PDF
ePub
Citation
Print
Introduction
Diabetes mellitus (DM) has risen as an important global health concern with a continued worldwide increase in its prevalence [1]. Type 2 diabetes mellitus (T2DM) in particular is expected to become more common and widespread in many countries, with the proliferation of obesity and the aging of the population [2]. Meanwhile, achieving a normal life expectancy for T2DM patients has become more attainable through effective modulation of the life-threatening cardiovascular and renal complications associated with the disease [1,3]. Accordingly, an improved understanding of debilitating diabetic complications is now of even greater importance for the long-term management of T2DM. In particular, T2DM-induced neurocognitive changes including cognitive decline and dementia may significantly affect the quality of life of patients and their caregivers and pose challenges to clinicians and researchers. In the present review, we will discuss the characteristic cognitive impairment observed in T2DM and present a summary of the outcomes from recent neuroimaging studies focusing on T2DM-related brain deficits in humans.
FEATURES OF NEUROCOGNITIVE CHANGES IN T2DM
DM in general is known to increase the risk of cognitive dysfunction and dementia [4,5,6]. A recent systematic review of longitudinal population-based studies revealed that the incidence rates of dementia were higher in diabetic populations than in nondiabetic populations [6]. A number of DM-related factors such as macrovascular and microvascular diseases, glucose toxicity, and hyperinsulinemia have been suggested to be involved in the emergence of cognitive impairment in DM [6], but the complexity of multifactorial influences makes it difficult to precisely delineate the exact determinants of the phenomenon.
The incidence rate of dementia in T2DM is suggested to be 1.5 to 2.5 times higher than that in the general population [2]. A meta-analysis based on nine population-based longitudinal studies in T2DM revealed a meaningful association between T2DM and progressive cognitive decline, including development of dementia [7]. Interestingly, the same cognitive domains that are impaired in T2DM are also observed to be affected in prediabetic stages [8], despite a smaller effect size [9]. These prediabetic cognitive deficits raise the possibility that glucose intolerance, which starts to arise in prediabetic or early diabetic stages, may play a role in the decline of cognitive function.
FACTORS INFLUENCING NEUROCOGNITIVE IMPAIRMENT IN T2DM
Complex factors including comorbid vascular diseases and the level of glycemic control are believed to influence the manifestation of cognitive impairment related to T2DM [2,10]. Here, we will briefly review some of the potential risk factors that have been repeatedly identified in the existing literature: hyperglycemia, hypoglycemia, and vascular diseases [2].
Hyperglycemia
Acute hyperglycemia induced by a glucose clamp technique in T2DM was reported to impede cognitive function in T2DM [11]. Also, in older women, an increased level of glycated hemoglobin (HbA1c), an indicator of average blood glucose levels over the past few months, was associated with the risk of developing cognitive impairment in later years [12]. These outcomes support the hypothesis that the hyperglycemic state is a major player in DM-related cognitive decline.
Determining the extent of the effect of hyperglycemia on cognitive decline in T2DM is complicated since many other factors associated with the diabetic condition also appear to be involved. For instance, according to a recent study with an elderly cohort, the effect of elevated fasting plasma glucose (FPG) levels on cognitive decline was not greater than that of hypertriglyceridemia or low high density lipoprotein levels [13]. Another study reported that the degree of hyperglycemia measured by the HbA1c level exhibited less predictive power as compared to the diabetic status itself [14]. It should be noted that using HbA1c or FPG as a marker for hyperglycemia in T2DM may have inherent limitations, as these measures do not fully reflect the short-term temporal variations in the hyperglycemic condition. Indeed, an investigation of the relationship between the T2DM-related cognitive decline and daily fluctuations in glucose levels found an independent correlation between glycemic instability and cognitive decline [15], indicating the potential importance of day-to-day glycemic control in DM-related cognitive impairment.
Hypoglycemia
Acute hypoglycemia, which may occur as a side effect of insulin administration, is known to disrupt brain functional efficiency [16]. In particular, an episode of severe hypoglycemia is known to be associated with neuronal damage in the brain areas crucial for learning and memory, including the hippocampus and cerebral cortices [17]. Hypoglycemia is thus likely to be influential in DM-related cognitive decline and related brain deficits. The excitotoxic effect of increased neuroactive amino acids has been suggested as a potential mechanism of neuronal damage induced by hypoglycemia [18]. The activation of neuronal nicotinamide adenine dinucleotide phosphate oxidase during glucose reperfusion, rather than hypoglycemia itself, has also been proposed as the direct cause of the neuronal damage [19].
Although hypoglycemia is more frequently observed in patients with T1DM, it is also common in patients with T2DM who are treated with insulin or orally-administered glucose-lowering drugs [20]. Thus far, investigations into the potential causal relationship between hypoglycemia and cognitive impairment in T2DM have resulted in mixed findings. In a study with an elderly population with T2DM, severe hypoglycemia was found to increase the risk of dementia [21]. Yet, there is evidence that cognitive dysfunction may actually be a cause of hypoglycemia rather than the result of it, as cognitive dysfunction may hinder adequate glycemic modulation through regular drug administration [22,23,24].
Vascular diseases
T2DM shows high comorbidity with many macrovascular and microvascular diseases [25]. A substantial amount of research suggests that vascular complications are a risk factor for cognitive decline in patients with DM [26,27]. DM has been suggested as an important risk factor in the pathogenesis of vascular dementia [28]. In patients with DM, brain alterations including white mater hyperintensity (WMH) and subcortical atrophy may arise as a result of the degeneration of cerebral small vessels [29]. These alterations are associated with the emergence of cognitive dysfunction [29]. Identification of a shared mechanism linking DM, cerebrovascular diseases, and cognitive decline would be an important research goal for the early detection of and intervention in DM-related neurocognitive impairment.
NEURAL CORRELATES OF NEUROCOGNITIVE IMPAIRMENT IN T2DM: EVIDENCE FROM NEUROIMAGING STUDIES
Both acute and chronic brain alterations in DM are presumed to be associated with the generation of cognitive impairment in DM [30]. Even though the exact pathophysiology linking DM and cognitive dysfunction remains unclear, recent brain imaging studies aided by state-of-the-art neuroscientific techniques have begun to shed light on related issues. In the following section, we will review the neuroimaging findings on the structural, functional, and metabolic changes that occur in the T2DM brain, to provide a more integrative understanding of the neurocognitive complications in T2DM.
Brain structural changes
DM is known to induce not only macroscopic but also microscopic changes in the brain. Several neuroimaging studies indicate that the brain structural alterations frequently observed in T2DM include global and regional atrophy in cortical and subcortical regions [31,32] and WMH [33].
Global brain atrophy and ventricular enlargements
Reductions in the total brain volume or cortical/subcortical brain volume in T2DM have been indicated by several neuroimaging studies [34]. Modest cerebral atrophy in the T2DM population has been consistently observed [32,35]. Furthermore, a significant association has been found between cerebral atrophy in DM and cognitive impairment, even after the effect of comorbid vascular diseases was adjusted [33]. Ventricular enlargement, an important index of cerebral atrophy and potential surrogate marker for dementia [36], has also been implicated in T2DM-related brain changes. A study with a middle-age population revealed a significant association between DM and greater ventricular size [37]. Furthermore, according to recent case-controlled studies, patients with T2DM showed greater increments in lateral ventricular volumes as well as shape alterations compared to control subjects [38,39]. With regard to the longitudinal trajectory of the changes, the rate of increase in lateral ventricular volumes in elderly patients with DM was faster than the rate of aging-related changes [40,41].
A number of factors have been suggested to be associated with brain atrophy in T2DM. Factors including retinopathy, extent of brain infarction, HbA1c level, and disease duration were associated with cortical atrophy [27,33]. In addition, several cross-sectional studies have demonstrated that indices such as comorbid hypertension, level of glycemic control, and history of hypoglycemic events are factors that may potentially affect the progression of cerebral atrophy in DM [35,40,42,43,44,45]. Results from animal models bring up the possibility that factors like glucose toxicity, hyperinsulinemia, and vascular damage have accelerating effects on the brain atrophy process [46,47]. Meanwhile, the use of cholesterol-lowering statin drugs is associated with less brain atrophy [27,44]. In sum, these findings propose the involvement of multiple factors in the development of DM-related brain structural alterations, although the inconsistency in research outcomes has not yet been entirely resolved.
Notably, along with the T2DM population, the prediabetic population also tends to exhibit cognitive dysfunction [48] and reduction in total brain volume [49]. As conditions like hyperinsulinemia and impaired glucose tolerance start to emerge during the prediabetic stage, cognitive dysfunction as well as brain alterations may occur even in the early stages of T2DM.
Regional brain atrophy
Animal studies have presented the associations of the diabetic state with the neuronal changes in specific brain regions such as the hippocampus and prefrontal cortex, as well as with cognitive deficits [50,51]. By implementing a variety of neuroimaging analysis methods including voxel-based morphometry and cortical thickness analysis, recent human studies also have yielded evidence of associations between DM and regional brain deficits [34]. The most consistently reported regional atrophy in the T2DM population is found in the medial temporal lobe, especially in the hippocampus [31,40,52]. For instance, cortical thickness and subcortical volumetric analysis of the T2DM group revealed prominent deficits in the hippocampus and middle temporal gyrus [40]. Another study demonstrated that T2DM was related to a higher risk of hippocampal atrophy [43]. Additionally, a volumetric reduction restricted to the hippocampus was observed along with a decline in memory function, starting from the early stages of T2DM [52]. Atrophic alterations in the medial temporal lobe, especially in the hippocampus, have also been reported as a robust neuroimaging finding in Alzheimer dementia (AD) [36] and thus suggest a possible mechanism connecting DM and dementia.
In addition to the brain volumetric reductions in the medial temporal lobe, deficits in the prefrontal regions have also been noted in a substantial body of research [53,54,55,56]. Specifically, patients with T2DM showed gray matter deficits in the prefrontal areas including the anterior cingulate and orbitofrontal regions, and the findings were robust even after adjusting for global brain atrophy [53,55]. The existing evidence suggests that prefrontal regions may act as neural correlates associated with depression and cognitive dysfunction [57,58]. Accordingly, prefrontal structural alterations in the diabetic brain may play a crucial role in central nervous complications, particularly in the depressive symptoms and cognitive impairment observed in T2DM.
White matter hyperintensity
WMH, a commonly observed phenomenon in the elderly brain, is known to increase with the normal aging process [59,60]. The findings on WMH in T2DM are rather inconsistent compared to those on cerebral atrophy [34]. Although some studies have identified DM as a risk factor accelerating the generation of WMH [61,62], in a longitudinal study no significant difference in the WMH progression rate was found between T2DM and control groups [40]. Furthermore, in a meta-analysis that inspected studies primarily conducted with a visual rating technique, no consistent association was found between T2DM and WMH [34]. While some researchers have presented additional negative findings [31,32,59], others have reported a significant association between WMH severity and DM [33,61]. In a study using volumetry analysis rather than visual rating, a prominent increase in the volume of white matter lesions was observed [38]. Some recent neuroimaging studies present region-specific effects, since subcortical white matter lesions, but not periventricular white matter regions, show a significant association with T2DM [33,63]. In these studies, WMH severity was associated with the magnitude of the cognitive impairment, suggesting a role for WMH in the cognitive changes associated with T2DM [35,63]. The inconsistency in the research outcomes on WMH in T2DM may be explained by the different methodologies used for WMH rating and the large interindividual variability in WMH [34]. Accordingly, further large-scale longitudinal studies with appropriate quantitative methods to evaluate WMH are warranted.
Although WMH is considered to have complex interactions with vascular factors, the precise mechanisms involved are yet to be elucidated. A number of studies in T2DM proposed indices such as macrovascular factors, illness duration, HbA1c levels, hyperinsulinemia, and hypertension to be related to WMH in T2DM [6,27,63,64]. The presence of WMH in the deep white matter is correlated with neurodegenerative changes as well as with vascular abnormalities [65,66,67].
Brain functional changes
Changes in brain functional connectivity can be induced by structural changes, but they may also emerge before clinically observable changes in cognitive function occur, preceding apparent structural alterations [68,69]. There have been reports suggesting a resemblance between alterations of brain functional connectivity in T2DM and those in mild cognitive impairment in AD [70,71,72]. Furthermore, decreased glucose metabolism in the frontal, parieto-temporal, and cingulate regions, a characteristic feature in AD, is also observed in the prediabetic state [73]. In patients with T2DM, the functional connectivity between the hippocampus and other brain regions is reduced [72,74], and the decline in cognitive performance in T2DM is associated with a reduction in functional connectivity [72].
In a resting-state functional magnetic resonance imaging study, patients with DM and microvascular complications showed a decrease in brain functional connectivity, whereas those without such complications and healthy controls did not [75]. Diabetic retinopathy, an example of microvascular complications, has been identified as an independent risk factor for cognitive decline in DM [73]. Given these findings, microvascular complications may be potential players in the development of brain functional abnormalities and the associated cognitive decline observed in T2DM.
Metabolic brain changes: results from magnetic resonance spectroscopy
Magnetic resonance spectroscopy (MRS) enables noninvasive inspection of metabolites in the brain. MRS is also valuable for the detection of brain damage or abnormalities that would appear normal on other modalities [76], and thus may prove particularly useful in investigating the neurocognitive changes in DM.
Both human and animal MRS studies of DM suggest a decrease in the N-acetylaspartate (NAA)/creatine (Cr) ratio in a relatively consistent manner, while there are mixed results for the choline (Cho)/Cr ratio [77]. In a study using a rat model of DM, reductions in the NAA/Cr ratio and the NAA/Cho ratio were observed in diabetic rats, whereas the Cr/Cho rate was not significantly changed [78]. In a few human MRS studies, the NAA/Cr ratio was also found to be reduced in patients with DM [79,80,81]. A study reported that the Cho/Cr ratio was decreased in the frontal and parietal lobes of patients with T2DM [81], while another study observed an increase in the occipital lobe [82]. In the most recent MRS study on T2DM, a decrease in the NAA/Cr ratio and an increase in the Cho/Cr ratio in the lenticular nucleus were noted [77]. The literature suggests an association between NAA and neuronal damage, as well as one between Cho and glial proliferation [83]. These abnormalities in NAA and Cho levels in T2DM are in accordance with the hypothesis that DM may incur aberrant changes in the brain metabolites and associated neuronal damage.
Glutamate is another key substance in the investigation of DM using MRS [84]. Glutamate is a highly important neurotransmitter involved in many crucial brain activities such as cognition [85,86] and can be particularly useful in the assessment of glucose metabolism within the cerebrum [87]. A number of MRS studies in T1DM illustrate associations between DM and glutamatergic alterations in the brain [56,88], yet the direction of change in glutamate levels appears to differ depending on the region of interest. While patients with T1DM show an elevated level of glutamate in the prefrontal region [56], the glutamate levels in T1DM have been reported to be lower in the occipital and parieto-occipital regions [88]. In depressed patients with T2DM, glutamine and glutamate concentrations were significantly reduced in the subcortical regions relative to healthy and diabetic control subjects [89]. A study based on a rat model found impaired glutamate-glutamine cycling in T2DM [90], while in a rat model of diabetic retinopathy, hyperglycemia was found to induce a dysfunction in glutamate transporters [91].
Since glutamate in the synaptic space would introduce a significant amount of noise to glutamatergic signal transduction, it is essential to uptake glutamate that has been released to the extracellular space [92,93]. Insulin not only enhances the glucose reuptake processes but also normalizes the altered glutamate uptaking properties in astrocytes [94]. Accordingly, the altered ability of glutamate reuptake in T2DM brain astrocytes indicates a potential impairment of brain signal transmission in DM [95]. As a considerable portion of the energy consumed in the brain is used in glutamatergic signaling [96], we propose that the diabetic brain, which shows lowered glucose metabolism, may also have difficulties in glutamate signaling. In any case, the investigation of DM-induced metabolite alterations in the human brain is still in its early stages, and further research should follow.
CONCLUSIONS
As T2DM has arisen as a global health problem affecting numerous individuals [1,3], the neurocognitive complications of T2DM including cognitive decline and dementia bring up important and challenging research questions in both clinical and research fields. Several conditions including hyperglycemia, hypoglycemia, and vascular diseases are suggested to be involved in the development of cognitive impairment in T2DM, although the precise mechanism is yet unclear due to the complexity of multifactorial influences. Recently, neuroimaging methods have been implemented in the investigation of the neural correlates underlying the association between T2DM and cognitive dysfunction. Studies have revealed changes in the T2DM brain at structural, functional, and metabolite levels and their potential role in cognitive changes. With further methodological developments, future neuroimaging studies in T2DM are expected to elucidate the underlying neural mechanisms of DM-related cognitive deficits.
XML Download