Edited by: Iris Nomikou, University of Portsmouth, United Kingdom
Reviewed by: Theano Kokkinaki, University of Crete, Greece; Anne Henning, SRH Hochschule für Gesundheit, Germany
This article was submitted to Developmental Psychology, a section of the journal Frontiers in Psychology
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
Infant-directed speech (IDS), the particular form of spontaneous language observed in interactions between parents and their infants, is a crucial aspect of the mother-infant interaction and an index of the attunement of maternal linguistic input to her infant communicative abilities and needs during dyadic interactions. The present study aimed to explore linguistic and pragmatic features of IDS during mother-infant interactions at 3-month of infant age. The effects of infant (birth status: preterm vs. full-term birth), maternal (perceived parenting stress) and dyadic (dyadic co-regulation) factors on IDS were explored. Results evidenced few differences between the groups on IDS linguistic characteristics. Moreover, observing the interaction of birth status and dyadic co-regulation, full-term mothers varied their IDS pragmatic features according to the quality of co-regulation while preterm mothers did not. Parenting stress was associated to specific linguistic IDS features independently from the birth status. Findings are discussed underling implications for the study of preterm dyads interactions and the importance to consider the interplay of several factors in affecting the quality of IDS.
Infant-directed speech (IDS) is the particular form of spontaneous language observed in interactions between parents and their infants (
Characteristics of IDS and associations with infant development have been studied extensively over the last few decades. The specific vocal patterns and linguistic features of IDS are an important part of mother-infant interactions, and play a role in regulating and attracting infant attention, making linguistic input more apparent and salient to infants, and helping infant interpretation of the emotional signals of adult speakers (
Patterns of child-directed speech have been observed from infancy to childhood (see for example
The specific features of IDS, its important role for infant and child development, its variations with infant age, make IDS a crucial aspect of the mother-infant interaction and an index of the attunement of maternal linguistic input to her infant and child communicative abilities and needs during dyadic interactions (
The main aim of the present study is to contribute to this field of research by exploring how select individual and dyadic factors, previously associated with the quality of the mother-infant interaction, are also associated with IDS features. We examined these associations during the preverbal age of 3 months of age. This age is crucial for several reasons, first because the growing communicative abilities of the infant, i.e., vocalizations, smiles and movements, make him/her contribution to the interaction very relevant, second because at this age maternal voice is considered to be one of the main interactive modalities during dyadic face-to-face interactions, all of this resulting in moments of dyadic shared affect and attention (
Preterm birth–birth before the 37th week of gestational age—is a non-normative birth experience, associated with perturbations in several areas of newborn development. One area of developmental vulnerability centers on the communicative abilities of preterm infants. Compared to infants born full-term, preterm infants, especially during the first months of life, show different communicative abilities. PT infants are reported to be less reactive to social cues than FT infants, and manifest interactions characterized by lower attentional control (
These interactive difficulties complicate the social and affective exchanges between PT infants and their mothers, who may find the infant’s cues and reactions difficult to understand (
Similar mixed findings have been reported by the few studies exploring differences in IDS between PT and FT mothers during dyadic interactions. Some studies found that mothers of PT infants demonstrated contingent vocalizations more frequently (
It is thought that mothers of children born PT may need to modulate their interactions to the appropriate level of linguistic stimulation to avoid over- or underwhelming the PT infant’s communication and arousal regulation capacities (
It is well known that the interactive difficulties of PT mothers are not homogeneous, but vary by conditions related to neonatal and maternal health. Interactional differences are attributed to both the level of neonatal risk, e.g., mothers of more at risk infants have demonstrated lower sensitivity to infants cues (
While it is well described that PT birth poses a risk for suboptimal mother-infant interactions, the impact of preterm birth on the quality of dyadic interactions and maternal wellbeing varies. As a consequence, the quality of maternal verbal communication varies. Considering the role of IDS in infant development, there is a growing need to explore the interactive effects of preterm birth with other potential sources of variability in the quality of mother-infant interactions and their associations with maternal communicative behavior in IDS.
The aim of the present study was to explore the interactive effects of PT birth, maternal parenting stress, and the quality of dyadic co-regulation on the linguistic and pragmatic features of IDS. We considered dyadic co-regulation as a form of dyadic process that consider both the infant and the mother behaviors. Co-regulation refers to a form of coordinated action between participants that involves a continuous mutual adjustment of actions and intentions (
One hundred and one mothers and their 3 month-old (corrected age for PT) infants (PT = 56 and FT = 55) participated in the study. Among those, 14 mothers of PT infants and 12 mothers of FT infants were excluded due to failure to complete study questionnaires (i.e., Parenting Stress Index). The final sample consisted of 86 dyads (PT = 42 and FT = 44).
Preterm infants born <37 weeks gestational age were included. Exclusion criteria for both preterm and full-term groups were the presence of genetic abnormalities, severe neurodevelopmental impairment, and/or neurosensory disabilities (e.g., blindness or deafness).
Most mothers (mean age: PT = 33.63, SD = 5.05; FT = 35.23, SD = 4.84) had a middle or high school level of education: 38.8% had a high school degree; 55.3% graduated college or had a master’s degree; 5.9% had less than a high school education. Preterm infants (43% Males; 78% First born) and FT infants (59% Males; 66% First born) had a mean gestational age of 30.71 (SD = 2.63) and 39.41 (SD = 1.18) weeks, and a mean birth weight of 1,379 (SD = 437.92) and 3,397 (SD = 406.97) grams, respectively. All infants were singletons.
Mothers were invited to participate with their infant in a videotaped observational session when their infant was 3 months old (for preterm infants the corrected age was used). FT dyads were recruited via public services and advertisements, PT dyads were recruited by nurses and doctors of the hospital where they were born. All mothers completed and signed a consensus form before participation. The session consisted of 3 min of face-to-face interaction with the infant seated on an infant seat, and the mother seated directly in front of the infant, facing a mirror which was located behind the infant’s seat so that both partners’ faces could be clearly seen. After a brief introduction, mothers were asked to interact with their infants, as they would do at home.
After the interactive episode, mothers were asked to complete the short version of the Parenting Stress Index. Each session was entirely transcribed according to the CHAT transcription system (Codes for the Human Analysis of Transcripts) of the CHILDES computational system (Child Language Data Exchange System) (
The study was approved by the Ethical Committee of the Department.
Mother–infant interactions were coded using the Revised Relational Coding System (
The coding system includes five global categories of communicative interactions: unilateral, asymmetrical, symmetrical, disruptive, and unengaged. In the present study we considered the global categories of
The co-regulation patterns were coded every second from the videotapes by a trained coder, using the Mangold Interact 18 software. The relative total duration of each pattern was computed. An independently trained coder processed 25% of the sessions to compute inter-observer reliability. The Kappa values were 0.86 for symmetrical co-regulation and 0.94 for unilateral co-regulation.
Mothers were asked to complete the PSI-Short Form questionnaire (PSI-SF; (
Maternal vocal productions were coded in order to analyze:
Verbosity: Rate per minute of utterances, word types and tokens (
Lexical variability: Type/token ratio (
Syntactic complexity:
In order to classify the pragmatic meaning of maternal productions, these were divided into the following categories and the percentages for each category over the total number of utterances considered were calculated (
Conversational: Sentences used to promote and maintain the conversation with the infant (i.e., emphatic sentences and comments “You look happy,” open questions “What are you looking at?”, comments on the present/past activity of the infant “We are playing together”).
Control: Sentences used to re-orient infant attention, to direct infant attention toward something (i.e., direct requests “Speak to me,” claiming infant attention “Ehy, look at me”).
Preverbal: Sounds and sentences using typical baby-talk words, repetition of infant’s sounds.
We first computed descriptive statistics and bivariate correlations among study variables in the full sample and separately in the two groups. The two birth groups were compared for mean values along the investigated variables. Afterward, to explore the single and additive role of birth status, dyadic co-regulation and parenting stress, and the interplay between dyadic co-regulation and parenting stress with birth status on IDS characteristics, we estimated and compared several multivariate models and then explored parameters of the best selected model More specifically, pertaining to predictors, first birth status was included in the model. Then, we considered the additive role of the psychological variables investigated (co-regulation and parenting stress). Finally, we included the interaction term between birth status and each of the dyadic/maternal variables considered to see if co-regulation and parenting stress differently predicted IDS depending on birth status. The Akaike Information Criterion (AIC) was used for model comparison, with lower values providing more support to a model against the others. The first group of multivariate models included as outcome variables IDS linguistic characteristics (verbosity, TTR and MLU), and then we considered IDS pragmatic characteristics (conversational, control and preverbal sentences) Regression parameters were explored for the best fitting model. Analyses were run using the statistical software R, using Lavaan package.
Means, SDs, and correlation values among variables of interest in the full sample are reported in
Descriptive and bivariate correlations for the full sample.
Mean (SD) | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | ||
1 | Verbosity | 26.16 (8.52) | − | ||||||||
2 | TTR | 0.40 (0.09) | −0.46 |
− | |||||||
3 | MLU | 4.07 (0.94) | −0.19 | 0.10 | − | ||||||
4 | % Conversational | 56.84 (15.36) | −0.19 | 0.01 | 0.46 |
− | |||||
5 | % Control | 30.89 (13.97) | 0.25 |
−0.22 |
−0.30 |
−0.68 |
− | ||||
6 | % Preverbal | 7.59 (10.77) | −0.08 | 0.25 |
−0.33 |
−0.42 |
−0.30 |
− | |||
7 | Parenting stress | 1.83 (0.56) | −0.16 | 0.28 |
0.22 |
0.08 | −0.14 | 0.15 | − | ||
8 | Unilateral co-regulation | 0.31 (0.26) | −0.09 | 0.21 | −0.11 | −0.09 | 0.19 | −0.05 | 0.08 | − | |
9 | Symmetrical co-regulation | 0.21 (0.16) | 0.18 | −0.16 | 0.06 | 0.02 | −0.18 | 0.15 | 0.03 | −0.56 |
− |
Exploration of bivariate associations among investigated variables run separately for the two birth status groups (see
Descriptive and bivariate correlations in the preterm (above the diagonal,
Mean (SD) PT | Mean (SD) FT | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | ||
1 | Verbosity | 21.91 (6.32) | 30.20 (8.43) | − | −0.43 |
−0.01 | −0.28 | 0.19 | 0.03 | −0.10 | −0.11 | 0.10 |
2 | TTR | 0.41 (0.09) | 0.39 0.09) | −0.55 |
− | −0.14 | 0.02 | −0.19 | 0.33 |
0.24 | 0.33 |
−0.19 |
3 | MLU | 4.33 (0.91) | 3.83 0.91) | −0.12 | 0.31 |
− | 0.47 |
−0.43 |
−0.17 | 0.18 | −0.17 | 0.18 |
4 | % Conversational | 60.77 (14.13) | 53.09 (15.69) | 0.05 | −0.03 | 0.39 |
− | −0.77 |
−0.29 | 0.17 | 0.10 | −0.08 |
5 | % Control | 29.02 (13.47) | 32.68 (14.35) | 0.24 | −0.24 | −0.13 | −0.59 |
− | −0.22 | −0.14 | −0.08 | 0.06 |
6 | % Preverbal | 5.27 (6.67) | 9.80 (13.30) | −0.31 |
0.26 | −0.37 |
−0.45 |
−0.41 |
− | 0.07 | 0.21 | −0.13 |
7 | Parenting Stress | 1.75 (0.54) | 1.90 (0.58) | −0.37 |
0.34 |
0.36 |
0.08 | −0.18 | 0.16 | − | 0.30 | −0.06 |
8 | Unilateral co-regulation | 0.34 (0.28) | 0.29 (0.23) | 0.01 | 0.06 | −0.11 | −0.34 |
0.52 |
−0.17 | −0.12 | − | −0.54 |
9 | Symmetrical co-regulation | 0.19 (0.17) | 0.23 (0.16) | 0.18 | −0.13 | 0.01 | 0.18 | −0.44 |
0.27 | 0.08 | −0.58 |
− |
The One Way ANOVA evidenced only few differences between PT and FT groups. When speaking to their infants, mothers of PT infants, talked less [verbosity:
Comparison of multivariate regression models (see
Model comparison, effects of birth status, and dyadic co-regulation: AIC.
IDS linguistic | IDS pragmatic | |
Model | AIC | AIC |
Model 1: Birth status | 262.51 | 1876.5 |
Model 2: Birth status, unilateral co-regulation | 1876.3 | |
Model 3: Birth status, unilateral co-regulation, birth status × unilateral co-regulation | 266.98 | |
Model 1: Birth status | 1876.5 | |
Model 2: Birth status, symmetrical co-regulation | 264.63 | 1877.5 |
Model 3: Birth status, symmetrical co-regulation, birth status × symmetrical co-regulation | 269.31 |
Multivariate analysis on IDS linguistic and pragmatic characteristics: Standardized estimated parameters of models 2 and 3, respectively.
Verbosity |
MLU |
Conversational |
Control |
Preverbal |
||
β |
β |
β |
β |
β |
β |
|
Birth status | 0.97 (<0.001) | −0.01 (0.712) | −0.52 (0.007) | 0.67 (0.889) | −6.85 (0.113) | 8.98 (0.011) |
Unilateral co-regulation | −0.15 (0.682) | 0.07 (0.05) | −0.49 (0.189) | 33.33 (0.068) | −39.68 (0.015) | 19.55 (0.139) |
Birth status × Unilateral co-regulation | − | − | − | −28.20 (0.021) | 35.95 (0.001) | −14.68 (0.096) |
Comparison of multivariate regression models (see
Interaction among birth status and unilateral co-regulation on IDS conversational pragmatic sentences. PT, preterm; FT, full-term.
Interaction among birth status and unilateral co-regulation on IDS control pragmatic sentences. PT, preterm; FT, full-term.
Comparison of multivariate regression models (see
Multivariate analysis on IDS linguistic and pragmatic characteristics: Standardized estimated parameters of models 1 and 3, respectively.
Verbosity |
MLU |
Conversational |
Control |
Preverbal |
||
β |
β |
β |
β |
β |
β |
|
Birth status | 0.97 (<0.001) | −0.01 (0.586) | −0.50 (0.011) | −12.87 (0.013) | 13.67 (0.003) | −1.52 (0.672) |
Symmetrical co-regulation | − | − | − | −30.36 (0.319) | 49.33 (0.068) | −32.64 (0.123) |
Birth status × Symmetrical co-regulation | − | − | − | 23.91 (0.220) | −44.74 (0.010) | 27.44 (0.042) |
Comparison of multivariate regression models (see
Interaction among birth status and symmetrical co-regulation on IDS control pragmatic sentences. PT, preterm, FT, full-term.
Interaction among birth status and symmetrical co-regulation on IDS preverbal pragmatic sentences. PT, preterm, FT, full-term.
Comparison of multivariate regression models (see
Model comparison, effects of birth status and parenting stress: AIC.
IDS linguistic | IDS pragmatic | |
Model | AIC | AIC |
Model 1: Birth status | 262.51 | |
Model 2: Birth status, parenting stress | 1877.4 | |
Model 3: Birth status, parenting stress, birth status × parenting stress | 256.26 | 1882.9 |
Multivariate analysis on IDS linguistic and pragmatic characteristics: Standardized estimated parameters of models 2 and 1, respectively.
Verbosity |
MLU |
Conversational |
Control |
Preverbal |
||
β |
β |
β |
β |
β |
β |
|
Birth status | 1.03 (<0.001) | −0.02 (0.346) | −0.56 (0.003) | −7.68 (0.016) | 3.66 (0.217) | 4.52 (0.045) |
Parenting stress | −0.40 (0.012) | 0.04 (0.005) | 0.44 (0.008) | − | − | − |
Comparison of multivariate regression models (see
The present study aimed to explore the intra- and inter-individual differences of the linguistic and pragmatic features of IDS, directed to 3-month-old full-term and preterm infants. We found evidence of interactive effects between individual (infant and maternal) and dyadic factors, on IDS characteristics. Recognizing that birth status (i.e., preterm birth) is an important condition affecting the infant, mother and dyad (
Regarding the role of dyadic co-regulation and IDS in PT versus FT infants, we found no differences between the groups concerning the quality of the dyadic, symmetrical and unilateral, co-regulation. Expressed differently, our results suggest that different patterns of IDS (between FT and PT) are not associated with fewer moments of co-regulated attention and affect. This is relevant because many studies considered the interactive qualities of PT mothers to be under- or over-stimulating, which was presumed to be suboptimal compared to what was observed in typical FT mothers (
One notable difference emerged between the groups when exploring how mothers vary their IDS with respect to the time spent in co-regulated interaction. We observed that the duration of shared versus un-shared moments was associated with different pragmatic features of IDS in FT, but not PT, mothers. In FT dyads, the greater time spent in moments of un-shared attention and affect (i.e., unilateral co-regulation) was associated with an IDS characterized by reduced conversational and increased control sentences. Conversely, the greater time spent in shared moments of attention and affect (i.e., symmetrical co-regulation) was associated with increased use of preverbal, and decreased use of control sentences. Conversational sentences have, as its primary purpose, to keep open the communicative channels between two individuals when they are engaged in the same subject or are having a shared emotional experience. This is manifest by, for example, making comments, offering compliments, and asking open-ended questions. In contrast, control sentences are used to redirect, modify, and capture the attention of another when the other individual is focused on something different. This is manifested by, for example, calling or giving orders (
Consistent with previous studies, we did not find differences in parenting stress between the two groups (
We would like to acknowledge some limitations of the current study. First, while our results were relevant, our sample sizes were not big and we included only 3-month-old infants. A larger sample followed longitudinally would have allowed us to explore additional interactive effects. An additional limitation is that our PT sample was quite homogeneous, composed of low-risk PT infants and of well educated low-risk mothers. Additional research should examine these associations in a more at-risk population of preterms to identify differential effects both at maternal, infant, and dyadic levels. With a larger and more at-risk sample, also the associations of IDS with the other co-regulation patterns, i.e., asymmetrical and unengaged, could be explored. Further studies should also consider the paternal dyadic communication and explore if these findings are replicable in father-infant dyads. Lastly, we did not explore infant vocal communication during the interaction. This information, as well as the inclusion of non-verbal infant and maternal cues, would help interpret our findings and should be the focus of future studies.
Despite these limits, this study presents several strengths. To the best of our knowledge, this is the first study to examine infant, maternal, and dyadic factors and their associations with characteristics of IDS. This study highlights the need to go beyond exploring IDS effects on language development, and to consider its potential importance when exploring the quality of dyadic interaction and its role in sharing attention, affect and meaning between the mother and the infant (
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.
The studies involving human participants were reviewed and approved by the Ethics Committee of the Department of Neuroscience, Imaging and Clinical Sciences of the University G. d’Annunzio Chieti-Pescara. Written informed consent to participate in this study was provided by the participants’ legal guardian/next of kin.
MS and MF conceptualized the study and organized the data collection. MS and FL wrote the first draft of the manuscript. FL ran the analyses and wrote the results section. MG, ML, and SP conducted the data collection and coding. SD and PC coordinated preterm data collection. MF and PS revised the manuscript. All authors contributed to revision of the final version of the manuscript.
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
This work was partially supported with PON-AIM funding (1811283-3) and by the “Departments of Excellence 2018-2022” initiative of the Italian Ministry of Education, University and Research for the Department of Neuroscience, Imaging and Clinical Sciences of the G. d’Annunzio University of Chieti-Pescara.
We thank the families who participated in the study and the students who contributed to the data collection.