Skip to main content
NCBI home page
JAMA Network logoLink to JAMA Network
. 2023 Feb 8;6(2):e2255125. doi: 10.1001/jamanetworkopen.2022.55125

Level of Attention to Motherese Speech as an Early Marker of Autism Spectrum Disorder

Karen Pierce 1,✉, Teresa H Wen 1, Javad Zahiri 1, Charlene Andreason 1, Eric Courchesne 1, Cynthia C Barnes 1, Linda Lopez 1, Steven J Arias 1, Ahtziry Esquivel 1, Amanda Cheng 1
PMCID: PMC9909502  PMID: 36753277

This diagnostic study investigates whether attention to high-affect motherese speech can be used as a diagnostic marker of autism spectrum disorder and language and social ability among toddlers aged 12 to 48 months.

Key Points

Question

Can levels of attention toward motherese speech, indexed through eye tracking, be used as a diagnostic marker of autism spectrum disorder (ASD), and are they associated with social and language abilities?

Findings

In this diagnostic study of 653 toddlers aged 12 to 48 months, levels of fixation on motherese speech were high in toddlers without ASD but significantly reduced in toddlers with ASD. If a toddler fixated on motherese speech at or below 30%, the probability of that toddler being accurately identified as ASD was 94% and also signified an association with reduced social and language abilities.

Meaning

The findings suggest that levels of attention to motherese speech may be a diagnostic and prognostic marker of ASD.

Abstract

Importance

Caregivers have long captured the attention of their infants by speaking in motherese, a playful speech style characterized by heightened affect. Reduced attention to motherese in toddlers with autism spectrum disorder (ASD) may be a contributor to downstream language and social challenges and could be diagnostically revealing.

Objective

To investigate whether attention toward motherese speech can be used as a diagnostic classifier of ASD and is associated with language and social ability.

Design, Setting, and Participants

This diagnostic study included toddlers aged 12 to 48 months, spanning ASD and non-ASD diagnostic groups, at a research center. Data were collected from February 2018 to April 2021 and analyzed from April 2021 to March 2022.

Exposures

Gaze-contingent eye-tracking test.

Main Outcomes and Measures

Using gaze-contingent eye tracking wherein the location of a toddler’s fixation triggered a specific movie file, toddlers participated in 1 or more 1-minute eye-tracking tests designed to quantify attention to motherese speech, including motherese vs traffic (ie, noisy vehicles on a highway) and motherese vs techno (ie, abstract shapes with music). Toddlers were also diagnostically and psychometrically evaluated by psychologists. Levels of fixation within motherese and nonmotherese movies and mean number of saccades per second were calculated. Receiver operating characteristic (ROC) curves were used to evaluate optimal fixation cutoff values and associated sensitivity, specificity, positive predictive value (PPV), and negative predictive value. Within the ASD group, toddlers were stratified based on low, middle, or high levels of interest in motherese speech, and associations with social and language abilities were examined.

Results

A total of 653 toddlers were included (mean [SD] age, 26.45 [8.37] months; 480 males [73.51%]). Unlike toddlers without ASD, who almost uniformly attended to motherese speech with a median level of 82.25% and 80.75% across the 2 tests, among toddlers with ASD, there was a wide range, spanning 0% to 100%. Both the traffic and techno paradigms were effective diagnostic classifiers, with large between-group effect sizes (eg, ASD vs typical development: Cohen d, 1.0 in the techno paradigm). Across both paradigms, a cutoff value of 30% or less fixation on motherese resulted in an area under the ROC curve (AUC) of 0.733 (95% CI, 0.693-0.773) and 0.761 (95% CI, 0.717-0.804), respectively; specificity of 98% (95% CI, 95%-99%) and 96% (95% CI, 92%-98%), respectively; and PPV of 94% (95% CI, 86%-98%). Reflective of heterogeneity and expected subtypes in ASD, sensitivity was lower at 18% (95% CI, 14%-22%) and 29% (95% CI, 24%-34%), respectively. Combining metrics increased the AUC to 0.841 (95% CI, 0.805-0.877). Toddlers with ASD who showed the lowest levels of attention to motherese speech had weaker social and language abilities.

Conclusions and Relevance

In this diagnostic study, a subset of toddlers showed low levels of attention toward motherese speech. When a cutoff level of 30% or less fixation on motherese speech was used, toddlers in this range were diagnostically classified as having ASD with high accuracy. Insight into which toddlers show unusually low levels of attention to motherese may be beneficial not only for early ASD diagnosis and prognosis but also as a possible therapeutic target.

Introduction

Caregivers have long captured the attention of their infants and toddlers by speaking in motherese, a playful speech style characterized by exaggerated intonation contours, simple grammar, high pitch, and slow tempo, also known as parentese or infant-directed speech.1 Infants prefer to listen to this speech style relative to adult-directed speech from the first days of life,2 a preference ubiquitous across cultural and geographical boundaries,3 suggesting a possible genetic basis.4 Studies have noted that this highly affective speech style stimulates joint attention and learning and is associated with improved language acquisition, affective engagement, and emotional reactivity.1,5,6

It has long been known that children with autism spectrum disorder (ASD) show unusual and sometimes absent responses to auditory information in their environment. For example, children may fail to respond when their name is called7,8 or exhibit a poor understanding of word meanings.9,10 At the neural level, studies commonly report significantly reduced functional brain activity10,11,12,13 or delayed timing in response to speech sounds.14,15,16 Using behavioral techniques such as turning the head toward sounds emitted from a tape recorder, Kuhl and colleagues17 showed that 70% of children with ASD preferred to listen to computer-generated sounds over motherese, a finding that has been replicated.18,19 Yet, reduced response to motherese speech is not commonly reported among children with non-ASD delays, suggesting possible specificity for ASD and utility as a diagnostic marker.20

In studies of typically and atypically developing children, associations between exposure to motherese speech and early word recognition,5 mean length of utterance,21 and receptive language ability22 have been reported. Given the experience-dependent mechanisms that support learning during the first years of life,23,24,25,26 infants who do not pay attention to motherese speech—or human speech in general—would likely have more impaired language ability than children who pay close attention to social speech. In a brain imaging study, toddlers with ASD who showed the lowest levels of attention to motherese speech also showed the lowest levels of neural functional activation in speech-processing regions and lower language abilities than toddlers who paid greater attention to motherese speech.27

Eye tracking is a relatively new approach for rapidly and accurately characterizing attention in ASD28,29,30,31,32,33,34,35,36,37 and holds promise as a screening and diagnostic tool.31,33 Eye-tracking studies, however, have been largely restricted to the visual domain and have demonstrated reduced attention to social visual images in ASD.30,38,39,40,41,42,43,44 Although there are some exceptions,45 few studies have leveraged eye-tracking technology to understand attention patterns within the auditory domain, particularly in response to speech sounds. In theory, this could be achieved through gaze-contingent technology wherein the location where a toddler looks on a monitor triggers specific speech and nonspeech sounds.

While visual attention is commonly indexed as a percent fixation metric (eg, percent fixation time on faces), other eye-tracking metrics, such as saccades, may also provide useful information. Unique saccade profiles can be associated with levels of anxiety46,47 but are not necessarily correlated with overall fixation quantities. For example, studies have shown dissociations between saccade and fixation levels in ASD, with some toddlers with ASD who exhibit high levels of fixation on nonsocial images also showing low saccade rates.31,32 Examining saccade metrics in conjunction with traditional percent fixation metrics could provide new insight regarding how toddlers with ASD visually and auditorily examine their world. This may be essential for improving the diagnostic classification accuracy of eye-tracking tests and help to uncover unique biological subtypes of ASD. The aim of this study was to investigate whether level of attention toward motherese speech can be used as a diagnostic classifier of ASD and is associated with language and social ability.

Methods

This diagnostic study was approved by the institutional review board at the University of California San Diego (UCSD). Parents gave written informed consent, and all testing occurred at the UCSD Autism Center. The study followed the Standards for Reporting of Diagnostic Accuracy (STARD) reporting guideline (eMethods in Supplement 1). Race and ethnicity information was obtained via parent report and was included as required for federal grants; toddlers’ race and ethnicity were categorized per National Institutes of Health definitions. Categories for race were African American/Black, American Indian/Alaska Native, Asian, Native Hawaiian/Pacific Islander, White, more than 1 race, and not reported, unknown, or other. Ethnicity was categorized as Hispanic or Latino, not Hispanic or Latino, and unknown or not reported.

Participants and Clinical Testing

Toddlers were recruited through 2 mechanisms: community referrals (eg, website) or a general population-based screening method called Get SET Early,48 using the Communication and Symbolic Behavior Scales Developmental Profile Infant-Toddler Checklist,49,50 that results in the detection of ASD as young as 12 months.48,51 Toddlers younger than 30 months who were referred were invited for a reevaluation every 9 to 12 months until their third birthday, when a final diagnosis was given.

Toddlers aged 12 to 48 months participated in 1 or more motherese eye-tracking tests: motherese vs traffic (ie, noisy, moving vehicles on a highway), motherese vs techno (ie, moving shapes with musical sounds), and motherese vs flat affect. Across all tests, 11% of the sample was excluded for a range of reasons such as inattention to the video. Details regarding recruitment, exclusion, and testing overlap are given in the eMethods and eFigures 1 to 3 in Supplement 1.

Toddlers and their parents participated in a series of tests including the Autism Diagnostic Observation Schedule (ADOS),52 the Mullen Scales of Early Learning,53 and the Vineland Adaptive Behavior Scales.54 All assessments were administered by ADOS-certified, licensed clinical psychologists (C.C.B. and S.J.A.) blind to eye-tracking results.

Based on diagnoses given at the final diagnosis age, toddlers were categorized into 5 contrast groups (ASD, ASD features, delay, typical development [TD], and typical sibling of an ASD proband). Details regarding diagnostic criteria and clinical scores are given in the Table and eFigure 4 and the eMethods in Supplement 1.

Table. Clinical Characteristics of Toddlers Who Completed 1 or More Motherese Eye-Tracking Tests.

Characteristic Toddlers (N = 653)a
ASD (n = 422) ASD features (n = 29) Delay or other (n = 75) TD (n = 115) TypSibASD (n = 12)
Sex
Female 100 (23.70) 9 (31.03) 26 (34.67) 34 (29.57) 4 (33.33)
Male 322 (76.30) 20 (68.97) 49 (65.33) 81 (70.43) 8 (66.67)
Age, mean (SD), mo
Motherese vs traffic 26.71 (7.38) 26.55 (9.54) 22.21 (8.67) 25.44 (8.52) 22.33 (9.85)
Motherese vs techno 26.94 (7.81) 27.04 (9.08) 24.07 (9.93) 25.35 (8.51) 23.33 (11.45)
Motherese vs flat affect 28.01 (8.66) 29.26 (8.81) 24.13 (10.04) 28.21 (8.15) 23.13 (9.69)
Ethnicity
Hispanic or Latino 169 (40.05) 10 (34.48) 30 (40.00) 15 (13.04) 1 (8.33)
Not Hispanic or Latino 219 (51.89) 17 (58.62) 37 (49.33) 86 (74.78) 10 (83.33)
Unknown or not reported 34 (8.06) 2 (6.90) 8 (10.67) 14 (12.17) 1 (8.33)
Race
African American/Black 7 (1.66) 1 (3.45) 1 (1.33) 6 (5.22) 0
American Indian/Alaska Native 5 (1.18) 1 (3.45) 2 (2.67) 0 0
Asian 69 (16.35) 0 5 (6.67) 8 (6.96) 0
Native Hawaiian/Pacific Islander 3 (0.71) 1 (3.45) 1 (1.33) 0 0
White 218 (51.66) 19 (65.52) 46 (61.33) 72 (62.61) 10 (83.33)
>1 Race 45 (10.66) 4 (13.79) 3 (4.00) 14 (12.17) 1 (8.33)
Not reported, unknown, or otherb 75 (17.77) 3 (10.34) 17 (22.67) 15 (13.04) 1 (8.33)
Mullen T score, mean (SD)53
Visual reception 34.73 (13.33) 50.62 (11.00) 45.19 (9.96) 55.23 (11.03) 52.33 (7.75)
Fine motor 34.91 (17.67) 43.93 (7.67) 44.75 (11.45) 51.90 (8.73) 50.00 (9.28)
Receptive language 29.10 (15.15) 44.86 (8.89) 40.55 (11.83) 51.33 (10.99) 54.83 (16.57)
Expressive language 28.00 (16.06) 46.28 (9.25) 34.20 (10.85) 51.17 (10.24) 57.58 (10.15)
Early learning composite 67.47 (21.04) 93.24 (11.05) 83.04 (14.34) 104.80 (14.61) 109.42 (13.23)
Vineland, mean (SD)54
Communication 77.45 (16.67) 94.34 (9.44) 87.56 (12.96) 100.55 (9.26) 100.00 (7.21)
Daily living 82.24 (13.22) 92.97 (12.73) 92.11 (10.83) 96.77 (11.42) 99.08 (8.98)
Socialization 81.80 (13.64) 91.72 (14.52) 90.67 (12.00) 100.23 (9.69) 100.42 (7.43)
Motor skills 90.66 (11.45) 97.41 (13.75) 94.12 (11.66) 99.31 (9.62) 100.36 (4.59)
Adaptive behavior composite 79.21 (11.60) 91.72 (11.13) 88.03 (11.23) 98.19 (10.38) 98.83 (8.22)
ADOS score, mean (SD)52
SA/CoSo 13.70 (4.31) 5.21 (3.72) 3.24 (2.29) 2.67 (1.94) 2.83 (1.27)
RRB 5.68 (1.75) 2.90 (1.47) 1.72 (1.28) 1.37 (1.21) 0.75 (1.06)
Total 19.39 (5.09) 8.10 (3.72) 4.96 (2.51) 4.04 (2.21) 3.58 (1.98)
Open in a new tab

Abbreviations: ADOS, Autism Diagnostic Observation Schedule; ASD, autism spectrum disorder; RRB, restricted and repetitive behavior; SA/CoSo, social affect and communication total score; TD, typical development; TypSibASD, typical sibling of an ASD proband.

a

Data are presented as the number (percentage) of toddlers unless otherwise indicated.

b

Other was not broken down further.

Apparatus and Procedure

Either a Tobii T120 (365 toddlers [55.90%]) or Tobii Pro Spectrum 600 Hz (288 toddlers [44.10%]) eye tracker in combination with E-Prime software, version 3.0 (Psychology Software Tools), was used to create gaze-contingent paradigms wherein the x-y coordinate of a toddler’s fixation triggered the appearance of a specific movie file. More details are given in the eMethods and eFigure 5 in Supplement 1.

Dog-Cat Gaze-Contingent Preexperiment Training

Toddlers were introduced to the concept of gaze contingency prior to the experiment. Details are given in the eMethods and eFigure 6 in Supplement 1.

Motherese Paradigms

Toddlers were presented with a series of 3 paradigms designed to quantify interest in motherese speech. Each of the paradigms consisted of 2 movies placed side by side, with 1 side always containing an actress speaking in a motherese style, using speech that contained high rates of questions, inflections, and emotional prosody.1,55 In the motherese vs traffic paradigm, the contrasting movie depicted cars moving on a busy highway with accompanying car horns and traffic noise, while in the motherese vs techno paradigm, the contrasting movie contained moving abstract shapes and numbers and was accompanied by musical chords. In the motherese vs flat affect paradigm, the same actress was shown in both movies, reading from identical scripts, but in the contrasting movie, she used a flat affect, reading the script in a monotone voice. The movies were available for approximately 60 seconds, and toddlers used the direction of their gaze to control the amount of time each movie was played. Areas of interest (AOIs) were drawn on either side and consisted of each movie in its entirety (eg, motherese AOI or traffic AOI). To assess test-retest reliability, eye-tracking tests were repeated in a subset of toddlers 1 to 12 or more months following their initial eye-tracking session.

Statistical Analyses

Primary analyses were conducted using R, version 4.0.2 (R Project for Statistical Computing), and SPSS, version 28 (IBM). Between-group differences were assessed using 1-way analyses of covariance with sex and age as covariates. The Tukey honestly significant difference test was used to perform post hoc, pairwise comparisons of group differences. Two-sided P < .05 indicated statistical significance. Ninety-five percent CIs of the mean differences between groups were calculated, and effect sizes are reported using η2 and Cohen d, as appropriate. Test-retest reliability was examined using intraclass correlations (ICCs; model 2, 1).56 Data were collected from February 2018 to April 2021 and were analyzed from April 2021 to March 2022.

Total Looking Time, Attention Toward Motherese Speech, and Number of Saccades

To assess whether overall attention differed between diagnostic groups, total looking time collapsed across AOIs was calculated for each paradigm. Time spent fixating within each AOI (eg, motherese or traffic) was divided by total looking time to determine a percent fixation value for each participant in each paradigm. The number of saccades per second was calculated for each participant as n – 1 fixations divided by the total looking time within motherese and nonmotherese AOIs.

Sensitivity, Specificity, Positive Predictive Value, Negative Predictive Value, and ROC Curve Analyses

For diagnostic classification accuracy analyses, toddlers were designated as having ASD or not having ASD (ie, ASD features, delay, typical sibling of an ASD proband, and TD) based on their most recent diagnostic evaluation. A receiver operating characteristic (ROC) curve was generated for each test and used to determine the optimal percent fixation and saccades per second associated with specificity of 96% or greater. A high specificity rate was chosen given the expense and anxiety associated with a false-positive result.57 To assess whether combining metrics enhanced classification accuracy, reduced error pruning was applied as the classifier within Waikato Environment for Knowledge Analysis (WEKA)58,59 using both percent fixation on motherese and saccades per second (eMethods in Supplement 1).

Association Between Motherese Fixation Levels, Social and Language Ability, and Age

Pearson correlations were used to examine motherese fixation and social and language associations (eMethods in Supplement 1). For interpretative simplicity, toddlers were also stratified into subgroups of high, medium, and low levels of fixation toward motherese, defined as follows: fixation of 30% or less, low motherese attention; fixation of 70% or more, high motherese attention; and the remaining toddlers, middle motherese attention. Differences in clinical profiles between each subgroup were examined.

Results

Of 736 toddlers who participated in 1 or more motherese eye-tracking tests, 653 (88.72%) were included in the final analyses. The mean (SD) age was 26.45 (8.37) months; 173 (26.49%) were female, and 480 (73.51%) were male. A total of 15 (2.30%) were African American/Black; 8 (1.23%), American Indian/Alaska Native; 82 (12.56%), Asian; 5 (0.77%), Native Hawaiian/Pacific Islander; 365 (55.90%), White; and 67 (10.26%), more than 1 race; for 111 (17.00%), race was unknown, not reported, or other. A total of 225 (34.46%) were Hispanic or Latino and 369 (56.51%), not Hispanic or Latino; for 59 (9.03%), ethnicity was unknown or not reported. Of the 653 toddlers, 422 (64.62%) had ASD, 104 (15.93%) had non-ASD delays, and 127 (19.45%) had TD; 588 (90.05%) participated in the motherese vs traffic paradigm, 479 (73.35%) in the motherese vs techno paradigm, and 389 (59.57%) in the motherese vs flat affect paradigm. Toddlers with ASD represented a wide range of ability and disability in terms of adaptive behavior, cognitive and language ability, and autism symptom severity. Details regarding clinical score distributions are given in the Table and eFigure 4 in Supplement 1.

Total Looking Time

There were no to minimal between-group differences in total looking time, depending on the paradigm. Details are given in the eResults in Supplement 1.

Attention Toward Motherese Speech as Indexed by Percent Fixation

Unlike toddlers without ASD, who almost uniformly attended to motherese speech with a median level of 82.85% and 80.75% across the traffic and techno tests, respectively, among toddlers with ASD, there was a wide range spanning from 0% to 100%. Differences in percent fixation levels were significant across the traffic (F4,581 = 28.08; P < .001; η2 = 0.15) and techno (F4,472 = 26.67; P < .001; η2 = 0.18) paradigms, with toddlers with ASD fixating on the motherese videos significantly less than all other diagnostic groups (except motherese vs traffic among toddlers with ASD vs typical siblings of an ASD proband). For example, in the techno paradigm, the mean (SD) motherese fixation value for toddlers with TD was 76.36% (19.73%), whereas for toddlers with ASD, the mean was 50.08% (27.56%) (ASD vs TD mean [SD] difference, 26.28% [2.78%]; P < .001; Cohen d, 1.0). However, some toddlers with ASD had high levels of fixation on motherese at or near 100% values. Overall associated effect sizes were large for ASD vs other diagnostic groups, ranging from 0.79 to 1.22 across both paradigms. In contrast, there were only small between-group differences in the flat affect paradigm (F4,382 = 3.49; P = .008; η2 = 0.03) with associated small effect sizes (Figure 1 and Figure 2). Age was a significant covariate for both the traffic and flat affect paradigms (traffic, F1,581 = 22.66; P < .001; η2 = 0.03; flat affect, F1,382 = 11.27; P < .001; η2 = 0.03). Sex was a significant covariate only for the motherese vs traffic paradigm (F1,581 = 9.01; P = .003; η2 = 0.01). Pairwise comparisons of group differences and sex differences are given in the eResults, eTable 1, and eFigure 7 in Supplement 1.

Figure 1. Motherese Eye-Tracking Test Performance.

Figure 1.

Open in a new tab

Box plots show the raw data for each of the motherese eye-tracking tests across 5 major diagnostic groups. Middle line indicates median value; upper and lower bounds of the box, IQR; whiskers, range; dots, individual participants; and triangles, the mean. Because tests are all preferential looking paradigms, points along the y-axes sum to 100%; for example, within the motherese vs traffic paradigm, if a toddler fixated on motherese 90% of the time, he or she had a corresponding value of 10% fixation in traffic. ASD indicates autism spectrum disorder; FP, false positive; TD, typical development; TP, true positive; and TypSibASD, typical sibling of an ASD proband.

Figure 2. Effect Size and Area Under Receiver Operating Characteristic (ROC) Curve (AUC).

Figure 2.

Open in a new tab

A-C, Effect sizes indicate between-group differences. D, A reduced error pruning tree classifier using 4 eye-tracking metrics (percent fixation on motherese from the techno and traffic paradigms and saccades per second viewing motherese) was trained to classify toddlers who had participated in both the techno and the traffic paradigms (n = 430) as ASD (n = 283) or non-ASD (n = 147). ASD indicates autism spectrum disorder; TD, typical development; and TypSibASD, typical sibling of an ASD proband.

Number of Saccades

Similar to our eye-tracking tests of social visual attention,32,33 toddlers in the low motherese attention group showed a significantly increased number of saccades when viewing the social motherese AOIs (motherese vs techno paradigm: low motherese attention vs TD, 2.34 saccades vs 1.56 saccades; F6,470 = 3.17; P = .005; η2 = 0.04; motherese vs traffic paradigm: low motherese attention vs TD, 2.39 saccades vs 1.68 saccades; F6,579 = 14.88; P < .001; η2 = 0.13). Conversely, across diagnostic groups, saccade rates were largely comparable when attending to nonsocial AOIs (Figure 3). Correlations across paradigms are given in the eResults and eFigure 8 in Supplement 1.

Figure 3. Saccade Profiles Stratified by Diagnostic Group.

Figure 3.

Open in a new tab

Circles indicate the mean saccade rate and error bars indicate the standard error of the mean. Toddlers with autism spectrum disorder (ASD) were additionally stratified based on strength of preference for motherese. AOI indicates area of interest; Mthr Attn, motherese attention; TD, typical development; and TypSibASD, typical sibling of an ASD proband.

Sensitivity, Specificity, Positive Predictive Value, Negative Predictive Value, and ROC Curve Analyses

Using a value of 30% fixation or less on motherese as the cutoff value resulted in sensitivity of 18% (95% CI, 14%-22%), specificity of 98% (95% CI, 95%-99%), positive predictive value (PPV) of 94% (95% CI, 86%-98%), and negative predictive value (NPV) of 40% (95% CI, 36%-45%) and an associated area under the ROC curve (AUC) of 0.733 (95% CI, 0.693-0.773) for the traffic paradigm and 29% (95% CI, 24%-34%) sensitivity, 96% (95% CI, 92%-98%) specificity, 94% (95% CI, 86%-98%) PPV, and 41% (95% CI, 36%-46%) NPV and an associated AUC of 0.761 (95% CI, 0.717-0.804) for the techno paradigm. In contrast, values were low for the flat affect paradigm, which had an AUC of 0.597 (95% CI, 0.541-650). Combining saccades per second with percent fixation metrics across the traffic and techno paradigms increased the AUC to 0.841 (95% CI, 0.805-0.877). Diagnostic accuracy cross-tabulation tables and 95% CIs are provided in eTables 2 and 3 in Supplement 1. Given the lack of diagnostic classification accuracy and small effect sizes, the flat affect paradigm was not considered in further analyses.

Association Between Motherese Fixation Levels, Social and Language Ability, and Age

Collapsed across all diagnostic groups as well as within the ASD group alone, percentage of time fixating on motherese speech during the techno paradigm was significantly correlated with receptive and expressive language ability, social affect scores, and the overall socialization domain score on the Vineland scales (Pearson r range for toddlers with ASD: 0.30-0.48; P < .001). When toddlers with ASD were stratified as having low, middle, or high levels of attention toward motherese speech, significant differences in social and language abilities between the subtypes were observed. For example, within the social domain, the mean (SD) Vineland score was 9.55 (1.78) points lower (Cohen d, 0.79; P < .001) and the mean (SD) social affect total score was 4.73 (0.63) points higher (indicating greater symptom severity) in the low motherese attention group than in the high motherese attention group (Cohen d, −1.12; P < .001). Within the expressive and receptive language domains, t scores were a mean (SD) of 16.03 (1.90) and 17.45 (1.84) points lower, respectively (expressive: Cohen d, 1.26 [P < .001]; receptive: Cohen d, 1.42 [P < .001]) (Figure 4). Similar results were obtained for the traffic paradigm (eFigure 9 in Supplement 1). Overall, there were inconsistent associations between fixation on motherese and age across diagnostic groups and paradigms (eResults, eTable 4, and eFigures 9 and 10 in Supplement 1).

Figure 4. Differences in Clinical Phenotype Among Toddlers With Autism Spectrum Disorder (ASD) With Low, Middle, and High Attention to Motherese Speech During the Techno Paradigm.

Figure 4.

Open in a new tab

A, Dots indicate percent fixation values on motherese for individual toddlers; whiskers, the range; and triangle, the mean. B-E, Differences in standardized test scores between subgroups. Results were similar for the traffic paradigm (eResults and eFigure 9 in Supplement 1). ADOS indicates Autism Diagnostic Observation Schedule; Mthr Attn, motherese attention; and SA, social affect.

Test-Retest Reliability

Percent fixation levels on motherese up to 1 year following initial testing were significantly correlated in the traffic paradigm, with ICC values ranging from 0.38 to 0.77. Fixation values associated with intervals greater than 1 year were significantly correlated only in the techno paradigm (ICC, 0.48; P = .001) (eResults and eTable 5 in Supplement 1).

Discussion

The animated speech style known as motherese can be found across cultures and languages,60 is evident between deaf mothers and their deaf children,61,62 and has even been noted to occur while fetuses are still in the womb.63 Likely genetic, motherese is more than a speech style; it is a tool for reciprocal engagement that enhances social and language development, with infants often matching the fundamental frequencies and voice pitch of their mothers.64 Using gaze-contingent eye tracking that allowed toddlers to control the images and sounds they heard, we found that similar to toddlers who are typically developing or non–ASD delayed, the majority of toddlers with ASD attended to motherese speech at high levels. In contrast, approximately 23% of toddlers with ASD showed low levels of attention toward motherese speech, with some as low as 1% to 2%. This subgroup instead preferred to listen to noises such as the sound of horns beeping and car engines during traffic. When fixation levels on motherese were at or below 30%, the PPV and specificity across the motherese vs traffic and motherese vs techno paradigms were high and ranged between 94% and 98%, depending on the test. Given that the tests are engaging for toddlers, result in objective quantitation of social and language attention, and require less than 5 minutes to complete, they may be useful as a screening or diagnostic tool to identify toddlers in most need of immediate attention.

Our conservative selection of a cutoff value of 30% fixation or less on motherese was deliberate to tune the tests to have high specificity and reduce false-positive results, which are unnecessarily stressful for parents.57 There is often a tradeoff, however, between specificity and sensitivity, and this deliberate tuning lowered sensitivity. The low sensitivity of the motherese-attention biomarkers and other ASD biomarkers31 likely reflects the true clinical and biological heterogeneity inherent in ASD.65,66,67 In terms of genetic findings, most single-gene sequence variations, such as in the Shank gene family, are found in approximately 1% of children with ASD.68

There was also an association between level of attention to motherese speech and social and language abilities, with children showing the lowest levels of interest in motherese speech also having the lowest social and language abilities. This is not surprising given the association of experience with brain development.25,69,70,71,72 Stratifying toddlers with ASD in the current study into low, middle, and high levels of attention to motherese speech also highlighted the reverse phenomenon: toddlers in the high–motherese attention group had receptive and expressive language standard scores 17.45 and 16.03 points, respectively, above those of toddlers with ASD in the low–motherese attention group as well as better social affect scores. Given that there are critical periods for language development that extend across the first years of life,73 it may be even more challenging for children with ASD in the low–motherese attention group to catch up once this critical period has passed. At least 1 previous eye-tracking study74 showed that toddlers with ASD who paid the least attention to social images at ages 1 to 3 years had greater symptom severity 5 to 9 years later.

Our data also suggest that the subgroup of the population with ASD and low attention to motherese may be more homogeneous at some intermediate psychobiological level, thereby potentially justifying the use of this test to stratify the population with autism in future studies of social attention and its correlates. A recent study27 revealed that toddlers with ASD who showed low attention levels to motherese speech as a group had lower levels of neural functional activity in classic speech-processing areas, such as the superior temporal gyrus, compared with other toddlers with ASD who paid greater attention to motherese speech.

Our findings also revealed a dissociation between saccade rates depending on what toddlers were looking at but only for the ASD subgroup that showed the lowest levels of attention to motherese. These toddlers showed a greater number of saccades during the brief periods that they viewed the motherese videos, possibly reflecting greater anxiety or challenges related to speech processing.

Interestingly, the motherese vs flat affect paradigm was not effective as a diagnostic classifier, showing only minimal between-group ASD vs TD differences. Given that null results are rarely reported,75 we believed this was important to underscore. One reason may be related to the nature of the flat affect video that contained an actress speaking in an unnatural computer-like voice that caused several toddlers to smirk and may have artificially increased attention to this video. Indeed, total looking time in the flat affect AOI was higher than in the nonmotherese contrast videos used in the other paradigms.

Strengths and Limitations

This study has several strengths. These include comprehensive diagnostic and psychometric evaluations conducted by licensed clinical psychologists blind to eye-tracking scores and the inclusion of multiple contrast groups including toddlers with non-ASD delays, an essential component for determining future clinical utility.

This study also has limitations. Given that the sample contained a combination of toddlers who first failed a parent-report screen at well-infant checkups and general community referrals, a key limitation relates to the unknown generalizability of the findings as a universal screening approach.

Conclusions

Overall, attention to motherese speech is a fundamental property of early development that can be rapidly measured using eye tracking. In this diagnostic study, a cutoff level of 30% fixation or less on motherese speech was accurate at identifying toddlers who were diagnostically classified as having ASD. Identifying toddlers who show unusually low levels of attention to motherese speech is potentially beneficial not only for early ASD screening, diagnosis, and prognosis but also for possible therapeutic targets and may be a key pathway toward precision medicine.

Supplement 1.

eMethods.

eFigure 1. Participation Flowchart

eFigure 2. Summary of Reasons for Exclusion

eFigure 3. Test Overlap and Intersection Plot

eFigure 4. Histograms Illustrating Distribution of Clinical Scores Across Diagnostic Groups

eFigure 5. Comparison of Motherese Fixation Levels Across Eye-Tracking Systems

eFigure 6. Preexperiment Dog-Cat Training Task

eResults.

eTable 1. Pairwise Comparisons of Diagnostic Group Differences Across Motherese Paradigms

eTable 2. Cross-tabulation Tables

eTable 3. Metrics of Diagnostic Accuracy with 95% CIs

eFigure 7. Motherese Fixations Levels Stratified by Sex

eFigure 8. Relationships Between Motherese Fixation Levels Across Motherese Paradigms

eTable 4. Correlations Between Motherese Fixation Levels and Social and Language Ability in the Techno and Traffic Paradigms

eFigure 9. Differences in Clinical Phenotype Between Toddlers with Low, Middle, and High Attention to Motherese Speech During the Traffic Paradigm

eFigure 10. Age-related Changes in Motherese Fixation Levels

eTable 5. Test-Retest Reliability

eReferences.

Click here for additional data file. (2MB, pdf)
Supplement 2.

Data Sharing Statement

Click here for additional data file. (16.7KB, pdf)

References

  • 1.Saint-Georges C, Chetouani M, Cassel R, et al. Motherese in interaction: at the cross-road of emotion and cognition? (a systematic review). PLoS One. 2013;8(10):e78103. doi: 10.1371/journal.pone.0078103 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Cooper RP, Aslin RN. Preference for infant-directed speech in the first month after birth. Child Dev. 1990;61(5):1584-1595. doi: 10.2307/1130766 [DOI] [PubMed] [Google Scholar]
  • 3.Sulpizio S, Kuroda K, Dalsasso M, et al. Discriminating between mothers’ infant- and adult-directed speech: cross-linguistic generalizability from Japanese to Italian and German. Neurosci Res. 2018;133:21-27. doi: 10.1016/j.neures.2017.10.008 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Falk D. Prelinguistic evolution in early hominins: whence motherese? Behav Brain Sci. 2004;27(4):491-503. doi: 10.1017/S0140525X04000111 [DOI] [PubMed] [Google Scholar]
  • 5.Singh L, Nestor S, Parikh C, Yull A. Influences of infant-directed speech on early word recognition. Infancy. 2009;14(6):654-666. doi: 10.1080/15250000903263973 [DOI] [PubMed] [Google Scholar]
  • 6.Schreiner MS, van Schaik JE, Sučević J, Hunnius S, Meyer M. Let’s talk action: infant-directed speech facilitates infants’ action learning. Dev Psychol. 2020;56(9):1623-1631. doi: 10.1037/dev0001079 [DOI] [PubMed] [Google Scholar]
  • 7.Miller M, Iosif AM, Hill M, Young GS, Schwichtenberg AJ, Ozonoff S. Response to name in infants developing autism spectrum disorder: a prospective study. J Pediatr. 2017;183:141-146.e1. doi: 10.1016/j.jpeds.2016.12.071 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Campbell K, Carpenter KL, Hashemi J, et al. Computer vision analysis captures atypical attention in toddlers with autism. Autism. 2019;23(3):619-628. doi: 10.1177/1362361318766247 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Lazenby DC, Sideridis GD, Huntington N, et al. Language differences at 12 months in infants who develop autism spectrum disorder. J Autism Dev Disord. 2016;46(3):899-909. doi: 10.1007/s10803-015-2632-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Key AP, D’Ambrose Slaboch K. Speech processing in autism spectrum disorder: an integrative review of auditory neurophysiology findings. J Speech Lang Hear Res. 2021;64(11):4192-4212. doi: 10.1044/2021_JSLHR-20-00738 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Anderson JS, Lange N, Froehlich A, et al. Decreased left posterior insular activity during auditory language in autism. AJNR Am J Neuroradiol. 2010;31(1):131-139. doi: 10.3174/ajnr.A1789 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Eyler LT, Pierce K, Courchesne E. A failure of left temporal cortex to specialize for language is an early emerging and fundamental property of autism. Brain. 2012;135(Pt 3):949-960. doi: 10.1093/brain/awr364 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Lombardo MV, Pierce K, Eyler LT, et al. Different functional neural substrates for good and poor language outcome in autism. Neuron. 2015;86(2):567-577. doi: 10.1016/j.neuron.2015.03.023 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Kuhl PK, Coffey-Corina S, Padden D, Munson J, Estes A, Dawson G. Brain responses to words in 2-year-olds with autism predict developmental outcomes at age 6. PLoS One. 2013;8(5):e64967. doi: 10.1371/journal.pone.0064967 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Williams ZJ, Abdelmessih PG, Key AP, Woynaroski TG. Cortical auditory processing of simple stimuli is altered in autism: a meta-analysis of auditory evoked responses. Biol Psychiatry Cogn Neurosci Neuroimaging. 2021;6(8):767-781. doi: 10.1016/j.bpsc.2020.09.011 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Chen J, Liang C, Wei Z, et al. Atypical longitudinal development of speech-evoked auditory brainstem response in preschool children with autism spectrum disorders. Autism Res. 2019;12(7):1022-1031. doi: 10.1002/aur.2110 [DOI] [PubMed] [Google Scholar]
  • 17.Kuhl PK, Coffey-Corina S, Padden D, Dawson G. Links between social and linguistic processing of speech in preschool children with autism: behavioral and electrophysiological measures. Dev Sci. 2005;8(1):F1-F12. doi: 10.1111/j.1467-7687.2004.00384.x [DOI] [PubMed] [Google Scholar]
  • 18.Paul R, Chawarska K, Fowler C, Cicchetti D, Volkmar F. “Listen my children and you shall hear”: auditory preferences in toddlers with autism spectrum disorders. J Speech Lang Hear Res. 2007;50(5):1350-1364. doi: 10.1044/1092-4388(2007/094) [DOI] [PubMed] [Google Scholar]
  • 19.Watson LR, Roberts JE, Baranek GT, Mandulak KC, Dalton JC. Behavioral and physiological responses to child-directed speech of children with autism spectrum disorders or typical development. J Autism Dev Disord. 2012;42(8):1616-1629. doi: 10.1007/s10803-011-1401-z [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Filipe MG, Watson L, Vicente SG, Frota S. Atypical preference for infant-directed speech as an early marker of autism spectrum disorders? a literature review and directions for further research. Clin Linguist Phon. 2018;32(3):213-231. doi: 10.1080/02699206.2017.1342694 [DOI] [PubMed] [Google Scholar]
  • 21.Venuti P, de Falco S, Esposito G, Zaninelli M, Bornstein MH. Maternal functional speech to children: a comparison of autism spectrum disorder, Down syndrome, and typical development. Res Dev Disabil. 2012;33(2):506-517. doi: 10.1016/j.ridd.2011.10.018 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Perryman TY, Carter AS, Messinger DS, Stone WL, Ivanescu AE, Yoder PJ. Brief report: parental child-directed speech as a predictor of receptive language in children with autism symptomatology. J Autism Dev Disord. 2013;43(8):1983-1987. doi: 10.1007/s10803-012-1725-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Majdan M, Shatz CJ. Effects of visual experience on activity-dependent gene regulation in cortex. Nat Neurosci. 2006;9(5):650-659. doi: 10.1038/nn1674 [DOI] [PubMed] [Google Scholar]
  • 24.Neville H, Bavelier D. Human brain plasticity: evidence from sensory deprivation and altered language experience. Prog Brain Res. 2002;138:177-188. doi: 10.1016/S0079-6123(02)38078-6 [DOI] [PubMed] [Google Scholar]
  • 25.Benasich AA, Choudhury NA, Realpe-Bonilla T, Roesler CP. Plasticity in developing brain: active auditory exposure impacts prelinguistic acoustic mapping. J Neurosci. 2014;34(40):13349-13363. doi: 10.1523/JNEUROSCI.0972-14.2014 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Quartz SR. The constructivist brain. Trends Cogn Sci. 1999;3(2):48-57. doi: 10.1016/S1364-6613(98)01270-4 [DOI] [PubMed] [Google Scholar]
  • 27.Xiao Y, Wen TH, Kupis L, et al. Neural responses to affective speech, including motherese, map onto clinical and social eye tracking profiles in toddlers with ASD. Nat Hum Behav. 2022;6(3):443-454. doi: 10.1038/s41562-021-01237-y [DOI] [PubMed] [Google Scholar]
  • 28.Macari S, Milgramm A, Reed J, et al. Context-specific dyadic attention vulnerabilities during the first year in infants later developing autism spectrum disorder. J Am Acad Child Adolesc Psychiatry. 2021;60(1):166-175. doi: 10.1016/j.jaac.2019.12.012 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.McPartland JC, Bernier RA, Jeste SS, et al. ; Autism Biomarkers Consortium for Clinical Trials . The Autism Biomarkers Consortium for Clinical Trials (ABC-CT): scientific context, study design, and progress toward biomarker qualification. Front Integr Neurosci. 2020;14:16. doi: 10.3389/fnint.2020.00016 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Jones W, Klin A. Attention to eyes is present but in decline in 2-6-month-old infants later diagnosed with autism. Nature. 2013;504(7480):427-431. doi: 10.1038/nature12715 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Pierce K, Marinero S, Hazin R, McKenna B, Barnes CC, Malige A. Eye tracking reveals abnormal visual preference for geometric images as an early biomarker of an autism spectrum disorder subtype associated with increased symptom severity. Biol Psychiatry. 2016;79(8):657-666. doi: 10.1016/j.biopsych.2015.03.032 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Moore A, Wozniak M, Yousef A, et al. The geometric preference subtype in ASD: identifying a consistent, early-emerging phenomenon through eye tracking. Mol Autism. 2018;9:19. doi: 10.1186/s13229-018-0202-z [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Wen TH, Cheng A, Andreason C, et al. Large scale validation of an early-age eye-tracking biomarker of an autism spectrum disorder subtype. Sci Rep. 2022;12(1):4253. doi: 10.1038/s41598-022-08102-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Alcañiz M, Chicchi-Giglioli IA, Carrasco-Ribelles LA, et al. Eye gaze as a biomarker in the recognition of autism spectrum disorder using virtual reality and machine learning: a proof of concept for diagnosis. Autism Res. 2022;15(1):131-145. doi: 10.1002/aur.2636 [DOI] [PubMed] [Google Scholar]
  • 35.Frazier TW, Uljarevic M, Ghazal I, et al. Social attention as a cross-cultural transdiagnostic neurodevelopmental risk marker. Autism Res. 2021;14(9):1873-1885. doi: 10.1002/aur.2532 [DOI] [PubMed] [Google Scholar]
  • 36.Frazier TW, Coury DL, Sohl K, et al. Evidence-based use of scalable biomarkers to increase diagnostic efficiency and decrease the lifetime costs of autism. Autism Res. 2021;14(6):1271-1283. doi: 10.1002/aur.2498 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Falck-Ytter T, Bölte S, Gredebäck G. Eye tracking in early autism research. J Neurodev Disord. 2013;5(1):28. doi: 10.1186/1866-1955-5-28 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Unruh KE, Sasson NJ, Shafer RL, et al. Social orienting and attention is influenced by the presence of competing nonsocial information in adolescents with autism. Front Neurosci. 2016;10:586. doi: 10.3389/fnins.2016.00586 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Elison JT, Sasson NJ, Turner-Brown LM, Dichter G, Bodfish JW. Age trends in visual exploration of social and nonsocial information in children with autism. Res Autism Spectr Disord. 2012;6(2):842-851. doi: 10.1016/j.rasd.2011.11.005 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Kaliukhovich DA, Manyakov NV, Bangerter A, et al. Social attention to activities in children and adults with autism spectrum disorder: effects of context and age. Mol Autism. 2020;11(1):79. doi: 10.1186/s13229-020-00388-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Wang Q, Chang J, Chawarska K. Atypical value-driven selective attention in young children with autism spectrum disorder. JAMA Netw Open. 2020;3(5):e204928. doi: 10.1001/jamanetworkopen.2020.4928 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Shic F, Bradshaw J, Klin A, Scassellati B, Chawarska K. Limited activity monitoring in toddlers with autism spectrum disorder. Brain Res. 2011;1380:246-254. doi: 10.1016/j.brainres.2010.11.074 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43.Sasson NJ, Touchstone EW. Visual attention to competing social and object images by preschool children with autism spectrum disorder. J Autism Dev Disord. 2014;44(3):584-592. doi: 10.1007/s10803-013-1910-z [DOI] [PubMed] [Google Scholar]
  • 44.Klin A, Jones W, Schultz R, Volkmar F, Cohen D. Visual fixation patterns during viewing of naturalistic social situations as predictors of social competence in individuals with autism. Arch Gen Psychiatry. 2002;59(9):809-816. doi: 10.1001/archpsyc.59.9.809 [DOI] [PubMed] [Google Scholar]
  • 45.Shic F, Wang Q, Macari SL, Chawarska K. The role of limited salience of speech in selective attention to faces in toddlers with autism spectrum disorders. J Child Psychol Psychiatry. 2020;61(4):459-469. doi: 10.1111/jcpp.13118 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46.Konovalova I, Antolin JV, Bolderston H, Gregory NJ. Adults with higher social anxiety show avoidant gaze behaviour in a real-world social setting: a mobile eye tracking study. PLoS One. 2021;16(10):e0259007. doi: 10.1371/journal.pone.0259007 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 47.Myles O, Grafton B, MacLeod C. Anxiety & inhibition: dissociating the involvement of state and trait anxiety in inhibitory control deficits observed on the anti-saccade task. Cogn Emot. 2020;34(8):1746-1752. doi: 10.1080/02699931.2020.1802229 [DOI] [PubMed] [Google Scholar]
  • 48.Pierce K, Gazestani V, Bacon E, et al. Get SET Early to identify and treatment refer autism spectrum disorder at 1 year and discover factors that influence early diagnosis. J Pediatr. 2021;236:179-188. doi: 10.1016/j.jpeds.2021.04.041 [DOI] [PubMed] [Google Scholar]
  • 49.Wetherby A, Prizant B. Communication and Symbolic Behavior Scales Developmental Profile, First Normed Edition. Brookes; 2002. [Google Scholar]
  • 50.Wetherby AM, Brosnan-Maddox S, Peace V, Newton L. Validation of the infant-toddler checklist as a broadband screener for autism spectrum disorders from 9 to 24 months of age. Autism. 2008;12(5):487-511. doi: 10.1177/1362361308094501 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 51.Pierce K, Gazestani VH, Bacon E, et al. Evaluation of the diagnostic stability of the early autism spectrum disorder phenotype in the general population starting at 12 months. JAMA Pediatr. 2019;173(6):578-587. doi: 10.1001/jamapediatrics.2019.0624 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 52.Lord C, Rutter M, DiLavore PC, Risi S. Autism Diagnostic Observation Schedule (ADOS). Western Psychological Services; 2001. [Google Scholar]
  • 53.Mullen EM. Mullen Scales of Early Learning. AGS ed. American Guidance Service Inc; 1995. [Google Scholar]
  • 54.Sparrow S, Cicchetti D, Balla D. Vineland-II: Vineland Adaptive Behavior Scales, Second Edition—Survey Forms Manual. American Guidance Service; 2005. [Google Scholar]
  • 55.Fernald A. Four-month-old infants prefer to listen to motherese. Infant Behav Dev. 1985;8(2):181-195. doi: 10.1016/S0163-6383(85)80005-9 [DOI] [Google Scholar]
  • 56.Shrout PE, Fleiss JL. Intraclass correlations: uses in assessing rater reliability. Psychol Bull. 1979;86(2):420-428. doi: 10.1037/0033-2909.86.2.420 [DOI] [PubMed] [Google Scholar]
  • 57.Al-Qabandi M, Gorter JW, Rosenbaum P. Early autism detection: are we ready for routine screening? Pediatrics. 2011;128(1):e211-e217. doi: 10.1542/peds.2010-1881 [DOI] [PubMed] [Google Scholar]
  • 58.Hall M, Frank E, Holmes G, Pfahringer B, Reutemann P, Witten IH. The WEKA data mining software: an update. SIGKDD Explor. 2009;11(1):10-18. doi: 10.1145/1656274.1656278 [DOI] [Google Scholar]
  • 59.Frank E, Hall M, Trigg L, Holmes G, Witten IH. Data mining in bioinformatics using WEKA. Bioinformatics. 2004;20(15):2479-2481. doi: 10.1093/bioinformatics/bth261 [DOI] [PubMed] [Google Scholar]
  • 60.Hilton CB, Moser CJ, Bertolo M, et al. Acoustic regularities in infant-directed speech and song across cultures. Nat Hum Behav. 2022;6(11):1545-1556. doi: 10.1038/s41562-022-01410-x [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 61.Masataka N. Perception of motherese in Japanese sign language by 6-month-old hearing infants. Dev Psychol. 1998;34(2):241-246. doi: 10.1037/0012-1649.34.2.241 [DOI] [PubMed] [Google Scholar]
  • 62.Reilly JS, Bellugi U. Competition on the face: affect and language in ASL motherese. J Child Lang. 1996;23(1):219-239. doi: 10.1017/S0305000900010163 [DOI] [PubMed] [Google Scholar]
  • 63.Parlato-Oliveira E, Saint-Georges C, Cohen D, et al. “Motherese” prosody in fetal-directed speech: an exploratory study using automatic social signal processing. Front Psychol. 2021;12:646170. doi: 10.3389/fpsyg.2021.646170 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 64.Shimura Y, Yamanoucho I. Sound spectrographic studies on the relation between motherese and pleasure vocalization in early infancy. Acta Paediatr Jpn. 1992;34(3):259-266. doi: 10.1111/j.1442-200X.1992.tb00956.x [DOI] [PubMed] [Google Scholar]
  • 65.Chang J, Gilman SR, Chiang AH, Sanders SJ, Vitkup D. Genotype to phenotype relationships in autism spectrum disorders. Nat Neurosci. 2015;18(2):191-198. doi: 10.1038/nn.3907 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 66.Beversdorf DQ; Missouri Autism Summit Consortium . Phenotyping, etiological factors, and biomarkers: toward precision medicine in autism spectrum disorders. J Dev Behav Pediatr. 2016;37(8):659-673. doi: 10.1097/DBP.0000000000000351 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 67.Lombardo MV, Lai MC, Baron-Cohen S. Big data approaches to decomposing heterogeneity across the autism spectrum. Mol Psychiatry. 2019;24(10):1435-1450. doi: 10.1038/s41380-018-0321-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 68.Leblond CS, Nava C, Polge A, et al. Meta-analysis of SHANK mutations in autism spectrum disorders: a gradient of severity in cognitive impairments. PLoS Genet. 2014;10(9):e1004580. doi: 10.1371/journal.pgen.1004580 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 69.Black JE, Jones TA, Nelson CA, Greenough WT. Neuronal plasticity and the developing brain. In: Alessi N, Coyle JT, Harrison SI, Eth S, eds. The Handbook of Child and Adolescent Psychiatry. John Wiley & Sons; 1998:31-53. [Google Scholar]
  • 70.Westermann G. Experience-dependent brain development as a key to understanding the language system. Top Cogn Sci. 2016;8(2):446-458. doi: 10.1111/tops.12194 [DOI] [PubMed] [Google Scholar]
  • 71.Mercure E, Evans S, Pirazzoli L, et al. Language experience impacts brain activation for spoken and signed language in infancy: insights from unimodal and bimodal bilinguals. Neurobiol Lang (Camb). 2020;1(1):9-32. doi: 10.1162/nol_a_00001 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 72.Toritsuka M, Makinodan M, Kishimoto T. Social experience-dependent myelination: an implication for psychiatric disorders. Neural Plast. 2015;2015:465345. doi: 10.1155/2015/465345 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 73.Cisneros-Franco JM, Voss P, Thomas ME, de Villers-Sidani E. Critical periods of brain development. Handb Clin Neurol. 2020;173:75-88. doi: 10.1016/B978-0-444-64150-2.00009-5 [DOI] [PubMed] [Google Scholar]
  • 74.Bacon EC, Moore A, Lee Q, Carter Barnes C, Courchesne E, Pierce K. Identifying prognostic markers in autism spectrum disorder using eye tracking. Autism. 2020;24(3):658-669. doi: 10.1177/1362361319878578 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 75.Franco A, Malhotra N, Simonovits G. Publication bias in the social sciences: unlocking the file drawer. Science. 2014;345(6203):1502-1505. doi: 10.1126/science.1255484 [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplement 1.

eMethods.

eFigure 1. Participation Flowchart

eFigure 2. Summary of Reasons for Exclusion

eFigure 3. Test Overlap and Intersection Plot

eFigure 4. Histograms Illustrating Distribution of Clinical Scores Across Diagnostic Groups

eFigure 5. Comparison of Motherese Fixation Levels Across Eye-Tracking Systems

eFigure 6. Preexperiment Dog-Cat Training Task

eResults.

eTable 1. Pairwise Comparisons of Diagnostic Group Differences Across Motherese Paradigms

eTable 2. Cross-tabulation Tables

eTable 3. Metrics of Diagnostic Accuracy with 95% CIs

eFigure 7. Motherese Fixations Levels Stratified by Sex

eFigure 8. Relationships Between Motherese Fixation Levels Across Motherese Paradigms

eTable 4. Correlations Between Motherese Fixation Levels and Social and Language Ability in the Techno and Traffic Paradigms

eFigure 9. Differences in Clinical Phenotype Between Toddlers with Low, Middle, and High Attention to Motherese Speech During the Traffic Paradigm

eFigure 10. Age-related Changes in Motherese Fixation Levels

eTable 5. Test-Retest Reliability

eReferences.

Click here for additional data file. (2MB, pdf)
Supplement 2.

Data Sharing Statement

Click here for additional data file. (16.7KB, pdf)

Articles from JAMA Network Open are provided here courtesy of American Medical Association

RESOURCES