SGLT-2 Inhibitors in the Prevention and Progression of Neurodegenerative Diseases: A Narrative Review

Abstract

Neurodegenerative diseases are among the most prevalent and debilitating disorders in aging populations. Despite growing insights into their complex pathophysiology, effective disease-modifying treatments remain limited. Sodium-glucose cotransporter 2 (SGLT-2) inhibitors, primarily used in type 2 diabetes mellitus, have recently gained attention for their potential neuroprotective effects. This narrative review aims to summarize the current preclinical and clinical evidence on the impact of SGLT-2 inhibitors on neurodegenerative diseases, exploring their mechanisms of action, therapeutic potential, and limitations. The authors reviewed experimental studies, animal models, clinical trials, and observational data focusing on the potential links between SGLT-2 inhibitors and neurodegeneration. We further analyzed proposed mechanisms—including metabolic, inflammatory, and vascular factors—in the context of their potential contribution to, or consequence of, neurodegenerative processes, emphasizing their interdependence rather than treating neurodegeneration as an isolated phenomenon. Preclinical studies consistently show that SGLT-2 inhibitors reduce neuroinflammation, improve mitochondrial function, enhance insulin sensitivity in the brain, and may mitigate amyloid and tau pathology. Observational clinical data suggest a lower incidence of dementia in patients treated with SGLT-2 inhibitors. However, cognitive outcomes have not been directly assessed in major randomized trials to date. SGLT-2 inhibitors hold promise as modulators of neurodegenerative processes, but robust clinical trials with cognitive endpoints are needed to confirm their therapeutic relevance. Their potential to bridge metabolic and neurodegenerative pathways highlights a novel avenue for future research and therapeutic development.

Key Summary Points

Sodium-glucose cotransporter-2 (SGLT-2) inhibitors, originally developed for type 2 diabetes, have been associated with potential neuroprotective effects.

Observational studies suggest an association with lower risk of dementia in patients with diabetes (hazard ratios approximately 0.6–0.8), but evidence remains limited and indirect.

Mechanistic insights are largely derived from preclinical models, highlighting effects on neuroinflammation, mitochondrial function, amyloid/tau metabolism, and cerebral perfusion.

Most available data derive from patients with diabetes and comorbidities, with short follow-up durations, limiting generalizability and causal inference.

Future research should incorporate long-term prospective trials with validated neuropsychological testing, biomarker panels (e.g., p-tau217, Aβ42/40, GFAP, NfL), and advanced neuroimaging modalities.

Introduction

Literature Search Methods

We performed a structured search of PubMed, Embase, and Web of Science covering January 2000 to July 2025 using the terms “SGLT-2 inhibitor,” “empagliflozin,” “dapagliflozin,” “canagliflozin,” “neurodegenerative disease,” “dementia,” “Alzheimer,” and “Parkinson.” Human studies and relevant preclinical studies were eligible for inclusion; case reports and non-English papers were excluded. References of relevant articles were also hand-searched.

Neurodegenerative diseases represent a growing global health burden, with significant medical, social, and economic implications. Among them, Alzheimer’s disease (AD) and Parkinson’s disease (PD) are the most prevalent and best-studied, particularly in the context of neuroinflammation and metabolic dysfunction. While research has traditionally focused on tauopathies, such as AD and progressive supranuclear palsy (PSP), and on synucleinopathies including PD, dementia with Lewy bodies (DLB), and multiple system atrophy (MSA), the pathological processes underlying these conditions remain only partially understood, and the role of shared mechanisms continues to be investigated.

Synucleinopathies, such as PD, DLB, and MSA, are characterized by pathological aggregation of alpha-synuclein, which leads to neuronal injury through mechanisms involving microglial activation, mitochondrial dysfunction, and oxidative stress. These mechanisms partially overlap with those observed in tauopathies, suggesting shared pathogenic pathways. Moreover, synucleinopathies also exhibit features of insulin resistance, metabolic dysfunction, and chronic inflammation—factors that may be amenable to modulation by sodium-glucose cotransporter-2 (SGLT-2) inhibitors. Therefore, their potential therapeutic role in this group of disorders merits further investigation.

Alpha-synuclein accumulation triggers microglial activation, mitochondrial impairment, and oxidative stress—pathophysiological pathways that overlap with those observed in tauopathies. Given the shared mechanisms, it is plausible that agents like SGLT-2 inhibitors which modulate neuroinflammatory and metabolic processes, may also have therapeutic relevance in the context of synucleinopathies. However, this potential remains underexplored and warrants further investigation. These diseases, along with chronic conditions like diabetes, share a common characteristic of being chronic inflammatory disorders, which contributes to their long-term progression and complicates treatment strategies [1–3]. These disorders are characterized by the progressive loss of structure and function of neurons, ultimately leading to cognitive and motor decline. Despite extensive research efforts, the precise etiology of neurodegenerative diseases remains only partially understood, and effective disease-modifying treatments are still lacking.

Emerging evidence suggests that metabolic dysfunction, particularly insulin resistance and impaired glucose metabolism in the brain, may play a key role in the pathogenesis of neurodegeneration. This dysfunction is closely linked to chronic inflammation, a common feature of both type 2 diabetes and neurodegenerative diseases. The relationship between inflammation and neurodegeneration remains complex and somewhat ambiguous, raising the question of whether inflammation is a cause or a consequence of neurodegeneration. Chronic low-grade inflammation is thought to contribute to both the onset and progression of neurodegenerative diseases [4, 5]. Conversely, accumulating evidence suggests that neurodegeneration may itself trigger inflammatory responses, notably via dysregulated and persistent microglial activation, thereby perpetuating a self-amplifying neurodegenerative process [6, 7].

The persistence of this inflammatory state in both diabetes and neurodegenerative diseases may act as a critical factor in accelerating disease progression, with implications for therapeutic strategies that target inflammation as a potential therapeutic approach [8]. This has led to the conceptualization of Alzheimer’s disease as “type 3 diabetes,” highlighting the importance of metabolic pathways in neurodegenerative processes [9]. Additionally, oxidative stress, neuroinflammation, and mitochondrial dysfunction are not only commonly observed in neurodegenerative diseases but also actively contribute to their progression through specific cellular mechanisms [10]. Neuroinflammation, in particular, is mediated by the activation of glial cells—especially microglia and astrocytes—which, upon chronic stimulation, release pro-inflammatory cytokines, reactive oxygen species (ROS), and other neurotoxic factors. This glial response, although initially protective, may become maladaptive, amplifying neuronal injury and accelerating neurodegeneration [11, 12]. Oxidative stress further disrupts cellular homeostasis, while mitochondrial dysfunction impairs neuronal energy metabolism, making neurons more susceptible to damage [13, 14]. These interrelated processes create a self-perpetuating cycle of cellular stress and degeneration, forming a critical pathophysiological basis for conditions such as Alzheimer’s disease, Parkinson’s disease, and other tauopathies. Understanding these mechanisms is essential for evaluating the therapeutic potential of agents like SGLT-2 inhibitors, which may modulate glial activity and metabolic stress.

SGLT-2 inhibitors are a relatively new class of antidiabetic drugs that lower blood glucose by promoting glucosuria through the inhibition of renal glucose reabsorption [15]. Beyond their glucose-lowering effects, SGLT-2 inhibitors have demonstrated a range of pleiotropic benefits, including cardiovascular and renal protection, anti-inflammatory properties, and potential metabolic reprogramming [16, 17]. These systemic effects may indirectly influence neurodegenerative processes by improving vascular function and kidney health—both of which are increasingly recognized as contributors to cognitive decline and neurodegeneration [18]. This narrative review aims to summarize and critically evaluate the current evidence on the potential impact of SGLT-2 inhibitors on neurodegenerative disorders. The review will explore mechanistic pathways linking SGLT-2 inhibition with neuroprotection, discuss findings from preclinical and clinical studies, and highlight areas where further research is needed to determine the translational relevance of these findings in the context of neurodegeneration.

Furthermore, accumulating clinical and experimental evidence supports a multifactorial interplay between neurodegeneration, metabolic dysfunction, and cardiovascular disease. Patients with neurodegenerative disorders often present with coexisting type 2 diabetes mellitus (T2DM), insulin resistance, and heart failure—conditions that share overlapping pathophysiological features such as chronic low-grade inflammation, oxidative stress, and mitochondrial impairment [19–21]. This triad of systemic dysfunction not only accelerates neuronal injury but also complicates therapeutic strategies due to its multifaceted nature.

Of particular interest is the bidirectional relationship between neurodegeneration and heart failure, which has been associated with cerebral hypoperfusion, impaired neurovascular coupling, and increased risk of vascular cognitive impairment [20, 22]. Emerging studies suggest that neurodegenerative syndromes such as corticobasal syndrome (CBS) may, in some cases, develop on a vascular substrate, highlighting the potential importance of cerebrovascular pathology in atypical parkinsonian disorders [23]. Moreover, hypertension—a well-established cardiovascular risk factor—has been independently associated with an increased risk of PSP, suggesting that vascular factors may play a role not only in classical vascular dementia but also in tauopathies [24].

Given this convergence of pathophysiological mechanisms, the role of SGLT-2 inhibitors may extend beyond glycemic control. These agents exhibit pleiotropic effects that include anti-inflammatory, antioxidative, and cardioprotective properties [19, 21, 25]. SGLT-2 inhibitors improve endothelial function, reduce oxidative stress, and modulate systemic inflammation, all of which may confer indirect neuroprotective benefits [19, 25, 26]. Their ability to reduce hospitalization rates for heart failure, preserve renal function, and improve metabolic efficiency suggests a broader therapeutic potential in patients with combined neurodegenerative, metabolic, and cardiovascular comorbidities [20, 22].

These observations support a growing rationale for investigating SGLT-2 inhibitors within a systems medicine framework, in which metabolic and vascular health are integral to neuroprotection. The intersection of metabolic reprogramming, glial modulation, and improved vascular function positions SGLT-2 inhibition as a compelling therapeutic avenue not only in Alzheimer’s disease and Parkinson’s disease, but potentially also in atypical parkinsonian syndromes with vascular or mixed pathologies [19, 21].

For clarity, in this review we use the term “neuroprotection” to mean modulation of pathological processes (e.g., inflammation, oxidative stress, insulin resistance, amyloid/tau dynamics) rather than a proven clinical slowing of disease; associations reported here should not be interpreted as evidence of disease modification in humans.

This article is based on previously conducted studies and does not contain any new studies with human participants or animals performed by any of the authors.

Characterization of SGLT-2 Inhibitors

Differences among agents may be relevant: canagliflozin partially inhibits SGLT-1; differences in protein binding and possible protein–glycoprotein (P-gp) interactions could affect central nervous system (CNS) exposure.

SGLT-2 inhibitors are a class of oral antidiabetic agents that lower blood glucose levels by inhibiting SGLT-2 proteins located in the proximal renal tubules, thereby reducing glucose reabsorption and promoting urinary glucose excretion [15]. The main agents in this class include canagliflozin, dapagliflozin, empagliflozin, and ertugliflozin, all of which have been approved for the treatment of T2DM.

Beyond their glucose-lowering effects, SGLT-2 inhibitors exert several pleiotropic benefits, including reductions in body weight, blood pressure, and albuminuria, as well as improvements in cardiovascular and renal outcomes [16, 17]. These effects are thought to result from both hemodynamic and metabolic mechanisms, such as natriuresis, ketogenesis, and modulation of inflammatory and oxidative pathways.

Recent studies have also suggested that SGLT-2 inhibitors may cross the blood–brain barrier and influence central nervous system function, potentially offering neuroprotective benefits [18]. This has prompted growing interest in evaluating their role beyond diabetes management, particularly in the context of neurodegenerative diseases.

In tauopathies—disorders marked by abnormal accumulation of hyperphosphorylated tau protein, including Alzheimer’s disease and progressive supranuclear palsy—SGLT-2 inhibitors may mitigate neuroinflammation, oxidative stress, and metabolic dysfunction, all of which contribute to tau pathology and neurodegeneration. Similarly, in synucleinopathies like Parkinson’s disease and multiple system atrophy, characterized by alpha-synuclein aggregation, SGLT-2 inhibitors might influence mitochondrial function, glial activation, and inflammatory cascades that exacerbate neuronal damage.

Therefore, the emerging role of SGLT-2 inhibitors extends beyond glycemic control into a promising therapeutic strategy targeting shared pathophysiological mechanisms across both tau- and synuclein-based neurodegenerative diseases. Further research is warranted to elucidate these mechanisms and to define patient populations who may derive the greatest benefit from such interventions [27–31].

Pathophysiological Mechanisms of Neurodegenerative Diseases and Related SGLT-2 Pathways

Most mechanistic inferences summarized below are based on preclinical models; confirmation in humans is currently limited and should be interpreted with caution.

Neurodegenerative diseases are multifactorial disorders with complex and overlapping pathophysiological mechanisms. Emerging evidence suggests that fluctuations in blood glucose levels, particularly insulin resistance and hyperglycemia, may contribute to the pathogenesis of neurodegenerative diseases not only in AD and PD but also in other conditions such as PSP. In PSP, diabetes has been identified as a significant risk factor, further highlighting the role of metabolic dysfunction in neurodegeneration [5, 26]. Chronic hyperglycemia, as commonly observed in diabetes, may lead to vascular changes, including endothelial dysfunction and increased blood–brain barrier permeability, which in turn could exacerbate neurodegeneration [32]. These vascular disturbances are thought to contribute to cognitive decline and the progression of neurodegenerative disorders [8]. Therefore, dysregulated glucose metabolism and the associated vascular complications may act as crucial determinants in the development and progression of these diseases, suggesting that metabolic control may have therapeutic potential in mitigating neurodegenerative pathologies [33]. Key contributors include oxidative stress, mitochondrial dysfunction, chronic neuroinflammation, and impaired glucose metabolism within the central nervous system (CNS) [10, 34]. These factors contribute to synaptic failure, protein misfolding (e.g., amyloid-beta, tau, and alpha-synuclein aggregation), and ultimately neuronal loss.

One increasingly recognized mechanism is brain insulin resistance. The brain relies heavily on glucose for energy, and insulin signaling plays a crucial role in synaptic plasticity, neuronal survival, and memory formation. In AD, decreased insulin receptor sensitivity and signaling lead to a cascade of neurodegenerative processes, including reduced glucose uptake, impaired mitochondrial function, and increased oxidative stress [35, 36]. These changes have led to the conceptualization of AD as a “type 3 diabetes” [9].

SGLT-2 inhibitors may influence several of these pathophysiological pathways. Although their primary site of action is the kidney, emerging evidence suggests that SGLT-2 inhibitors may exert systemic metabolic effects that extend to the brain. First, by promoting mild ketosis through enhanced fat oxidation and decreased insulin levels, SGLT-2 inhibitors increase circulating ketone bodies such as β-hydroxybutyrate, which can serve as an efficient alternative fuel for neurons, potentially improving mitochondrial function and cognitive performance [37].

Second, SGLT-2 inhibitors have demonstrated anti-inflammatory and antioxidant effects. These drugs have been shown to reduce levels of pro-inflammatory cytokines (e.g., IL-6, TNF-α) and inhibit the activation of microglia and astrocytes, which are key mediators of neuroinflammation [18, 38]. This may slow or prevent neuronal damage in neurodegenerative settings.

Third, SGLT-2 inhibitors may indirectly improve insulin signaling in the brain by enhancing peripheral insulin sensitivity and lowering systemic insulin levels, thereby potentially alleviating central insulin resistance [32]. In animal models of AD, SGLT-2 inhibitors have been associated with reduced amyloid plaque formation, improved synaptic function, and decreased neurodegeneration [39, 40].

Finally, some studies suggest that SGLT-2 inhibitors may play a role in preserving the integrity of the blood–brain barrier (BBB), a highly selective endothelial interface that protects the central nervous system from harmful substances while regulating the exchange of nutrients and metabolic waste [41]. In neurodegenerative diseases, the BBB is often disrupted, leading to increased permeability that permits the infiltration of pro-inflammatory cytokines, peripheral immune cells, and neurotoxic proteins into the brain parenchyma. This dysfunction exacerbates neuroinflammation and contributes to the accumulation of pathological aggregates such as amyloid-β and hyperphosphorylated tau, accelerating neuronal damage.

Evidence for clinically meaningful BBB penetration of SGLT-2 inhibitors in humans remains uncertain; existing data are predominantly preclinical or indirect.

The potential of SGLT-2 inhibitors to protect BBB integrity may be mediated by several mechanisms. These include improvements in endothelial function through reductions in oxidative stress and systemic inflammation, as well as better regulation of glucose and lipid metabolism, which indirectly benefits cerebrovascular health. In preclinical studies, SGLT-2 inhibitors have been shown to reduce microglial activation and protect against BBB breakdown in models of diabetes and Alzheimer’s disease. Moreover, by influencing mitochondrial efficiency and reducing reactive oxygen species in brain endothelial cells, these drugs may help maintain tight junction protein expression and barrier stability.

Understanding the precise cellular and molecular pathways through which SGLT-2 inhibitors affect BBB function remains a subject of ongoing investigation. Nonetheless, current evidence suggests that preserving BBB integrity may be an important, though indirect, mechanism by which SGLT-2 inhibitors exert neuroprotective effects in both metabolic and neurodegenerative contexts [42, 43].

These multimodal effects highlight the therapeutic potential of SGLT-2 inhibitors as modulators of neurodegenerative pathways, although further mechanistic and clinical studies are needed to confirm these benefits in humans.

Preclinical Evidence of the Neuroprotective Effects of SGLT-2 Inhibitors

A growing body of preclinical studies supports the potential neuroprotective role of SGLT-2 inhibitors in models of neurodegenerative diseases. These studies, conducted both in vitro and in vivo, have demonstrated beneficial effects on cognitive function, neuropathology, and neuroinflammation.

In rodent models of AD, treatment with dapagliflozin or empagliflozin has been associated with significant improvements in learning and memory performance, often measured using the Morris water maze or novel object recognition tests [40, 44]. These cognitive improvements are frequently accompanied by reduced deposition of amyloid-beta plaques and phosphorylated tau protein, key pathological hallmarks of AD [45].

Additionally, SGLT-2 inhibitors have been shown to attenuate neuroinflammation and oxidative stress in diabetic and non-diabetic animal models. In particular, empagliflozin administration in high-fat diet-induced obese mice reduced the activation of microglia and astrocytes in the hippocampus, along with decreased levels of pro-inflammatory cytokines such as IL-1β and TNF-α [38, 46]. These changes were accompanied by improved mitochondrial function and reduced neuronal apoptosis, suggesting a direct neuroprotective effect. These findings are particularly relevant in the context of neurodegenerative diseases, where distinct inflammatory profiles underlie disease-specific pathophysiology. For instance, Alzheimer’s disease is characterized by chronic microglial activation and elevated levels of IL-1β and TNF-α in response to amyloid-β plaques, while diseases like progressive supranuclear palsy (PSP) show more localized astroglial tau pathology and differential cytokine signatures. Given this heterogeneity, it is plausible to hypothesize that SGLT-2 inhibitors may exert varying degrees of efficacy depending on the dominant inflammatory mechanisms in each condition. For example, in AD, suppression of pro-inflammatory microglial activation might slow amyloid-induced neurotoxicity, whereas in PSP, modulation of astrocytic function and tau propagation might be more relevant. Therefore, further disease-specific studies are warranted to determine how the immunomodulatory effects of SGLT-2 inhibitors translate into clinical benefit across different neurodegenerative disorders [47–50].

In a streptozotocin-induced model of diabetes, which mimics insulin resistance and cognitive decline, dapagliflozin treatment resulted in increased expression of brain-derived neurotrophic factor (BDNF) and synaptic markers such as synaptophysin, indicating enhanced synaptic plasticity [40]. Furthermore, markers of oxidative stress such as malondialdehyde (MDA) were reduced, while antioxidant enzymes (e.g., superoxide dismutase [SOD], catalase) were upregulated, supporting the antioxidative properties of SGLT-2 inhibitors [51].

Although studies in PD models are limited, initial evidence suggests that SGLT-2 inhibitors may protect dopaminergic neurons and mitigate motor dysfunction through similar mechanisms involving mitochondrial preservation and reduced inflammation [46].

Collectively, these preclinical findings provide a compelling rationale for the investigation of SGLT-2 inhibitors as potential therapeutic agents in neurodegenerative diseases. However, translation of these results into clinical benefit requires confirmation in well-designed human studies (Table 1).

Table 1.

Preclinical models of SGLT-2 inhibitors in neurodegeneration

Compound	Model	Dose	Duration	Observed effects
Dapagliflozin	Transgenic AD mice	1 mg/kg	8 weeks	↓ Aβ plaques; ↑ memory performance
Empagliflozin	MPTP-induced PD mouse model	10 mg/kg	4 weeks	↑ dopaminergic neuron survival; ↓ oxidative stress
Canagliflozin	High-fat diet mouse model	5 mg/kg	12 weeks	↓ hippocampal inflammation; ↑ synaptic markers

Clinical Evidence Supporting Neuroprotective Effects of SGLT-2 Inhibitors

It is difficult to disentangle putative direct CNS effects from indirect benefits mediated via cardiovascular and renal improvements, which themselves influence cognitive trajectories.

Although SGLT-2 inhibitors were initially developed for the management of T2DM, growing clinical interest has emerged around their potential effects on cognitive health and the risk of neurodegenerative diseases. Several epidemiological studies and post hoc analyses of cardiovascular outcome trials suggest that SGLT-2 inhibitors may offer cognitive protection, although direct evidence from randomized clinical trials remains limited [52, 53].

Recent preclinical studies have begun to elucidate the underlying mechanisms by which SGLT-2 inhibitors may influence cognitive function, particularly in relation to hippocampal health. The hippocampus, a critical brain region for learning and memory, is highly susceptible to metabolic and inflammatory insults. SGLT-2 inhibitors such as empagliflozin and canagliflozin have been shown to reduce neuroinflammation and oxidative stress in the hippocampus, which are key contributors to synaptic dysfunction and cognitive decline. For instance, empagliflozin administration in high-fat diet-induced obese mice resulted in decreased activation of microglia and astrocytes in the hippocampus, along with reduced levels of pro-inflammatory cytokines such as IL-1β and TNF-α [47].

Additionally, SGLT-2 inhibitors may modulate brain energy metabolism, shifting the brain's fuel preference from glucose to ketone bodies, which can enhance mitochondrial function and reduce oxidative stress. This metabolic shift has been associated with improved synaptic plasticity and cognitive performance in animal models [54].

These findings suggest that SGLT-2 inhibitors may exert neuroprotective effects through multiple pathways, including modulation of neuroinflammation, oxidative stress, and energy metabolism. However, further research, particularly randomized clinical trials, is necessary to confirm these effects and determine the clinical relevance of SGLT-2 inhibitors in the prevention and treatment of cognitive decline and neurodegenerative diseases.

Observational Studies and Real-World Data

Potential biases include immortal time bias, lack of active comparator groups, and time-varying confounding (e.g., HbA1c, eGFR, blood pressure, weight). These limit causal inference.

Variables such as ketone bodies, albuminuria, blood pressure, weight, and uric acid are likely to lie on the causal pathway linking SGLT-2 inhibitor use to neurodegenerative outcomes. Adjusting for such variables as confounders may introduce bias and underestimate the true effect. In contrast, classical confounders (e.g., age, sex, baseline comorbidities, socioeconomic status) should be appropriately controlled. Future observational studies should incorporate formal mediation analyses to better delineate causal pathways and quantify the indirect vs direct effects of SGLT-2 inhibitors on neurocognitive outcomes.

Another important limitation is that current observational and post hoc analyses rarely stratify patients by SGLT-2 inhibitor monotherapy versus use in combination with other antidiabetic agents (e.g., glucagon-like peptide-1 receptor agonists [GLP-1 RAs], insulin, dipeptidyl peptidase-4 [DPP-4] inhibitors). Such polypharmacy is common in clinical practice and may confound interpretation of cognitive outcomes. Future studies should explicitly distinguish between monotherapy and combination therapy to clarify whether observed associations are attributable to SGLT-2 inhibitors themselves or to additive/synergistic effects with other agents.

We note that outcome definitions in these datasets frequently rely on administrative diagnostic codes rather than standardized neuropsychological testing, which limits inference about cognition.

A number of large-scale retrospective cohort studies have investigated the relationship between SGLT-2 inhibitor use and the incidence of dementia. In a nationwide South Korean study of over 200,000 individuals with T2DM, SGLT-2 inhibitors were associated with significantly reduced risk of all-cause dementia compared to dipeptidyl peptidase-4 inhibitors (DPP-4is), with a hazard ratio (HR) of 0.79 (95% CI 0.74–0.84) [55, 56]. Importantly, this protective effect was consistent across both Alzheimer’s disease and vascular dementia subtypes.

Similarly, a US-based study utilizing real-world data from the Veterans Health Administration analyzed outcomes in over 70,000 patients and found that those treated with SGLT-2 inhibitors had a 30–40% lower risk of incident dementia compared to those receiving sulfonylureas or insulin [58]. These findings support the hypothesis that SGLT-2 inhibitors may offer benefits beyond glycemic control, possibly through mechanisms such as reduced insulin resistance, anti-inflammatory effects, or improved cerebrovascular function.

Another study from Taiwan compared SGLT-2 inhibitors with GLP-1 receptor agonists and found that both drug classes were associated with a reduced risk of dementia; however, SGLT-2 inhibitors showed a slightly greater protective effect, particularly in patients with coexisting cardiovascular disease [57]. While these findings are encouraging, the study’s observational design limits the ability to draw definitive causal conclusions. Additionally, potential confounders—such as differences in baseline characteristics, medication adherence, or access to healthcare—may have influenced the results. Nevertheless, the consistency of the protective trend across subgroups strengthens the hypothesis that SGLT-2 inhibitors may offer neuroprotective benefits, especially in high-risk populations (Table 2).

Table 2.

Comparative summary of human observational and post hoc studies of SGLT-2 inhibitors and cognitive outcomes (study design, sample size, comparator, follow-up, HR [95% CI], adjustments, limitations)

Study	Design	Population	Cognitive outcome definition		Comparator	Follow-up	HR (95% CI)	Adjustments	Key limitations
Wium-Andersen 2022	Nationwide cohort	> 200,000 patients with diabetes	DPP-4 inhibitors		All-cause dementia	5 years	0.79 (0.74–0.84)	Age, sex, comorbidities	Administrative codes, no cognitive testing
Veterans Health Admin 2023	Retrospective	~ 70,000 patients with diabetes	Incident dementia		Sulfonylurea/insulin	4 years	0.60–0.70	Demographics, cardiovascular risk factors	Selection bias, limited generalizability
Lin 2021	Taiwan cohort	30,000 patients with diabetes + cardiovascular disease	Dementia		GLP-1 RAs	3 years	0.82 (0.70–0.96)	Cardiovascular risk factors	Short follow-up

Secondary Analyses of Clinical Trials

Although none of the major cardiovascular outcome trials (CVOTs) were designed to assess cognitive outcomes as primary endpoints, secondary analyses have provided indirect insights. In the EMPA-REG OUTCOME trial, empagliflozin significantly reduced the risk of cardiovascular death and hospitalization for heart failure, outcomes that are known to correlate with cognitive decline in older adults [16]. Similarly, the CANVAS and DECLARE–TIMI 58 trials demonstrated reductions in major adverse cardiovascular events and improvements in renal function—factors that are increasingly recognized as contributors to vascular cognitive impairment and mixed dementia [17, 58].

A recent meta-analysis of observational and clinical trial data concluded that SGLT-2 inhibitor use is associated with a reduced risk of stroke and cognitive impairment compared to other glucose-lowering therapies, although the certainty of evidence was rated as moderate due to heterogeneity and indirect outcome measures [60]. These neurovascular benefits may be partly explained by the pleiotropic effects of SGLT-2 inhibitors, including improved blood pressure control, reduction of arterial stiffness, and attenuation of endothelial dysfunction. Additionally, their anti-inflammatory and antioxidant properties—as well as the modulation of cerebral energy metabolism—may further contribute to neuroprotection and the reduction of stroke risk [61–64].

Ongoing and Planned Randomized Controlled Trials

Recognizing the need for more robust evidence, several prospective randomized controlled trials (RCTs) have been initiated to directly examine the cognitive effects of SGLT-2 inhibitors. One of the most notable is the DAPA-MIND trial (NCT03801642), which aims to evaluate the effect of dapagliflozin on cerebral perfusion and cognitive function in patients with heart failure with preserved or reduced ejection fraction. The study employs advanced imaging techniques (e.g., arterial spin labeling magnetic resonance imaging [MRI]) alongside neurocognitive testing to assess the direct impact of SGLT-2 inhibition on brain function.

Additional trials are being designed to explore the use of SGLT-2 inhibitors in populations with early cognitive decline, mild cognitive impairment (MCI), and even established Alzheimer’s disease, particularly in patients with metabolic syndrome or insulin resistance. These studies are expected to provide critical data on whether the neuroprotective effects observed in preclinical models and retrospective cohorts translate into measurable cognitive benefit in humans.

Limitations of Current Evidence

Generalizability is further limited by underrepresentation of non-white populations, low- and middle-income countries, and non-diabetic cohorts. Future work should use culturally validated cognitive assessments.

Generalizability remains limited because most human data derive from patients with diabetes and substantial comorbidity burden; non-diabetic populations and individuals with isolated neurodegenerative diseases are underrepresented. Follow-up in available studies is often shorter than 5 years, which is insufficient to capture slow neurodegenerative trajectories. Publication bias cannot be excluded.

Despite the promising observational data, several limitations must be acknowledged. Most available studies are retrospective and subject to confounding by indication, patient comorbidities, and concomitant medication use. Cognitive outcomes are rarely standardized across studies, and many rely on administrative diagnostic codes rather than validated neuropsychological assessments. Additionally, the populations studied are often limited to individuals with T2DM, leaving the applicability of these findings to non-diabetic individuals uncertain. Moreover, the potential neuroprotective effects of SGLT-2 inhibitors may vary depending on the underlying pathophysiology of different neurodegenerative diseases. For instance, inflammatory mechanisms differ significantly across conditions such as Alzheimer’s disease, Parkinson's disease, and atypical parkinsonian syndromes like PSP, which is characterized by distinct neuroinflammatory and tau-related processes. These differences may influence the extent to which SGLT-2 inhibitors confer benefit, underscoring the need for disease-specific investigations.

Furthermore, while improvements in cardiovascular and renal function may indirectly support cognitive health, causality cannot be inferred from current data. Long-term, placebo-controlled trials with cognitive endpoints are needed to definitively establish the role of SGLT-2 inhibitors in neuroprotection.

Proposed Neuroprotective Mechanisms of SGLT-2 Inhibitors

The potential neuroprotective effects of SGLT-2 inhibitors may be explained by a combination of metabolic, vascular, anti-inflammatory, and mitochondrial mechanisms that extend beyond their glucose-lowering properties. Several biological pathways have been proposed to underlie these effects, many of which intersect with the known pathophysiology of neurodegenerative diseases (Table 3).

Table 3.

Potential neuroprotective mechanisms of SGLT-2 inhibitors

Mechanism	Description	Clinical relevance	References
1. Improved brain energy metabolism	Induction of mild, sustained ketosis and increased circulating β-hydroxybutyrate (BHB) levels via enhanced fatty acid oxidation and reduced insulin levels	Enhances mitochondrial adenosine triphosphate (ATP) production, reduces oxidative stress, and supports neuronal function under insulin-resistant conditions	[9, 36, 37, 65, 71]
2. Reduction of neuroinflammation	Inhibits microglial and astrocyte activation, decreases pro-inflammatory cytokines (e.g., IL-6, TNF-α, IL-1β), and downregulates NF-κB signaling pathways	Prevents or slows neuronal damage, protects synaptic integrity, and mitigates progression of neurodegenerative pathology	[38, 46, 66, 75]
3. Amelioration of insulin resistance and oxidative stress	Improves systemic and possibly central insulin sensitivity; reduces mitochondrial reactive oxygen species (ROS); upregulates antioxidant defenses (e.g., SOD, catalase)	Supports synaptic plasticity, inhibits tau hyperphosphorylation, and reduces oxidative injury to neurons	[34, 35, 39, 51, 69]
4. Enhancement of cerebral perfusion and vascular health	Lowers blood pressure, reduces arterial stiffness, and improves endothelial function	Promotes cerebral blood flow and may reduce risk of vascular cognitive impairment, especially in mixed dementia	[58, 67, 70]
5. Potential modulation of amyloid and tau pathways	May reduce amyloid-β accumulation and tau phosphorylation via enhanced autophagy, improved insulin signaling, and reduced oxidative stress in animal models	Targets core pathological processes in Alzheimer’s disease; potential for disease modification	[40, 45, 68]
6. Preservation of blood–brain barrier (BBB) integrity	Decreases systemic inflammation and oxidative stress; improves endothelial stability	Helps prevent neuroinflammation and restricts entry of neurotoxins into the CNS, especially relevant in metabolic and vascular comorbidities	[41]

Limitations and Future Directions

Biomarker and Imaging Plan

Future prospective studies should incorporate a multimodal biomarker and imaging panel to improve mechanistic understanding. Recommended plasma biomarkers include phosphorylated tau (p-tau217), amyloid-β 42/40 ratio (Aβ42/40), glial fibrillary acidic protein (GFAP), and neurofilament light chain (NfL). Neuroimaging should include arterial spin labeling magnetic resonance imaging (ASL-MRI) for cerebral perfusion and volumetric MRI for structural assessment. Measurements should be collected at baseline, 6, 12, and 24 months, with strict assay quality control procedures and predefined minimal-detectable-change thresholds. Such biomarker-informed designs will enable detection of early neurobiological signals and clarify potential disease-modifying effects.

When positioning SGLT-2 inhibitors in the broader context of antidiabetic therapies, it is important to note that glucagon-like peptide-1 receptor agonists (GLP-1 RAs) and metformin also demonstrate neuro-metabolic effects. GLP-1 RAs have been associated with reduced amyloid deposition, improved insulin signaling, and possible cognitive benefits, while metformin may exert effects on mitochondrial function and inflammation. However, the mechanistic profile of SGLT-2 inhibitors is distinct, with greater emphasis on modulation of energy metabolism, ketone body availability, and vascular function. Future head-to-head and combination studies will be essential to clarify the relative and complementary contributions of these agents to neurocognitive outcomes.

Despite encouraging preclinical data and emerging clinical evidence, the role of SGLT-2 inhibitors as neuroprotective agents in neurodegenerative diseases remains far from conclusive. Several methodological and conceptual limitations must be considered before translating these findings into clinical practice or formal therapeutic recommendations (Table 4).

Table 4.

Ongoing/registered trials of SGLT-2 inhibitors with cognitive/biomarker endpoints

Trial	Population	Intervention	Comparator	Endpoints	Biomarkers/imaging
DAPA-MIND (NCT03801642)	Patients with heart failure	Dapagliflozin	Placebo	Cognitive testing; cerebral perfusion	ASL-MRI
EMPA-KIDNEY cognition substudy	Patients with chronic kidney disease	Empagliflozin	Placebo	MoCA; incident dementia	NfL; GFAP; volumetry
Planned AD phase II pilot	Early AD + insulin resistance	Dapagliflozin	Placebo	ADAS-Cog	p-tau217; Aβ42/40; ASL-MRI

Lack of Cognitive Endpoints in Clinical Trials

Most large-scale randomized controlled trials (RCTs) of SGLT-2 inhibitors have focused on cardiovascular and renal outcomes, with cognitive function not included as a predefined endpoint. Although post hoc analyses and observational studies suggest cognitive benefit, these findings are inherently limited by the lack of standardized neuropsychological assessments, unmeasured confounders, and potential indication bias [57, 58]. Current evidence is insufficient to establish causality or determine the magnitude of neuroprotective effects.

Short Duration and Population Bias

Many studies have limited follow-up durations (often < 5 years), which may not be sufficient to capture the slow progression of neurodegenerative diseases like Alzheimer’s or Parkinson’s disease. In addition, most study populations include patients with type 2 diabetes, cardiovascular disease, or chronic kidney disease—groups at high risk for cognitive decline, but not fully representative of the broader population at risk for neurodegeneration [56, 59]. Thus, the generalizability of current findings to non-diabetic or younger populations is uncertain.

Unclear Mechanistic Pathways in Humans

While animal models have demonstrated multiple potential mechanisms of SGLT-2-mediated neuroprotection (e.g., reduced inflammation, improved mitochondrial function, enhanced autophagy), the extent to which these findings translate to humans remains unclear. There is limited direct evidence confirming these effects in the human brain, particularly in the context of Alzheimer’s or Parkinson’s pathology [40, 68].

Unknown Class Effects and Drug-Specific Differences

It remains uncertain whether the observed neuroprotective effects are a class effect of all SGLT-2 inhibitors or are more prominent in specific agents (e.g., dapagliflozin vs. empagliflozin). Differences in pharmacokinetics, blood–brain barrier penetration, and off-target effects may lead to variable outcomes [30, 34]. Comparative studies are needed to identify whether certain agents are more suitable for cognitive protection.

Cognitive Safety vs. Therapeutic Benefit

A critical distinction must be made between cognitive safety and true cognitive enhancement or disease modification. Even if SGLT-2 inhibitors reduce dementia risk in patients with diabetes, it is unclear whether they meaningfully alter the course of neurodegenerative diseases or improve quality of life. Ongoing trials such as DAPA-MIND (NCT03801642) and EMPA-KIDNEY cognition substudies are expected to provide key insights.

Future Research Directions

To address these limitations, several avenues should be prioritized in future investigations:

• Dedicated RCTs with cognitive endpoints: Large, prospective studies should evaluate cognitive performance over time using validated tools (e.g., Montreal Cognitive Assessment [MoCA], Alzheimer’s Disease Assessment Scale–Cognitive Subscale [ADAS-Cog]) in both diabetic and non-diabetic populations at risk for neurodegeneration.

• Biomarker-driven studies: Incorporation of cerebrospinal fluid (CSF) biomarkers, neuroimaging (e.g., amyloid positron emission tomography [PET], fluorodeoxyglucose [FDG]–PET, tau PET tracers such as AV1451 and PI2620), and blood-based markers (e.g., plasma Aβ, tau, neurofilament light chain [NfL]) will help clarify mechanisms and patient subgroups most likely to benefit.

• Mechanistic human studies: Translational studies, including brain imaging and post-mortem analyses, are needed to determine whether observed benefits in animal models are replicable in humans.

• Comparative effectiveness research: Head-to-head comparisons between SGLT-2 inhibitors and other metabolically active agents (e.g., GLP-1 receptor agonists, metformin) will help delineate the specific role of SGLT-2 inhibition in brain health.

• Exploration in other neurodegenerative diseases: Although most data have focused on Alzheimer’s disease, future studies should evaluate the effects of SGLT-2is in Parkinson’s disease, vascular dementia, and mixed cognitive impairment.

Ultimately, while the evidence to date is promising, it remains preliminary. Rigorous clinical trials and mechanistic studies will be essential to establish whether SGLT-2 inhibitors can be repositioned as effective agents for neuroprotection and dementia prevention.

Related Disorders Without Effective Treatments

Hypothesis Box: Application to atypical parkinsonisms (PSP, CBS, MSA) remains hypothetical, with no supportive human data; these statements should be interpreted as speculative only.

Hypothesis note: The following considerations about atypical parkinsonisms (PSP, CBS/corticobasal degeneration [CBD], MSA) are presented as hypotheses; direct human evidence is currently lacking.

While current clinical evidence for the use of SGLT-2 inhibitors in neurodegenerative diseases predominantly focuses on AD and PD, there is growing interest in exploring these agents as potential therapeutic options for atypical parkinsonisms. However, it is important to underscore that atypical parkinsonisms are a heterogeneous group of disorders with distinct pathological substrates. For example, PSP and CBD are classified as tauopathies, whereas multiple system atrophy (MSA) is considered a synucleinopathy, similar to PD. These differences may significantly influence the potential utility and mechanisms of action of SGLT-2 inhibitors across subtypes. Moreover, emerging evidence suggests that CBS, one of the clinical correlates of CBD, may develop on a vascular substrate, further complicating its pathophysiology and therapeutic approach. Additionally, arterial hypertension has been identified as a potential risk factor for the development of progressive supranuclear palsy, indicating a possible vascular contribution to its pathogenesis [72–74].

Hypotheses for Future Research

Future investigations should carefully differentiate between atypical parkinsonism subtypes, acknowledging their divergent pathological mechanisms and clinical profiles, particularly regarding cognition. Promising research avenues include:

• Subtype-specific trials evaluating SGLT-2 inhibitors as adjunct therapies, with separate arms for tauopathies (e.g., PSP, CBD) and synucleinopathies (e.g., MSA).

• Cognitive endpoints in tauopathy-dominant diseases, where cognitive decline is a defining feature and may be more amenable to metabolic or inflammatory modulation.

• Biomarker-driven stratification, using markers like tau, α-synuclein, or neurofilament light chain (NfL) to assess drug impact in a mechanistically informed way.

While there is no direct evidence to support the use of SGLT-2 inhibitors in atypical parkinsonisms, the pathophysiological overlap with other neurodegenerative diseases and the promising preclinical data suggest that these agents could serve as potential disease-modifying therapies. However, rigorous clinical trials are needed to test these hypotheses and evaluate the safety and efficacy of SGLT-2 inhibitors in patients with atypical parkinsonisms. Such trials would not only expand our understanding of these disorders but also offer new therapeutic avenues for a group of diseases with currently limited treatment options.

Conclusions

SGLT-2 inhibitors, originally developed for glycemic control in type 2 diabetes mellitus, have emerged as promising agents with pleiotropic effects that may extend to the central nervous system. Preclinical evidence consistently demonstrates neuroprotective actions through multiple mechanisms, including improved cerebral energy metabolism, attenuation of neuroinflammation, enhancement of insulin signaling, reduction of oxidative stress, and preservation of the blood–brain barrier.

Observational clinical studies suggest a potential association between SGLT-2 inhibitor use and reduced risk of dementia, particularly in patients with diabetes and cardiovascular comorbidities. However, causality remains unproven, and current data are limited by the absence of dedicated cognitive endpoints in randomized clinical trials.

While these findings offer an exciting opportunity for therapeutic repurposing, further research is essential. Prospective, biomarker-supported clinical trials with standardized cognitive assessments are needed to determine the true clinical impact of SGLT-2 inhibitors on neurodegenerative disease progression. Additionally, mechanistic studies in humans will be crucial to validate the biological plausibility of their neuroprotective properties.

In summary, SGLT-2 inhibitors represent a novel and potentially valuable avenue in the prevention or modulation of neurodegenerative diseases. Their expanding role underscores the importance of metabolic health in brain aging and highlights the need for integrated, multidisciplinary approaches in tackling cognitive decline.

Author Contributions

Paulina Kostrzewska, Pawel Kuca, Przemysław Witek, Jolanta Małyszko, Natalia Madetko-Alster, and Piotr Alster contributed to the study conception and design. Literature analysis and synthesis were performed by Piotr Alster, Przemysław Witek, and Jolanta Małyszko. The first draft of the manuscript was written by Paulina Kostrzewska, Piotr Alster, and Natalia Madetko-Alster, and all authors reviewed and approved the final version.

Funding

No funding or sponsorship was received for this study. The journal’s publication fee was funded by the authors.

Data Availability

Data sharing is not applicable to this article as no datasets were generated or analyzed during the current study.

Declarations

Conflict of Interest

Paulina Kostrzewska, Pawel Kuca, Przemysław Witek, Jolanta Małyszko, Natalia Madetko-Alster, and Piotr Alster declare that they have no conflicts of interest.

Ethical Approval

This article is based on previously conducted studies and does not contain any new studies with human participants or animals performed by any of the authors.