Search

Background Psychological stress is associated with accelerated cellular aging and increased risk for aging-related diseases, but the underlying molecular mechanisms are unclear. Methods We examined the effect of stress on a DNA methylation age predictor that was shown to correlate strongly with chronological age across human tissues (Horvath 2013). Genome-wide DNA methylation was measured in peripheral blood using the 450K Illumina array in three independent cohorts: the Grady Trauma Project/GTP (N=366); a panic disorder case/control sample recruited at the Max Planck Institute of Psychiatry/MPI-P (N=318); and the Conte Center for the Psychobiology of Early-Life Trauma/Conte (N=42). Age acceleration was calculated by subtracting chronological age from age predicted by DNA methylation. Psychiatric symptomatology and stressors were assessed using standard questionnaires. Results DNA methylation age strongly correlated with chronological age in all samples (r=0.9, p=2.5x10 -133 ). Cumulative lifetime stress but not childhood or current stress predicted age acceleration in GTP (p=0.012) and MPI-P (p=0.021). Moreover, epigenetic age acceleration predicted depression (GTP: p=0.002; Conte: p=0.014) and panic disorder (p=0.007). In secondary analyses, we examined the effect of lifetime stress on individual CpGs of the DNA methylation age predictor. After correcting for multiple comparisons, we identified in both GTP and MPI-P a stress-regulated CpG near MCAM , a gene implicated in aging-related diseases, including cardiovascular disease and cancers. Conclusions Cumulative lifetime stress, but not childhood or current stress, and psychiatric phenotypes are associated with accelerated epigenetic aging. Our findings may explain the accelerated cellular aging and increased disease risk associated with chronic stress and psychiatric disorders.

Background:In light of the ongoing debate whether agoraphobia [AG] should be viewed as a severe phobic disorder similar to specific phobia [SPE] or as a complication of panic disorder [PD] we aim to study the vulnerability structure of PD, AG and SPE.Methods:3021 14-24 year-olds from the general population were followed-up over 10 years. DSM-IV syndromes were assessed via computerized M-CIDI interview and vulnerability factors via questionnaires. Associations were assessed with odds ratios from logistic regression. Latent class analysis (LCA) regressed on vulnerability factors was used to derive classes that underlie panic and phobic syndromes and to assess their associations with vulnerability factors.Results:1.Vulnerability patterns were largely similar between PD, AG and SPE.2.The LCA resulted in a best fitting model with 4 classes: a healthy class, a class with moderate frequency of phobias without PD, a class characterized by PD and AG and moderate frequency of SPE (PDAG class) and one class characterized by high frequency of AG and SPE situational type and lower frequency of PD (AGSIT class).3.All classes showed different associations with multiple vulnerability measures. Subjects in the PDAG class reported less SPE in parents (OR=0.2; 95% CI=0.0-0.6) and older onset-age of any psychopathology (OR=2.0; 95% CI=1.07-3.6) than the AGSIT class.Discussion:We found indications for separate latent classes underlying PD and phobias that were characterized by different vulnerability factors. We interprete the different classes as different vulnerability clusters and evidence of multiple pathways leading to panic and phobias.