Trajectories of expressive language development are highly heterogeneous in autism. Yoder et al. found that parental responsiveness, child response to joint attention, child communicative intent and consonant inventory were unique predictors of expressive language growth in minimally verbal preschoolers 16 months later ( n = 87). This study applied these predictors to an independent sample, over a 12-month period ( n = 27). A broader measure of phonetic repertoire, combining reported, elicited and observed speech sounds, was included to further understand the contribution of speech production skills. Expressive language growth was highly variable: 65% remained minimally verbal at mean age 5;2, while 7% gained over 340 words. Contrary to expectations, communicative intent, parent responsiveness and response to joint attention were not found to predict expressive language growth or outcome. In contrast, both consonant inventory and phonetic repertoire were significant predictors (adjusted R2 = 0.29 and 0.45). These results underscore the contribution of speech production abilities to expressive language development in this population, which may reflect an additional deficit rather than a consequence of core autism symptoms. Future work should include those with the most persistently limited expressive language, so that findings can be generalized and additional barriers to communication identified and addressed. Lay abstract What is already known about the topic? Language skills vary enormously in autism: while some autistic individuals can engage in sophisticated conversations, many remain minimally verbal, meaning they use few or no words regularly for communication. We do not know what causes this variation, but we do know that certain child and family characteristics can be measured when a child is young, and this information can improve our prediction of how expressive language might develop over time. What this article adds? We examined four characteristics, which have already been found to predict language development in young minimally verbal autistic children. We followed the expressive language progress of 27 minimally verbal children, aged three to five, for a year. One-third no longer met the minimally verbal criteria at the end of the study. In this sample, only one factor predicted language progress, which was the child’s initial speech skills (the number of different speech sounds that the child made during an interaction). This finding adds to the evidence that speech skills contribute to language development in autism. In some cases, persistent and severe expressive language difficulties may reflect an additional deficit in speech production, rather than a consequence of core autism features. Implications for practice, research or policy Our findings suggest that there are factors other than social skills that influence language development in autism. Careful assessment of speech production should be considered when language does not develop as expected. Future research should evaluate speech skills interventions for minimally verbal autistic individuals, as well as promoting the use of alternative communication systems.
Peer Review #2 of "Detecting communicative intent in a computerised test of joint attention (v0.2)"