We examined the trajectory of association among SCZ PRS with EEG coherence (theta, alpha, and beta). Association trajectories of PRS with EEG coherence were calculated as described in detail previously22,28. Given previous research demonstrating sex differences both in measures of EEG28, the clinical presentation of SCZ29, and polygenic effects on neuropsychiatric illnesses22,30,31, sex-specific models were examined. Covariates in the model included genotyping array, ancestral principal components (PCs 1–3), and a sex × PRS interaction term. Weights for individuals were adjusted to account for multiple observations on single individuals and co-presence of related individuals (e.g., siblings). Given the mixed evidence regarding the influence of substance use on the onset of psychosis45, and the potential for confounding introduced by the enrichment of substance use/disorders within the COGA derived sample, we examined models adjusted for the regular use of alcohol and nicotine, and ever use of cannabis on PRS-EEG coherence associations. We next examined specific genetic variants included in the PRS to determine whether any individual SCZ loci are driving this finding. Note that 101 of 108 loci reported by the PGC were available for analysis in the COGA sample.
