A Multi-modality Magnetic Resonance Imaging Study Of Cognitive Impairment In Type 2 Diabetes Mellitus

Part I.Gray Matter Alterations Study in Type 2 Diabetes Mellitus Patients with Early-stage Mild Cognitive ImpairmentBackground and Purpose:Diabetes mellitus(DM)is a series of metabolic disorders syndrome combined with insulin resistance(IR)or impaired islet function,resulting from various pathogenic factors such as genetic factors,immune system dysfunction and neuropsychiatric factors.It could lead to chronic metabolic diseases with high blood sugar level as the main characteristic.In recent studies,it has been released that the number of DM patients in the world has been over400 million,and in China the number of diabetes people has been more than 100 million.In China,DM has aroused leading concerns.Meanwhile,the prevailing one in China is type 2diabetes.In recent years,much has been made to compare T2 DM patients’ cognitive impairment with normal people of the same age and the conclusion has been reached that T2 DM patients have much more cognitive impairment.When the cognitive impairment is mild,it is called mild cognitive impairment(MCI),and when the cognitive impairment is severe,it can lead to dementia.A number of studies have suggested that T2 DM with MCI could be relevant to both the degenerative changes caused by normal aging process and the chronic diseases related structural changes in the brain.Due to the extreme complexity of human brain,it is difficult and challenging to study the structure and function of human brain.Therefore,if we explore and discover the T2 DM with MCI patients’ pathogenic factors and early pathogenesis,it can offer the chances for early-stage clinical diagnosis and timely interference,so as to decrease the happening rate of T2DM-MCI and dementia.Due to the lack of non-invasive research methods in human brain,former researches on T2 DM with mild cognitive impairment are not comprehensive.MRI is a non-harmful means to investigate the human brain without invasive methods.Former studies on cognitive impairment were limited to the scoring and judgment of the cognitive function scale,which was relatively limited and not objective.However,the relationship and mechanism of the structural alterations and cognitive impairment in T2 DM patients are still unclear.Hence,it is crucial to investigate the brain structural alterations in T2 DM.Based on Free Surfer analysis method,this study analyzed and compared the changes of cortical and subcortical gray matter volume in each brain region in patients with early-onset mild cognitive impairment,type 2diabetes mellitus,and patients with type 2 diabetes mellitus with early-onset mild cognitive impairment.Correlation analysis was conducted with various biomarkers,cognitive function and depression scales such as Montreal Cognitive Assessment Scale(Mo CA),Simplified Mental State Examination Scale(MMSE),Daily Living Ability Scale(ADL),Trail Making Test(TMT)and Hamilton Depression Scale(HAMD).The relationship was analyzed between patients with early-onset mild cognitive impairment,patients with type 2 diabetes,and patients with type 2 diabetes and early-onset mild cognitive impairment.With in-depth exploration of the cortical and subcortical gray matter volume changes and pathogenesis of various brain regions of the patients,this study could offer more scientific and effective support for accurate clinical evaluation,early diagnosis,timely intervention and efficient treatment.And in-depth exploration of the causes of type 2 diabetes with early pathophysiological imaging markers of onset mild cognitive impairment provides new research ideas and scientific direction for exploring and understanding the molecular biological mechanism of type 2 diabetes mellitus with early-onset mild cognitive impairment.Materials and methods:1.General date collectionGeneral data were collected from 28 T2 DM patients(T2DM group),28 MCI patients(MCI group),28 T2DM-MCI patients(T2DM with MCI group)and 28 healthy controls group matched in age,gender and education years.All subjects’ blood biochemical data and scores of Mo CA,MMSE,ADL,TMT and HAMD were collected.2.MRI data collection and analysisTo keep out of the intracranial organic injuries,3.0T MRI was used to collect FLAIR,T1 WI,and T2 WI images,and high-resolution T1 structural image data were collected by using MRI.Then,Free Surfer software was used to analyze and process the high-resolution T1 structural image data,and cortical volume data of each brain region were obtained.At the same time,the volume of subcortical gray matter was extracted by reconstruction segmentation calculation and analysis.The volume of subcortical areas included thalamus,caudate,putamen,pallidum,hippocampus,amygdala,nucleus accumbens.Apart from that,we further segmented the hippocampus so as to observe the role of structural changes in the hippocampal subregions.The segmentation regions of hippocampus included the hippocampus tail(HT),cornu ammonis(CA1),hippocampal fissure(HF),CA3,CA4,subiculum,molecular layer(ML),parasubiculum,fimbria,presubiculum,granule cell layer of dentate gyrus(DG),and hippocampus-amygdala-transition-area(HATA).We adopted the left posterior cingulate cortex(PCC),a key hub of the default mode network(DMN),to calculate its structural connectivity with 10 other key subsegments of DMN,containing the bilateral temporal pole,precuneus,medial prefrontal cortex,lateral temporal cortex,and hippocampus,based on gray matter volume.3.Statistical Analysis MethodsStatistical analysis was performed by adopting SPSS22.0 software.Demographic characteristics,standard clinical laboratory testing measurements and neuropsychological scale scores were compared among the four groups using χ2 test,independent two-sample t-test,one-way analysis of variance(ANOVA)or Kruskal-Wallis one-way ANOVA.Among the four groups,differences in cortical volume were assessed using a general linear model(GLM).Monte Carlo simulation method was used to conduct a whole-brain statistical threshold correction,and p-value <0.05 after clusterwise-correction was considered having statistical significance.To evaluate the differences in subcortical areas and hippocampal subfields in volumes among the four groups,one-way ANOVA was performed,followed by least-squares difference(LSD)post hoc test or Kruskal-Wallis one-way ANOVA,followed by all pairwise corrections.Based on gray matter volume,Pearson correlation analysis was adopted to evaluate the structural covariance.In order to evaluate the differences of correlation coefficient between groups,Snedecor’s method was applied to transform r value into z value.In appropriate circumstances,the Bonferroni correction was used to correct multiple comparisons involving multiple brain regions,and a Bonferroni-correction p-value<0.05 was considered statistical significance.Using partial Pearson correlation analysis or partial Spearman correlation analysis,age,sex and education level as covariables,the correlations amid neuropsychological scale scores,blood biochemical index and gray matter volume were analyzed.In order to evaluate the potential trends,we did not conduct multiple comparison correction,and the p level was set at <0.01 to be statistical significance.ResultsIn our study,overlapping and distinct cortical and subcortical gray matter atrophy was observed in patients with MCI,T2 DM,and T2DM-MCI,with T2DM-MCI patients having lower volumes in several regions than MCI or T2 DM patients.Volume loss in subcortical gray matter areas including the thalamus,hippocampus,and putamen was associated with cognitive impairment in MCI and T2DM-MCI patients,whereas volume loss in cortical regions was not associated with cognitive impairment.No correlation was found between blood biochemical measurements and volume loss in this study.In addition,DMN structural connectivity was disrupted in both MCI and T2DM-MCI patients.Conclusions:These findings add to evidence that T2 DM may exacerbate the atrophy in specific gray matter regions,which may be primarily related to MCI.Gray matter volume impairment associated with T2 DM or MCI was not associated with cardiovascular risk factors.For MCI and T2DM-MCI patients,subcortical volume atrophy may play a more critical role than cortical volume changes in cognitive impairment.Enhanced DMN connectivity in patients with T2DM-MCI may be a compensatory mechanism for chronic neurodegenerative changes.Part II.White Matter Alterations Study in Type 2 Diabetes Mellitus Patients with Mild Cognitive ImpairmentBackground and Purpose:The number of people with diabetes mellitus in the world has exceeded 400 million,with more than 100 million people suffering from diabetes in China.Type 2 diabetes mellitus(T2DM)patients are highly susceptible to developing dementia,especially for those with mild cognitive impairment(MCI),but its underlying cause is still unclear.In this study,the subcortical white matter volume(WMV)of each brain region in T2 DM patients was statistically analyzed based on the structural images which were carried out by using Free Surfer software.Each brain region’s white matter volume’s relationship with the neuropsychological scale scores were tested.These findings will further help to investigate the changes of WMV in each brain region and the pathophysiological mechanism of T2 DM and T2DM-MCI patients,and provide potential imaging biomarkers for the diagnosis and treatment of T2DM-MCI patients.Materials and methods:1.General data collectionThe statistics of 30 T2 DM with normal cognition patients(T2DM-NC)and 30 T2 DM with MCI patients(T2DM-MCI)and 30 normal controls matched in education years,age,and gender.Laboratory blood biochemical data of all subjects were collected,and MMSE,Mo CA,Complex Figure Test(CFT),Digit Symbol Coding Test(DSCT),Digit Span Test(DST),Verbal Fluency Test(VFT),Auditory Verbal Learning Test(AVLT),TMT and HAMD were conducted.2.MRI data acquisition and processingTo keep out of intracranial organic injuries,3.0T MRI scanner was used to obtain the images of FLAIR,T1 WI,and T2 WI,and high-resolution T1 structural image data were collected by MRI.By using Free Surfer software,the brain structural image data were analyzed and processed.Each cerebral hemisphere was separated into thirty-four brain areas in line with Christopher and Desikan-Killian neuroanatomical categorization and the subcortical white matter volume data in sixty-eight brain areas were processed.3.Statistical analysis methodsSPSS software was adopted to conduct statistical analysis.For WMV,comparisons among the three groups were performed by using ANOVA test,with the level of significance setting at p < 0.05,false discovery rate(FDR)corrected.For the T2 DM patients,the correlations between the WMV of each brain region and the neuropsychological scale scores were tested using Pearson correlation analysis.For receiver operating characteristic(ROC)analysis,AUCs were adopted to assess the diagnostic value of each marker.Generally,an AUC greater than 0.9 indicated excellent diagnostic efficacy,between 0.7 and 0.9 indicated good diagnostic efficacy,between 0.5 and 0.7 indicated poor diagnostic efficacy,and no more than 0.5 indicated the lack of a diagnostic value.In this study,Med Calc Statistical Software was adopted to compare differences in AUCs.The statistical significance was set as p < 0.05.ResultsThis study analyzed WMV of the whole brain of T2DM-MCI,T2DM-NC,and HC groups.There were significant differences in WMV in several regions in patient with T2DM-MCI compared with patients with T2DM-NC.Areas of white matter atrophy in patients with T2DM-MCI patients included the left insula,posterior cingulate,and precuneus,right lateral orbitofrontal gyrus,pars orbitalis gyrus,rostral middle frontal gyrus,and temporal pole.We examined the relationship between WMV in these regions and neuropsychological scale data in all T2 DM patients.The results showed that Mo CA scores were significantly correlated with the white matter volume in the left posterior cingulate,precuneus,insula,right rostral middle frontal gyrus,and temporal pole.According to the ROC results,when used as diagnostic criteria alone,the white matter volume of the left posterior cingulate,precuneus,insula,and right rostral middle frontal gyrus had high diagnostic value for MCI detection in T2 DM patients,among which white matter volume value of left precuneus had the highest diagnostic value.ConclusionsT2DM could give rise to the white matter atrophy in several brain regions.Each WMV of left posterior cingulate,precuneus,insula,and right rostral middle frontal gyrus can serve as an independent imaging biomarker for early cognitive impairment in T2 DM patients and is of great importance in its pathophysiological mechanism.Part III.Alterations of Brain Structural Network Connectivity in Type 2 Diabetes Mellitus Patients with Mild Cognitive ImpairmentBackground and Purpose:In the past 40 years,with the aging of the population and the well-off lifestyle,the prevalence of diabetes in China has risen rapidly from 0.67% in 1980 to 10.4% in 2013,and the number of Diabetes Mellitus(DM)patients is still increasing.The number of patients with DM has reached 109 million,and more than 90% of patients with diabetes diabetes in China are type 2 diabetes(T2DM).T2 DM patients are highly susceptible to developing dementia compared with normal people,especially for those with mild cognitive impairment(MCI).This study aims to investigate the changes of white matter structural network in T2 DM patients with MCI,and to assess the correlational relationship between cognitive impairments and alterations of white matter structural network in patients with T2 DM.Materials and methods:1.General data collectionThe clinical statistics of 30 T2 DM with normal cognition patients(T2DM-NC),30 T2 DM with MCI patients(T2DM-MCI)and 30 normal controls 30 normal controls matched in education years,age and gender.Laboratory blood biochemical data and neuropsychological scale data,including MMSE,AVLT,DSCT,DST,TMT,CFT,VFT,Mo CA,HAMD and other scale scoring data,were collected for all subjects.2.MRI data acquisition and processingDiffusion tensor imaging(DTI)data and high-resolution structural image data were collected by a 3.0T MRI scanner.By using the Pipeline for Analyzing Brain Diffusion Images Toolkit(PANDA)based on the FSL(FMRIB Software Library),the DTI data were preprocessed and analyzed.Then,SPM,FACT and GRETNA are used to further process the high-resolution structural image data and DTI data.In this study,the following global network parameters of the brain structural network are calculated: Clustering coefficient(Cp),characteristic path length(Lp),normalized Cp(γ),normalized Lp(λ),small-worldness(σ),global efficiency(Eglob)and local efficiency(Eloc).Betweenness centrality and nodal efficiency were also calculated to demonstrate regional characteristics of the structural network.In addition,we use a network-based statistic(NBS)to analyze the altered structural connections in patients.3.Statistical analysis methodsSPSS software was adopted to carry out data analysis.For neuropsychological and demographic testing,a one-way analysis of variance(ANOVA)test was used to make comparison among the three groups.We performed post-hoc tests with Bonferroni correction after carefully observing statistical differences among groups.An χ2 test was adopted to compare sex variables.The group differences in AUC(area under the curve)values of nodal properties(nodal efficiency and nodal betweenness centrality)and global network metrics(Cp,Lp,Eglob,and Eloc)were examined with one-way ANOVA,while gender and age were adjusted as potential confounders.A significant threshold of each test was set to p < 0.05 and was applied to each test.Moreover,the false discovery rate(FDR)correction was applied for multiple comparison corrections.Furthermore,Spearman’s correlation analysis was used to investigate the correlation between the topological properties and cognitive scale scores in the T2DM-MCI patient groups to explore the relationships between the topological properties of the network measures and clinical cognitive functional outcomes.ResultsThe neuropsychological cognitive function scale scores exhibited obvious differences among the three groups.For patients with T2 DM,the structural network was remarkably disrupted in both regional and global levels.In T2DM-MCI group,global network efficiency was more severely impaired with lower nodal efficiency and less connectivity within multiple regions like the basal ganglia,limbic system and several cortical structures.In addition,the impaired subnetwork in patients with T2DM-MCI was characterized by cortical-limbic fibers,commissural fibers and pathways within the frontal,temporal,and occipital lobes.These changes in global and nodal parameters in T2DM-MCI patients were remarkably correlated with changes in cognitive function.Particularly,working memory impairment and executive dysfunction in patients with T2DM-MCI were correlated with triangular part of the inferior frontal gyrus and nodal efficiency in the right opercular part.ConclusionsTo sum up,this study brought to light the disruption of brain network measures in patients with T2 DM.Compared with T2DM-NC patients,sporadic impairments of structural network were shown in T2DM-MCI patients,mainly located in the limbic system,basal ganglia,and parts of the frontal,temporal,and parietal lobes.Furthermore,these network abnormalities were significantly associated with cognitive function performance in T2DM-MCI patients.The research indicated that the working memory impairment and executive dysfunction in patients with T2DM-MCI connected with the right triangular part and opercular part of the inferior frontal gyrus of the reduced nodal efficiency,which implied that white matter in those areas may serve as potential biomarkers for T2 DM correlated with MCI detection.Our experimental investigation offers an original insight to explore the neuropathological effects in white matter network disruption in T2DM-induced early-onset cognitive impairments.