Synaptic density imaging associates with performance but not self-report in autism

Autism is a common neurodevelopmental condition with considerable between-person variability and unknown biological etiology. Diagnosis of autism in adults relies on measurement of performance (e.g., clinician rating of social behavior, standardized assessment of cognition) and self-report of autism features. Our prior work revealed a strong relationship between clinician ratings of social function in autism and in vivo measures of synaptic density as measured by [ ¹¹ C]UCB-J PET (Matuskey et al., 2025). However, it is not understood how the measured aspects of the autism phenotype index underlying biology. Here, in a sample of 12 autistic adults, we address this question by relating synaptic density imaging to a range of performance measures, such as cognition and face recognition, in contrast to self-report measures. Our results identify a dissociation between these measurement types in that synaptic density correlates strongly with measures of performance but shows virtually no relationship with self-report. These findings suggest that, while self-report and performance measures share variance in characterizing phenotypes in autism, they diverge when mapping onto the underlying biology. Results suggest that higher synaptic density may index a general mechanism that facilitates social and cognitive performance across a range of tasks. These data also indicate that performance and self-report measures are not interchangeable in capturing brain-behavior correspondences in autism, a finding that can guide measure selection in future biomarker studies and clinical research.