| Background and purposeType 2 diabetes mellitus(T2DM)is a chronic metabolic disease characterized by hyperglycemia.According to the International Diabetes Federation(IDF),the number of people with diabetes in the world in 2017 was about 425 million,90% of which was T2 DM.China is the country with the largest number of T2 DM patients in the world,and the prevalence rate is increasing year by year.Studies have shown that T2 DM patients have reduced functions in multiple cognitive areas,such as executive function,motor function,memory function and information processing function,and so on.The risk of dementia increases,which causes huge economic burden to society and families.Revealing the neurophysiological mechanism of cognitive damage in T2 DM will contribute to the diagnosis,prevention and treatment of the disease and the reduction in incidence of dementia.However,the mechanism of T2 DM cognitive impairment has not been elucidated.Resting-state functional magnetic resonance imaging(rs-fMRI),as a non-invasive imaging technique,has been widely used in the study of brain dysfunction in patients with T2 DM.However,the current research on subgroup studies of T2 DM with and without cognitive impairment is rare,and most researches focused on the comparative study of T2 DM and healthy subjects,lacks of interpretation of cognitive changes in brain function,which seriously affects the further understanding of its neural mechanisms.Cerebral iron deposition can induce cell damage through oxidative stress pathway.It has been con firmed that brain iron deposition exists in neurodegenerative diseases such as Parkinson’s disease,Alzheimer’s disease and multiple sclerosis,and is related to cognitive performance.However,there are few reports on the presence or absence of brain iron deposition in T2 DM.Quantitative magnetic susceptibility mapping(QSM)has obvious advantages over magnetic sensitive weighted imaging,T2* and phase images in measuring brain iron content.The first two parts of the study were to use two methods in rs-fMRI: whole brain voxel-based and brain network-based methods to analyze brain function changes of T2 DM patients with and without mild cognitive impairment(MCI)to explore the characteristics and rules of brain function alteration.The last part of this study intended to use the QSM method to quantitatively measure the iron content of the gray matter nucleus in the two groups of patients,to investigate the alteration of brain iron content in T2 DM patients with MCI,and the correlation with cognitive performance.Materials and methods1.T2 DM was diagnosed with the criterions published by WHO,in 1999.According to neuropsychologic-scale rating and current literature,we divided patients into two groups: T2 DM with MCI and T2 DM without MCI,each group has 27 patients.Twenty-seven age,gender and education background-matched normal controls were also included.Demographic and blood biochemical data of all three groups were collected.The resting-state fMRI data were processed using Data Processing Assistant for Resting-State fMRI(DPARSF,v2.2),then we used group ICA of fMRI Toolbox(GIFT,v2.0a)for ICA(independent component analysis)to identify different brain network components.The correlation analysis among networks was conducted,we got the correlation value of the brain network that had differences in T2 DM with MCI group,and then correlated it with blood biochemical data and neuropsychological data.2.Twenty-seven T2 DM patients without mild cognitive impairment(MCI),27 T2 DM patients with MCI and 27 healthy controls were enrolled in the study.Data acquisition was consistent with the first part.Resting-State fMRI data preprocessing and amplitude of low-frequency fluctuation(ALFF)values calculation were conducted with DPARSF v2.2.Variance analysis was used to compare the differences of ALFF among three groups.The potential correlations of abnormal region in T2 DM patients with MCI among the imaging indices,neuropsychological data and biochemical results were also explored.3.We divided patients into 3 groups: T2 DM with MCI,T2 DM without MCI and normal controls,each group has 30 patients.Susceptibility weighted imaging(SWI)data was collected by using a Siemens 3.0T magnetic resonance imaging system.The post-processing data analysis to obtain QSM images was conducted by Susceptibility Mapping and Phase Artifacts Removal Toolbox(SMART)software.The values of bilateral parietal cortex,frontal white matter,caudate nucleus(CN),putamen(PU),globus pallidus,thalamus,red nucleus,substantia nigra(SN),hippocampus(HP)and the susceptibility values of the dentate nucleus were measured by Signal Processing in Nuclear(SPIN)magnetic resonance software and then partial correlations between these data and the neuropsychological scale score were analyzed.Result:1.There were no significant differences in age,gender,education level,Body Mass Index,systolic and diastolic blood pressure,total cholesterol,LDL cholesterol,Auditory verbal learning test(AVLT)delayed recognition(5 min and 20min),Digital Span Test(DST)forwards,Verbal Fluency Test(VFT),Trail Making Test-B and Mini-mental state examination(MMSE)(all P > 0.05).Twenty components were identified using independent component analysis.Total 8 components were identified after exclusion of noise,including anterior-default mode network(aDMN),auditory network(AN),left frontal-parietal lobe network(LFPN),right frontal-parietal lobe network(RFPN),post-default mode network(pDMN),sensorimotor network(SMN),salience network(SN),visual network(VN).There were significant differences in coherence of AN-VN(frequency band 53-62,frequency 0.102-0.119Hz)and AN-aDMN(frequency band 42-49,frequency 0.08-0.094Hz)among three groups.Coherence of AN-VN was significantly larger in T2 DM patients with MCI than in those two groups(P < 0.05).There were no significant differences between the latter groups in coherence of AN-VN(P > 0.05).Coherence of AN-aDMN was significantly larger in T2 DM patients with MCI and without MCI than in HC(P < 0.05).Coherence of AN-VN were significantly associated with Complex Figure Test(CFT)and Montreal cognitive assessment(MoCA)values,coherence of AN-aDMN were significantly associated with Complex Figure Test(CFT)values in T2 DM patients with MCI(all P < 0.05).2.There were significant differences of ALFF values in bilateral inferior temporal gyrus(BA20),left precuneus(BA7),right precentral gyrus(BA4),and right posterior cingulate gyrus(BA30)among three groups.In right posterior cingulate gyrus,ALFF value was highest in T2 DM patients with MCI,and lowest in HC(P < 0.05).In bilateral inferior temporal gyrus,ALFF value was significantly higher in T2 DM patients with MCI than in T2 DM patients without MCI and HC.In left precuneus,ALFF values of T2 DM patients with MCI were similar with those of HC,but both were lower than those of T2 DM patients without MCI.In right precentral gyrus,ALFF value was significantly lower in T2 DM patients with MCI than in T2 DM patients without MCI and HC.In T2 DM patients with MCI group,AVLT delayed recognition(5 min)significantly correlated with ALFF value of right posterior cingulate gyrus,right inferior temporal gyrus,and left precuneus(P < 0.05).3.Compared with HC,T2 DM patients without MCI showed significantly elevated susceptibility values in the left HP,whereas T2 DM patients with MCI exhibited significantly elevated susceptibility values in the bilateral CN,HP,left PU and right SN.Compared with T2 DM patients without MCI,T2 DM patients with MCI showed significantly elevated susceptibility values in the right CN,SN and left PU.In T2 DM patients with MCI group,the right CN susceptibility values were closely correlated with the AVLT-delayed recall(5 min)scores(r =-0.450,P = 0.019),the AVLT-delayed recall(20 min)scores(r =-0.474,P = 0.013),the CFT(Complex Figure Test)-immediate recall(20 min)scores(r =-0.396,P = 0.041),and the CFT-delayed recall(20 min)scores(r =-0.418,P = 0.030);the left PU susceptibility values were closely correlated with the Verbal Fluency Test scores(r =-0.469,P = 0.014);and the right SN susceptibility values were closely correlated with the CFT-delayed recall(20 min)scores(r =-0.458,P = 0.0016)and the Digit Symbol Coding Test scores(r =-0.533,P = 0.004).Conclusion1.There were abnormal brain functional connectivity in T2 DM patients with and without MCI.The decrease of AN-VN coherence value can reflect the degree of cognitive decline of T2 DM,suggesting the neural activity disorders between visual cortex and auditory cortex in T2 DM patients with MCI.2.There were abnormal spontaneous neuronal activity in multiple brain regions in T2 DM patients with and without MCI.Short-term memory impairment in T2 DM patients with MCI may be related to spontaneous nerve activity abnormalities in the right posterior cingulate gyrus,right inferior temporal gyrus,and left precuneus.3.Iron deposits were found in multiple gray matter nucleus in both T2 DM patients with and without MCI,and correlated with cognitive performance.QSM may be a very useful imaging technology for detecting and evaluating the cognitive impairment of T2 DM. |