Randomized controlled trial of a book-sharing intervention in a deprived South African community: effects on carer-infant interactions, and their relation to infant cognitive and socioemotional outcome

Abstract

Background: Consistent with evidence from high-income countries (HICs), we previously showed that, in an informal peri-urban settlement in a low-middle income country, training parents in book sharing with their infants benefitted infant language and attention (Vally, Murray, Tomlinson, & Cooper, ). Here, we investigated whether these benefits were explained by improvements in carer-infant interactions in both book-sharing and non-book-sharing contexts. We also explored whether infant socioemotional development benefitted from book sharing.

Methods: We conducted a randomized controlled trial in Khayelitsha, South Africa. Carers of 14-16-month-old infants were randomized to 8 weeks' training in book sharing (n = 49) or a wait-list control group (n = 42). In addition to the cognitive measures reported previously, independent assessments were made at base line and follow-up of carer-infant interactions during book sharing and toy play. Assessments were also made, at follow-up only, of infant prosocial behaviour in a 'help task', and of infant imitation of doll characters' nonsocial actions and an interpersonal interaction. Eighty-two carer-infant pairs (90%) were assessed at follow-up. (Trial registration ISRCTN39953901).

Results: Carers who received the training showed significant improvements in book-sharing interactions (sensitivity, elaborations, reciprocity), and, to a smaller extent, in toy-play interactions (sensitivity). Infants in the intervention group showed a significantly higher rate of prosocial behaviour, and tended to show more frequent imitation of the interpersonal interaction. Improvements in carer behaviour during book sharing, but not during toy play, mediated intervention effects on all infant cognitive outcomes, and tended to mediate intervention effects on infant interpersonal imitation.

Conclusions: Training in book sharing, a simple, inexpensive intervention that has been shown to benefit infant cognitive development in a low-middle income country, also shows promise for improving infant socioemotional outcomes in this context. Benefits are mediated by improvements in carer-infant interactions, particularly in book-sharing contexts.

Keywords: Parent-child interactions; attention; language; parent training; prosocial behaviour.

Figure 1

CONSORT Diagram Showing Flow of Participants through Trial

Figure 2

Multiple Mediator and Multiple Outcome Model, with Standardised Weights for Significant Effects (shown in bold). Paths from baseline to follow-up measures, and covariances were estimated but, for clarity, are not shown.