“Native and Non-Native Speech-Evoked Responses in High-Risk Infant Siblings”

Background: Studies of infants who have older siblings with autism indicate that 30-50% exhibit abnormal sensory behaviors, language and/or social delays as early as 12-months of age (Zwaigenbaum et al., 2005; Landa & Garrett-Mayer, 2006; Mitchell et al., 2006). Although children with autism are a densely heterogeneous population, communication and/or language delays are part of the fundamental triad of impairments contributing to a diagnosis of autism. Previous research has established that young infants can discriminate phonemes used in all languages (Werker & Tees, 1984; Best & McRoberts, 2003), but that between 10 and 12 months of age, native-language phonetic perception ability increases, while the detection of unfamiliar patterns in non-native languages declines (Cheour et al., 1998; Kuhl et al., 2006). Our lab has also demonstrated a predictive relationship between measures of native-language speech perception recorded at 7 months of age and the number of words produced in early childhood. Individual children who had better discrimination of an English syllable in infancy produced more words at 18 and 24 months of age (Kuhl et al., 2008).  

Objectives: The goal of our research is to investigate if an abnormal pattern of speech perception is a risk marker that can be used to facilitate an earlier diagnosis of language delay or an autism spectrum disorder. We will also investigate the presence of a predictive relationship between early measures of speech discrimination and subsequent behavior in high-risk infants.

Methods: EEG data was recorded from high-risk infant siblings and from no-risk infants at 7 months of age. After data collection, high-risk infants were randomly assigned to either the early intervention group or the no treatment group. All infants were re-assessed at 12 months of age. At both ages, English and Chinese syllable contrasts were each presented in an oddball paradigm to test discrimination.

Results: Preliminary results indicate that before 8-months of age, at-risk infants have different physiological responses to English speech syllables compared to no-risk infants. The mismatch negativity is significantly different between groups at electrode FZ. Compared to typically developing infants, high-risk infants continue to process English syllables differently at one year of age; however, both groups of infants continue to process Chinese syllables similarly.  As a group, participation in a relationship-based form of early intervention did not normalize speech-evoked responses to English syllables at one year of age.

Conclusions: Taken together, these results may indicate that some infants who will develop autism have an auditory processing disorder that hinders the acquisition of critical language skills.