Poster #6 - Resting-state functional networks and ERP responses to auditory stimuli in 4-month-old infants

• In Event: Poster Session 13

Abstract

How the brain responds to novelty may differ as a function of early-life temperaments, such as negative reactivity and BI (Marshall et al., 2009; Barker et al., 2013). However, what remain underspecified are the mechanisms for how brain responses to novelty translate into stable temperamental traits. This is partly due to event-related potential (ERP) measures having limited spatial resolution, but it also remains unclear if these transient signals map onto individual differences in the brain’s intrinsic functional connectivity. To evaluate if this is the case, we explored the association between two ERP responses to novelty (Mismatch Negativity (MMN) and P3) and resting-state functional connectivity (rs-FC). The MMN reflects automatic change-detection of auditory input, whilst the P3 reflects attention allocation and top-down processing.
Twenty-nine infants underwent neuroimaging including fMRI (Mage= 4.77 months) and EEG (Mage= 5.04 months). The auditory oddball task was administered during EEG data collection, which includes repetitive ‘standard’ tones, infrequent ‘deviant’ tones, and ‘novel’ tones which are unique and more complex than deviant tones. MMN scores were calculated as the ERP difference between responses to deviant and standard tones, and P3 scores as the mean amplitude ERP to novel tones. Rs-FC was estimated among 200 parcels. Network-level enrichment analyses were performed to identify associations between rs-FC and MMN (Model 1) and P3 (Model 2). Bonferroni corrections were applied to correct for two models.
Rs-FC between and within several functional networks were associated with both MMN and P3, including within the control network (p<0.021), between control–limbic (p<0.002), limbic–somatosensory (p<0.017), and control–default mode networks (p<0.024). Rs-FC associated with only MMN included within-network ventral attention network (VAN) connectivity (p=0.021, Figure 1). Rs-FC associated with only P3 included connectivity with the dorsal attention network (DAN) (p<0.019) and visual central network (p<0.020, See Table 1 for details).
Findings showed decreased rs-FC within the VAN is specifically related to increased MMN response, which is in line with research indicating VAN to be activated during bottom-up processing (Onofrj et al., 2022). Further, decreased VAN rs-FC has been associated with increased behavioural inhibition and anxiety in children (Sylvester et al., 2013; 2018), whilst increased MMN has been related to childhood social withdrawal (Bar-Haim et al., 2003) and behavioural inhibition in infants (Marshall et al., 2009).
Alternatively, the DAN was shown to be specifically associated with P3 response to novel stimuli, which is in line with previous research implicating DAN in focusing attention and goal-directed, top-down processing of attention (Corbetta & Shulman, 2002). Several resting-state functional networks were significantly related to both MMN and P3, including the control, limbic, somatosensory, and default mode networks.
This work further solidifies our understanding of the different underlying networks implicated in the development of immediate responses to stimuli. Future analyses will test associations with temperament to determine the behavioural significance of individual differences in neural signal, as measured by multimodal neuroimaging techniques.

Authors

• Dana Kanel, National Institute of Mental Health; University of Maryland, College Park
  Presenting Author

• Courtney Filippi, New York University School of Medicine

• Hannah Hardiman, National Institute of Mental Health; University of Maryland - College Park

• Anderson Winkler

• Santiago Morales, University of Southern California

• Chad Sylvester

• Daniel S. Pine, National Institutes of Health

• Nathan A Fox, University of Maryland - College Park