All brain cell types express insulin receptors, with the highest densities localized to the olfactory bulb, hypothalamus, hippocampus, cerebral cortex, striatum, and cerebellum (11, 12). Emerging evidence indicates that insulin influences cerebral bioenergetics, enhances synaptic viability and dendritic spine formation, increases the turnover of neurotransmitters, and facilitates clearance of amyloid β peptide while modulating tau phosphorylation (13, 14). T2DM, the predominant form of diabetes mellitus, is generally characterized by chronic hyperglycemia, hyperinsulinemia, dyslipidemia, as well as lipotoxicity, which result in progressive deterioration of insulin secretion and insulin action (15–18). Insulin resistance (IR) is defined as the lack or decreased response of the target tissues to insulin (19, 20). Notably, evidence has shown that peripheral IR results in loss of brain function, which in turn is strongly associated with brain degeneration, cognitive dysfunction, depression, and AD (21–24). Similarly, brain insulin resistance (bIR) can be defined as the failure of brain cells to respond to insulin as they normally would, resulting in impairments in synaptic, metabolic, and immune response functions (25). Individuals with relatively diminished brain insulin sensitivity have a particularly high risk for an AD-like brain pattern (26). Indeed, preclinical and clinical findings support the hypothesis that bIR underlies the basic neuropathological mechanism of cognitive impairment in the aging-related, T2DM-associated, and neurodegenerative context (26). Through these multiple pathways, we hypothesize that insulin resistance could contribute to neurodegeneration, which in turn mediates and promotes the development of AD, vascular cognitive impairment, and other dementias. Notably, metabolites of the intestinal flora, such as bile acids (BAs), short-chain fatty acids (SCFAs) and amino acids (AAs) may influence to some extent the decreased insulin sensitivity associated with T2DM dysfunction and regulate metabolic as well as immune homeostasis (27).

The gut microbiome is known for playing a major role in human health as well as being increasingly recognized as being involved in the pathogenesis of metabolic diseases. Accumulating preclinical and clinical data over the past years has shown that alterations in the gut microbiota affect many organs involved in T2DM and the clinical onset of hyperglycemia (28, 29). Multi-omics (OMICs) studies have shown that single-dose streptozotocin (STZ) -induced hyperglycemia (HG) is sufficient to induce and exacerbate intestinal dysbiosis through modulation of the cecum metabolite pool by analyzing the taxonomic composition, transcriptional activity, and small molecule libraries of the cecum (30). A two-stage case-control metagenome-wide association study (MGWAS) based on deep next-generation shotgun sequencing showed that patients with T2DM were characterized by gut microbial dysbiosis, a decrease in the abundance of some universal butyrate-producing bacteria and an increase in various opportunistic pathogens, as well as an enrichment of other microbial functions conferring sulfate reduction and oxidative stress resistance (31). Intestinal dysbiosis promotes insulin resistance and inflammation, exacerbating diabetes; diabetes further worsens the intestinal dysbiosis, creating a mutually reinforcing mechanism. GM dysbiosis synergistically results in (i) a general increase in pro-inflammatory bacteria and a decrease in anti-inflammatory bacteria; (ii) impair intestinal tight junction integrity by increasing production of inflammatory metabolites and intestinal inflammation; and (iii) induced neuroinflammation, accelerated Aβ fibrillogenesis and parenchymal plaque burden (32). Several studies report that a variety of intestinal bacteria responsible for the production of lipopolysaccharides (LPS), a neurotoxin that disrupts paracellular barriers by cleaving intercellular proteins, such as E-cadherin in epithelial cells, leading to the “leaky gut” phenomenon (33, 34). LPS activates microglia and astrocytes, triggering neuroinflammation that promotes amyloid precursor protein accumulation, Aβ42 fibrillogenesis, plaque formation, and ultimately neuronal loss—a key pathway in neurodegeneration (35–37). Dysbiosis of gut microbiota promotes the harmful intestinal substances enter the systemic circulation through a compromised intestinal barrier, triggering systemic inflammation, which in turn destroys the blood–brain barrier and activates the TLR4/NF-κB signaling pathway in the brain, causing neuroinflammation (38, 39).

An increasing number of studies have indicated that increased oxidative stress is associated with neuronal damage and is a key factor contributing to the onset and progression of DCAD (40–42). ROS are normally produced as by-products of oxygen metabolism, but various factors can elevate its production. Among them, diabetes is a major cause of increased ROS generation by auto-oxidation of glucose, protein glycation, and through the polyol pathway (43). Oxidative stress has been shown to be a major causal factor compromising neuronal loss and synaptic disruption by impairing brain mitochondrial homeostasis, as seen in diabetic mice models, ultimately having deleterious effects on cognitive performance (44). In addition, a wide range of clinical studies have noted that oxidative stress is contribute significantly to the pathogenesis and progression of cognitive dysfunction (23, 45–47). On the other hand, oxidative stress has also been shown to be a facilitator of neuroinflammation, which is another primary contributor to the progression of cognitive decline in T2DM (48).

During T2DM, influenced by HG, microglial activation can exacerbate cytotoxicity and neuronal damage. Previous studies have revealed that in T2DM rat models, microglia activation in the brain is evident, resulting in overexpression of proinflammatory cytokines in the brains of these model rats, along with a marked decline in their learning and memory abilities (49). In contrast, treatment of T2DM model mice with drugs significantly suppressed the over-activation of microglia in the CNS, accompanied by a notable downregulation of pro-inflammatory cytokines and a significant amelioration of cognitive impairment symptom in the mice (50, 51). In the CNS, the functions of microglia are highly dynamic and can adopt different phenotypes based on the microenvironmental signals they receive. This ability to change phenotype, known as polarization, is a key characteristic of microglia, enabling them to adapt to various physiological and pathological conditions (52). However, they are generally classified into two main phenotypes: M1 and M2 (53). M1 phenotype microglia are typically considered pro-inflammatory, playing a key role in immune responses and inflammatory reactions. On the other hand, M2 phenotype microglia are mainly involved in neuroprotection and anti-inflammatory responses. In db/db diabetic mouse model, microglia are polarized into a pro-inflammatory M1 phenotype, along with low levels of a neuroprotective M2 phenotype, and significant cognitive impairment was observed through the Morris water maze test (54). The link between regulatory T (Treg) function and microglia polarization is well established in the brain (55), dipeptidyl peptidase-4 (DPP4)-mediated impairment of Tregs function polarize microglia toward a pro-inflammatory phenotype and subsequently lead to neuroinflammation and cognitive dysfunction in T2DM patients (54). In the T2DM mouse model, the pharmacological intervention inhibited the overactivated microglia and reversed the polarization of microglial phenotypes under T2DM conditions, shifting them from the proinflammatory M1 type to the anti-inflammatory M2 type, with concomitant improvement of cognitive impairment in T2DM mice (56, 57). IF, gut flora dysbiosis, neuroinflammation, and oxidative stress form a self-reinforcing network that underlies DACD.