Developmental Changes in Brain Function Underlying Inhibitory ...

Developmental Changes in Brain Function Underlying InhibitoryControl in Autism Spectrum DisordersAarthi Padmanabhan, Krista Garver, Kirsten O’Hearn, Natalie Nawarawong, Ran Liu, Nancy Minshew,John Sweeney, and Beatriz Luna

The development of inhibitory control—the ability to suppress inappropriate actions in order to make goal-directedresponses—is often impaired in autism spectrum disorders (ASD). In the present study, we examined whether theimpairments in inhibitory control evident in ASD reflect—in part—differences in the development of the neuralsubstrates of inhibitory control from adolescence into adulthood. We conducted a functional magnetic resonanceimaging (fMRI) study on the anti-saccade task, a probe of inhibitory control, in high-functioning adolescents and adultswith ASD compared to a matched group of typically developing (TD) individuals. The ASD group did not show theage-related improvements in behavioral performance from adolescence to adulthood evident in the typical group,consistent with previous behavioral work. The fMRI results indicated that much of the circuitry recruited by the ASDgroup was similar to the TD group. However, the ASD group demonstrated some unique patterns, including: (a) a failureto recruit the frontal eye field during response preparation in adolescence but comparable recruitment in adulthood; (b)greater recruitment of putamen in adolescence and precuneus in adolescence and adulthood than the TD group; and (c)decreased recruitment in the inferior parietal lobule relative to TD groups. Taken together, these results suggest that braincircuitry underlying inhibitory control develops differently from adolescence to adulthood in ASD. Specifically, theremay be relative underdevelopment of brain processes underlying inhibitory control in ASD, which may lead toengagement of subcortical compensatory processes. Autism Res 2014, ••: ••–••. © 2014 International Society forAutism Research, Wiley Periodicals, Inc.

Keywords: autism; fMRI; inhibitory control; antisaccade; development; adolescence

Introduction

Autism spectrum disorders (ASD) are a class ofneurodevelopmental disorders characterized in part byatypical brain development, which may lead to abnor-malities in cognitive behaviors. In typical development,substantial improvements in inhibitory control of behav-ior and the brain systems that underlie it continue fromadolescence into early adulthood [Luna, Garver, Urban,Lazar, & Sweeney, 2004; Velanova, Wheeler, & Luna,2008, 2009]. Inhibitory control, defined as the ability tosuppress automatic but inappropriate responses in favorof a goal-directed behavior [Bjorklund & Harnishfeger,1995; Dempster, 1992] may be a likely candidate for cog-nitive differences in ASD, though the results are mixed,suggesting that deficits may be paradigm specific[Goldberg et al., 2005; Ozonoff & Strayer, 1997; Ozonoff,Strayer, McMahon, & Filloux, 1994; Schmitz et al., 2006][Agam, Joseph, Barton, & Manoach, 2010; Luna, Doll,Hegedus, Minshew, & Sweeney, 2007; Minshew, Luna, &Sweeney, 1999; Ozonoff & Strayer, 1997; Russell, 1997] or

highlight the need to study the development of theseprocesses [Luna et al., 2007; Solomon et al., 2014]. Asadolescence is a time of significant change in brain andbehavior [Spear, 2000], the transition from adolescence toadulthood has become focus of investigation and repre-sents a potential window of plasticity in the brain, whichmay be impacted in ASD.

On the inhibitory tasks shown to differ in ASD, studieshave demonstrated that both children and adults withASD have poorer performance than matched typical con-trols, suggesting that abnormalities may persist through-out development [Agam et al., 2010; Goldberg et al., 2002;Luna et al., 2007; Minshew et al., 1999; Mosconi et al.,2009; Solomon et al., 2014]. One index of inhibitorycontrol, the anti-saccade(AS) task, is a well-establishedoculomotor paradigm that has previously been used tomodel the developmental trajectory of inhibitory controlthrough adolescence into adulthood [Fischer, Biscaldi, &Gezeck, 1997; Fukushima, Hatta, & Fukushima, 2000;Klein & Foerster, 2001; Luna et al., 2001, 2004; Munoz,Broughton, Goldring, & Armstrong, 1998; Velanova et al.,

From Psychiatry, University of Pittsburgh, Pittsburgh, Pennsylvania (A.P., K.G., K.O., N.N.); Psychology, Carnegie Mellon University, Pittsburgh,Pennsylvania (R.L.); Psychiatry and Neurology, University of Pittsburgh Medical School, Pittsburgh, Pennsylvania (N.M.); University of Texas HealthScience Center, Psychiatry and Pediatrics, Dallas, Texas (J.S.); Psychiatry and Psychology, University of Pittsburgh, Pittsburgh, Pennsylvania (B.L.)

Received February 18, 2013; accepted for publication May 28, 2014Address for correspondence and reprints: Aarthi Padmanabhan, Laboratory of Neurocognitive Development, Loeffler Building, Room 108, 121 Meyran

Avenue, Pittsburgh, PA 15213. E-mail: [email protected] online in Wiley Online Library (wileyonlinelibrary.com)DOI: 10.1002/aur.1398© 2014 International Society for Autism Research, Wiley Periodicals, Inc.

RESEARCH ARTICLE

INSAR 1Autism Research ••: ••–••, 2014

mailto:[email protected]

http://wileyonlinelibrary.com

2008, 2009]. During the AS task, participants must sup-press an automatic reaction to a peripheral stimulus andgenerate a planned oculomotor response in the oppositedirection. Prior evidence has suggested that TD adults andadolescents engage a widely distributed set of brainregions during performance of the AS task, includingcortical and subcortical regions such as the frontal, supple-mentary, and parietal eye fields (FEF, SEF, and PEF, respec-tively), basal ganglia, dorsolateral prefrontal cortex(DLPFC), and cerebellum [for review, see Munoz &Everling, 2004]. Our previous fMRI studies using the anti-saccade task to examine normative development indicatesignificant developmental changes in brain function sup-porting anti-saccade performance and improved error pro-cessing. [Luna et al., 2001; Velanova et al., 2008, 2009].

Robust inhibitory control deficits on the AS task havebeen documented in adults with ASD [Agam et al., 2010;Luna et al., 2007; Minshew et al., 1999; Mosconi et al.,2009; Thakkar et al., 2008]. Adults with ASD demonstrateworse performance, decreased recruitment of the anteriorcingulate cortex (ACC) and the FEF following correcttrials, and reduced functional connectivity between theACC and premotor regions compared to typical adults[Agam et al., 2010; Thakkar et al., 2008]. This is consis-tent with other inhibitory control tasks that result inreduced activation and functional connectivity in ACC,PFC, and posterior parietal regions, in adults with ASDrelative to typically developed individuals [Kana, Keller,Minshew, & Just, 2007; Schmitz et al., 2006; Solomonet al., 2009]. We previously examined the development ofinhibitory control behavior in ASD using the AS task in across-sectional sample from childhood to adulthood. Theresults indicated that despite performing worse than typi-cally developing individuals in both adolescence andadulthood, the group with ASD showed significant age-related improvements similar to typically developingindividuals [Luna et al., 2007]. That is, individuals withASD show parallel improvements in inhibitory controlthrough adolescence but do not catch up with typicallydeveloping peers. These behavioral findings suggest thatsimilar brain maturation processes may be occurring inASD from adolescence to adulthood, but to date, little isknown about the development of the neural substrates ofinhibitory control in ASD.

Identifying the neural substrates of the developmentaldelays, arrests, and abnormalities in ASD will provideinsight into limitations in inhibitory control, and help toidentify time points during which intervention will beparticularly successful. Furthermore, assessing develop-mental changes into adulthood will identify cognitivedeficits that may be late appearing and highlights howthe final stage of development may deviate from typicaldevelopment.

In the present study, we examined the neural correlatesof the AS task in adolescents and adults diagnosed with

ASD relative to age and IQ-matched typically developingadolescent and adult participants using functional mag-netic resonance imaging (fMRI). Similar to our previousbehavioral results [Luna et al., 2007], we predicted thatthe ASD group would show age-related improvements inanti-saccade performance from adolescence to adulthoodsimilar to the typically developing group. However, wehad no directional hypotheses for the interactionbetween age group and diagnosis group, given the lack ofprevious studies in this area.

MethodsParticipants

Forty-two participants were recruited for the study, 14typical adults (ages 18–31, M = 23.4 (+ /− 4.0); twofemales), and nine typical adolescents (ages 12–17,M = 13.9 (+ /− 1.4); two females) (CON groups) and eightadults with ASD (ages 19–33, M = 24.9 (+ /− 6.9),and eleven adolescents with ASD (ages 13–17, M = 15.1(+ /− 1.2); two females) (ASD groups). The Autism Diag-nostic Interview-Revised [Lord, Rutter, & Couteur, 1994]and the Autism Diagnostic Observation Schedule-General[Lord et al., 2000; Lord, Rutter, & Goode, 1989] were usedto diagnose ASD. Diagnosis was confirmed by expertclinical opinion (NJM). Potential subjects were excludedif known to have an associated disorder such as tuberoussclerosis or fragile-X syndrome. All participants wererequired to have full scale and verbal IQ scores of 80 orabove (Table 1). Medication status of the ASD participantsis reported in Supporting Information Table S1. All par-ticipants were free of medications known to affect eyemovements and had no history of seizure disorder.

There were no significant differences between groups inIQ (WASI) (Adult CON: 109.5 + /− 13.9, Adult ASD:98.9 + /− 13.9, Adolescent CON: 109.7 + /− 10.1, Adoles-cent ASD: 106.2 + /− 12.1, Ps > .05). All but one partici-pant were right handed. CON individuals had nopersonal or first-degree relative history of neurologicaldisease or neuropsychiatric illness as determined by inter-view. Vision was normal (as stated by the participant orguardian for minor participants) or corrected to normalusing MRI compatible glasses or contact lenses. Experi-mental procedures for this study complied with the Codeof Ethics of the World Medical Association (1964 Decla-ration of Helsinki) and the Institutional Review Board atthe University of Pittsburgh. Subjects were paid for theirparticipation in the study. All participants and guardiansprovided informed consent prior to participation.

Paradigm

Participants performed an event-related task, whichincluded AS and visually guided saccade (VGS) trials.

INSAR2 Padmanabhan et al./Development of inhibitory control in autism spectrum disorders

Immediately prior to scanning, participants were trainedon the AS and VGS tasks in a training room outside thescanner. During the AS trials, a red central fixation crossappeared on the screen for 2.5 sec signaling participantsto prepare for presentation of the target stimulus. Whenthe red cross disappeared, a peripheral target appeared atan unpredictable location on the horizontal meridian at3, 6, or 9 degrees of visual angle to the left and right of thecenter fixation cross for 2.5 sec. Participants wereinstructed not to look at the stimulus, but instead makean eye movement to the mirror location. Peripheraltarget location was randomized within each run. Duringthe VGS trials, participants were presented with a greenfixation cross and instructed to look toward theperipheral stimulus when it appeared. These VGS trialswere randomly interspersed between the AS trials(Fig. 1).

To uniquely estimate the hemodynamic response forthe preparatory and response phases of the AS trials, ourexperimental design included additional AS partial“catch” trials, randomly inserted along with jitteredintertrial intervals [Ollinger, Corbetta, & Shuldman,2001; Ollinger, Shulman, & Corbetta, 2001]. Thisapproach ensured that there were a sufficient number ofindependent linear equations to separately estimate theBlood-Oxygen-Level Dependent (BOLD) response associ-ated with the preparatory and saccade response phases ofeach AS trial. The catch trials consisted of the presenta-tion of a red central fixation cross, with no followingtarget. The intertrial fixation period was uniformly dis-tributed and jittered between intervals of 2.5, 5, or 7.5 secand consisted of a white cross on a black background that

subjects were instructed to fixate on in between trials.This allowed us to isolate activation specific to the fixa-tion period when the inhibitory processes begin prior tomaking a correct inhibitory response [Everling, Dorris, &Munoz, 1998; Everling, Spantekow, Krappmann, & Flohr,1998]. The three trial types (prep only, whole AS, wholeVGS) were presented in a random order, separated by theintertrial intervals. Participants performed two functionalruns of the task (duration = 6 min 50 sec each) for a totalof 34 AS trials, 16 VGS trials, and 18 catch AS trials.

Eyetracking

Eye movement measurements during the scan using along-range optics eye-tracking system (Model R-LRO6,Applied Science Laboratories, Bedford, MA). Nine-pointcalibrations were performed at the beginning of thesession. Stimuli were presented using E-Prime (Psychol-ogy Software Tools, Inc., Pittsburgh, PA) projected onto aflat screen positioned behind the magnet. Participantsviewed the screen using a mirror mounted on the headcoil. Eye-movement data were analyzed and scoredoffline using ILAB [Gitelman, 2002] in conjunction withan in-house scoring program. The behavioral variables ofinterest were error rates for AS and VGS trials (the numberof accurate eye movements/ total number of scorabletrials). A correct response in the VGS task was one inwhich the first eye movement during the saccaderesponse epoch was made toward the peripheral cue. Acorrect response in the AS task was one in which the firsteye movement during the saccade response epoch wasmade toward the mirror location of the peripheral cue

Table 1. Demographic Information

ASD CON

Adolescent Adult Adolescent Adultn = 11 n = 8 n = 9 n = 14

Males 9 8 7 12Right handed, n (%) 36 (88)

ANOVAMean (SD) Mean (SD) Mean (SD) Mean (SD) P

Age (years) 15.10 (1.19) 24.88 (6.94) 13.89 (1.36) 23.36 (3.99) 0.797Full scale IQ 108.10 (10.94) 98.88 (13.93) 109.71 (10.13) 109.53 (6.63) 0.925ADOSCommunication 4.50 (1.58) 4.38 (1.41) – –Social 9.50 (2.27) 10.5 (1.69) – –Total 14.00 (4.43) 14.88 (2.53) – –ADISocial 23.90 (3.28) 19.38 (4.87) – –Communication 18.20 (3.42) 15.25 (4.33) – –Restricted, repetitive

behaviors6.30 (2.45) 6.25 (1.98) – –

Abnormal 3.4 (1.43) 2.62 (1.60) – –

dashes mean not relevant (CON individuals were not administered the ADOS or ADI).

3Padmanabhan et al./Development of inhibitory control in autism spectrum disordersINSAR

and extended beyond a 2.5 degrees/visual angle centralfixation zone. Inhibitory errors occurred when the firstsaccade during the saccade response epoch was directedtoward the suddenly appearing peripheral stimulus thatthey were to ignore and exceeded the 2.5 degrees/visualangle central fixation zone.

Image Acquisition

Imaging data were acquired using a 3.0 Tesla whole-bodyMR scanner (SIGNA; General Electric Medical Systems,Milwaukee, WI) with echo-planar imaging capability anda commercial head radiofrequency coil. The followinggradient echo echo-planar acquisition parameters wereused: echo time = 25 ms; repetition time = 2,500 ms;single shot; full k space; 64 × 64 acquisition matrixwith field of view = 20 × 20 cm3, voxel dimen-sion = 3.125 × 3.125 × 6 mm. Twenty-three 5-mm axialslices were prescribed with a 1-mm gap in order to imagethe whole brain. The first four volumes in each run werediscarded to allow stabilization of longitudinal magneti-zation. A three-dimensional spoiled gradient recalled(SPGR) pulse sequence with 124 slices (6 mm each) wasused to acquire structural images in the axial plane. Theseanatomical images were used to register to functionalimages and to localize regions of activation.

Image Preprocessing

Imaging data were preprocessed using FSL (FunctionalMRI of the Brain (FMRIB) Software Library; [Smith et al.,2004]. Our preprocessing procedures included the follow-ing: SPGR images were warped into Talairach space[Talairach & Tournoux, 1988] using a 12-parameter affinetransformation in FSL [Jenkinson & Smith, 2001]. Func-tional images were first slice-time corrected, adjusting forinterleaved slice acquisition. Images were then rigid-bodymotion corrected by aligning all volumes with thevolume acquired in the middle of the fMRI session. Rota-tional and translational head movement estimates werecalculated. Following brain extraction, functional imageswere affine registered and warped to structural SPGRimages in Talairach space [Talairach & Tournoux, 1988].Images were resampled to voxel sizes of 3 × 3 × 3 mm. Nosubjects were excluded due to motion: instead the tem-poral derivative of the relative displacement from themiddle volume for each run was calculated for eachvolume in the x, y, and z directions. Magnitude of thevelocity was then calculated by taking the square root ofthe sum of squares of the x, y, and z components for eachvolume. Volumes with a velocity of over 1.2 mm/TR wereremoved from subsequent analyses. Participant groupsdid not differ in number of volumes removed due to

Figure 1. Anti-saccade (AS) task schematic. A red fixation cross appeared for 2.5 sec. A peripheral light appeared for 2.5 sec duringwhich participants were instructed to generate a saccade to its mirror location.


excessive motion (P > 0.05). Total number of volumesremoved for each group were the following: Adult ASD: 2,Adult CON: 1, Teen ASD: 15, Teen CON: 8. Images werethen spatially smoothed with a 5 mm full-width at halfmaximum Gaussian smoothing kernel, and high-pass fil-tered (sigma = 30 sec) to remove low frequency drift.Finally, data from each run were scaled to a mean of onehundred.

Image Analyses

Analysis and Visualization of Functional Neuroimages(AFNI, Bethesda, MD, USA) software [Cox, 1996] was usedfor individual subject deconvolution, followed by groupanalyses (see Supporting Information Fig. S1).Deconvolution methods followed steps delineated byWard [2002]. Our model consisted of three orthogonalregressors of interest; the preparatory period of the AS(catch and correct trials only), the correct AS saccaderesponse, and the correct VGS trials, as well as regressorsfor incorrect AS and VGS trials. Linear and nonlineartrends and six motion parameters were also included asnuisance regressors. A unique estimated impulse responsefunction for each regressor of interest was determined bya weighted linear sum of sine basis functions, multipliedby data determined least squares estimated beta weights.The estimated impulse response function reflects the esti-mated BOLD response to a type of trial after controllingfor variations in the BOLD signal due to other regressors.We specified the duration of the estimated response (ASprep and response: 20 sec; VGS: 24 sec post stimulusonset), a sufficient time window for the hemodynamicresponse to peak and return to baseline [Boynton, Engel,Glover, & Heeger, 1996; Buckner, 1998; Dale & Buckner,1997]. This procedure produced one-time course estimateper voxel per condition of interest. No assumptions aboutthe specific shape of the BOLD response were madebeyond using zero as the start point. The finite impulseresponse method and deconvolution allowed for com-parison of the shapes of the estimated time courses atseveral different time points in order to choose the peaktime point. This approach allowed us to characterizepotential differences (if any) in the shape of the responserelated to development or ASD that an assumed responseshape via the classical hemodynamic response modelingmight miss. Goodness-of-fit statistics were calculatedincluding partial F-statistics for each regressor andt-scores comparing each of the 5 estimated beta weights(from each basis function) with zero.

For group analyses, impulse response function valuesassociated with AS prep and response epochs from eachsubject were entered into voxel-wise linear mixed effectsmodels (one for prep and one for saccade), with “sub-jects” as a random factor and time course values (prepand saccade: 6 time points) as a within-group factor, and

“age-group” (adolescents, adults) and “diagnosis group”(patients, controls) as between-group fixed factors. Theresulting statistical map produced “main effect of time”maps that were used as base images from which func-tional regions of interest (ROIs) were defined (Figures 3and 5). The “main effect of time” map reveals all theregions that demonstrate a significant modulation ofsignal from baseline across groups and conditions withineach epoch, making this approach unbiased with respectto the effects of interest across groups. Prior researchsuggests that this method is reliable in delineating thebasic circuitry recruited for the AS task Velanova et al.,2008, Geier et al., 2010. We also ran a second linearmixed effects model including estimated time pointsassociated with correct VGS trials only in order to delin-eate the circuitry involved in eye movement control.

The main effect of time map was corrected for multiplecomparisons using a combination of cluster size and indi-vidual voxel probabilities, with the parameters deter-mined by a Monte Carlo simulation using AFNI’sAlphaSim program. This analysis specified that 23 con-tiguous voxels (27 mm3 voxels) along with a single-voxelthreshold of P < 0.001 was required to achieve a cor-rected, cluster-level alpha value of 0.05, and these regionswere designated as clusters. Next, peak voxels withinthese significant clusters that were more than 12 mmapart in the corrected main effect of time map were iden-tified using an automatic search algorithm. Twelve-millimeter diameter spheres were then centered on thesepeak voxels, resulting in a “sphere map.” Finally, a con-junction of the “sphere map” and the corrected maineffect of time map yielded a functional ROI map, witheach ROI consisting of at least 23 contiguous voxels. Theresulting ROI map was used as a mask for subsequentlyextracting time course values for each participant.

We focused our analyses on functionally defined clus-ters that fell within the boundaries of a priori anatomicalregions of interest that are known to be involved inoculomotor control. These included the SEF adjacent tothe paracentral lobule, the FEF on the superior aspect ofthe precentral sulcus, the PEF on the inferior parietalsulcus, the putamen, the ACC, and the DLPFC [Curtis &Connolly, 2008; Luna et al., 1998] (Table 2).

Mean estimated time courses from each participantwere extracted from each ROI of interest. The values fromthese mean estimated time courses at each time point foreach subject response were entered into a repeated mea-sures analysis of variance (ANOVA) using age group anddiagnosis group as between-subjects factors and time as awithin-subject factor. Below, we report regions that dem-onstrated significantly different modulations across timeby age group, diagnosis group, and/or an age group bydiagnosis group interaction. We also ran regression analy-ses with peak time course values for each subject responseusing age as a continuous variable, and diagnosis group as


a categorical variable and report any significant effects asa linear function of age below. In Figures 4 and 6, weshow graphs of average peak time course values for eachgroup.

ResultsBehavior

We ran an ANOVA with two factors (age group and diag-nosis) on performance (percent correct values). There wasa significant main effect of age group, and an age groupby diagnosis interaction. The CON group performed sig-nificantly better than the ASD group (t40 = −2.325,P = 0.025). Collapsed across diagnosis groups, there wasno effect of age group (t40 = 1.999, P = 0.052). Plannedcomparison two tailed t-tests revealed that AS perfor-mance improved significantly between CON adolescentsand CON adults (t21 = 2.384, P = .027) but not betweenASD adolescents and ASD adults (t17 = −.122, P = .904).Adolescents in the ASD group performed similarly to theCON adolescent group (t18 = −.192, P = .85). In contrast,adults with ASD performed significantly worse thanadults in the typical group (t20 = −3.356, P = .003), consis-tent with prior findings [Agam et al., 2010; Luna et al.,2007; Mosconi et al., 2009; Thakkar et al., 2008] (Fig. 2).When considering age as a continuous variable, there wasno significant effect of age or diagnosis group, or age bydiagnosis group interactions. No differences were foundbetween any of the groups in VGS error rates.

Imaging Results

VGS. All groups similarly recruited a widely distributedset or regions during the VGS trials including the corticaleye fields (FEF, SEF, and PEF) and subcortical regions(caudate, putamen, thalamus). There were no diagnosisor age group effects.

AS response preparation. During the preparatoryperiod of correct whole trials and all catch trials, all fourparticipant groups similarly recruited a distributed set ofbrain regions known to support AS response preparation,including the FEF, SEF, DLPFC and IPL (Fig. 3). Acrossdiagnosis groups, in the SEF, adults demonstrated greateractivation than adolescents (F6,240 = 2.219, P = .042)(Fig. 4). In the left FEF, a significant age group by diagno-sis group interaction (F6,228 = 2.468, P = .024) revealedthat the ASD adolescents demonstrated reduced activa-tion relative to the ASD adults (F6,102 = 2.687, P = .018)and CON adolescent (F6,108 = 2.925, P = .011) groups(Fig. 4). Secondary regression analyses confirmed the ageeffect in SEF (β = 0.015, t = 2.433, P = 0.020), showinglinear increases in activation with age. In the left FEF,there were no significant effects when considering age asa continuous variable.

AS response. All groups recruited a set of regionsknown to be involved in making a correct AS responseincluding the FEF, SEF, IPL, DLPFC, and striatum (Fig. 5).A significant age group effect was observed in SEF(F8,304 = 3.305, P = .001), with adults demonstratingincreased activation relative to adolescents in both ASDand typical groups (Fig. 6I). A significant age group effectwas also observed in left FEF (F8,304 = 2.080, P = .038),again with CON and ASD adults demonstrating increasedactivation relative to CON and ASD adolescents (Fig. 6I).In left precuneus, there was a significant main effect ofdiagnosis (F8,304 = 2.532, P = .011), revealing that indi-viduals with ASD displayed increased activation relativeto the typical groups (Fig. 6II). In the left putamen, adiagnosis group by age group interaction (F8,304 = 2.452,P = .014 ) indicated that adolescents with ASD demon-strated increased activation relative to typical adolescents

Table 2. Region Locations and Characteristics for SignificantRegions of Interest During Preparatory and OutcomeAnti-saccade (AS) Epochs

Region BA

Talairach coordinates

Epoch Max Fx y z

L. FEF 6 22 5 55 Prep 57.1SEF 6 1 −1 49 Prep 49.8L. FEF 6 22 11 55 Outcome 51.2SEF 6 4 −1 49 Outcome 49.9R. IPL 39, 40 −32 59 40 Outcome 50.1L.Precuneus 17, 18 −1 62 28 Outcome 27.6L.Putamen N/A 19 −4 7 Outcome 29.3

FEF, frontal eye field; SEF, supplementary eye field; IPL, inferior parietallobule; IPL, inferior parietal lobule; L, left; R, right; BA, Brodmann’s Area.

Figure 2. Behavioral results. Anti-saccade (AS) performance (%correct) for adolescents (left) and adults (right) in both CON (solidline, open circles) and ASD groups (dashed line, closed circles).*P < .05. Error bars denote standard error.


and adults with ASD (Fig. 6III). In right IPL, there was asignificant age group by diagnosis group interaction(F8,304 = 11.819, P = .0000), revealing an age-relatedincrease in activation in the CON group but not in theASD group (Fig. 6III). There were no significant effectswhen considering age as a continuous variable in SEF, FEF,IPL, precuneus, or putamen.

Secondary analyses. We ran secondary analyses usingequal groups (eight participants in each group) that werematched for age and IQ to ensure that our findings for ASperformance and in the right IPL were not due differencesin group sizes. These results revealed a significant effect ofgroup on performance (F3,28 = 3.504, P = 0.028). Post-hoctests confirmed that adult ASD participants performedsignificantly worse than adult CON (t14 = −2.867,P = 0.012). Adolescent ASD participants performed simi-larly to Adolescent CON (t14 = 0.096, P = 0.925) and AdultASD participants (t14 = .118, P = 0.908). Adolescent CONparticipants performed significantly worse than AdultCON participants (t14 = 2.689, P = 0.018). In right IPL, wealso found a significant age-group by diagnosis interac-tion (F8,224 = 7.33, P = 0.01).

Discussion

We aimed to characterize age-related differences ininhibitory control in ASD during adolescence and

Figure 3. Activation map for main effect of time during responsepreparation epoch collapsed across age group and diagnosis group.Threshold set at P < .001(corrected). Right side of image = rightbrain.

Figure 4. Peak values from time courses during response prepa-ration. Error bars represent + /− 1 standard error of the mean. Forvisualization purposes, filled black circles indicate location of themasks above slices of the Analysis and Visualization of FunctionalNeuroimages (AFNI) Talairach atlas, drawn using AFNI. The circlesdo not reflect the actual shape of the mask.

Figure 5. Activation map for main effect of time during anti-saccade (AS) response epoch collapsed across age group and diag-nosis group. Threshold set at P < .001(corrected). Right side ofimage = right brain.


adulthood. We used the AS task, which reliably shows aprotracted maturation through adolescence and a behav-ioral deficit in ASD [Goldberg et al., 2002; Luna et al.,2001, 2007]. Overall, our findings suggest that braindevelopment continues between adolescence and adult-hood in individuals both with and without ASD, asexpected on the basis of behavior.

While the ASD adolescent group performed at equiva-lent levels to the CON adolescent group, there were nochanges between adolescence and adulthood in ASD,which was unlike our previous behavioral findings [Lunaet al., 2007]. The CON groups showed characteristic age-related improvements in AS performance (e.g. [Geier,et al., 2010; Luna et al., 2001; Padmanabhan, 2011;Velanova et al., 2008]). The lack of age-related improve-ment in behavior in the ASD group suggests that devel-opment is delayed or arrested. The current studyemployed an interleaved paradigm, mixing VGS trialswith AS trials, adding a task-switching component to thetask. Prior research has suggested that mixing VGS andAS trials results in worse AS performance relative to pureblocks of AS trials [Cherkasova, Manoach, Intriligator, &Barton, 2002; Manoach et al., 2002; Reuter, Herzog, &Kathmann, 2006]. This added complexity may have par-ticularly affected age-related improvements in individu-als with ASD and suggests particular limitations inhigher order cognitive processes may be delayed orarrested.

Contrary to prior research showing brain activationdifferences during VGS trials in ASD relative to TD[Takarae, Minshew, Luna, & Sweeney, 2007], we found noVGS differences across diagnostic groups. This differencefrom previous work may be due to the fact that this taskhad a basic sensorimotor task (VGS) in the context of amore cognitively demanding inhibition task, whichengages the same sensorimotor regions, but may requiremore effort in processing. Our findings of similar VGSperformance and activation patterns between ASD andTD mirror recent findings demonstrating that, in thecontext of AS trials, VGS performance and brain activa-tion is similar in ASD [Agam et al., 2010]. Prior researchhas also suggested that cortical eye movement controlregions such as the FEF are recruited to a higher degree forinhibitory processes (AS) relative to sensorimotor pro-cesses (VGS), and disruptions in FEF and pre-supplementary motor area (SMA) have been found toimpair inhibitory control while keeping simple saccadicprocesses intact [Muggleton, Chen, Tzeng, Hung, & Juan,2010].

All participant groups recruited a largely similarnetwork of relevant brain regions during both the prepa-ration and response components of AS trials, includingthe FEF, SEF, IPL, ACC, and DLPFC, suggesting that thecore circuitry required for making inhibitory responses isin place by adolescence and is functional in ASD.

Figure 6. Peak values from time courses showing regions thatdemonstrated an effect during anti-saccade (AS) response. Seematerials and methods for how data were extracted. Error barsrepresent + /− 1 standard error of the mean. For visualizationpurposes only, filled black circles indication location of the masksare schematically shown above slices of the Analysis and Visual-ization of Functional Neuroimages (AFNI) Talairach atlas, drawnusing AFNI. The circles do not reflect the actual shape of the mask.


However, there were distinct age and diagnosis groupdifferences in activation during both the preparatory andsaccade response phases of the task that are discussed indetail below.

Three major patterns of differences were observed inbrain activation during AS trials relative to baseline fixa-tion. First, results revealed that the right IPL demon-strated an age group by diagnosis group interaction,mirroring the behavioral findings, providing evidence fordistinct developmental trajectories between ASD andtypically developing individuals. Second, the ASD groupsdemonstrated increased activation in precuneus relativeto typical groups, suggesting increased reliance onattention-specific mechanisms to perform the task. Third,results showed that—in adulthood—both ASD andtypical groups demonstrated increased recruitment rela-tive to their adolescent counterparts in FEF and SEF, sug-gesting parallel age related change between diagnosisgroups.

Although the typical groups demonstrated an age-related increase in lateral IPL activation, the ASD groupsdid not. This finding mirrored the behavioral findings ofa lack of developmental improvement in AS performancein ASD relative to typically developing individuals. Thelateral IPL has been previously associated with AS genera-tion and in transforming sensory information to a motorplan [Barash & Zhang, 2006]. The ASD groups recruitedthis region to a lesser degree than the typical groupsduring correct anti-saccades, which may point to ageneral deficit in sensorimotor programming resulting inmore errors overall. These results remained significanteven with smaller group sizes, suggesting that they areunlikely to be due to unequal groups.

Furthermore, the groups with ASD demonstratedincreased recruitment of left precuneus relative to thetypical groups during the generation of a voluntaryresponse. The precuneus has direct connections toregions of the IPL that are involved in visuo-spatial pro-cessing [Andersen, Asanuma, Essick, & Siegel, 1990;Leichnetz, 2001; Selemon & Goldman-Rakic, 1988], aswell as the SMA and dorsal premotor areas [Cavada &Goldman-Rakic, 1989; Petrides & Pandya, 1984] includ-ing oculomotor regions such as the FEF [Cavanna &Trimble, 2006; Leichnetz, 2001; Leichnetz & Goldberg,1988; Leichnetz & Gonzalo-Ruiz, 1996; Tian & Lynch,1996a, 1996b]. Evidence suggests that the precuneus isengaged when attending to visuospatial tasks [Bermanet al., 1999] and the response phase of the AS task[Brown, Goltz, Vilis, Ford, & Everling, 2006; Brown, Vilis,& Everling, 2007]. In the current study, increased relianceon the precuneus suggests the ASD group may rely onattention-related circuitry to generate a voluntaryresponse as a compensatory mechanism for possible limi-tations in utilizing the cortical eye fields, specifically theIPL. The precuneus is also part of the putative “default-

mode” network (a network of regions that are suppressedduring goal-directed processes) [Raichle & Snyder,2007]. Thus, it is also possible that a failure to attenuatedefault mode activation hampers inhibitory controlperformance.

Both ASD and typical groups showed age-relatedincreased activation in SEF during response preparation,and SEF and FEF during the generation of the AS. Single-cell recordings demonstrate that neurons in the SEF andFEF are active prior to making correct AS responses, sug-gesting that these regions support planned responses andthat response preparation is an essential component inthe AS task [Schlag & Schlag-Rey, 1987]. Furthermore, theSEF and FEF have connections to oculomotor regions inthe superior colliculus, with direct routes for planningeye movements [De Weijer et al., 2010; Everling &Munoz, 2000; Huerta & Kaas, 1990; Shook, Schlag-Rey, &Schlag, 1990]. Neurons in the FEF are also engaged duringresponse preparation, supporting the ability to inhibitsaccade motor neurons in the superior colliculus, thusstopping the prepotent motor response to the target[Munoz & Everling, 2004; Schall, Stuphorn, & Brown,2002].

Our previous developmental studies have shownincreased recruitment of SEF and FEF with age acrossblocks of AS trials, paralleled with developmental differ-ences in performance [Luna et al., 2001]. Although priorevent-related fMRI studies have demonstrated that ado-lescents do not differ from adults in recruitment of cor-tical eye fields during correct trials [Velanova et al., 2008],the added demand of task-switching in the current taskmay have tapped into a process that is not yet mature intypical adolescent populations. Furthermore, the parallelsin age-related increases in the recruitment of FEF and SEFduring correct AS responses from adolescence to adult-hood in both typical and ASD groups suggest that somebrain maturational processes underlying AS performancemay be preserved in ASD.

In the FEF, adolescents with ASD demonstrateddecreased preparatory activation relative to the otherthree participant groups. This suggests a maturationaldelay in recruiting key inhibitory regions required toperform successful anti-saccades, as ASD adults showedthe same magnitude of activation as CON adults.However, given that this activation difference in the FEFwas found during correct AS trials and adolescents withASD demonstrated the same level of performance astypical adolescents, compensatory mechanisms may havebeen recruited to perform at equivalent levels. Adoles-cents with ASD demonstrated increased engagement ofputamen during correct AS trials relative to adults withASD and typical adolescents. The putamen has directprojections to the FEF and is integral to oculomotor pro-cessing [O’Driscoll et al., 1995; Petit et al., 1993].Increased activation in the putamen in the adolescent


ASD group may reflect a compensatory process in order toovercome possible delayed development in cortical (i.e.,FEF) control in order to generate a successful AS response.There is little fMRI evidence showing activation differ-ences in putamen in ASD, with one study linking reducedactivation in putamen with repetitive behavior symp-toms and verbal fluency in ASD [Kenworthy et al., 2013].However, converging lines of research from PET andstructural MRI have suggested that the putamen may be alocus of dysfunction in ASD. Adults with ASD have alsodemonstrated decreased glucose metabolism in putamen[Haznedar et al., 2006] and decreased white matter con-nectivity strength between putamen and prefrontalcortex in ASD relative to typical individuals [Langenet al., 2011]. Individual differences in putamen shapehave previously been correlated with deficits in motorskills in ASD [Qiu, Adler, Crocetti, Miller, & Mostofsky,2010] and differences in putamen volume with repetitiveand stereotyped behavior [Estes et al., 2011, Hollander etal., 2005]. There is however, a lack of developmentalstudies in this area, with one study showing differences inrates of change over age in putamen volume in individu-als with ASD relative to typically developing individualsbetween childhood and early adulthood [Langen et al.,2009].

Conclusions

The present study was a cross-sectional study. Our resultsas such provide indirect evidence of developmentalchanges, which in the future may benefit from a longi-tudinal approach with a larger cohort especially givenknown variation in this spectrum. Furthermore, thenumber of trials per condition limited our ability to lookat AS error versus correct trials. Given prior results sug-gesting limitations in AS circuitry in ASD, such as thedorsal ACC [Agam et al., 2010] that are involved in errorprocessing over adolescence, an interesting future direc-tion would be to explore this further in ASD. While thesample size is also a limitation, given the scarcity ofdevelopmental studies of inhibitory control in ASD thatcurrently exist in the literature, the results from thepresent study can inform future work in this area.

We demonstrate that inhibitory control is affected inASD and that age related change in both brain functionand behavior from adolescence to adulthood may bedistinct relative to typically developing individuals.These results suggest that there are underlying limitationsin ASD in the ability to recruit the optimal circuitry tosupport inhibitory control. Furthermore, individualswith ASD may not show the optimal transitions in brainfunction through adolescence evident in typical matura-tion. Lastly, much of the circuitry supporting cognitivecontrol was intact in ASD changes. Taken together, the

results suggest that the well-delineated circuitry underly-ing the AS task is recruited in ASD, but specific refine-ments in the underlying circuitry evident in typicallydeveloping individuals may be compromised in ASD. Fur-thermore, age related change in AS performance andunderlying brain function in key areas such as the FEFand IPL may be affected in ASD.

Acknowledgments

We thank Melanie Wilds for assistance with participantrecruitment and data collection. We also thank all par-ticipants and families who volunteered for this study.

References

Agam, Y., Joseph, R.M., Barton, J.J., & Manoach, D.S. (2010).Reduced cognitive control of response inhibition by the ante-rior cingulate cortex in autism spectrum disorders.Neuroimage, 52, 336–347. doi: 10.1016/j.neuroimage.2010.04.010

Andersen, R.A., Asanuma, C., Essick, G., & Siegel, R.M. (1990).Corticocortical connections of anatomically and physiologi-cally defined subdivisions within the inferior parietal lobule.The Journal of Comparative Neurology, 296, 65–113.

Barash, S., & Zhang, M. (2006). Switching of sensorimotor trans-formations: Antisaccades and parietal cortex. Novartis Foun-dation Symposium, 270, 59–71.

Berman, R.A., Colby, C.L., Genovese, C.R., Voyvodic, J.T., Luna,B., et al. (1999). Cortical networks subserving pursuit andsaccadic eye movements in humans: An fMRI study. HumanBrain Mapping, 8, 209–225.

Bjorklund, D.F., & Harnishfeger, K.K. (1995). The evolution ofinhibition mechanisms and their role in human cognitionand behavior. In F.N. Dempster & C.J. Brainerd (Eds.), Inter-ference & inhibition in cognition (pp. 141–173). San Diego,CA: Academic Press, Inc.

Boynton, G.M., Engel, S.A., Glover, G.H., & Heeger, D.J. (1996).Linear systems analysis of functional magnetic resonanceimaging in human V1. The Journal of Neuroscience, 16,4207–4221.

Brown, M.R., Goltz, H.C., Vilis, T., Ford, K.A., & Everling, S.(2006). Inhibition and generation of saccades: Rapid event-related fMRI of prosaccades, antisaccades, and nogo trials.Neuroimage, 33, 644–659.

Brown, M.R., Vilis, T., & Everling, S. (2007). Frontoparietal acti-vation with preparation for antisaccades. Journal of Neuro-physiology, 98, 1751–1762.

Buckner, R.L. (1998). Event-related fMRI and the hemodynamicresponse. [Clinical Trial Randomized Controlled Trial].Human Brain Mapping, 6, 373–377.

Cavada, C., & Goldman-Rakic, P.S. (1989). Posterior parietalcortex in rhesus monkey: II. Evidence for segregatedcorticocortical networks linking sensory and limbic areaswith frontal lobe. The Journal of Comparative Neurology,287, 422–445.

Cavanna, A.E., & Trimble, M.R. (2006). The precuneus: A reviewof its functional anatomy and behavioural correlates.


[Research Support, Non-U.S. Gov’t Review]. Brain: A Journalof Neurology, 129(Pt 3), 564–583.

Cherkasova, M.V., Manoach, D.S., Intriligator, J.M., & Barton,J.J.S. (2002). Antisaccades and task-switching: Interactions incontrolled processing. Experimental Brain Research, 144,528–537.

Cox, R.W. (1996). AFNI: Software for analysis and visualizationof functional magnetic resonance neuroimages. Computersand Biomedical Research, 29, 162–173.

Curtis, C.E., & Connolly, J.D. (2008). Saccade preparation signalsin the human frontal and parietal cortices. Journal of Neuro-physiology, 99, 133–145.

Dale, A.M., & Buckner, R.L. (1997). Selective averaging of rapidlypresented individual trials using fMRI. Human BrainMapping, 5, 329–340.

De Weijer, A.D., Mandl, R.C., Sommer, I.E., Vink, M., Kahn, R.S.,& Neggers, S.F. (2010). Human fronto-tectal and fronto-striatal-tectal pathways activate differently during anti-saccades. Frontiers in Human Neuroscience., 4, 41.

Dempster, F.N. (1992). The rise and fall of the inhibitory mecha-nism: Toward a unified theory of cognitive development andaging. Developmental Review, 12, 45–75.

Estes, A., Shaw, D. W., Sparks, B. F., Friedman, S., Giedd, J. N.,Dawson, G., . . . Dager, S. R. (2011). Basal ganglia morphom-etry and repetitive behavior in young children with autismspectrum disorder. Autism research : official journal of theInternational Society for Autism Research, 4, 212–220. doi:10.1002/aur.193

Everling, S., Dorris, M.C., & Munoz, D.P. (1998). Reflex suppres-sion in the anti-saccade task is dependent on prestimulusneural processes. Journal of Neurophysiology, 80, 1584–1589.

Everling, S., & Munoz, D.P. (2000). Neuronal correlates for pre-paratory set associated with pro-saccades and anti-saccades inthe primate frontal eye field. Journal of Neuroscience, 20,387–400.

Everling, S., Spantekow, A., Krappmann, P., & Flohr, H. (1998).Event-related potentials associated with correct and incorrectresponses in a cued antisaccade task. Experimental BrainResearch, 118, 27–34.

Fischer, B., Biscaldi, M., & Gezeck, S. (1997). On the develop-ment of voluntary and reflexive components in humansaccade generation. Brain Research Brain Research Reviews,754, 285–297.

Fukushima, J., Hatta, T., & Fukushima, K. (2000). Developmentof voluntary control of saccadic eye movements. I. Age-related changes in normal children. Brain & Development,22, 173–180.

Geier, C.F., Terwilliger, R., Teslovich, T., Velanova, K., & Luna, B.(2010). Immaturities in reward processing and its influenceon inhibitory control in adolescence. Cereb Cortex, 20, 1613–1629. doi: 10.1093/cercor/bhp225

Gitelman, D.R. (2002). ILAB: A program for postexperimentaleye movement analysis. Behavior Research Methods, Instru-ments, & Computers: A Journal of the Psychonomic Society,Inc, 34, 605–612.

Goldberg, M.C., Lasker, A.G., Zee, D.S., Garth, E., Landa, R.J., &Tien, A. (2002). Deficits in the initiation of eye movements inthe absence of a visual target in adolescents with high func-tioning autism. Neuropsychologia, 40, 2039–2049.

Goldberg, M.C., Mostofsky, S.H., Cutting, L.E., Mahone, E.M.,Astor, B.C., et al. (2005). Subtle executive impairment in chil-dren with autism and children with ADHD. Journal of Autism& Developmental Disorders, 35, 279–293.

Haznedar, M.M., Buchsbaum, M.S., Hazlett, E.A., LiCalzi, E.M.,Cartwright, C., & Hollander, E. (2006). Volumetric analysisand three-dimensional glucose metabolic mapping of thestriatum and thalamus in patients with autism spectrumdisorders. American Journal of Psychiatry, 163, 1252–1263.

Hollander, E., Anagnostou, E., Chaplin, W., Esposito, K.,Haznedar, M. M., Licalzi, E., . . . Buchsbaum, M. (2005). Stria-tal volume on magnetic resonance imaging and repetitivebehaviors in autism. Biol Psychiatry, 58, 226–232.

Huerta, M.F., & Kaas, J.H. (1990). Supplementary eye field asdefined by intracortical microstimulation: Connections inmacaques. The Journal of Comparative Neurology, 293, 299–330.

Jenkinson, M., & Smith, S. (2001). A global optimisation methodfor robust affine registration of brain images. Medical ImageAnalysis, 5, 143–156.

Kana, R.K., Keller, T.A., Minshew, N.J., & Just, M.A. (2007).Inhibitory control in high-functioning autism: Decreasedactivation and underconnectivity in inhibition networks.Biological Psychiatry, 62, 198–206.

Kenworthy, L., Yerys, B.E., Weinblatt, R., Abrams, D.N., &Wallace, G.L. (2013). Motor demands impact speed of infor-mation processing in autism spectrum disorders. Neuropsy-chology, 27, 529–536. doi: 10.1037/a0033599

Klein, C., & Foerster, F. (2001). Development of prosaccade andantisaccade task performance in participants aged 6 to 26years. Psychophysiology, 38, 179–189.

Langen, M., Leemans, A., Johnston, P., Ecker, C., Daly, E.,Murphy, C.M., . . . Murphy, D.G. (2011). Fronto-striatal cir-cuitry and inhibitory control in autism: Findings from diffu-sion tensor imaging tractography. Cortex; a journal devotedto the study of the nervous system and behavior. doi:10.1016/j.cortex.2011.05.018

Langen, M., Schnack, H.G., Nederveen, H., Bos, D., Lahuis, B.E.,et al. (2009). Changes in the developmental trajectories ofstriatum in autism. Biological Psychiatry, 66, 327–333.

Leichnetz, G.R. (2001). Connections of the medial posterior pari-etal cortex (area 7 m) in the monkey. [Comparative Study].The Anatomical Record, 263, 215–236.

Leichnetz, G.R., & Goldberg, M.E. (1988). Higher centers con-cerned with eye movement and visual attention: Cerebralcortex and thalamus. [Review]. Reviews of OculomotorResearch, 2, 365–429.

Leichnetz, G.R., & Gonzalo-Ruiz, A. (1996). Prearcuate cortex inthe Cebus monkey has cortical and subcortical connectionslike the macaque frontal eye field and projects to fastigial-recipient oculomotor-related brainstem nuclei. [ComparativeStudy Research Support, U.S. Gov’t, Non-P.H.S.]. BrainResearch Bulletin, 41, 1–29.

Lord, C., Risi, S., Lambrecht, L., Cook, E.H. Jr., Leventhal, B.L.,et al. (2000). The autism diagnostic observation schedule-generic: A standard measure of social and communicationdeficits associated with the spectrum of autism. Journal ofAutism and Developmental Disorders, 30, 205–223.


Lord, C., Rutter, M., & Couteur, A.L. (1994). Autism diagnosticinterview-revised: A revised version of a diagnostic interviewfor caregivers of individuals with possible pervasive develop-mental disorders. Journal of Autism and Developmental Dis-orders, 24, 659–685.

Lord, C., Rutter, M., & Goode, S. (1989). Autism diagnosticobservation schedule: A standardized observation of commu-nicative and social behavior. Journal of Autism and Develop-mental Disorders, 19, 185–212.

Luna, B., Doll, S., Hegedus, S.J., Minshew, N., & Sweeney, J.(2007). Maturation of executive function in autism. Biologi-cal Psychiatry, 61, 474–481.

Luna, B., Garver, K.E., Urban, T.A., Lazar, N.A., & Sweeney, J.A.(2004). Maturation of cognitive processes from late child-hood to adulthood. Child Development, 75, 1357–1372.

Luna, B., Thulborn, K.R., Munoz, D.P., Merriam, E.P., Garver,K.E., et al. (2001). Maturation of widely distributed brainfunction subserves cognitive development. Neuroimage, 13,786–793.

Luna, B., Thulborn, K.R., Strojwas, M.H., McCurtain, B.J.,Berman, R.A., et al. (1998). Dorsal cortical regions subservingvisually-guided saccades in humans: An fMRI study. CerebralCortex, 8, 40–47.

Manoach, D.S., Lindgren, K.A., Cherkasova, M.V., Goff, D.C.,Halpern, E.F., et al. (2002). Schizophrenic subjects show defi-cient inhibition but intact task switching on saccadic tasks.Biological Psychiatry, 51, 816–826.

Minshew, N.J., Luna, B., & Sweeney, J.A. (1999). Oculomotorevidence for neocortical systems but not cerebellar dysfunc-tion in autism. Neurology, 52, 917–922.

Mosconi, M.W., Kay, M., D’Cruz, A.M., Seidenfeld, A., Guter, S.,et al. (2009). Impaired inhibitory control is associated withhigher-order repetitive behaviors in autism spectrum disor-ders. Psychological Medicine, 39, 1559–1566.

Muggleton, N.G., Chen, C.Y., Tzeng, O.J., Hung, D.L., & Juan,C.H. (2010). Inhibitory control and the frontal eye fields. JCogn Neurosci, 22, 2804–2812. doi: 10.1162/jocn.2010.21416

Munoz, D.P., Broughton, J.R., Goldring, J.E., & Armstrong, I.T.(1998). Age-related performance of human subjects on sac-cadic eye movement tasks. Experimental Brain Research, 121,391–400.

Munoz, D.P., & Everling, S. (2004). Look away: The anti-saccadetask and the voluntary control of eye movement. NatureReviews. Neuroscience, 5, 218–228.

O’Driscoll, G.A., Alpert, N.M., Matthysse, S.W., Levy, D.L.,Rauch, S.L., & Holzman, P.S. (1995). Functional neuro-anatomy of antisaccade eye movements investigated withpositron emission tomography. Proceedings of the NationalAcademy of Sciences of the United States of America, 92,925–929.

Ollinger, J.M., Corbetta, M., & Shuldman, G.L. (2001). Separat-ing processes within a trial in event-related functional MRI:Part II. Neuroimage, 13, 218–229.

Ollinger, J.M., Shulman, G.L., & Corbetta, M. (2001). Separatingprocesses within a trial in event-related functional MRI: PartI. Neuroimage, 13, 210–217.

Ozonoff, S., & Strayer, D.L. (1997). Inhibitory function innonretarded children with autism. Journal of Autism andDevelopmental Disorders, 27, 59–77.

Ozonoff, S., Strayer, D.L., McMahon, W.N., & Filloux, F. (1994).Executive function abilities in autism and Tourette syndrome:An information processing approach. Journal of Child Psy-chology and Psychiatry, 35, 1015–1032.

Padmanabhan, A., Geier, C.F., Ordaz, S.J., Teslovich, T., & Luna,B. (2011). Developmental changes in brain function under-lying the influence of reward processing on inhibitorycontrol. Dev Cogn Neurosci, 1, 517–529. doi: 10.1016/j.dcn.2011.06.004

Petit, L., Orssaud, C., Tzourio, N., Salamon, G., Mazoyer, B., &Berthoz, A. (1993). PET study of voluntary saccadic eye move-ment in humans: Basal ganglia-thalamocortical system andcingulate cortex involvement. Journal of Neurophysiology,69, 1009–1017.

Petrides, M., & Pandya, D.N. (1984). Projections to the frontalcortex from the posterior parietal region in the rhesusmonkey. The Journal of Comparative Neurology, 228, 105–116.

Qiu, A., Adler, M., Crocetti, D., Miller, M.I., & Mostofsky, S.H.(2010). Basal ganglia shapes predict social, communication,and motor dysfunctions in boys with autism spectrum disor-der. Journal of the American Academy of Child and Adoles-cent Psychiatry, 49, 539–551.

Raichle, M.E., & Snyder, A.Z. (2007). A default mode of brainfunction: A brief history of an evolving idea. Neuroimage, 37,1073–1082.

Reuter, B., Herzog, E., & Kathmann, N. (2006). Antisaccade per-formance of schizophrenia patients: Evidence of reducedtask-set activation and impaired error detection. [ResearchSupport, Non-U.S. Gov’t]. Journal of Psychiatric Research, 40,122–130.

Russell, J. (1997). Autism as an executive disorder. New York:Oxford University Press.

Schall, J.D., Stuphorn, V., & Brown, J.W. (2002). Monitoring andcontrol of action by the frontal lobes. Neuron, 36, 309–322.

Schlag, J., & Schlag-Rey, M. (1987). Evidence for a supplemen-tary eye field. Journal of Neurophysiology, 57, 179–200.

Schmitz, N., Rubia, K., Daly, E., Smith, A., Williams, S., &Murphy, D.G. (2006). Neural correlates of executive functionin autistic spectrum disorders. Biological Psychiatry, 59, 7–16.

Selemon, L.D., & Goldman-Rakic, P.S. (1988). Common corticaland subcortical targets of the dorsolateral prefrontal and pos-terior parietal cortices in the rhesus monkey: Evidence for adistributed neural network subserving spatially guided behav-ior. Journal of Neuroscience, 8, 4049–4068.

Shook, B.L., Schlag-Rey, M., & Schlag, J. (1990). Primate supple-mentary eye field: I. Comparative aspects of mesencephalicand pontine connections. The Journal of Comparative Neu-rology, 301, 618–642.

Smith, S.M., Jenkinson, M., Woolrich, M.W., Beckmann, C.F.,Behrens, T.E., et al. (2004). Advances in functional and struc-tural MR image analysis and implementation as FSL.Neuroimage, 23(Suppl. 1), S208–S219.

Solomon, M., Ozonoff, S.J., Ursu, S., Ravizza, S., Cummings, N.,et al. (2009). The neural substrates of cognitive control defi-cits in autism spectrum disorders. Neuropsychologia, 47,2515–2526.

Solomon, M., Yoon, J.H., Ragland, J.D., Niendam, T.A., Lesh,T.A., Fairbrother, W., & Carter, C.S. (2014). The development


of the neural substrates of cognitive control in adolescentswith autism spectrum disorders. Biol Psychiatry, 76, 412–421.doi: 10.1016/j.biopsych.2013.08.036

Spear, L.P. (2000). Neurobehavioral changes in adolescence.Current Directions in Psychological Science, 9, 111–114.

Takarae, Y., Minshew, N.J., Luna, B., & Sweeney, J.A. (2007).Atypical involvement of frontostriatal systems during senso-rimotor control in autism. Psychiatry Research, 156, 117–127.

Talairach, J., & Tournoux, P. (1988). Co-planar stereotaxic atlas ofthe human brain. New York: Thieme Medical Publishers.

Thakkar, K.N., Polli, F.E., Joseph, R.M., Tuch, D.S., Hadjikhani,N., et al. (2008). Response monitoring, repetitive behaviourand anterior cingulate abnormalities in autism spectrum dis-orders (ASD). Brain: A Journal of Neurology, 131(Pt 9), 2464–2478.

Tian, J.R., & Lynch, J.C. (1996a). Corticocortical input to thesmooth and saccadic eye movement subregions of the frontaleye field in cebus monkeys. Journal of Neurophysiology, 76,2754–2771.

Tian, J.R., & Lynch, J.C. (1996b). Functionally defined smoothand saccadic eye movement subregions in the frontal eyefield of cebus monkeys. Journal of Neurophysiology, 76,2740–2753.

Velanova, K., Wheeler, M.E., & Luna, B. (2008). Maturationalchanges in anterior cingulate and frontoparietal recruitmentsupport the development of error processing and inhibitorycontrol. Cerebral Cortex, 18, 2505–2522.

Velanova, K., Wheeler, M.E., & Luna, B. (2009). The maturationof task set-related activation supports late developmentalimprovements in inhibitory control. Journal of Neuroscience,29, 12558–12567.

Ward, B.D. (2002). Deconvolution analysis of fMRI timeseries data: documentation for the AFNI software package.Available at: http://afni.nimh.nih.gov/pub/dist/doc/manual/3dDeconvolve.pdf.

Supporting information

Additional Supporting Information may be found in theonline version of this article at the publisher’s web-site:

Figure S1. Schematic of single subject and group dataanalysis steps. Images for the single subject analysis weretaken from a representative participant.Table S1. Medication status of ASD participants.


Developmental Changes in Brain Function Underlying Inhibitory ...

Documents