3.1 Hyperglycemia

Hyperglycemia is the most prevalent symptom among diabetic patients. Chronic hyperglycemia contributes to cognitive impairment by inducing oxidative stress, neuroinflammation, and advanced glycation end-product (AGE) accumulation, leading to neuronal damage [11]. Three clinical studies found that patients with diabetes frequently experience acute and temporary cognitive impairment associated with hyperglycemia, which could impact their quality of life and daily activities [35]. A clinical study reported that the visual reaction time of patients with T1DM was obviously delayed during a clamp study when blood glucose levels reached 16.7 mmol/L [36]. Similarly, blood glucose levels in the 20–30 mmol/L range were associated with a 9.5% decrease in intelligence quotient performance among children with T1DM [37]. Another clinical study conducted a hyperinsulinic glucose clamp study on patients with T2DM at 14.5 and 16 mmol/L, revealing notable disruptions in performance on complex cognitive function tests during hyperglycemic episodes [38]. Additionally, Clinical studies assessed daily cognitive and motor function in patients with T1DM and T2DM, finding that some patients exhibited obvious cognitive dysfunction at a hyperglycemia threshold of approximately 15 mmol/L. While the adverse effects of hyperglycemia varied among diabetic patients, the resulting increase in errors and slower responses during basic verbal and mathematical tasks affected various daily functions, such as managing finances, calculating insulin doses, and performing at school or work [35].

3.2 Hypoglycemia

Hypoglycemia, a common occurrence in the stringent management of diabetes, is recognized as an important risk factor for dementia [11]. Recurrent hypoglycemia is associated with impaired neuronal energy metabolism and increased dementia risk due to excitotoxicity and synaptic dysfunction [39]. Two cohort studies established a clear link between episodes of hypoglycemia and cognitive decline in patients with diabetes [40, 41]. Elderly diabetic patients are particularly vulnerable to hypoglycemia due to various comorbidities, undernutrition, polypharmacy, and renal or hepatic damage [42]. In a retrospective study of elderly patients with T2DM, the risk of dementia increased by 26% to 94% with a higher number of severe hypoglycemia episodes [43]. Another observational cohort study found that elderly patients with T2DM who experienced hypoglycemia had a greater risk (HR: 2.689) of dementia compared to those without such episodes [44]. Additionally, diabetic patients with dementia or cognitive dysfunction are much more likely to be hospitalized for hypoglycemia than those with normal cognitive function [45-47]. However, a prospective cohort and a randomized controlled trial (RCT) found that hypoglycemia does not necessarily heighten the risk of cognitive dysfunction in diabetic patients [48, 49]. This may be attributed to the fact that individuals experiencing more hypoglycemia may be under stricter glycemic control, which may mitigate the neurocognitive damage associated with hypoglycemia [50].

3.3 Fluctuation of Blood Glucose

Glucose fluctuations further exacerbate cognitive deficits by triggering vascular endothelial dysfunction and blood–brain barrier disruption [51]. Recent reviews indicated that fluctuations in glucose levels may be linked to cognitive dysfunction as well as dementia risk in diabetic patients [52, 53]. An observational study found that the glycoalbumin (GA) /hemoglobin A1c (HbA1c) ratio, an effective marker for fluctuations of blood glucose levels, was negatively associated with cognitive function scores in elderly patients with T2DM [54]. Similarly, a cross-sectional, observational study found that varying glucose levels were associated with cognitive impairment in patients with T2DM to some degree [55]. A retrospective study also reported a consensus on this issue [56]. Observational studies found that fluctuations in blood glucose were associated with increased oxidative stress and vascular endothelial dysfunction in T2DM [57], contributing to the onset and progression of dementia [58].

3.4 Diabetic Retinopathy

Diabetic retinopathy (DR) is a common microvascular complication of diabetes [59], and an increasing body of research links it to cognitive decline [15, 60]. A longitudinal study found that diabetic patients with DR had over twofold higher odds of experiencing worsening cognitive impairment, while those with moderate or severe DR had threefold higher odds compared to diabetic patients without DR [61]. An observational study identified DR as a contributing risk factor for cognitive dysfunction in T2DM [15]. Moreover, a current meta-analysis suggested that DR is associated with an increased risk of cognitive dysfunction [62], with a positive correlation between the severity of DR and cognitive dysfunction [63]. Reviews showed a strong connection between retinopathy and cerebral microvascular injury, suggesting that microvascular dysfunction was a key mechanism driving cognitive decline in diabetes [64, 65]. Thus, we hypothesize DR may serve as an important indicator for identifying individuals at high risk of cognitive dysfunction.

3.5 Diabetic Nephropathy

Diabetic nephropathy is another common microvascular complication of diabetes, and observational studies found a close relationship between diabetic nephropathy and cognitive function in diabetic patients [66, 67]. A retrospective study found that both the MMSE and MoCA scores were negatively correlated with urinary albumin excretion rates and positively correlated with the estimated glomerular filtration rates (eGFR) in patients with type 2 diabetic nephropathy [68]. Similarly, decreased cognitive function was associated with kidney disease in diabetic patients, as assessed by the albumin/creatinine ratio (ACR), which reflected microvascular endothelial damage, and cystatin C, a marker of eGFR [66]. Studies suggested that cognitive dysfunction related to diabetic nephropathy may arise because brain and kidney damage share similar microvascular lesions [69, 70]. Additionally, the kidneys may serve as the primary clearance pathway for Aβ; the impairment of their function may further aggravate cognitive dysfunction [68, 71].

3.6 Cardiovascular Diseases

Systematic reviews consistently showed that patients with a history of cardiovascular diseases, including coronary heart disease, heart failure, and atrial fibrillation, had an increased risk of cognitive impairment [72, 73]. An observational study found that endothelial dysfunction was associated with cognitive impairment in elderly patients with cardiovascular disease [74]. This association is also supported by a prospective study, which found that a history of stroke (OR: 1.523) and cardiovascular diseases (OR: 1.258) was associated with a higher risk of dementia in patients with T2DM [75]. In addition, cardiovascular risk factors, particularly hypertension and dyslipidemia, may contribute to cognitive dysfunction in patients with T2DM [11, 52, 53]. The presence of arterial hypertension in diabetic patients also affects cognitive function [76]. A large network meta-analysis found that antihypertensive therapy improved cognitive function, except in the language domain [77]. However, lowering blood pressure may potentially reduce cerebral perfusion, leading to an increased risk of cardiovascular complications [78]. Therefore, we think the effectiveness of antihypertensive strategies for preventing cognitive decline remains controversial.

3.7 Cerebrovascular Diseases

Cerebrovascular diseases, which often coexist with Alzheimer's disease, include atherosclerosis, arteriosclerosis, and cerebral amyloid angiopathy. Both arteriosclerosis and cerebral amyloid angiopathy fall under the category of small vessel disease, an important vascular contributor to dementia [79, 80]. Chronic hyperglycemia, insulin resistance, and vascular dysfunction contribute to the acceleration of cerebrovascular pathology, increasing susceptibility to cognitive impairment [64]. A RCT showed that in patients with type 1 diabetes and proliferative retinopathy, baseline white matter lesions and reduced skin capillary perfusion were associated with a decline in general cognitive ability over time, regardless of age, sex, HbA1c levels, and severe hypoglycemic events [81]. Moreover, atherosclerosis could lead to cognitive impairment and dementia through thromboembolic stroke in patients with T2DM [28]. An observational study found that diabetes was associated with poor performance on cognitive tests measuring information-processing speed and executive function, with this relationship partially mediated by markers of cerebrovascular disease [82]. In a cohort study of seniors with diabetes, the risk of dementia was highest (HR: 2.03) among those with a history of cerebrovascular disease [40]. Generally, cognitive decline associated with cerebrovascular disease tends to occur gradually and in a stepwise manner, slowly impacting processing speed, complex attention, and frontal executive function [83].

3.8 Genetic Predisposition

Genetics is a vital factor influencing cognitive impairment and Alzheimer's disease [84-86]. Notably, the 4 allele of apolipoprotein E (APOE) stands out as the most reliable and critical genetic risk factor [87]. The effects of APOE4 include damaged integrity of blood–brain barrier [88], increased accumulation of Aβ [89], and alterations in Aβ metabolism [90]. Moreover, APOE4 makes the cerebrovascular system more vulnerable to damage from diabetes and exacerbates its effects [91]. A cross-sectional study found that the impact of diabetes on cognitive performance was more pronounced in individuals carrying one or more APOE ε4 alleles [92]. In a study based on mice, APOE4 mice exhibited greater vulnerability to the detrimental cognitive effects of high-fat diet-induced insulin resistance, likely due to APOE subtype-specific differences in brain metabolism [93]. Furthermore, a clinical cohort study found that rs391300 single-nucleotide polymorphism in the serine racemase gene, which linked to a higher risk of type 2 diabetes, was also associated with the progression from MCI to probable Alzheimer's disease [94]. However, a population-based cohort study found that APOE4 genotypes were not associated with cognitive dysfunction in patients with T1DM [95]. More diabetes-specific genetic studies are needed to clarify these associations and their clinical implications for DACD.

3.9 Depression

Patients with diabetes are at an increased risk of developing depression, which is a risk factor for dementia [96, 97]. A systematic review reported that the prevalence of depression among diabetic patients was more than 2–3 times higher than that of those without diabetes [96]. Moreover, a population-based cohort study showed that the risk ratio for developing dementia in patients with depression was 1.83, for those with diabetes was 1.2, and for those with both conditions was 2.17, indicating that the combined effect of depression and diabetes on dementia risk was greater than simply additive [97]. Another cohort study similarly found that diabetic patients with depression had a two-fold increased risk of developing dementia compared to those with diabetes alone [98]. Although some debate exists, most structural imaging studies demonstrated that a smaller hippocampal volume was associated with memory dysfunction in depression [99, 100]. However, we think that the underlying cerebral pathology contributing to cognitive dysfunction in depression still requires further investigation.

3.10 Other Risk Factors

Beyond the previously mentioned risk factors, several common factors also influence cognitive function in diabetes. Advanced age is a well-established risk factor for cognitive decline. Two cross-sectional studies indicated that in patients with T2DM, MMSE score was significantly negatively correlated with age. MMSE score was positively correlated with education level [60, 101]. Education may serve as a protective factor by promoting cognitive reserve, which helps delay the onset of cognitive impairment. Additionally, cognitive decline is positively associated with a longer duration of diabetes [60, 102, 103]. Furthermore, physical inactivity, smoking, and frequent alcohol consumption have also been associated with cognitive impairment in diabetic patients [104, 105]. Sedentary behavior and lack of exercise are also positively associated with cognitive impairment in diabetic patients. Physical activity can improve insulin sensitivity, reduce neuroinflammation, and enhance cognitive function, suggesting that an active lifestyle may mitigate the impact of diabetes on cognition [106]. Moreover, smoking and excessive alcohol consumption accelerate vascular damage and oxidative stress, both of which contribute to DACD [107]. Furthermore, sleep disturbances are also positively associated with impaired cognitive function in patients with T2DM, possibly through mechanisms involving insulin resistance and amyloid deposition [108]. Consequently, these modifiable lifestyle and demographic factors play an important role in the development of DACD. We believe that encouraging healthy behaviors, such as regular physical activity, smoking cessation, moderated alcohol intake, and cognitive engagement, may help reduce cognitive decline in diabetic patients.

4.1 Insulin Resistance

It is widely recognized that insulin mainly functions to regulate blood glucose and energy metabolism in the periphery. Two studies based on rodents found that insulin in the brain promoted the growth and development of nerve cells, regulated the release of neurotransmitters, and played a crucial role in cognitive functions such as learning and memory [109, 110]. Insulin and its receptors are extensively expressed in neurons and glial cells, particularly in the cerebral cortex and hippocampus, areas closely related to cognition [111, 112].

Insulin resistance (IR) is characterized by a decreased sensitivity of insulin in target organs. Prolonged hyperinsulinemia resulting from IR can impair the function of blood–brain barrier (BBB) and insulin activity [113], thereby promoting IR in the brain [112, 114]. Additionally, mitochondrial dysfunction and elevated production of reactive oxygen species, observed in the hippocampus in rodent models of T2DM, also contribute to hippocampal IR. Neuroinflammation is another potential mechanism underlying hippocampal IR in T2DM [30].

Brain IR exposes neurons to high insulin levels for extended periods, resulting in neuronal degeneration and memory impairment [115]. Increasing evidence supports the concept that brain IR is crucial to the pathophysiologic mechanisms of cognitive dysfunction in diabetic patients and may serve as a bridge between diabetes and Alzheimer's disease. A cross-sectional study found that type 2 diabetic patients developed abnormal functional connectivity in the posterior cingulate cortex, which was associated with insulin resistance in specific brain regions and could play a key role in evaluating the cognitive dysfunction in T2DM [116]. Brain insulin resistance also worsens cognitive function by disrupting brain networks [117]. Notable IR was detected in the hippocampus of diabetic rat models with learning and memory deficits, generated by injection of streptozotocin, and insulin treatment improved the impaired cognitive function in these diabetic rats while increasing the expression levels of phosphorylated insulin receptor and insulin receptor substrate 1 in the hippocampus [118]. Moreover, insulin resistance in hippocampal microvascular may play a critical role in the development of cognitive impairment in T2DM [119]. A study based on mice found that the combination of insulin resistance and hyperinsulinemia did not accelerate plaque formation or memory abnormalities in Alzheimer's disease mice carrying the insulin receptor mutation, but these mutations reduced oxidative damage in mice brains [120]. However, another study based on mice found that mice with a brain-specific knockout of insulin exhibited brain mitochondrial dysfunction characterized by decreased mitochondrial oxidative activity, increased levels of reactive oxygen species, elevated lipid, and protein oxidation levels in the striatum and nucleus accumbens, along with age-related anxiety and depressive-like behaviors [121]. Moreover, brain IR drives pro-apoptosis, pro-inflammatory, pro-phosphorylation of tau protein, and pro-Aβ cascades [122, 123].

Therefore, the insulin-signaling pathway is critical for cognitive functions. It primarily activates two key pathways: the phosphatidylinositide 3-kinase (PI3K)/Akt pathway and the mitogen-activated protein kinase (MAPK)/extracellular signal-regulated kinase (ERK) kinase pathway [30, 124]. In high-fat diet-induced obesity model mice, exercise improved cognitive function by enhancing the activity of hippocampal insulin signaling (PI3K/Akt) [125]. The activation of the PI3K/Akt pathway phosphorylates and inactivates glycogen synthase kinase 3β (GSK3β), thereby decreasing the ability of GSK3β to phosphorylate the microtubule-associated protein tau [30]. When insulin signaling is diminished, normal soluble tau levels decrease, while hyperphosphorylated tau accumulates, exacerbating neuronal cytoskeletal collapse, neurite retraction, and impaired synapse formation [126, 127].

We conclude that these suggest that restoring insulin activity in the brain could be an effective strategy to alleviate the cognitive decline associated with T2DM.

4.2 Neuroinflammation

Chronic low-grade neuroinflammation is widely considered a mechanism underlying brain changes in T2DM. Studies based on diabetic rodents found that impaired memory was associated with increased levels of proinflammatory cytokines (TNF-α, IL-1β, and IL-6), while inhibiting neuroinflammation improved cognitive outcomes, suggesting a relationship between neuroinflammation and memory impairment [128-131]. Causes of nervous system inflammation in diabetes include the entry of peripheral inflammatory cytokines into the brain through disrupted BBB and the activation of microglia triggered by neuronal damage [102, 132, 133]. Several signaling molecules, such as toll-like receptor 4 (TLR4)/nuclear factor-κB (NF-κB) and triggering receptor expressed on myeloid cells 2 (TREM2), play a critical role in regulating microglial activation and neuroinflammation [134-136]. NF-κB, a transcription factor involved in the expression of inflammatory cytokines and chemokines, regulated the inflammatory cascade enhancers in brain cells of a mice model of Alzheimer's disease [137]. Moreover, overexpression of TREM2 in the hippocampus reduced neuroinflammation and microglial activation, improving cognitive impairment caused by a high-fat diet through the NF-κB signaling pathway [138]. Deficiency of peroxisomal proliferator-activated receptor β/δ (PPARβ/δ) also involved in the activation of astrocytes and microglia, increasing neuroinflammatory markers [139]. Additionally, lipin 2 upregulation in the hippocampus of diabetic encephalopathy mice reduced NLRP3 inflammasome-mediated inflammation and improved cognitive function by inhibiting the JNK/ERK signaling pathway [140]. Astrocytes, alongside microglia, are important immune regulatory cells in the central nervous system (CNS), capable of initiating inflammation by secreting cytokines when stimulated [141]. DPP4 (CD26) bound to the insulin-like growth factor 2-receptor on Treg cells, impairing Tregs function, polarizing microglia toward a pro-inflammatory phenotype in the hippocampus, and ultimately leading to neuroinflammation and cognitive impairment in T2DM [142]. Glycation end products induced by hyperglycemia are also implicated in diabetes-related neuroinflammation. Neuroinflammation can disrupt brain mitochondria and neurotransmitter functions, ultimately leading to neuronal damage and subsequent cognitive decline [143, 144].

4.3 Microvascular Dysfunction

The primary role of cerebral microcirculation is to supply nutrients and energy while clearing waste products in response to local neural activity. Cerebral microvascular dysfunction is a common issue in T2DM, and growing evidence suggests it may be a key mechanism in DACD [60, 64]. Two observational studies found that higher scores of microvascular dysfunction—based on MRI-detected cerebral small vessel disease, retinal arteriolar and venular dilation response to flicker light, albuminuria, and plasma biomarkers—were associated with lower cognitive function [145, 146]. In T2DM, cerebral microvascular pathology is characterized by basement membrane thickening, increased angiogenesis, increased BBB permeability, and altered blood flow regulation, all of which may contribute to cognitive dysfunction [64, 132, 147]. Additionally, cerebral blood flow reduction in T2DM is likely associated with a higher prevalence of cerebrovascular risk factors [148, 149]. Microvascular complications are also associated with cognitive dysfunction in T1DM, where microvascular issues tend to experience both cognitive decline and its acceleration [2, 150]. A review highlighted cerebral microvascular senescence as an important factor in DACD [151].

BBB dysfunction is a significant indicator of cerebral vascular disease. The endothelial function of BBB is essential for regulating ion balance, nutrient transport, blocking neurotoxic molecules, and maintaining brain homeostasis, making it pivotal to the development and progression of DACD [129, 152]. Increased BBB permeability was associated with hyperglycemia and memory deficits in both T1DM and T2DM mouse models [129].

Key drivers of diabetes-associated cerebral microvascular dysfunction include hyperglycemia, obesity, insulin resistance, and hypertension [64]. Due to similar embryologic and anatomic properties, examining retinal vessels serves as a valuable alternative marker for assessing cerebral microvascular disease [153].

4.4 Oxidative Stress

Oxidative stress is recognized as a vital factor in the development and progression of cognitive impairment in diabetes [154]. It impacts insulin resistance, neuroinflammation, neurotransmitters, and lipid metabolism in the diabetic brain [131, 154, 155]. An observational study found that elevated glucose and HbA1c levels were negatively associated with glutathione (GSH) levels in the anterior cingulate of prediabetic/diabetic patients with cognitive impairment, indicating a potential role for oxidative stress in its pathophysiology [156]. In diabetic rats, oxidative and nitrosative stress worked together to damage brain mitochondria through increased reactive oxygen species (ROS) and nitric oxide (NO), and decreased GSH levels [157]. The generation of ROS in the diabetic brain activated multiple cellular pathways, including the AGE, polyol, and protein kinase C pathways, leading to neuroinflammation and neurodegeneration [1, 158]. Superoxide stimulates the production of other ROS and the chelation of NO, disrupting cerebrovascular tension and NO-dependent dilation. Antioxidant treatments were shown to fully reverse Aβ-induced cerebrovascular dysfunction in elderly amyloid precursor protein (APP) mice, suggesting oxidative stress as a deleterious mechanism through which Aβ affected cerebral vascular reactivity [159]. Studies suggested that antioxidative stress, including activating sirtuin 1 (SIRT1) [160, 161], Nrf2 [161], GPR55 [162], ChemR23 signaling [163], or carnosine [164], improved cognitive function in diabetic rodent models by reducing inflammatory cytokines and malondialdehyde (MDA) levels while increasing levels of GSH, superoxide dismutase (SOD), and catalase.

4.5 Neurotransmitter Disorders

Neurotransmitters are biologically active molecules essential for maintaining the physiological function of the brain, and changes in their metabolism are closely associated with various neurodegenerative diseases [165]. In the CNS, insulin plays a key role in learning and memory by regulating the secretion and reuptake of neurotransmitters, such as acetylcholine (ACh), norepinephrine, and epinephrine, and promoting the accumulation of γ-Aminobutyric acid (GABA) receptors to the post-synaptic membrane [1]. Reduced sensitivity to insulin in the hippocampus and cortex can disrupt neurotransmitter synthesis, contributing to cognitive dysfunction. Moreover, pro-inflammatory mediators may also influence neurotransmitter synthesis [158]. Therefore, we think that alterations in neurotransmitter levels may be central to DACD.

GABA, the primary inhibitory neurotransmitter in the adult mammalian brain, is decreased in DACD [166, 167]. In addition, hyperglycemia made GABAergic neurons in the hippocampus more vulnerable, and their dysfunction, combined with overexcitation of glutamatergic neurons, caused glucose-induced neurotoxicity during diabetic encephalopathy in mouse models of diabetes [168]. In diabetic rat models, ACh levels were reduced in the hippocampus and hydrolyzed by acetylcholinesterase (AChE) [167]. A study found a significant decrease in ACh in the CSF of diabetic rats with major cognitive damage, which was possibly associated with changes in Aβ42, P-tau, and IL-6 [169]. Moreover, increased cholinesterase activities impaired neuronal function, a phenomenon observed in chronic diabetes patients and rodent models, while decreased activities of cholinesterase improved cognitive impairment in diabetic rodent models [170-172].

Dopamine, a key neurotransmitter, binds to regulatory presynaptic autoreceptors or postsynaptic receptors in the synaptic cleft to trigger action potential, and the presence of diabetes impacts dopaminergic neurotransmission [173]. The increased dopamine clearance and decreased dopamine signaling were observed in rodent models with a brain-specific knockout of the insulin receptor, suggesting that insulin resistance in the brain altered dopamine turnover [121]. Additionally, obviously reduced levels of serotonin in the hippocampus also contributed to DACD in diabetic rodent models [165, 173, 174]. Furthermore, excitatory neurotransmitters like glutamate and glycine were elevated in DACD [167].

5.1 β-Amyloid (Aβ) Deposition

The primary pathology underlying dementia is the accumulation of Aβ; however, the Aβ hypothesis remains controversial in the context of DACD. A population-based cohort study found that diabetic patients were at a higher risk of having a positive amyloid status [175]. Two studies showed that diabetes might contribute to cognitive dysfunction by exacerbating Aβ accumulation in the hippocampus of mice [176, 177]. This accumulation was associated with the upregulation of the receptor for advanced glycation end products (RAGE) and AGE, the downregulation of low-density lipoprotein receptor-related protein and vacuolar protein sorting-associated protein 26a, or insulin resistance [178-182]. In diabetic rodent models, Aβ expression in the hippocampus was obviously higher than controls, and Aβ deposition was positively associated with autophagy markers, while being negatively associated with lysosome function markers [183-185]. This indicates that the autophagy-lysosome pathway may be involved in Aβ deposition in diabetic cognitive impairment. Moreover, neuronal calcium overload is an important factor in cognitive dysfunction. Knocking out the transient receptor potential cation channel 6, which is highly correlated with intracellular Ca²⁺ levels, decreased Aβ production in T2DM mice [186]. This suggests that intracellular Ca²⁺ concentrations regulate Aβ deposition in the brain. However, a study observed that diabetes accelerated cognitive dysfunction through cerebrovascular Aβ deposition rather than brain Aβ accumulation in an Alzheimer mouse model with diabetes [187].

Antagonizing Aβ in the hippocampus of diet-induced obesity rats could reverse cognitive deficits, highlighting the important role of Aβ in the pathophysiology of cognitive dysfunction [188]. Furthermore, the improvement of cognitive function in diabetic rodent models was accompanied by a reduction in hippocampus Aβ levels [189, 190].

However, a cross-sectional study reported that T2DM was not associated with CSF Aβ levels regardless of whether participants had diabetes or MCI [21]. Similarly, a study in mice with T2DM and Alzheimer's disease found that cognitive impairment caused by diabetes was unlikely to be directly dependent on Aβ deposition [191]. Therefore, we suggest that further research is needed to better understand the role of Aβ in DACD.

5.2 Tau Protein Hyperphosphorylation

Tau is a microtubule-associated protein that is highly abundant in the CNS [192]. Hyperphosphorylation of tau is another feature of the diabetic brain, mainly driven by the activation of GSK-3β, which is regulated by the insulin-PI3K-Akt signaling pathway [193, 194]. Compared to the control group, diabetic patients and rodent models exhibited elevated levels of P-tau in their CSF [21, 169]. Diabetes exacerbated cognitive dysfunction by increasing P-tau in the hippocampus of mouse models [176, 177, 195]. Protein phosphatase 2A is considered the main tau phosphatase responsible for regulating tau phosphorylation, and its methylation can reduce tau phosphorylation [194, 196]. In addition, activation of the mTOR pathway also led to hyperphosphorylation of tau in the hippocampus of diabetic rodent models [197, 198]. Chronic hyperglycemia in diabetic mice induced tau hyperphosphorylation by O-GlcNAc transferase-involved O-GlcNAcylation, both in vivo and in vitro, leading to DACD [199]. High glucose levels also promoted tau hyperphosphorylation in diabetic rodent models via ALKBH5-mediated demethylation of Dgkh m⁶A [200] and downregulation of VPS26a [182], further contributing to DACD. Moreover, insulin deficiency also promoted tau hyperphosphorylation in an Alzheimer's disease mice model [180]. Improvements in cognitive function in diabetic rodent models were accompanied by a reduction in tau hyperphosphorylation, suggesting that tau hyperphosphorylation is a critical pathological change in DACD [201-203]. However, some observational studies and diabetic animal model studies argue that P-tau might not play a major role in the negative impact of diabetes on cognitive performance [204-207]. Therefore, we recommend that further research is needed to clarify the role of tau hyperphosphorylation in DACD.

5.3 Synapses and Dendritic Spines Damage

Synaptic plasticity, which refers to the activity-dependent changes in strength of neuronal connections and the modification of synaptic transmission, is a critical element in learning and memory [208]. Impairments in synaptic plasticity contribute significantly to DACD. These impairments are often assessed by long-term potentiation (LTP) and various markers of synapses and axons, such as SNAP-25, synaptophysin (SYP), postsynaptic density protein 95 (PSD 95), GAP-43, and NfL [16, 209, 210]. Diabetic conditions can aggravate synaptic loss, leading to cognitive impairment [177, 211, 212], potentially through the enhancement of the interaction between F1F0 ATP synthase and Cyclophilin D [213, 214]. In diabetic rodent models with cognitive dysfunction, the number of synapses, the length of postsynaptic densities, and dendritic spine density were reduced in the hippocampus, along with axon growth and synaptic proteins such as GAP 43, SYP, and PSD 95. High glucose exposure reduced the length of the longest neurite as well as the axon initial segment [215-217]. High glucose also contributed to synaptic protein loss via impairing thrombospondin-1 secretion from astrocytes [218]. A study in diabetic mouse models found that elevated tau levels exacerbated synaptic impairments in T1DM [195]. Some diabetes medications, while improving cognitive dysfunction, also enhance synaptic dysfunction by upregulating synaptic proteins including PSD 95, brain-derived neurotrophic factor (BDNF), SYP, and synapsin-1 (SYN1) in DACD rodent models [219-221]. We conclude that these findings reveal the function of synaptic damage in the development of DACD.

5.4 Microglia Proinflammatory Polarization

Microglia, the resident macrophages of the CNS, contribute to the development of neurodegenerative diseases [222, 223]. In a diabetic rat model, knocking down microglia reduced neuroinflammation and improved cognitive function [133]. However, excessive activation of microglia damages surrounding healthy neural tissue, and the factors released by dying or damaged neurons further exacerbate chronic microglial activation, causing progressive loss of neurons [224]. Traditionally, microglia are categorized into the proinflammatory M1 phenotype, the alternatively activated or deactivated phenotype, and the immunosuppressive M2 phenotype [224, 225]. Elevated levels of proinflammatory cytokines in the diabetic brain indicate that the innate immune system, particularly activated microglia, plays a crucial role in neuronal damage in both diabetic animals and patients [102, 226]. Moreover, several factors drive microglial activation toward the proinflammatory M1 phenotype under diabetic conditions [138, 142, 227, 228]. Interventions that reduce neuroinflammation could attenuate DACD by promoting M2 polarization of microglia and inhibiting M1 polarization [229, 230].

6.1 Antidiabetic Drugs

6.2 Lifestyle and Behavioral Interventions

Increasing evidence suggests that modifiable lifestyle factors play a critical role in mitigating cognitive decline in diabetic patients. These non-pharmacological approaches can complement pharmacological treatments and may offer neuroprotective benefits.

6.3 Other Drugs

Recent studies have explored the potential of various compounds in mitigating cognitive dysfunction in diabetic rodent models.

In some synthetic compounds, AB-38b was reported to enhance cognitive function in diabetic mice by modulating oxidative stress pathways [312], while Gypenoside LXXV (GP-75) increased the brain glucose uptake and hippocampal GLUT4 expression levels [313]. Moreover, some herbal extracts and phytochemicals, such as rolipram [128], Forsythoside B [220] and Rosa canina L. [314], and ginsenoside Rb1 [315] extract, showed neuroprotective effects by reducing Aβ aggregation and inhibiting hippocampal neuroinflammation in streptozotocin (STZ)-induced diabetic rodent models. Traditional Chinese Medicines were also demonstrated potential in attenuating cognitive dysfunction in diabetic rodent models. Berberine showed to improve insulin sensitivity and reduce oxidative stress, potentially delaying diabetes-associated cognitive decline [171, 316]. Huang-Lian-Jie-Du was demonstrated antioxidative and anti-inflammatory effects in diabetic rats, improving memory performance [317]. Jiawei Shengmai San formula was reported to reduce neuronal cell necrosis and inflammatory cell infiltration in the hippocampus in diabetic rats [318]. Zi Shen Wan Fang modulated gut microbiota and brain-gut axis interactions, contributing to cognitive improvement in diabetic mice [319, 320]. Furthermore, treatments such as blocking voltage-gated potassium channels [321], various inhibitors [189, 322], carnosine [164], vitamins [323, 324], fibroblast growth factor 21 [325], erythropoietin [219, 326, 327], melatonin [328, 329], and other therapeutic approaches may also play important roles in improving DACD through improving hippocampal and synaptic damage, or attenuating oxidative stress and neuroinflammation.

Although these compounds exhibit promising effects in preclinical studies, their precise mechanisms remain under investigation, and further validation through clinical trials is required before translation into human applications.