Class Notes (807,344)
United States (312,496)
Psychology (228)
Pascal (8)

Foss Feig Autism.pdf

7 Pages
Unlock Document

New York University

TheJournalofNeuroscience,May8•33(19):8243–8249 • 8243 Behavioral/Cognitive ASubstantialandUnexpectedEnhancementofMotion PerceptioninAutism JenniferH.Foss-Feig, *DujeTadin, *KimberlyB.Schauder, andCarissaJ.Cascio4 2,4 1DepartmentofPsychologyandHumanDevelopment,and VanderbiltKennedyCenter,VanderbiltUniversity,Nashville,Tennessee37203, Centerfor3 VisualScience,DepartmentofBrainandCognitiveSciencesandDepartmentofOphthalmology,UniversityofRochester,Rochester,NewYork14627,and 4DepartmentofPsychiatry,VanderbiltUniversity,Nashville,Tennessee37212 Atypical perceptual processing in autism spectrum disorder (ASD) is well documented. In addition, growing evidence supports the hypothesisthatanexcitatory/inhibitoryneurochemicalimbalancemightunderlieASD.Hereweinvestigatedputativebehavioralconse- quencesoftheexcitatory/inhibitoryimbalanceinthecontextofvisualmotionperception.Asstimulussizeincreases,typicalobservers exhibit marked impairments in perceiving motion of high-contrast stimuli. This result, termed “spatial suppression,” is believed to reflect inhibitory motion-processing mechanisms. Motion processing is also affected by gain control, an inhibitory mechanism that underliessaturationofneuralresponsesathighcontrast.Motivatedbythesebehavioralcorrelatesofinhibitoryfunction,weinvestigated motion perception in human children with ASD (n ⫽ 20) and typical development (n ⫽ 26). At high contrast, both groups exhibited similar impairments in motion perception with increasing stimulus size, revealing no apparent differences in spatial suppression. However,therewasasubstantialenhancementofmotionperceptioninASD:childrenwithASDexhibitedaconsistenttwofoldimprove- ment in perceiving motion. Hypothesizing that this enhancement might indicate abnormal weakening of response gain control, we repeatedourmeasurementsatlowcontrast,wheretheeffectsofgaincontrolshouldbenegligible.Atlowcontrast,weindeedfoundno groupdifferencesinmotiondiscriminationthresholds.Theselow-contrastresults,however,revealedweakerspatialsuppressioninASD, suggesting the possibility that gain control abnormalities in ASD might have masked spatial suppression differences at high contrast. Overall,wereportapatternofmotionperceptionabnormalitiesinASDthatincludessubstantialenhancementsathighcontrastandis consistentwithanunderlyingexcitatory/inhibitoryimbalance. (Plaisted et al., 1998; Joseph et al., 2009), have been observed. Introduction Dating from the earliest clinical observations (Kanner, 1943), However,theseenhancementsarelargelyspecifictoprocessingof atypical perceptual processing has been documented widely in static stimuli. The integrity of motion processing has been con- autism spectrum disorder (ASD; Ben-Sasson et al., 2009). Per- troversial. Whereas some studies suggest widespread impair- ments indicating broad dorsal stream abnormalities (Spencer et ceptual differences likely play an important role in complex behavioral, social, and cognitive deficits that define ASD al., 2000; Pellicano et al., 2005), other research argues deficits (Zwaigenbaum et al., 2005; Mottron et al., 2006; Samson et al., occur only with “complex” moving stimuli such as biological 2011). Therefore, working toward a mechanistic understanding motion (Bertone et al., 2003; Kaiser and Shiffrar, 2009). A parallel line of research has focused on neurobiological ofperceptualalterationsmightrevealkeyinsightsintotheneural underpinnings of ASD. Much attention has focused on visual causes of the ASD phenotype. One prominent model posits that perception in ASD (Simmons et al., 2009). Both visual impair- animbalancebetweenexcitatoryandinhibitoryneuralmechanisms ments, such as face perception deficits (Simmons et al., 2009), might explain widespread abnormalities in ASDR(ubenstein and and enhancements, such as superior perception of local visual Merzenich,2003).Indeed,genetic,neuroanatomical,andneuro- features (Dakin and Frith, 2005) and enhanced visual search imagingevidencesupportsanincreaseinneuralexcitability.This changeappearstobedrivenbyreducedefficacyoftheGABAergic ReceivedMarch29,2012;revisedMarch23,2013;acceptedApril1,2013. system (Buxbaum et al., 2002; Fatemi et al., 2009; Oblak et al., Authorcontributions:J.H.F.-F.,D.T.,andC.J.C.designedresearch;J.H.F.-F.andK.B.S.performedresearch;J.H.F.-F.,D.T.,ayalsocontribute K.B.S.,andC.J.C.analyzeddata;J.H.F.-F.,D.T.,andC.J.C.wrotethepaper. ThisworkwassupportedbytheNationalEyeInstitute–NationalInstitutesofHealth(CoreGrantsP30EY0013192). and P30 EY08126, Grant R01 EY019295 to D.T., and Grant K01-MH090232 to C.J.C.) and the National Center for motion perception, we explored ResearchResources–NationalInstitutesofHealth(Grant1UL1RR024975).J.H.F.-F.wassupportedbyanAutismgnatures of excitatory/inhibitory imbal- SpeaksDennisWeatherstonePredoctoralFellowship. ance.Specifically,weinvestigatedeffectsofstimulussizeandcon- Theauthorsdeclarenocompetingfinancialinterests. *J.H.F.-F.andD.T.contributedequallytothiswork. trast on motion perception in ASD. We showed previously that, Correspondence should be addressed to Jennifer H. Foss-Feig, 1601 23 Ave South, PMB #552, Nashville, TNsize increases, motion direction of 37212.E-mail:[email protected] high-contrast patterns becomes markedly harder to perceive DOI:10.1523/JNEUROSCI.1608-12.2013 (Tadin et al., 2003). This behavioral result, termed “spatial sup- Copyright©2013theauthors 0270-6474/13/338243-07$15.00/0 pression,” is believed to reflect center-surround inhibition, likely 8244• J.Neurosci.,May8,•33(19):8243–8249 Foss-Feig,Tadinetal.•EnhancedMotionPerceptioninAutism Table1.Participantdemographics Groupmeans Statistics Variable ASD TD tstatisticp-value High-contrastexperiment Age(years) 12.7 ⫾ 2.9 12.4 ⫾ 3.1 0.31 0.76 Full-scaleIQ 116.7 ⫾ 12.4 107.9 ⫾ 15.1 1.78 0.09 ADOSalgorithmtotal 15.7 ⫾ 4.6 — — — ADI-Rsummaryscore 44.5 ⫾ 10.5 — — — Low-contrastexperiment Age(years) 11.4 ⫾ 2.2 10.7 ⫾ 2.4 0.75 0.46 Full-scaleIQ 111.4 ⫾ 20.5 116.7 ⫾ 18.2 ⫺0.66 0.52 ADOSalgorithmtotal 12.6 ⫾ 3.3 — — — ADI-Rsummaryscore 42.3 ⫾ 13.3 — — — aAutismDiagnosticObservationSchedule(ADOS)algorithmtotalscore(Gothametal.,2007)summarizingcommu nication,social,andrepetitivebehaviordomains. bAutismDiagnosticInterview-Revised(ADI-R)scorecollapsedacrossalgorithmdomainsforcommunication,recip rocalsocialinteraction,andrepetitivebehavior. within cortical area MT (Tadin et al., 2003; Churan et al., 2008; Tadin et al., 2011). Abnormally weak spatial suppression, re- flectedinreducedeffectsofincreasingstimulussize,characterizes populations exhibiting deficits in cortical inhibition; for exam- ple, patients with schizophrenia (Tadin et al., 2006b), those with Figure1. Stimuliandtask.A,Sequenceofeventsconstitutingasingletrial.B,Space–time a history of major depression (Golomb et al., 2009), and older illustrationshowingasmallstimulus(1°)movingtotheright.Thedepictedduration(25.6ms) adults (Betts et al., 2005; Betts et al., 2009). Given evidence for an equalstheaverageASDthresholdforthisconditionathighcontrast(Fig.2A).C,Stimulussizes excitatory/inhibitory imbalance in ASD, we expected similar re- usedinthestudy(radius⫽1°,2.5°,and6°).Onlyonestimuluswasshownpertrial. sults in our population. Motionprocessingisalsoaffectedbyresponsegaincontrol,an contrast was 98%. For all stimuli, the temporal envelope was a hybrid- Gaussian (Tadin et al., 2011). Specifically, for very brief stimuli ( inhibitory mechanism underlying saturation of neural responses at high contrasts (Albrecht and Hamilton, 1982; Katzner et al., ms), the temporal contrast envelope was Gaussian. Longer temporal en- 2011). To detect possible abnormalities in response gain control, velopes were trapezoid-like: flanks were half-Gaussians and the central portion was set to the maximum contrast. Fine temporal precision was we examined motion direction discriminations at both low and obtainedbyadjustingtheSDofhalf-Gaussianflanks(withaconstraintof high contrast. If gain control mechanisms are impaired in ASD (Pei et al., 2012), we expected that increasing stimulus contrast ␴ ⬍ 15 ms) and transferring “excess” contrast to the flat central portion. Thishybridenvelopeallowedfinetemporalprecisionofbriefstimuliand would have larger effects on participants with ASD. Finally, fo- avoided protracted fade-in/fade-out periods associated with prolonged cusing on motion perception enabled us to address outstanding temporalGaussians.Gaussianflanksallowsubframesampling(atempo- questions about the integrity of motion processing in ASD. ralequivalentofsubpixelsampling),permittingaccuratepresentationof briefstimuliusingonlyafewmonitorframes(Tadinetal.,2006a;Lappin MaterialsandMethods et al., 2009). Stimulus duration was defined as the width at half-height of Subjects. We studied children and adolescents with ASD and typically the temporal envelope. developing (TD) controls between 8 and 17 years of age. A total of 20 The participants’ task was to discriminate motion direction of briefly participants with ASD (all male) and 26 TD controls (21 male) partici- presented moving gratings (Fig. 1A,B). For each trial, a stimulus was pated in the study. Thirty-two participants completed the high-contrast presented at the central location, with its size chosen pseudorandomly. experiment (15 ASD and 17 TD); 23 participants completed the low- Participants indicated the perceived direction (left or right) by a key contrast experiment (10 ASD and 13 TD). Five and four children over- press. Auditory feedback followed correct responses. The subsequent lapped across experiments for the ASD and TD groups, respectively. For trial started 1 s after each response. Participants were instructed to fixate both experiments, ASD and TD participants were matched for age and on the center of the screen. To facilitate fixation, we used the following IQ score, as measured by the Wechsler Abbreviated Scales of Intelligence sequence: a fixation circle (0.8° radius) appeared after each key press (Weschsler, 1999; Table 1). For children with ASD, diagnoses were con- response and, 500 ms later, the circle shrank to 0.13° over 100 ms, re- firmed with the Autism Diagnostic Observation Schedule (Lord et al., mained at that size for 250 ms, and then disappeared 150 ms before 2000) and the Autism Diagnostic Interview-Revised (Lord et al., 1994); stimulus onset. Based on our subjective impressions and feedback from all participants in the ASD group met the diagnostic criteria for both child participants, this dynamic sequence was very effective in guiding measures.Descriptivestatisticsfordiagnosticmeasuresalsoarereported eye gaze to the center of the screen before the onset of a static fixation in Table 1. Children in the TD group were free of psychiatric, learning, cross. It is important to note that no dynamic fixation stimuli appeared and neurological disorders and had no first-degree relatives with ASD. within the 400 ms preceding stimulus onset; a conventional fixation The protocol was approved by the Vanderbilt University Institutional target followed by a blank screen was presented during that time. Review Board. All participants and their parents gave written informed Three conditions, varying in stimulus size, were used (Fig. 1C): small assentandconsent,respectively,andallwerepaidfortheirparticipation. (1° radius), medium (2.5° radius), and large (6° radius). These stimulus High-contrast experiment. Stimuli were created in MATLAB and Psy- sizes were selected on the basis of previous results (Tadin et al., 2003) chophysicsToolbox(Brainard,1997)andshownonalinearizedmonitor with the goal of including both small stimuli for which performance is (24-inch Sony GDM-FW900 CRT, 1024 ⫻ 640 resolution, 120 Hz). high and large stimuli for which spatial suppression reduces motion Viewing was binocular at 77 cm and was enforced using a chin rest. sensitivity. Task difficulty was controlled by adjusting the stimulus du- Ambient and background illumination were 0.9 and 56.3 cd/m 2. ration to evaluate the minimum stimulus presentation duration for Stimuli were drifting gratings (Fig. 1; 1 cycle/degree, 4°/s, starting which each participant could reliably judge motion direction at a given phase randomized) presented in a stationary two-dimensional raised stimulus size. Specifically, psychophysical thresholds were measured by cosine envelope, the radius of which defined the stimulus size. Stimulus adaptive QUEST staircases that adjusted log10(stimulus duration) and Foss-Feig,Tadinetal.•EnhancedMotionPerceptioninAutism J.Neurosci.,May8,20•33(19):8243–8249 • 8245 stimuli (Tadin et al., 2003; Betts et al., 2005; Tadin et al., 2006b; Betts et al.,2009;Tadinetal.,2011).Specifically,suppressionindex⫽log 10(large stimulus threshold) ⫺ log (10all stimulus threshold). Note that this definition of the suppression index assumes that the thresholds are ex- pressed in milliseconds (as reported in the Results section). Using this convention,highersuppressionindexvaluescorrespondtostrongerspa- tialsuppression(i.e.,worseningofperformancewithincreasingstimulus size), whereas negative values indicate spatial summation (i.e., an im- provement with increasing size). Typically, spatial suppression is found at high contrast and its strength decreases as the contrast decreases. At low contrast, spatial summation is typically found (Tadin et al., 2003). It is worth noting that the suppression index is not a log-transform mea- sure, but rather is simply the difference between the two thresholds. The QUEST staircase procedure is designed to “work” in the log space (Watson and Pelli, 1983), directly estimating log 10hreshold). In other words, data presented here are not log transformed, but rather were natively estimated in the log space (for ease of interpretation, thresholds are reported in milliseconds in the Results section and were obtained by calculating the 10^threshold). Before conducting the statistical analysis, we confirmed that datasets did not significantly deviate from the normal distribution and that vari- ancesofASDandTDdatasetsdidnotsignificantlydifferfromeachother. To analyze duration threshold results for each experiment, group differ- ences were evaluated using a 2 (group) ⫻ 3 (stimulus size) repeated- measures ANOVA. Greenhouse-Geisser correction for nonsphericity was used for low-contrast data only. Follow-up analyses with IQ score and age entered as covariates yielded similar results to the primary mo- tion discrimination and suppression index analyses, with all group dif- Figure2. Motiondiscriminationperformanceandspatialsuppressionstrengthathighcontrast. remaining highly significant. Results also held when female participantswereeliminatedfromthecontrolgrouptomatchthegroups A, Comparison of duration thresholds for discriminating high-contrast motion for gender. To analyze the results between low- and high-contrast exper- ASDandTD.ResultsrevealasubstantialenhancementofmotionperceptioninASDthatisindiments,weconducteda2(group)⫻2(contrast)ANOVA,withrepeated, dent of stimulus size. The insert shows data for participants who completed both low- and high- contrast experiments.B, Suppression index scores (computed as the10ifference of logdurationsmeasuresoncontrast.Forthisanalysis,onlythesubsetof thresholdsforlargeversussmallmotionstimuli)donotdifferbetweenASDandTDgroups.ErrorbarsTD participants that completed both high- and low-contrast experiments was included. Finally, to evaluate the clinical relevance of inAandBrepresentSEM.C,Box-and-whiskerplotsshowingrobustseparationofresultsforourfindings,exploratorycorrelationalanalyseswereconductedtoexam- pants with ASD and TD. Thick lines indicate median performance. Boxes indicate the interquartile rangeandwhiskersshowthedatarangeexclusiveofoutliers.Outliers(filledcircles)aredefinedasdatask performance was related to clinical symptoms in the points⬎1.5timestheinterquartilerangebeyondthefirstandthethirdquartiles.Foreachstimulus. size, the median TD performance approximately equals the threshold exhibited by the worst- performingparticipantwithASD.Notethediy-taxisrangefortherightmostplot. Results High-contrast experiment convergedto82%correct(WatsonandPelli,1983).Aftera66-trialprac- Forhigh-contraststimuli,boththeTDandASDgroupsexhibited tice block, thresholds were obtained in three successive 132-trial blocks, the characteristic increase in direction discrimination thresholds with increasing stimulus size, a result indicative of spatial sup- separated by breaks. Each block contained six interleaved staircases and yielded two thresholds for each of three size conditions (i.e., small, me- pression of large moving stimuli (Fig. 2A; F (2,30)⫽ 98.8, p ⬍ dium,andlarge).Therefore,acrossthreeblocksoftrials,weestimated18 10 ⫺9 ). Specifically, as stimulus size increased, participants in thresholds (i.e., six for each stimulus size). For each participant, the bothgroupsrequiredlongerstimulusexposuretimestocorrectly lowest and highest of the six duration thresholds for each stimulus size identify motion direction. This threshold increase with increas- were eliminated; the average of the remaining four measurements con- ing stimulus size, however, did not differ between groups stituted a threshold estimate. This helped to eliminate occasional ex- treme values. Participants completed all trials within a single session (F (2,30) 0.72, p ⫽ 0.49); both groups showed equal spatial sup - (lasting ⬃25 min) and were monitored by an examiner throughout. pression strength. This result was further supported by a between-groups comparison of the suppression index (Fig. 2B; Low-contrast experiment. The low-contrast experiment was identical to the above-described high-contrast experiment, except that the stimu- t30⫺ 0.18, p ⫽ 0.86). luscontrastwasloweredto2.9%.Themotivationforthisexperimentwas The surprising result was the substantial superiority of partic- to determine whether the results we found at high contrast, namely a ipants with ASD in their ability to discriminate motion direction pronounced ASD enhancement in motion perception (Fig. 2), could be (Fig. 2A; F (1,30)⫽ 19.2, p ⫽ 0.0001). Specifically, across all stim - generalized to low-contrast stimuli. See Discussion for additional ulus sizes, we found a consistent, twofold decrease in thresholds information. for participants with ASD. This ASD advantage was considerable Data analysis. First, to assess the reliability of our measurements, for and was not driven by outliers: for all three stimulus sizes, the each of six conditions (two experiments with three size conditions in third quartile for the ASD group was approximately equal to the each), we computed split-half reliabilities for the ASD and TD groups. first quartile for the TD group (Fig. 2C). Moreover, the perfor- Theaveragesplit-halfreliabilitiesforparticipantswithASDandTDwere 0.989 and 0.984, respectively. These high interparticipant correlations mance of the worst participant with ASD approximately equaled showthatourapproachwasveryreliableatmeasuringmotionsensitivity themedianTDperformanceforeachstimulussize(Fig.2C).The main finding from the high-contrast experiment was that partic- in children and adolescents. To quantify spatial suppression strength, we calculated the suppres- ipants with ASD were markedly better at discriminating motion sion index, defined as the difference of thresholds for large versus small direction across all stimulus sizes. 8246• J.Neurosci.,May8,•33(19):8243–8249 Foss-Feig,Tadinetal.•EnhancedMotionPerceptioninAutism Figure4. EffectsofcontrastonmotiondirectiondiscriminationsforparticipantswithTDand Figure3. MotiondiscriminationperformanceandspatialsummationstrengthatlowASD. A, Effect of increasing stimulus contrast on direction discrimination thresholds for small trast.A,Comparisonofdurationthresholdsfordiscriminatinglow-contrastmotionmovingstimuli.ThisplotisconstructedbycombiningtheleftmostdatapointsinFigure2Aand pants with ASD and TD. Unlike high-contrast results, there are no substantial differences betweengroupsexceptanincreaseintheTD(butnotASD)thresholdsforthelargeststimuluse3Aandreplottingthesedataasafunctionofstimuluscontrast.B,SameasinA,except size(shownbythearrow).Theinsertshowsdataforparticipantswhocompletedbothlow-andnlydatafromparticipantswhocompletedbothlow-andhigh-contrastexperimentsare included.ErrorbarsinbothpanelsrepresentSEM. high-contrastexperiments.B,Suppressionindexscoreswerenegativeforbothgroups,indicat- ing spatial summation at low contrast. However, participants with TD also exhibited signs of spatialsuppression(A,arrow).Thiswasevidentasasignificantdifferenceinsuppressionindex scorebetweenthetwogroups.ErrorbarsinbothpanelsrepresentSEM. est stimulus size, a size at which the effects of spatial suppression should be minimal regardless of contrast level (Tadin et al., Low-contrast experiment 2003). This allowed us to better isolate contrast-dependent For low-contrast stimuli, we saw the characteristic decrease in mechanisms. For the smallest size, the two groups were nearly directiondiscriminationthresholdsw⫺5hincreasingstimulussize identical in their duration thresholds at low contrast (Fig. 4A). (Fig. 3A; F (2,21) 43.2, p ⫽ 10 ). This result indicates spatial Increasing contrast improved performance for both groups, but summation,animprovementinmotionperceptionwithincreas- this improvement was considerably stronger for participants ing stimulus size (Anderson and Burr, 1991; Tadin et al., 2003). with ASD than for those with TD (40.8 vs 19.1 ms, respectively). Contrary to our findings at high contrast, we found no overall As a result, at high contrast, participants with ASD performed group differences in participants’ ability to discriminate motion substantially better than those with TD. Comparing group per- direction at low contrast (F ⫽ 0.83, p ⫽ 0.37). In another formance across contrast levels, participants with ASD exhibited (1,21) greaterperformancegainswithincreasingcontrast.Forstatistical contrast-specific finding, the effect of increasing stimulus size differed between the two groups, as indicated by a significant analysis, we only considered participants who completed both group by stimulus size interaction (F (2,21)⫽ 6.50, p ⫽ 0.011). experiments (Fig. 4B). These results were consistent with the full ParticipantswithTDexhibitedthebeginningsignsofspatialsup- dataset, revealing a significant main effect of contrast (F ⫽ ⫺7 (1,7) pression, reflected in a threshold increase for the largest stimulus 534.3, p ⫽ 10 ) and a significant group-by-contrast intera
More Less

Related notes for PSYCH-UA 25

Log In


Don't have an account?

Join OneClass

Access over 10 million pages of study
documents for 1.3 million courses.

Sign up

Join to view


By registering, I agree to the Terms and Privacy Policies
Already have an account?
Just a few more details

So we can recommend you notes for your school.

Reset Password

Please enter below the email address you registered with and we will send you a link to reset your password.

Add your courses

Get notes from the top students in your class.