Mendelian Randomization in IMAGEN Cohort Links Three Genes to Shared Risk Across Six Neuropsychiatric Disorders

Researchers first mapped SNPs influencing DNA methylation and gene expression in the IMAGEN sample, then applied two-step Mendelian randomization and colocalization across large GWAS summary statistics for the six disorders. This yielded 73 causal CpG sites and 62 genes, with three genes showing transdiagnostic effects and enrichment in both psychiatric and autoimmune pathways.

The design integrates quantitative trait loci with causal inference to move past single-disorder candidate-gene approaches that have dominated the field. By using blood from disease-free adolescents, the study isolates upstream molecular signatures rather than secondary consequences of illness or medication, offering a potential scaffold for unifying comorbidity patterns documented in epidemiological registries.

Replication in independent adult cohorts, functional validation in induced pluripotent stem cell models, and longitudinal tracking of methylation trajectories remain essential before any biomarker panel can inform stratified trials or diagnostic algorithms.