ERP Correlates of Emo3on Processing in Adults with Au3sm and Schizophrenia Emily Levy, Adam Naples, Jennifer Foss-­‐Feig, Rachael Tillman, Hannah Reuman, Karen Law, Hannah Samson, Caroline Diehl, Charles Schleifer, Nicole Santamauro, Alan An3cevic, Vinod Srihari & James McPartland McPartland Lab, Yale Child Study Center, New Haven, CT Method Au3sm Spectrum Disorder (ASD) and Schizophrenia (SZ) are neurodevelopmental disorders characterized by difficul3es with social cogni3on, such as eye contact, emo3on recogni3on, and theory of mind.1,2 Prior electrophysiological studies suggest atypical structural encoding of faces, as indexed by N170 amplitude and latency,3 and emo3onal processing, as indexed by P300 amplitude,4 in both clinical groups compared to typically developing (TD) adults. Despite parallel findings in separate literatures, no studies have directly compared electrophysiological correlates of emo3on processing in individuals with ASD, SZ, and TD. The current study inves3gated N170 and P300 ERP response to happy and fearful faces in adults with ASD, SZ, and TD and relates them to behavioral measures of emo3on iden3fica3on. By integra3ng eye-­‐tracking, we developed a gaze-­‐con3ngent ERP paradigm to study emo3on processing in the context of interac3ve faces that respond to par3cipant gaze. Preliminary Results Experimental Paradigm •  A gaze-­‐con3ngent EEG paradigm was used to measure emo3on processing. Upon par3cipant fixa3on to the eyes of a neutral face, it displayed an emo3onal expression (happy or fearful). •  With this design, par3cipants were fixa3ng on the eyes of the emo3onal face when it appeared onscreen. TD ASD SZ Data Collec9on and Processing •  Behavioral measures of SZ and ASD symptomatology and emo3on recogni3on collected during visit: Au3sm Diagnos3c Observa3on Schedule (ADOS), Posi3ve and Nega3ve Syndrome Scale (PANSS), and Reading the Mind in the Eyes Test (RMET). •  EEG data were recorded with 128-­‐channel HydroCel Geodesic Sensor Nets at 1000Hz. Eye-­‐tracking data were recorded concurrently with EEG using SR-­‐
Eyelink at 500Hz. Ocular ar3fact was removed from EEG data with independent component analysis. •  ERP waveforms were segmented to emo3on onset. N170 latency and mean amplitude were extracted from electrode groups over the lej and right hemisphere. P300 mean amplitude was extracted from midline parietal electrodes. •  Amplitude and latency were analyzed with Repeated Measures ANOVAs (within-­‐subject factors: Emo3on, Hemisphere; between-­‐subject factor: Group). Follow-­‐up analyses were performed to explore direc3on of interac3ons. 0 300 400 500 600 TD ASD SZ amplitude (μV) -­‐3 163ms -­‐4 150ms 155ms 1 Amplitude •  No significant effects or interac3ons in the right hemisphere. Latency •  Main effect of Group: In the right hemisphere, the SZ group showed a marginally delayed N170 response compared to the TD group. •  F(2,39) = 2.667, p = .081 TD ASD SZ 180 160 140 120 8 12 16 20 RMET score 24 28 32 N170: LeD Hemisphere 2 1 0 -­‐100 0 100 200 300 400 500 600 -­‐1 -­‐2 (#
!'#
-­‐3 !&#
!%#
-­‐4 !$#
!"#
-­‐5 %#$"
%"
0.5 !#$"
3me (ms) 0 100 200 300 400 500 600 !"
'("
)*("
*+"
-­‐0.5 200 Correla9ons •  Across groups, N170 amplitude and latency over the right hemisphere correlated with RMET score, such that a faster and more nega3ve N170 was associated with beher emo3on iden3fica3on. •  Latency: r = -­‐.410, p = .007 •  Amplitude: r = -­‐.324, p = .038 amplitude (µ V) 4. Kestenbaum, R. (1991). Neural and behavioral correlates of emo*on recogni*on in children and adults. J Exp Child Psychol, 54(1): 1-­‐18. 200 -­‐2 1. Abdi, Z., & Sharma, T. (2004). Social cogni*on and its neural correlates in schizophrenia and au*sm. CNS Spectr, 9(5): 335-­‐343. 3. Ben3n, S., Allison, T., Puce, A., Perez, E., & McCarthy, G. (1996). Electrophysiological studies of face percep*on in humans. J Cogni3ve Neurosci, 8(6): 551-­‐565. 100 -­‐1 References 2. Sasson, N., Pinkham, A., Carpenter, K., & Belger, A. (2011). The benefit of directly comparing au*sm and schizophrenia for revealing mechanisms of social cogni*ve impairment. J Neurodevelop Disord, 3: 87-­‐100. 3me (ms) N170 latency (ms) FSIQ (SD) 109.7 (21.5) 103.8 (17.5) 96.1 (12.0) 1.5 &"
,-../"
01-2345"
'67-5"
-­‐1 0 -­‐5 TD Happy TD Fearful ASD Happy ASD Fearful SZ Happy SZ Fearful 2 N170: Right Hemisphere 1 -­‐100 &#$"
0 2 Method Age (SD) 26.5 (7.2) 23.7 (5.4) 30.5 (8.6) + -­‐100 We predicted that: 1) In response to gaze-­‐con3ngent shijs in facial displays of emo3on, adults with ASD and SZ would show ahenuated and delayed right-­‐
hemisphere N170 and smaller P300 amplitude compared to TD adults, and 2) Behavioral measures of emo3on recogni3on would correlate with strength of N170 and P300 responses. N (F) 16 (3F) 12 (2F) 14 (1F) 2.5 Preliminary Results Par9cipants •  42 adults aged 18-­‐48 completed EEG, ques3onnaires, and neuropsychological assessments. •  Groups were matched on FSIQ, age, and sex. P300 amplitude (µ V) Background )*#
+,*#
Amplitude •  Main effect of Emo3on: Across groups, fearful faces elicited a less nega3ve N170 TD Happy TD Fearful response than happy faces. ASD Happy •  F(2,39) = 20.325, p < .0005 ASD Fearful SZ Happy •  Group by Emo3on interac3on: In the ASD SZ Fearful group, fearful faces elicited a less nega3ve N170 than happy faces; there were no differences by emo3on in the TD or SZ ,-#
groups. •  F(2,39) = 5.490, p = .008 Latency •  Main effect of Emo3on: Across groups, happy ./001#
faces elicited a faster N170 than fearful faces. 23/4567#
•  F(2,39) = 5.040, p = .031 -­‐1.5 -­‐2 Amplitude •  Group by Emo3on interac3on: Fearful faces elicited more robust P300 responses than happy faces in the ASD and TD groups. There was no modula3on by emo3on in the SZ group. •  F(2,39) = 4.948, p = .012 •  Main effect of Group: Across emo3ons, P300 amplitude was marginally ahenuated in the SZ group, par3cularly rela3ve to the ASD group. •  F(2,39) = 2.733, p = .077 Conclusions Summary of Findings •  A gaze-­‐con3ngent face processing paradigm revealed dis3nct neural responses to interac3ve happy and fearful emo3onal expressions. •  Over lej hemisphere, par3cipants with ASD showed modula3on of N170 amplitude by emo3on, such that fearful faces elicited an ahenuated response rela3ve to happy faces. •  Face and emo3on processing deficits were observed in the SZ group rela3ve to the TD group, as indexed by: •  Delayed right-­‐hemisphere N170 across emo3ons. •  No modula3on of P300 amplitude by emo3on. •  Correla3ons between emo3on iden3fica3on and N170 amplitude and latency suggested a rela3onship between behavior and neural response spanning diagnoses. Interpreta9ons and Future Direc9ons •  Due to the gaze-­‐con3ngent nature of the current paradigm, par3cipants were forced to make eye contact with the s3muli. Because individuals with ASD tend to make less eye contact, these s3muli may be especially salient for them, resul3ng in a heightened neural response rela3ve to passive experimental paradigms. •  Future research will con3nue to iden3fy diagnos3c convergence and divergence, as indexed by behavioral, eye-­‐tracking, and ERP measures. The ul3mate goal of this work is to iden3fy biomarkers of social performance to enable targeted treatments based on social func3on and dysfunc3on rather than diagnos3c category. Funding Sources NIMH R01 MH100173 (McPartland) NIMH R01 MH100173-­‐02S1 (McPartland) Au3sm Science Founda3on Research Enhancement Grant (Foss-­‐Feig) NARSAD Young Inves3gator Award (Foss-­‐Feig) Au3sm Speaks Transla3onal Postdoctoral Fellowship (Naples) NIMH R01 MH103831 (Srihari) Na3onal Ins3tutes of Health Grants DP50D012109-­‐02 (An3cevic) NARSAD Young Inves3gator Award (An3cevic) Yale Center for Clinical Inves3ga3on (An3cevic)