Is reduced behavioral orienting to motherese speech in toddlers with ASD related to reduced neural responses to motherese?

In this study, for the first time, we are seeing what the possible brain impact is for toddlers with autism who fail to pay attention to social information.
Like

When interacting with a baby, most adults will slow down their speech, increase pitch and intonation, and repeat the same words in a singsong fashion over and over again, usually accompanied by exaggerated facial expression. This style of speech is termed as motherese, or infant-directed speech, and is used by parents across the globe. In fact, motherese is regarded as a key tool that may augment the human-special, caregiver–infant interactive loop and is thought to have a genetic basis that emerged evolutionarily in humans.

It has long been theorized that motherese is an experience-expectant, human-specific, and innate form of parent speech that enhances social and language learning as well as affect and emotion development in infants and toddlers. This occurs because infants are thought to have an innate neural response to motherese speech that supports the mother-infant social interactive loop. In turn, the mother-infant interactive loop increases motivation and enriches social learning.

However, within the first years of life, infants and toddlers with autism spectrum disorder (ASD) often exhibit challenges in maintaining social interactions and have been shown to pay reduced attention to social auditory information, such as a reduced orienting to their own name. Importantly, some children with ASD achieve good outcomes by the time they reach adulthood and are able to maintain close relationships and earn a college degree, while others struggle to communicate, make friends, and find employment. The reasons for such wide-ranging outcomes in ASD is unclear. We hypothesized that the varying levels of successful outcomes in ASD may be due to varying degrees of neural responding to socially relevant information during the first years of life, including to motherese speech.

To test this hypothesis, in our study, we devised a novel experiment that combined functional magnetic resonance imaging (fMRI) that measured how the toddler’s brain responds to affective speech, including motherese, during natural sleep, along with clinical measures of social and language abilities. To objectively measure each toddler’s preference for motherese speech as a measure of external validation, we used gaze contingent eye tracking, wherein the toddler’s visual point of gaze triggered either motherese speech or non-human sounds.

We first examined our brain imaging data at the group level and observed reduced neural responses to affective speech including motherese in ASD toddlers as compared to typically developing (TD) toddlers. Next, we correlated neural responses to affective speech with a toddlers’ social and communication abilities as measured by a standardized measure – the Vineland. Here we found significant relationships between overall neural responding to affective speech and social and language ability across ASD and TD toddlers.

To further examine relationships between neural responses to motherese speech and social and language abilities, we determined if there were unique neural-clinical subgroups by using Similarity Network Fusion (SNF), a powerful method for integrating multi-modality data, and then using a clustering approach to generate biological-, behavioural- or clinical-relevant subgroup clusters. Specifically, SNF combines diverse data types including functional brain imaging and in the current study, SNF combined three different speech paradigms stratified by emotional intensity and standardized clinical measures from ASD and TD toddlers. The properties of SNF make it ideal for our study to integrate multiple layers of neural functional responses to three different levels of emotional speech (totalling 200 fMRI acquisitions) including motherese and several types of symptom, cognitive, and adaptive behaviour data to form neural-clinical-relevant ASD and TD toddler subgroups.

SNF analysis revealed 4 subgroups (i.e., ‘Clusters’): two largely TD clusters and two clusters with only ASD. Toddlers who fell within Cluster 1 (12 TD/2 ASD toddlers) exhibited the strongest neural responses to motherese speech, highest levels of attention towards motherese speech as indexed by eye tracking, and strong social and language skills, while toddlers in Cluster 4 (11 ASD toddlers) had reduced neural responsiveness to motherese speech, very poor social and language abilities, and reduced eye-tracking-related attention to motherese (Figure 1). A distinctive finding among toddlers with ASD in Cluster 4 was that motherese stimuli evoked weaker neural responses than did moderately affective speech, which is in direct contrast with TD cluster 1 toddlers, who exhibited increased neural responses to motherese. TD and ASD toddlers who fell within Clusters 2 and 3 had somewhat intermediate neural–clinical phenotypes.

Figure 1. TD and ASD subgroups with distinct fMRI-clinical patterns. a, Similarity Network Fusion and Louvain algorithm revealed 4 fMRI-clinical distinct subgroups. b, The left barplot displays average percent signal change across language paradigms for each cluster. The right barplot shows differences in percent signal change between motherese vs moderate affect speech in four clusters. c, Social and language ability across TD and ASD clusters. Toddlers in TD clusters (Clusters 1 and 2) had significantly higher percentage fixation towards motherese versus non-social computer “techno” images and sounds than ASD Cluster 4 toddlers.
Figure 1. TD and ASD subgroups with distinct fMRI-clinical patterns. a, Similarity Network Fusion and Louvain algorithm revealed 4 fMRI-clinical distinct subgroups. b, The left barplot displays average percent signal change across language paradigms for each cluster. The right barplot shows differences in percent signal change between motherese vs moderate affect speech in four clusters. c, Social and language ability across TD and ASD clusters. d, Toddlers in TD clusters (Clusters 1 and 2) had significantly higher percentage fixation towards motherese versus non-social computer “techno” images and sounds than ASD Cluster 4 toddlers. 

These findings suggest varying phenotypic characteristics across TD and ASD clusters, which indicate that social preference and language development are intertwined across a wide spectrum of social and language ability and disability. Moreover, because the brain imaging data was collected during natural sleep, neural responsiveness effects were un-confounded by volition, arousal, interest, motivation, attention, awareness, expectation, and cooperation. Further, clustering results suggest that TD toddlers with high neural responses and ASD toddlers with low neural responses to motherese speech stand at opposite ends of the neural–affective–response and social–language ability spectrum. Overall, these results demonstrate the multi-dimensionality of biology and behaviour among both TD and ASD toddlers.

In sum, our study points to the early-age neural bases of the core social deficits and reduced responsiveness to parental affective speech that first emerge in infants and toddlers who develop ASD. These data offers unique evidence that neural activity elicited by affective speech such as motherese may be important in driving infants to engage with caregivers, thus facilitating social and language learning. We speculate that enhanced neural responsiveness leads to such learning, while weaker neural responsiveness may impede or preclude it. Thus, we predict that the several toddlers with ASD in our study who showed the strongest neural responses to motherese speech may go on to have more successful outcomes as they move through childhood. This hypothesis also predicts that neural and behavioural deficits together may be a biomarker of foundational dysregulation of social–emotional neural development and learning, which contributes to life-long challenges for ASD children. Finally, identification of neural–clinical–behavioural ASD subgroups may be critical for development of more targeted and effective interventions.