Junying Zhang,1,2 Yunxia Wang,1,2 Jun Wang,1,2 Xiaoqing Zhou,1,2 Ni Shu,1,2 Yongyan Wang,2,3
and Zhanjun Zhang1,2
White Matter IntegrityDisruptions Associated WithCognitive Impairments in Type 2Diabetic PatientsDiabetes 2014;63:3596–3605 | DOI: 10.2337/db14-0342
Type 2 diabetes mellitus (T2DM) is associated witha twofold increased risk of dementia and can affectmany cognitive abilities, but its underlying cause is stillunclear. In this study, we used a combination of a batteryof neuropsychological tests and diffusion tensor imag-ing (DTI) to explore how T2DM affects white matter(WM) integrity and cognition in 38 T2DM patients and 34age-, sex-, and education-matched normal control sub-jects. A battery of neuropsychological tests was used toassess a wide range of cognitive functions. Tract-basedspatial statistics combined with region of interest–wise(ROI-wise) analysis of mean values of DTI metrics inROIs was used to compare group differences of DTImetrics on WM skeletons to identify severely disruptedWM tracts in T2DM. We found that T2DM patientsshowed 1) various cognitive impairments, including ex-ecutive function, spatial processing, attention, and work-ing memory deficits; 2) widespread WM disruptions,especially in the whole corpus callosum, the left anteriorlimb of the internal capsule (ALIC.L), and external capsule(EC); and 3) a positive correlation between executivefunction and WM integrity in the ALIC.L and the left EC.In conclusion, T2DM patients show various cognitiveimpairments and widespread WM integrity disruptions,which we attribute to demyelination. Moreover, executivedysfunction closely correlates with WM abnormalities.
Type 2 diabetes mellitus (T2DM) in the elderly is a majorpublic health problem. In China, .20.4% of the elderly
population has diabetes, among which T2DM accountsfor ;90% (1). As a risk factor for cognitive decline,T2DM is associated with a twofold increased risk of de-mentia (2) and can affect a wide range of cognitive abil-ities (3,4). However, its underlying cause is still unclear.
A number of studies were carried out to investigate themechanisms of T2DM-induced dementia and attributedthe cognition decline to cerebral microstructure impair-ments, which included gray matter (GM) and whitematter (WM) destruction (5,6). Despite the importanceof GM atrophy in T2DM, WM microstructure abnormal-ities played a distinct and irreplaceable role in cognitionimpairments induced by T2DM (7,8). WM has a vital rolefor transferring information between GM regions, andits efficiency depends on WM microstructural integrity(9). However, the relationship between WM microstruc-tural changes and T2DM is debated and contradictory.Many studies suggested a close correlation betweenT2DM and WM lesion (WML) severity or progression(5,10), but others did not (11,12). This uncertainty couldbe partly due to the insufficient sensitivity of conven-tional MRI modalities in detecting subtle brain WMchanges or assessing the severity of WM hyperintensities(WMH) (7).
Diffusion tensor imaging (DTI), a new type of MRI, hasbeen developed as a powerful noninvasive technique toinvestigate WM microstructures and integrity (13,14).Fractional anisotropy (FA) and mean diffusivity (MD)
1State Key Laboratory of Cognitive Neuroscience and Learning and IDG/McGovernInstitute for Brain Research, Beijing Normal University, Beijing, People’s Republicof China2BABRI Centre, Beijing Normal University, Beijing, People’s Republic of China3Institute of Basic Research in Clinical Medicine, China Academy of TraditionalChinese Medicine, Beijing, People’s Republic of China
describe fiber density, axonal diameter, and myelinationin WM based on quantitative measure of the degree ofdiffusion anisotropy (15,16). As reported previously basedon the DTI technique, decreased FA or increased MD invarieties of cognitive dysfunction related to Alzheimerdisease (AD) (17,18), amnestic mild cognitive impairment(MCI) (19), schizophrenia (20,21), and type 1 diabetes(22), etc. Thus, DTI techniques are important for explor-ing more sensitive imaging-based biomarkers in preven-tion and early treatment of cognitive dysfunction inducedby T2DM. However, there were only a few DTI-basedinvestigations in WM disruption in T2DM (7,16), andHsu et al. (7) reported a remarkable FA decrease in thebilateral frontal WM and increased MD in T2DM patientsby assessing the DTI with global and voxel-based analyses(VBA).
However, VBA methods that many previous researchesperformed may encounter atrophy-induced WM tractmisregistration, which causes an unexpected decrease inaccuracy. Tract-based spatial statistics (TBSS), a corre-sponding fully automated whole-brain analysis techniquebased on DTI images, improves the sensitivity, objectivity,and interpretability of analysis of multisubject diffusionimaging and applies voxel-wise statistics to diffusionmeasures in various studies (23,24).
In our present study, we assessed cognitive functionswith a battery of neuropsychological tests, measuredWM structural integrity quantitatively using the TBSSmethod based on DTI data in T2DM subjects, and evaluatedthe correlation between WM disruptions and cognitivedysfunctions.
RESEARCH DESIGN AND METHODS
ParticipantsParticipants in this study were all from the Beijing AgeingBrain Rejuvenation Initiative (BABRI), which is an ongo-ing, longitudinal study investigating aging and cognitiveimpairment in urban elderly people in Beijing, China. Thecurrent study included 38 patients with diabetes (18male) and 34 age-, sex-, and education-matched healthycontrol subjects (17 male). All of the T2DM patients werediagnosed by a physician and had a history of using oralantidiabetic medications or insulin. Of the 38 T2DMsubjects, 13 subjects are under treatment with insulinand 27 subjects control blood glucose using oral hypogly-cemic agents. The duration of diabetes was defined as thenumber of years since diagnosis. All subjects had a medicalhistory and physical examination, during which height,weight, and BMI were recorded. Fasting plasma glucose(FPG), glycosylated hemoglobin (HbA1c), and cholesterollevels were measured with standard laboratory testing.Exclusion criteria for the groups were a previous historyof neurological or psychiatric disease including stroke,dementia, or transient ischemic attack and unsuitabilityfor MRI (e.g., due to a pacemaker, prosthetic heart valve,or claustrophobia). All participants provided written in-formed consent to our protocol that was approved by the
ethics committee of the State Key Laboratory of Cogni-tive Neuroscience and Learning, Beijing Normal Univer-sity. All subjects were right-handed and native Chinesespeakers.
Neuropsychological TestingAll of the participants received a battery of neuropsycho-logical tests assessing general mental status and othercognitive domains, such as episodic memory, tested usingthe Auditory Verbal Learning Test (AVLT) (25) and Rey-Osterrieth Complex Figure recall test; attention, testedusing the Trail Making Test Part a (TMT-a) and the Sym-bol Digit Modalities Test (SDMT) (26); visuospatial abil-ity, tested using the Rey-Osterrieth Complex Figure copytest and the Clock Drawing Test (CDT); language, testedusing the Category Verbal Fluency Test (CVFT) and theBoston Naming Test (BNT); and executive function,tested using the TMT-b and the Stroop Test. Verbal work-ing memory was assessed with the digit span of theWechsler Adults Intelligence Scale–Chinese revision. Log-ical reasoning was assessed with the Similarities subtestof the Wechsler Adult Intelligence Scale–Chinese revision.
Image AcquisitionThe MRI data were acquired on a 3.0T Siemens Tim MRIscanner in the Imaging Center for Brain Research, BeijingNormal University. Participants lay supine with the headsnugly fixed by a belt and foam pads to minimize headmotion. Two sets of DTI data scans were acquired forevery subject and then averaged during the data prepro-cessing. DTI images covering the whole brain were acquiredby using a single-shot, twice-refocused, diffusion-weightedecho planar imaging sequence with the following scanparameters: TR = 9,500 ms; TE = 92 ms; 30 diffusion-weighted directions with a b value of 1,000 s/mm2, anda single image with a b value of 0 s/mm2; slice thickness =2 mm; no interslice gap; 70 axial slices; matrix size = 1283128; field of view = 256 3 256 mm2; and voxel size = 232 3 2 mm3.
DTI Image PreprocessingAll the DTI image preprocessing and analyses describedbelow were implemented using a pipeline tool fordiffusion MRI (PANDA) (27). First, the DICOM files ofall subjects were converted into NIfTI images using thedcm2nii tool embedded in MRIcron. Second, the brainmask was estimated, and this step yielded the brainmask, which was required for the subsequent processingsteps. Third, the nonbrain space in the raw images was cutoff, leading to a reduced image size, reducing the memorycost, and speeding up the processing in subsequent steps.Fourth, each diffusion-weighted image was coregisteredto the b0 image using an affine transformation to correctthe eddy current–induced distortions and simple head-motion artifacts. The diffusion gradient directions wereadjusted accordingly. Fifth, a voxel-wise calculation of thetensor matrix and the diffusion tensor metrics wereyielded for each subject, including FA, MD, axial diffusiv-ity (l1), and radial diffusivity (l23).
diabetes.diabetesjournals.org Zhang and Associates 3597
NormalizingTo compare across subjects, location correspondence hasto be established. To this end, registration of the in-dividual images to a standardized template is alwaysapplied. Here, PANDA nonlinearly registered all of theindividual images in native space to a standardizedtemplate in the MNI space (27).
Voxel-Wise TBSS Statistical AnalysisThe voxel-wise statistical analysis in TBSS compares groupdifferences only on the WM skeleton, so that it providesbetter sensitivity, objectivity, and interpretability ofanalysis for multisubject DTI studies (23). The TBSSanalyses of FA, MD, l1, and l23 images were carried outusing the FMRIB software library (FSL 4.1.4; http://www.fmrib.ox.ac.uk/fsl). In brief, first the following five-step process on the FA images was performed: 1) the FAimage of each subject was aligned to a preidentified targetFA image (FMRIB58_FA) by nonlinear registration, 2) allof the aligned FA images were transformed onto theMNI152 template by affine registration, 3) a mean FAimage and its skeleton (mean FA skeleton) were createdfrom the images of all the subjects, 4) individual subject FAimages were projected onto the skeleton, and 5) voxel-wisestatistics across subjects were calculated for each point onthe common skeleton.
The voxel-wise statistics in TBSS were carried out usinga permutation-based inference tool for nonparametricstatistical thresholding (the “randomize” tool, part of FSL).In this study, voxel-wise group comparisons were performedusing nonparametric, two-sample Student t tests betweenthe T2DM and control groups. The mean FA skeleton wasused as a mask (thresholded at a mean FA value of 0.2),
and the number of permutations was set to 5,000. Thesignificance threshold for between-group differences wasset at P , 0.05 (family-wise error [FWE] corrected formultiple comparisons) using the threshold-free cluster en-hancement option in the “randomize” permutation-testingtool in FSL.
Data for MD, l1, and l23 were generated by applyingthe above FA transformations to additional diffusivitymaps and projecting them onto the skeleton with projec-tion vectors that were identical to the vectors inferredfrom the original FA data. The statistical analyses of thesediffusion tensor metrics were performed similarly to theFA analysis.
ROI-Wise Statistical AnalysisRecently, a few WM atlases (e.g., the ICBM-DTI-81 WMlabels atlas [see http://cmrm.med.jhmi.edu/] and the JHUWM tractography atlas) have been proposed (28). TheseWM atlases in the standard space allow for parcellation ofthe entire WM into multiple region of interests (ROIs),each representing a labeled region in the atlas. In ourcurrent study, to investigate the diffusion changes in spe-cific tracts, the ICBM-DTI-81 WM labels atlas was used toparcel the entire WM into 48 ROIs, and only the 40 ROIsin cerebral regions (we focused on the 40 WM tractswithin the cerebrum and did not consider the other 8ROIs within the cerebellum and brain stem) were usedfor the analysis (Fig. 1 and Supplementary Table 1).Then, the regional diffusion metrics (i.e., FA, MD, l1,and l23) were calculated by averaging the values withineach region of the WM atlas.
Linear regression analyses were performed to comparethe resultant ROI-based data between the T2DM and
Figure 1—The 40 WM tract ROIs based on the ICBM-DTI-81 WM labels atlas within cerebral regions. For the abbreviations of WM tracts,see Supplementary Table 1.
3598 White Matter Disruptions in Diabetes Diabetes Volume 63, November 2014
control groups. Age, sex, and education years were treatedas covariates in the regression analysis. A false discoveryrate (FDR) was applied to correct for multiple comparisons.
Demographic, clinical, neuropsychological, and behav-ioral data were analyzed in SPSS 17.0 (SPSS, Inc.). Thecomparisons of demographic and clinical data betweentwo groups were performed using Student t tests or x2
test. P , 0.05 was considered significant. In the twogroups separately, multiple linear regression analysiswas used to calculate the correlation between the resul-tant between-group different ROI-based data and behav-ioral performance, with age, sex, and education years ascovariates. The threshold value for establishing signifi-cance of group effects was set at P , 0.05 (uncorrectedfor multiple comparisons).
RESULTS
Demographics and Neuropsychological TestingThere were no significant differences in age, sex, years ofeducation, BMI, total cholesterol, triglycerides, HDL cho-lesterol, or LDL cholesterol between T2DM and controlsubjects. As expected, HbA1c (P , 0.001) and FPG levels(P , 0.001) were elevated in the T2DM group (Table 1). InT2DM patients, cognitive function in the domains of exec-utive function, spatial processing, attention, and workingmemory were significantly worse than healthy control sub-jects. Demographic data and neuropsychological testing arepresented in Table 2.
WM Skeleton Voxel-Wise TBSS ComparisonsVoxel-wise TBSS statistical analyses revealed significantlydecreased FA in widespread WM tracts in T2DM patientscompared with control subjects, including the wholecorpus callosum (CC), the bilateral corona radiata, in-ternal capsule (IC), posterior thalamic radiation, cingulum(cingulate gyrus), superior longitudinal fasciculus, inferior
longitudinal fasciculus, superior fronto-occipital fascicu-lus, external capsule (EC), fornix/striaterminalis, uncinatefasciculus, tapetum, and left cingulum (hippocampus)(P , 0.05, FWE corrected) (Fig. 2 and Table 3). Mean-while, significantly increased MD and l23 were found inonly some of the above tracts in T2DM patients comparedwith control subjects at P , 0.05 (FWE corrected). In-creased MD was observed in the whole CC, bilateral co-rona radiata, IC (except the right posterior limb of it), EC,posterior thalamic radiation, cingulum (cingulate gyrus),superior longitudinal fasciculus, and tapetum (Fig. 2 andTable 3). Increased l23 was observed in the whole CC,bilateral corona radiata, IC (except the right posteriorlimb), posterior thalamic radiation, cingulum (cingulate gy-rus), superior longitudinal fasciculus, inferior longitudinalfasciculus, EC, fornix/striaterminalis, uncinate fasciculus,tapetum, and right superior fronto-occipital fasciculus(Fig. 2 and Table 3). However, there were no significanttract-specific l1 differences between the two groups at P,0.05 (FWE corrected).
To identify the more severely impaired WM tracts inT2DM patients, we conducted a further investigation byusing more strict correction criterions of P , 0.01 andP , 0.002 (FWE corrected) in the voxel-wise TBSS statis-tical analysis. This process was only carried out on FAmaps. For P , 0.01 (FWE corrected), significantly de-creased FA was observed in the whole CC, bilateral coronaradiata, IC (except the right posterior limb), posteriorthalamic radiation, superior longitudinal fasciculus, infe-rior longitudinal fasciculus, superior fronto-occipital fas-ciculus, EC, cerebral peduncle, tapetum, left cingulum(hippocampus), fornix/striaterminalis, and uncinate fas-ciculus in T2DM patients compared with control subjects(Fig. 3 and Table 3). For P , 0.002 (FWE corrected),significantly decreased FA was found in the CC, bilateral
Table 1—Demographics and clinical characteristics for each group
Values are mean 6 SD. The comparisons of demographics between the two groups were performed using Student t tests. P , 0.05was considered significant. #Other diseases were excluded except hypertension. *The P value for sex distribution in the two groups wasobtained using a x2 test.
diabetes.diabetesjournals.org Zhang and Associates 3599
corona radiata, left anterior limb of the IC (ALIC.L), EC,and posterior thalamic radiation (Fig. 3 and Table 3).
Group Differences of Atlas-Based Tract ROIsFigure 4 illustrates the mean diffusion metrics of eachatlas-based tract ROIs with significant between-group dif-ferences after FDR correction in the T2DM and controlgroups. Compared with control subjects, T2DM patientshad significantly lower FA in the whole CC, ALIC.L, sag-ittal stratum, EC, uncinate fasciculus, right superiorfronto-occipital fasciculus, and bilateral tapetum (P ,0.05, FDR corrected). However, there were no significantdifferences in MD, l1, and l23 between the two groupsafter FDR correction.
Correlations Between ROI-Wise Diffusion Metrics andBehaviorsWe next examined the relationship between regionaldiffusion metrics of ROIs with significant group effects(P , 0.05, FDR corrected, i.e., the FA values of those 10ROIs in ROI-wise analysis) and neuropsychological scoreswith significant group differences (P , 0.05, i.e., R-Ocopy, Stroop C-B time, SDMT, and backward recall) intwo groups. In the T2DM group, only a negative correla-tion between the mean FA value of the left EC (EC.L) andStroop C-B time was found (r =20.376, P = 0.026) (Fig. 5).In addition, a marginally significant correlation in T2DMpatients between the mean FA of the ALIC.L and the
Stroop C-B time score (r = 20.323, P = 0.059) was alsotaken into consideration (Fig. 5).
DISCUSSION
In the current study, we found that T2DM patientsshowed 1) various cognitive impairments, including exec-utive function, spatial processing, attention, and workingmemory; 2) widespread WM disruptions, especially in thewhole CC, ALIC.L, and EC.L; and 3) a positive correlationbetween executive function and WM integrity in ALIC.Land EC.L. Considering that WM microstructural impair-ments underlie common mechanisms of cognitive dys-function and disrupt the large-scale distributed braincognitive networks, our findings imply that extensiveWM destruction plays a distinct role in cognition declineassociated with T2DM. Our investigation provides novelinsight into the neuropathological effects of WM integrityreduction on cognition in T2DM.
FA and MD are the primary DTI-derived metricsbelieved to reflect overall WM health, maturation, andorganization (15,16). In addition to these primary DTImeasures, l1, which reflects axon integrity, and l23,which reflects myelin sheath integrity, can be useful inunderstanding the underlying physiology (8,29). Basedon the voxel-wise TBSS statistical analysis, we found de-creased FA in almost the whole cerebral WM skeleton inT2DM (P , 0.05, FWE corrected) and increased MD in
Table 2—Neuropsychological test results for T2DM and control groups
All of the subjects (control subjects and mild WML subjects) were matched for age, sex, and education. Values are mean 6 SD.P , 0.05 was considered significant. The differences in neuropsychological scores between the two groups were tested for signif-icance with ANOVA adjusted for age, sex, and education. MMSE, Mini-Mental State Examination; R-O recall, Rey-Osterrieth Complex Figurerecall tests; R-O copy, Rey-Osterrieth Complex Figure copy test; Stroop C-B time, difference in time between Stroop test parts B and A.
3600 White Matter Disruptions in Diabetes Diabetes Volume 63, November 2014
the whole CC, bilateral corona radiata, IC (except the rightposterior limb of it), EC, posterior thalamic radiation,cingulum (cingulate gyrus), superior longitudinal fascicu-lus, and tapetum. These widespread WM microstructuralimpairments disrupt the large-scale distributed brain cog-nitive networks and underlie the various cognitive dys-functions in T2DM. To date, there are several studiessuggesting various WM impairments in diabetic patients.Partly consistent with our results, Kodl et al. (22) sug-gested that diabetic subjects had significantly lower FAthan control subjects in the posterior corona radiataand the optic radiation. Hsu et al. (7) reported a remark-able FA decrease in the bilateral frontal WM in T2DMpatients by assessing the DTI with VBA. Hoogenboomet al. (8) found that T2DM patients showed lower FA inthe cingulum bundle and the uncinate fasciculus. Despitethese findings, disrupted WM in T2DM patients in ourstudy exhibited significantly decreased FA, and increasedMD and l23, with no striking between-group differencesin l1. Accordingly, the alteration of FA and MD might bemainly attributed to the increase in l23 but not l1. Giventhat l23 is a specific marker of myelin alterations whereasl1 is more related to axonal injury (30), our data showedthat demyelination may be an important contributingfactor to WM abnormalities in T2DM.
To identify the more severely impaired WM tractsin T2DM patients, we used a stricter correction criterion(P , 0.002, FWE corrected) in the voxel-wise TBSS sta-tistical analysis and found significant differences between
the T2DM and control group in the FA values of eightWM tracts. Furthermore, we measured the mean value ofDTI metrics of ROIs by ROI-wise statistical analysis. Sig-nificantly reduced mean FA values were detected in 10WM ROIs in T2DM patients (after FDR correction). Com-bining the results of the two analyses (both the ROI-wisestatistical analysis and the voxel-wise TBSS statisticalanalysis), T2DM patients showed WM microstructuralimpairments in WM tracts of whole CC, the ALIC.L, andthe EC.L when compared with the control group. The CCplays an important role in interhemispheric functionalintegration. In AD (31,32) and MCI (32,33), the loss ofWM connectivity and regionally specific atrophy of the CCare observed. It follows that CC abnormalities may berelated to cognitive deficits. The ALIC.L is a portion ofthe IC that connects the medial and anterior thalamicnuclei with the prefrontal cortex and the cingulate gyrus.Several studies have found that the ALIC also plays animportant role in cognition function. The lower FA valueof the ALIC correlates significantly with impaired declar-ative/episodic memory performance (34) and cognitivelevels (35). The EC is a route for cholinergic fibers fromthe basal forebrain to the cerebral temporal cortex, trans-mitting auditory and polymodal sensory information. Asreported, the EC is crucial for cognition, including mem-ory and executive functioning (36). Some studies (37)suggested that T2DM exacerbates the damage of WM in-tegrity in the IC and EC bilaterally, which can result inexecutive dysfunction and memory deficits and may be
Figure 2—Voxel-wise TBSS analysis results of FA, MD, l1, and l23 images between T2DM and control groups. Green represents the meanWM skeleton of all subjects. Blue–light blue (thickened for better visibility) represents regions with decreased FA in T2DM group (P < 0.05,FWE corrected). Red-yellow (thickened for better visibility) represents regions with increased MD and l23 in T2DM patients compared withcontrol subjects (P < 0.05, FWE corrected).
diabetes.diabetesjournals.org Zhang and Associates 3601
a downstream consequence of poor learning efficiencysecondary to the executive dysfunction. Therefore, thesethree WM tracts may play particularly important roles inT2DM-induced cognitive dysfunction.
To further understand the brain mechanism of cogni-tive impairment, we performed a correlation analysisbetween WM impairment and cognition decline. Wefound FA in the EC.L and ALIC.L positively correlatedwith Stroop C-B time in the T2DM group, whichrepresents executive function. Executive functions are
high-level cognitive functions for the management ofa series of cognitive processes (38), including workingmemory, problem solving, planning, and execution (39).Several studies reported a close relationship between ex-ecutive dysfunction and WM impairment. Sun et al. (40)found declines in executive function in mild WML sub-jects. Zheng et al. (41) revealed that executive functiondirectly correlated with FA in frontal WM tracts, espe-cially the ALIC. Smith et al. (42) found that WMH volumein the ALIC was strongly inversely associated with
Table 3—Diffusion changes in the WM tracts in diabetic patients during the voxel-wise and ROI-wise statistical analysis
Tracts
Voxel-wise TBSS statistical analysisROI-wise
statistical analysis
FA (P , 0.05,FWE)
MD (P , 0.05,FWE)
l1 (P , 0.05,FWE)
l23 (P , 0.05,FWE)
FA (P , 0.01,FWE)
FA (P , 0.002,FWE)
FA (P , 0.05,FDR)
CC g, b, s g, b, s — g, b, s g, b, s g, b, s* g, b, s*
SLF Bilateral Bilateral — Bilateral Bilateral — Right
SFOF Bilateral — — Right — — —
UF Bilateral — — Bilateral — — Left
TAP Bilateral Bilateral — Bilateral — — Bilateral
g, b, s represents genu, body, and splenium of the CC. For the abbreviations of WM tracts, see Supplementary Table 1. *Disrupted WMtracts in both voxel-wise TBSS analysis and ROI-wise analysis.
Figure 3—Further results of the TBSS analysis using stricter correction criterions (all after FWE correction). Green represents the mean WMskeleton of all subjects. Blue–light blue (thickened for better visibility) represents regions with decreased FA in T2DM group.
3602 White Matter Disruptions in Diabetes Diabetes Volume 63, November 2014
executive function. Similarly, subjects with WM integrityimpairments of the EC, reflected by decreased FA, alsoshowed significantly poorer performance on executivefunction tasks (43). Therefore, we propose that executivefunction might be partially supported by the WM integ-rity of the ALIC and EC. On the other hand, consideringthat there were no significant associations between WMintegrity and cognition in the control group, we believethat executive function may be more sensitive to WMintegrity variations in T2DM. WM integrity could bemore beneficial to neuropsychological performance inT2DM subjects than in control subjects.
As a major risk factor for dementia, diabetes isreported to be closely associated with the pathologicalprocesses of various dementias such as AD and fronto-temporal dementia (FTD). AD can be regarded as a cortical
disconnection syndrome that affects not only the corticalneurons but also the axons and dendrites in the cerebralWM (18). MCI, the prodromal phase of AD, impairs mem-ory function and disrupts widespread WM tracts (19).Although extensive WM alterations were also exhibitedin T2DM patients in the current study, some studies sug-gested that T2DM might not be a driving factor of AD butmight add to the damaging effects of AD on cognition.Correspondingly, executive dysfunctions are closely re-lated to WM abnormalities in T2DM. At the same time,many investigations have found that FTD affects executivefunction through regional disruption of the frontal andtemporal lobes (44), where the more severely disruptedWM tracts were found in T2DM. Therefore, we predictthat T2DM might be associated with FTD. However, therelationships between T2DM and AD or FTD are not clear
Figure 4—Mean diffusion metrics and group differences of each atlas-based tract ROIs in T2DM and control groups. All the ROIs shown inthe figure were significantly different after FDR correction. For the abbreviations of WM tracts, see Supplementary Table 1.
Figure 5—The significant correlations between ROI-wise diffusion metrics and behaviors in T2DM and control groups, respectively. For theabbreviations of WM tracts, see Supplementary Table 1.
diabetes.diabetesjournals.org Zhang and Associates 3603
to date. An interaction analysis of T2DM and MCI, as wellas follow-up investigations, is worthy of further study inour future research.
It is also interesting that we found that there weremore abnormal WM tracts in the left hemisphere than inthe right. Thompson et al. (45) suggest that the shiftingGM deficits in AD were asymmetric (left . right hemi-sphere) and correlated with progressively declining cogni-tive status. The HAROLD (hemispheric asymmetryreduction in older adults) model states that age-relatedhemispheric asymmetry reductions may function to com-pensate for the lower prefrontal activity lateralizationduring cognitive performances in older adults than inyounger adults (46). Accordingly, the mechanisms of howT2DM affects hemispheric symmetry of WM integrityare worthy of further investigation. Additionally, ourstudy is limited by its cross-sectional design, which shouldbe interpreted cautiously, and longitudinal studies areneeded to investigate the conversion of T2DM to demen-tia and to evaluate clinical values of WM metrics to pre-dict longitudinal changes. Further, we only focused on theWM changes in T2DM patients in this study. The rela-tionship between these functional and structural changes,however, is still unclear. In future studies, the putativeconnection between functional connectivity and WM de-generation could be examined.
In conclusion, there are widespread WM integritydisruptions, which are likely caused by demyelination,and various cognitive impairments in T2DM patients,among which executive dysfunction closely correlateswith WM abnormalities. The current investigation willcontribute to a better understanding of the neuropatho-logical process in T2DM and may lead to better imaging-based biomarkers for prevention and early treatment ofcognitive dysfunction caused by T2DM.
Funding. This work was supported by the Beijing New Medical DisciplineBased Group (grant 100270569), the Natural Science Foundation of China (grants30873458 and 81173460), the Institute of Basic Research in Clinical Medicine,China Academy of Chinese Medical Sciences (grant Z0175), and the program forNew Century Excellent Talents in University (grant NCET-10-0249).Duality of Interest. No potential conflicts of interest relevant to this articlewere reported.Author Contributions. J.Z. performed the experiments and wrote themanuscript. Yu.W. performed the experiments, analyzed the results, and wrotethe manuscript. J.W., X.Z., and N.S. performed the experiments. Yo.W. and Z.Z.designed the experiments and interpreted the results. Z.Z. is the guarantor of thiswork and, as such, had full access to all the data in the study and takesresponsibility for the integrity of the data and the accuracy of the data analysis.
References1. Yang W, Lu J, Weng J, et al.; China National Diabetes and MetabolicDisorders Study Group. Prevalence of diabetes among men and women in China.N Engl J Med 2010;362:1090–11012. Biessels GJ, Staekenborg S, Brunner E, Brayne C, Scheltens P. Risk ofdementia in diabetes mellitus: a systematic review. Lancet Neurol 2006;5:64–743. Munshi M, Grande L, Hayes M, et al. Cognitive dysfunction is associatedwith poor diabetes control in older adults. Diabetes Care 2006;29:1794–1799
4. Kodl CT, Seaquist ER. Cognitive dysfunction and diabetes mellitus. EndocrRev 2008;29:494–5115. van Harten B, Oosterman J, Muslimovic D, van Loon BJP, Scheltens P,Weinstein HC. Cognitive impairment and MRI correlates in the elderly patientswith type 2 diabetes mellitus. Age Ageing 2007;36:164–1706. Reijmer YD, van den Berg E, Ruis C, Kappelle LJ, Biessels GJ. Cognitive dys-function in patients with type 2 diabetes. Diabetes Metab Res Rev 2010;26:507–5197. Hsu J-L, Chen Y-L, Leu J-G, et al. Microstructural white matter abnor-malities in type 2 diabetes mellitus: a diffusion tensor imaging study. Neuroimage2012;59:1098–11058. Hoogenboom WS, Marder TJ, Flores VL, et al. Cerebral white matter in-tegrity and resting-state functional connectivity in middle-aged patients with type2 diabetes. Diabetes 2014;63:728–7389. Hagmann P, Cammoun L, Gigandet X, et al. Mapping the structural core ofhuman cerebral cortex. PLoS Biol 2008;6:e15910. Manschot SM, Biessels GJ, de Valk H, et al.; Utrecht Diabetic Encepha-lopathy Study Group. Metabolic and vascular determinants of impaired cognitiveperformance and abnormalities on brain magnetic resonance imaging in patientswith type 2 diabetes. Diabetologia 2007;50:2388–2397
11. Schmidt R, Launer LJ, Nilsson L-G, et al.; CASCADE Consortium. Magneticresonance imaging of the brain in diabetes: the Cardiovascular Determinants ofDementia (CASCADE) Study. Diabetes 2004;53:687–69212. van Harten B, de Leeuw F-E, Weinstein HC, Scheltens P, Biessels GJ. Brainimaging in patients with diabetes: a systematic review. Diabetes Care 2006;29:2539–254813. Jones DK, Leemans A. Diffusion tensor imaging. In Magnetic ResonanceNeuroimaging. Springer, 2011, p. 127–14414. Le Bihan D, Mangin JF, Poupon C, et al. Diffusion tensor imaging: conceptsand applications. J Magn Reson Imaging 2001;13:534–54615. Basser PJ, Mattiello J, LeBihan D. MR diffusion tensor spectroscopy andimaging. Biophys J 1994;66:259–26716. Reijmer YD, Brundel M, de Bresser J, Kappelle LJ, Leemans A, Biessels GJ;Utrecht Vascular Cognitive Impairment Study Group. Microstructural white matterabnormalities and cognitive functioning in type 2 diabetes: a diffusion tensorimaging study. Diabetes Care 2013;36:137–14417. Chen SQ, Kang Z, Hu XQ, Hu B, Zou Y. Diffusion tensor imaging of the brainin patients with Alzheimer’s disease and cerebrovascular lesions. J Zhejiang UnivSci B 2007;8:242–247
18. Stahl R, Dietrich O, Teipel SJ, Hampel H, Reiser MF, Schoenberg SO. Whitematter damage in Alzheimer disease and mild cognitive impairment: assessmentwith diffusion-tensor MR imaging and parallel imaging techniques. Radiology2007;243:483–492
19. Li H, Liang Y, Chen K, et al. Different patterns of white matter disruptionamong amnestic mild cognitive impairment subtypes: relationship with neuro-psychological performance. J Alzheimers Dis 2013;36:365–37620. Hoptman MJ, Ardekani BA, Butler PD, Nierenberg J, Javitt DC, Lim KO. DTIand impulsivity in schizophrenia: a first voxelwise correlational analysis. Neu-roreport 2004;15:2467–2470
21. Lim KO, Hedehus M, Moseley M, de Crespigny A, Sullivan EV, PfefferbaumA. Compromised white matter tract integrity in schizophrenia inferred from dif-fusion tensor imaging. Arch Gen Psychiatry 1999;56:367–37422. Kodl CT, Franc DT, Rao JP, et al. Diffusion tensor imaging identifies deficitsin white matter microstructure in subjects with type 1 diabetes that correlate withreduced neurocognitive function. Diabetes 2008;57:3083–308923. Smith SM, Jenkinson M, Johansen-Berg H, et al. Tract-based spatial statistics:voxelwise analysis of multi-subject diffusion data. Neuroimage 2006;31:1487–150524. Smith SM, Johansen-Berg H, Jenkinson M, et al. Acquisition and voxelwiseanalysis of multi-subject diffusion data with tract-based spatial statistics. NatProtoc 2007;2:499–50325. Rosenberg SJ, Ryan JJ, Prifitera A. Rey Auditory-Verbal Learning Testperformance of patients with and without memory impairment. J Clin Psychol1984;40:785–787
3604 White Matter Disruptions in Diabetes Diabetes Volume 63, November 2014
26. Sheridan LK, Fitzgerald HE, Adams KM, et al. Normative Symbol DigitModalities Test performance in a community-based sample. Arch Clin Neuro-psychol 2006;21:23–2827. Cui Z, Zhong S, Xu P, He Y, Gong G. PANDA: a pipeline toolbox for analyzingbrain diffusion images. Front Hum Neurosci 2013;7:4228. Mori S, Oishi K, Jiang H, et al. Stereotaxic white matter atlas based ondiffusion tensor imaging in an ICBM template. Neuroimage 2008;40:570–58229. Budde MD, Kim JH, Liang HF, et al. Toward accurate diagnosis of whitematter pathology using diffusion tensor imaging. Magn Reson Med 2007;57:688–69530. Song S-K, Sun S-W, Ju W-K, Lin S-J, Cross AH, Neufeld AH. Diffusiontensor imaging detects and differentiates axon and myelin degeneration in mouseoptic nerve after retinal ischemia. Neuroimage 2003;20:1714–172231. Yamauchi H, Fukuyama H, Nagahama Y, et al. Comparison of the pattern ofatrophy of the corpus callosum in frontotemporal dementia, progressive supra-nuclear palsy, and Alzheimer’s disease. J Neurol Neurosurg Psychiatry 2000;69:623–62932. Di Paola M, Spalletta G, Caltagirone C. In vivo structural neuroanatomy ofcorpus callosum in Alzheimer’s disease and mild cognitive impairment usingdifferent MRI techniques: a review. J Alzheimers Dis 2010;20:67–9533. Anstey KJ, Mack HA, Christensen H, et al. Corpus callosum size, reactiontime speed and variability in mild cognitive disorders and in a normative sample.Neuropsychologia 2007;45:1911–192034. Levitt JJ, Alvarado JL, Nestor PG, et al. Fractional anisotropy and radialdiffusivity: diffusion measures of white matter abnormalities in the anterior limbof the internal capsule in schizophrenia. Schizophr Res 2012;136:55–6235. Wang S, Fan G, Xu K, Wang C. Potential of diffusion tensor MR imaging inthe assessment of cognitive impairments in children with periventricular leuko-malacia born preterm. Eur J Radiol 2013;82:158–164
36. Jung RE, Chavez RS, Flores RA, Qualls C, Sibbitt WL Jr, Roldan CA. Whitematter correlates of neuropsychological dysfunction in systemic lupus erythe-matosus. PLoS ONE 2012;7:e2837337. Yau P, Hempel R, Tirsi A, Convit A. Cerebral white matter and retinal arterialhealth in hypertension and type 2 diabetes mellitus. Int J Hypertens 2013;2013:32960238. Elliott R. Executive functions and their disorders. Br Med Bull 2003;65:49–5939. Chan RC, Shum D, Toulopoulou T, Chen EY. Assessment of executivefunctions: review of instruments and identification of critical issues. Arch ClinNeuropsychol 2008;23:201–21640. Sun X, Liang Y, Wang J, et al. Early frontal structural and functional changesin mild white matter lesions relevant to cognitive decline. J Alzheimers Dis 2014;40:123–13441. Zheng Z, Shemmassian S, Wijekoon C, Kim W, Bookheimer SY, Pouratian N.DTI correlates of distinct cognitive impairments in Parkinson’s disease. HumBrain Mapp 2014;35:1325–133342. Smith EE, Salat DH, Jeng J, et al. Correlations between MRI white matterlesion location and executive function and episodic memory. Neurology 2011;76:1492–149943. Wang XQ, Lang SY, Hong LU, Lin MA, Yan-ling MA, Yang F. Changes inextratemporal integrity and cognition in temporal lobe epilepsy: a diffusion tensorimaging study. Neurol India 2010;58:891–89944. Seelaar H, Rohrer JD, Pijnenburg YA, Fox NC, van Swieten JC. Clinical,genetic and pathological heterogeneity of frontotemporal dementia: a review.J Neurol Neurosurg Psychiatry 2011;82:476–48645. Thompson PM, Hayashi KM, de Zubicaray G, et al. Dynamics of gray matterloss in Alzheimer’s disease. J Neurosci 2003;23:994–100546. Cabeza R. Hemispheric asymmetry reduction in older adults: the HAROLDmodel. Psychol Aging 2002;17:85–100
diabetes.diabetesjournals.org Zhang and Associates 3605
