Neonatal sensitivity to vocal emotions: A milestone at 37 weeks of gestational age

Reviewer #1 (Public Review):

Summary:
This manuscript aimed to investigate the emergence of emotional sensitivity and its relationship with gestational age. Using an oddball paradigm and event-related potentials, the authors conducted an experiment in 120 healthy neonates with a gestational age range of 35 to 40 weeks. A significant developmental milestone was identified at 37 weeks gestational age, marking a crucial juncture in neonatal emotional responsiveness.

Strengths:
This study has several strengths, by providing profound insights into the early development of social-emotional functioning and unveiling the role of gestational age in shaping neonatal perceptual abilities. The methodology of this study demonstrates rigor and well-controlled experimental design, particularly involving matched control sounds, which enhances the reliability of the research. Their findings not only contribute to the field of neurodevelopment, but also showcase potential clinical applications, especially in the context of autism screening and early intervention for neurodevelopmental disorders.

Comments on the revised version:

After reviewing the authors' response letter and the revised manuscript, I believe they have done a commendable job in addressing my comments.
Additionally, I concur with the concerns raised by Reviewer #2 regarding several potential confounding factors that require better control in their experimental design. These include the differences in physical properties between vocal and nonvocal stimuli, as well as the infant's exposure to the speech/auditory environment. These concerns should be thoroughly and explicitly discussed in the manuscript, ensuring a clearer understanding for the readers.

Reviewer #2 (Public Review):

This is an important and very interesting report on a change in newborns' neural abilities to distinguish auditory signals as a function of the gestational age (GA) of the infant at birth (from 35 weeks GA to 40 weeks GA). The authors tested neural discrimination of sounds that were labeled 'happy' vs 'neutral' by listeners that represent two categories of sound, either human voices or auditory signals that mimic only certain properties of the human vocal signals. The finding is that a change occurs in neural discrimination of the happy and neutral auditory signals for infants born at or after 37 weeks of gestation, and not prior (at 35 or 36 weeks of gestation), and only for discrimination of the human vocal signals; no change occurs in discrimination of the nonhuman signals over the 35- to 40-week gestational ages tested. The neural evidence of discrimination of the vocal happy-neutral distinction and the absence of the discrimination of the control signals is convincing. The authors interpret this as a 'landmark' in infants' ability to detect changes in emotional vocal signals, and remark on the potential value of the test as a marker of the infants' interest in emotional signals, underscoring the fact that children at risk for autism spectrum disorder may not show the discrimination. Although the finding is novel and interesting, additional discussion is essential so that readers understand two potential caveats affecting this interpretation.

Comments on the revised version:

The revised manuscript does discuss the limitations of the control stimuli, as well as the limitations with regard to conclusions that can be drawn from this data set. I therefore expected the authors to temper a bit their recommendation that this could be a 'screening' signal for autism because these data are not sufficiently strong to make that recommendation. Also, in the same vein, perhaps the title might be adjusted somewhat to suggest less certainty, for example, by using the word "change" rather than "milestone"'? The data are of interest, but the limitations are genuine limitations.

Author response:

The following is the authors’ response to the original reviews.

Public Reviews:

Reviewer #1 (Public Review):

More details should be provided in terms of inclusion and exclusion criteria for the participants, as well as missing data due to the non-cooperation of newborns during the experimental process. Potential differences between preterm and full-term infants are worth exploring. Several aspects of EEG data analyses and data interpretation should be better clarified.

Here I have several comments and questions to improve the manuscript.

(1) It would be wise to know whether there was any missing data due to the non-cooperation of newborns during the experimental process.

Thank you for the suggestion. While our initial aim was to include 120 neonates in the final data analysis, we actually recruited 198 neonatal participants for this study. The 78 EEG datasets were excluded from the data analysis due to non-cooperation of neonates (n = 75) or technical issues (n = 3). We have incorporated this detailed information in the Subjects subsection (lines 375-383) in the revised manuscript.

(2) The authors investigated the impact of gestational age on emotional perceptual sensitivity in newborns by grouping infants of varying gestational ages in the experiment. The methods section mentions that the study conducted experiments within 24 hours after the birth of the newborns. When do preterm infants (with a gestational age of 35 and 36 weeks) begin to exhibit emotional discrimination comparable to full-term newborns?

This is indeed an intriguing question that merits exploration. However, in our study, we recruited relatively healthy preterm neonates, many of whom were discharged from the hospital with their mothers within 3-5 days after birth. It would have been challenging to arrange for another EEG testing session once these preterm infants reached full-term age, as their parents were unwilling to return to the hospital.

(3) When analyzing EEG data, excluding artifacts with peak deviations exceeding ±200 μV is a relatively lenient criterion, potentially resulting in the retention of some large-amplitude artifacts or noise. What is the rationale behind the author's choice of this criterion? Or, in other words, what considerations led to this specific selection?

In our standard practice, we typically employ a stricter threshold of ±100 μV for artifact removal in studies involving healthy adults and a median threshold of ±150 μV for data from adult patients, such as those with schizophrenia. However, when analyzing neonatal data, we often resort to the loosest criterion of ±200 μV. This decision is primarily due to the inherent challenges associated with neonatal EEG recordings, as we cannot expect newborns to cooperate or remain quiet during the recording process. Consequently, neonatal EEG data tend to contain more artifacts compared to those from healthy adults. Furthermore, the excitability of the newborn brain is notably elevated. This heightened excitability arises from an imbalance in the distribution and function of excitatory and inhibitory neurotransmitter systems. Typically, the expression of excitatory neurotransmitters and their receptors surpasses that of inhibitory neurotransmitters, resulting in increased excitability in the immature brain. This heightened excitability can occasionally lead to the occurrence of paroxysmal electrical activity. As a result, neonatal EEG recordings may at times display large amplitudes, exceeding even 100 μV. In this revision, we have referenced other neonatal/infant EEG studies or technique pipelines that have used the threshold of ±200 μV to support this criterion (lines 483-484).

(4) In the Discussion section, the authors mentioned the biomarkers, such as the fusiform gyrus and hippocampus, which have been identified as potential predictors of autism risk. It is suggested that the authors briefly elucidate the crucial role of these biomarkers in processing social information, which would enhance the readability and logicality of this manuscript.

Thank you for the thoughtful suggestion. We have expanded the discussion concerning the involvement of the fusiform gyrus and hippocampus in social information processing (lines 314-319).

Reviewer #2 (Public Review):

First, readers need to see spectrograms that show the 0-4000 Hz in more detail, rather than what is now shown (0-10,000 Hz). The vocal signals in clearer spectrograms will show I believe the initial consonant burst and formant frequencies that are unique to human speech and give rise to the perception of the consonant sounds in the vocal signals like 'dada' and 'tutu' that were tested. The control signals will presumably not show these abrupt acoustic changes at their onset, even though they appear (from the oscillograms) to approximate the amplitude envelope. The primary cue distinguishing the happy and neutral signals in both the vocal and control signals is the pitch of the signals (high vs low), but the burst of energy representing the consonants is only contained in the vocal signals; it has no comparable match in the control signals. It is possible that the presence of a sharp acoustic onset (a unique characteristic of consonants in human speech) is especially alerting to the infants, and that this acoustic cue, in the context of the pitch change, enhances discrimination in the vocal case. One way to test this would be to use only vowel sounds to represent the vocal signals, without consonants.

Thank you for your expert comments and considerations. We have redrawn Figure 3 using Praat software with a frequency range of 0-5000 Hz, as suggested by Praat’s default parameters. Based on the spectrograms, we acknowledge the potential role of consonants in accounting for differences in stimuli. Consequently, we have included this consideration as one of the limitations of our study in this revised version (lines 325-330).

Another critical detail that the authors need to include about the signals is an explanation of how the control signals were generated. The text states that the Fo and amplitude envelope of the vocal signals were mimicked in the control signals, but what was the signal used for the controls? Was a pure tone complex modulated, or was pink noise used to generate the control signals? Or were the original vocal signals simply filtered in some way to create the controls, which would preserve the Fo and amplitude envelope? If merely filtered, the control signals still may be perceived as 'vocal' signals, rather than as nonspeech (the Supplement contains the sounds, and some of the control sounds can be perceived, to my ear, as 'vocal' signals).

We sincerely appreciate your attention to detail regarding the generation of control signals. As a non-specialized laboratory in audio editing, our approach involved filtering the original vocal sounds around the fundamental frequency (f0) and ensuring a balanced mean intensity between vocal and nonvocal stimuli (as now stated in lines 432-437). However, it became evident that certain “vocal” components persisted in the control sounds, particularly noticeable in the sound “tutu”. In this revision, we openly acknowledge this oversight (lines 331-333). We extend our gratitude once again for highlighting the importance of meticulous consideration when generating control sounds for a study.

Second, there is no information in the manuscript or supplement about the auditory environment of the participants, nor discussion of the fetus' ability to hear in the womb. In the womb, infants are listening to the mothers' bone-conducted speech (which is full of consonant sounds), and we know from published studies that infants can discern differences not only in the prosody of the speech they hear in the womb, but the phonetic characteristics of the mother's speech. The ability at 37 weeks GA or beyond to discriminate the pitch changes in the vocal, but not control signals, could thus be due to additional experience in utero to speech. Another experiential explanation is that the infants born at 37 weeks GA and beyond may be exposed to greater amounts of speech after birth, when compared to those born at 35 and 36 weeks GA, from the attending nurses and from their caregivers, and this speech is also full of consonant sounds. What these infants hear is likely to be 'infant-directed speech,' which is significantly higher in pitch, mirroring the signals tested here. At 37 weeks GA, infants are likely more robust, may sleep less, and are likely more alert. If infants' exposure to speech, either after birth, or their auditory ability to discern differences in speech in utero, is enhanced at 37 weeks GA and beyond, then an 'experience-related' explanation is a viable alternative to a maturational explanation, and should be discussed. Perhaps both are playing a role. As the authors state, many more signals need to be tested to discern how the effect should be interpreted, and other viable interpretations of the current results discussed.

We acknowledge the importance of considering the auditory environment of participants and the fetus' ability to hear in the womb. In our study, neonates were exposed to a native language environment both before and after birth (as added in lines 385-386), and we took efforts to minimize their exposure to speech stimuli other than those used in the experiment. Specifically, all neonates participated the experiment and underwent EEG recording within the first 24 hours after birth (lines 386-387). They were promptly transported to a dedicated testing room for EEG recording as soon as their condition stabilized after birth. During recording sessions, they were separated from their mothers to minimize exposure to natural speech (as added in lines 459-461). As a result, we believe that both preterm and term neonates were exposed to comparable amounts of speech after birth and before the experiment. We also ensured that all participants were in a natural sleep state during EEG recording. However, it is possible that term neonates slept less and were more attentive to the limited speech stimuli in their environment before the experiment compared to preterm newborns.

The debate surrounding nature versus nurture in neonate and infant development persists. We recognize the potential impact of prenatal auditory experiences on neonatal perceptual sensitivity. Therefore, we have added a brief discussion regarding innate- or experience-related explanations for emotional prosodic discrimination in neonates, aiming to shed light on future research directions (lines 343-351).