Intrinsic Brain Abnormalities in Attention Deficit Hyperactivity Disorder: A Resting-State Functional MR Imaging Study

Orig

inal

res

earc

h n

Neu

rora

diol

ogy

514 radiology.rsna.org n Radiology: Volume 272: Number 2—August 2014

intrinsic Brain abnormalities in attention Deficit hyperactivity Disorder: A Resting-State Functional MR Imaging Study1

Fei Li, PhDNing He, MDYuanyuan Li, MDLizhou Chen, MDXiaoqi Huang, PhDSu Lui, PhDLanting Guo, MDGraham J. Kemp, MA, DMQiyong Gong, MD, PhD

Purpose: To explore alterations of regional and network-level neural function using resting-state functional magnetic resonance (MR) imaging in children and adolescents with attention deficit hyperactivity disorder (ADHD) and to assess the association between these alterations of intrinsic neural activity and executive dysfunction in ADHD.

Materials and Methods:

This prospective study was approved by the local ethical committee, and written informed consent was obtained from guardians of all participants. Thirty-three boys with ADHD who were not receiving medication and who were without comorbidity (aged 6–16 years) and 32 healthy control subjects (aged 8–16 years) underwent imaging by using resting-state functional MR imaging. Amplitude of low-frequency fluctuation (ALFF) and seed-based func-tional connectivity (FC) were calculated to examine re-gional neural function and functional integration, respec-tively, and were compared between patients and control subjects by using the voxel-based two-sample t test, while Pearson correlation analyses were performed to identify neural correlates of executive function measured with the Wisconsin Card Sorting Test and the Stroop Color-Word Test.

Results: Relative to healthy control subjects, patients with ADHD showed impaired executive function (P , .05), along with the following: lower ALFF in the left orbitofrontal cortex (P = .004) and the left ventral superior frontal gyrus (P = .003); higher ALFF in the left globus pallidus (P = .004), the right globus pallidus (P = .002), and the right dor-sal superior frontal gyrus (P = .025); lower long-range FC in the frontoparietal and frontocerebellar networks; and higher FC in the frontostriatal circuit that correlated across subjects with ADHD with the degree of executive dysfunction (P , .05).

Conclusion: These findings of focal spontaneous hyper- and hypo-function, together with altered brain connectivity in the large-scale resting-state networks, which correlates with executive dysfunction, point to a connectivity-based path-ophysiologic process in ADHD.

q RSNA, 2014

Online supplemental material is available for this article.

1 From the Huaxi MR Research Center, Department of Radiology (F.L., L.C., X.H., S.L., Q.G.), and Department of Psychiatry, the State Key Laboratory of Biotherapy (N.H., Y.L., L.G.), West China Hospital of Sichuan University, 37 Guo Xue Xiang, Chengdu, Sichuan 610041, China; and Magnetic Resonance and Image Analysis Research Centre and Institute of Ageing and Chronic Disease, University of Liverpool, Liverpool, England (G.J.K.). Received July 11, 2013; revision requested September 26; final revision received January 8, 2014; accepted February 3; final version accepted February 18. Supported by the National Natural Science Foundation (grants 81030027, 81227002, 30970903, and 81220108013), National Key Technologies R&D Program (Program No. 2012BAI01B03), and Program for Changjiang Scholars and Innovative Research Team in University of China. Q.G. supported by a CMB Distinguished Professorship Award (Award No. F510000/G16916411) administered by the Institute of International Education, USA. Address correspondence to Q.G. (e-mail: [email protected]).

q RSNA, 2014

Note: This copy is for your personal non-commercial use only. To order presentation-ready copies for distribution to your colleagues or clients, contact us at www.rsna.org/rsnarights.

Radiology: Volume 272: Number 2—August 2014 n radiology.rsna.org 515

NEURORADIOLOGY: Brain Abnormalities in Attention Deficit Hyperactivity Disorder Li et al

and to assess the association between these alterations of intrinsic neural activity and executive dysfunction in ADHD.

Materials and Methods

ParticipantsThis prospective study was approved by the local ethical committee of the West China Hospital of Sichuan University (Sichuan, China), and written informed consent was obtained from guardians of all participants. Subjects were recruit-ed and studied between June 2009 and December 2011 at the Mental Health Center, West China Hospital of Sichuan University (Chengdu, Sichuan, China): Forty-two right-handed male patients with ADHD who were not receiving medication and who were without co-morbidity and 36 healthy right-handed male control subjects of similar age and education were recruited from the local area.

Diagnosis of ADHD was determined by three experienced clinical psychia-trists (L.G., Y.L., N.H., with 27, 6, and 5 years of experience in clinical psychia-try, respectively) by using the Structured Clinical Interview for DSM-IV (Patient

MR imaging, in which spontaneous blood oxygen level–dependent fluctua-tions are thought to reflect spontane-ous neural function in the resting state (6,7). Two derived measurement in-dexes, regional amplitude low-frequen-cy fluctuation (ALFF) and functional connectivity (FC), provide information on regional activity and network-level brain function, respectively (8,9). In contrast to the task-based approach, resting-state functional MR imaging is not susceptible to potential perfor-mance confounds and is relatively easy to implement in a clinical setting (10). Evidence from resting-state functional MR imaging has underpinned the de-velopment of models of ADHD that encompass a number of large-scale resting-state networks, suggesting that aberrant frontosubcortical circuits are not a sufficient explanation (5).

However, the specific neuropsy-chological dysfunction underlying such atypical brain network profiles remains poorly understood (11). Patients with ADHD often exhibit deficiencies in cog-nitive function, particularly in executive functions such as strategic planning, set switching, cognitive flexibility, and interference inhibition (12), which can be defined by neuropsychological tests such as the Wisconsin Card Sorting Test (WCST) (13) and the Stroop Color-Word Test (14). A link to executive dys-function is central to models involving aberrant frontosubcortical circuits; how-ever, there has been much debate about what core deficit of brain function might cause the impairments of ADHD (2).

The aims of the present study were to explore alterations of regional and network-level neural function by using resting-state functional MR imaging in children and adolescents with ADHD

A ttention deficit hyperactivity dis-order (ADHD), characterized by age-inappropriate degrees of inat-

tention, hyperactivity, and impulsivity, affects about 5% of children and adoles-cents worldwide (1). Researchers who have performed functional magnetic resonance (MR) imaging studies ex-ploiting blood oxygen level–dependent signal contrast have detected abnormal-ities in specific brain areas in ADHD, leading to the suggestion that frontos-triatal dysfunction is important in the pathophysiologic functional changes of the disorder (2). However, reported results have been inconsistent: For ex-ample, in functional MR imaging with a response-inhibition task, patients with ADHD have been reported as showing both decreased (3) and increased (4) activation in prefrontal regions. In view of this complexity, rather than focus on the dysregulation of particular areas in particular tasks, a better approach to the pathophysiologic functional changes of ADHD may be to consider the whole set of brain systems (5).

A task-independent approach to assessing regional and network-level brain function is resting-state functional

Implication for Patient Care

n Exploration of the association between intrinsic neural activity and executive function might be useful in better characterizing patients with ADHD and in un-derstanding the pathophysiologic abnormalities underlying this condition.

Advances in Knowledge

n Using resting-state functional MR imaging, we found altered re-gional brain function located in the prefrontal cortex and globus pallidus, as well as aberrant func-tional connectivity (FC) in large-scale networks, including fronto-striatal, frontoparietal, and frontocerebellar circuits, in chil-dren and adolescents with atten-tion deficit hyperactivity disorder (ADHD) (P , .05, familywise error corrected).

n The associations between altered FC within the frontostriatal cir-cuit and executive dysfunction (P , .05, two-tailed test) suggested that the characteristics of the brain’s resting-state functional architecture are relevant to un-derstanding relationships between neural substrate and executive function in ADHD.

Published online before print10.1148/radiol.14131622 Content codes:

Radiology 2014; 272:514–523

Abbreviations:ADHD = attention deficit hyperactivity disorderALFF = amplitude low-frequency fluctuationBA = Brodmann areaFC = functional connectivityWCST = Wisconsin Card Sorting Test

Author contributions:Guarantors of integrity of entire study, F.L., N.H., Y.L., L.C., L.G., Q.G.; study concepts/study design or data acquisition or data analysis/interpretation, all authors; manuscript drafting or manuscript revision for important intellectual content, all authors; approval of final version of submitted manuscript, all authors; literature research, F.L., N.H., Y.L., L.C., L.G., G.J.K., Q.G.; clinical studies, F.L., N.H., Y.L., L.C., X.H., S.L., L.G., Q.G.; experimental studies, F.L., N.H., Y.L., L.C., S.L., L.G., Q.G.; statistical analysis, F.L., N.H., Y.L., L.C., L.G., Q.G.; and manuscript editing, F.L., N.H., Y.L., L.C., S.L., L.G., G.J.K., Q.G.

Conflicts of interest are listed at the end of this article.

516 radiology.rsna.org n Radiology: Volume 272: Number 2—August 2014

NEURORADIOLOGY: Brain Abnormalities in Attention Deficit Hyperactivity Disorder Li et al

University, Beijing, China; http://www.restfmri.net) (21). For each participant, the first five volumes were discarded to allow for imaging unit stabilization and subject familiarization, and the remain-ing images were section-time correct-ed, realigned to the middle volume, and unwarped to correct for susceptibility-by-movement interaction. Realigned images were spatially normalized to the Montreal Neurological Institute tem-plate, and each voxel was resampled to 3 3 3 3 3 mm3. Previous work has in-dicated that normalization of children’s MR images to standard adult templates is acceptable for statistical group com-parisons (22,23). Head translation movement for all participants was less than 2 mm, and rotation was less than 2°. Analysis of head motion parameters in statistical parametric mapping soft-ware (SPM8; Wellcome Trust Centre for Neuroimaging, University College London) did not reveal differences in motion correction parameters between patients with ADHD versus control subjects: For translation, the mean was 0.34 mm 6 0.14 versus 0.29 mm 6 0.12 (P = .209); for rotation, the mean was 0.3° 6 0.16 versus 0.27° 6 0.13 (P = .357). To regress out the brain volume effects that might confound the neural functional activity, with details included in Appendix E1 (online), we also acquired a high-spatial-resolution three-dimensional T1-weighted image by using a spoiled gradient-recalled echo sequence (1.9/2.26; flip angle, 9°; number of axial sections, 176; section thickness, 1 mm; gap, none; field of view, 256 3 256 mm2; and data matrix, 256 3 256).

Amplitude of Low-Frequency FluctuationALFF values were calculated by using software (Data Processing Assistant for Resting-State Functional MR Im-aging, version 2.3; State Key Labora-tory of Cognitive Neuroscience and Learning, Beijing Normal Univcrsity) with a procedure similar to that used in our earlier study (24). In brief, after linear-trend removal, the time series was transformed to the frequency do-main by using fast-Fourier transform (taper percentage, zero; fast-Fourier

scores from the Child Behavior Check-list (rated by parents) (17) and the hy-peractivity index from the revised Con-ners’ Parent Rating Scale (18).

The modified WCST and the Stroop Color-Word Test were used to assess executive function after the imaging ses-sion. The widely used WCST measures aspects of executive function, including strategic planning, cognitive shifting, and perseveration (19). Five raw scores were used for statistical analysis: total correct, total errors, perseverative er-rors, nonperseverative errors, and cat-egories completed. The Stroop Color-Word Test is used to assess executive inhibitory control (20). Four raw scores were used for statistical analysis: right numbers, error numbers, correction numbers, and total time.

MR Data Acquisition and PreprocessingAll subjects underwent resting-state functional MR imaging by using a 3-T MR imaging system (Trio; Siemens, Er-langen, Germany) with an eight-channel phased-array head coil. Blood oxygen level–dependent–sensitive MR images were obtained with a gradient-echo echo-planar imaging sequence with the following parameters: repetition time msec/echo time msec, 2000/30; flip angle, 90°; section thickness, 5 mm; in-tersection gaps, none; matrix size, 64 3 64; field of view, 240 3 240 mm2; and voxel size, 3.75 3 3.75 3 5 mm3. Each brain volume comprised 30 axial sec-tions, and each functional imaging ses-sion contained 200 image volumes pre-ceded by five dummy volumes. During imaging, subjects were instructed to re-lax with their eyes closed without falling asleep and without directed, systematic thought (confirmed by subjects imme-diately after the experiment). Func-tional image preprocessing and statis-tical analysis were performed by using statistical parametric mapping (SPM8; Wellcome Trust Centre for Neuroimag-ing, University College London, Lon-don, England; http://www.fil.ion.ucl.ac.uk/spm/) and the Data Processing Assistant for Resting-State Functional MR Imaging software (version 2.3; State Key Laboratory of Cognitive Neu-roscience and Learning, Beijing Normal

Edition) to exclude conduct disorder, op-positional defiant disorder, and Tourette disorder, or any other axis I psychiatric comorbid disorders. The following were exclusion criteria: (a) history of stimulant or other medication to treat inattention problems; (b) head motion more than 2 mm or 2° during resting-state functional MR imaging, or inability to complete the executive function tests; (c) left-handed-ness, as assessed with the Annett Hand Preference Questionnaire (15); (d) full-scale IQ less than 90 according to an age-appropriate Wechsler Intelligence Scale for Children–Chinese Revision (16); or (e) any previous head trauma, neurologic disorders or psychosurgery, and any sub-stantial physical illness. A patient would have to meet just one of the exclusion cri-teria to be excluded.

The healthy control subjects were screened by using the Structured Clin-ical Interview for DSM-IV (Nonpatient Edition); they had no history of psy-chiatric illness in first-degree relatives. Exclusion criteria for the control sub-jects were the same as for the ADHD group, and the IQ did not differ signifi-cantly between patients with ADHD and control subjects (P . .05). Diagnostic-quality MR images were reported by two experienced neuroradiologists (S.L. and F.L., with 11 and 6 years of experience in neuroimaging, respectively). We ex-cluded six patients with ADHD and four control subjects who showed excessive head movement more than 2 mm or 2° during imaging and three patients with ADHD who proved unable to complete the executive function tests. The final study consisted of 33 right-handed male patients with ADHD who were not re-ceiving medication (mean age, 10.1 years 6 2.6 [standard deviation]; range, 6–16 years; 22 with the combined sub-type [both inattention and hyperactiv-ity and/or impulsivity symptoms] and 11 with the inattentive subtype) and 32 right-handed male healthy control subjects (mean age, 10.9 years 6 2.6; range, 8–16 years).

Behavioral Measures and Executive Function TestsPrimary behavioral measures were as-sessed by using attention problems

Radiology: Volume 272: Number 2—August 2014 n radiology.rsna.org 517

NEURORADIOLOGY: Brain Abnormalities in Attention Deficit Hyperactivity Disorder Li et al

of altered ALFF and FC were described in terms of standard Talairach coordi-nates, after conversion from Montreal Neurological Institute space.

In the ADHD group, two-tailed Pearson correlation analyses were per-formed in software (SPSS 16.0; SPSS, Chicago, Ill) to assess the relationship between characteristics of executive function tests (WCST and Stroop Color-Word Test) and averaged eigenvalues of altered ALFF and FC extracted by using the eigenvariate option in SPM8.

Differences in behavioral measures and executive function tests between patients with ADHD and healthy con-trol subjects were examined by using the two-sample t test in software (SPSS 16.0; SPSS).

Results

Behavioral Measures and Executive Function TestsDemographic variables, including age (P = .246), sex (male), and handedness (right handed), were not significantly different between patients with ADHD and healthy control subjects.

Results of behavior measures and executive function tests are shown in Table 2. Patients with ADHD had high-er attention problem scores from the Child Behavior Checklist and the hyper-activity index from the revised Conners’ Parent Rating Scale than control sub-jects (P , .001). Furthermore, patients with ADHD performed much worse than control subjects on executive

applied with an 8-mm full-width-at-half-maximum Gaussian kernel.

Statistical AnalysisImage preprocessing and statistical analyses were performed by three au-thors (F.L., N.H., and S.L., with 6, 4, and 11 years of experience in MR re-search, respectively). Three-dimension-al T1-weighted MR data were analyzed by using a toolbox (VBM8), implement-ed in SPM8, as described in Appendix E1 (online). By using the two-sample t test, there were no significant differ-ences (P = .89) in total brain volume between patients with ADHD (1437 mm3 6 137) and healthy control sub-jects (1432 mm3 6 126), but we in-cluded total brain volume as a covariate of no interest in ALFF and FC analyses to ensure that their variability did not underlie the observed effects.

The analyses of ALFF and FC dif-ferences between patients with ADHD and healthy control subjects were per-formed voxel by voxel by using a two-sample t test in SPM8, with age and total brain volume as covariates, taking a significance threshold of P = .05, cor-rected for multiple comparisons with familywise error correction at the clus-ter level. The two-sample t test with covariates of no interest violates the assumption of sphericity, and different variances (heteroscedasticity) induce different error covariance components that are estimated by using restricted maximum likelihood in SPM8. The peak voxel coordinates with the high-est significance within the brain areas

transform length, shortest) without bandpass filtering. The power spectrum was obtained and transformed to the square root and then averaged across 0.01–0.08 Hz at each voxel. This aver-aged square root was taken as the ALFF for each voxel. Then, with the standard-ization procedure, the ALFF of each voxel was divided by the global mean ALFF value of the individual (analogous to approaches used in positron emis-sion tomography [25]). Finally, spatial smoothing was applied with an 8-mm full-width-at-half-maximum Gaussian kernel.

Functional ConnectivityAs reported below, significant ALFF ab-normalities in patients with ADHD com-pared with control subjects were dem-onstrated in five brain regions (Table 1, Fig 1), and these five regions were used as seeds for FC analysis. After bandpass filtering (0.01–0.08 Hz) and linear-trend removal, a reference time series for each seed was extracted by averaging the resting-state functional MR imaging time series of voxels within each seed. Correlation analysis was performed voxel by voxel between each of these five time series and the filtered time series in the rest of the brain. Nuisance covari-ates, including head motion parameters, global mean signal intensity, white mat-ter signal intensity, and cerebrospinal fluid signal intensity, were regressed out. The correlation coefficients in each voxel were transformed to z scores using the Fisher r-to-z transformation to improve normality. Finally, spatial smoothing was

Table 1

Significant Differences of Regional ALFF between Patients with ADHD and Healthy Control Subjects

Side Brain Region

Talairach CoordinatesNo. of Clusters

Analysis P Value* x y z

Regions showing increased ALFF in ADHD patients relative to healthy control subjects Right Globus pallidus 18 6 2 64 .002 Left Globus pallidus 212 6 2 56 .004 Right Dorsal superior frontal gyrus, BA9 21 31 34 29 .025Regions showing decreased ALFF in ADHD patients relative to healthy control subjects Left Orbitofrontal cortex, BA11 224 46 212 32 .004 Left Ventral superior frontal gyrus, BA10 218 70 2 49 .003

* P , .05 with familywise error correction at cluster level, with age and total brain volume as covariates.

518 radiology.rsna.org n Radiology: Volume 272: Number 2—August 2014

NEURORADIOLOGY: Brain Abnormalities in Attention Deficit Hyperactivity Disorder Li et al

the five seed areas between the two groups, including both positive and neg-ative connectivity: The results (P , .05, familywise error corrected) are shown in Table 3 and Figure 2.

With the seeds placed in the bilat-eral globus pallidus, where higher ALFF was detected, patients with ADHD showed significantly higher FC between the seeds and bilateral orbitofrontal cortex (BA47) than control subjects; with the seed placed in the right dor-sal superior frontal gyrus (BA9), where

area is positive) mainly involved the ex-ecutive control and salience network, including prefrontal cortex and parietal cortex, insula, and subcortical regions; by contrast, the negative FC patterns (ie, correlation coeffiient r between the time series of the seed and connected area is negative) mainly involved the posterior brain regions, including the default mode network (Figs E1, E2 [online]).

Two-sample t tests were performed to investigate the differences of FC of

function tests: In the WCST, patients with ADHD achieved fewer total correct numbers (P = .005) and completed cat-egories (P = .036) and made more total errors (P = .016), perseverative errors (P = .016), and nonperseverative errors (P = .039); in the Stroop Color-Word Test, patients with ADHD scored fewer right numbers (P , .001) and more er-ror numbers (P , .001) and correction numbers (P , .001) and longer total time (P , .001) than control subjects.

Regional Neural FunctionResults of ALFF analysis are shown in Table 1 and Figure 1. Relative to control subjects, patients with ADHD showed higher ALFF in the right globus pal-lidus (P = .002), left globus pallidus (P = .004), and right dorsal superior frontal gyrus (Brodmann area [BA] 9) (P = .025) and lower ALFF in the left orbitofrontal cortex (BA11) (P = .004) and left ventral superior frontal gyrus (BA10) (P = .003) (familywise error corrected). Figure E3 (online) details the distribution of ALFF.

Neural Network FCThe within-group FC patterns of the five seed areas with altered ALFF were revealed by one-sample t tests; details are in Appendix E1 (online). In both pa-tients with ADHD and control subjects, the positive resting-state FC patterns (ie, correlation coefficient r between the time series of the seed and connected

Figure 1

Figure 1: Parametric maps superimposed on the three-dimensional T1-weighted template images from software (MRIcro, version 1.39, Chris Rorden’s Neuropsychology Laboratory, University of South Carolina, Columbia, SC; http://www.mricro.com) show ALFF differences between patients with ADHD and control subjects. Regions that show increased (warm color [red, orange, yellow]) and decreased (cool color [white, light blue, dark blue]) ALFF values in patients with ADHD relative to healthy control subjects (P , .05, with familywise error correction) are depicted. At right, color bars indicate t values from global voxel-based two-sample t test analysis.

Table 2

Behavioral Measures and Executive Function Tests in Patients with ADHD and Control Subjects

Characteristics Patients with ADHD Healthy Control Subjects P Value

Child Behavior Checklist attention problem scores

8.6 (3.6) 3.4 (3.5) ,.001

Revised Conners’ Parent Rating Scale hyperactivity index

13.2 (6.9) 4.8 (4.2) ,.001

WCST Total correct 28.5 (11.1) 35.7 (7.4) .005 Total errors 16.9 (13.5) 9.3 (8.3) .016 Perseverative errors 6.0 (7.7) 2.2 (2.8) .016 Nonperseverative errors 10.9 (8.1) 7.2 (6.1) .039 Categories completed 4.2 (2.1) 5.4 (1.8) .036Stroop Color-Word Test Right numbers 95.3 (11.5) 105.3 (6.6) ,.001 Error numbers 15.4 (9.9) 5.1 (6.4) ,.001 Correction numbers 9.2 (6.4) 2.6 (1.8) ,.001 Total time (sec) 266.2 (104.6) 160.0 (39.7) ,.001

Note.—Data are means. Numbers in parentheses are standard deviations.

Radiology: Volume 272: Number 2—August 2014 n radiology.rsna.org 519

NEURORADIOLOGY: Brain Abnormalities in Attention Deficit Hyperactivity Disorder Li et al

Table 3

Significant Differences of FC between Patients with ADHD and Healthy Control Subjects

Seed Connected Area

Talairach Coordinates

No. of ClustersAnalysis P Value*x y z

Regions with increased FC in ADHD patients relative to healthy control subjects

Right globus pallidus Left orbitofrontal cortex, BA47 221 16 221 15 .037 Right globus pallidus Right orbitofrontal cortex, BA47 15 11 221 20 .031 Left globus pallidus Left orbitofrontal cortex, BA47 221 16 221 38 .022 Left globus pallidus Right orbitofrontal cortex, BA47 15 11 221 14 .043 Right dorsal superior frontal gyrus, BA9 Left ventral medial prefrontal cortex, BA10 29 61 26 31 .007 Left orbitofrontal cortex, BA11 Left superior frontal gyrus, BA8 29 35 51 87 .005 Left orbitofrontal cortex, BA11 Right superior frontal gyrus, BA8 9 37 51 42 .039 Left ventral superior frontal gyrus, BA10 Left superior frontal gyrus, BA10 234 47 11 48 .038 Left ventral superior frontal gyrus, BA10 Right superior frontal gyrus, BA10 15 50 14 45 .018 Left ventral superior frontal gyrus, BA10 Left orbitofrontal cortex, BA11 219 44 218 83 .01 Left ventral superior frontal gyrus, BA10 Right orbitofrontal cortex, BA11 23 43 216 63 .034Regions with decreased FC in ADHD patients

relative to healthy control subjects Right dorsal superior frontal gyrus, BA9 Right superior temporal gyrus, BA42 62 23 6 73 .008 Right dorsal superior frontal gyrus, BA9 Left precentral gyrus, BA4 236 29 58 212 ,.001 Right dorsal superior frontal gyrus, BA9 Right precentral gyrus, BA4 33 218 45 191 .001 Right dorsal superior frontal gyrus, BA9 Left supramarginal gyrus, BA40 245 228 24 100 .005 Right dorsal superior frontal gyrus, BA9 Right supramarginal gyrus, BA40 62 222 18 86 .01 Right dorsal superior frontal gyrus, BA9 Left putamen 230 211 3 70 .003 Right dorsal superior frontal gyrus, BA9 Right putamen 30 212 22 21 .009 Left orbitofrontal cortex, BA11 Left cerebellum 239 253 220 27 .018 Left ventral superior frontal gyrus, BA10 Left supramarginal gyrus, BA40 236 244 40 45 .023 Left ventral superior frontal gyrus, BA10 Right angular gyrus, BA40 39 256 39 26 .011

* P , .05, with familywise error correction at cluster level, with age and total brain volume as covariates.

higher ALFF was detected, patients with ADHD showed significantly higher FC with the left ventral medial prefron-tal cortex (BA10) and significantly lower FC with the bilateral precentral gyrus (BA4), supramarginal gyrus (BA40) and putamen, and right superior temporal gyrus (BA42); with the seed placed in the left orbitofrontal cortex (BA11), where lower ALFF was detected, pa-tients with ADHD showed significantly higher FC with bilateral superior frontal gyrus (BA8) and significantly lower FC with the left cerebellum; last, with the seed placed in the left ventral superior frontal gyrus, where lower ALFF was detected, patients with ADHD showed significantly higher FC with bilateral su-perior frontal gyrus (BA10) and orbito-frontal cortex (BA11) and significantly lower FC with left supramarginal gyrus (BA40) and right angular gyrus (BA40).

Correlational AnalysisFor the patients with ADHD, the higher FC between the right globus pallidus and the left orbitofrontal cortex (BA47) was positively correlated with total errors (r = 0.37, P = .036) of the WCST and neg-atively correlated with the total correct (r = 20.36, P = .046) and completed categories (r = 20.35, P = .048) of the WCST; the higher FC between the right globus pallidus and the right orbitofron-tal cortex (BA47) was positively corre-lated with error numbers (r = 0.36, P = .045) and negatively correlated with right numbers (r = 20.39, P = .026) of the Stroop Color-Word Test. The higher FC between the left globus pallidus and the left orbitofrontal cortex (BA47) was positively correlated with total errors (r = 0.39, P = .026) of the WCST and negatively correlated with the total correct (r = 20.42, P = .018) and the

completed categories (r = 20.4, P = .024) of the WCST; the higher FC be-tween the left globus pallidus and right orbitofrontal cortex (BA47) was nega-tively correlated with the total correct (r = 20.35, P = .048) of the WCST and right numbers (r = 20.39, P = .029) of the Stroop Color-Word test. The lower FC between the right dorsal superior frontal gyrus and the left putamen was negatively correlated with correction numbers (r = 20.36, P = .042) of the Stroop Color-Word Test; the lower FC between the right dorsal superior frontal gyrus and right putamen was negatively correlated with correction numbers (r = 20.38, P = .034) of the Stroop Color-Word Test. Details for the correlations are depicted in the scatter diagrams in Figure E4 (online).

There were no significant correla-tions between altered ALFF values and

520 radiology.rsna.org n Radiology: Volume 272: Number 2—August 2014

NEURORADIOLOGY: Brain Abnormalities in Attention Deficit Hyperactivity Disorder Li et al

was consistent with these earlier find-ings (34–36).

Recently, the pathologic focus of ADHD has shifted from regional brain abnormalities to dysfunctions in dis-tributed networks (5). Stimulant medi-cation not only modulates regional acti-vation but also normalizes dysfunctional connectivity in ADHD (37). The present study revealed altered FC in large-scale networks, extending the established frontostriatal model of ADHD. Fur-thermore, in normal maturation, local connectivity decreases while long-range connectivity increases (38–41), raising the possibility that both the increased FC within local prefrontal lobe and striatum and the decreased long-range connectivity in frontoparietal and fron-tocerebellar networks that we observed in ADHD might reflect delayed or dys-functional maturation.

Investigators in previous studies in ADHD found deficits in FC within the frontostriatal network (42) and in-creases in nodal efficiency in the ven-tral frontal gyrus and globus pallidus (43). The increased FC we observed in the frontostriatal circuit is consis-tent with diffusion-tensor imaging evi-dence of increased frontostriatal white matter connectivity in ADHD (44,45).

in motivation, reward-related executive function, and social disinhibition and impulse control (28,29). Consistent with this finding, the decreased ALFF in the left orbitofrontal cortex and the left ventral superior frontal gyrus in ADHD in the present study might also suggest prefrontal hypofunction.

Additional abnormalities in the frontostriatal circuit in children with ADHD include reduced volume of the globus pallidus (30,31) and decreased N-acetylaspartate–creatine ratio (32). In the present study, the apparent con-tradiction between hypofrontality and the increased ALFF in the bilateral globus pallidus suggests a pattern of compensatory hypo- and hyperfunction within the frontostriatal network.

The dorsal superior frontal gyrus mediates executive function in the fron-toparietal network (28), abnormalities that have been suggested as part of the pathophysiologic process in ADHD (33). Researchers in previous studies (34,35) found increased microstruc-tural integrity of the right superior frontal white matter in children with ADHD, correlated with the executive dysfunction and clinical symptoms. The present finding of increased ALFF of the right dorsal superior frontal gyrus

any characteristics of the WCST and the Stroop Color-Word test in patients with ADHD.

Discussion

The present study characterized alter-ations of distributed brain function at resting state in children and adolescents with ADHD. Compared with healthy control subjects, patients with ADHD showed higher ALFF in bilateral glo-bus pallidus and right dorsal superior frontal gyrus and lower ALFF in the left orbitofrontal cortex and ventral supe-rior frontal gyrus, as well as increased FC within the frontostriatal circuit and decreased long-range connectivity in the frontoparietal and frontocerebellar networks. The increased frontostriatal connectivity was correlated with ex-ecutive dysfunction assessed by using the WCST and the Stroop Color-Word Test, pointing to a connectivity-based pathophysiologic process in ADHD.

Abnormalities in the prefrontostria-tal circuit are an important correlate of ADHD (5). Structural MR imaging depicts decreased cortical thickness across orbital and ventral frontal cortex in children with ADHD (26,27), areas whose hypofunction underpins deficits

Figure 2

Figure 2: Anatomic replicas show differences of FC between patients with ADHD and healthy control subjects. The red nodes represent the seed areas of FC. The yellow nodes and orange lines and blue nodes and green lines represent, respectively, increased and decreased FC in patients with ADHD relative to healthy control subjects. Left: BA8.L = BA8, left; BA8.R = BA8, right; BA10.L = BA10, left; BA10.R = BA10, right; dSFG.R = dorsal superior frontal gyrus, right; GP.L = globus pallidus, left; GP.R = globus pallidus, right; OFC.L = orbitofrontal cortex, left; OFC.R = orbitofrontal cortex, right; vMPFC.L = ventral medial prefrontal cortex, left; vSFG.L = ventral superior frontal gyrus, left. Right: AG.R = angular gyrus, right; CER.L = cerebellum, left; dSFG.R = dorsal superior frontal gyrus, right; OFC.L = orbitofrontal cortex, left; PG.L = precentral gyrus, left; PG.R = precentral gyrus, right; PUT.L = putamen, left; PUT.R = putamen, right; SG.L = supramarginal gyrus, left; SG.R = supramarginal gyrus, right; STG.R = superior temporal gyrus, right; vSFG.L = ventral superior frontal gyrus, left.

Radiology: Volume 272: Number 2—August 2014 n radiology.rsna.org 521

NEURORADIOLOGY: Brain Abnormalities in Attention Deficit Hyperactivity Disorder Li et al

the relatively low sampling rate (repe-tition time, 2 seconds) for multisection acquisitions. Simultaneous recording of heart rate, respiratory rate, and depth during resting-state functional MR im-aging might help further reduce phys-iologic noise artifacts (61). Last, the mixed subtypes in our ADHD group might have confounded the results, and future studies could usefully focus on these.

In summary, we found that chil-dren and adolescents with ADHD have altered regional brain function and ab-errant FC in large-scale networks that were associated with executive dysfunc-tion, suggesting that the characteristics of the brain’s resting-state functional architecture are relevant to understand-ing relationships between neural sub-strate and executive function in ADHD.

Disclosures of Conflicts of Interest: F.L. No relevant conflicts of interest to disclose. N.H. No relevant conflicts of interest to disclose. Y.L. No relevant conflicts of interest to disclose. L.C. No relevant conflicts of interest to dis-close. X.H. No relevant conflicts of interest to disclose. S.L. No relevant conflicts of interest to disclose. L.G. No relevant conflicts of inter-est to disclose. G.J.K. No relevant conflicts of interest to disclose. Q.G. No relevant conflicts of interest to disclose.

References 1. Polanczyk G, de Lima MS, Horta BL, Bie-

derman J, Rohde LA. The worldwide prev-alence of ADHD: a systematic review and metaregression analysis. Am J Psychiatry 2007;164(6):942–948.

2. Biederman J, Faraone SV. Attention-deficit hyperactivity disorder. Lancet 2005;366(9481):237–248.

3. Rubia K, Overmeyer S, Taylor E, et al. Hy-pofrontality in attention deficit hyperactivity disorder during higher-order motor control: a study with functional MRI. Am J Psychia-try 1999;156(6):891–896.

4. Vaidya CJ, Austin G, Kirkorian G, et al. Se-lective effects of methylphenidate in atten-tion deficit hyperactivity disorder: a func-tional magnetic resonance study. Proc Natl Acad Sci U S A 1998;95(24):14494–14499.

5. Castellanos FX, Proal E. Large-scale brain systems in ADHD: beyond the pre-frontal-striatal model. Trends Cogn Sci 2012;16(1):17–26.

6. Biswal B, Yetkin FZ, Haughton VM, Hyde JS. Functional connectivity in the motor cortex

suggesting that dysconnectivity of the dorsal frontostriatal circuit may lead to impairments in executive functions in ADHD (51). Furthermore, because the frontoparietal network interacts with areas such as the superior tem-poral gyrus and the sensorimotor cortex to support cognitive-motor processing (28), the abnormalities of the superior temporal gyrus (char-acterized by the decreased temporal cortex thickness (52) and white mat-ter abnormalities in the frontotem-poral circuit [53]) may contribute to the symptoms of ADHD. The present study extends this factor by adding that the frontotemporal dysconnectiv-ity might also be involved in the path-ophysiologic process of ADHD.

The cerebellum, once considered as principally involved in motor con-trol and coordination, is implicated in a range of neuropsychiatric condi-tions (54). Researchers in a structural meta-analysis (55) have reported re-duced cerebellar volume in children with ADHD, and children with ADHD have reduced white matter connectiv-ity in the frontocerebellar tract (37). Supporting this idea, we observed reduced FC between the right orbi-tofrontal cortex and cerebellum in ADHD. These abnormalities of in-terregional FC in long-range brain networks indicate more widespread brain alterations in ADHD.

This study has some limitations. Global signal regression, a common yet controversial technique in resting-state functional MR image processing, is known to increase negative correla-tions (56) and may distort group dif-ferences in intrinsic FC (57). Neverthe-less, global signal regression correction is often used to account for several potential sources of physiologic noise and is effective in reducing the effects of head motion (58,59). Because of the hyperactivity of children with ADHD, we, like others (60), used global signal regression as part of our seed-based FC analysis. Although we temporally bandpass-filtered all resting-state func-tional MR imaging data (0.01–0.08 Hz), noise of cardiac and respiratory origin could not be completely filtered, given

Investigators in behavioral studies com-monly report impairment in set shift-ing measured by using the WCST (46) and in selective attention and inhibi-tory control measured by the Stroop Color-Word Test (47), which have been linked with the dysfunction of prefron-tal cortex and striatum in children and adolescents with ADHD (28,48). In the present work, we have directly ad-dressed this issue by investigating the FC correlates of executive function. Our finding of altered frontostriatal FC correlated with the central factors of executive function (such as cognitive flexibility reflected by the results of the WCST and interference inhibition pro-cessing assessed by using the Stroop Color-Word Test) sheds light on the neural substrate for executive control, in accordance with the speculation that the FC deficits within the frontostria-tal network might have a direct effect on ADHD-related executive dysfunction (42).

Increased frontostriatal local con-nectivity combined with long-range dysconnectivity in ADHD suggests a shift of the regular networks, which has attracted interest as a potential substrate of the disorder (5,42). There are several reports of decreased cor-tical thickness in the inferior parietal lobule, including the supramarginal gy-rus and the angular gyrus (26,49), and hypoactivation in frontoparietal circuits during attention control (33) in ADHD. Consistent with diffusion-tensor im-aging studies demonstrating reduced white matter structural integrity and connectivity in the frontoparietal tract connecting inferior parietal cortex and dorsolateral prefrontal cortex (37), we observed decreased FC between the superior frontal gyrus and the bilateral inferior parietal lobule in ADHD. The accumulating evidence suggests that ab-normalities in the frontoparietal circuit have a role in ADHD.

We also found decreased FC be-tween the right superior frontal gyrus and the bilateral putamen, associated with executive dysfunction of inter-ference inhibition, consistent with the results of a previous resting-state functional MR imaging study (50),

522 radiology.rsna.org n Radiology: Volume 272: Number 2—August 2014

NEURORADIOLOGY: Brain Abnormalities in Attention Deficit Hyperactivity Disorder Li et al

ADHD: voxel-based meta-analysis exploring the effects of age and stimulant medication. Am J Psychiatry 2011;168(11):1154–1163.

31. Frodl T, Skokauskas N. Meta-analysis of structural MRI studies in children and adults with attention deficit hyperactivity disorder indicates treatment effects. Acta Psychiatr Scand 2012;125(2):114–126.

32. Jin Z, Zang YF, Zeng YW, Zhang L, Wang YF. Striatal neuronal loss or dysfunction and choline rise in children with attention-deficit hyperactivity disorder: a 1H-magnetic reso-nance spectroscopy study. Neurosci Lett 2001;315(1-2):45–48.

33. Rubia K. “Cool” inferior frontostriatal dys-function in attention-deficit/hyperactivity disorder versus “hot” ventromedial orbito-frontal-limbic dysfunction in conduct disor-der: a review. Biol Psychiatry 2011;69(12): e69–e87.

34. Peterson DJ, Ryan M, Rimrodt SL, et al. Increased regional fractional anisotropy in highly screened attention-deficit hyperactiv-ity disorder (ADHD). J Child Neurol 2011; 26(10):1296–1302.

35. Li Q, Sun J, Guo L, et al. Increased frac-tional anisotropy in white matter of the right frontal region in children with atten-tion-deficit/hyperactivity disorder: a diffu-sion tensor imaging study. Neuroendocrinol Lett 2010;31(6):747–753.

36. Yang H, Wu QZ, Guo LT, et al. Abnormal spontaneous brain activity in medication-naïve ADHD children: a resting state fMRI study. Neurosci Lett 2011;502(2):89–93.

37. Konrad K, Eickhoff SB. Is the ADHD brain wired differently? a review on structural and functional connectivity in attention defi-cit hyperactivity disorder. Hum Brain Mapp 2010;31(6):904–916.

38. Fair DA, Posner J, Nagel BJ, et al. Atypical default network connectivity in youth with attention-deficit/hyperactivity disorder. Biol Psychiatry 2010;68(12):1084–1091.

39. Fair DA, Cohen AL, Dosenbach NU, et al. The maturing architecture of the brain’s default network. Proc Natl Acad Sci U S A 2008;105(10):4028–4032.

40. Fair DA, Cohen AL, Power JD, et al. Functional brain networks develop from a “local to distributed” organization. PLOS Comput Biol 2009;5(5):e1000381. http://www.ploscompbiol.org/article/info%3Adoi%2F10.1371%2Fjournal .pcbi.1000381. Published May 1, 2009. Ac-cessed November 9, 2009.

41. Kelly AM, Di Martino A, Uddin LQ, et al. Development of anterior cingulate functional connectivity from late childhood to early

21. Chao-Gan Y, Yu-Feng Z. DPARSF: A MAT-LAB toolbox for “pipeline” data analysis of resting-state fMRI. Front Syst Neurosci 2010;4:13. eCollection 2010. http://journal .frontiersin.org/Journal/10.3389/fnsys .2010.00013/full. Published May 14, 2010. Accessed July 10, 2010.

22. Burgund ED, Kang HC, Kelly JE, et al. The feasibility of a common stereotactic space for children and adults in fMRI studies of development. Neuroimage 2002;17(1): 184–200.

23. Kang HC, Burgund ED, Lugar HM, Petersen SE, Schlaggar BL. Comparison of functional activation foci in children and adults using a common stereotactic space. Neuroimage 2003;19(1):16–28.

24. Lui S, Li T, Deng W, et al. Short-term ef-fects of antipsychotic treatment on cerebral function in drug-naive first-episode schizo-phrenia revealed by “resting state” func-tional magnetic resonance imaging. Arch Gen Psychiatry 2010;67(8):783–792.

25. Raichle ME, MacLeod AM, Snyder AZ, Powers WJ, Gusnard DA, Shulman GL. A default mode of brain function. Proc Natl Acad Sci U S A 2001;98(2):676–682.

26. McLaughlin KA, Sheridan MA, Winter W, Fox NA, Zeanah CH, Nelson CA. Wide-spread reductions in cortical thickness following severe early-life deprivation: a neurodevelopmental pathway to attention-deficit/hyperactivity disorder. Biol Psychi-atry doi:10.1016/j.biopsych.2013.08.016. http://www.sciencedirect.com/science/article/pii/S0006322313007695?via=ihub. Published September 30, 2013. Accessed January 10, 2014.

27. Hoekzema E, Carmona S, Ramos-Quiroga JA, et al. Laminar thickness alterations in the fronto-parietal cortical mantle of pa-tients with attention-deficit/hyperactivity disorder. PLoS ONE 2012;7(12):e48286. http://www.plosone.org/article/info%3Adoi %2F10.1371%2Fjournal.pone.0048286. Published December 11, 2012. Accessed March 1, 2013.

28. Bush G. Attention-deficit/hyperactivity dis-order and attention networks. Neuropsy-chopharmacology 2010;35(1):278–300.

29. Cubillo A, Halari R, Smith A, Taylor E, Ru-bia K. A review of fronto-striatal and fronto-cortical brain abnormalities in children and adults with attention deficit hyperactivity disorder (ADHD) and new evidence for dys-function in adults with ADHD during mo-tivation and attention. Cortex 2012;48(2): 194–215.

30. Nakao T, Radua J, Rubia K, Mataix-Cols D. Gray matter volume abnormalities in

of resting human brain using echo-planar MRI. Magn Reson Med 1995;34(4):537–541.

7. Logothetis NK, Pauls J, Augath M, Trinath T, Oeltermann A. Neurophysiological inves-tigation of the basis of the fMRI signal. Na-ture 2001;412(6843):150–157.

8. Fox MD, Raichle ME. Spontaneous fluctu-ations in brain activity observed with func-tional magnetic resonance imaging. Nat Rev Neurosci 2007;8(9):700–711.

9. Vincent JL, Kahn I, Snyder AZ, Raichle ME, Buckner RL. Evidence for a frontoparietal control system revealed by intrinsic functional connectivity. J Neurophysiol 2008;100(6): 3328–3342.

10. Lui S, Deng W, Huang X, et al. Association of cerebral deficits with clinical symptoms in an-tipsychotic-naive first-episode schizophrenia: an optimized voxel-based morphometry and resting state functional connectivity study. Am J Psychiatry 2009;166(2):196–205.

11. Chabernaud C, Mennes M, Kelly C, et al. Dimensional brain-behavior relationships in children with attention-deficit/hyperactivity disorder. Biol Psychiatry 2012;71(5):434–442.

12. Kipp K. A developmental perspective on the measurement of cognitive deficits in attention-deficit/hyperactivity disorder. Biol Psychiatry 2005;57(11):1256–1260.

13. Nelson HE. A modified card sorting test sensi-tive to frontal lobe defects. Cortex 1976;12(4): 313–324.

14. Golden CJ. Identification of brain disorders by the Stroop Color and Word Test. J Clin Psychol 1976;32(3):654–658.

15. Dragovic M, Hammond G. A classification of handedness using the Annett Hand Prefer-ence Questionnaire. Br J Psychol 2007;98(pt 3):375–387.

16. Dai X, Gong Y. Wechsler intelligence scale for children, Chinese version. Hunan, Chi-na: Hunan Medical University, 1990.

17. Achenbach T. Manual for the Child Behavior Checklist/4-18 and 1991 Profile. Burlington, Vt: Department of Psychiatry, University of Vermont, 1991.

18. Conners CK. Clinical use of rating scales in diagnosis and treatment of attention-deficit/hyperactivity disorder. Pediatr Clin North Am 1999;46(5):857–870.

19. Nyhus E, Barceló F. The Wisconsin Card Sorting Test and the cognitive assessment of prefrontal executive functions: a critical update. Brain Cogn 2009;71(3):437–451.

20. Stroop JR. Studies of interference in serial verbal reactions. J Exp Psychol 1935;18(6): 643–662.

Radiology: Volume 272: Number 2—August 2014 n radiology.rsna.org 523

NEURORADIOLOGY: Brain Abnormalities in Attention Deficit Hyperactivity Disorder Li et al

adulthood. Cereb Cortex 2009;19(3):640–657.

42. Liston C, Malter Cohen M, Teslovich T, Lev-enson D, Casey BJ. Atypical prefrontal con-nectivity in attention-deficit/hyperactivity disorder: pathway to disease or pathologi-cal end point? Biol Psychiatry 2011;69(12): 1168–1177.

43. Wang L, Zhu C, He Y, et al. Altered small-world brain functional networks in children with attention-deficit/hyperactivity disorder. Hum Brain Mapp 2009;30(2):638–649.

44. Silk TJ, Vance A, Rinehart N, Bradshaw JL, Cunnington R. White-matter abnormalities in attention deficit hyperactivity disorder: a diffusion tensor imaging study. Hum Brain Mapp 2009;30(9):2757–2765.

45. Tamm L, Barnea-Goraly N, Reiss AL. Dif-fusion tensor imaging reveals white matter abnormalities in Attention-Deficit/Hyperac-tivity Disorder. Psychiatry Res 2012;202(2): 150–154.

46. Willcutt EG, Doyle AE, Nigg JT, Faraone SV, Pennington BF. Validity of the executive function theory of attention-deficit/hyperac-tivity disorder: a meta-analytic review. Biol Psychiatry 2005;57(11):1336–1346.

47. Schwartz K, Verhaeghen P. ADHD and Stroop interference from age 9 to age 41 years: a meta-analysis of developmental ef-fects. Psychol Med 2008;38(11):1607–1616.

48. Castellanos FX, Sonuga-Barke EJ, Milham MP, Tannock R. Characterizing cognition

in ADHD: beyond executive dysfunction. Trends Cogn Sci 2006;10(3):117–123.

49. Narr KL, Woods RP, Lin J, et al. Wide-spread cortical thinning is a robust ana-tomical marker for attention-deficit/hyper-activity disorder. J Am Acad Child Adolesc Psychiatry 2009;48(10):1014–1022.

50. Cao X, Cao Q, Long X, et al. Abnormal rest-ing-state functional connectivity patterns of the putamen in medication-naïve children with attention deficit hyperactivity disorder. Brain Res 2009;1303:195–206.

51. Sonuga-Barke EJ. Causal models of attention-deficit/hyperactivity disorder: from common simple deficits to multiple developmental pathways. Biol Psychiatry 2005;57(11):1231– 1238.

52. Shaw P, Eckstrand K, Sharp W, et al. At-tention-deficit/hyperactivity disorder is char-acterized by a delay in cortical maturation. Proc Natl Acad Sci U S A 2007;104(49): 19649–19654.

53. van Ewijk H, Heslenfeld DJ, Zwiers MP, Buite-laar JK, Oosterlaan J. Diffusion tensor imag-ing in attention deficit/hyperactivity disorder: a systematic review and meta-analysis. Neuro-sci Biobehav Rev 2012;36(4):1093–1106.

54. O’Halloran CJ, Kinsella GJ, Storey E. The cerebellum and neuropsychological func-tioning: a critical review. J Clin Exp Neuro-psychol 2012;34(1):35–56.

55. Valera EM, Faraone SV, Murray KE, Seidman LJ. Meta-analysis of structural imaging find-

ings in attention-deficit/hyperactivity disor-der. Biol Psychiatry 2007;61(12):1361–1369.

56. Murphy K, Birn RM, Handwerker DA, Jones TB, Bandettini PA. The impact of global sig-nal regression on resting state correlations: are anti-correlated networks introduced? Neuroimage 2009;44(3):893–905.

57. Saad ZS, Gotts SJ, Murphy K, et al. Trouble at rest: how correlation patterns and group differences become distorted after global signal regression. Brain Connect 2012;2(1): 25–32.

58. Yan CG, Craddock RC, Zuo XN, Zang YF, Milham MP. Standardizing the intrinsic brain: towards robust measurement of inter-individual variation in 1000 functional con-nectomes. Neuroimage 2013;80:246–262.

59. Yan CG, Cheung B, Kelly C, et al. A com-prehensive assessment of regional variation in the impact of head micromovements on functional connectomics. Neuroimage 2013;76:183–201.

60. Beucke JC, Sepulcre J, Talukdar T, et al. Ab-normally high degree connectivity of the orbi-tofrontal cortex in obsessive-compulsive disor-der. JAMA Psychiatry 2013;70(6):619–629.

61. Birn RM, Diamond JB, Smith MA, Bandet-tini PA. Separating respiratory-variation-related fluctuations from neuronal-activity-related fluctuations in fMRI. Neuroimage 2006;31(4):1536–1548.

©RSNA, 2014

Appendix E1

Voxel-based Morphometric Analysis

Three-dimensional T1-weighted MR data were analyzed by using a toolbox (VBM8) developed

by Christian Gaser, MEE, PhD (Department of Psychiatry, University of Jena, Jena, Germany)

implemented in statistical parametric mapping software (SPM8; Wellcome Trust Centre for

Neuroimaging, University College London) with default parameters and bias corrected; then the

tissue was classified into gray and white matter and cerebrospinal fluid. Results of tissue

segmentation analyses were multiplied by the nonlinear components derived from the

normalization matrix to preserve actual modulated brain volumes. The resulting tissue partition

was spatially normalized to the Dartel template provided by the VBM8 toolbox by using linear

(12-parameter affine) transformation and nonlinear high-dimensional warping, within a unified

model (62), and then was smoothed by using a Gaussian kernel of 8-mm full width at half

maximum. There were no significant differences (P = .89) in total brain volume, as determined

by using a two-sample t test, between ADHD patients (1437 mm3

± 137) and healthy control

subjects (1432 mm3 ± 126). Using a two-sample t test in SPM8, we found no significant

differences in regional gray matter volume between the ADHD patients and control subjects (P >

.05, with familywise error correction with the same threshold as used for functional data). But we

included the total brain volume as covariate of no interest in ALFF and FC analyses to ensure

that their variability did not underlie the observed effects.

Resting-state FC Patterns of the Five Seeds That Showed Altered ALFF between the Two Groups

For both the patients with ADHD and the healthy control subjects, the within-group resting-state

FC z score maps were analyzed with a one-sample t test to identify voxels showing a significant

positive or negative correlation with the seed time series; the threshold for the correlations was P

< .05, with familywise error correction for multiple comparisons.

The resting-state FC patterns of the bilateral globus pallidus in the two groups were not

significantly different. With the seed located in the bilateral globus pallidus, both the patients

with ADHD and the healthy control subjects showed significant positive correlations between

the seeds and activity in the globus pallidus, the caudate, the putamen, the thalamus, the

temporopolar cortex, the insula, the operculum, the orbitofrontal cortex, the midcingulate cortex,

and the dorsomedial prefrontal cortex, all bilaterally; both groups also showed negative

correlations, with activity in the precuneus, the cuneus, and the visual cortex, all bilaterally (Fig

E1).

With the seed located in the right dorsal superior frontal gyrus, both the patients with

ADHD and the healthy control subjects showed significant positive correlations, with activity in

the bilateral dorsal superior frontal gyrus, the bilateral ventromedial prefrontal cortex, the

bilateral posterior and anterior cingulate cortex, the right middle temporal gyrus, and the right

angular gyrus; the patients with ADHD showed significant negative correlations, with activity in

the bilateral supplementary motor area, the bilateral postcentral and precentral gyrus, the bilateral

occipital lobe, the right cerebellum, and the bilateral supramarginal gyrus extending to bilateral

superior temporal gyrus, insula, and putamen; the healthy control subjects showed significant

negative correlations, with activity in the bilateral occipital lobe and the right cerebellum (Fig

E2).

With the seed located in left ventral superior frontal gyrus, both the patients withADHD

and the healthy control subjects showed significant positive correlations, with activity in the

ventral superior frontal gyrus, the orbitofrontal cortex, the posterior cingulate cortex, the inferior

temporal gyrus, and the angular gyrus, bilaterally; the patients with ADHD showed significant

negative correlations, with activity in the bilateral midcingulate cortex, the right middle frontal

gyrus, the bilateral supramarginal gyrus, the right lingual gyrus, and the left insula; the healthy

control subjects showed significant negative correlations, with activity in the bilateral

midcingulate cortex, the right middle frontal gyrus, the right supramarginal gyrus, the left lingual

gyrus, the bilateral precentral gyrus, and the bilateral insula (Fig E2).

With the seed located in the left orbitofrontal cortex, the patients with ADHD showed

significant positive correlations, with activity in the bilateral orbitofrontal cortex, the bilateral

dorsomedial prefrontal cortex, the left superior frontal gyrus, and the bilateral inferior parietal

lobule; the healthy control subjects showed significant positive correlations, with activity in the

bilateral orbitofrontal cortex; the patients with ADHD showed significant negative correlations,

with activity in the right supramarginal gyrus, the bilateral lingual gyrus, and the cerebellum; the

healthy control subjects showed significant negative correlations, with activity in the bilateral

lingual gyrus and the cerebellum (Fig E2).

Thus the positive resting-state FC patterns of both the patients with ADHD and the

healthy control subjects mainly involve the executive control and salience network, including the

prefrontal and parietal cortex, the insula, and the subcortical regions; by contrast, the negative

connectivity patterns mainly involve the posterior brain regions, including the default mode

network (Figs E1, E2). The salience network, which responds to behaviorally salient events,

consists of three main cortical areas: the dorsal anterior cingulate cortex, the left and anterior

right insula, and the adjacent inferior frontal gyrus (63,64). Cognitive control, or more formally

the ‘executive control function,’ includes several cognitive functions, such as working memory

and response selection and/or suppression (65). The executive control network consists of

regions across the prefrontal and parietal cortices and the corticosubcortical circuit, such as the

cingulate cortex, the dorsolateral and the inferior frontal cortex (involved in monitoring, conflict

processing, and response inhibition), the visual cortex, and the parietal lobes (associated with

top-down visuospatial control, especially of working memory), and also motor and subcortical

areas (9, 65,66).

Lack of Resting-state FC Difference in the Anterior Cingulate Cortex between the Two Groups

The within-group resting state analyses in this study revealed that the anterior cingulate cortex

had a significant positive FC with the globus pallidus and superior frontal gyrus in both the

ADHD and control groups (Figs E1, E2), consistent with the notion that the anterior cingulate

cortex comprises the main components of the cingulofrontoparietal executive-control network. In

apparent contrast, previous resting-state functional MR imaging studies in which the researchers

chose the anterior cingulate cortex as a seed to perform the FC analysis found decreased FC

between the anterior cingulate cortex and components in the default mode network in adults with

ADHD (67) and also in medicated children with ADHD (68). However, published studies have

yielded inconsistent results in ADHD, such as increased (69) and decreased (36) ALFF in the

anterior cingulate cortex, which may be attributable to small sample size (fewer than 20 patients

with ADHD in both studies [36,69]), medication exposure, and comorbidity. A resting-state

functional MR imaging study of 23 male children with ADHD who were not receiving drug

treatment and were noncomorbid found no abnormality of ALFF in the anterior cingulate cortex

(70), which is consistent with our result. Furthermore, our finding of no FC difference in the

anterior cingulate cortex between patients with ADHD and control subjects converges with

structural local alterations identified in ADHD (31): Meta-analysis reveals decreased anterior

cingulate cortex volume in adult patients with ADHD but not in children (31). Thus, a possible

explanation of our finding is that the intrinsic regional function of the anterior cingulate cortex is

simply not abnormal in children with ADHD. However, direct comparison is difficult between

our work and published FC analyses which used the anterior cingulate cortex as a seed, as we did

not.

References

62. Ashburner J, Friston KJ. Unified segmentation. Neuroimage 2005;26(3):839–851.

63. Seeley WW, Menon V, Schatzberg AF, et al. Dissociable intrinsic connectivity networks for

salience processing and executive control. J Neurosci 2007;27(9):2349–2356.

64. Ham T, Leff A, de Boissezon X, Joffe A, Sharp DJ. Cognitive control and the salience

network: an investigation of error processing and effective connectivity. J Neurosci

2013;33(16):7091–7098.

65. Coderre EL, van Heuven WJ. Modulations of the executive control network by stimulus

onset asynchrony in a Stroop task. BMC Neurosci 2013;14(1):79.

66. Menon V. Large-scale brain networks and psychopathology: a unifying triple network model.

Trends Cogn Sci 2011;15(10):483–506.

67. Castellanos FX, Margulies DS, Kelly C, et al. Cingulate-precuneus interactions: a new locus

of dysfunction in adult attention-deficit/hyperactivity disorder. Biol Psychiatry 2008;63(3):332–

337.

68. Sun L, Cao Q, Long X, et al. Abnormal functional connectivity between the anterior

cingulate and the default mode network in drug-naïve boys with attention deficit hyperactivity

disorder. Psychiatry Res 2012;201(2):120–127.

69. Zang YF, He Y, Zhu CZ, et al. Altered baseline brain activity in children with ADHD

revealed by resting-state functional MRI. Brain Dev 2007;29(2):83–91.

70. An L, Cao QJ, Sui MQ, et al. Local synchronization and amplitude of the fluctuation of

spontaneous brain activity in attention-deficit/hyperactivity disorder: a resting-state fMRI study.

Neurosci Bull 2013;29(5):603–613.