Posttraumatic Stress Disorder and Epigenetic Aging: A Longitudinal Meta-Analysis

PTSD and Epigenetic Aging: A Longitudinal Meta-analysis (Pre-publication corresponding)

PTSD and Epigenetic Aging: A Longitudinal Meta-analysis Pre-publication Corresponding

 

Authors: Xiang Zhao1, Seyma Katrinli2, Beth M McCormick3,4, Mark W Miller3,4, Nicole R Nugent5,6,7, Agaz H Wani8, Anthony S Zannas9,10,11,12, Allison E Aiello13, Dewleen G Baker14, Marco P Boks15, Chia-Yen Chen16, Catherine B Fortier17,18,19, Joel Gelernter20,21, Elbert Geuze22,23, Karestan C Koenen24,25,26, Sarah D Linnstaedt12,27, Jurjen J Luykx28,29,30, Adam X Maihofer14,31,32, Samuel A McLean11,12, William P Milberg17,18,19, Andrew Ratanatharathorn25,33, Kerry J Ressler17,34,35, Victoria B Risbrough14,31,32, Bart P F Rutten24,36, Jordan W Smoller24,26,37, Murray B Stein14,38,39, Robert J Ursano40, Eric Vermetten41,42, Christiaan H Vinkers28,43,44, Erin B Ware45, Derek E Wildman8, Ying Zhao12,27, PGC-PTSD Epigenetics Workgroup, Mark W Logue1,3,4,46, Caroline M Nievergelt14,31,32, Alicia K Smith2,34,47, Monica Uddin8, Erika J Wolf3,4

1Boston University School of Public Health, Department of Biostatistics, Boston, MA, US, 2Emory University, Department of Gynecology and Obstetrics, Atlanta, GA, US, 3Boston University Chobanian & Avedisian School of Medicine, Department of Psychiatry, Boston, MA, US, 4VA Boston Healthcare System, National Center for PTSD, Boston, MA, US, 5Alpert Brown Medical School, Department of Emergency Medicine, Providence, RI, US, 6Alpert Brown Medical School, Department of Pediatrics, Providence, RI, US, 7Alpert Brown Medical School, Department of Psychiatry and Human Behavior, Providence, RI, US, 8University of South Florida College of Public Health, Genomics Program, Tampa, FL, US, 9University of North Carolina at Chapel Hill, Carolina Stress Initiative, Chapel Hill, NC, US, 10University of North Carolina at Chapel Hill, Department of Genetics, Chapel Hill, NC, US, 11University of North Carolina at Chapel Hill, Department of Psychiatry, Chapel Hill, NC, US, 12University of North Carolina at Chapel Hill, Institute for Trauma Recovery, Chapel Hill, NC, US, 13Columbia University, Robert N Butler Columbia Aging Center, Department of Epidemiology, New York, NY, US, 14University of California San Diego, Department of Psychiatry, La Jolla, CA, US, 15Brain Center University Medical Center Utrecht, Department of Psychiatry, Utrecht, UT, NL, 16Biogen Inc., Translational Sciences, Cambridge, MA, US, 17Harvard Medical School, Department of Psychiatry, Boston, MA, US, 18VA Boston Healthcare System, Geriatric Research, Education and Clinical Center (GRECC), Boston, MA, US, 19VA Boston Healthcare System, Translational Research Center for Traumatic Brain Injury and Stress Disorders (TRACTS), Boston, MA, US, 20VA Connecticut Healthcare Center, Psychiatry Service, West Haven, CT, US, 21Yale University School of Medicine, Department of Genetics and Neuroscience, New Haven, CT, US, 22Netherlands Ministry of Defence, Brain Research and Innovation Centre, Utrecht, UT, NL, 23UMC Utrecht Brain Center Rudolf Magnus, Department of Psychiatry, Utrecht, UT, NL, 24Broad Institute of MIT and Harvard, Stanley Center for Psychiatric Research, Cambridge, MA, US, 25Harvard T.H. Chan School of Public Health, Department of Epidemiology, Boston, MA, US, 26Massachusetts General Hospital, Psychiatric and Neurodevelopmental Genetics Unit (PNGU), Boston, MA, US, 27University of North Carolina at Chapel Hill, Department of Anesthesiology, Chapel Hill, NC, US, 28Amsterdam University Medical Center, Amsterdam Neuroscience Research Institute, Mood, Anxiety, Psychosis, Stress & Sleep Program, Amsterdam, NH, NL, 29Amsterdam University Medical Center, Amsterdam Public Health Research Institute, Mental Health Program, Amsterdam, NH, NL, 30Amsterdam University Medical Center, Department of Psychiatry, Amsterdam, NH, NL, 31Veterans Affairs San Diego Healthcare System, Center of Excellence for Stress and Mental Health, San Diego, CA, US, 32Veterans Affairs San Diego Healthcare System, Research Service, San Diego, CA, US, 33Columbia University Mailman School of Public Health, Department of Epidemiology, New York, NY, US, 34Emory University, Department of Psychiatry and Behavioral Sciences, Atlanta, GA, US, 35McLean Hospital, Division of Depression and Anxiety, Belmont, MA, US, 36Maastricht Universitair Medisch Centrum, School for Mental Health and Neuroscience, Department of Psychiatry and Neuropsychology, Maastricht, Limburg, NL, 37Massachusetts General Hospital, Department of Psychiatry, Boston, MA, US, 38University of California San Diego, School of Public Health, La Jolla, CA, US, 39Veterans Affairs San Diego Healthcare System, Psychiatry Service, San Diego, CA, US, 40Uniformed Services University of Health Sciences, Center for the Study of Traumatic Stress, Department of Psychiatry, Bethesda, Maryland, US, 41Leiden University Medical Center, Department of Psychiatry, Leiden, ZH, NL, 42New York University School of Medicine, Department of Psychiatry, New York, NY, US, 43Amsterdam UMC location Vrije Universiteit Amsterdam, Department of Anatomy and Neurosciences, Amsterdam, NH, NL, 44Amsterdam UMC location Vrije Universiteit Amsterdam, Department of Psychiatry, Amsterdam, Holland, NL, 45University of Michigan, Survey Research Center, Ann Arbor, MI, US, 46Boston University Chobanian & Avedisian School of Medicine, Department of Biomedical Genetics, Boston, MA, US, 47Emory University, Department of Human Genetics, Atlanta, GA, US

 

Translated by: Specialist Psychologist Gizem Pozam

 

 

Posttraumatic Stress Disorder and Epigenetic Aging: A Longitudinal Meta-Analysis

1. Introduction

Epigenome-wide scores derived from DNA methylation (DNAm) have emerged as leading biomarkers of biological age and mortality (e.g., Horvath, 2013; Lu et al., 2019) and are referred to as “DNAm age.” DNAm age estimates may differ from chronological age, indicating a DNAm age that is more advanced than chronological age in some individuals. Advanced DNAm age has been associated with age-related diseases and adverse health outcomes, including metabolic syndrome, inflammation, neuropathology, and mortality (Chen et al., 2016; Christiansen et al., 2016; Gassen, Chrousos, Binder, & Zannas, 2017; Hillary et al., 2021; Levine et al., 2015; Marioni et al., 2016, 2015; Meier, Mitchell, Karadimas, & Faul, 2023; Miller & Sadeh, 2014; Reed, Carroll, Marsland, & Manuck, 2022), highlighting the practical importance of DNAm age as a biomarker of biological aging and its potential to identify individuals at risk for the early onset of age-related diseases.

The two DNAm age algorithms that have received the most attention to date are Horvath’s (Horvath, 2013) marker of biological age and Lu et al.’s (Lu et al., 2019) index of time to death, termed “GrimAge.” Both algorithms were trained using machine learning approaches to select a set of CpG sites from the epigenome. The regression coefficients for the selected CpG sites are used as weights to calculate the corresponding DNAm age estimates in new datasets. Age-adjusted DNAm age is typically defined as the residuals obtained by regressing DNAm age on chronological age; negative values indicate decelerated epigenetic age, whereas positive values indicate advanced epigenetic age relative to chronological age.

Many studies have used cross-sectional methods to examine psychiatric diagnoses and symptoms associated with advanced DNAm age in blood samples; examples include studies of depression (Liu et al., 2022) and alcohol use disorders (Rosen et al., 2018). Posttraumatic stress disorder (PTSD) has received particular attention in this context, with numerous individual studies (Jovanovic et al., 2017; Katrinli et al., 2020; Roberts et al., 2017; Wolf et al., 2016; Wolf, Logue et al., 2018) and a meta-analysis (Wolf, Maniates et al., 2018) finding cross-sectional associations between PTSD and advanced DNAm age. Specifically, in our previous meta-analysis of cross-sectional data from 9 cohorts contributing to the Psychiatric Genomics Consortium (5 of which provided longitudinal data for these analyses), we found associations between increased epigenetic age and lifetime PTSD severity, but not PTSD diagnosis, and childhood trauma exposure (when measured using a consistent trauma measure across studies) (Wolf, Maniates et al., 2018). This is consistent with the hypothesis that the cumulative burden of lifetime PTSD symptom severity, particularly its chronic effects such as physiological reactions, poor sleep, anger, and arousal, is more relevant to the risk of accelerated aging. Some studies have also linked PTSD-related advanced DNAm age to biomarkers of inflammation, metabolic pathology, and neuropathology (Morrison et al., 2019; Wolf et al., 2023), suggesting that individuals with PTSD may be at greater risk for the early onset of these conditions. However, cross-sectional designs cannot address questions about the directionality of associations between PTSD and advanced DNAm age or the temporal stability of advanced epigenetic age estimates over time, nor can they track how PTSD-related epigenetic age changes accumulate and contribute to long-term health outcomes. Examining the longitudinal relationship between epigenetic aging and PTSD is therefore critical to understanding how trauma-related disorders contribute to premature aging and long-term health decline over time, and is also necessary for developing interventions targeted to the pathophysiology contributing to the early onset of these health conditions.

To date, only a few studies have examined the relationship between PTSD and advanced DNAm age longitudinally, and approaches to modeling changes in DNAm age estimates over time have varied across studies, as have their findings. In a longitudinal cohort of 96 male Dutch soldiers assessed before deployment to a combat zone and 6 months after deployment, intervening combat trauma was associated with an increase in raw (i.e., not accounting for chronological age) Horvath DNAm age estimates over time (Boks et al., 2015). However, increased postdeployment PTSD symptoms were negatively associated with changes in raw DNAm age between time points (Boks et al., 2015), suggesting potentially different effects of trauma exposure and PTSD.

Using an alternative analytical design, a study of 40 paramedic students assessed before and after work-related trauma exposure found that baseline Horvath DNAm age residuals were positively associated with PTSD severity at follow-up (approximately 1–2 months after trauma exposure) (Mehta et al., 2022). However, all participants reported a history of trauma at baseline, raising the possibility that baseline DNAm age may have been advanced in association with this history and/or preexisting psychiatric symptoms. Consistent with this possibility, the same study found that baseline PTSD severity was associated with larger Horvath and GrimAge DNAm age residuals at follow-up (Mehta et al., 2022). Most critically, the analyses did not account for baseline DNAm age acceleration, obscuring the extent to which DNAm age changed from baseline to follow-up. Another study (Yang et al., 2021) modeled changes in DNAm age estimates using correlated change scores for GrimAge residuals and PTSD symptom severity in trauma-exposed male military veterans. Among those with PTSD at baseline, changes in PTSD severity over three years were positively associated with changes in GrimAge residuals, but this was based on only 26 participants. Small sample size (and the associated limitations in statistical power and representativeness) is a concern applicable to all longitudinal studies of PTSD and epigenetic aging conducted to date.

Other studies have attempted to address the challenges of modeling relationships between time between assessments, DNAm age, and change by evaluating the rate of change in raw DNAm age estimates. One study (Sumner et al., 2023) followed 171 children and adolescents over 2 years and found that the rate of change in DNAm age was greater (more positive) among those who experienced more adverse stressful life events. Two other studies examining the rate of change among military personnel with chronic PTSD found that baseline PTSD symptom severity and diagnosis predicted accelerated Horvath DNAm age over approximately 2 (Wolf et al., 2019) and 5.5 (Hawn et al., 2023) years, respectively. These studies, with sample sizes ranging from 26 to 179, highlight the challenges of modeling changes in DNAm age over time and raise the possibility that different analytical strategies may be needed to best address questions about how new-onset PTSD diagnoses and chronic symptoms relate to changes in DNAm age over time.

 

1.1 Aims and Hypotheses
 This study aimed to examine PTSD-related health changes in terms of changes in epigenetic aging over time. Using a meta-analysis of longitudinal data from seven cohorts contributing to the Psychiatric Genomics Consortium (PGC), we examined the relationship between PTSD and subsequent changes in epigenetic aging (Ratanatharathorn et al., 2017). The diverse methodological structures of these datasets allowed us to ask multiple questions about the relationship between PTSD and epigenetic age. Specifically, our first aim was to assess whether the development of new-onset PTSD between two time points altered the association between Time 1 (T1) and Time 2 (T2) DNAm age residuals. We expected the association between DNAm age residuals to be more positive among those with new-onset PTSD at follow-up. Our second aim was to examine whether the association between DNAm age residuals at the two time points varied as a function of change in PTSD symptom severity (T2 – T1). We also hypothesized that the association between DNAm age residuals at the two time points would become stronger as the increase in PTSD symptom severity became greater. Both hypotheses were tested by modeling the associations of T2 DNAm age residuals with the interaction terms between T1 DNAm age residuals and new-onset PTSD, and between T1 DNAm age residuals and change in PTSD symptom severity; these associations were evaluated while adjusting for main effects and other covariates. These aims were examined in a total of 1,367 individuals, each assessed twice for DNAm. The cohorts included civilian and military samples, with research designs focused on pre/post military deployment or pre/post trauma exposure, as well as studies focused on chronic PTSD symptoms. The former design allowed us to examine new-onset PTSD diagnoses over time, whereas the latter allowed us to examine the association between changes in PTSD symptom severity and changes in epigenetic age residuals. Due to the structure of the contributing datasets, we could not examine changes in the rate of epigenetic aging over time (e.g., Wolf et al., 2019), as these datasets largely included cohorts defined by pre/post trauma exposure.

 

The analyses focused on Horvath age (Horvath, 2013) and GrimAge (Lu et al., 2019), considered the most widely used and robust measures of biological age and time to death, respectively (sometimes referred to as first- and second-generation clocks). These are also the only DNAm age indices previously associated with PTSD and have been used in small longitudinal studies (Boks et al., 2015; Hawn et al., 2023; Mehta et al., 2022; Sumner et al., 2023; Wolf et al., 2023; Yang et al., 2021). We also aimed to reduce the burden of multiple-testing corrections by limiting the number of tests, and therefore chose to restrict the analyses to the strongest epigenetic age measures with prior evidence of longitudinal associations with PTSD.

 

 

1. Methods and Materials
   2.1 Participating Studies
   Seven participating cohorts were included in the meta-analysis. The mean time between assessments in each study ranged from 5.7 months to 5.6 years. The cohorts were: 1) the Longitudinal National Center for PTSD (NCPTSD) Cohort of Trauma-Exposed Veterans (Wolf et al., 2023), a study involving trauma-exposed veterans, many with chronic PTSD, assessed twice an average of 5.6 years apart; 2) the Translational Research Center for Traumatic Brain Injury (TBI) and Stress Disorders (TRACTS) Cohort (McGlinchey, Milberg, Fonda, & Fortier, 2017), a study of post-9/11 veterans assessed twice an average of 1.9 years apart; 3) the Army Study to Assess Risk and Resilience in Servicemembers (Army STARRS) (Ursano et al., 2014), a study of servicemembers assessed before and after deployment to Afghanistan over an average of 9.6 months; 4) the Marine Resiliency Study (MRS) Cohort (Baker et al., 2012; Nievergelt et al., 2015), a study of male U.S. Marines assessed before and after deployment to Iraq or Afghanistan over an average of 12.4 months; 5) the Prospective Research in Stress-Related Military Operations (PRISMO) Cohort (Eekhout, Reijnen, Vermetten, & Geuze, 2016; Reijnen, Rademaker, Vermetten, & Geuze, 2015), a study of Dutch soldiers assessed before and after deployment to Afghanistan over an average of 14.4 months; 6) the Detroit Neighborhood Health Study (DNHS) Cohort (Goldmann et al., 2011; Uddin et al., 2010), a study of trauma-exposed Detroit residents (some with chronic psychiatric symptoms) assessed twice an average of 16.3 months apart; 7) the Advancing Understanding of Recovery After Trauma (AURORA) Study (McLean et al., 2020), a study of individuals assessed in the emergency department following trauma exposure and reassessed an average of 5.7 months after emergency department treatment. Table 1 lists the demographic and clinical characteristics of each cohort. Each study site obtained local IRB approval, and all participants provided written informed consent. The VA Boston Healthcare System IRB approved the meta-analyses of the summary data. Individuals seeking access to the data should contact the principal investigator of each cohort to determine data availability.

 

2.2 DNA and DNAm Procedures
 DNA was extracted from the buffy coat of peripheral blood samples. DNAm was measured at two time points using the Illumina Infinium EPIC BeadChip. DNAm data were processed following a quality control (QC) pipeline developed by the PGC PTSD Epigenetics Workgroup (Ratanatharathorn et al., 2017; and for the updated version, https://github.com/PGC-PTSD-EWAS/EPIC_QC). Details of the QC procedures are provided in the Supplementary Materials. DNAm is cell-type specific, and DNAm levels can therefore vary depending on the composition of the cell types from which the DNA was extracted. It is thus important to include proportional estimates of white blood cell types as covariates in the analyses (a standard practice in DNA methylation analyses). The proportions of six cell types (B cells, CD4+T cells, CD8+T cells, natural killer cells, monocytes, and neutrophils) were estimated directly from DNAm data using a reference library of CpG sites aligned with sorted cells in the Bioconductor package EpiDISH (Teschendorff, Breeze, Zheng, & Beck, 2017). Of these, B cells, CD4+T cells, CD8+T cells, natural killer cells, and monocytes were included as covariates in the analysis (neutrophils were excluded because, given the proportional nature of the estimates, they were collinear with all the other cell types included in the model). Blood-based methylation is strongly influenced by smoking habits, so a DNAm smoking score was calculated based on 39 smoking-associated CpGs (Li et al., 2018; Supplementary Material). Because the focus of our study was epigenetic age acceleration at T2, the T2 DNAm smoking score was included as an additional covariate in a sensitivity analysis to determine whether the effects of smoking, which can co-occur with PTSD, influenced our findings.

 

Genotyping was performed on various arrays. Genotype data cleaning was completed at each site according to procedures previously described in the original publications for each study (Table 1). Ancestry was determined from genotype data (when available) using the pipeline developed by PGC PTSD (Nievergelt et al., 2019). Ancestry-based principal components (PCs) were calculated from 100,000 randomly selected common single-nucleotide polymorphisms (SNPs) with a minor allele frequency (MAF) greater than 0.05%. When genotype data were unavailable (e.g., DNHS), methylation probes within 1 base pair of SNPs were used to generate a set of DNAm-based ancestry PCs to serve as proxies for genotype-based PCs in determining ancestry (Barfield et al., 2014). As noted in our previous meta-analysis, DNAm-derived PCs are significantly associated with genotype-based PCs (Ratanatharathorn et al., 2017; Wolf, Maniates, et al., 2018). Either the first three genotype-based PCs (1–3) or, if genotype-based PCs were unavailable, DNAm-based PCs (2–4) were used as covariates in the analyses to account for ancestry.

 

2.3 Measures
 2.3.1 PTSD Measures
 Current PTSD diagnosis and symptom severity were assessed using the various measures listed in Table 1. Following a previous PGC publication (Sumner et al., 2021), we harmonized PTSD symptom severity across these different measures by scaling the raw current PTSD severity score to a range of 0 to 1, representing the score as a percentage of the maximum possible score on each measure (i.e., 1 indicates that the individual experienced all symptoms at the most severe level, whereas 0 indicates that the individual had no symptoms). In this study, we examined both new-onset PTSD diagnosis (e.g., the difference between those without PTSD at T1 who were PTSD-positive at T2 and those without PTSD at both T1 and T2) and change in PTSD symptom severity (T2 – T1) using the harmonized PTSD score. Analyses of new-onset PTSD diagnosis included only individuals who were PTSD-negative at T1; this reduced the sample size in some studies that included participants with chronic PTSD (Table 1). Analyses of change in PTSD severity included all participants across cohorts with PTSD symptom severity data, regardless of baseline PTSD diagnostic status.

 

2.3.2 DNAm Age Indices
 Horvath DNAm age and GrimAge estimates were calculated by uploading DNAm data to Dr. Horvath’s website (https://dnamage.genetics.ucla.edu/) or, when local regulations did not permit uploading methylation data, by running scripts provided by Dr. Horvath and Dr. Lu (Horvath, 2013; Lu et al., 2019). Because Horvath’s algorithm includes its own normalization and imputation step, raw DNAm values were used as input for the Horvath age calculation. Values from 353 probes were used to generate Horvath age (Horvath, 2013). However, 17 of the 353 CpGs (4.8 percent) are missing from the EPIC chip (Dhingra et al., 2019). Some additional missing probes were identified in each cohort and are summarized in the Supplementary Materials. GrimAge estimates were generated using normalized and imputed DNAm data. For the GrimAge calculation, 30,084 probes were used as input (Lu et al., 2019). DNAm age residuals were calculated by saving the unstandardized residuals from a linear model regressing DNAm age on chronological age.

Both Horvath age and GrimAge residuals were calculated by saving the unstandardized residuals from a linear model regressing raw DNAm age on chronological age. This procedure was performed separately at the two time points for each cohort, so that the age residuals in each cohort had a mean of 0 at each time point. An R script developed by the first author was tested with the authors and then sent to the data analyst at each participating cohort site, ensuring that the same computational and analytical approaches were applied in each cohort. Summary statistics from each cohort were meta-analyzed to combine results across studies.

 

First, we examined the Pearson correlations of Horvath age and GrimAge with chronological age at each time point. We then evaluated correlations over time for Horvath age, GrimAge, Horvath age residuals, and GrimAge residuals. Correlations over time for cell types and correlations between the two DNAm age residual measures and cell types were also evaluated. Correlation coefficients collected from each group were meta-analyzed using the metacor package (Laliberté E, 2022) in R.

We performed two main regression analyses. The first examined how a new-onset PTSD diagnosis at T2 only (compared with being PTSD-negative at both T1 and T2) and its interaction with T1 DNAm age residuals predicted T2 DNAm age residuals, adjusting for T1 DNAm age residuals and covariates. The second analysis replaced new-onset PTSD diagnosis with change in PTSD symptom severity over time. In both analyses, we first fitted a regression model without the interaction term to capture the main effects of all predictors (all reported main-effect coefficients were taken from this initial model) and then added the interaction term to the model. The interaction term examined how the association between T1 and T2 DNAm age residuals differed with changes in PTSD (diagnosis or severity). A positive interaction term would indicate that T2 DNAm age residuals became more extreme than would be predicted from T1 DNAm age residuals alone, reflecting the moderating effect of changes in PTSD status. This more extreme change could occur at either end of the DNAm age residual distribution. Specifically, the predictors in the main linear regression models predicting T2 age residuals were: DNAm age residuals at T1, new-onset PTSD diagnosis at T2 (or, separately, change in PTSD symptom severity), the interaction between T1 DNAm age residuals and new-onset PTSD diagnosis (or change in PTSD severity), and the following covariates: sex (excluded if there was no variation in sex within a given sample), three ancestry PCs, and five estimated cell-type proportions at T2. Significant associations were further examined in sensitivity analyses that included DNAm-based smoking scores at T2 as an additional covariate. Follow-up models evaluated the association between childhood trauma (before T1) and changes in epigenetic age over time (Supplementary Materials).

Each unstandardized parameter estimate from the regression models (except ancestry PCs) was meta-analyzed using an inverse-variance-weighted random-effects model with the metafor package in R (Viechtbauer, 2010). Results for each term were adjusted for multiple testing across the two age algorithms using false discovery rate (FDR) correction (Benjamini & Hochberg, 1995).

 

1. Results
   3.1 Associations Between Chronological Age, DNAm Age, and DNAm Age Residuals
   Chronological age was strongly correlated with Horvath age and GrimAge at both time points (meta rs = 0.86-0.89, meta ps < 0.001; Table S1). At the individual cohort level, lower correlations were observed in ArmySTARRS and MRS (Table S1), which may have resulted from the smaller variance in chronological age in these cohorts (Table 1). Raw Horvath age and GrimAge were strongly associated with each other at both time points (meta rs = 0.80 and 0.78 at T1 and T2, respectively; Table S1). However, Horvath age and GrimAge residuals were weakly correlated with each other at both time points (both meta rs = 0.09; Table S1). The meta-analysis revealed strong correlations between T1 and T2 estimates of raw Horvath age and GrimAge (meta rs = 0.91 and 0.96, meta ps <0.001; Table 2). Residuals were also consistent over time: T1 and T2 Horvath DNAm age residuals had a meta-r = 0.65, and GrimAge residuals had a meta-r = 0.88 (Table 2).

3.2 Cell-Type Proportions and Their Associations with DNAm Age Residuals
 Estimated cell-type proportions were strongly associated with themselves over time. The estimated proportion of B cells showed the strongest correlation over time (meta r = 0.75, meta p < 0.001), followed by CD8+T cells (meta r = 0.71, meta p < 0.001), natural killer cells (meta r = 0.61, meta p < 0.001), CD4+T cells (meta r = 0.61, meta p < 0.001), and monocytes (meta r = 0.55, meta p < 0.001; Table 2). In addition, cell-type estimates were weakly correlated with both Horvath age residuals and GrimAge residuals at both time points, with meta-correlations ranging from -0.08 to 0.11 for Horvath age residuals and from -0.21 to 0.05 for GrimAge residuals (Table S2).

 

3.3 New-Onset PTSD Diagnosis and Change in DNAm Age Residuals Over Time
 We examined how T1 DNAm age residuals, new-onset PTSD diagnosis (between T1 and T2), and their interaction predicted T2 DNAm age residuals (n = 745). The meta-analysis revealed a significant effect of T1 age residuals on T2 age residuals for both the Horvath algorithm (meta b = 0.63, meta SE = 0.05, meta p = 3.63 × 10^-37) and the GrimAge algorithm (meta b = 0.85, meta SE = 0.02, meta p < 2.23 × 10^-308; Table 3). New-onset PTSD diagnosis was not associated with T2 Horvath age residuals (meta b = -0.26, meta SE = 0.24, meta p = 0.27) or T2 GrimAge residuals (meta b = 0.22, meta SE = 0.13, meta p = 0.10; Table 3). The meta-analysis also showed that the interaction term between T1 Horvath age residuals and new-onset PTSD diagnosis was significantly associated with T2 Horvath age residuals after correction for multiple testing across both algorithms (meta b = 0.16, meta SE = 0.07, meta p = 0.02, p-adj = 0.03, I2 = 0; Table 3 and Figure 1). This indicates that the positive association between Horvath age residuals at the two time points was stronger among those with a new-onset PTSD diagnosis. With additional adjustment for the smoking score, the interaction effect remained significant (meta b = 0.17, meta SE = 0.07, meta p = 0.02), and the smoking score was not associated with T2 Horvath age residuals (meta p = 0.11). This association was not significant for the GrimAge algorithm (meta p = 0.99; Table 3). The association between the interaction term and T2 Horvath DNAm age residuals was not driven by differences in baseline Horvath age residuals between new-onset PTSD cases and those who did not develop PTSD (meta p = 0.16; Table S3). This suggests that those who developed PTSD at T2 did not simply have higher DNAm age residuals at baseline. Full results for each individual cohort can be found in Tables S4 and S5. Follow-up analyses revealed that childhood trauma was not associated with T2 DNAm age residuals (Supplementary Materials).

 

3.4 Change in PTSD Symptom Severity and Change in Horvath DNAm Age Residuals Over Time
 Next, as a follow-up to the new-onset PTSD diagnosis analysis, we examined whether the association between Horvath DNAm age residuals over time was moderated by change in PTSD symptom severity (n = 1191). We included the main and interaction effects of baseline Horvath age residuals and change in PTSD symptom severity over time as predictors of T2 Horvath age residuals (along with covariates). The association between T1 and T2 Horvath age residuals was significant (meta b = 0.63, meta SE = 0.05, meta p = 6.42 × 10^-31), but the PTSD symptom severity change score was not associated with T2 Horvath age residuals (meta b = 0.47, meta SE = 0.63, meta p = 0.46). The interaction between T1 Horvath age residuals and change in PTSD symptom severity was significantly associated with T2 Horvath DNAm age residuals (meta b = 0.24, meta SE = 0.12, meta p = 0.05, I2 = 0; Table 4 and Figure 2). The association between DNAm age residuals over time became stronger (a steeper slope) among those who experienced the greatest increase in PTSD symptom severity from T1 to T2. The interaction effect remained significant with additional adjustment for the smoking score (meta b = 0.25, meta SE = 0.12, meta p = 0.04), but the smoking score was not associated with T2 Horvath age residuals (meta p = 0.28). The association between the interaction term and T2 Horvath age residuals was not explained by baseline differences in T1 Horvath age residuals as a function of change in PTSD symptom severity (meta p = 0.41; Table S3). Full results for each individual study are listed in Table S6. Although we did not find a significant interaction between GrimAge residuals and new-onset PTSD diagnosis, for completeness we performed the PTSD severity change analysis for GrimAge residuals and report the (null) result in the Supplementary Materials.

 

1. Discussion
   The main objectives of this study were, first, to test whether individuals who developed new-onset PTSD between two assessments showed greater epigenetic aging than expected based on their epigenetic aging at a time when they did not meet criteria for PTSD. The second objective was to examine whether changes in PTSD symptom severity moderated the strength of the association in epigenetic aging between two time points among individuals with and without PTSD. We expected individuals with new-onset PTSD and those whose PTSD symptom severity increased over time to have higher DNAm age residuals at follow-up relative to their baseline DNAm age residuals. We evaluated this question through a meta-analysis of 7 cohorts. For each one-year increment in advanced Horvath DNAm age, we found that new-onset PTSD cases showed 0.16 years more epigenetic aging according to the Horvath algorithm at T2 compared with longitudinal controls who did not develop PTSD. In other words, those with new-onset PTSD aged an additional 1.9 months faster over the interval, which averaged approximately 18.9 months. Furthermore, an individual experiencing the maximum possible change in symptom severity (from having no symptoms to experiencing all PTSD symptoms at the most severe level) would experience 0.24 years more Horvath age acceleration at T2, meaning an additional 2.9 months of faster aging compared with an individual whose symptom severity score did not change; the interval for the symptom severity analyses averaged approximately 21.3 months.

To better interpret our findings, we also estimated the pooled standard deviation of changes in harmonized PTSD symptom severity scores across cohorts (SDpooled = 0.23). This allows the interaction term to be interpreted as follows: for each one-year increase in Horvath age residuals, an individual experiencing a one-pooled-SD (23%) increase in PTSD symptom severity at T2 would experience an additional 0.68 months of Horvath age acceleration over an average interval of 21.3 months compared with an individual whose symptom severity did not change. Although the observed effect sizes were relatively small, the cumulative effect of PTSD on epigenetic aging may become substantial over the lifespan, given that this meta-analysis used a large sample that included young veterans in their early 30s (e.g., the ArmySTARRS, MRS, and TRACTS cohorts).

These results are consistent with our hypotheses that age acceleration increases among individuals with new-onset PTSD and increasing symptom severity. Previous studies have not examined interactions between new-onset PTSD (or change in symptom severity) and baseline age acceleration, but the structure of these data, with numerous cohorts defined by pre- and post-trauma exposure, made it critical to ask how new-onset diagnoses would affect changes in measures of advanced DNAm age. In these cohorts, it would not have been possible to use baseline PTSD diagnoses to predict changes in DNAm age residuals (or the rate of DNAm age change per intervening year), because most individuals were PTSD-negative before trauma exposure.

Previous studies have found that PTSD symptom change is associated with changes in Horvath DNAm age as well as in DNAm loci that contribute to the Horvath DNAm age calculation, and these changes may be partly related to the Horvath DNAm age changes observed in this study. Specifically, Katrinli et al. (Katrinli et al., 2022) modeled the association between postdeployment DNAm and change in PTSD symptom severity (PTSS), conditional on predeployment DNAm, and found 15 differentially methylated regions (DMRs) associated with PTSS change. The guanine nucleotide-binding protein, alpha-stimulating activity polypeptide (GNAS) complex locus was one of the 15 DMRs associated with PTSS change, and this locus contains cg14597908, a CpG site that also contributes to the calculation of Horvath age (Horvath, 2013). Differential methylation at GNAS has been associated with maternal stress (Vangeel et al., 2015) and anxiety (Alisch et al., 2014, 2017), suggesting that it may be sensitive to traumatic stress. In a porcine model, GNAS has also been associated with cellular aging (Jeon et al., 2012). Thus, PTSD-related changes in GNAS over time may influence the advanced epigenetic age changes observed in this study. Katrinli et al. also reported that PTSS was associated with DNAm changes in genes related to immune processes and oxidative stress (Katrinli et al., 2022). Alterations in these biological pathways may affect the aging process and contribute to the effects observed in this study, given their known associations with accelerated aging (Cevenini, Monti, & Franceschi, 2013).

Our main interaction effects indicated that PTSD moderates the trajectory of age acceleration over time, raising the question of whether PTSD treatment could slow cellular aging and reduce the risk of early-onset age-related diseases. However, there are several important challenges in addressing this question. First, the strongest predictor of DNAm age acceleration at follow-up was baseline DNAm age acceleration, so after accounting for this effect, only a small amount of variance may remain to be predicted by treatment. This means that the effects of PTSD treatment on cellular aging are likely to be small and that large sample sizes (probably larger than those in most intervention trials) would be needed to observe them. Second, DNAm age estimates may not be sensitive enough to accurately reflect changes in cellular age within a typical 12- or 18-week trial. Studies would need sufficiently long follow-up periods to observe reliable changes in the DNA methylome. Nevertheless, pharmacological studies offer some promising findings regarding treatment effects on epigenetic aging. For example, a study of 30 individuals with bipolar disorder and 30 healthy controls showed that combination treatment with mood stabilizers (lithium carbonate, sodium valproate, and carbamazepine) was associated with reduced Horvath age acceleration compared with monotherapy or no medication (Okazaki et al., 2020). In human neuroblastoma cells, lithium, valproate, and carbamazepine induced hypermethylation at 377, 70, and 66 CpG sites, respectively, and hypomethylation at 145, 37, and 14 CpG sites, respectively (Asai et al., 2013). Similar studies in PTSD samples are needed to explore PTSD treatments and their potential to alter the aging methylome.

Our study also found high correlations of DNAm age estimates and age residuals with themselves over time, demonstrating the robustness of the two age calculators. Both GrimAge and GrimAge residuals showed stronger correlations with themselves over time than did the Horvath algorithm. This may be due to its unique two-stage calculation, the inclusion of age and sex in the model (Lu et al., 2019), or the greater stability afforded by the number of GrimAge probes compared with the smaller probe set included in the Horvath model. The stronger correlation between GrimAge residuals over time may be one reason we did not observe a significant interaction between T1 GrimAge residuals and PTSD. After adjusting for the T1 GrimAge main effect in the model, little variance may have remained in the outcome. In addition, the two models may capture different elements of the aging process reflected in methylation and may differ in their relevance to PTSD-related pathologies.

This study also provided an opportunity to test the stability of estimated white blood cell types over time. We found that cell-type proportion estimates were strongly correlated with themselves over time, with particularly high correlations for CD8+T cells and B cells. This highlights the robustness of the cell-type estimation algorithm and its value for examining aging-related changes in cell-type composition.

4.1 Study Limitations
 Our results should be considered in light of several limitations. We did not account for intervening time in any analysis because there was little or no variability in time between assessments in the pre- and postdeployment military cohorts. This makes it impossible to determine how epigenetic age measures correlate over shorter versus longer intervals, and how severe, chronic PTSD (i.e., more symptoms over a longer period) might alter accelerated aging. Data were available at only two time points, so we could not test whether change in epigenetic age was constant or nonlinear over time. The absence of PTSD cases at baseline in the pre/postdeployment samples made it impossible to explore the association of epigenetic aging with baseline diagnosis and severity.

We did not assess intervening trauma or life stress between the two time points across all cohorts, so we could not adjust for new trauma exposure/life stressors or substitute them for new-onset PTSD in our analyses. Consequently, we could not distinguish the effects of PTSD from those of trauma or life stress, a common challenge because these variables are often correlated. However, many previous studies have suggested that PTSD is more strongly associated with advanced epigenetic age than trauma exposure alone (which is nearly universal across populations) (Wolf et al., 2016, 2019; Wolf, Logue, et al., 2018). We believe that the ongoing chronic psychological and physiological stress associated with traumatic stress in PTSD (e.g., startle, arousal, anger, poor sleep, etc.) may have a more direct and sustained effect on the physiological processes associated with epigenetic aging.

Because information on participants’ medical and pharmacy records was unavailable, we could not determine their associations with epigenetic aging or account for them in the analyses. We also could not adjust for lifestyle factors such as body mass index and substance misuse because these data were not consistently available across cohorts. However, sensitivity analyses revealed that smoking did not affect our main association findings. The error associated with the Horvath age algorithm (a median absolute difference of 3.6 years between predicted and actual age in the original test data; Horvath, 2013) was greater than the average time between assessments in some studies (e.g., the mean interval of 5.7 months in the AURORA study). Analyses in cohorts with short intervals may be more susceptible to measurement error because the algorithm may not be sensitive enough to detect small changes in methylation over a short period. We chose not to analyze the recently developed DunedinPACE metric (Belsky et al., 2022) because it estimates the pace of aging using DNAm data from a single time point, whereas we wanted to calculate observed DNAm age at two time points and measure the change between them. Because of the limited sample size of civilian participants, we could not examine how military and civilian samples might influence our findings. The participants included in this study were diverse in terms of sex, race, ethnicity, military and civilian status, and geography, but they may still not fully represent any particular population, which may limit the generalizability of our findings. Finally, although our data are longitudinal, we cannot claim a causal relationship between PTSD and change in epigenetic age.

 

4.2 Conclusions
 This was the first meta-analysis and the largest study to date examining the association between PTSD and changes in DNAm age over time. Using meta-analytic data from seven cohorts encompassing both military and civilian samples, we found that new-onset PTSD diagnosis and increases in PTSD symptom severity were associated with greater age acceleration than expected based on baseline age acceleration. This adds to a growing body of evidence suggesting that stress-related disorders may accelerate cellular aging and potentially contribute to the association between traumatic stress and the early onset of age-related diseases, such as cardiovascular disease (Vidal et al., 2018) and dementia (Yaffe et al., 2010). It underscores the importance of understanding the integration of mental and physical health, even at the cellular level, and highlights the substantial personal costs associated with traumatic stress.

 

 

Note. NCPTSD = National Center for PTSD Study; TRACTS = Translational Research Center for Traumatic Brain Injury (TBI) and Stress Disorders Study; Army STARRS = Army Study to Assess Risk and Resilience in Servicemembers; MRS = Marine Resiliency Study; PRISMO = Prospective Research in Stress-Related Military Operations; DNHS = Detroit Neighborhood Health Study; AURORA = Advancing Understanding of Recovery After Trauma Study; IT = Intervening Time; PTSD = Posttraumatic Stress Disorder; DX = Diagnosis; T1 = Time 1; T2 = Time 2; ED = Emergency Department; EUR = European ancestry; AAM = African American ancestry; LAT = Latino ancestry; OTH = Other ancestries; PCL = PTSD Checklist for DSM-IV; PCL-5 = PTSD Checklist for DSM-5; PCL-C = PTSD Checklist–Civilian Version; DSM = Diagnostic and Statistical Manual of Mental Disorders; CIDI-SC = Composite International Diagnostic Interview screening scales; CAPS = Clinician-Administered PTSD Scale for DSM-IV; CAPS-5 = Clinician-Administered PTSD Scale for DSM-5; SRIP = Self-Rating Inventory for PTSD; △ = Change in harmonized PTSD symptom severity (T2-T1); SD = Standard deviation. *: Sample sizes for calculating cross-sectional and longitudinal correlations between DNAm age, DNAm age residuals, and DNAm-based cell-type proportions; ¨: Sample sizes for the new-onset PTSD diagnosis analysis and the symptom severity change analysis can be found in Figure 1 and Figure 2. Compared with the total N, the sample size for the new-onset PTSD diagnosis analysis was reduced due to the exclusion of baseline PTSD cases and missing values for PTSD diagnosis and covariates. The sample size for the symptom severity change analysis was reduced due to missing values for PTSD symptom severity and covariates; †: 6-item screening version of the PCL; §: Abbreviated (six-item) civilian version of the PCL-5; ‡: CAPS was used to determine PTSD diagnosis, whereas PCL-5 was used to assess PTSD severity. Superscripts indicate references for each cohort and PTSD measure: 1. Wolf ve ark. (2023); 2. McGlinchey ve ark. (2017); 3. Ursano ve ark. (2014); 4. Baker ve ark. (2012); Nievergelt ve ark. (2015); 5. Eekhout ve ark. (2016); Reijnen ve ark. (2015); 6. Goldmann ve ark. (2011); Uddin ve ark. (2010); 7. McLean ve ark. (2020); 8. Blake ve ark. (1995); 9. Weathers ve ark. (2018); 10. Weathers, Litz, Herman, Huska ve Keane (1993); 11. Kessler ve ark. (2013); 12. Hovens, Bramsen ve Van Der Ploeg (2002); 13. Blevins, Weathers, Davis, Witte ve Domino (2015).

 

Note. CD8+T = CD8+ T cell; CD4+T = CD4+ T cell; NK = natural killer cell; Mono = monocyte; CI = confidence interval; Army STARRS = Army Study to Assess Risk and Resilience in Servicemembers; DNHS = Detroit Neighborhood Health Study; NCPTSD = National Center for PTSD Study; MRS = Marine Resiliency Study; PRISMO = Prospective Research in Stress-Related Military Operations; TRACTS = Translational Research Center for Traumatic Brain Injury (TBI) and Stress Disorders Study; AURORA = Advancing Understanding of Recovery After Trauma Study. Correlations were calculated using Pearson correlation coefficients for the same variable between T1 and T2. The 95% confidence intervals were calculated for meta-analytic correlations. All meta-analytic correlations were significantly different from zero at the p-value threshold of 0.001, as determined by a t-test.

 

 

Note. The first three ancestry principal components (PCs) from each cohort were also included in the model. CD8+T = CD8+ T cell; CD4+T = CD4+ T cell; NK = natural killer cell; Mono = monocyte; DX = diagnosis; SE = standard error; NA = not applicable; p-adj = p-value adjusted using the false discovery rate (FDR) procedure to correct for multiple testing across the two age algorithms.

Note. Meta-analytic results from multiple linear regression models in individual cohorts. Meta-analytic main effects were derived from models containing only main (and covariate) effects in each cohort. Meta-analytic interaction effects were derived from models containing both main and interaction effects. The first 3 ancestry principal components were also included in the meta-analysis. CD8+T = CD8+ T cell; CD4+T = CD4+ T cell; NK = natural killer cell; Mono = monocyte; SE = standard error; Δ = change (T2 – T1).

 

References

Alisch, R. S., Chopra, P., Fox, A. S., Chen, K., White, A. T. J., Roseboom, P. H., … Kalin, N. H. (2014). Differentially methylated plasticity genes in the amygdala of young primates are linked to anxious temperament, an at risk phenotype for anxiety and depressive disorders. Journal of Neuroscience, 34(47). doi: 10.1523/JNEUROSCI.3338-14.2014

 

Alisch, R. S., Van Hulle, C., Chopra, P., Bhattacharyya, A., Zhang, S. C., Davidson, R. J., … Goldsmith, H. H. (2017). A multi-dimensional characterization of anxiety in monozygotic twin pairs reveals susceptibility loci in humans. Translational Psychiatry, 7(12). Doi: 10.1038/s41398-017-0047-9

 

Asai, T., Bundo, M., Sugawara, H., Sunaga, F., Ueda, J., Tanaka, G., … Iwamoto, K. (2013). Effect of mood stabilizers on DNA methylation in human neuroblastoma cells. International Journal of Neuropsychopharmacology, 16(10). doi:

10.1017/S1461145713000710

Baker, D. G., Nash, W. P., Litz, B. T., Geyer, M. A., Risbrough, V. B., Nievergelt, C. M., … Webb-Murphy, J. A. (2012). Predictors of Risk and Resilience for Posttraumatic Stress Disorder Among Ground Combat Marines: Methods of the Marine Resiliency Study. Preventing Chronic Disease, 9(5). doi: 10.5888/pcd9.110134

 

Barfield, R. T., Almli, L. M., Kilaru, V., Smith, A. K., Mercer, K. B., Duncan, R., … Conneely, K. N. (2014). Accounting for population stratification in DNA methylation studies. Genetic Epidemiology, 38(3). doi: 10.1002/gepi.21789

 

Belsky, D. W., Caspi, A., Corcoran, D. L., Sugden, K., Poulton, R., Arseneault, L., … Moffitt, T. E. (2022). DunedinPACE, A DNA methylation biomarker of the Pace of Aging. ELife, 11. doi: 10.7554/eLife.73420

 

Benjamini, Y., & Hochberg, Y. (1995). Controlling the False Discovery Rate: A Practical and Powerful Approach to Multiple Testing. Journal of the Royal Statistical Society: Series B (Methodological), 57(1). doi: 10.1111/j.2517-6161.1995.tb02031.x

 

Blake, D. D., Weathers, F. W., Nagy, L. M., Kaloupek, D. G., Gusman, F. D., Charney, D. S., & Keane, T. M. (1995). The development of a Clinician-Administered PTSD Scale. Journal of Traumatic Stress, 8(1). doi: 10.1007/BF02105408

 

Blevins, C. A., Weathers, F. W., Davis, M. T., Witte, T. K., & Domino, J. L. (2015). The Posttraumatic Stress Disorder Checklist for DSM-5 (PCL-5): Development and Initial Psychometric Evaluation. Journal of Traumatic Stress, 28(6). doi: 10.1002/jts.22059

 

Boks, M. P., Mierlo, H. C. van, Rutten, B. P. F., Radstake, T. R. D. J., De Witte, L., Geuze, E., … Vermetten, E. (2015). Longitudinal changes of telomere length and epigenetic age related to traumatic stress and post-traumatic stress disorder. Psychoneuroendocrinology, 51. doi: 10.1016/j.psyneuen.2014.07.011

 

Cevenini, E., Monti, D., & Franceschi, C. (2013). Inflamm-ageing. Current Opinion in Clinical Nutrition and Metabolic Care, Vol. 16. doi: 10.1097/MCO.0b013e32835ada13

 

Chen, B. H., Marioni, R. E., Colicino, E., Peters, M. J., Ward-Caviness, C. K., Tsai, P. C., … Horvath, S. (2016). DNA methylation-based measures of biological age: Meta-analysis predicting time to death. Aging, 8(9). doi: 10.18632/aging.101020

 

Christiansen, L., Lenart, A., Tan, Q., Vaupel, J. W., Aviv, A., Mcgue, M., & Christensen, K. (2016). DNA methylation age is associated with mortality in a longitudinal Danish twin study. Aging Cell, 15(1). doi: 10.1111/acel.12421

 

Dhingra, R., Kwee, L. C., Diaz-Sanchez, D., Devlin, R. B., Cascio, W., Hauser, E. R., … Ward-Caviness, C. K. (2019). Evaluating DNA methylation age on the Illumina MethylationEPIC Bead Chip. PLoS ONE, 14(4). doi: 10.1371/journal.pone.0207834

 

Eekhout, I., Reijnen, A., Vermetten, E., & Geuze, E. (2016). Post-traumatic stress symptoms 5 years after military deployment to Afghanistan: An observational cohort study. The Lancet Psychiatry, 3(1). doi: 10.1016/S2215-0366(15)00368-5

 

Gassen, N. C., Chrousos, G. P., Binder, E. B., & Zannas, A. S. (2017). Life stress, glucocorticoid signaling, and the aging epigenome: Implications for aging-related diseases. Neuroscience and Biobehavioral Reviews, Vol. 74. doi: 10.1016/j.neubiorev.2016.06.003

 

Goldmann, E., Aiello, A., Uddin, M., Delva, J., Koenen, K., Gant, L. M., & Galea, S. (2011). Pervasive exposure to violence and posttraumatic stress disorder in a predominantly African American Urban Community: The Detroit neighborhood health study. Journal of Traumatic Stress, 24(6). doi: 10.1002/jts.20705

 

Hawn, S. E., Zhao, X., Miller, M. W., Wallander, S., Govan, C., Stone, A., … Wolf, E. J. (2023). PTSD and Alcohol Use Disorders Predict the Pace of Cellular Aging. Journal of Mood and Anxiety Disorders, 0(0), 100026. doi: 10.1016/J.XJMAD.2023.100026

 

Hillary, R. F., Stevenson, A. J., Cox, S. R., McCartney, D. L., Harris, S. E., Seeboth, A., … Marioni, R. E. (2021). An epigenetic predictor of death captures multi-modal measures of brain health. Molecular Psychiatry, 26(8). doi: 10.1038/s41380-019-0616-9

 

Horvath, S. (2013). DNA methylation age of human tissues and cell types. Genome Biology, 14(10). doi: 10.1186/gb-2013-14-10-r115

 

Hovens, J. E., Bramsen, I., & Van Der Ploeg, H. M. (2002). Self-rating Inventory for Posttraumatic Stress Disorder: Review of the psychometric properties of a new brief Dutch screening instrument. Perceptual and Motor Skills, Vol. 94. doi:10.2466/pms.2002.94.3.996

 

Jeon, H. Y., Jeong, Y. W., Kim, Y. W., Jeong, Y. I., Hossein, S. M., Yang, H., … Hwang, W. S. (2012). Senescence is accelerated through donor cell specificity in cloned pigs. International Journal of Molecular Medicine, 30(2). doi: 10.3892/ijmm.2012.1003

 

Jovanovic, T., Vance, L. A., Cross, D., Knight, A. K., Kilaru, V., Michopoulos, V., … Smith, A. K. (2017). Exposure to Violence Accelerates Epigenetic Aging in Children. Scientific Reports, 7(1). doi: 10.1038/s41598-017-09235-9

 

Katrinli, S., Maihofer, A. X., Wani, A. H., Pfeiffer, J. R., Ketema, E., Ratanatharathorn, A., … Uddin, M. (2022). Epigenome-wide meta-analysis of PTSD symptom severity in three military cohorts implicates DNA methylation changes in genes involved in immune system and oxidative stress. Molecular Psychiatry, 27(3). doi: 10.1038/s41380-021-01398-2

 

Katrinli, S., Stevens, J., Wani, A. H., Lori, A., Kilaru, V., van Rooij, S. J. H., … Smith, A. K. (2020). Evaluating the impact of trauma and PTSD on epigenetic prediction of lifespan and neural integrity. Neuropsychopharmacology, 45(10). doi: 10.1038/s41386-020-0700-5

 

Kessler, R. C., Santiago, P. N., Colpe, L. J., Dempsey, C. L., First, M. B., Heeringa, S. G., … Ursano, R. J. (2013). Clinical reappraisal of the Composite International Diagnostic Interview Screening Scales (CIDI-SC) in the Army Study to Assess Risk and Resilience in Servicemembers (Army STARRS). International Journal of Methods in Psychiatric Research, 22(4). doi: 10.1002/mpr.1398

 

Laliberté E. (2022). metacor: Meta-analysis of correlation coefficients (R package version 1.0-2.1).

 

Levine, M. E., Hosgood, H. D., Chen, B., Absher, D., Assimes, T., & Horvath, S. (2015). DNA methylation age of blood predicts future onset of lung cancer in the women’s health initiative. Aging, 7(9), 690–700. doi: 10.18632/AGING.100809

 

Li, S., Wong, E. M., Bui, M., Nguyen, T. L., Joo, J. H. E., Stone, J., … Hopper, J. L. (2018). Causal effect of smoking on DNA methylation in peripheral blood: A twin and family study. Clinical Epigenetics, 10(1). doi: 10.1186/s13148-018-0452-9

 

Liu, C., Wang, Z., Hui, Q., Goldberg, J., Smith, N. L., Shah, A. J., … Vaccarino, V. (2022). Association between depression and epigenetic age acceleration: A co-twin control study. Depression and Anxiety, 39(12). doi: 10.1002/da.23279

 

Lu, A. T., Quach, A., Wilson, J. G., Reiner, A. P., Aviv, A., Raj, K., … Horvath, S. (2019). DNA methylation GrimAge strongly predicts lifespan and healthspan. Aging, 11(2). doi:10.18632/aging.101684

 

Marioni, R. E., Harris, S. E., Shah, S., McRae, A. F., von Zglinicki, T., Martin-Ruiz, C., … Deary, I. J. (2016). The epigenetic clock and telomere length are independently associated with chronological age and mortality. International Journal of Epidemiology, 45(2). doi:10.1093/ije/dyw041

 

Marioni, R. E., Shah, S., McRae, A. F., Chen, B. H., Colicino, E., Harris, S. E., … Deary, I. J. (2015). DNA methylation age of blood predicts all-cause mortality in later life. Genome Biology, 16(1). doi: 10.1186/s13059-015-0584-6

 

McGlinchey, R. E., Milberg, W. P., Fonda, J. R., & Fortier, C. B. (2017). A methodology for assessing deployment trauma and its consequences in OEF/OIF/OND veterans: The TRACTS longitudinal prospective cohort study. International Journal of Methods in Psychiatric Research, 26(3). doi: 10.1002/mpr.1556

 

McLean, S. A., Ressler, K., Koenen, K. C., Neylan, T., Germine, L., Jovanovic, T., … Kessler, R. (2020). The AURORA Study: a longitudinal, multimodal library of brain biology and function after traumatic stress exposure. Molecular Psychiatry, 25(2). doi: 10.1038/s41380-019-0581-3

 

Mehta, D., Bruenig, D., Pierce, J., Sathyanarayanan, A., Stringfellow, R., Miller, O., … Shakespeare-Finch, J. (2022). Recalibrating the epigenetic clock after exposure to trauma: The role of risk and protective psychosocial factors. Journal of Psychiatric Research, 149. doi: 10.1016/j.jpsychires.2021.11.026

 

Meier, H. C. S., Mitchell, C., Karadimas, T., & Faul, J. D. (2023). Systemic inflammation and biological aging in the Health and Retirement Study. GeroScience, 45(6). doi:10.1007/s11357-023-00880-9

 

Miller, M. W., & Sadeh, N. (2014). Traumatic stress, oxidative stress and post-Traumatic stress disorder: Neurodegeneration and the accelerated-aging hypothesis. Molecular Psychiatry, Vol. 19. doi: 10.1038/mp.2014.111

 

Morrison, F. G., Logue, M. W., Guetta, R., Maniates, H., Stone, A., Schichman, S. A., … Wolf, E. J. (2019). Investigation of bidirectional longitudinal associations between advanced epigenetic age and peripheral biomarkers of inflammation and metabolic syndrome. Aging, 11(11). doi: 10.18632/aging.101992

 

Nievergelt, C. M., Maihofer, A. X., Klengel, T., Atkinson, E. G., Chen, C. Y., Choi, K. W., … Koenen, K. C. (2019). International meta-analysis of PTSD genome-wide association studies identifies sex- and ancestry-specific genetic risk loci. Nature Communications, 10(1). doi: 10.1038/s41467-019-12576-w

 

Nievergelt, C. M., Maihofer, A. X., Mustapic, M., Yurgil, K. A., Schork, N. J., Miller, M. W., … Baker, D. G. (2015). Genomic predictors of combat stress vulnerability and resilience in U.S. Marines: A genome-wide association study across multiple ancestries implicates PRTFDC1 as a potential PTSD gene. Psychoneuroendocrinology, 51. doi:10.1016/j.psyneuen.2014.10.017

 

Okazaki, S., Numata, S., Otsuka, I., Horai, T., Kinoshita, M., Sora, I., … Hishimoto, A. (2020). Decelerated epigenetic aging associated with mood stabilizers in the blood of patients with bipolar disorder. Translational Psychiatry, 10(1). doi: 10.1038/s41398-020-0813-y

 

Ratanatharathorn, A., Boks, M. P., Maihofer, A. X., Aiello, A. E., Amstadter, A. B., Ashley- Koch, A. E., … Smith, A. K. (2017). Epigenome-wide association of PTSD from heterogeneous cohorts with a common multi-site analysis pipeline. American Journal of Medical Genetics, Part B: Neuropsychiatric Genetics, 174(6). doi: 10.1002/ajmg.b.32568

 

Reed, R. G., Carroll, J. E., Marsland, A. L., & Manuck, S. B. (2022). DNA methylation-based measures of biological aging and cognitive decline over 16-years: preliminary longitudinal findings in midlife. Aging, 14(23). doi: 10.18632/aging.204376

 

Reijnen, A., Rademaker, A. R., Vermetten, E., & Geuze, E. (2015). Prevalence of mental health symptoms in Dutch military personnel returning from deployment to Afghanistan: A 2-year longitudinal analysis. European Psychiatry, 30(2). doi: 10.1016/j.eurpsy.2014.05.003

 

Roberts, A. L., Koenen, K. C., Chen, Q., Gilsanz, P., Mason, S. M., Prescott, J., … Kubzansky, L. D. (2017). Posttraumatic stress disorder and accelerated aging: PTSD and leukocyte telomere length in a sample of civilian women. Depression and Anxiety, 34(5). doi:10.1002/da.22620

 

Rosen, A. D., Robertson, K. D., Hlady, R. A., Muench, C., Lee, J., Philibert, R., … Lohoff, F. W. (2018). DNA methylation age is accelerated in alcohol dependence. Translational Psychiatry, 8(1). doi: 10.1038/s41398-018-0233-4

 

Sumner, J. A., Gao, X., Gambazza, S., Dye, C. K., Colich, N. L., Baccarelli, A. A., … McLaughlin, K. A. (2023). Stressful life events and accelerated biological aging over time in youths. Psychoneuroendocrinology, 151. doi: 10.1016/j.psyneuen.2023.106058

 

Sumner, J. A., Maihofer, A. X., Michopoulos, V., Rothbaum, A. O., Almli, L. M., Andreassen, O. A., … Wolf, E. J. (2021). Examining Individual and Synergistic Contributions of PTSD and Genetics to Blood Pressure: A Trans-Ethnic Meta-Analysis. Frontiers in Neuroscience, 15. doi: 10.3389/fnins.2021.678503

 

Teschendorff, A. E., Breeze, C. E., Zheng, S. C., & Beck, S. (2017). A comparison of reference-based algorithms for correcting cell-type heterogeneity in Epigenome-Wide Association Studies. BMC Bioinformatics, 18(1). doi: 10.1186/s12859-017-1511-5

 

Uddin, M., Aiello, A. E., Wildman, D. E., Koenen, K. C., Pawelec, G., De Los Santos, R., … Galea, S. (2010). Epigenetic and immune function profiles associated with posttraumatic stress disorder. Proceedings of the National Academy of Sciences of the United States of America, 107(20). doi: 10.1073/pnas.0910794107

 

Ursano, R. J., Colpe, L. J., Heeringa, S. G., Kessler, R. C., Sehoenbaum, M., & Stein, M. B. (2014). The army study to assess risk and resilience in servicemembers (Army STARRS). Psychiatry (New York), 77(2). doi: 10.1521/psyc.2014.77.2.107

 

Vangeel, E. B., Izzi, B., Hompes, T., Vansteelandt, K., Lambrechts, D., Freson, K., & Claes, S. (2015). DNA methylation in imprinted genes IGF2 and GNASXL is associated with prenatal maternal stress. Genes, Brain and Behavior, 14(8). doi: 10.1111/gbb.12249

 

Vidal, C., Polo, R., Alvarez, K., Falgas-Bague, I., Wang, Y., Lê Cook, B., & Alegría, M. (2018). Co-Occurrence of Posttraumatic Stress Disorder and Cardiovascular Disease among Ethnic/Racial Groups in the United States. Psychosomatic Medicine, 80(7). doi:10.1097/PSY.0000000000000601

 

Viechtbauer, W. (2010). Conducting meta-analyses in R with the metafor. Journal of Statistical Software, 36(3).

 

Weathers, F., Litz, B., Herman, D., Huska, J., & Keane, T. (1993). The PTSD Checklist (PCL): Reliability, Validity, and Diagnostic Utility. Paper Presented at the Annual Convention of the International Society for Traumatic Stress Studies, San Antonio, TX.

 

Weathers, F. W., Bovin, M. J., Lee, D. J., Sloan, D. M., Schnurr, P. P., Kaloupek, D. G., … Marx, B. P. (2018). The clinician-administered ptsd scale for DSM-5 (CAPS-5): Development and initial psychometric evaluation in military veterans. Psychological Assessment, 30(3). doi: 10.1037/pas0000486

 

Wolf, E. J., Logue, M. W., Hayes, J. P., Sadeh, N., Schichman, S. A., Stone, A., … Miller, M. W. (2016). Accelerated DNA methylation age: Associations with PTSD and neural integrity. Psychoneuroendocrinology, 63. doi: 10.1016/j.psyneuen.2015.09.020

 

Wolf, E. J., Logue, M. W., Morrison, F. G., Wilcox, E. S., Stone, A., Schichman, S. A., … Miller, M. W. (2019). Posttraumatic psychopathology and the pace of the epigenetic clock: A longitudinal investigation. Psychological Medicine, 49(5). doi:

10.1017/S0033291718001411

 

Wolf, E. J., Logue, M. W., Stoop, T. B., Schichman, S. A., Stone, A., Sadeh, N., … Miller, M. W. (2018). Accelerated DNA Methylation Age: Associations with Posttraumatic Stress Disorder and Mortality. Psychosomatic Medicine, 80(1). doi:10.1097/PSY.0000000000000506

 

Wolf, E. J., Maniates, H., Nugent, N., Maihofer, A. X., Armstrong, D., Ratanatharathorn, A., … Logue, M. W. (2018). Traumatic stress and accelerated DNA methylation age: A meta- analysis. Psychoneuroendocrinology, 92. doi: 10.1016/j.psyneuen.2017.12.007

 

Wolf, E. J., Miller, M. W., Hawn, S. E., Zhao, X., Wallander, S. E., McCormick, B., … Logue, M. W. (2023). Longitudinal study of traumatic-stress related cellular and cognitive aging. Brain, Behavior, and Immunity. doi: 10.1016/J.BBI.2023.11.009

 

Yaffe, K., Vittinghoff, E., Lindquist, K., Barnes, D., Covinsky, K. E., Neylan, T., … Marmar, C. (2010). Posttraumatic stress disorder and risk of dementia among US veterans. Archives of General Psychiatry, 67(6). doi: 10.1001/archgenpsychiatry.2010.61

 

Yang, R., Wu, G. W. Y., Verhoeven, J. E., Gautam, A., Reus, V. I., Kang, J. I., … Wolkowitz, O. M. (2021). A DNA methylation clock associated with age-related illnesses and mortality is accelerated in men with combat PTSD. Molecular Psychiatry, 26(9). doi:10.1038/s41380-020-0755-z