UCLA

Autism spectrum disorder (autism) is a heterogeneous, highly heritable neurodevelopmental condition. Language delays are highly prevalent among children with autism or with an elevated likelihood of autism, and delays in spoken language development are often the earliest concerns reported by parents of autistic children. However, there is remarkable heterogeneity in language skills and trajectories among autistic children. There is thus intense interest in identifying early markers that may shed light on language variability in autism. This dissertation sought to fill these gaps by studying individual variability in receptive and expressive language abilities, cross-sectionally and longitudinally, among community-referred 12- to 23-month-olds with autism symptoms. The studies reported here utilized a multimodal approach to identify behavioral and neural mechanisms associated with variability in language abilities in emerging autism.Study 1 examined receptive-expressive language phenotypes (i.e., the extent to which receptive and expressive language abilities were of a similar developmental level). Nearly half of children exhibited an atypical “expressive advantage” language profile characterized by stronger expressive language skills than receptive language skills. Contrary to hypotheses, there was no evidence for significant concurrent associations between receptive-expressive language phenotypes and proposed demographic, cognitive, social communication, and behavioral predictors. However, receptive-expressive language phenotype at T1 did significantly predict rate of expressive language growth: children with a greater expressive language advantage (i.e., a greater delay in receptive language relative to their own expressive language level) exhibited significantly slower expressive language growth over 12 months than children with a receptive language level more similar to or exceeding their own expressive language level at T1. Leveraging task-free electroencephalography (EEG) data acquired at T1, Study 2 examined EEG correlates of concurrent receptive-expressive language phenotypes and individual differences in receptive and expressive language growth. Data-driven analyses revealed significant, positive associations between spontaneous theta (3-6 Hz) power and receptive-expressive phenotypes and significant associations between spontaneous alpha power (6-9 Hz) and rate of both receptive language and expressive language growth. These findings suggest that early receptive-expressive language profiles are a meaningful prognostic marker of language delay, and EEG-based metrics may be sensitive to individual differences in neurocognitive mechanisms that shape language growth.