Genetic Susceptibility to Neurodevelopmental Conditions Is Associated With Neonatal DNA Methylation Patterns in the...

Autism spectrum disorder (ASD), attention-deficit/hyperactivity disorder (ADHD), and schizophrenia (SCZ) are highly heritable and linked to disruptions in fetal neurodevelopment. Epigenetic processes, such as DNA methylation (DNAm), are considered a key pathway of interest. However, it is unclear whether 1) genetic susceptibility to neurodevelopmental conditions (NDCs) is associated with DNAm patterns already at birth, 2) DNAm patterns are unique or shared across conditions, and 3) neonatal DNAm patterns can be leveraged to enhance genetic prediction of neurodevelopmental outcomes.

We conducted epigenome-wide meta-analyses of genetic susceptibility to ASD, ADHD, and SCZ (measured with polygenic scores [PGSs]) and cord blood DNAm in 4 European population-based cohorts (n_pooled = 5802; 50.2% female). We estimated DNAm pattern overlap between PGSs using heterogeneity statistics. Furthermore, we built methylation profile scores for each PGS to test incremental variance explained over genetic data alone in 130 developmental outcomes from birth to 14 years.

In probe-level analyses, the SCZ PGS was associated with neonatal DNAm at 246 loci (p < 9 × 10^-8), predominantly in the major histocompatibility complex, supporting an early-origins perspective on SCZ. Functional characterization confirmed strong genetic effects, blood-brain concordance, and enrichment for immune-related pathways. Eight loci were identified for the ASD PGS (mapping to FDFT1 and MFHAS1) and none for the ADHD PGS. Differentially methylated regions were detected across PGSs (130-166 regions). Overall, DNAm signals were largely distinct between conditions. Incorporating neonatal DNAm data in genetic prediction models nominally increased the explained variance for several cognitive and motor outcomes.

Genetic susceptibility to NDCs, particularly SCZ, is detectable in cord blood DNAm in the general population.