Genetic Testing Abnormalities in Children with Autism Spectrum Disorder: Prevalence, Patterns, and Clinical Associations
Abstract
Background: Genetic testing in autism spectrum disorder (ASD) can reveal a wide range of chromosomal and sequence-level abnormalities, yet large real-world neurology cohorts rarely report the full spectrum of findings alongside clinical correlates and patterns of testing. We characterized genetic findings in an 1884-patient ASD cohort and examined factors associated with both abnormal findings and the use of genetic testing. Methods: This retrospective cross-sectional study included 1884 patients. Genetic testing was performed in 1011 patients; the primary outcome was an abnormal versus normal genetic result among those tested. All 371 abnormal findings were catalogued in a de-identified supplement. The primary multivariable model included age, sex, seizure history, and EEG abnormality (N = 901); an MRI-inclusive sensitivity model used 422 complete cases, and a separate full-cohort model evaluated factors associated with genetic testing. Results: Abnormal genetic findings were documented in 371/1011 tested patients (36.7%, 95% CI 33.7–39.8%). In the primary model, female sex (aOR 1.483, 95% CI 1.089–2.018; p = 0.012) and seizure history (aOR 1.634, 95% CI 1.136–2.352; p = 0.008) were independently associated with abnormal findings, whereas EEG abnormality was not significant after adjustment (aOR 1.373, p = 0.084). In the MRI-inclusive sensitivity model, female sex and seizure history remained significant, while MRI abnormality was not independently associated (aOR 1.235, p = 0.302). Genetic testing was more common among patients who also underwent EEG or MRI (both p < 0.001). Conclusions: More than one-third of tested patients had an abnormal genetic finding, spanning a broad range of copy-number, sequence, homozygosity, Fragile X-related, chromosomal, and syndromic findings. Seizure history and female sex were the principal adjusted correlates of abnormal findings. Genetic testing clustered with EEG and MRI use, reflecting observed patterns of neurological work-up within this cohort rather than temporal or causal relationships.