EEG functional connectivity in infants at elevated familial likelihood for autism spectrum disorder.

The study was underpowered to detect small or moderate effects; the authors note that only very large differences would have been detected. The analysis was post hoc and exploratory, lacking pre-registration, which increases the risk of false-positive findings in the secondary analyses. Infant EEG is inherently noisy, and a significant portion of recordings was rejected due to artifacts or outlier status, potentially introducing bias. The elevated-likelihood group is defined by sibling status, which may not represent all pathways to autism (e.g., those without family history). Results varied between sites, and many exploratory findings failed to replicate across the two independent datasets.

Funding was provided by the Brain & Behavior Research Foundation, National Institutes of Health, and Medical Research Council. No specific conflicts of interest regarding commercial entities or personal financial gains were declared in the provided text.

Do not interpret the exploratory correlations with symptom severity as evidence that EEG connectivity can currently be used for diagnosis or prognosis. These findings were not significant in the primary analysis, failed statistical correction for multiple testing, and were not consistent across the two study sites.