PTSD has long been linked to accelerated biological aging and a higher risk of age-related diseases such as cardiovascular conditions and dementia. However, previous studies were limited by small samples (fewer than 2,200 people) and varying definitions of PTSD and epigenetic clocks. This work stands apart: it uses data from more than 45,000 veterans, which is 20 times larger than previous efforts. That allows for a reliable estimate of the link between PTSD and epigenetic age and a test of which PTSD measures and which epigenetic clocks are most sensitive.
The study is based on data from 45,091 veterans in the Million Veteran Program (MVP) who had both DNA methylation data and electronic medical records available. PTSD was defined in three ways: diagnosis from medical records (with the requirement that the first diagnosis was made at least a year before blood collection), symptom severity from the PCL-C questionnaire (for 22,835 participants), and PTSD duration as the proportion of life with the diagnosis. Epigenetic age was calculated using five algorithms: Horvath, Hannum, GrimAge, PhenoAge, and DunedinPACE. For each algorithm, residuals were computed after adjusting for chronological age. Regression analysis was then performed with adjustment for sex, blood cell composition, principal components of genetic ancestry, and technical factors. To account for multiple comparisons, the false discovery rate (FDR) correction was applied.
All three PTSD measures were positively associated with all five epigenetic clocks after correction for multiple comparisons. The strongest and most statistically significant associations were seen for PTSD symptom severity, especially with the GrimAge clock. For example, the difference between the maximum (85) and minimum (17) severity score corresponded to an acceleration of epigenetic age by 23.7 months on GrimAge. A PTSD diagnosis was linked to an acceleration of 4.68 months on the Hannum clock, and the maximum PTSD duration (42% of life) to 10.7 months on Hannum. After adding body mass index (BMI) and diabetes to the models, the associations weakened but remained significant for Hannum and GrimAge. Mediation analysis showed that BMI partially mediates the link between PTSD and epigenetic age (13–20% of the effect), especially in younger veterans. The interaction with age was significant: in younger veterans (21–44 years), the link between PTSD and PhenoAge and GrimAge was stronger than in older ones. Analysis by ethnic group revealed that associations were more pronounced in veterans of European ancestry than in African Americans, where only the links between PTSD severity and GrimAge and DunedinPACE remained significant.
This work is the first to show in a huge sample that PTSD is linked to accelerated epigenetic aging regardless of how it is measured. This supports the hypothesis that psychological stress can speed up biological aging. Unlike previous studies, it used five epigenetic clocks and three definitions of PTSD at once, allowing their sensitivity to be compared. The most robust were the GrimAge and Hannum clocks, and the strongest predictor was symptom severity, not simply the presence of a diagnosis. This underscores the importance of using quantitative scales in clinical and epidemiological research.
Future studies could explore which biological mechanisms link PTSD to accelerated aging, such as inflammation, oxidative stress, or changes in the immune system. It is also important to test whether successful treatment of PTSD slows the pace of epigenetic aging. A promising avenue is studying individual components of epigenetic clocks that reflect aging of specific organ systems, to understand which organs are most vulnerable in PTSD.
The results could influence clinical practice in mental health, emphasizing the need for early detection and treatment of PTSD to prevent premature aging. They are also important for health systems serving veterans, as they point to potentially higher healthcare needs in patients with PTSD.
The next step could be a longitudinal study tracking changes in epigenetic age in veterans with PTSD over time and in response to therapy. It is also necessary to replicate the analysis in more diverse populations, including civilians and women, to test the generalizability of the results.
The study addresses the fundamental problem of the link between mental health and biological aging, which remains unresolved in psychiatry and gerontology. It shows that PTSD is a risk factor for accelerated aging, but the causal mechanisms require further investigation. Another open question is why the associations are weaker in African Americans, which may point to differences in biology or in the accuracy of epigenetic clocks for different populations.
The GrimAge epigenetic clock was trained to predict time to death, so its link to PTSD may reflect a real risk of premature mortality in veterans with this disorder.
In Neal Stephenson's novel Seveneves, psychological stress after a catastrophe leads to accelerated aging in survivors, echoing the idea that trauma affects biological age.
Not published on arXiv. Here are its page, the PDF and the formalization: you can check it yourself, not only read our account.