Lack of behavioral attention to motherese speech in ASD involves impaired development of innate temporal cortical neural systems


Motherese is a form of simplified, exaggerated melodic speech that parents use to communicate with newborns and young toddlers. A horse becomes horsie; a dog becomes doggie; parents become mama and dada. The tendency to speak in such short sing-song phrases is universal across cultures.

Previous research has shown that infants prefer to listen to motherese, more formally known as infant-directed speech, over adult-like speech; that it more effectively holds their attention and is an important component of emotional bonding and fosters learning experiences between child and parents.

An early sign of autism spectrum disorder (ASD) in children is a reduced response to motherese speech and challenges in sustained attention to social information in general.

In a new study, published January 3, 2022 in the journal Nature Human Behavior, researchers at University of California San Diego School of Medicine employed a number of techniques to pinpoint the regions of the brain responsible for a child’s response to baby talk.

“This new study, which combined state-of-the-art brain imaging, eye-tracking and clinical testing, opens the door toward precision medicine in autism,” said senior author Eric Courchesne, PhD, professor of neuroscience at UC San Diego School of Medicine.

Courchesne said the approach generates new insights into how the brain is developing in children with autism related to objective information about social preference and social attention.

“For the first time, we are seeing what the possible brain impact is for children with autism who fail to pay attention to social information,” he said.

Typically developing infants prefer motherese to other forms of adult speech, and previous studies have suggested their brains may process motherese differently from non-speech sounds. But research is scant regarding how and why infants with ASD do not consistently respond to motherese speech and what the long-term consequences might be when they “tune out.”

Courchesne, with colleagues at the Autism Center of Excellence at UC San Diego, hypothesized that ASD infants and toddlers experience impaired development of innately driven neural mechanisms that respond to motherese. To investigate, they conducted a series of tests involving 200 datasets from 71 toddlers and 41 datasets from 14 adults:

  • Using functional magnetic resonance imaging (fMRI) of sleeping toddlers, they measured brain activity to motherese and other forms of social affective speech.
  • They conducted clinical assessments of social and language development.
  • And they utilized eye-tracking technology to measure responses to females speaking motherese versus non-speech computer sounds and images. Earlier research at UC San Diego and elsewhere has shown that toddlers with ASD show less interest in social activities and stimuli that would normally attract a young child’s attention, such as watching other children play, sing or dance.

The researchers found that individual differences in early-age social and language development correlated with a child’s neural responses to speech, and that ASD infants and toddlers with the poorest neural responses to motherese also displayed the most severe social symptoms, poorest language outcomes and greatest impairment of behavioral preference and attention toward motherese.

Conversely, infants and toddlers with typical development showed the strongest neural responses and affinity to motherese.

Using a computational precision medicine method for integrating data called similarity network fusion, they correlated eye-gaze patterns to neural and behavioral responses, further confirming their findings.

The researchers noted that the superior temporal cortex, a region of the brain that processes sounds and language, responded more weakly to motherese and emotion speech in ASD children, who also had the poorest social abilities and lowest eye-tracking attention to motherese.

The opposite was true among typically developing children, who displayed strong superior temporal neural response to motherese and emotion speech. A small number of toddlers with ASD showed strong brain activation and interest in motherese speech, as determined by eye-tracking.

“Our conclusion is that lack of behavioral attention to motherese speech in ASD involves impaired development of innate temporal cortical neural systems that normally would automatically respond to parental emotional speech,” said study co-author Karen Pierce, PhD, professor of neurosciences at UC San Diego School of Medicine and co-director of Autism Center of Excellence with Courchesne.

“The fact that a few children with autism did show strong brain activation and good attention to motherese speech is encouraging for two reasons: First, because it suggests that these particular toddlers with autism are likely to have good outcomes, a newly discovered and important subgroup. And second, it suggests a novel avenue for treatment.

The authors said their findings, based upon data-driven, empirical evidence, may be useful in developing further diagnostic tools and biomarkers for early identification of ASD and in further clarifying how ASD affects toddlers in widely and dramatically different ways.

Co-authors include: Yaqiong Xiao, Teresa H. Wen, Lisa Eyler, Disha Goel and Nathan E. Lewis, all at UC San Diego; Lauren Kupis, University of Miami; Keith Vaux, UC San Diego Health Physician Network; and Michael V. Lombardo, Instituto Italiano di Tecnoligia and University of Cambridge.

Autism Spectrum Disorder (ASD) is characterised by difficulties in social communication and restrictive and repetitive behaviours (DSM-5 [2];). A range of underlying genetic and environmental aetiologies have been identified, but the pathways that link these distal causal factors to the later emergence of diagnostic symptoms remain unclear [16]. Recently, sensory symptoms have been recognised as a core part of the diagnostic profile [2].

Indeed, up to 90% of individuals with ASD report difficulties in sensory processing [65] that include both hyper and hyposensitivity to auditory, tactile, and visual stimulation [90]. Since sensory systems mature very early in postnatal development, it is possible that early sensory atypicalities have a cascading effect on developmental trajectories and contribute to later emerging behavioural symptoms [55, 68]. However, there remains little direct investigation of this possibility.

Identifying causal pathways to symptom emergence requires prospective longitudinal studies of infants with an elevated likelihood of developing ASD. One common approach has been to study infants with an older sibling with ASD [56, 84, 91]. Such studies have identified some evidence of early sensory atypicalities in infants with later ASD, including faster identification of visual differences [18, 40], slower latency of pupillary responses to luminance changes [80], increased behavioural responses to perceptual change [19], and elevated cortical reactivity to repeated sounds [60].

This range of work suggests that early disruptions in sensory processing may be detectable in infants with later ASD prior to the onset of other behavioural symptoms, consistent with a causal model. Observations from familial designs, however, are limited to the 10–20% of children whose ASD is associated with the accumulation of multiple common genetic variants of small effect [64, 94]. It is not currently clear whether similar effects are present in the 5–11% of autistic children who present with a monogenic or more penetrant cause of ASD [108], or indeed in idiopathic cases of ASD where there is a non-familial route to the disorder.

A complementary approach to familial designs is thus to study infants who have an elevated likelihood of developing ASD due to the presence of a monogenic disorder that can be identified in infancy. A strong candidate monogenic condition associated with ASD is Neurofibromatosis Type 1 (NF1), an autosomal dominant neurocutaneous disorder with a birth incidence of 1:2700 [28].

Fifty percent of cases of NF1 are inherited, while the rest arise de novo due to spontaneous loss-of-function mutation of the NF1 gene located on chromosome 17q11.2 [20]. The NF1 gene encodes for neurofibromin, a large 2818-amino acid negative RAS GTPase-regulating protein [97, 98].

Although the physical phenotype of NF1 may include neurofibromas, café-au-lait macules, Lisch nodules and abnormalities within the skeleton and the central nervous system [54], the main challenges reported by parents and children with NF1 in clinical settings are cognitive, social and behavioural difficulties [44, 75]. Indeed, up to 25% of children with NF1 may meet criteria for ASD, up to 45% may experience broader autism symptomatology [34, 36, 87, 109] and up to 50% receive a diagnosis of ADHD [53, 61].

NF1 is a suitable monogenic condition for studying early developmental pathways to ASD for several key reasons. First, the phenotypic profile of ASD in NF1 is broadly similar to idiopathic ASD [37], with a similar male bias in prevalence of ASD [35], making insights from NF1 more likely to be generalisable to the understanding of ASD as a whole.

Second, NF1 is typically identified early in development through either cord blood testing in familial cases or through its cutaneous manifestations in both inherited and de novo cases (particularly café-au-lait spots). This makes prospective studies from early infancy feasible. Third, NF1 is not associated with profound developmental delays, but rather with a more subtle shift in IQ [53].

Due to this, comparisons with typically developing infants are less confounded by selection bias and developmental challenges within the NF1 group [3]. Finally, there are good animal models of NF1 that may facilitate the subsequent investigation of neural mechanisms underlying particular phenotypes in human infants [22, 43, 103]. Such investigations may highlight new paths in animal-to-human translation.

One promising domain of investigation in infants with NF1 is low-level auditory processing. Important to pathways to translational research, auditory paradigms can be meaningfully reproduced across the lifespan and in animal models [8, 99] where the importance of auditory processing is high across species (unlike visual processing, which is far more important in primates than in rodents).

In particular, the suppression of neural responses following repetition and an increase in response when change is detected are suitable for studying auditory brain development across the lifespan [76, 86]. Failure to attenuate responses to repetition or to respond selectively to a change in auditory input could compromise language development [6, 7] and may also relate to broader aspects of cognitive inflexibility, which has been noted in several neurodevelopmental disorders including ASD, Prader-Willi, Rett Syndrome and Fragile X [2, 66, 70, 81]. Further, failure in these basic learning mechanisms may indicate alterations in neural organisation, consistent with observations of atypical neural connectivity in developmental disorders as well as differences in the co-ordination of excitation and inhibition [24, 27, 67, 92].

Previous work has indeed identified alterations in auditory responses in children with NF1. A study of 22 children and adults with NF1 found that while peripheral acoustic hearing was within the normal range, differences emerged in temporal auditory processing during standardised tests, including phonological processing and temporal resolution [4].

Difficulties in auditory processing were further associated with degree of language impairment and communication disorders in the sample. Additionally, Chaix et al. [14] found that children with NF1 (n = 75) scored lower on a phoneme deletion task, indicating impaired phonological processing. However, there has been no work on early auditory development in infants with NF1, and no efforts to examine the relationship between auditory processing and the presence of ASD within the sample.

One common method that has been used to measure responses to auditory repetition and change in the developing brain is electroencephalography, or EEG. EEG is suitable for infants and children of all ages because it is relatively non-invasive and does not require verbal or behavioural responses. The high temporal resolution of EEG also allows it to accurately capture the time-course of neural correlates of auditory processing.

EEG studies have shown that neural responses to auditory stimuli decrease with repetition and that the magnitude of this effect increases with age [14]. A related phenomenon is sensory gating, in which a pair of stimuli are presented in quick succession after a period of silence and the second stimulus elicits a smaller neural response than the first [45]. Importantly, individual differences in responses to repetition have been related to future cognitive abilities of the infant [69], as well as neurodevelopmental conditions such as ASD [46, 60, 67, 100]. Atypical neural responses to repetition in ASD have also been linked to severity of behavioural symptoms [83].

Another fundamental feature of low-level auditory processing is the ability to detect when changes occur in a sequence of repeated tones. The mismatch negativity is a frontocentral negative deflection in an event-related waveform, elicited by subtracting responses to an infrequent “deviant” from a repeatedly presented “standard” tone [5, 17, 48, 62, 63, 96]. Two- to 4-month-old infants show age-related differences in responses to deviant versus standard tones, with a slow positive wave at 2 months that becomes an adult-like negativity in 3–4-month-olds [48].

When embedded within a train of repeating stimuli, infants have shown sensitivity to both pitch [32, 48, 110] and frequency [9, 82] change, which have been associated with increasing brain specialisation to language processing [7]. Dysregulation of deviance detection, as both suppression or enhancement of the event-related response, has been described as atypical neurological function [79].

More importantly, these early differences have been associated with downstream effects on social and non-social sound processing and language production [26, 49, 113]. Taken together, auditory habituation and deviance detection could provide important insights into basic perceptual processing in infants with NF1, and help us understand whether disrupted low-level auditory processing is predictive of ASD diagnosis in toddlerhood or appear as a cumulative risk factor across several neurodevelopmental conditions [29, 30, 46, 58].

In the current study, we examined age-related changes in auditory repetition suppression and change detection responses in infants with typical development (TD) and those diagnosed with NF1. We assessed infants at 5 and 10 months, which is a sensitive period for auditory development, as well as the comprehension and production of speech [95, 111]. Further, these processes occur prior to the onset of behavioural symptoms associated with neurodevelopmental conditions including ASD and ADHD, which holds great promise for early detection and implementation of effective interventions during the period of high brain plasticity.

We presented infants with trains composed of three repeated vowels and then one deviant (either a change in the Vowel category or a change in its pitch), separated by an inter-train interval jittered between 3 and 5 seconds (simplified from a more complex design [23]). We indexed repetition suppression by comparing neural responses to the first and the second standard, and change detection by comparing the second standards with the deviants (reducing the influence of orienting effects from the first standard and preparatory effects from the third).

As responses to the third standard could have additional influences from a recovery response [85], changes between the second and third repeated standard stimulus were examined as a secondary question in the SM. In some previous studies examining auditory processing and habituation the decrease in neural response between the first and second standards has been referred to as the fast decay response [93].

In the typically developing population, we expected a clear reduction in response between repetition of the first and second standard vowel sound at 5 and 10 months of age. We additionally expected a stronger change detection response (increased ERP negativity) towards pitch change at 5 months, and towards Vowel category change at 10 months, to reflect increasing specialisation to language processing [7].

Further, we predicted age-related changes in the localisation of repetition suppression and change detection responses between 5 and 10 months, to reflect increasing specialisation for sound processing towards the mature frontally driven response [88]. We predicted that infants with NF1 would show reduced or absent repetition suppression and change detection responses relative to typically developing infants at both 5 and 10 months of age.

Finally, we examined whether individual differences in auditory change responses related to variation in three key phenotypes at a follow-up visit at 14 months: (a) language ability [107]; (b) ASD-relevant early behaviours as measured by the Autism Observation Scale for Infants [11]; and (c) early ADHD-relevant behaviours as measured by the Infant Behaviour Questionnaire-Revised [89], as a test of specificity [102];). We predicted there would be a positive association between a weaker repetition suppression and deviance detection response, reduced language ability and ASD-relevant behaviours.

reference link :


Please enter your comment!
Please enter your name here

Questo sito usa Akismet per ridurre lo spam. Scopri come i tuoi dati vengono elaborati.