic dysfunction, which then interact and reinforce one another. Updated diagnostic criteria place greater emphasis on the severity of cognitive impairment, the underlying vascular etiology and the contribution of coexisting disorders such as Alzheimer’s disease than on older fixed subtypes or vascular risk factors alone.

MRI remains important for identifying cerebrovascular lesions, including lacunar infarcts, white-matter hyperintensities, cortical infarcts and cerebral microbleeds. Fluid biomarkers can supply complementary information: neurofilament light chain may reflect axonal injury, while cerebrospinal fluid or plasma amyloid-β and phosphorylated tau can help identify concomitant Alzheimer’s disease pathology. The review describes these measures as supportive rather than stand-alone diagnostic markers. Candidate markers associated with complement activation, immune responses, vascular dysfunction and neuronal or glial injury are also being investigated, but most have not received sufficient external validation and may be affected by patient heterogeneity and mixed vascular–degenerative pathology. Clinical phenotyping, cerebrovascular imaging and complementary biomarker profiles may therefore offer the most informative basis for diagnosis and stratification.

The proposed therapeutic framework divides disease modification into two overlapping dimensions. Neuroprotection aims to limit secondary brain injury and preserve surviving neurons, white matter, blood–brain barrier and neurovascular units. Neural repair extends beyond injury control by supporting vascular remodeling, endogenous neurogenesis, synaptic plasticity and recovery of neural networks. These are not discrete stages: limiting oxidative damage or inflammation can preserve the cellular substrates needed for later repair, while vascular and synaptic remodeling depends partly on the local tissue environment. Mechanistic evidence is strongest when a molecular change is accompanied by protection of tissue structure, improved perfusion or better functional performance, rather than a shift in the abundance of one protein alone.

Experimental findings indicate that natural products can affect redox stress through several upstream mechanisms. In models of chronic cerebral hypoperfusion, GJ-4 was associated with reduced white-matter injury and improved cognitive performance alongside activation of the Keap1/Nrf2/HO-1 antioxidant axis. Fisetin combined Nrf2/HO-1 activation with suppression of NF-κB/NLRP3 signaling in vitro, while puerarin acted on a redox-sensitive TRPM2/NMDAR pathway in bilateral common carotid artery occlusion rats and was associated with lower reactive oxygen species production. Fo-Shou-San and an ethanol extract of Rubia yunnanensis were reported to improve glutathione-related antioxidant capacity or biochemical defense measures. These are model observations, and reductions in reactive oxygen species or malondialdehyde alone do not identify a specific therapeutic target or demonstrate benefit in patients. A small clinical combination study involving Dengzhan Shengmai Capsules and butylphthalide associated treatment with lower peripheral lipid-peroxidation markers and better cognition, but its design and blood-based measurements did not directly establish cerebral antioxidant mechanisms.

Anti-inflammatory effects are similarly varied. Baicalein and puerarin have been linked in experimental systems to suppression of TLR4/MyD88/NF-κB signaling, while resveratrol was associated with reduced STING/TBK1/IRF3 signaling in chronic cerebral hypoperfusion. In a two-vessel-occlusion rat model, 20(S)-protopanaxadiol reduced Iba-1 expression, shifted microglial polarization toward an anti-inflammatory phenotype and attenuated NLRP3 inflammasome-mediated neuroinflammation. Quercetin and QBT have also been associated with reduced glial activation, demyelination or neuroinflammation through pathways including CXCL12/CXCR4. These findings suggest that the relevant goal may be to limit sustained damaging inflammation while preserving reparative glial functions, rather than indiscriminately suppressing all immune activity. The clinical link remains uncertain: a randomized trial of Sailuotong found a cognitive difference at 26 weeks that was no longer statistically significant at week 52, and the trial did not directly measure microglial states or inflammasome activity.

Natural products have also been studied in relation to apoptosis, autophagy and barrier protection. In two-vessel-occlusion rats, resveratrol decreased Bax and caspase-3 levels and increased Bcl-2, while imperatorin produced a similar molecular pattern alongside reduced hippocampal injury and improved synaptic structure. Macelignan was linked to mTOR activation, restoration of mitochondrial function, reduced oxidative stress and attenuation of neuronal apoptosis. Autophagy findings depend strongly on cell type and disease context: β-asarone and gastrodin were reported to restrain excessive or dysregulated neuronal autophagy, whereas the ginseng-derived saponin PF11 enhanced mTOR-mediated autophagy and lysosomal function in oligodendrocytes. Changes in static autophagy markers do not by themselves demonstrate restored autophagic flux. Other experiments associate melatonin and an Artemisia annua preparation with preservation of endothelial, pericyte and tight-junction markers involved in blood–brain barrier integrity. Clinical studies have not directly measured autophagic flux or barrier permeability, leaving these mechanisms largely preclinical.

Repair-oriented evidence includes experimental effects on angiogenesis, neurogenesis and synaptic function. Dihydromyricetin increased cerebral blood flow, promoted angiogenesis and improved cognitive performance in a bilateral common carotid artery stenosis mouse model, with target-level and inhibition experiments implicating VEGFR2. NBP produced angiogenic and functional improvements associated with Shh/Ptch1 signaling in two-vessel-occlusion rats, and partial loss of these effects after Shh silencing supported a functional contribution from that pathway. Floralozone, Xinshubao tablets and QNYZD have been linked to hippocampal neurogenesis in models of cerebral hypoperfusion. Dendrobium polysaccharides, naringenin and tetrahydrocurcumin have improved synaptic protein expression, ultrastructural measures or cognitive scores through pathways including NMDAR signaling and Wnt/β-catenin/BDNF/TrkB/CREB. These findings are encouraging within their models, but improved clinical cognition cannot by itself demonstrate restored angiogenesis, neurogenesis or synaptic architecture in patients.

Clinical evidence has not established consistent, patient-important benefits from natural-product interventions. A systematic review cited by the article identified 16 randomized controlled trials involving 5,668 participants, although only a small proportion evaluated natural or herbal interventions. EGb 761 was evaluated in three trials, with a possible signal in the SKT cognitive test but substantial heterogeneity and very low certainty of evidence; Tianzhi granule was examined in only one trial, leaving its efficacy uncertain. A separate 24-week study reported significant improvement in VADAS-cog and CIBIC-plus scores with Tianzhi granule compared with placebo, while long-term efficacy and disease-modifying effects remained unestablished. Cannabidiol did not significantly improve cognitive or functional outcomes in a small, four-week trial, despite changes in neuropsychiatric scales. Smaller combination studies were limited by adjunctive designs, short follow-up, participant dropout and reliance on peripheral biomarkers rather than direct measurements of brain mechanisms.

Translation is further complicated by pharmacokinetics, safety, manufacturing consistency and model limitations. Natural products may have limited oral bioavailability, restricted blood–brain barrier penetration, uncertain metabolic stability and incompletely characterized interactions with conventional medicines. Antiplatelet or anticoagulant activity could increase bleeding risk when combined with antithrombotic therapy, and possible neurotoxicity requires assessment for central nervous system use. Two-vessel occlusion and bilateral common carotid artery stenosis reproduce important aspects of chronic cerebral hypoperfusion, but their severity, temporal profile, lesion distribution and behavioral phenotypes do not match the heterogeneous clinical spectrum of vascular cognitive impairment. Isolated compounds permit clearer target and dose–response assessment, whereas botanical extracts and herbal formulations require chemical fingerprints, quality markers, identification of active constituents and evidence of batch consistency. The review therefore identifies natural products as credible experimental candidates for vascular dementia, not established disease-modifying treatments. The decisive test is whether carefully controlled human studies can connect exposure and target engagement with tissue responses and clinically meaningful outcomes.

Subject of Research: Natural-product neuroprotection and neural repair strategies in vascular dementia

Article Title: Treatment of vascular dementia based on neuroprotection and neural repair strategies: from pathophysiology to natural products

Article References: Luo, Q., Li, Z., Liao, Y., Chen, W., Xu, Z., You, X., Xue, G., Duan, Y., Li, J., & Li, S. (2026). Treatment of vascular dementia based on neuroprotection and neural repair strategies: from pathophysiology to natural products. Frontiers in Pharmacology, 17, Article 1942636. https://doi.org/10.3389/fphar.2026.1942636

Image Credits: AI Generated

DOI: 10.3389/fphar.2026.1942636

Keywords: vascular dementia, natural products, neuroprotection, neural repair, neuroinflammation, cerebral hypoperfusion, blood–brain barrier, neurogenesis, synaptic plasticity, clinical translation, Treatment, vascular