Prosodic and Linguistic Features of Child-Directed Speech in Word Learning Contexts: A Comparison Between Preterm and Full-Term Children

Background: Caregiver–child interactions, and specifically child-directed speech, play a central role in early language acquisition by scaffolding children's attentional and lexical development. The features and developmental effects of child-directed speech may vary depending on the child's developmental status and communicative needs, especially in clinical risk conditions such as preterm birth. Aims: This study examined maternal child-directed speech in a semi-structured context, focusing on prosodic cues (mean pitch, pitch range, and speech rate) and structural features (lexical diversity, verbosity, and syntactic complexity) in preterm (PT) and full-term (FT) 18-month-old toddlers interacting with their mothers. We explored (a) differences in child-directed speech between PT and FT groups, (b) the modulation of prosody based on word familiarity, and (c) associations between child-directed speech characteristics, children's attentional engagement (visual exploration of familiar/unfamiliar objects), and vocabulary skills. Methods and Procedures: Forty-three mother–child dyads (20 PT, 23 FT) participated in the study. Maternal speech was transcribed and analysed for linguistic features, while target word tokens were acoustically analysed with PRAAT for prosodic measures. Child visual exploration of objects was coded from video recordings, and vocabulary skills were assessed using validated measures. Outcomes and Results: All mothers modulated prosody to highlight novel words, using higher pitch, wider pitch range, and slower speech rate for unfamiliar versus familiar referents. Mothers of FT children produced more frequent and lexically diverse speech compared to PT counterparts, though verbosity did not differ between groups. In the FT group, slower maternal speech rate was associated with reduced visual exploration of objects, suggesting dynamic adaptations to children's attention. No such association was found in PT dyads. In the PT group, receptive vocabulary index was positively correlated with maternal verbosity, lexical diversity and syntactic complexity. Conclusions and Implications: Findings highlight both shared and group-specific patterns in maternal child-directed speech effects. While prosodic adjustments to word familiarity are preserved in PT dyads, structural aspects of maternal input appear more closely linked to receptive vocabulary in PT children, possibly reflecting adaptive scaffolding strategies. These results underscore the importance of considering both prosodic and structural features of child-directed speech and child conditions in informing targeted early interventions. WHAT THIS PAPER ADDS: What is already known on this subject CDS is essential for early language acquisition by enhancing attention and vocabulary. However, CDS characteristics may vary with child developmental status, and little is known about differences in maternal speech between preterm and full-term toddlers in relation to word learning. What this paper adds to existing knowledge Both preterm and full-term mothers modulate prosody to mark unfamiliar words, yet full-term mothers use more frequent and lexically diverse speech. Structural aspects of maternal input correlate more strongly with receptive vocabulary in preterm children, indicating possible adaptive strategies. What are the potential or actual clinical implications of this work? These results support the need for early interventions tailored to preterm children that target both prosodic and structural features of maternal speech to optimise language development.