This report summarizes epigenome-wide association studies (EWAS) of posttraumatic stress disorder (PTSD) conducted within the United States Veterans Affairs Million Veteran Program (MVP). The EWAS aimed to identify DNA methylation differences associated with PTSD diagnosis and symptom severity, evaluate replication of prior PTSD EWAS findings, and compare peripheral methylation with previously reported brain methylation differences.
The investigators evaluated both the 17-item PTSD Checklist (PCL-17) total score and three symptom subdomains—re-experiencing, avoidance, and hyperarousal—alongside lifetime PTSD diagnosis derived from electronic health records using lasso regression.
The analytic cohort consisted of 22,141 MVP participants with PCL-17 data. Ancestry-stratified sample sizes were: 17,674 European ancestry (EUR), 3,430 African ancestry (AFR), and 1,037 admixed American descent (AMR).
Using lasso regression applied to electronic health record data, the study defined lifetime PTSD cases and controls. Across ancestries there were 9,413 lifetime PTSD cases and 27,257 controls in total. Breakdown by ancestry reported in the source: EUR had 5,184 cases and 19,382 controls; AFR had 3,095 cases and 6,240 controls; AMR had 1,134 cases and 1,635 controls.
The analyses examined epigenome-wide DNA methylation at CpG sites in relation to multiple PTSD-related traits: PCL-Total, the three symptom subdomains, and lifetime PTSD diagnosis. The study assessed epigenome-wide significance across these traits and performed ancestry-stratified assessments.
The investigators also used clinical covariates including smoking status in stratified analyses, and evaluated DNA methylation age measures to test for accelerated epigenetic aging associated with PTSD case status.
Across the evaluated PTSD traits, the MVP EWAS identified a total of 241 epigenome-wide-significant CpG sites associated with PTSD phenotypes.
The study attempted replication of previously reported PTSD EWAS findings from the Psychiatric Genomics Consortium. Of 11 CpG sites reported in that prior study, five were replicated in the MVP EWAS.
These findings extend prior reports by identifying a larger set of PTSD-associated CpG sites across multiple ancestries and by assessing symptom severity as well as case-control status.
The MVP EWAS results were compared to a human postmortem brain study that reported differentially methylated CpG sites in subregions of the amygdala and hippocampus from individuals with PTSD versus controls. At the gene level, 10 genes identified in the brain study overlapped with 192 genes identified in the MVP EWAS. The source reports this gene-level overlap but does not provide the specific gene names in the summary.
The comparison assessed overlap at both the CpG and gene levels; details about the specific overlapping CpG sites or effect directions were not reported in the source abstract.
Gene-set and cell-type enrichment analyses indicated significant enrichment of genes downregulated in somatostatin interneurons and in brain endothelial cells among the genes implicated by the EWAS. This suggests potential relevance of interneuron and vascular-related pathways in the epigenetic architecture of PTSD, as inferred from peripheral methylation signals and gene-level overlap with brain findings.
The source does not provide mechanistic validation or causal inference but highlights these cell-type signals as biologically informative annotations of the EWAS results.
Regression analyses reported in the source show that PTSD case status was significantly associated with accelerated DNA methylation aging. The source reports this association but does not include effect sizes, specific methylation age metrics used, or model details in the abstract.
Prior PTSD EWAS have reported associations between PTSD and methylation of the AHRR gene. In the MVP analyses, stratification by smoking status suggested that the observed association between PTSD and AHRR methylation may be confounded by smoking. The source notes this discrepancy with earlier findings and indicates smoking status as an important covariate to consider in PTSD methylation studies.
Key strengths reported include the large sample size, multi-ancestry design, evaluation of symptom severity domains in addition to lifetime diagnosis, and replication of several prior EWAS findings. This MVP EWAS is described as the largest-scale PTSD EWAS to date and the first large EWAS to examine PTSD symptom severity.
Limitations implicit in the summary include lack of detailed effect estimates and CpG-level information in the abstract, and the potential for residual confounding such as by smoking. The source abstract does not report full methodological details or the full list of implicated CpG sites and genes; those are expected in the full manuscript and deposited data.
All MVP and meta-analysis summary statistics will be made available on dbGaP under accession phs001672 upon publication, as stated in the source. The Million Veteran Program and the study were approved by the Veterans Affairs Central Institutional Review Board. Competing interests declared by some authors are reported in the source.
The MVP PTSD EWAS identified 241 epigenome-wide significant CpG associations across PTSD diagnostic and symptom traits, replicated some prior EWAS signals, showed gene-level overlap with brain methylation findings, found enrichment in somatostatin interneuron and brain endothelial cell–related genes, and associated PTSD with accelerated DNA methylation age. Smoking appears to confound at least some previously reported associations (for example, AHRR methylation). Full results and summary statistics will be available through dbGaP for follow-up analyses.