Reduced right frontal cortical thickness in children, adolescents and adults with ADHD and its correlation to clinical variables: A cross-sectional study

lable at ScienceDirect

Journal of Psychiatric Research 44 (2010) 1214e1223

Contents lists avai

Journal of Psychiatric Research

journal homepage: www.elsevier .com/locate/psychires

Reduced right frontal cortical thickness in children, adolescents and adultswith ADHD and its correlation to clinical variables: A cross-sectional study

Luis G. Almeida a,b,*, Josefina Ricardo-Garcell b, Hugo Prado a, Lázaro Barajas c,Antonio Fernández-Bouzas b, David Ávila b, Reyna B. Martínez a

aCentro Estatal de Salud Mental, Servicios de Salud del Estado de Querétaro, Avenida 5 de Febrero 105, Los Virreyes. C.P, 76170 Querétaro, Méxicob Instituto de Neurobiología, Universidad Nacional Autónoma de México, Campus Juriquilla. Boulevard Juriquilla 3001. C.P. 76230. Querétaro, Méxicoc Instituto Tecnológico de Monterrey, Campus Querétaro. Epigmenio González 500, Fraccionamiento San Pablo. C.P. 76130. Querétaro, México

a r t i c l e i n f o

Article history:Received 2 February 2010Received in revised form31 March 2010Accepted 20 April 2010

Keywords:ADHDMRICortical thicknessSuperior frontal gyrusSeverity of the illness

* Corresponding author. Centro Estatal de SalNeurosciences, Avenida 5 de Febrero 105, Los VirMéxico. Tel./fax: þ52 442 242 4381

E-mail address: [email protected] (L.G. Alm

0022-3956/$ e see front matter � 2010 Elsevier Ltd.doi:10.1016/j.jpsychires.2010.04.026

a b s t r a c t

Objective: Some longitudinal magnetic resonance imaging (MRI) studies have shown reduced volume orcortical thickness (CT) in the frontal cortices of individuals with attention-deficit/hyperactivity disorder(ADHD). These studies indicated that the aforementioned anatomical abnormalities disappear duringadolescence. In contrast, cross-sectional studies on adults with ADHD have shown anatomical abnor-malities in the frontal lobe region. It is not known whether the anatomical abnormalities in ADHD area delay or a deviation in the encephalic maturation. The aim of this study was to compare CT in thefrontal lobe of children, adolescents and adults of both genders presenting ADHD with that in corre-sponding healthy controls and to explore its relationship with the severity of the illness.Method: An MRI scan study was performed on never-medicated ADHD patients. Twenty-one children(6e10 year-olds), twenty adolescents (14e17 year-olds) and twenty adults (25e35 year-olds) werematched with healthy controls according to age and sex. CT measurements were performed using theFreesurfer image analysis suite.Results: The data showed regions in the right superior frontal gyrus where CT was reduced in children,adolescents and adults with ADHD in contrast to their respective healthy controls. The CT of theseregions correlated with the severity of the illness.Conclusions: In subjects with ADHD, there is a thinning of the cortical surface in the right frontal lobe,which is present in the children, adolescents and in adults.

� 2010 Elsevier Ltd. All rights reserved.

1. Introduction

Several published studies have used MRI (magnetic resonanceimaging) on subjects with attention-deficit/hyperactivity disorder(ADHD) (Valera et al., 2007). These studies have reported structuralabnormalities in individuals with ADHD, such as a decrease in the-volume of various encephalic structures or in the cortical thickness(CT) of certain cortical regions, particularly in the frontal lobe.(Hes-slinger et al., 2002,Castellanos et al., 2002,Shaw et al., 2007a;Makris et al., 2007). However, most of the studies have beencarried out on male children and few works have included adults.Additionally, female subjectshavebeenextremelyunderrepresented

ud Mental, Psychiatry andreyes. C.P, 76170 Querétaro,

eida).

All rights reserved.

in these studies (Valera et al., 2007; Ellison-Wright et al., 2008). Thisfact precludes the generalisation of the results derived from thosestudies to older subjects and females with ADHD. Consequently,further research that includes adults and female individuals isnecessary to create a broader picture of the developmental andgender characteristics of subjects with ADHD.(Valera et al., 2007).

Some longitudinal studies have found structural abnormalities,such as a decrease in volume or CT of several brain regions analysedin children with ADHD. Nearly all of these abnormalities persistedthrough adolescence, at which point they disappeared (Castellanoset al., 2002; Shaw et al., 2006, 2007a, 2007b). Nevertheless, theselongitudinal studies have several problems: they do not includesubjects older than 18 years and they do include subjectsmedicatedwith stimulants, which can affect CT measurements (Shaw et al.,2009). Furthermore, for a significant proportion of the subjects,the subsequent CTmeasurements were predicted based on oneMRIstudy instead of the actual values. This was due to a loss of subjectsduring the follow-up, a problem inherent in the longitudinal design.

L.G. Almeida et al. / Journal of Psychiatric Research 44 (2010) 1214e1223 1215

On the other hand, some studies have reported reduced volumeor CT in diverse encephalic regions in adults with ADHD (Hesslingeret al., 2002; Seidman et al., 2006; Makris et al., 2007; Biedermanet al., 2008). These divergent data raise the question: “Is ADHDa delay or a deviation in encephalic maturation?” Based on thesecontradictory data, we believe this issue remains unclear andtherefore warrants further research. On the other hand, the pres-ence of anatomical abnormalities in adults with ADHD probablyunderlies the clinical features observed in adults with ADHD (e.g.,only 10% of individuals with ADHD reach functional clinicalremission at age 20) (Biederman et al., 2000).

We chose to analyse the frontal lobe for several reasons. First, itis one of the cortical regions more frequently found to be smaller orthinner in its CT in subjects with ADHD (Castellanos et al., 2002;Shaw et al., 2006, 2007a, 2007b; Valera et al., 2007; Hesslingeret al., 2002; Seidman et al., 2006; Makris et al., 2007; Biedermanet al., 2008; Ellison-Wright et al., 2008). The frontal lobe is alsoone of the cortical regions most related to the physiology ofattention (Fan et al., 2005; Halperin and Schulz, 2006; Posner et al.,2006, Stahl, 2008), and it is one of the cortical regions thatconsistently showed abnormalities in studies using functionalmagnetic resonance imaging (fMRI) (Rubia et al., 2008, 2009). Ourgoal in analysing this cortical region was to determine whetherthere is a persistent structural abnormality in the different stages oflife in individuals with ADHD. Additionally, we opted for a cross-sectional design to avoid the inherent problems of the longitudinaldesign mentioned above.

2. Objective

To compare frontal CT between healthy individuals and indi-viduals with ADHD in three different age groups.

3. Method

This study included three age groups of never-medicatedsubjectswith combined typeADHD:21children (6e10year-olds),18adolescents (14e17-year-olds) and 20 adults (25e35-year-olds), aswell as three age groups of healthy individuals (22 children, 20adolescents and 20 adults). ADHD and control participants werematched by age, sex, height, weight, body mass index (BMI) and IQ(intellectual quotient). The diagnosis of ADHD was made accordingto the DSM-IV-TR (Diagnostic and Statistical Manual of MentalDisorders-Fourth Edition, Text Revision) (American PsychiatricAssociation, 2000) criteria using a checklist of symptoms. The MiniInternational Neuropsychiatric Structured Interview (MINI) wasalso used: Spanish version 4was applied to children (“MINI-Kid” forsubjects under 13 years of age) and version 5 (“MINI-Plus”) wasapplied to adolescents and adults (Sheehan et al., 1998). Comor-bidities were assessed with theMINI-Plus andMINI-Kid interviews.

The following individuals were excluded from the study: thosewho presented a history of current neurological or psychiatricdisease (judged by two experienced clinicians to cause moredistress or impairment to global function than ADHD); those withpredominantly inattentive or hyperactive/impulsive subtypes ofADHD; those showing evidence of general medical illness; thosewith IQ< 85 points; those with past or current use of substances orstimulants; and those with abnormalities in MRI scans. All of theMRI scans were evaluated by a certified neuroradiologist who wasunaware of the identity of the individual in question. The controland ADHD individuals were chosen from the general community inthe urban areas of Querétaro and Mexico City. Recruitmentoccurred by means of open invitation posters in public places andreferrals from public and private schools. The entire duration of thestudy spanned from March of 2006 to February of 2008. After

having fully explained the study to all participants, writteninformed consent was obtained (in accordance with general healthregulations inMexico). Certified clinicians carried out a full medicalhistory and physical examination on every participant. Examina-tions of children and adolescents were conducted by a paediatri-cian and neuropaediatrician; examinations of adults wereperformed by a certified internist and neurologist.

The Spanish version of the Edinburg Handedness Inventory(EHI) was used to assess handedness (Oldfield, 1971). A subject wasconsidered right-handed if his/her total score on the EHI was>þ40points. Subjects scoring <�40 points were considered left-handed,and those scoring between �40 and þ40 points were consideredambidextrous.

The Institutional Review Board (IRB) of the General Hospital ofQuerétaro, México, approved the protocol of the study and theinformed consent forms. There was no financial sponsorship fromany medical equipment or pharmaceutical industry, and there is noneed to report any conflicts of interest.

The appropriate age version of the MINI interview was admin-istered to all participants by two experienced clinicians (by twocertified child psychiatrists in the case of children and adolescentsand by two certified adult psychiatrists in the case of adults). The kfor the diagnosis of ADHD according the MINI interview wask ¼ 0.97 for children, k ¼ 0.93 for adolescents and k ¼ 0.93 foradults. The two clinicians who interviewed the subjects to make orto discard the diagnosis of ADHD were also responsible foradministering the MINI interview. The results of the diagnosismade by one clinician remained unknown to the other clinicianthroughout the entire study.

The Friederichsen, Almeida, Serrano, Cortes Test (FASCT) ratingscale (Almeida et al., 2006) was used to assess adult participants.This scale has two versions: self-reported (FASCT-SR) and observer-reported (FASCT-O). Participants completed the FASCT-SR them-selves, and the mothers of each participant completed the FASCT-O.Although the FASCT was developed to screen and measure theseverity of ADHD in Hispanic adults, the agreement kappa index forthe diagnosis of ADHD betweenMINI-plus and FASCT-SR score� 23pointswas k¼0.82. The sensitivity for ADHDdiagnosis of the FASCT-SR scale for individuals with a total score� 23 points was 80.36. Thespecificity index ¼ 97.99, the positive predictive value ¼ 93.75, andthe negative predictive value ¼ 92.99. The k for the diagnosis ofADHD between the MINI-plus and FASCT-O score � 23 points was0.88.The sensivity for the ADHD diagnosis of the FASCT-O scale forindividuals with a total score � 23 points was 95.45, the specificityvalue ¼ 96.39, positive predictive value ¼ 87.50, and negativepredictive value ¼ 98.77. More detailed information about theclinimetric properties of the FASCT-SR and FASCT-O scales havebeenpublished elsewhere (Almeida et al., 2006).

The Wechsler Intelligence Scale for Children, Third Edition(WISC-R) (Wechsler, 1981), was administered to individualsbetween the ages of 6 and 15 years. TheWechsler Adult IntelligenceScale, Third Edition (Wechsler, 1991), was administered to indi-viduals 16 years of age and older. The Conners’ Parent Rating Scale-Revised (CPRS-R) (Conners et al., 1998b) and the Conners’ TeacherRating Scale-Revised (CTRS-R) (Conners et al., 1998a) wereadministered to groups of children and adolescents. In addition,adolescent participants completed the Conners-Wells AdolescentSelf-Report Scale: Long Version (Conners et al.,1997).

3.1. MRI acquisition

The1.0 T PhilipsNew Intera release 10.3 (PhilipsMedical SystemsNetherlands) was used for all individuals. Fast field echo sequencewas used. A T1e3D volumetric pulse sequence produced 190continuous coronal slices of 1.0 mm thickness, with no gap in

L.G. Almeida et al. / Journal of Psychiatric Research 44 (2010) 1214e12231216

between, following theseparameters: echo time¼6.9ms, repetitiontime ¼ 25 ms, flip angle 30�, acquisition matrix ¼ 230 � 230 mm,field of view¼ 256mm and voxel size¼ 1.0 mm3. TheMRI scanwaslocated at the Instituto de Neurobiología de la Universidad NacionalAutónoma de México (U.N.A.M.), Juriquilla campus, Querétaro,México. There were no variations in the acquisition settings of theequipment, or in the use of different scanners, scanner upgrades, orpulse sequence changes. The acquisition time was 20e30 min perindividual. The clinimetric and behavioural data were obtained onthe same day as the MRI study. The MRI data were collected in aninterleaved fashion instead of by blocks. There was no equipmentcalibration during the MRI acquisition phase (lasting six months).

3.2. Image analysis

Cortical reconstruction and volumetric segmentation was per-formed with FreeSurfer Image Analysis Suite. The technical details ofthese procedures are described elsewhere (Dale et al., 1999, Dale andSereno, 1993; Fischl and Dale, 2000; Fischl et al., 2001, 2002, 2004a,1999a,b, 2004b; Han et al., 2006; Jovicich et al., 2006; Segonne et al.,2004). Described briefly, this processing includes motion correctionand averaging of multiple volumetric T1 weighted images; theremoval of non-brain tissue using a hybrid watershed/surface defor-mation procedure (Segonne et al., 2004), automated Talairach trans-formation; segmentation of the subcortical white matter and deepgray matter volumetric structures (Fischl et al., 2002 and Fischl et al.,2004a,b); intensitynormalisation (Sledetal.,1998); tessellationof thegraymatterewhitematter boundary; automated topology correction(Fischl et al., 2001, Segonne et al., 2007); and surface deformationfollowing intensity gradients to optimally place the gray/white andgray/cerebrospinal fluid borders at the location where the greatestshift in intensity defines the transition to the other tissue class (Daleet al., 1999, Dale and Sereno, 1993, Fischl and Dale, 2000). Once thecortical modelswere completed, a number of deformable procedurescouldbeperformedfor furtherdataprocessingandanalysis, includingsurface inflation (Fischl et al.,1999a,b), registration to a spherical atlasthat utilised individual cortical folding patterns to match corticalgeometry across subjects (Jovicich et al., 2006), parcellation of thecerebral cortex into parts based ongyral and sulcal structure (Desikanetal., 2006andFischletal., 2004a), andcreationofavarietyof surface-baseddata includingmapsof curvature and sulcaldepth. Thismethoduses both intensity and continuity information from the entire three-dimensional MRI volume in segmentation and deformation proce-dures to produce representations of CT, calculated as the distancefrom the gray/white boundary to the gray/CSF boundary at eachvertex on the tessellated surface (Fischl et al., 2004b). These proce-dures for the measurement of CT have been validated by histologicalanalyses (Rosas et al., 2002) and manual measurements (Kuperberget al., 2003 and Salat et al., 2004). Furthermore, the Freesurfer suiteproduces a table of statistical measures such as the intracranialvolume. Freesurfer morphometric procedures have been shown toproduce good testeretest reliability across scanner manufacturersand across magnetic field strengths (Han et al., 2006).

3.3. Statistical analysis

For the general data (e.g. gender, perinatal complication, schoolproblems,medical historyetc.), either theX2 test or Fisher’s exact testwas used to compare nominal variables between healthy and ADHDsubjects. Fisher’s exact test was used for nominal variables with anyexpected frequency (EF) of less than 2, or if more than the half of theEF was less than 5. For the general data t-tests, previous normalitytests (KolmogoroveSmirnov test with Lilliefors Significance Correc-tion and the ShapiroeWilk test) and a test for equality of variances(Levene’s test), were used to compare numerical variables (e.g., age

differences between healthy and ADHD subjects). In the case of non-normally distributed data, the ManneWhitney “U” test was used.

Pearson or Spearman correlation coefficients were used toassess the covariance between CT and the number of DSM-IV-TRcriteria for ADHD, as well as the covariance between total scores onthe FASCT- SR and FASCT-O rating scales.

Two-way ANOVA was conducted in order to reveal an interac-tion between age group and clinical status (Healthy or ADHD).

Stepwise linear regressionwas employed to determine whetherthe CT in the frontal lobe (where the General Linear Model (GLM)and the Monte Carlo test simulation bootstrap (MCTSB) (see below)showed differences between healthy and ADHD subjects) was pre-dicted by the diagnosis (coded as a dummy variable: 0 ¼ healthy,1 ¼ ADHD), controlling for the effects of gender (coded as dummyvariable 0 ¼ Female, 1 ¼ Male), age, body mass index (BMI), intra-cranial volume and the presence of comorbidities (coded asa dummy variable: 0 ¼ presence, 1 ¼ absence). This analysis wascarried out because we were comparing CT in different age groupsand with different patients in each group. This creates a lot of vari-ability in CT measures and can produce flawed statistical results.

The significance value of enter was set at 0.01. The Durbin Wat-son value was calculated to measure the independence of theresiduals, and a value between 1.5 and 2.5was considered adequate.The colinearity between the variables in the models of regressionwas measured by the auto-engine value and the condition index.Suitable values were placed at > 0.0 and <15.0, respectively. Allprobability (“p”) values were two-tailed. These analyses were con-ducted using the SPSS� version 17 statistical package.

The Freesurfer suite uses the GLM to compare CT betweengroups. The MCTSB was used to correct for the effect of multiplecomparisons and to prove the stability of the significance valueinitially given by the GLM. The MCTSB was thresholded at p < 0.05and 10,000 iterations. This procedure produces a colour-codedstatistical map (Fig. 2) that represents whole-brain differences in CTbetween healthy and ADHD subjects. Given our hypothesis aboutthe frontal lobe, we only analysed the frontal lobe and we did notexamined the occipital, temporal or parietal lobes.

After the MCTSB produced a map of statistical differencesbetween groups, the regions were bounded by hand and filled inusing the Qdec environment tools included in the Freesurfer suite.This was done to build a mask of each frontal regionwhere the GLMshowed significant differences between ADHD and healthysubjects. This mask was applied to each individual’s reconstructedbrain. This procedure calculated the following statistics for theregion of interest: area (mm2), gray matter volume (ml) andaverage CT (mm). t-tests were used to compare CT between healthyand ADHD subjects. (Fig. 1).

The effect sizes for the result of t test (Cohen’s d) were calculatedaccording the method published by Cohen (1988), Rosenthal andRosnow (1991), Rosnow and Rosenthal (1996).

4. Results

4.1. General characteristics of the sample

The characteristics of the subjects in the three age groups arepresented in Tables 1, 2 and 3. All subjects in all three groups wereright-handed (EHI score > þ40 points).

4.2. Comorbidities

Among the groups of children, the presence of other psychiatricdisorders was more frequent in ADHD individuals than in healthyindividuals: 13 (61.9%) vs. 2 (9.5%) (X2 ¼ 12.48, df ¼ 1, p ¼ 0.000).Two (9.5%) of the healthy children met the criteria for Separation

Fig. 1. Differences in cortical thickness between ADHD and healthy control individuals.

Table 1General, somatometric, sociodemographic, psychometric and clinimetric profile of the sample of children.

Healthy ADHD Statistical value df P value

n ¼ 22 n ¼ 21

Mean (�s.d) or n (%) Mean (�s.d) or n (%)

Age (years) 7.63 (�1.39) 7.61 (�1.28) MeW U ¼ 230, z ¼ �0.02 e 0.980Gender Males 10 (45.5%) Males 10 (47.6%) X2 ¼ 0.200 1 0.800

Females 12 (54.5%) Females 11 (52.4%)BMI 17.24 (�2.96) 17.23 (�2.44) MeWU ¼ 229, z ¼ �0.04 e 0.960Intracranial volume

(ml)1323.50 (�132.35) 1340.8 (�151.55) t ¼ �0.136 41 0.893

Right-handed 22 (100%) 21 (100%) e

Hispanic race 22 (100%) 21 (100%) e

Birth weight (g) 3211 (�96.85) 3315 (�104) t ¼ �0.72 41 0.470Perinatal complications 5 (23.8%) 6 (28.6%) X2 ¼ 0.123 1 0.726History of hospitalisationa 1 (4.8%) 0 (0%) Fisher’s exact test, p ¼ 1.00History of convulsive events 0 (0%) 0 (0%) e

School problemsb 1 (4.8%) 17 (81.0%) X2 ¼ 0.24.88 1 0.000Average academic

achievementc9.02 (�0.11) 7.80 (�0.36) t ¼ 3.01 41 0.004

DSM-IVd

Inattention 1.09 (�1.26) 8.38 (�0.920) MeW U ¼ 0.000, Z ¼ �5.738 e 0.000Hyperactivity 1.22 (�1.26) 4.66 (�1.23) MeW U ¼ 16.5, Z ¼ �5.270 0.000Impulsivity 0.54 (�0.73) 2.86 (�2.83) MeW U ¼ 18.0, Z ¼ �5.396 0.000

WISC-R total score 115.61 (�3.47) 107.23 (�3.03) t ¼ 1.87 41 0.077Parents GCIe 6.22 (�1.71) 19.31 (�1.52) t ¼ �6.50 41 0.000Teachers GCIf 6.33 (�5.36) 15.25 (�1.80) t ¼ �0.495 41 0.007Relatives with probable

ADHD2 (9.5%) 10 (47.6%) X2 ¼ 7.46 1 0.006

2693e6462 USDg 12 (54.5%) 12 (54.14%) X2 ¼ 0.29 1 0.864>6539 USD 10 (45.5%) 9 (42.90%)

MeWU ¼ ManneWhitney test. BMI ¼ body mass index.a Due to any medical cause.b History of school problems: Failed academic years, school expulsion, physical and verbal aggression towards peers and teachers.c Last academic year, scale 0e10.d Number of DSM-IV criteria for ADHD.e Parents GCI: Conner’s Global Index of the Conner’s Parent Rating Scale-Revised: Long Version (CPRS-R).f Teachers GCI: Conner’s Global Index of the Conner’s Teachers Rating Scale-Revised: Long Version (CTRS-R).g Average monthly family income in US dollars.

L.G. Almeida et al. / Journal of Psychiatric Research 44 (2010) 1214e1223 1217

L.G. Almeida et al. / Journal of Psychiatric Research 44 (2010) 1214e12231218

Anxiety Disorder. Among the children with ADHD, 1 (4.8%) childmet the criteria for Adjustment Disorder, 8 (38%) met the criteriafor Conduct Disorder, 11 (52.4%) met the criteria for OppositionalDefiant Disorder, and 1 (4.8%) met the criteria for GeneralisedAnxiety Disorder (GAD).

In the adolescent groups, the presence of other psychiatricdisorders was more frequent in ADHD individuals than in healthyindividuals: 9 (50%) vs.1 (5%) (X2¼ 9.89, df¼ 1, p¼ 0.002). One (5%)healthy adolescent met the criteria for Adjustment Disorder.Among the adolescents with ADHD, 2 (11.5%) individuals met thecriteria for Adjustment Disorder, 4 (22.2%) met the criteria forConduct Disorder, 6 (33.%) met the criteria for Oppositional DefiantDisorder, 2 (11%) met the criteria for GAD, and 1 (5.6%) met thecriteria for once having had Dysthymia.

In the adult groups, the presence of other psychiatric disorderswas again more frequent in ADHD individuals. Five (25%) ADHDparticipants met the criteria for other psychiatric disorders, asopposed to only 1 (5%) healthy participant, who met the criteria forGAD (Fisher’s exact test, p ¼ 0.091). In the ADHD group, 1 (5%) indi-vidual met the criteria for Premenstrual Dysphoric Disorder, 3 (15%)met the criteria for GAD, 3 (15%)met the criteria for having once hadMajorDepressiveDisorder (MDD),1 (15%)met thecriteria for SpecificPhobia, and 1 (15%) met the criteria for having once had Dysthymia.

4.3. Cortical thickness

CT differences were found in diverse regions of the frontal lobesof both hemispheres. When the MCTSB was applied, however, onlythe differences located in the right hemisphere reached statistical

Table 2General, somatometric, sociodemographic, psychometric and clinimetric profile of the sa

Healthy ADHD

n ¼ 20 n ¼ 18

Mean (�s.d) or n (%) Mean (�s.d) or

Age (years) 14.75 (�1.06) 15.44 (�1.24)Gender Males 10 (50%) Males 10 (55.6%

Females 10 (50%) Females 8 (44.4BMI 21.27 (�3.25) 22.42 (�3.05)Intracranial volume (ml) 1530 (�150) 1521 (�226)Right-handed 20 (100%) 18 (100%)Hispanic race 20 (100%) 18 (100%)Birth weight (g) 2992 (�123.34) 2919 (�151.57Perinatal complications 6 (30.0%) 7 (38.9%)History of hospitalisationa 6 (30.0%) 7 (38.9%)History of convulsive events 0 (0%) 0 (0%)School problemsb 1 (5%) 13 (72.2%)Average academic achievementc 8.68 (�1.15) 7.08 (�1.21)DSM-IVd

Inattention 1.05 (�0.99) 6.72 (�0.25)Hyperactivity 0.70 (�0.80) 3.94 (�1.73)Impulsivity 0.45 (�0.68) 2.33 (�0.90)

WAIS III total score 104.15 (�2.52) 100.55 (�1.91)Parents GCI e 9.25 (�4.44) 22.00 (�3.76)Teachers GCI f 12.20 (�3.35) 30.00 (�3.63)Personal GCIg 9.42 (�2.65) 30.66 (�6.07)Relatives with probable ADHD 5 (25%) 10 (55%)2693e6462 USDh 11 (55%) 11 (55.9%)>6539 USD 9 (45%) 7 (38.9%)

MeW U ¼ ManneWhitney test. BMI ¼ body mass index.a Due to any medical cause.b History of school problems: failed academic years, school expulsion, physical and vec Last academic year, scale 0e10.d Number of DSM-IV criteria for ADHD.e Parents GCI: Conner’s Global Index of the Conner’s Parent Rating Scale-Revised: Lonf Teachers GCI: Conner’s Global Index of the Conner’s Teacher Rating Scale-Revised: Lh Conner’s Global Index of the Conners-Wells Self-Report Scale: Long Version (CWSRSg Average monthly family income in US dollars.

significance. In the child groups, three zones showed reduced CT inthe frontal lobes of subjects with ADHD. The same was true for onezone in adolescents and one zone in adults (Fig. 1). These regionswere identified according to the labelling and parcellation of thecerebral cortex described by Desikan et al. (2006): They are locatedin the superior frontal region (green region in Fig. 2) and themiddlefrontal region (purple region in Fig. 2).

In anteroposterior order, the first zone that showed differencesbetween groups was located in the rostral middle frontal area(RMFA) and the superior frontal area (SFA). The Talairach coordi-nates at which the statistical difference between groups reachedtheminimum “p” value were: X¼ 18.5, Y¼ 59.8, Z¼ 0.2; Brodmannarea #10 (RMFA-SFA a in Fig. 2). This zone had a mean area of410 mm2 (s.d. � 91.00) and a mean volume of 1.90 ml (s d � 0.43).The average of the difference in CT between healthy and ADHDsubjects was 0.33 mm (Fig. 1).

The second zone that showed differences between healthy andADHD children was located on the medial surface of the SFA (SFAb in Fig. 2). The Talairach coordinates at which the statisticaldifference between groups reached the minimum “p” value were:X ¼ 12.7, Y ¼ 53.6, Z ¼ 22; Brodmann area #9 (SFA b in Fig. 2). Thiszone had a mean area of 115mm2 (s.d.� 24.70) and a mean volumeof 0.44 ml (s d � 0.13). The average of the difference in CT betweenhealthy and ADHD subjects was 0.28 mm (Fig. 1).

The third zone that showed differences between healthy andADHD children was located on the medial surface of the SFA (SFAcin Fig. 2). The Talairach coordinates at which the statistical differ-ence between groups reached the minimum “p” value were:X ¼ 9.5, Y ¼ 33.5, Z ¼ 22.7; Brodmann area #9 (SFA c in Fig. 2). This

mple of adolescents.

Statistical value df P value

n (%)

MeW U ¼ 120,50, z ¼ �1.88 e 0.082) X2 ¼ 0.117 1 0.757%)

MeW U ¼ 146.00, z ¼ �0.994 e 0.326t ¼ 0.138 36 0.891e

e

) t ¼ 0.375 36 0.708X2 ¼ 0.333 1 0.734X2 ¼ 0.333 1 0.734e

Fisher's exact test, p ¼ 0.000t ¼ 4.13 36 0.000

MeW U ¼ 0.000, z ¼ �5.360 e 0.000MeW U ¼ 24.00, z ¼ �4.650 0.000MeW U ¼ 27.50, z ¼ �4.640 0.000

t ¼ 1.09 36 0.429t ¼ �2.19 36 0.013t ¼ �3.60 36 0.007t ¼ �3.38 36 0.006X2 ¼ 23.702 1 0.054X2 ¼ 0.145 1 0.703

rbal aggression towards peers and teachers.

g Version (CPRS-R).ong Version (CTRS-R).-L).

L.G. Almeida et al. / Journal of Psychiatric Research 44 (2010) 1214e1223 1219

zone had amean area of 232mm2 (s.d.� 40.70) and amean volumeof 0.74 ml (s d � 0.19). The average of the difference in CT betweenhealthy and ADHD subjects was 0.35 mm (Fig. 1).

A unique zone that showed differences between healthy andADHD adolescents was located on the superior and medial surfaceof the SFA (SFA d in Fig. 2). The Talairach coordinates at which thestatistical difference between groups reached the lowest p valuewere: X¼ 6.9, Y¼ 52.7, Z¼ 25.7; Brodmann area #9 (SFA d in Fig. 2).This zone had a mean area of 332 mm2 (s.d. � 33.50) and a meanvolume of 1.20 ml (s d � 0.16). The average of the difference in CTbetween healthy and ADHD subjects was 0.36 mm (Fig. 1).

Finally, the only zone that showed differences between healthyand ADHD adults was located on the superior, lateral and medialsurface of the SFA (SFAe in Fig. 2). The Talairach coordinates atwhichthe statistical difference between groups reached the lowest “p”value were: X ¼ 7.2, Y ¼ 49.6, Z ¼ 25.4; Brodmann area #9 (SFA d inFig. 2). This zone had a mean area of 750 mm2 (s.d. � 101.90) andamean volume of 2.82 ml (s d� 0.38). The average of the differencein CT between healthy and ADHD subjects was 0.33 mm (Fig. 1).

4.4. Interaction between age group and clinical status

Two-way ANOVA did not show any interaction between thesetwo factors. (Interaction age group*clinical status:F ¼ 0.204,df ¼ 2,p ¼ 0.816).

4.5. Correlation between CT and the DSM-IV-TR criteria for ADHD

All regions that showed statistical differences in CT betweenhealthy and ADHD subjects in the three age groups showeda negative and significant correlation with the number of criteria

Table 3General, somatometric, sociodemographic, psychometric and clinimetric profile of the sa

Healthy ADHD

n ¼ 20 n ¼ 20

Mean (�s.d) or n (%) Mean (�Age (years) 27.57 (�2.6) 28.95 (Gender Male: 10 (50%) Male: 10

Female: 10 (50%) Female:BMI 24.42 (�2.35) 25.64 (Intracranial volume (ml) 1499 (�133) 1512 (Right-handed 20 (100%) 20 (Hispanic race 20 (100%) 20 (History of hospitalisationa 10 (50%) 5 (History of convulsive events 0 (0%) 0School problemsb 1 (5%) 8 (Junior High schoolc 0 (0%) 2 (High school c 2 (10%) 4 (College/Universityc 11 (55.0%) 8 (Postgraduatec 7 (35%) 6 (DSM-IVd

Inattention 0.40 (�0.68) 7.65 (Hyperactivity 0.85 (�1.38) 4.35 (Impulsivity 0.45 (�0.75) 2.03 (

WAIS III total 104.05 (�1.86) 103.85 (FASCT-SR 11.10 (�1.31) 34.25 (FASCT-O 11.15 (�3.03) 29.25 (Relatives with probable

ADHD3 (15%) 4 (

2693e6462 USDe 16 (80%) 13 (>6539 USD 4 (20%) 7 (

M.W.U ¼ ManneWhitney test. BMI ¼ body mass index.a Due to any medical cause.b History of school problems: failed academic years, school expulsion, physical and vec Highest degree achieved.d Number of DSM-IV criteria for ADHD.e Average monthly family income in US dollars.

for ADHD (according to the DSM-IV-TR). The highest correlationswere with the number of criteria for inattention and hyperactivity,and the lowest correlation was with the number of criteria forimpulsivity (Table 4).

The regression analysis showed that, in the three age groups, CTis predicted by the variable group controlling for the effects of age,gender, intracranial volume, BMI and the presence of comorbidities(for which respective values of beta did not reach statisticalsignificance) (Table 5).

5. Discussion

This study shows that CT in regions of the right superior frontalarea is thinner in children, adolescents and adults with ADHD.Furthermore, findings from the present work indicate that there isa correlation between the CTof these regions and the severity of thedisorder. The regression model shows that, in the three age groups,a diagnosis of ADHD predicts CT in the right superior frontalregions, where statistical differences between healthy and ADHDsubjects were observed (controlling for age, gender, intracranialvolume and BMI). This finding is in concordance with resultsobtained in studies of adults with ADHD (Hesslinger et al., 2002;Seidman et al., 2006; Makris et al., 2007 and Biederman et al.,2008) in which a reduction of the frontal lobe was demonstrated.Nonetheless, our findings are inconsistent with other studies (Shawet al., 2007a, Shaw et al., 2007b and Shaw et al., 2006) indicatingthat the differences in CT between healthy and ADHD childrendisappear during adolescence. These conflicting findings may bedue to dissimilarities in themethodologies employed in the studies.For example, one study (Shaw et al., 2007a) used one MRI scan in40% of the participants and used these data to estimate the cortical

mple of adults.

Statistical value df P value

s.d) or n (%)

�4.01) t ¼ �1.25 38 0.217(50%) X2 ¼ 0.000, df ¼ 1, p ¼ 1.010 (50%)�2.44) t ¼ �1.305 38 0.200�132) t ¼ �0.316 38 0.754100%) e

100%) e

25%) X2 ¼ 2.66 1 0.102(10%) e

40%) Fisher’s exact test, p ¼ 0.02010%) Fisher’s exact test, p ¼ 0.48720%) Fisher’s exact test, p ¼ 0.66140%) Fisher’s exact test, p ¼ 0.34230%) Fisher’s exact test, p ¼ 0.736

�1.18) MeW U ¼ 0.000, z ¼ �5.360�1.72) MeW U ¼ 34.50, z ¼ �4.570�1.30) MeW U ¼ 41.00, z ¼ �4.520

�1.86) t ¼ �0.790 38 0.043�1.45) t ¼ 11.80 38 0.000�1.87) t ¼ 6.54 38 0.00020%) Fisher’s exact test, p ¼ 1.00

72%) X2 ¼ 1.120 1 0.28835%)

rbal aggression towards peers and teachers.

Table 4Correlation coefficients for the cortical thickness of each region where statisticaldifferences were found between healthy and ADHD subjects and the number ofDSM-IV-TR criteria for ADHD.

Children Adolescents Adults

RMFA-SFA SFA SFA SFA SFA

a b c d e

Inattention rs ¼ �0.666 rs ¼ �0.664 rs ¼ �0.613 rs ¼ �0.509 rs ¼ �0.481p ¼ 0.000 p ¼ 0.000 p ¼ 0.0000 p ¼ 0.001 p ¼ 0.002

Hyperactivity rs ¼ �0.626 rs ¼ �0.628 rs ¼ �0.572 rs ¼ �0.369 rs ¼ �0.630p ¼ 0.000 p ¼ 0.000 p ¼ 0.000 p ¼ 0.023 p ¼ 0.000

Impulsivity rs ¼ �0.559 rs ¼ �0.559 rs ¼ �0.471 rs ¼ �0.348 rs ¼ �0.404p ¼ 0.000 p ¼ 0.000 p ¼ 0.001 p ¼ 0.002 p ¼ 0.010

Talairach coordinates in vertexmaximumdifferences between groups [RMFA-SFA a:x¼ 18.5, y ¼ 59.8, z¼ 0.2; SFA b: x¼ 12.7, y ¼ 53.6, z¼ 22.5; SFA c: x¼ 9.5, y¼ 33.5,z ¼ 22.7; SFA d: x ¼ 6.9, y ¼ 52.7, z ¼ 25.7; SFA e: x ¼ 7.2, y ¼ 49.6, z ¼ 25.4]. RMFA-SFA ¼ rostral middle frontal area and superior frontal area, SFA ¼ superior frontalarea.

L.G. Almeida et al. / Journal of Psychiatric Research 44 (2010) 1214e12231220

development trajectories. Hence, the results are predicted values ofCT instead of actual values. In addition, 36e41% of individuals weremedicated with stimulants at the time of the MRI scan. (Shaw et al.,2006). In two different studies published by Shaw et al. (2007b,2006), 66% of the participants received prior stimulant treatment(Shaw et al., 2007b) and 62e81% of the participants were under-going treatment at the beginning of the study or during follow-up.The same research group demonstrated that the stimulants have aneffect on CT in adolescents with ADHD, making their CT moresimilar to that of healthy individuals (Shaw et al., 2009). Becauseweexcluded patients with previous or current stimulant use, it ispossible that these effects may account for the differences betweentheir results and our own.

Regarding laterality (we found differences in CT of the righthemisphere but not the left), our results are in agreement withthose of studies analysed by Valera et al. (2007). The results of theirmeta-analysis show that the right hemisphere and frontal lobe areamong the brain regions most frequently assessed, and they displaythe largest and most significant area and volumetric reductionsrelative to control subjects. Additionally, some studies have showna reduction in CT of children and adolescents in the right frontallobe (Shaw et al., 2007b). On the other hand, several studies haveshown a reduction in the volume of the left frontal orbital lobe(Hesslinger et al., 2002) or in the frontal left superior gyrus(Biederman et al., 2008) of adults with ADHD. Moreover, somestudies report a reduction in CT or volume in the left and rightfrontal lobe in children (Shaw et al., 2007a) and adults with ADHD(Makris et al., 2007; Seidman et al., 2006). Based on these data, itseems that there is no clear hemispheric laterality in the anatomical

Fig. 2. Locations, areas and volumes (in the right hemisphere) where ct was thin

abnormalities found in the frontal lobe of subjects with ADHD.Although the right hemisphere has a greater role in attention,visuospatial functions and arousal (Rhawn, 1996) d and lesions ofthe right hemisphere can result in visuospatial, attentional andemotional disorders (Heilman et al., 1986) and deficits in inhibitorycontrol (Pliszka et al., 2000) d neither hemisphere is dominant inthe absolute sense. In addition, although each appears to be spe-cialised for different functions, most functions require the cooper-ation of both hemispheres (Heilman et al., 1986). However, ourresults are in agreement with the anatomical and functionalfeatures of the right frontal cortex.

ner in adhd individuals than in healthy individuals (according to age group).

Table 5Results of regression models of each anatomical area by age group. The predicted variable was the cortical thickness and the predictors were age, gender, intracranial volume,BMI and comorbidity.

R2 ANOVA b t p DurbineWatson AE IC

Childrena RMFA-SFA 0.214 F ¼ 12.5, 0.001 �0.48 �3.53 0.001 2.15 0.300 2.30b SFA 0.217 F ¼ 12.6, 0.001 �0.48 �3.55 0.001 2.30 0.301 2.37c SFA 0.316 F ¼ 18.9, 0.000 �0.56 �4.35 0.001 2.56 0.301 2.37Adolescentsd SFA 0.261 F ¼ 12.7, 0.001 �0.51 �3.56 0.001 2.88 0.312 2.32Adultse SFA 0.354 F ¼ 20.8, 0.000 �0.59 �4.56 0.000 2.52 0.293 2.41

In each age group the R2 and the ANOVA for the model are shown. The value of b (the standardised regression coefficient of the predictor variable “group”) with t-test andsignificance values are also displayed. Finally DurbineWatson¼ value of the residuals. Colinearity: AE¼ auto-engine, IC¼ condition index for each model are also shown. Theexcluded variables in each model were: age, gender, intracranial volume and body mass index (BMI). Talairach coordinates in vertex maximum differences between groups[RMFA-SFA a: x¼ 18.5, y¼ 59.8, z¼ 0.2; SFA b: x¼ 12.7, y¼ 53.6, z¼ 22.5; SFA c*: x¼ 9.5, y¼ 33.5, z¼ 22.7; SFA d: x¼ 6.9, y¼ 52.7, z¼ 25.7; SFA e: x¼ 7.2, y¼ 49.6, z¼ 25.4].RMFA-SFA ¼ rostral middle frontal area and superior frontal area, SFA ¼ superior frontal area. In each model the dependent variable was mean CT.

L.G. Almeida et al. / Journal of Psychiatric Research 44 (2010) 1214e1223 1221

This cortical region and its connections with diverse subcor-tical encephalic structures (e.g., the basal ganglia and thalamus)are involved in the control of motor activity, selective and sus-tained attention, problem solving and the control of impulsivity(Fan et al., 2005; Halperin and Schulz, 2006; Posner et al., 2006).All of these cognitive functions are disrupted in patients withADHD.

In the child and adolescent samples, 52.4% and 33% met theDSM-IV criteria for conduct disorder, respectively. There aredifferences in the brain regions involved in ADHD and conductdisorder. Studies utilising fMRI have used reward and attentiontasks to compare encephalic activation in children and adolescentswith ADHD and conduct disorder. A decrease in activation of theprefrontal cortex was found in ADHD individuals. In contrast,individuals with conduct disorder showed a reduction in activationof the paralimbic regions of the insula, hippocampus and anteriorcingulate (Rubia et al., 2009). Kruesi et al. (2004) discovered thatthe volume of the gray matter in the right temporal lobe is smallerin subjects with conduct disorder when compared to healthycontrols. They did not observe statistical differences in the volumeof gray matter in the frontal lobe. Lorberboym et al. (2004) usedsingle-photon computed emission tomography in conjunctionwith a performance test and reported that children and adoles-cents with comorbid ADHD, conduct disorder and oppositionaldefiant disorder showed a decrease in blood perfusion of thetemporal lobe. In contrast, individuals with non-comorbid ADHDshowed a decrease in blood perfusion of the frontal lobe. Anotherstudy (Bussing et al., 2002), however, did not find any volumetricdifferences between children and adolescents with ADHD andconduct disorder. Huebner et al. (2008) used voxel-basedmorphometry to compare children and adolescents with comorbidconduct disorder and ADHD to healthy controls. The affectedindividuals showed reduced gray matter volume in the left orbitalregion and in the temporal lobe bilaterally, including the amygdalaand hippocampus of the left side. Finally, Sterzer et al. (2007)found a reduction in the volume of the cortex in the insula ofadolescents with conduct disorder. Altogether, these findingssupport the idea that the neurobiological correlates of conductdisorder are different from those found in ADHD. Consequently, wecan conclude that the comorbidity between ADHD and conductdisorder in our sample did not affect the results. A diagnosis ofGAD was present in 4.8%, 11% and 15% of the children, adolescentsand adults with ADHD, respectively. Various studies have shownthat the structural abnormalities of this disorder are located in theorbitofrontal cortex (Mholman et al., 2009), amygdala (De Belliset al., 2000) and superior temporal gyrus (De Bellis et al., 2002).These data suggest that the encephalic regions involved in GAD aredifferent from those found in this study, which are presumably

associated with ADHD. One patient (15%) from the sample of adultswith ADHD had a history of major depressive disorder. Numerousstudies have shown that the brain regions associated with thisdisorder are the amygdala, orbitofrontal cortex, prefrontal cortex,anterior cingulate gyrus, hippocampus and basal ganglia. It seemsthat the encephalic structures involved in MDD are different fromthose found in this study, which are presumably associated withADHD.

6. Limitations

In the present study we used a cross-sectional design intendedto test the hypothesis that CT alterations persist in ADHD fromchildhood to adulthood. A longitudinal approach would have beenan alternative design since we were comparing CT in different agegroups and with different patients in each group, thereby creatinga great deal of variability. However, our cross-sectional designallowed us to avoid some of the problems inherent to longitudinalstudies (e.g., by including patients older than 18 years, obtaining anactual measure instead of estimated measures of CT, and includingsubjects free of stimulant treatment). We believe that cross-sectional and longitudinal designs each have their own advantagesand disadvantages.

In conclusion, the current study suggests that there is a thinningof the cortex located in regions of the frontal superior gyrus inchildren, adolescents and adults with ADHD. Moreover, the degreeof CT thinning is related to the severity of the disorder.

7. Role of the funding source

Funding for this study was provided by Servicios de Salud delEstado de Querétaro, México (Department of Public Health).

Servicios de Salud del Estado de Querétaro had no further role inthe study design; in the collection, analysis and interpretation ofdata; in the writing of the report; or in the decision to submit thepaper for publication.

Contributors

Luis G. Almeida. M.D., MSc, Ph.D. Psychiatry and NeurosciencesAddress 1: Centro Estatal de Salud Mental, Servicios de Salud delEstado de Querétaro. Avenida 5 de Febrero 105, Los Virreyes. C.P.76170. Querétaro México. Phone/Fax (52e442) [email protected].

Address 2: Instituto de Neurobiología, Universidad NacionalAutónoma de México, Campus Juriquilla. Boulevard Juriquilla 3001.C.P. 76230. Querétaro, México.

JosefinaRicardo-Garcell. Ph.D. Neurosciences andNeurophysiology.

L.G. Almeida et al. / Journal of Psychiatric Research 44 (2010) 1214e12231222

Instituto de Neurobiología, Universidad Nacional Autónoma deMéxico, Campus Juriquilla. Boulevard Juriquilla 3001. C.P. 76230.Querétaro, México.

Hugo Prado. M.D. Centro Estatal de Salud Mental, Servicios deSalud del Estado de Querétaro. Avenida 5 de Febrero 105, Los Vir-reyes. C.P. 76170. Querétaro México.

Lázaro Barajas. Ph.D. Physics, Mathematics and Statistics Insti-tuto Tecnológico de Monterrey, Campus Querétaro. EpigmenioGonzález 500, Fraccionamiento San Pablo. C.P. 76130. Querétaro,México.

Antonio Fernández-Bouzas. Ph.D. Neuroradiology.Instituto de Neurobiología, Universidad Nacional Autónoma de

México, Campus Juriquilla. Boulevard Juriquilla 3001. C.P. 76230.Querétaro, México.

David Ávila M.D. Neuroradiology.Instituto de Neurobiología, Universidad Nacional Autónoma de

México, Campus Juriquilla. Boulevard Juriquilla 3001. C.P. 76230.Querétaro, México.

Reyna B. Martínez. M.D. Centro Estatal de Salud Mental, Servi-cios de Salud del Estado de Querétaro. Avenida 5 de Febrero 105,Los Virreyes. C.P. 76170. Querétaro México.

Disclosure of competing interests

Drs. Almeida, Ricardo-Garcell, Prado, Barajas, Fernández, Ávila &Martínez state that: “In the present study there are no competinginterests of any kind in the last 10 years.”

Acknowledgements

We thank all the individuals who participated in this study, tothe Consejo Nacional de Ciencia y Tecnología (CONACYT), al pro-grama de posgrado en ciencias biomédicas de la UniversidadNacional Autónoma de México y a los servicios de salud del estadode Querétaro (SESEQ)

References

Almeida L, Friederichsen A, Hernández A, Rodriguez R, Cortés J. Construction, val-idity and reliability, of the screening scale “FASCT” for attention deficit hyper-activity disorder in adults (Self-reported and observer versions). ActasEspañolas de Psiquiatría 2006;34(4):231e8.

American Psychiatric Association. Attention-deficit and disruptive behavior disor-ders. Text Revision. In: American psychiatric association. Diagnostic andstatistical manual of Mental disorders. 2nd ed. Washington DC: AmericanPsychiatric Publishing INC; 2000.

Biederman J, Makris N, Valera E, Monuteaux M, Goldstein J, Buka S, et al. Towardsfurther understanding of the co-morbidity between attention deficit hyperac-tivity disorder and bipolar disorder: a MRI study of brain volumes. Psycholog-ical Medicine 2008;38(7):1045e56.

Biederman J, Mick E, Faraone S. Age-dependent decline of symptoms of attentiondeficit hyperactivity disorder: impact of remission definition and symptomtype. The American Journal of Psychiatry 2000;157(5):816e8.

Bussing R, Grudnik J, Mason D, Wasiak M, Leonard C. ADHD and conduct disorder:an MRI study in a community sample. The World Journal of Biological Psychi-atry 2002;3(4):216e20.

Castellanos F, Lee P, Sharp W, Jeffries N, Greenstein D, Clasen L, et al. Developmentaltrajectories of brain volume abnormalities in children and adolescents withattention-deficit/hyperactivity disorder. Journal of the American MedicalAssociation 2002;288(14):1740e8.

Cohen J. Statistical power analysis for the behavioral sciences. 2nd ed. Hillsdale, NJ:Lawrence Earlbaum Associates; 1988.

Conners C, Sitarenios G, Parker J, Epstein J. Revision and re-standardization of theConners teacher rating scale (CTRS-R): factor structure, reliability and criterionvalidity. Journal of Abnormal Child Psychology 1998a;26(4):279e91.

Conners C, Sitarenios G, Parker J, Epstein J. The revised Conners’ parent rating scale(CPRS-R): factor structure, reliability and criterion validity. Journal of AbnormalChild Psychology 1998b;26(4):257e68.

Conners C, Wells K, Parker J, Sitarenios G, Diamond J, Powell J. A new self-reportscale for the assessment of adolescent psychopathology: factor structure, reli-ability, validity, and diagnostic sensitivity. Journal of Abnormal ChildPsychology 1997;25(6):487e97.

Dale A, Fischl B, Sereno M. Cortical surface-based analysis. I. Segmentation andsurface reconstruction. Neuroimage 1999;9(2):179e94.

Dale A, Sereno M. Improved localization of cortical activity by combining EEG andMEG with MRI cortical surface reconstruction: a linear approach. Journal ofCognitive Neuroscience 1993;5(2):162e76.

De Bellis M, Casey B, Dahl R, Birmaher B, Williamson D, Thomas K, et al. A pilotstudy of amygdala volumes in pediatric generalized anxiety disorder. BiologicalPsychiatry 2000;48(1):51e7.

De Bellis M, Keshavan M, Shifflett H, Iyengar S, Dahl R, Axelson D, et al. Superiortemporal gyrus volumes in pediatric generalized anxiety disorder. BiologicalPsychiatry 2002;51(7):553e62.

Desikan R, Ségonne F, Fischl B, Quinn B, Dickerson B, Blacker D, et al. An automatedlabeling system for subdividing the human cerebral cortex on MRI scans intogyral based regions of interest. Neuroimage 2006;31(3):968e80.

Ellison-Wright I, Ellison-Wright Z, Bullmore E. Structural brain changes in atten-tion-deficit hyperactivity disorder identified by meta-analysis. Bio Med Central(BMC) Psychiatry 2008;8(51):1e8.

Fan J, McCadliss B, Fosella J, Flombaum J, Posner M. The activation of attentionalnetworks. Neuroimage 2005;26(2):471e9.

Fischl B, Dale A. Measuring the thickness of the human cerebral cortex frommagnetic resonance images. Proceedings of the National Academy of Sciencesof the United States of America 2000;97(20):11050e5.

Fischl B, Liu A, Dale A. Automated manifold surgery: constructing geometricallyaccurate and topologically correct models of the human cerebral cortex. IEEETransactions on Medical Imaging 2001;20(1):70e80.

Fischl B, Salat D, Busa E, Albert M, Dieterich M, Haselgrove C, et al. Whole brainsegmentation: automated labeling of neuroanatomical structures in the humanbrain. Neuron 2002;33(3):341e55.

Fischl B, Salat D, van der Kouwe A, Makris N, Ségonne F, Quinn B, et al. Sequence-independent segmentation of magnetic resonance images. Neuroimage2004a;23(1 Suppl):S69e84.

Fischl B, Sereno M, Dale A. Cortical surface-based analysis. II: inflation, flattening,and a surface-based coordinate system. Neuroimage 1999a;9(2):195e207.

Fischl B, Sereno M, Tootell R, Dale A. High-resolution intersubject averaging anda coordinate system for the cortical surface. Human Brain Mapping 1999b;8(4):272e84.

Fischl B, van der Kouwe A, Destrieux C, Halgren E, Ségonne F, Salat D, et al. Automat-ically parcellating the human cerebral cortex. Cerebral Cortex 2004b;14(1):11e22.

Halperin J, Schulz K. Revisiting the role of the prefrontal cortex in the pathophys-iology of attention-deficit/hyperactivity disorder. Psychopharmacology Bulletin2006;132(4):560e81.

HanX, Jovicich J, SalatD, vanderKouweA, QuinnB, Czanner S, et al. Reliability ofMRI-derived measurements of human cerebral cortical thickness: the effects of fieldstrength, scanner upgrade and manufacturer. Neuroimage 2006;32(1):180e94.

Heilman K, Bowers D, Valenstein E, Watson R. The right hemisphere: neuro-psychological functions. Journal of Neurosurgery 1986;64(5):693e704.

Hesslinger B, Tebartz van Elst L, Thiel T, Haegele K, Henning J, Ebert D. Fronto-orbital volume reductions in adult patients with attention deficit hyperactivitydisorder. Neuroscience Letters 2002;328(3):319e21.

Huebner T, Vloet T, Marx I, Konrad K, Fink G, Herpertz S, et al. Morphometric brainabnormalities in boys with conduct disorder. Journal of the American Academyof Child and Adolescent Psychiatry 2008;47(5):540e7.

Jovicich J, Czanner S, Greve D, Haley E, van der Kouwe A, Gollub R, et al. Reliabilityin multi-site structural MRI studies: effects of gradient non-linearity correctionon phantom and human data. Neuroimage 2006;30(2):436e43.

Kruesi M, Casanova M, Mannheim G, Johnson-Bilder A. Reduced temporal lobevolume in early onset conduct disorder. Psychiatry Research 2004;132(1):1e11.

Kuperberg G, Broome M, McGuireP David A, Eddy M, Ozawa F, Goff D, et al.Regionally localized thinning of the cerebral cortex in schizophrenia. Archivesof General Psychiatry 2003;60(9):878e88.

Lorberboym M, Watemberg N, Nissenkorn A, Nir B, Lerman-Sagie T. Technetium99m ethylcysteinate dimer single-photon emission computed tomography(SPECT) during intellectual stress test in children and adolescents with pureversus comorbid attention-deficit hyperactivity disorder (ADHD). Journal ofChild Neurology 2004;19(2):91e6.

Makris N, Biederman J, Valera E, Bush G, Kaiser J, Kennedy D, et al. Cortical thinningof the attention and executive function networks in adults with attention-deficit/hyperactivity disorder. Cerebral Cortex 2007;17(6):1364e75.

Mholman J, Price R, Eldreth D, Chazin D, Glover D, Kates W. The Relation of worry toprefrontal cortex volume in older adults with and without generalized anxietydisorder. Psychiatry Research 2009;173(2):121e7.

Oldfield R. The assessment and analysis of handedness: the Edinburgh inventory.Neuropsychologia 1971;9(1):97e113.

Pliszka S, Liotti M, Woldorff M. Inhibitory control in children with attention-deficit/hyperactivity disorder: event-related potentials identify the processingcomponent and timing of an impaired right-frontal response-inhibitionmechanism. Biological Psychiatry 2000;48(3):238e46.

Posner M, Sheese B, Odludas Y, Tang Y. Analyzing and shaping human attentionalnetworks. Neural Netwoks 2006;19(9):1422e9.

Rhawn J. The right cerebral hemisphere. In: Rhawn J, editor. Neuropsychiatry,Neuropsychology, and clinical Neuroscience. Williams & Wilkins; 1996. p.75e117.

Rosas H, Liu A, Hersch S, Glessner M, Ferrante R, Salat D, et al. Regional andprogressive thinning of the cortical ribbon in Huntington’s disease. Neurology2002;58(5):695e701.

L.G. Almeida et al. / Journal of Psychiatric Research 44 (2010) 1214e1223 1223

Rosenthal R, Rosnow RL. Essentials of behavioral research: Methods and dataanalysis. 2nd ed. New York: McGraw Hill; 1991.

Rosnow RL, Rosenthal R. Computing contrast, effect sizes, and counternulls onpeople’s published data: general procedures for research consumers. Psycho-logical Methods 1996;1(4):331e40.

Rubia K, Halari R, Smith A, Mohammed M, Scott S, Giampietro V, et al. Dissociatedfunctional brain abnormalities of inhibition in boys with pure conduct disorderand in boys with pure attention deficit hyperactivity disorder. American Journalof Psychiatry 2008;165(7):889e97.

Rubia K, Smith A, Halari R, Matsukura F, Mohammad M, Taylor E, et al. Disorder-specific dissociation of orbitofrontal dysfunction in boys with pure conductdisorder during reward and ventrolateral prefrontal dysfunction in boys withpure ADHD during sustained attention. American Journal of Psychiatry2009;166(1):83e94.

Salat D, Buckner R, Snyder A, Greve D, Desikan R, Busa E, et al. Thinning of thecerebral cortex in aging. Cerebral Cortex 2004;14(7):721e30.

Ségonne F, Dale A, Busa E, Glessner M, Salat D, Hahn H, et al. A hybrid approach tothe skull-stripping problem in MRI. Neuroimage 2004;22(3):1060e75.

Ségonne F, Pacheco J, Fischl B. Geometrically accurate topology-correction ofcortical surfaces using nonseparating loops. IEEE Transactions on MedicalImaging 2007;26(4):518e29.

Seidman L, Valera E, Makris N, Monuteaux M, Boriel D, Kelkar K, et al. Dorsolateralprefrontal and anterior cingulated cortex volumetric abnormalities in adultswith attention-deficit/hyperactivity disorder identified by magnetic resonanceimaging. Biological Psychiatry 2006;60(10):1071e80.

Shaw P, Eckstrand K, Sharp W, Blumenthal J, Lerch J, Greenstein D, et al. Attention-deficit/hyperactivity disorder is characterized by a delay in cortical maturation.Proceedings of the National Academy of Sciences of the United States ofAmerica 2007a;104(49):19649e54.

Shaw P, Gornick M, Lerch J, Addington A, Seal J, Greenstein D, et al. Polymorphismsof the dopamine D4 receptor, clinical outcome, and cortical structure in

attention-deficit/hyperactivity disorder. Archives of General Psychiatry2007b;64(8):921e31.

Shaw P, Lerch J, Greenstein D, Sharp W, Clasen L, Evans A, et al. Longitudinalmapping of cortical thickness and clinical outcome in children and adolescentswith attention-deficit/hyperactivity disorder. Archives of General Psychiatry2006;63(5):540e9.

Shaw P, Sharp W, Morrison M, Eckstrand K, Greenstein D, Clasen L, et al. Psy-chostimulant treatment and the developing cortex in attention-deficit/hyperactivity disorder. The American Journal of Psychiatry 2009;166(1):58e63.

Sheehan D, Lecrubier Y, Sheehan K, Amorim P, Janavs J, Weiller E, et al. The miniinternational neuropsychiatric interview (MINI): the development and valida-tion of a structured diagnostic interview for DSM-IV and ICD-10. Journal ofClinical Psychiatry 1998;20(59 Suppl):22e33.

Sled J, Zijdenbos A, Evans A. A nonparametric method for automatic correction ofintensity nonuniformity in MRI data. IEEE Transactions on Medical Imaging1998;17(1):87e97.

Stahl S. Attention deficit hyperactivity disorder and its treatment. In: Grandy M,editor. Stahl’s Essential Psychopharmacology: Neuroscientific Basis and Prac-tical Applications. 3rd ed. New York: Cambridge University Press; 2008. p.863e97.

Sterzer P, Stadler C, Poustka F, Kleinschmidt A. A structural neural deficit inadolescents with conduct disorder and its association with lack of empathy.Neuroimage 2007;37(1):335e42.

Valera E, Faraone S, Murray K, Seidman L. Meta-analysis of structural imagingfindings in attention-deficit/hyperactivity disorder. Biological Psychiatry2007;61(12):1361e9.

Wechsler D. Manual for Wechsler adult Intelligence scale. 3rd ed. San Antonio,Texas: The Psychological Corporation; 1981. WAIS-III.

Wechsler D. Manual for Wechsler Intelligence scale for children. 3rd ed. SanAntonio, Texas: The Psychological Corporation; 1991. WISC-III.