2026 ECNP Congress

PS03-3257

Epigenomic embedding of childhood adversity links to disease risk and chronic immune changes

Darina Czamara — D. Juntilla, L. Casten, A. Nounu, A.L. Weihs, M. Erhart, L. Maier, M. Leskien, J. Tuhkanen, A.D. Scherff, N. Yusupov, J. Hagenberg, A.S. Fröhlich, V.N. Karlbauer, S. van der Auwera, T. Brückl, J. Kopf-Beck, S. Egli, H. Völzke, U. Völker, A. Weihs, A. Teumer, J. Winkelmann, H. Prokisch, J. Knauer-Arloth, L. Feldmann, L. Miranda, B. Müller-Myhsok, B.W. Dunlop, W.E. Craighead, J.C. Felger, A.H. Miller, C.B. Nemeroff, F. Hoffmann, S. Entringer, C. Buss, S.M. Winter, N. Matosin, H.S. Mayberg, C.M. Heim, M. Standl, J. Lahti, K. Räikkönen, E. Greimel, G. Schulte-Körne, H.J. Grabe, A. Pete, M. Waldenberger, E.B. Binder · Max-Planck-Institute of Psychiatry, Department Genes and Environment, Munich, Germany

2026 ECNP Congress

Background

Childhood adversity (CA) is a major environmental risk factor for a wide range of psychiatric and somatic disorders across the lifespan. Marked inter-individual differences in outcomes suggest that genetic background moderates biological sensitivity to adversity. Gene × environment (G×E) interplay may shape vulnerability and resilience following early-life stress. DNA methylation (DNAm), a key epigenetic mechanism regulating gene expression, is a plausible molecular mechanism through which CA becomes biologically embedded. Immune-related epigenetic alterations are particularly relevant, given the central role of immune dysregulation in the long-term health consequences of CA. We aimed to characterize the epigenetic embedding of CA in the context of common genetic variation by identifying contextual methylation quantitative trait loci (contmeQTLs), defined as genetic variants that modulate DNAm responses to CA exposure.

Methods

We performed an epigenome-wide G×E analysis in peripheral blood across six cohorts comprising 3,471 adolescents and adults. Linear regression models tested interactions between CA exposure and single nucleotide polymorphisms (SNPs) on DNAm at CpG sites, adjusting for relevant biological and technical covariates. Cohort-specific results were combined using random-effects meta-analysis. Replication was performed in three independent childhood cohorts (N=780). Significant contmeQTLs were functionally annotated for genomic context, regulatory features, and disease relevance. Pathway enrichment analyses were conducted, and putative causal relevance for complex traits was assessed using Mendelian randomization (MR) and genome-wide association study (GWAS) annotations. We further derived a polygenic interaction risk score indexing genetically moderated DNAm reactivity to CA and tested its association with inflammatory plasma profiles and disease burden in UK Biobank. To examine evolutionary persistence, we tested signatures of selection in >15,000 ancient human genomes from the Allen Ancient DNA Resource, comparing allele frequency trajectories of human-derived versus ancestral contmeQTL alleles and modeling allele persistence over time.

Results

We identified 5,120 contmeQTLs associated with CA. Twenty percent replicated in childhood cohorts, supporting robustness across developmental stages. Functional annotation showed enrichment of contmeQTL SNPs and CpGs in immune-related regulatory regions, particularly within the major histocompatibility complex. These loci were enriched for associations with psychiatric and somatic disorders in GWAS and were linked via MR to disease-relevant traits. contmeQTL-associated CpGs also overlapped with differential DNAm signatures observed in post-mortem brain tissue from individuals with psychiatric disorders. In UK Biobank, the polygenic interaction risk score was associated with a pro-inflammatory plasma profile exclusively in CA-exposed individuals, this profile predicted greater multimorbidity. Evolutionary analyses showed that human-derived contmeQTL alleles carried signatures of recent positive selection. A higher polygenic interaction risk score predicted later age at death in ancient humans, suggesting a survival advantage. These same variants have been linked to higher cognitive performance, but also increased liability for psychosis and substance use.

Conclusions

Common genetic variation shapes the epigenetic embedding of childhood adversity through context-dependent effects on DNAm, particularly in immune-related pathways. contmeQTLs provide a mechanistic framework linking early adversity to later-life psychiatric and somatic disease risk through genetically moderated epigenetic plasticity. Their evolutionary persistence suggests a trade-off in which heightened environmental sensitivity may have conferred adaptive advantages in ancestral environments while increasing psychiatric vulnerability in modern contexts.

References

No references