Advertisement

Posttraumatic Stress Disorder, Depression, and Accelerated Aging: Leukocyte Telomere Length in the Nurses’ Health Study II

Open AccessPublished:June 01, 2022DOI:https://doi.org/10.1016/j.bpsgos.2022.05.006

      Abstract

      Background

      Exposure to trauma, posttraumatic stress disorder (PTSD), and depression have been independently associated with leukocyte telomere length (LTL), a cellular marker of aging associated with mortality and age-related diseases. However, the joint contributions of trauma and its psychological sequelae on LTL have not been examined.

      Methods

      We conducted an analysis of LTL in a subset of women from the Nurses’ Health Study II (N = 1868). Lifetime exposure to traumatic events, PTSD, and depression was assessed with validated measures. DNA was extracted from peripheral blood leukocytes and telomere repeat copy number to single gene copy number was determined by quantitative real-time polymerase chain reaction telomere assay. Linear regression models assessed the association of trauma, PTSD, and depression with LTL after adjustment for health behaviors and medical conditions.

      Results

      Trauma, PTSD, and depression were not independently associated with LTL in mutually adjusted models. However, individuals with severe psychological distress—characterized by comorbid PTSD and depression—had shorter LTL equivalent to being 7.62 years older (95% CI, 0.02 to 17.97) than participants who had never experienced a traumatic event and were not depressed. Further examination found only an association among individuals with the highest number of PTSD symptoms and comorbid depression equivalent to 9.71 additional years of aging (95% CI, 1.36 to 20.49). No effect was found among individuals meeting the minimum threshold for probable PTSD with comorbid depression.

      Conclusions

      Severe psychological distress, as indicated by the presence of comorbid PTSD and depression, may be associated with shorter LTL.
      Previous studies have found that lifetime stress is associated with shorter telomere length, indicating that telomere length may be a marker of psychosocial stress exposure (
      • Mathur M.B.
      • Epel E.
      • Kind S.
      • Desai M.
      • Parks C.G.
      • Sandler D.P.
      • et al.
      Perceived stress and telomere length: A systematic review, meta-analysis, and methodologic considerations for advancing the field.
      ,
      • Willis M.
      • Reid S.N.
      • Calvo E.
      • Staudinger U.M.
      • Factor-Litvak P.
      A scoping systematic review of social stressors and various measures of telomere length across the life course.
      ). Telomeres are repetitive noncoding sequences of TTAGGG at the end of chromosomes that protect chromosomes from deterioration during cell division (
      • Moyzis R.K.
      • Buckingham J.M.
      • Cram L.S.
      • Dani M.
      • Deaven L.L.
      • Jones M.D.
      • et al.
      A highly conserved repetitive DNA sequence, (TTAGGG) n, present at the telomeres of human chromosomes.
      ). The erosion or shortening of telomeres, which occurs naturally during aging and as a response to toxic environmental exposures such as cigarette smoke (
      • Valdes A.M.
      • Andrew T.
      • Gardner J.P.
      • Kimura M.
      • Oelsner E.
      • Cherkas L.F.
      • et al.
      Obesity, cigarette smoking, and telomere length in women.
      ), leads to reproductive senescence and apoptosis of cells (
      • Wong L.M.
      • van der Harst P.
      • de Boer R.
      • Huzen J.
      • van Gilst W.
      • van Veldhuisen D.
      Aging, telomeres and heart failure.
      ). Shorter telomeres have been associated with chronic diseases, including cardiovascular disease (
      • Le Carolyn M.H.
      • Neylan T.C.
      • Na B.
      • Regan M.
      • Zhang Q.
      • Cohen B.E.
      Lifetime trauma exposure and prospective cardiovascular events and all-cause mortality: Findings from the Heart and Soul Study.
      ,
      • Edmondson D.
      • Kronish I.M.
      • Shaffer J.A.
      • Falzon L.
      • Burg M.M.
      Posttraumatic stress disorder and risk for coronary heart disease: A meta-analytic review.
      ,
      • Van der Kooy K.
      • Van Hout H.
      • Marwijk H.
      • Marten H.
      • Stehouwer C.
      • Beekman A.
      Depression and the risk for cardiovascular diseases: Systematic review and meta analysis.
      ) and type 2 diabetes (
      • Knol M.
      • Twisk J.W.
      • Beekman A.T.
      • Heine R.
      • Snoek F.J.
      • Pouwer F.
      Depression as a risk factor for the onset of type 2 diabetes mellitus. A meta-analysis.
      ,
      • Mezuk B.
      • Eaton W.W.
      • Albrecht S.
      • Golden S.H.
      Depression and type 2 diabetes over the lifespan: A meta-analysis.
      ,
      • Vancampfort D.
      • Rosenbaum S.
      • Ward P.B.
      • Steel Z.
      • Lederman O.
      • Lamwaka A.V.
      • et al.
      Type 2 diabetes among people with posttraumatic stress disorder: Systematic review and meta-analysis.
      ,
      • Zhao J.
      • Miao K.
      • Wang H.
      • Ding H.
      • Wang D.W.
      Association between telomere length and type 2 diabetes mellitus: A meta-analysis.
      ), as well as with mortality (
      • Chwastiak L.A.
      • Rosenheck R.A.
      • Desai R.
      • Kazis L.E.
      The association of psychiatric illness and all-cause mortality in the national department of Veterans Affairs Health Care System.
      ,
      • Vancampfort D.
      • Rosenbaum S.
      • Ward P.B.
      • Steel Z.
      • Lederman O.
      • Lamwaka A.V.
      • et al.
      Type 2 diabetes among people with posttraumatic stress disorder: Systematic review and meta-analysis.
      ,
      • Cuijpers P.
      • Vogelzangs N.
      • Twisk J.
      • Kleiboer A.
      • Li J.
      • Penninx B.W.
      Comprehensive meta-analysis of excess mortality in depression in the general community versus patients with specific illnesses.
      ), and are considered a marker of biological age (
      • Aviv A.
      Telomeres and human somatic fitness.
      ). Recent meta-analyses have found that posttraumatic stress disorder (PTSD) (
      • Darrow S.M.
      • Verhoeven J.E.
      • Révész D.
      • Lindqvist D.
      • Penninx B.W.
      • Delucchi K.L.
      • et al.
      The association between psychiatric disorders and telomere length: A meta-analysis involving 14,827 persons.
      ,
      • Li X.
      • Wang J.
      • Zhou J.
      • Huang P.
      • Li J.
      The association between post-traumatic stress disorder and shorter telomere length: A systematic review and meta-analysis.
      ) and depression (
      • Darrow S.M.
      • Verhoeven J.E.
      • Révész D.
      • Lindqvist D.
      • Penninx B.W.
      • Delucchi K.L.
      • et al.
      The association between psychiatric disorders and telomere length: A meta-analysis involving 14,827 persons.
      ,
      • Ridout K.K.
      • Ridout S.J.
      • Price L.H.
      • Sen S.
      • Tyrka A.R.
      Depression and telomere length: A meta-analysis.
      ) are each associated with shorter telomere length. For example, one meta-analysis comprising 5 studies of PTSD and telomere length found that individuals with versus without PTSD had shorter telomeres equivalent to a standardized mean difference of −1.27 (95% CI, −2.12 to −0.43), while individuals with versus without depression had a standardized mean difference of −0.21 (95% CI, −0.29 to −0.12) (
      • Darrow S.M.
      • Verhoeven J.E.
      • Révész D.
      • Lindqvist D.
      • Penninx B.W.
      • Delucchi K.L.
      • et al.
      The association between psychiatric disorders and telomere length: A meta-analysis involving 14,827 persons.
      ). However, while both forms of psychological distress commonly occur following exposure to trauma, previous studies have generally focused on PTSD and depression independently.
      These meta-analyses have reported substantial heterogeneity in effects across studies (e.g., PTSD I2 = 94%; depression I2 = 96%) (
      • Darrow S.M.
      • Verhoeven J.E.
      • Révész D.
      • Lindqvist D.
      • Penninx B.W.
      • Delucchi K.L.
      • et al.
      The association between psychiatric disorders and telomere length: A meta-analysis involving 14,827 persons.
      ) as well as indicated the inconsistency of associations between PTSD or depression with telomere length across study designs (
      • Li X.
      • Wang J.
      • Zhou J.
      • Huang P.
      • Li J.
      The association between post-traumatic stress disorder and shorter telomere length: A systematic review and meta-analysis.
      ). There are 3 plausible reasons for inconsistency in the findings. First, meta-analyses of both forms of distress included studies with relatively small sample sizes [e.g., a mean of 44 participants per study (
      • Darrow S.M.
      • Verhoeven J.E.
      • Révész D.
      • Lindqvist D.
      • Penninx B.W.
      • Delucchi K.L.
      • et al.
      The association between psychiatric disorders and telomere length: A meta-analysis involving 14,827 persons.
      ,
      • Boks M.P.
      • van Mierlo H.C.
      • Rutten B.P.
      • Radstake T.R.
      • De Witte L.
      • Geuze E.
      • et al.
      Longitudinal changes of telomere length and epigenetic age related to traumatic stress and post-traumatic stress disorder.
      ,
      • Wolf E.J.
      • Morrison F.G.
      • Sullivan D.R.
      • Logue M.W.
      • Guetta R.E.
      • Stone A.
      • et al.
      The goddess who spins the thread of life: Klotho, psychiatric stress, and accelerated aging.
      ,
      • Connolly S.L.
      • Stoop T.B.
      • Logue M.W.
      • Orr E.H.
      • De Vivo I.
      • Miller M.W.
      • et al.
      Posttraumatic stress disorder symptoms, temperament, and the pathway to cellular senescence.
      ); or half the studies with <60 participants (
      • Darrow S.M.
      • Verhoeven J.E.
      • Révész D.
      • Lindqvist D.
      • Penninx B.W.
      • Delucchi K.L.
      • et al.
      The association between psychiatric disorders and telomere length: A meta-analysis involving 14,827 persons.
      )]. Second, despite findings that 52% of individuals with PTSD report co-occurring depression (
      • Rytwinski N.K.
      • Scur M.D.
      • Feeny N.C.
      • Youngstrom E.A.
      The co-occurrence of major depressive disorder among individuals with posttraumatic stress disorder: A meta-analysis.
      ), previous studies have generally examined these forms of distress independently without accounting for either overlap between them or shared effects, which may result in obscuring their associations with telomere length (
      • Wolf E.J.
      • Morrison F.G.
      • Sullivan D.R.
      • Logue M.W.
      • Guetta R.E.
      • Stone A.
      • et al.
      The goddess who spins the thread of life: Klotho, psychiatric stress, and accelerated aging.
      ,
      • Connolly S.L.
      • Stoop T.B.
      • Logue M.W.
      • Orr E.H.
      • De Vivo I.
      • Miller M.W.
      • et al.
      Posttraumatic stress disorder symptoms, temperament, and the pathway to cellular senescence.
      ). Third, variations in the prevalence of comorbid PTSD and depression across studies could further explain differing associations. Previous studies have found that depression severity, but not timing of depression onset (e.g., time since onset and leukocyte telomere length [LTL] measurement), has been associated with shorter LTL (
      • Ridout K.K.
      • Ridout S.J.
      • Price L.H.
      • Sen S.
      • Tyrka A.R.
      Depression and telomere length: A meta-analysis.
      ). Because individuals who meet criteria for both PTSD and depression report higher levels of depressive symptoms than those with depression alone (
      • Gros D.F.
      • Price M.
      • Magruder K.M.
      • Frueh B.C.
      Symptom overlap in posttraumatic stress disorder and major depression.
      ,
      • Shalev A.Y.
      • Freedman S.
      • Peri T.
      • Brandes D.
      • Sahar T.
      • Orr S.P.
      • et al.
      Prospective study of posttraumatic stress disorder and depression following trauma.
      ), variation in the prevalence of PTSD and the accompanying severe distress (either comorbid PTSD and depression or higher depression severity) could account for some of the observed variation in associations.
      To date, no study has examined the joint effects of both disorders, as large sample sizes with sufficient comorbidity are needed to disentangle these relationships (
      • Rytwinski N.K.
      • Scur M.D.
      • Feeny N.C.
      • Youngstrom E.A.
      The co-occurrence of major depressive disorder among individuals with posttraumatic stress disorder: A meta-analysis.
      ). Of note, in our own previous work in the same cohort used in the present study, the Nurses’ Health Study II (NHS II), we found in sensitivity analyses some initial suggestion that women with comorbid PTSD and depression had shorter telomeres than those with either disorder alone (
      • Roberts A.L.
      • Koenen K.C.
      • Chen Q.
      • Gilsanz P.
      • Mason S.M.
      • Prescott J.
      • et al.
      Posttraumatic stress disorder and accelerated aging: PTSD and leukocyte telomere length in a sample of civilian women.
      ). However, our previous study was underpowered (N = 116), there was only 1 participant in the group with trauma and depression but no PTSD, and the interaction analyses did not reach statistical significance (
      • Roberts A.L.
      • Koenen K.C.
      • Chen Q.
      • Gilsanz P.
      • Mason S.M.
      • Prescott J.
      • et al.
      Posttraumatic stress disorder and accelerated aging: PTSD and leukocyte telomere length in a sample of civilian women.
      ). Fourth, prior studies vary widely in the confounders included. For example, few studies consider possible effects of childhood abuse in these associations, even though child abuse has been associated with shorter telomere length and PTSD/depression (
      • Li X.
      • Wang J.
      • Zhou J.
      • Huang P.
      • Li J.
      The association between post-traumatic stress disorder and shorter telomere length: A systematic review and meta-analysis.
      ,
      • Ridout K.K.
      • Ridout S.J.
      • Price L.H.
      • Sen S.
      • Tyrka A.R.
      Depression and telomere length: A meta-analysis.
      ,
      • Brewin C.R.
      • Andrews B.
      • Valentine J.D.
      Meta-analysis of risk factors for posttraumatic stress disorder in trauma-exposed adults.
      ,
      • Nanni V.
      • Uher R.
      • Danese A.
      Childhood maltreatment predicts unfavorable course of illness and treatment outcome in depression: A meta-analysis.
      ,
      • Ridout K.K.
      • Levandowski M.
      • Ridout S.J.
      • Gantz L.
      • Goonan K.
      • Palermo D.
      • et al.
      Early life adversity and telomere length: A meta-analysis.
      ). As a result, child abuse may be a confounder or PTSD and depression may serve to mediate the observed associations with telomere length. Some recent studies following up on these meta-analyses have failed to replicate associations of PTSD or depression with telomere length, finding either no effect (
      • Wium-Andersen M.K.
      • Ørsted D.D.
      • Rode L.
      • Bojesen S.E.
      • Nordestgaard B.G.
      Telomere length and depression: Prospective cohort study and Mendelian randomisation study in 67 306 individuals.
      ) or an effect only among subgroups of participants [e.g., Whites vs. African Americans or Hispanics (
      • Wignall N.D.
      • Deschner M.
      • Evans H.M.
      • Brown E.S.
      Relationship between current depressive symptoms and telomere length in a large, multiethnic sample.
      ), females (
      • Liu J.J.
      • Wei Y.B.
      • Forsell Y.
      • Lavebratt C.
      Stress, depressive status and telomere length: Does social interaction and coping strategy play a mediating role?.
      ), and males (
      • Whisman M.A.
      • Richardson E.D.
      Depressive symptoms and salivary telomere length in a probability sample of middle-aged and older adults.
      )].
      In the current article, we build on our previous analysis by using a sample over 10 times larger to address these limitations and disentangle the contributions of trauma, PTSD, and depression to telomere length. We also address the limitations of prior work by conducting sensitivity analyses accounting for childhood abuse and by capitalizing on the rich, longitudinal data of the NHS II to adjust for a wide range of potential covariates. We hypothesized that while trauma, PTSD, and depression would each be independently associated with shorter telomere length after mutual adjustment, comorbid PTSD and depression would be associated with shorter telomere length as compared with trauma alone or each disorder independently.

      Methods and Materials

      Participants for the current study included women from the NHS II, an ongoing cohort of 116,430 female nurses recruited in 1989 and assessed every 2 years via questionnaires assessing health-related factors. Participants were ages 24 to 44 years at baseline. Across follow-up, women are invited to complete additional questionnaires and substudies. Blood samples from a subset of the full cohort were collected at 2 time points, between 1996 and 1999 (blood draw 1) or between 2010 and 2012 (blood draw 2). A total of 92,888 women who had completed the 1995 biennial questionnaire and had no previous diagnoses of cancer were invited to participate in the first blood draw, 29,611 (31.9%) of whom participated. All women who completed the first blood draw were invited to participate in the second blood draw, 16,424 (55.5%) of whom participated. For both draws, women were mailed collection kits with all necessary supplies and blood sample–related questionnaires. Samples were returned to the NHS II laboratory via overnight courier and were processed by the lab; separated into plasma, red blood cell, and white blood cell components; and stored in liquid nitrogen freezers. Prior studies with these data suggest that women who participated in the blood sample collection were similar to the whole NHS II cohort (
      • Sumner J.A.
      • Chen Q.
      • Roberts A.L.
      • Winning A.
      • Rimm E.B.
      • Gilsanz P.
      • et al.
      Cross-sectional and longitudinal associations of chronic posttraumatic stress disorder with inflammatory and endothelial function markers in women.
      ,
      • Sumner J.A.
      • Chen Q.
      • Roberts A.L.
      • Winning A.
      • Rimm E.B.
      • Gilsanz P.
      • et al.
      Posttraumatic stress disorder onset and inflammatory and endothelial function biomarkers in women.
      ). Our sample included women who participated in a PTSD substudy of NHS II (2008), which assessed trauma, PTSD, and depression symptoms (
      • Koenen K.C.
      • De Vivo I.
      • Rich-Edwards J.
      • Smoller J.W.
      • Wright R.J.
      • Purcell S.M.
      Protocol for investigating genetic determinants of posttraumatic stress disorder in women from the Nurses' Health Study II.
      ), and had LTL measured as part of their participation in additional NHS II substudies including ones on childhood abuse (
      • Mason S.M.
      • Prescott J.
      • Tworoger S.S.
      • DeVivo I.
      • Rich-Edwards J.W.
      Childhood physical and sexual abuse history and leukocyte telomere length among women in middle adulthood.
      ), religion and spirituality (Spirituality Study) (
      • Spence N.D.
      • Farvid M.S.
      • Warner E.T.
      • VanderWeele T.J.
      • Tworoger S.S.
      • Argentieri M.A.
      • et al.
      Religious service attendance, religious coping, and risk of hypertension in women participating in the Nurses’ Health Study II.
      ), or lifetime stress (Mind-Body Study) (
      • Huang T.
      • Trudel-Fitzgerald C.
      • Poole E.M.
      • Sawyer S.
      • Kubzansky L.D.
      • Hankinson S.E.
      • et al.
      The Mind–Body Study: Study design and reproducibility and interrelationships of psychosocial factors in the Nurses’ Health Study II.
      ). Mailed return of the questionnaires implied consent. This study was approved by the Partners HealthCare Human Research Committee.

      Lifetime Trauma

      Lifetime trauma was assessed using the Brief Trauma Questionnaire, which listed 15 types of traumatic events plus a 16th “other” option, with participants asked to endorse any event they experienced (
      • Sumner J.A.
      • Chen Q.
      • Roberts A.L.
      • Winning A.
      • Rimm E.B.
      • Gilsanz P.
      • et al.
      Cross-sectional and longitudinal associations of chronic posttraumatic stress disorder with inflammatory and endothelial function markers in women.
      ,
      • Roberts A.L.
      • Dohrenwend B.P.
      • Aiello A.
      • Wright R.J.
      • Maercker A.
      • Galea S.
      • et al.
      The stressor criterion for posttraumatic stress disorder: Does it matter?.
      ). Participants were then asked which traumatic event happened first, which traumatic event they considered their worst, and at what ages each occurred. Participants were considered trauma exposed if they endorsed any type of traumatic event, including an event identified via the other category. Informed consent was received from all participants and the study protocol was approved by the Institutional Review Boards of Brigham and Women's Hospital and the Harvard T.H. Chan School of Public Health.

      Posttraumatic Stress Disorder

      The Short Screening Scale for DSM-IV Posttraumatic Stress Disorder was used to assess 5 PTSD avoidance symptoms (criterion C) and 2 hyperarousal symptoms (criterion D) in relation to a participant’s worst trauma (
      • Breslau N.
      • Peterson E.L.
      • Kessler R.C.
      • Schultz L.R.
      Short screening scale for DSM-IV posttraumatic stress disorder.
      ). Participants that reported a history of 4 or more symptoms in relation to their worst traumatic event were considered to have a probable diagnosis of PTSD, with onset identified according to the date of their worst traumatic event. Prior work has demonstrated that a cutoff of 4 symptoms on this measure classified PTSD cases with a sensitivity of 85%, a specificity of 93%, a positive predictive value of 68%, and a negative predictive value of 98% in a validation study (
      • Lee Y.C.
      • Agnew-Blais J.
      • Malspeis S.
      • Keyes K.
      • Costenbader K.
      • Kubzansky L.D.
      • et al.
      Post-traumatic stress disorder and risk for incident rheumatoid arthritis.
      ). In the NHS II, the reliability of self-reported age of onset of trauma and PTSD has been excellent when comparing the screener with the reported age of onset in an interview within a subset of participants (intraclass correlation coefficient = 0.95) (
      • McCabe C.
      • Thomas K.
      • Brazier J.
      • Coleman P.
      Measuring the mental health status of a population: A comparison of the GHQ–12 and the SF–36 (MHI–5).
      ).

      History of Depression

      Two different classification algorithms for lifetime history of depression were used, based on available measures. For the first blood collection (1996–1999), following prior work, participants with a score ≤60 on the 5-item Mental Health Index in the survey prior to their blood collection (either 1993 or 1997) were considered to have a history of depression (
      • McCabe C.
      • Thomas K.
      • Brazier J.
      • Coleman P.
      Measuring the mental health status of a population: A comparison of the GHQ–12 and the SF–36 (MHI–5).
      ). For the second blood collection (2010–2012), we used 3 indicators for history of depression: 1) whether participants experienced 2 or more weeks of depressive symptoms (asked every 2 years starting in 2001); 2) if they were ever diagnosed with depression by a doctor (asked every 2 years starting in 2003); and 3) the Center for Epidemiologic Studies Depression Scale, 10-Item Version (asked in 2008) (
      • Irwin M.
      • Artin K.H.
      • Oxman M.N.
      Screening for depression in the older adult: Criterion validity of the 10-item Center for Epidemiological Studies Depression Scale (CES-D).
      ). Participants answering yes to either question or scoring ≥10 on the Center for Epidemiologic Studies Depression Scale, 10-Item Version were classified as having probable depression (
      • Irwin M.
      • Artin K.H.
      • Oxman M.N.
      Screening for depression in the older adult: Criterion validity of the 10-item Center for Epidemiological Studies Depression Scale (CES-D).
      ).

      Co-occurrence of Trauma, PTSD, and Depression

      Based on the dichotomous lifetime trauma, lifetime PTSD, and depression variables, joint occurrence of trauma, PTSD, and depression was coded as follows: 1) no trauma, no depression (reference); 2) no trauma, depression; 3) trauma, no depression, no PTSD; 4) trauma, depression, no PTSD; 5) trauma, no depression, PTSD; and 6) trauma, depression and PTSD.

      Leukocyte Telomere Length

      LTL was assessed in batches of assays across the 3 NHS II substudies (childhood abuse substudy, Spirituality Study, Mind-Body Study) with all assays conducted at the same laboratory using the same procedures, which have been described elsewhere (
      • McGrath M.
      • Wong J.Y.
      • Michaud D.
      • Hunter D.J.
      • De Vivo I.
      Telomere length, cigarette smoking, and bladder cancer risk in men and women.
      ). Briefly, genomic DNA was extracted from peripheral blood leukocytes using the QIAmp (Qiagen) 96-spin blood protocol. PicoGreen DNA quantitation was conducted using a Molecular Devices 96-well spectrophotometer. Genomic DNA was subsequently dried down and resuspended. The telomere repeat copy number to single gene copy number (T/S) ratio was determined by a modified version of the quantitative real-time polymerase chain reaction telomere assay run on the Applied Biosystems 7900HT PCR System. Triplicate reactions of each assay were done on each sample. As previously reported in the childhood abuse substudy, the coefficient of variation (CV) for the exponentiated T/S ratio of blinded quality control (QC) samples across batches ranged from 23% to 33%, while within-replicate CVs for participants and blinded QC samples were lower within plates (range, 10.7%–19.4%), further dropping to 14% to 26% after excluding QC samples with within-triplicate CVs >20%. When nonblinded laboratory QC samples were included across plates, CVs in these samples averaged 11.5% (
      • Roberts A.L.
      • Koenen K.C.
      • Chen Q.
      • Gilsanz P.
      • Mason S.M.
      • Prescott J.
      • et al.
      Posttraumatic stress disorder and accelerated aging: PTSD and leukocyte telomere length in a sample of civilian women.
      ). Based on these results, we concluded that most variation occurred between batches.
      LTL, reported as the exponentiated T/S ratio, was log transformed, as values deviated significantly from normality. The Rosner batch-correction method (
      • Rosner B.
      • Cook N.
      • Portman R.
      • Daniels S.
      • Falkner B.
      Determination of blood pressure percentiles in normal-weight children: Some methodological issues.
      ) was used to correct for possible batch effects by first regressing the log-LTL on indicators for laboratory batch and potential predictors of LTL that might vary by batch, including participant’s year of birth; father’s age at birth; and age, smoking status, and body mass index (BMI) at time of blood draw. A correction parameter for each batch was estimated by subtracting the average of all batch parameter estimates from the parameter estimate for each individual batch. A batch-adjusted log-LTL for each sample was then created by adjusting the log-LTL by subtracting that batch’s correction parameter. Although quantitative real-time polymerase chain reaction method is known to introduce variation in LTL measurement, the variation is nondifferential across outcomes, as assays were conducted blinded to PTSD and depression status.

      Covariates

      Age at blood draw was determined by self-reported birthdate and the date of the blood draw. Self-reported BMI was collected at the time of the blood draw (
      • Rimm E.B.
      • Stampfer M.J.
      • Colditz G.A.
      • Chute C.G.
      • Litin L.B.
      • Willett W.C.
      Validity of self-reported waist and hip circumferences in men and women.
      ). Lifetime smoking status from the preceding survey was coded as never versus ever. Participants’ diets from the most recent survey prior to the blood draw were classified into quintiles of healthiness based on a validated food frequency questionnaire (
      • Willett W.
      • Sampson L.
      • Browne M.
      • Stampfer M.J.
      • Rosner B.
      • Hennekens C.H.
      • et al.
      The use of a self-administered questionnaire to assess diet four years in the past.
      ) and the Alternative Healthy Eating Index (
      • McCullough M.L.
      • Feskanich D.
      • Stampfer M.J.
      • Giovannucci E.L.
      • Rimm E.B.
      • Hu F.B.
      • et al.
      Diet quality and major chronic disease risk in men and women: Moving toward improved dietary guidance.
      ). Past month alcohol consumption was coded as 0 drinks/mo, 1 to 3 drinks/mo, 1 drink/wk, 2 to 4 drinks/wk, or 5+ drinks/wk. Frequency of exercising heavily enough to sweat was assessed in the closest biennial questionnaire prior to blood draws as <1, 1, 2 to 3, or 4+ times/wk. Self-reported physician diagnoses of high cholesterol, high blood pressure, and type 2 diabetes were assessed in the biennial survey prior to blood draw. Childhood physical and emotional abuse before age 12 was assessed using 5 questions from the physical and emotional abuse subscales of the Childhood Trauma Questionnaire included as part of the violence substudy in 2001 (
      • Bernstein D.P.
      • Fink L.
      • Handelsman L.
      • Foote J.
      • Lovejoy M.
      • Wenzel K.
      • et al.
      Initial reliability and validity of a new retrospective measure of child abuse and neglect.
      ). The frequency of family members 1) hitting so hard it left bruises, 2) punishing in a way that seemed cruel, 3) insulting, 4) screaming and yelling, and 5) punishing with a belt or other hard object was queried. Following prior work in this sample (
      • Roberts A.L.
      • Chen Y.
      • Slopen N.
      • McLaughlin K.A.
      • Koenen K.C.
      • Austin S.B.
      Maternal experience of abuse in childhood and depressive symptoms in adolescent and adult offspring: A 21-year longitudinal study.
      ), responses of never (scored as 0), rarely (scored as 1), sometimes (scored as 2), often (scored as 3), or very often (scored as 4) were summed, and then overall scores were divided into quartiles. Sexual abuse was assessed with the Sexual Maltreatment Scale of the Conflict Tactics Scales (
      • Straus M.A.
      • Hamby S.L.
      • Finkelhor D.
      • Moore D.W.
      • Runyan D.
      Identification of child maltreatment with the Parent-Child Conflict Tactics Scales: Development and psychometric data for a national sample of American parents.
      ), which consists of 2 questions regarding unwanted sexual touching and forced or coerced sexual contact by an adult or older child before age 12 years and 2 questions about experiences between ages 12 and 17. Participants could respond, “No this never happened” (scored as 0), “Yes this happened once” (scored as 1), or “Yes this happened more than once” (scored as 2). Responses were summed and individuals were then classified as having experienced no sexual abuse (score = 0), infrequent sexual abuse (score = 1 or 2), moderately frequent sexual abuse (score = 3 or 4), or frequent sexual abuse (score ≥ 5).

      Statistical Analyses

      We examined the distribution of LTL and prevalence of covariates by the joint trauma, PTSD, and depression exposure. Differences in the means and standard deviations of age and BMI by exposure group were examined using Kruskal-Wallis tests. Prevalence of smoking, diet, alcohol use, high cholesterol, high blood pressure, type 2 diabetes, and frequency of exercise in each group was calculated, and group differences were assessed using χ2 tests.
      We fit a series of linear regressions estimating the association of each level of trauma, PTSD, and depression with log-transformed, batch-corrected LTL, adjusting for 1) demographics (e.g., age, race) and substudy; 2) demographics plus health behaviors including BMI, smoking, alcohol use, exercise, and diet; and 3) demographics, health behaviors, and health risk factors, including type 2 diabetes, high blood pressure, and the use of cholesterol-lowering drugs. We expanded on this main analysis in 3 additional analyses outlined in Table 1. First, we assessed whether the effect of trauma, PTSD, or depression varied by the participant’s age by fitting an additional model including an interaction between age and trauma, PTSD, depression, and their joint effects. Second, to assess whether the severity of PTSD was associated with shorter LTL and not just probable PTSD, we recategorized participants as having no PTSD symptoms, 1 to 3 PTSD symptoms (subclinical), 4 to 5 PTSD symptoms (moderate), and 6 to 7 PTSD symptoms (high). The joint contribution of trauma, depression, and PTSD severity was assessed by creating indicator variables as follows: no trauma/no depression (reference), trauma/no depression/no PTSD symptoms, trauma/no depression/1 to 3 PTSD symptoms, trauma/no depression/4 to 5 PTSD symptoms, trauma/no depression/6 to 7 PTSD symptoms, no trauma/depression, trauma/depression/no PTSD symptoms, trauma/depression/1 to 3 PTSD symptoms, trauma/depression/4 to 5 PTSD symptoms, and trauma/depression/6 to 7 PTSD symptoms. We also conducted a test of trend across levels of PTSD exposure by creating a continuous variable of PTSD severity (range, 0–3) among individuals with and without depression to assess a possible dose-response association with PTSD severity. Third, we similarly assessed whether severity of depression symptoms was associated with shorter LTL by categorizing individuals into quartiles of either 5-item Mental Health Index scores (for LTL assessed in blood draw 1) or the 2008 Center for Epidemiologic Studies Depression Scale, 10-Item Version (for LTL assessed in blood draw 2). We then stratified depression quartile by trauma and probable PTSD status resulting in 12 categories of exposure. Models were fit as above, and we also conducted a test of trend across levels of depression exposure separately among individuals with and without trauma and PTSD to assess a possible dose-response association.
      Table 1Trauma, Depression, and PTSD Categories and Number of Exposure Groups Across Analyses
      AnalysisTraumaDepressionPTSDNo. of Categories
      MainExposedProbable depressionProbable PTSD6
      UnexposedNo probable depressionNo probable PTSD
      Age InteractionExposedProbable depressionProbable PTSD11
      UnexposedNo probable depressionNo probable PTSD
      PTSD SeverityExposedProbable depressionNo PTSD symptoms10
      UnexposedNo probable depression1–3 PTSD symptoms
      4–5 PTSD symptoms
      6–7 PTSD symptoms
      Depression SeverityExposedDepression quartilesProbable PTSD12
      UnexposedNo probable PTSD
      PTSD, posttraumatic stress disorder.
      To aid the interpretation of our results, we estimated the equivalent number of years of aging for each level of joint trauma, PTSD, and depression variable by dividing their β coefficients by the β coefficient for age, which reflects the effect of 1 year of aging, from a model fit without trauma, PTSD, or depression (
      • Monroe D.M.
      • Goldstein R.L.
      • Teylan M.A.
      • Hart J.E.
      • DeVivo I.
      • Orr E.H.
      • et al.
      Clinical associations with telomere length in chronic spinal cord injury.
      ). A set of 1000 bootstraps was conducted to estimate a 95% confidence interval. To estimate the degree to which unmeasured confounders could account for the associations between trauma, PTSD, depression, and their joint effects on LTL, we calculated the minimum associations that the confounder would need to have in order to fully account for the association of trauma, PTSD, and depression with LTL using the E-value (
      • VanderWeele T.J.
      • Ding P.
      Sensitivity analysis in observational research: introducing the E-value.
      ).
      We conducted 4 sensitivity analyses to assess the robustness and of our findings as well as the potential selection bias. First, the history of depression of participants whose blood was collected in 1996 to 1999 may have been underdiagnosed, as the 5-item Mental Health Index assessed depressive symptoms in the last 4 weeks only. At the time of this blood collection, participants were ages 31 to 51, and 12% met thresholds for probable depression. Nationally representative studies have found a lifetime prevalence of 23.2% for depression, with the onset of 90% of cases by age 55 (
      • Kessler R.C.
      • Berglund P.
      • Demler O.
      • Jin R.
      • Merikangas K.R.
      • Walters E.E.
      Lifetime prevalence and age-of-onset distributions of DSM-IV disorders in the National Comorbidity Survey Replication.
      ). While differences in prevalence between our sample and other studies could be due to real differences between the NHS II cohort and the general U.S. population, we nevertheless tested the effect of misclassifying 9% of the sample as not having a depression diagnosis by randomly assigning 9% of these participants (n = 73) to have depression, refitting our models, and pooling results across 1000 iterations.
      Second, as childhood abuse has been associated with PTSD (
      • Brewin C.R.
      • Andrews B.
      • Valentine J.D.
      Meta-analysis of risk factors for posttraumatic stress disorder in trauma-exposed adults.
      ), depression (
      • Norman R.E.
      • Byambaa M.
      • De R.
      • Butchart A.
      • Scott J.
      • Vos T.
      The long-term health consequences of child physical abuse, emotional abuse, and neglect: A systematic review and meta-analysis.
      ), and LTL (
      • Darrow S.M.
      • Verhoeven J.E.
      • Révész D.
      • Lindqvist D.
      • Penninx B.W.
      • Delucchi K.L.
      • et al.
      The association between psychiatric disorders and telomere length: A meta-analysis involving 14,827 persons.
      ,
      • Zhao J.
      • Miao K.
      • Wang H.
      • Ding H.
      • Wang D.W.
      Association between telomere length and type 2 diabetes mellitus: A meta-analysis.
      ) and represents a potential common cause, we fit a model further adjusted for childhood physical, emotional, and sexual abuse to assess potential confounding. Third, we excluded participants (n = 116) who were included in the previous NHS II analysis conducted by Roberts et al. (
      • Roberts A.L.
      • Koenen K.C.
      • Chen Q.
      • Gilsanz P.
      • Mason S.M.
      • Prescott J.
      • et al.
      Posttraumatic stress disorder and accelerated aging: PTSD and leukocyte telomere length in a sample of civilian women.
      ) to assess whether our results were driven primarily by this previous sample. Fourth, the childhood abuse substudy selected participants based on cardiovascular disease status, which previous studies have found to be associated with both PTSD and shorter telomere length (
      • Sumner J.A.
      • Kubzansky L.D.
      • Elkind M.S.
      • Roberts A.L.
      • Agnew-Blais J.
      • Chen Q.
      • et al.
      Trauma exposure and posttraumatic stress disorder symptoms predict onset of cardiovascular events in women.
      ,
      • Weischer M.
      • Bojesen S.E.
      • Cawthon R.M.
      • Freiberg J.J.
      • Tybjærg-Hansen A.
      • Nordestgaard B.G.
      Short telomere length, myocardial infarction, ischemic heart disease, and early death.
      ). Thus, using participants from the childhood abuse substudy may artificially induce an association between PTSD/depression and LTL. To address whether the selection criteria used to recruit participants into the different substudies affected our results, we fit models stratified by substudy (e.g., childhood abuse substudy, Mind-Body Study, and Spirituality Study).

      Results

      Significant differences in age, BMI, smoking status, exercise, high blood pressure, type 2 diabetes, and cholesterol drug use were found by PTSD/depression exposure status (Table 2). Women who experienced trauma and reported PTSD and/or depression were more likely to have been lifetime smokers (p < .001) and have higher BMI (p < .001) than women without PTSD/depression regardless of trauma history. Women with both PTSD and depression were more likely to report high blood pressure (p < .001) and type 2 diabetes than women who reported only PTSD or depression.
      Table 2Descriptive Statistics by Trauma, PTSD, and Depression Status
      T–P–D–, n = 294T+P–D–, n = 834T–P–D+, n = 62T+P+D–, n = 238T+P–D+, n = 265T+P+D+, n = 175Total, N = 1868p Value
      Age at Blood Draw, Years, Mean (SD)49.69 (7.39)50.59 (7.14)54.12 (6.74)48.48 (6.57)55.12 (6.67)54.02 (6.83)51.26 (7.32)<.001
      Race, n (%).723
       Other
      “Other” includes participants identifying as American Indian, Asian, Black, Hawaiian, Multiracial, or Other/Unknown.
      15 (5.1%)38 (4.6%)1 (1.6%)13 (5.5%)9 (3.4%)9 (5.1%)85 (4.6%)
       White279 (94.9%)796 (95.4%)61 (98.4%)225 (94.5%)256 (96.6%)166 (94.9%)1783 (95.4%)
      Substudy, n (%)<.001
       Childhood abuse152 (51.7%)408 (48.9%)20 (32.3%)143 (60.1%)49 (18.5%)44 (25.1%)816 (43.7%)
       Mind-Body Study24 (8.2%)81 (9.7%)2 (3.2%)31 (13.0%)10 (3.8%)9 (5.1%)157 (8.4%)
       Spirituality Study118 (40.1%)345 (41.4%)40 (64.5%)64 (26.9%)206 (77.7%)122 (69.7%)895 (47.9%)
      BMI, Mean (SD)25.65 (5.33)25.93 (5.26)25.14 (3.25)26.63 (5.98)27.18 (5.93)27.58 (6.24)26.28 (5.55)<.001
      Lifetime Smoker, n (%)71 (24.1%)264 (31.7%)20 (32.3%)97 (40.8%)84 (31.7%)77 (44.0%)613 (32.8%)<.001
      Past Month Alcohol Use, n (%).589
       0 drinks/mo113 (38.4%)321 (38.5%)16 (25.8%)86 (36.1%)105 (39.6%)65 (37.1%)706 (37.8%)
       1–3 drinks/mo67 (22.8%)175 (21.0%)14 (22.6%)67 (28.2%)51 (19.2%)33 (18.9%)407 (21.8%)
       1 drink/wk30 (10.2%)85 (10.2%)7 (11.3%)26 (10.9%)24 (9.1%)22 (12.6%)194 (10.4%)
       2–4 drinks/wk46 (15.6%)129 (15.5%)15 (24.2%)29 (12.2%)40 (15.1%)27 (15.4%)286 (15.3%)
       5+ drinks/wk38 (12.9%)124 (14.9%)10 (16.1%)30 (12.6%)45 (17.0%)28 (16.0%)275 (14.7%)
      Exercise, n (%).025
       <1 time/wk71 (24.1%)194 (23.3%)14 (22.6%)75 (31.5%)70 (26.4%)41 (23.4%)465 (24.9%)
       1 time/wk45 (15.3%)124 (14.9%)12 (19.4%)39 (16.4%)20 (7.5%)29 (16.6%)269 (14.4%)
       2–3 times/wk94 (32.0%)284 (34.1%)18 (29.0%)73 (30.7%)82 (30.9%)61 (34.9%)612 (32.8%)
       4+ times/wk84 (28.6%)232 (27.8%)18 (29.0%)51 (21.4%)93 (35.1%)44 (25.1%)522 (27.9%)
      Diet Quintile, n (%).375
       158 (19.7%)138 (16.5%)11 (17.7%)37 (15.5%)41 (15.5%)20 (11.4%)305 (16.3%)
       259 (20.1%)155 (18.6%)16 (25.8%)35 (14.7%)48 (18.1%)29 (16.6%)342 (18.3%)
       358 (19.7%)167 (20.0%)8 (12.9%)60 (25.2%)59 (22.3%)45 (25.7%)397 (21.3%)
       462 (21.1%)166 (19.9%)9 (14.5%)50 (21.0%)55 (20.8%)34 (19.4%)376 (20.1%)
       557 (19.4%)208 (24.9%)18 (29.0%)56 (23.5%)62 (23.4%)47 (26.9%)448 (24.0%)
      High Blood Pressure, n (%)46 (15.6%)126 (15.1%)13 (21.0%)25 (10.5%)55 (20.8%)48 (27.4%)313 (16.8%)<.001
      Type 2 Diabetes, n (%)5 (1.7%)9 (1.1%)3 (4.8%)4 (1.7%)14 (5.3%)11 (6.3%)46 (2.5%)<.001
      Cholesterol Drug Use, n (%)34 (11.6%)120 (14.4%)14 (22.6%)24 (10.1%)80 (30.2%)36 (20.6%)308 (16.5%)<.001
      LTL, Mean (SD)−0.62 (0.28)−0.62 (0.26)−0.61 (0.23)−0.62 (0.27)−0.61 (0.24)−0.65 (0.25)−0.62 (0.26).712
      Plus or minus signs indicate presence of each condition (e.g., T+P+D– indicates participants who were trauma exposed, met criteria for lifetime PTSD, and did not have probable depression). Kruskal-Wallis tests were used to compare means of continuous variables across groups and χ2 tests for categorical variables.
      BMI, body mass index; D, depression; LTL, leukocyte telomere length; P, posttraumatic stress disorder; PTSD, posttraumatic stress disorder; T, trauma.
      a “Other” includes participants identifying as American Indian, Asian, Black, Hawaiian, Multiracial, or Other/Unknown.
      In the model adjusting for age, race, and substudy, women with comorbid PTSD and depression had shorter LTL (β = −0.05; 95% CI, −0.10 to −0.01) compared with women with no trauma exposure, PTSD, or depression (see model 1 in Table 3). This association remained after further adjustment for health behaviors (model 2) or for additional health risk factors (model 3). This difference in LTL was equivalent to women with trauma exposure, PTSD, and depression being 7.62 years older (95% CI, 0.02 to 17.97) than women without trauma or depression. In E-value analyses, we calculated that an unmeasured confounder would have to be equally associated with the joint distress variable and LTL at a relative risk of 1.69 to account for the observed association. No significant interaction was found between age and any PTSD/depression category (see Table S1).
      Table 3Linear Association of Trauma, PTSD, and Depression With Leukocyte Telomere Length
      nModel 1 (95% CI)
      Adjusted for age, race, and substudy.
      Model 2 (95% CI)
      Model 1 further adjusted for body mass index, smoking, alcohol use, exercise, and diet.
      Model 3 (95% CI)
      Model 2 further adjusted for blood pressure, diabetes, and cholesterol drug use.
      No Trauma or Depression294ReferenceReferenceReference
      Trauma, No PTSD or Depression8340.00 (−0.03 to 0.03)0.00 (−0.03 to 0.03)0.00 (−0.04 to 0.03)
      Depression, No Trauma or PTSD62−0.01 (−0.08 to 0.06)−0.01 (−0.08 to 0.06)−0.01 (−0.08 to 0.06)
      Trauma and PTSD, No Depression2380.02 (−0.03 to 0.06)0.02 (−0.02 to 0.06)0.02 (−0.02 to 0.06)
      Trauma and Depression, No PTSD265−0.02 (−0.06 to 0.02)−0.02 (−0.06 to 0.03)−0.02 (−0.06 to 0.03)
      Trauma, PTSD, and Depression175−0.05 (−0.10 to −0.01)
      p < .05.
      −0.05 (−0.1 to 0.00)
      p < .05.
      −0.05 (−0.1 to 0.00)
      p < .05.
      All models mutually adjusted for trauma, depression, and PTSD status.
      PTSD, posttraumatic stress disorder.
      a Adjusted for age, race, and substudy.
      b Model 1 further adjusted for body mass index, smoking, alcohol use, exercise, and diet.
      c Model 2 further adjusted for blood pressure, diabetes, and cholesterol drug use.
      d p < .05.
      In analyses examining PTSD severity, participants with the highest number of PTSD symptoms (i.e., 6–7 symptoms) and depression had shorter LTL equivalent to 9.71 years of aging (95% CI, 1.37 to 20.49) (Figure 1) versus participants with no lifetime trauma, depression, or PTSD (Table 4). Trend analyses of PTSD symptom severity were not significant among either those with (p = .11) or without (p = .97) depression. In analyses examining depression severity, no significant association was found between those with higher levels of depression symptoms with or without comorbid PTSD or experience of trauma compared with those with the lowest level of depression severity and no experience of trauma or PTSD (see Table S2). Results from sensitivity analyses were consistent with those from the main results (see Figure S1 and Table S3). When considering effects of potentially underestimating prevalence of depression history, the association of comorbid PTSD and depression with LTL was consistent with the primary analysis (β = −0.04; 95% CI, −0.09 to 0.00). Similarly, after adjusting the experience of childhood physical, emotional, and sexual abuse, the association between comorbid PTSD and depression compared with no experience of trauma, depression, or PTSD remained consistent though not significant (β = −0.04; 95% CI, −0.10 to 0.01). Analyses excluding participants in the prior NHS II publication (
      • Roberts A.L.
      • Koenen K.C.
      • Chen Q.
      • Gilsanz P.
      • Mason S.M.
      • Prescott J.
      • et al.
      Posttraumatic stress disorder and accelerated aging: PTSD and leukocyte telomere length in a sample of civilian women.
      ), and stratified by substudy, also led to attenuation in the effect estimates, but the direction of effects was consistent with findings from the main analyses.
      Figure thumbnail gr1
      Figure 1Equivalent number of years of aging associated with trauma and posttraumatic stress disorder (PTSD) symptom severity, stratified by lifetime depression status. Years of aging were estimated by weighting the effect of each trauma, PTSD, and depression variable with the effect of age from a model fit without trauma, PTSD, or depression. A set of 1000 bootstraps was conducted to estimate the 95% confidence interval.
      Table 4Linear Association of the Joint Occurrences of Trauma, Depression, and PTSD Symptoms With Leukocyte Telomere Length
      nModel 1 (95% CI)
      Adjusted for age, race, and substudy,
      Model 2 (95% CI)
      Model 1 further adjusted for body mass index, smoking, alcohol use, exercise, and diet.
      Model 3 (95% CI)
      Model 2 further adjusted for blood pressure, diabetes, and cholesterol drug use.
      No Trauma or Depression294ReferenceReferenceReference
      Depression, No Trauma62−0.01 (−0.08 to 0.06)−0.01 (−0.08 to 0.06)−0.01 (−0.08 to 0.06)
      Trauma, No Depression, PTSD Sx = 04270.02 (−0.02 to 0.05)0.02 (−0.02 to 0.05)0.02 (−0.02 to 0.05)
      Trauma, No Depression, PTSD Sx = 1–3407−0.02 (−0.06 to 0.02)−0.02 (−0.06 to 0.02)−0.02 (−0.06 to 0.02)
      Trauma, No Depression, PTSD Sx = 4–51650.02 (−0.03 to 0.07)0.02 (−0.03 to 0.07)0.02 (−0.03 to 0.07)
      Trauma, No Depression, PTSD Sx = 6–7730.01 (−0.06 to 0.07)0.01 (−0.05 to 0.08)0.01 (−0.05 to 0.08)
      Trauma, Depression, PTSD Sx = 095−0.01 (−0.07 to 0.05)−0.01 (−0.07 to 0.05)−0.01 (−0.07 to 0.05)
      Trauma, Depression, PTSD Sx = 1–3170−0.02 (−0.07 to 0.02)−0.02 (−0.07 to 0.03)−0.02 (−0.07 to 0.03)
      Trauma, Depression, PTSD Sx = 4–597−0.04 (−0.10 to 0.02)−0.04 (−0.10 to 0.02)−0.04 (−0.10 to 0.02)
      Trauma, Depression, PTSD Sx = 6–778−0.07 (−0.13 to −0.01)
      p < .05.
      −0.07 (−0.13 to −0.00)
      p < .05.
      −0.06 (−0.13 to 0.00)
      p < .05.
      All models mutually adjusted for trauma, depression, and PTSD status.
      PTSD, posttraumatic stress disorder; Sx, symptoms.
      a Adjusted for age, race, and substudy,
      b Model 1 further adjusted for body mass index, smoking, alcohol use, exercise, and diet.
      c Model 2 further adjusted for blood pressure, diabetes, and cholesterol drug use.
      d p < .05.

      Discussion

      We report here that PTSD with comorbid depression is associated with shorter LTL, equivalent to 7.62 years of aging, compared with women with no trauma exposure or depression. When adjusted for the other conditions, trauma, PTSD, and depression were not associated with shorter LTL (Table 3). Prior studies have mixed findings regarding the association of PTSD and depression with LTL, with recent studies finding either no effect (
      • Vincent J.
      • Hovatta I.
      • Frissa S.
      • Goodwin L.
      • Hotopf M.
      • Hatch S.L.
      • et al.
      Assessing the contributions of childhood maltreatment subtypes and depression case-control status on telomere length reveals a specific role of physical neglect.
      ) or an effect only for a subgroup of participants [e.g., older participants (
      • Wang X.
      • Sundquist K.
      • Hedelius A.
      • Palmér K.
      • Memon A.A.
      • Sundquist J.
      Leukocyte telomere length and depression, anxiety and stress and adjustment disorders in primary health care patients.
      ) or males (
      • Whisman M.A.
      • Richardson E.D.
      Depressive symptoms and salivary telomere length in a probability sample of middle-aged and older adults.
      ,
      • Wang X.
      • Sundquist K.
      • Hedelius A.
      • Palmér K.
      • Memon A.A.
      • Sundquist J.
      Leukocyte telomere length and depression, anxiety and stress and adjustment disorders in primary health care patients.
      ) for depression; severely trauma-exposed participants for PTSD (
      • Kim T.Y.
      • Kim S.J.
      • Choi J.R.
      • Lee S.-T.
      • Kim J.
      • Hwang I.S.
      • et al.
      The effect of trauma and PTSD on telomere length: An exploratory study in people exposed to combat trauma.
      )]. Our results indicate that severe distress as indicated by comorbid PTSD and depression is associated with shorter LTL specifically among women with the highest number of PTSD symptoms (6–7 symptoms; equivalent years of aging = 9.71), but more severe depression symptoms even in the presence of PTSD was not associated with shorter LTL (Table S2). These results were consistent even after adjusting for the experience of childhood physical, emotional, and sexual abuse, which have been associated both with psychological distress and shorter LTL (
      • Li X.
      • Wang J.
      • Zhou J.
      • Huang P.
      • Li J.
      The association between post-traumatic stress disorder and shorter telomere length: A systematic review and meta-analysis.
      ,
      • Ridout K.K.
      • Ridout S.J.
      • Price L.H.
      • Sen S.
      • Tyrka A.R.
      Depression and telomere length: A meta-analysis.
      ,
      • Brewin C.R.
      • Andrews B.
      • Valentine J.D.
      Meta-analysis of risk factors for posttraumatic stress disorder in trauma-exposed adults.
      ,
      • Nanni V.
      • Uher R.
      • Danese A.
      Childhood maltreatment predicts unfavorable course of illness and treatment outcome in depression: A meta-analysis.
      ,
      • Ridout K.K.
      • Levandowski M.
      • Ridout S.J.
      • Gantz L.
      • Goonan K.
      • Palermo D.
      • et al.
      Early life adversity and telomere length: A meta-analysis.
      ). If only severely symptomatic individuals have shorter LTL, as PTSD and depression are highly comorbid (
      • Wolf E.J.
      • Morrison F.G.
      • Sullivan D.R.
      • Logue M.W.
      • Guetta R.E.
      • Stone A.
      • et al.
      The goddess who spins the thread of life: Klotho, psychiatric stress, and accelerated aging.
      ,
      • Connolly S.L.
      • Stoop T.B.
      • Logue M.W.
      • Orr E.H.
      • De Vivo I.
      • Miller M.W.
      • et al.
      Posttraumatic stress disorder symptoms, temperament, and the pathway to cellular senescence.
      ,
      • Monsees G.M.
      • Tamimi R.M.
      • Kraft P.
      Genome-wide association scans for secondary traits using case-control samples.
      ), previous results in the literature of PTSD and depression alone being associated with shorter LTL may be capturing the comorbidity implicitly when classifying cases by only one disorder. Furthermore, individuals with comorbid PTSD and depression may also represent those with the greatest severity for each disorder (
      • Rytwinski N.K.
      • Scur M.D.
      • Feeny N.C.
      • Youngstrom E.A.
      The co-occurrence of major depressive disorder among individuals with posttraumatic stress disorder: A meta-analysis.
      ,
      • Kendler K.S.
      • Gardner C.O.
      • Prescott C.A.
      Toward a comprehensive developmental model for major depression in women.
      ), which supports previous findings that depression severity is associated with shorter LTL (
      • Ridout K.K.
      • Ridout S.J.
      • Price L.H.
      • Sen S.
      • Tyrka A.R.
      Depression and telomere length: A meta-analysis.
      ).
      These results are subject to several limitations. First, the data are cross-sectional, which precludes testing direction of effect. It may be that shorter LTL leads to increased risk for PTSD or depression rather than the reverse, though prospective studies of PTSD and depression indicate that prior PTSD/depression is associated with subsequent shortened LTL (
      • Darrow S.M.
      • Verhoeven J.E.
      • Révész D.
      • Lindqvist D.
      • Penninx B.W.
      • Delucchi K.L.
      • et al.
      The association between psychiatric disorders and telomere length: A meta-analysis involving 14,827 persons.
      ,
      • Ridout K.K.
      • Ridout S.J.
      • Price L.H.
      • Sen S.
      • Tyrka A.R.
      Depression and telomere length: A meta-analysis.
      ). Second, LTL was assessed in multiple batches across several substudies. While we adjusted for substudy in our models, residual differences may remain. Third, PTSD was assessed retrospectively and incompletely for participants whose blood was drawn after 2008, while current and not lifetime depression was assessed for participants in first blood draw. In our sensitivity analysis of depression misclassification, we found an effect of comorbid PTSD and depression on LTL consistent with our main analysis, even with 9% of participants’ depression status misclassified. These results imply that the associations between comorbid PTSD and depression are robust to underreporting of depression. Fourth, while the NHS II sample size would be one of the largest studies of LTL and PTSD/depression conducted, some levels of the exposure had a relatively small number of women in them (e.g., lifetime depression, no trauma/PTSD group had n = 62). Fifth, while we examined increasing PTSD symptoms stratified across depression status (Table 4) and vice versa, our screening measures assessed only seven PTSD symptoms and half of the DSM-IV depression symptoms, which limited our ability to assess whether severity of PTSD or depression alone accounted for LTL shortening. Our use of screening measures also prevents us from examining the association of PTSD symptom dimensions with telomere length.
      Further research explicating the relationship between PTSD and depression with LTL would provide important insight into how PTSD and depression affect physical health. Prospective studies, which can directly address issues of temporality, should be conducted and assess all PTSD and depression symptoms to determine whether joint distress, severe symptoms of a single disorder, or a shared component is associated with shortened LTL. Untangling the association between PTSD/depression and LTL would aid in the identification of at-risk groups based on specific symptomology for poorer physical health outcomes with implications for both treatment and prevention.

      Acknowledgments and Disclosures

      This work was supported by National Institute of Mental Health Grant No. R01 MH101269 (to KCK, LDK) and Department of Defense Grant No. W81XWH-17-1-0153 (to LDK) . The NHS II is funded in part by National Institutes of Health Grant Nos. UM1 CA176726, R01 CA163451, and R01 CA67262. KWC was supported in part by a NARSAD Brain and Behavior Foundation Young Investigator Award.
      We acknowledge the Channing Division of Network Medicine, Department of Medicine, Brigham and Women's Hospital and Harvard Medical School for its management of the Nurses' Health Study II. We thank the participants and staff of the Nurses' Health Study II for their valuable contributions.
      The authors report no biomedical financial interests or potential conflicts of interest.

      References

        • Mathur M.B.
        • Epel E.
        • Kind S.
        • Desai M.
        • Parks C.G.
        • Sandler D.P.
        • et al.
        Perceived stress and telomere length: A systematic review, meta-analysis, and methodologic considerations for advancing the field.
        Brain Behav Immun. 2016; 54: 158-169
        • Willis M.
        • Reid S.N.
        • Calvo E.
        • Staudinger U.M.
        • Factor-Litvak P.
        A scoping systematic review of social stressors and various measures of telomere length across the life course.
        Ageing Res Rev. 2018; 47: 89-104
        • Moyzis R.K.
        • Buckingham J.M.
        • Cram L.S.
        • Dani M.
        • Deaven L.L.
        • Jones M.D.
        • et al.
        A highly conserved repetitive DNA sequence, (TTAGGG) n, present at the telomeres of human chromosomes.
        Proc Natl Acad Sci U S A. 1988; 85: 6622-6626
        • Valdes A.M.
        • Andrew T.
        • Gardner J.P.
        • Kimura M.
        • Oelsner E.
        • Cherkas L.F.
        • et al.
        Obesity, cigarette smoking, and telomere length in women.
        Lancet. 2005; 366: 662-664
        • Wong L.M.
        • van der Harst P.
        • de Boer R.
        • Huzen J.
        • van Gilst W.
        • van Veldhuisen D.
        Aging, telomeres and heart failure.
        Heart Fail Rev. 2010; 15: 479-486
        • Le Carolyn M.H.
        • Neylan T.C.
        • Na B.
        • Regan M.
        • Zhang Q.
        • Cohen B.E.
        Lifetime trauma exposure and prospective cardiovascular events and all-cause mortality: Findings from the Heart and Soul Study.
        Psychosom Med. 2013; 75: 849-855
        • Edmondson D.
        • Kronish I.M.
        • Shaffer J.A.
        • Falzon L.
        • Burg M.M.
        Posttraumatic stress disorder and risk for coronary heart disease: A meta-analytic review.
        Am Heart J. 2013; 166: 806-814
        • Van der Kooy K.
        • Van Hout H.
        • Marwijk H.
        • Marten H.
        • Stehouwer C.
        • Beekman A.
        Depression and the risk for cardiovascular diseases: Systematic review and meta analysis.
        Int J Geriatr Psychiatry. 2007; 22: 613-626
        • Knol M.
        • Twisk J.W.
        • Beekman A.T.
        • Heine R.
        • Snoek F.J.
        • Pouwer F.
        Depression as a risk factor for the onset of type 2 diabetes mellitus. A meta-analysis.
        Diabetologia. 2006; 49: 837
        • Mezuk B.
        • Eaton W.W.
        • Albrecht S.
        • Golden S.H.
        Depression and type 2 diabetes over the lifespan: A meta-analysis.
        Diabetes Care. 2008; 31: 2383-2390
        • Vancampfort D.
        • Rosenbaum S.
        • Ward P.B.
        • Steel Z.
        • Lederman O.
        • Lamwaka A.V.
        • et al.
        Type 2 diabetes among people with posttraumatic stress disorder: Systematic review and meta-analysis.
        Psychosom Med. 2016; 78: 465-473
        • Zhao J.
        • Miao K.
        • Wang H.
        • Ding H.
        • Wang D.W.
        Association between telomere length and type 2 diabetes mellitus: A meta-analysis.
        PLoS One. 2013; 8e79993
        • Chwastiak L.A.
        • Rosenheck R.A.
        • Desai R.
        • Kazis L.E.
        The association of psychiatric illness and all-cause mortality in the national department of Veterans Affairs Health Care System.
        Psychosom Med. 2010; 72: 817-822
        • Cuijpers P.
        • Vogelzangs N.
        • Twisk J.
        • Kleiboer A.
        • Li J.
        • Penninx B.W.
        Comprehensive meta-analysis of excess mortality in depression in the general community versus patients with specific illnesses.
        Am J Psychiatry. 2014; 171: 453-462
        • Aviv A.
        Telomeres and human somatic fitness.
        J Gerontol A Biol Sci Med Sci. 2006; 61: 871-873
        • Darrow S.M.
        • Verhoeven J.E.
        • Révész D.
        • Lindqvist D.
        • Penninx B.W.
        • Delucchi K.L.
        • et al.
        The association between psychiatric disorders and telomere length: A meta-analysis involving 14,827 persons.
        Psychosom Med. 2016; 78: 776
        • Li X.
        • Wang J.
        • Zhou J.
        • Huang P.
        • Li J.
        The association between post-traumatic stress disorder and shorter telomere length: A systematic review and meta-analysis.
        J Affect Disord. 2017; 218: 322-326
        • Ridout K.K.
        • Ridout S.J.
        • Price L.H.
        • Sen S.
        • Tyrka A.R.
        Depression and telomere length: A meta-analysis.
        J Affect Disord. 2016; 191: 237-247
        • Boks M.P.
        • van Mierlo H.C.
        • Rutten B.P.
        • Radstake T.R.
        • De Witte L.
        • Geuze E.
        • et al.
        Longitudinal changes of telomere length and epigenetic age related to traumatic stress and post-traumatic stress disorder.
        Psychoneuroendocrinology. 2015; 51: 506-512
        • Wolf E.J.
        • Morrison F.G.
        • Sullivan D.R.
        • Logue M.W.
        • Guetta R.E.
        • Stone A.
        • et al.
        The goddess who spins the thread of life: Klotho, psychiatric stress, and accelerated aging.
        Brain Behav Immun. 2019; 80: 193-203
        • Connolly S.L.
        • Stoop T.B.
        • Logue M.W.
        • Orr E.H.
        • De Vivo I.
        • Miller M.W.
        • et al.
        Posttraumatic stress disorder symptoms, temperament, and the pathway to cellular senescence.
        J Trauma Stress. 2018; 31: 676-686
        • Rytwinski N.K.
        • Scur M.D.
        • Feeny N.C.
        • Youngstrom E.A.
        The co-occurrence of major depressive disorder among individuals with posttraumatic stress disorder: A meta-analysis.
        J Trauma Stress. 2013; 26: 299-309
        • Gros D.F.
        • Price M.
        • Magruder K.M.
        • Frueh B.C.
        Symptom overlap in posttraumatic stress disorder and major depression.
        Psychiatry Res. 2012; 196: 267-270
        • Shalev A.Y.
        • Freedman S.
        • Peri T.
        • Brandes D.
        • Sahar T.
        • Orr S.P.
        • et al.
        Prospective study of posttraumatic stress disorder and depression following trauma.
        Am J Psychiatry. 1998; 155: 630-637
        • Roberts A.L.
        • Koenen K.C.
        • Chen Q.
        • Gilsanz P.
        • Mason S.M.
        • Prescott J.
        • et al.
        Posttraumatic stress disorder and accelerated aging: PTSD and leukocyte telomere length in a sample of civilian women.
        Depress Anxiety. 2017; 34: 391-400
        • Brewin C.R.
        • Andrews B.
        • Valentine J.D.
        Meta-analysis of risk factors for posttraumatic stress disorder in trauma-exposed adults.
        J Consult Clin Psychol. 2000; 68: 748-766
        • Nanni V.
        • Uher R.
        • Danese A.
        Childhood maltreatment predicts unfavorable course of illness and treatment outcome in depression: A meta-analysis.
        Am J Psychiatry. 2012; 169: 141-151
        • Ridout K.K.
        • Levandowski M.
        • Ridout S.J.
        • Gantz L.
        • Goonan K.
        • Palermo D.
        • et al.
        Early life adversity and telomere length: A meta-analysis.
        Mol Psychiatry. 2018; 23: 858-871
        • Wium-Andersen M.K.
        • Ørsted D.D.
        • Rode L.
        • Bojesen S.E.
        • Nordestgaard B.G.
        Telomere length and depression: Prospective cohort study and Mendelian randomisation study in 67 306 individuals.
        Br J Psychiatry. 2017; 210: 31-38
        • Wignall N.D.
        • Deschner M.
        • Evans H.M.
        • Brown E.S.
        Relationship between current depressive symptoms and telomere length in a large, multiethnic sample.
        J Clin Psychiatry. 2017; 78: 1331-1336
        • Liu J.J.
        • Wei Y.B.
        • Forsell Y.
        • Lavebratt C.
        Stress, depressive status and telomere length: Does social interaction and coping strategy play a mediating role?.
        J Affect Disord. 2017; 222: 138-145
        • Whisman M.A.
        • Richardson E.D.
        Depressive symptoms and salivary telomere length in a probability sample of middle-aged and older adults.
        Psychosom Med. 2017; 79: 234-242
        • Sumner J.A.
        • Chen Q.
        • Roberts A.L.
        • Winning A.
        • Rimm E.B.
        • Gilsanz P.
        • et al.
        Cross-sectional and longitudinal associations of chronic posttraumatic stress disorder with inflammatory and endothelial function markers in women.
        Biol Psychiatry. 2017; 82: 875-884
        • Sumner J.A.
        • Chen Q.
        • Roberts A.L.
        • Winning A.
        • Rimm E.B.
        • Gilsanz P.
        • et al.
        Posttraumatic stress disorder onset and inflammatory and endothelial function biomarkers in women.
        Brain Behav Immun. 2018; 69: 203-209
        • Koenen K.C.
        • De Vivo I.
        • Rich-Edwards J.
        • Smoller J.W.
        • Wright R.J.
        • Purcell S.M.
        Protocol for investigating genetic determinants of posttraumatic stress disorder in women from the Nurses' Health Study II.
        BMC Psychiatry. 2009; 9: 29
        • Mason S.M.
        • Prescott J.
        • Tworoger S.S.
        • DeVivo I.
        • Rich-Edwards J.W.
        Childhood physical and sexual abuse history and leukocyte telomere length among women in middle adulthood.
        PLoS One. 2015; 10e0124493
        • Spence N.D.
        • Farvid M.S.
        • Warner E.T.
        • VanderWeele T.J.
        • Tworoger S.S.
        • Argentieri M.A.
        • et al.
        Religious service attendance, religious coping, and risk of hypertension in women participating in the Nurses’ Health Study II.
        Am J Epidemiol. 2020; 189: 193-203
        • Huang T.
        • Trudel-Fitzgerald C.
        • Poole E.M.
        • Sawyer S.
        • Kubzansky L.D.
        • Hankinson S.E.
        • et al.
        The Mind–Body Study: Study design and reproducibility and interrelationships of psychosocial factors in the Nurses’ Health Study II.
        Cancer Causes Control. 2019; 30: 779-790
        • Roberts A.L.
        • Dohrenwend B.P.
        • Aiello A.
        • Wright R.J.
        • Maercker A.
        • Galea S.
        • et al.
        The stressor criterion for posttraumatic stress disorder: Does it matter?.
        J Clin Psychiatry. 2012; 73: 264-270
        • Breslau N.
        • Peterson E.L.
        • Kessler R.C.
        • Schultz L.R.
        Short screening scale for DSM-IV posttraumatic stress disorder.
        Am J Psychiatry. 1999; 156: 908-911
        • Lee Y.C.
        • Agnew-Blais J.
        • Malspeis S.
        • Keyes K.
        • Costenbader K.
        • Kubzansky L.D.
        • et al.
        Post-traumatic stress disorder and risk for incident rheumatoid arthritis.
        Arthritis Care Res (Hoboken). 2016; 68: 292-298
        • McCabe C.
        • Thomas K.
        • Brazier J.
        • Coleman P.
        Measuring the mental health status of a population: A comparison of the GHQ–12 and the SF–36 (MHI–5).
        Br J Psychiatry. 1996; 169: 517-521
        • Irwin M.
        • Artin K.H.
        • Oxman M.N.
        Screening for depression in the older adult: Criterion validity of the 10-item Center for Epidemiological Studies Depression Scale (CES-D).
        Arch Int Med. 1999; 159: 1701-1704
        • McGrath M.
        • Wong J.Y.
        • Michaud D.
        • Hunter D.J.
        • De Vivo I.
        Telomere length, cigarette smoking, and bladder cancer risk in men and women.
        Cancer Epidemiol Biomarkers Prev. 2007; 16: 815-819
        • Rosner B.
        • Cook N.
        • Portman R.
        • Daniels S.
        • Falkner B.
        Determination of blood pressure percentiles in normal-weight children: Some methodological issues.
        Am J Epidemiol. 2008; 167: 653-666
        • Rimm E.B.
        • Stampfer M.J.
        • Colditz G.A.
        • Chute C.G.
        • Litin L.B.
        • Willett W.C.
        Validity of self-reported waist and hip circumferences in men and women.
        Epidemiology. 1990; 1: 466-473
        • Willett W.
        • Sampson L.
        • Browne M.
        • Stampfer M.J.
        • Rosner B.
        • Hennekens C.H.
        • et al.
        The use of a self-administered questionnaire to assess diet four years in the past.
        Am J Epidemiol. 1988; 127: 188-199
        • McCullough M.L.
        • Feskanich D.
        • Stampfer M.J.
        • Giovannucci E.L.
        • Rimm E.B.
        • Hu F.B.
        • et al.
        Diet quality and major chronic disease risk in men and women: Moving toward improved dietary guidance.
        Am J Clin Nutr. 2002; 76: 1261-1271
        • Bernstein D.P.
        • Fink L.
        • Handelsman L.
        • Foote J.
        • Lovejoy M.
        • Wenzel K.
        • et al.
        Initial reliability and validity of a new retrospective measure of child abuse and neglect.
        Am J Psychiatry. 1994; 151: 1132-1136
        • Roberts A.L.
        • Chen Y.
        • Slopen N.
        • McLaughlin K.A.
        • Koenen K.C.
        • Austin S.B.
        Maternal experience of abuse in childhood and depressive symptoms in adolescent and adult offspring: A 21-year longitudinal study.
        Depress Anxiety. 2015; 32: 709-719
        • Straus M.A.
        • Hamby S.L.
        • Finkelhor D.
        • Moore D.W.
        • Runyan D.
        Identification of child maltreatment with the Parent-Child Conflict Tactics Scales: Development and psychometric data for a national sample of American parents.
        Child Abuse Neglect. 1998; 22: 249-270
        • Monroe D.M.
        • Goldstein R.L.
        • Teylan M.A.
        • Hart J.E.
        • DeVivo I.
        • Orr E.H.
        • et al.
        Clinical associations with telomere length in chronic spinal cord injury.
        Spinal Cord. 2019; 57: 1084-1093
        • VanderWeele T.J.
        • Ding P.
        Sensitivity analysis in observational research: introducing the E-value.
        Ann Intern Med. 2017; 167: 268-274
        • Kessler R.C.
        • Berglund P.
        • Demler O.
        • Jin R.
        • Merikangas K.R.
        • Walters E.E.
        Lifetime prevalence and age-of-onset distributions of DSM-IV disorders in the National Comorbidity Survey Replication.
        Arch Gen Psychiatry. 2005; 62: 593-602
        • Norman R.E.
        • Byambaa M.
        • De R.
        • Butchart A.
        • Scott J.
        • Vos T.
        The long-term health consequences of child physical abuse, emotional abuse, and neglect: A systematic review and meta-analysis.
        PLoS Med. 2012; 9e1001349
        • Sumner J.A.
        • Kubzansky L.D.
        • Elkind M.S.
        • Roberts A.L.
        • Agnew-Blais J.
        • Chen Q.
        • et al.
        Trauma exposure and posttraumatic stress disorder symptoms predict onset of cardiovascular events in women.
        Circulation. 2015; 132: 251-259
        • Weischer M.
        • Bojesen S.E.
        • Cawthon R.M.
        • Freiberg J.J.
        • Tybjærg-Hansen A.
        • Nordestgaard B.G.
        Short telomere length, myocardial infarction, ischemic heart disease, and early death.
        Arterioscler Thromb Vasc Biol. 2012; 32: 822-829
        • Vincent J.
        • Hovatta I.
        • Frissa S.
        • Goodwin L.
        • Hotopf M.
        • Hatch S.L.
        • et al.
        Assessing the contributions of childhood maltreatment subtypes and depression case-control status on telomere length reveals a specific role of physical neglect.
        J Affect Disord. 2017; 213: 16-22
        • Wang X.
        • Sundquist K.
        • Hedelius A.
        • Palmér K.
        • Memon A.A.
        • Sundquist J.
        Leukocyte telomere length and depression, anxiety and stress and adjustment disorders in primary health care patients.
        BMC Psychiatry. 2017; 17: 148
        • Kim T.Y.
        • Kim S.J.
        • Choi J.R.
        • Lee S.-T.
        • Kim J.
        • Hwang I.S.
        • et al.
        The effect of trauma and PTSD on telomere length: An exploratory study in people exposed to combat trauma.
        Sci Rep. 2017; 7: 4375
        • Monsees G.M.
        • Tamimi R.M.
        • Kraft P.
        Genome-wide association scans for secondary traits using case-control samples.
        Genet Epidemiol. 2009; 33: 717-728
        • Kendler K.S.
        • Gardner C.O.
        • Prescott C.A.
        Toward a comprehensive developmental model for major depression in women.
        Focus. 2005; 159: 1133-1197