Posttraumatic stress disorder and epigenetic signatures of inflammation in middle-aged women.

backgroundPost-traumatic stress disorder (PTSD) has been linked to elevated serological and plasma concentrations of inflammatory biomarkers in large cross-sectional studies. Methylation signatures for inflammation provide an assessment of inflammatory state beyond the time of blood draw, but such signatures have not been examined in relation to PTSD. In this study, we hypothesized that PTSD would be associated with inflammation-related epigenetic signatures and changes in these levels over time.

methodsData are from 831 trauma-exposed women in the Nurses' Health Study II. Lifetime PTSD diagnosis and symptom severity were evaluated using a structured interview. Epigenetic signatures of C-reactive protein (CRP) and interleukin-6 (IL-6) were assessed using DNAm measured in two blood samples collected after individuals experienced trauma exposure, 11-16 years apart. Epigenetic signatures were derived using previously published models trained in independent samples (primary signatures). Additional signatures were assessed to evaluate the consistency of associations across signatures (secondary signatures). Measurements of plasma CRP and IL-6 levels using conventional assays were also available from the same blood collections for a subset of women (CRP: n = 345, IL-6: n = 135). We evaluated the correlation between epigenetic signatures and the concentrations measured with conventional assays, and estimated associations between PTSD and inflammation (epigenetic and plasma) levels and changes over time using linear mixed-effects models.

resultsEpigenetic signatures of CRP and IL-6 showed small to medium positive correlations with their corresponding plasma biomarkers from the same blood collections (Pearson r = 0.16-0.33 for CRP and r = -0.02-0.31 for IL-6). After adjusting for sociodemographic factors (e.g., age at blood draw, race and ethnicity), PTSD symptom severity and diagnosis were associated with higher levels of multiple CRP epigenetic signatures (e.g., b=0.194, 95 % CI: 0.047-0.340, for the association between PTSD diagnosis and primary signature of CRP) and plasma CRP concentration at comparable magnitudes. There was suggestive evidence that both PTSD symptom severity and PTSD diagnosis were associated with a faster increase in one of the three epigenetic CRP signatures over time, but not in rate of change in plasma CRP. Weak positive associations were observed between both measures of PTSD and two epigenetic IL-6 signatures, achieving statistical significance only after adjusting for white blood cell proportions and not observed with plasma IL-6. PTSD was not associated with changes in either epigenetic or plasma IL-6 levels.

conclusionPTSD was associated with higher levels of multiple CRP epigenetic signatures, as well as a more rapid increase over time in the level of one of the epigenetic signatures. While analysis suggested a weak association between PTSD and IL-6 epigenetic signatures, further research is needed to confirm these findings and clarify their implications.