Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2023 Mar 1.
Published in final edited form as: Psychol Trauma. 2022 Jul 11;15(1):131–139. doi: 10.1037/tra0001304

The Unique Association of Posttraumatic Stress Disorder with Hypertension Among Veterans: A Replication of Kibler et al. (2009) Using Bayesian Estimation and Data from the United States-Veteran Microbiome Project

Daniel J Reis 1,2, Alexander M Kaizer 3, Adam R Kinney 1,4, Nazanin H Bahraini 1,2,4, Jeri E Forster 1,3,4, Lisa A Brenner 1,2,4,5
PMCID: PMC9976482  NIHMSID: NIHMS1877170  PMID: 35816586

Abstract

Objective:

In 2009, Kibler et al. reported that hypertension was related to posttraumatic stress disorder (PTSD) independent of depression. These two conditions have significant diagnostic overlap. The present study sought to conceptually replicate this work with a veteran sample, utilizing Bayesian estimation to directly update past results, as well as examine symptom severity scores in relation to hypertension.

Method:

This was a secondary analysis of data obtained from the United States-Veteran Microbiome Project. Lifetime diagnoses of PTSD and major depressive disorder (MDD) were obtained from a structured clinical interview and hypertension diagnoses were extracted from electronic medical records. PTSD and depressive symptom severity were obtained from self-report measures. Logistic regressions with Bayesian estimation were used to estimate the associations between hypertension and: 1) psychiatric diagnostic history; and, 2) symptom severity scores.

Results:

Compared to veterans without lifetime diagnoses of either disorder, the PTSD-only group was estimated to have a 29% increase in hypertension risk, and the PTSD+MDD group was estimated to have a 66% increase in hypertension risk. Additionally, higher levels of PTSD symptom severity were associated with a higher risk of hypertension.

Conclusion:

PTSD diagnosis and symptom severity are uniquely associated with hypertension, independent of MDD or depressive symptom severity. These results support previous findings that PTSD may be a modifiable risk factor for the prevention and treatment of hypertension.

Keywords: hypertension, posttraumatic stress disorder, major depressive disorder, Bayesian estimation, relative risk

Military service is associated with high rates of posttraumatic stress disorder (PTSD; Yehuda et al., 2015). Recently, combat operations in support of Operations Enduring Freedom, Iraqi Freedom, and New Dawn have been found to be associated with increased PTSD incidence among military veterans (Cameron et al., 2019). Furthermore, among veterans seeking care within the Veterans Health Administration (VHA), PTSD has been identified as one of the most prevalent forms of mental illness (Trivedi et al., 2015; Zulman et al., 2015), suggesting that trauma-related health needs will remain a VHA priority for the foreseeable future. Considerable scientific investment has been devoted to understanding the psychosocial consequences of PTSD, such as its link to disrupted relationships (Campbell & Renshaw, 2018) and death by suicide (Holliday et al., 2020). In addition, emergent research regarding the mind-body connection—the relationship between mental and physical health—is uncovering the contribution that either mental health conditions like PTSD, or shared underlying mechanisms (e.g., inflammation), make to the development and maintenance of chronic and debilitating physical health conditions (e.g., heart disease; Minhas et al., 2021). Co-morbid physical and mental illness may pose a substantial burden for both patients and the VHA system, as such patterns of comorbidity are associated with high levels of VHA utilization and costs (Yoon et al., 2014). Therefore, it is critical to understand the physical sequelae that may arise among those with PTSD, the management of which may have long-term physical and mental health benefits.

One physical health condition linked to PTSD is hypertension (i.e., high blood pressure). Hypertension is highly prevalent in the general VHA veteran population (Zulman et al., 2015) and strongly associated with adverse outcomes such as heart failure (Messerli et al., 2017). PTSD is associated with dysregulation of multiple physiological pathways implicated in hypertension onset, including inflammation, the hypothalamic-pituitary-adrenal (HPA) axis, and the sympathetic nervous system (Burford et al., 2017; Fisher & Paton, 2012; Fonkoue et al., 2020; Sherin & Nemeroff, 2011). For example, recurring cardiovascular perturbations in response to traumatic reminders, such as elevated heart rate and blood pressure, are common among those with PTSD (Sherin & Nemeroff, 2011).

While previous research has found a positive association between PTSD and hypertension (Burg & Soufer, 2016; Kibler et al., 2009; Stevelink et al., 2020), other studies have found equivocal results (Nichter et al., 2019; Tsai & Shen, 2017). Notably, only a few of the studies controlled for depression, which may help explain the discrepant findings. Depression is also associated with hypertension (Burnette et al., 2020; Nichter et al., 2019; Stevelink et al., 2020) and, like PTSD, has been theoretically implicated in the pathogenesis of hypertension through disruption of processes such as inflammation and HPA-axis activity (Peters et al., 2004; Yamagata et al., 2017). Importantly, PTSD and depressive disorders are highly comorbid (Trivedi et al., 2015) and have considerable diagnostic overlap (Barbano et al., 2019), which may obscure the unique contribution of either disorder to hypertension risk. Because PTSD and depressive disorders are addressed using different treatments (U.S. Department of Veterans Affairs & U.S. Department of Defense, 2016, 2017), understanding these individual contributions, if present, could have a significant impact on how these modifiable risk factors for hypertension are approached in clinical practice. This is particularly relevant within the VHA, which is charged with meeting the needs of a growing trauma-exposed patient population (Cameron et al., 2019). Moreover, within the VHA, holistic care and preventative practices have been prioritized (U.S. Department of Veterans Affairs, 2019). Nonetheless, mental health conditions are not addressed in guidelines for hypertension management (U.S. Department of Veterans Affairs & U.S. Department of Defense, 2020).

Recently, Kibler et al. (2009) performed the most direct and rigorous evaluation of the unique associations that PTSD and major depressive disorder (MDD) have with hypertension. Using structured interview data from over 3,000 adults, the authors assessed whether lifetime diagnoses of PTSD or MDD were related to self-reported hypertension. They found that a history of 1) both PTSD and MDD or 2) PTSD alone was associated with increased odds of self-reported hypertension (Kibler et al., 2009). However, this study made use of a sample from the general population. Given the unique needs of veterans receiving VHA care (Yoon et al., 2014), it is important to assess whether such findings generalize to veterans. Additionally, Kibler et al. (2009) noted as a study limitation the inability to assess continuous symptom measures in addition to dichotomous diagnoses, which may have overshadowed sub-threshold diagnostic presentations. Accordingly, in the present study the team sought to conceptually replicate and extend the Kibler et al. (2009) findings within a veteran sample. This was conducted using a Bayesian analytical approach with both interview-based mental health diagnoses and self-reported symptom severity scores. Bayesian estimation is a natural fit for replication research, as it provides a framework for directly assessing and updating past findings in the light of novel data (Wagenmakers et al., 2018). Additionally, the use of Bayesian priors that are appropriately informed by past research can enhance estimation precision (Morris et al., 2015), further clarifying the relationships under study.

Method

Participants

This was a secondary analysis of data obtained from the United States-Veteran Microbiome Project, which has been previously described (Brenner et al., 2018). In brief, all local U.S. veterans are eligible to enroll in this ongoing project, in which data are collected through structured clinical interviews, self-report measures, and electronic medical records. The study began in May 2016 (Brenner et al., 2018). For the present analyses, all veterans who had completed the baseline assessment and had data available through the VHA electronic medical record were included (see the Results section for more sample details). No restrictions (e.g., combat-related only) were placed on the type of trauma required for PTSD diagnosis (see below for details on diagnostic process). Additionally, veterans who met criteria for psychiatric diagnoses other than PTSD or MDD were not excluded. This study was conducted according to guidelines laid down in the Declaration of Helsinki and all procedures involving human participants were approved by a local institutional review board. All veterans provided informed consent prior to engaging in study activities.

Study Variables

PTSD and Depression

Lifetime diagnoses of PTSD and MDD were established using the Structured Clinical Interview for DSM-5 Disorders Research Version (SCID-5), which is a well-accepted method for assessing psychiatric conditions in research studies (First et al., 2015). This is equivalent to the approach used by Kibler et al. (2009), who also used lifetime diagnostic data derived from a structured clinical interview. PTSD symptom severity was measured using the PTSD Checklist for DSM-5 (PCL-5; Blevins et al., 2015) and depressive symptom severity was measured using the Beck Depression Inventory-II (BDI-II; Beck et al., 1996), both of which are validated and reliable self-report symptom scales (Beck et al., 1996; Blevins et al., 2015). The PCL-5 consists of 20 items and patients are asked to rate how much a symptom has bothered them in the past month ranging from zero (“Not at all”) to four (“Extremely”; Blevins et al., 2015). Total severity scores range from 0–80, with higher scores reflective of more severe PTSD symptoms, and a cut score of 31–33 is typically used to indicate clinically elevated symptomatology (U.S. Department of Veterans Affairs, 2021). For this study, veterans self-reported on symptoms related to highly stressful experiences; however, the measure did not confirm that such experiences were traumatic in nature. The BDI-II consists of 21 items and patients are asked to rate how much a symptom has bothered them in the past two weeks, with total scores ranging from 0–63 and higher scores reflective of more severe depressive symptoms (Beck et al., 1996). Initial validation of the measure categorized total scores from 14–19 as mild depression, 20–28 as moderate depression, and 29–63 as severe depression (Beck et al., 1996), although it should be noted that the BDI-II has been found to overestimate depressive symptom severity in medical populations and veterans (Brown et al., 2012; Reis et al., 2020).

Hypertension

Information on hypertension was extracted from VHA electronic medical records via the Corporate Data Warehouse, a national repository containing medical data from the past several decades. For each veteran included in this study, all available medical records for the 20 years prior to study enrollment and baseline assessment were considered. A veteran was coded as having a history of hypertension if they met at least one of the following conditions: 1) they had an ICD-9 or ICD-10 code indicating a primary diagnosis of hypertension that was associated with an inpatient hospital encounter or at least two outpatient encounters within a one-year span, with the two outpatient encounters separated by at least 30 days; or 2) they had at least two blood pressure measurements, within a one-year span, in which systolic pressure was equal to 140 or higher and diastolic pressure was equal to 90 or higher. This algorithm was designed to minimize “rule-out” or erroneous diagnoses and was based on previous research extracting hypertension and other conditions from medical records (Klabunde et al., 2000) and the Department of Veterans Affairs Centralized Interactive Phenomics Resource, an internal platform for sharing algorithms based on electronic medical record data.

Covariates

Study covariates, used in all analyses, included sex, age, body mass index (BMI), race, ethnicity, marital status, education history, smoking history, history of alcohol use disorder, history of substance use disorder, service era, history of combat zone deployment, and number of unique deployments. Smoking status (coded as never smoked, smoked in the past, or currently smoking) was extracted from medical record health factors (McGinnis et al., 2011). BMI was also extracted from medical records—the BMI recorded closest to study baseline was captured for the present analyses. Lifetime diagnoses of alcohol and substance use disorder were established using the SCID-5. Finally, information for all other covariates was collected using a custom demographics questionnaire.

Statistical Analyses

Analyses were performed using R (R Core Team, 2021) and the “brms” package, which uses a no-U-turn Hamiltonian Monte Carlo sampler for Bayesian estimation (Bürkner, 2017). Two logistic regressions were conducted to assess the association between hypertension history and PTSD- and MDD-related variables, both using hypertension as the dependent variable (DV). The first analysis used psychiatric diagnostic status as independent variables (IVs) and the second used continuous symptom scores. Each model used four chains with 1,000 warm-up iterations and another 1,000 iterations that contributed to the posterior, for a total of 4,000 final iterations per model, which are the default “brms” settings. Model convergence was assessed using visual inspection of trace plots, the R^ statistic, effective sample size (ESS), and a graphical posterior predictive check of model-generated data compared to observed data. The threshold for an appropriate R^ was set to 1.01 and the thresholds for bulk- and tail-ESS were each set to 400, in accordance with recommendations by Vehtari et al. (2021). As noted below for each respective model (see Results), the default settings were sufficient to achieve adequate convergence. Additionally, results for each model were essentially identical after doubling the number of warm-up and final iterations (data not shown), further indicating that convergence was achieved.

Lifetime diagnoses of PTSD, MDD, and their interaction were the primary IVs in the first analysis. To build upon the work by Kibler et al. (2009), each IV was given an informative prior based on this previous study. Each IV was normally distributed and centered on the log odds of the corresponding Kibler et al. (2009) odds ratio estimate as reported in the first table of said paper. The approximate standard deviation for each reported estimate, determined by assuming that the 95% confidence interval was normally distributed for each reported odds ratio, was chosen as the standard deviation for the prior distribution. As such, the final priors for the first analysis were as follows: the PTSD prior was distributed as Normal (0.940,0.212); the MDD prior was distributed as Normal (0.445,0.147); and the interaction prior was distributed as Normal (0.952,0.220). Because Kibler et al. (2009) did not provide an adjusted prevalence rate for hypertension without PTSD or MDD, the intercept was given a weakly informative prior distributed as Normal (0,1). This distribution was selected because it permits a broad range of possible odds ratios—approximately 95% of prior probabilities fall between an odds ratio of 0.14 and 7.39 (and approximately 50% between 0.51 and 1.96)—with more extreme values possible but unlikely (Figure S1). In other words, the hyperparameters of this distribution broadly cover the range of expected odds ratios without placing significant probability on any specific value (i.e., weakly informative). Additionally, although asymmetrical, this distribution places most of the probability around an odds ratio of 1 (i.e., no effect), requiring the data to provide sufficient evidence otherwise.

A series of sensitivity analyses, altering the means and standard deviations of the informative priors, were conducted to evaluate the impact of the informative priors on findings. Sensitivity analysis models are described in Table S1. For additional information, models were also compared using leave-one-out-cross-validation to examine relative model fits (Sivula et al., 2020).

The second analysis used continuous symptom severity scores from baseline study assessment as IVs in place of the dichotomous diagnoses. Standardized scores on the PCL-5, BDI-II, and their interaction, as well as the intercept, were given weakly informative priors distributed as Normal (0,1) for the above-mentioned reasons. For both analyses, continuous covariates were standardized, and all covariates were also given weakly informative priors distributed as Normal (0,1).

Given that relative risk is a more intuitive effect measure than an odds ratio, model results were then used to estimate adjusted hypertension risks for separate diagnostic or symptom severity groups. Marginal hypertension probabilities, weighted for observed covariates as recommended by Muller and MacLehose (2014), were estimated by generating four new datasets for each analysis that contained all of the observed covariate data for the study veterans. The primary IVs for each dataset were changed to correspond to the targeted group. For the first analysis, the four groups/datasets were based on diagnostic status and were “Neither diagnosis,” “PTSD-only,” “MDD-only,” and “PTSD+MDD.” Note that the “Neither diagnosis” group could include veterans without any history of mental illness as well as those with a history of psychiatric diagnoses other than PTSD or MDD. For the second analysis, hypothetical groups were created corresponding to “high” and “low” PCL-5 and BDI-II symptom severity scores (i.e., one standard deviation above or below the mean of the symptom measure), with the four groups being “Low PCL-5 / Low BDI-II,” “High PCL-5 / Low BDI-II,” “High BDI-II / Low PCL-5,” and “High BDI-II / High PCL-5.” This was done to further explore any potential interaction between the two measures. Then, for every iteration of the respective model estimation, hypertension probabilities were generated for each veteran in each of the four new datasets and these probabilities were averaged to obtain the mean probability for the iteration. This resulted in posterior distributions of estimated hypertension risks for the four target groups and relative risk distributions were obtained by dividing the appropriate risk distributions. Means and equal tail 95% credible intervals were then summarized for each estimated risk and relative risk posterior distribution.

Results

Sample characteristics are presented in Table 1. In total, data on PTSD, MDD, and hypertension diagnoses, as well as covariates, were available from 374 veterans, of whom 15.8% had a history of PTSD but not MDD, 17.4% had a history of MDD but not PTSD, and 31.8% had a history of both PTSD and MDD. The average sample age was 47 years and 39% of veterans had served during Operations Enduring Freedom, Iraqi Freedom, and New Dawn. Only two veterans endorsed smoking in the past (but not presently), so the smoking status variable was collapsed prior to analyses into a binary variable with two levels: positive or negative smoking history. Similarly, marital status was also collapsed into “married” or “not married”. Due to missing PCL-5 and BDI-II data, the second analysis (using symptom score as IVs) included data from 337 veterans.

Table 1.

Sample Characteristics

No PTSD or MDD (N = 131) PTSD (N = 59) MDD (N = 65) PTSD+MDD (N = 119) Total (N = 374)
Hypertension 58 (44.3) 24 (40.7) 32 (49.2) 54 (45.4) 168 (44.9)
Lifetime Alcohol Use Disorder 50 (38.2) 32 (54.2) 43 (66.2) 76 (63.9) 201 (53.7)
Lifetime Substance Use Disorder 44 (33.6) 16 (27.1) 25 (38.5) 37 (31.1) 122 (32.6)
Identify as Hispanic/Latino(a) 18 (13.7) 12 (20.3) 7 (10.8) 22 (18.5) 59 (15.8)
Combat Zone Deployment 53 (40.5) 41 (69.5) 24 (36.9) 72 (60.5) 190 (50.8)
Service Era: OEF/OIF/OND 46 (35.1) 33 (55.9) 25 (38.5) 42 (35.3) 146 (39.0)
Smoking History
 Never 29 (22.1) 11 (18.6) 9 (13.8) 25 (21.0) 74 (19.8)
 Former 1 (0.8) 1 (1.7) 0 (0) 0 (0) 2 (0.5)
 Current 101 (77.1) 47 (79.7) 56 (86.2) 94 (79.0) 298 (79.7)
Sex
 Male 115 (87.8) 50 (84.7) 59 (90.8) 83 (69.7) 307 (82.1)
 Female 15 (11.5) 9 (15.3) 6 (9.2) 36 (30.3) 66 (17.6)
 Intersex 1 (0.8) 0 (0) 0 (0) 0 (0) 1 (0.3)
Racial Background
 Caucasian/White 82 (62.6) 39 (66.1) 44 (67.7) 90 (75.6) 255 (68.2)
 Black or African American 31 (23.7) 10 (16.9) 11 (16.9) 12 (10.1) 64 (17.1)
 Native American/Alaskan Native 1 (0.8) 1 (1.7) 0 (0) 4 (3.4) 6 (1.6)
 Asian 3 (2.3) 0 (0) 1 (1.5) 0 (0) 4 (1.1)
 Multiracial 4 (3.1) 5 (8.5) 3 (4.6) 7 (5.9) 19 (5.1)
 Other 10 (7.6) 4 (6.8) 6 (9.2) 6 (5.0) 26 (7.0)
Marital Status
 Married 43 (32.8) 29 (49.2) 19 (29.2) 50 (42.0) 141 (37.7)
 Single 41 (31.3) 14 (23.7) 21 (32.3) 19 (16.0) 95 (25.4)
 Cohabitating 8 (6.1) 2 (3.4) 3 (4.6) 6 (5.0) 19 (5.1)
 Widowed 5 (3.8) 1 (1.7) 4 (6.2) 5 (4.2) 15 (4.0)
 Divorced/Separated 34 (26.0) 13 (22.0) 18 (27.7) 39 (32.8) 104 (27.8)
Highest Education Achieved
 High School or Less 17 (13.0) 7 (11.9) 11 (16.9) 10 (8.4) 45 (12.0)
 Some College 71 (54.2) 28 (47.5) 33 (50.8) 60 (50.4) 192 (51.3)
 Bachelor’s Degree 26 (19.8) 13 (22.0) 15 (23.1) 36 (30.3) 90 (24.1)
 Graduate Degree 17 (13.0) 11 (18.6) 6 (9.2) 13 (10.9) 47 (12.6)
BMI Mean (SD) 28.1 (5.65) 29.7 (6.52) 28.9 (6.08) 30.0 (6.61) 29.1 (6.21)
PCL Total Scorea Mean (SD) 12.8 (15.5) 32.0 (19.1) 23.8 (19.0) 40.9 (18.0) 26.6 (21.0)
BDI Total Scoreb Mean (SD) 7.09 (8.21) 15.1 (11.5) 16.6 (13.9) 23.6 (13.1) 15.3 (13.3)
Age in years Mean (SD) 50.7 (14.8) 44.2 (13.3) 46.8 (12.6) 45.6 (12.2) 47.4 (13.6)
# of Deployments Mean (SD) 2.38 (6.74) 2.95 (4.55) 1.34 (1.60) 1.87 (1.99) 2.13 (4.58)
Open in a new tab

Note. Unless otherwise noted, data are presented as N (%). BDI = Beck Depression Inventory-II; BMI = Body mass index; MDD = Major Depressive Disorder; OEF/OIF/OND = Operation Enduring Freedom/Operation Iraqi Freedom/Operation New Dawn; PCL = PTSD Checklist-5; PTSD = Posttraumatic Stress Disorder; SD = Standard deviation.

a

Data missing for 17 participants.

b

Data missing for 23 participants.

Psychiatric Diagnoses

The original model assessing the association between hypertension and diagnostic group status using informative priors showed similar relative fit compared to sensitivity analysis models (Table S2; Sivula et al., 2020). Furthermore, the sensitivity analyses revealed that choice of prior hyperparameters had very little impact on final results (Table S3). Risk and relative risk estimates were nearly identical across models. Estimates that did show the greatest difference across choice of priors, such as the relative risk for hypertension in the group with a history of both PTSD and MDD (PTSD+MDD) compared to the group with neither diagnosis, varied only in magnitude, not in substantive interpretation. As such, the findings from the original model are considered robust with regards to selection of prior distributions.

Diagnostic checks for the original model showed excellent trace plot chain mixing (Figure S2), as well as appropriate R^ and ESS statistics (Table S4) for each parameter. Additionally, the posterior predictive check was supportive of good model fit, as generated DV data matched what was observed in the data set (Figure S3). The PTSD+MDD group demonstrated the highest risk of hypertension, followed by the PTSD-only group, the MDD-only group, and finally the group with neither diagnosis (Table 2). Looking at pairwise relative risk comparisons, the PTSD+MDD group was found to have increased risk of hypertension (i.e., relative risk > 1) compared to all three of the other diagnostic groups—for each of these comparisons, the posterior probability that the relative risk was greater than 1.0 exceeded 99.9%. In other words, there was a greater than 99.9% chance that the PTSD+MDD group had a higher risk of hypertension than the other three diagnostic groups. The PTSD-only group was also found to have a higher risk of hypertension than the group with neither diagnosis, with the posterior probability that the relative risk was greater than 1.0 exceeding 99.9% as well. The MDD-only group had a 99.3% chance of increased hypertension risk versus the group with neither diagnosis. Finally, the PTSD-only group had a 94.1% chance of increased hypertension risk versus the MDD-only group.

Table 2.

Adjusted Hypertension Risk Estimates and Pairwise Comparisons for Diagnostic Groups

Neither PTSD-only MDD-only PTSD+MDD
Neither 35.05 (30.50, 39.70) -- -- --
PTSD-only 1.29 (1.13, 1.45) 45.05 (39.49, 50.63) -- --
MDD-only 1.14 (1.02, 1.26) 0.89 (0.76, 1.03) 39.80 (34.84, 44.88) --
PTSD+MDD 1.66 (1.43, 1.94) 1.29 (1.14, 1.45) 1.46 (1.28, 1.68) 57.80 (52.63, 63.08)
Open in a new tab

Note: Numbers in the lower triangles represent pairwise relative risks (row/column). The diagonals represent the percentage risk of hypertension for the group. Values within the parentheses represent 95% credible intervals. The risk and relative risk estimates are estimated from their own respective posterior distributions—therefore the relative risk is not equal to the exact ratio of any two corresponding risk estimates. MDD = Major Depressive Disorder; PTSD = Posttraumatic Stress Disorder.

Symptom Severity Scores

The model assessing the association between hypertension risk and symptom severity scores also displayed excellent chain mixing (Figure S4), appropriate R^ and ESS statistics (Table S5), and an adequate posterior predictive check (Figure S5). Full summaries of regression coefficients on the log-odds scale are provided in Table S5. The 95% credible intervals for BDI-II score, PCL-5 score, and their interaction all included zero, indicating inconclusive evidence of an association with hypertension odds. However, it should be noted that more severe PCL-5 scores had a fairly high posterior probability of being associated with increased odds of hypertension (96.7%). Higher BDI-II scores and the interaction term only had marginally elevated posterior probabilities (74.4% and 67.3%, respectively) of being associated with decreased odds of hypertension. Examination of symptom severity subgroups helps to elucidate the interaction between these two measures (Table 3). Veterans with high PCL-5 and low BDI-II scores were estimated to have the highest risk of hypertension, whereas veterans with high BDI-II and low PCL-5 scores were estimated to have the lowest, which may be due to the small probability that higher BDI-II scores were associated with a decreased risk of hypertension.

Table 3.

Adjusted Hypertension Risk Estimates and Pairwise Comparisons for Symptom Severity Subgroups

Low PCL-5 / Low BDI-II High PCL-5 / Low BDI-II High BDI-II / Low PCL-5 High PCL-5 / High BDI-II
Low PCL-5 / Low BDI-II 39.85 (33.01, 47.24) -- -- --
High PCL-5 / Low BDI-II 1.42 (1.01, 1.89) 56.10 (40.70, 70.89) -- --
High BDI-II / Low PCL-5 0.93 (0.45, 1.51) 0.68 (0.28, 1.25) 36.55 (18.92, 55.85) --
High PCL-5 / High BDI-II 1.23 (0.95, 1.57) 0.88 (0.67, 1.19) 1.42 (0.89, 2.47) 48.55 (42.08, 54.75)
Open in a new tab

Note. Numbers in the lower triangle represent pairwise relative risks (row/column). The diagonal represents the percentage risk of hypertension for the group. Values within the parentheses represent 95% credible intervals. The risk and relative risk estimates are estimated from their own respective posterior distributions—therefore the relative risk is not equal to the exact ratio of any two corresponding risk estimates. BDI-II = Beck Depression Inventory-II; PCL-5 = PTSD Checklist-5.

Discussion

Diagnostic overlap between PTSD and MDD has complicated the relationship between these disorders and hypertension. The goal of the present study was to conceptually replicate the work of Kibler et al. (2009) by assessing the unique associations between PTSD, MDD, and hypertension with a veteran sample. One strength of the present analyses is the use of a Bayesian framework to directly update previous findings. Notably, study results were broadly consistent with those of Kibler et al. (2009), in that a history of PTSD-only and PTSD+MDD were most strongly associated with hypertension history. Compared to veterans without a history of PTSD or MDD, veterans with a history of only PTSD had an estimated 29% increase in hypertension risk. A history of both PTSD and MDD was associated with the highest risk of hypertension, as well as an elevated risk when compared to the other diagnostic groups. Finally, a history of MDD-only was also associated with elevated hypertension risk (14%) compared to the group with neither diagnosis. These findings were further supported by the performed sensitivity analyses—while some variation was observed in specific risk estimates, an overall association with increased hypertension risk remained consistent for all diagnostic groups.

Additionally, this study extended the work of Kibler et al. (2009) by assessing the association between hypertension and symptom severity scores. Results were consistent with the diagnostic status findings—the greatest risk of hypertension was estimated in veterans with high PCL-5 scores, regardless of accompanying scores on the BDI-II. Veterans scoring one standard deviation above the mean on the PCL-5—corresponding to a score of 48, which is well above the cut score for probable PTSD (U.S. Department of Veterans Affairs, 2021)—were estimated to have a 42% increased risk of hypertension compared to veterans reporting little-to-no PTSD and MDD symptoms. This level of risk is much higher than the average age-adjusted estimate of hypertension prevalence in U.S. adults (Ostchega et al., 2020). Importantly, the 95% credible interval for the relative risk estimate comparing veterans with high PCL-5 scores to those who were relatively asymptomatic did not include 1.0, indicating a high certainty that hypertension risk is elevated in this group. This provides support that increased severity of PTSD symptomatology is associated with increased risk of hypertension. However, data were only available for PTSD and depression symptom severity at the time of study participation, meaning that the obtained scores may not have accurately reflected the past experiences of the veterans. For example, veterans with a history of PTSD or depression may have experienced more severe symptoms in the past, particularly if their conditions were remitted at the time of the study. Additional research is needed to determine if PTSD or depression symptom severity is prospectively associated with hypertension risk.

Taken together, these two analyses provide support for an association between PTSD and elevated hypertension risk, independent of MDD. Biologically, PTSD may lead to or be associated with hypertension, at least in part, through disruption of the HPA-axis, sympathetic hyperactivity, and excessive inflammation (Fonkoue et al., 2020; Sherin & Nemeroff, 2011). PTSD is associated with abnormal regulation of the HPA axis, such as a blunted response to corticotropin-releasing hormone and disrupted immune and metabolic response to stress (Sherin & Nemeroff, 2011). Dysregulation of the HPA axis is linked with a broad array of pathologies, including hypertension (Burford et al., 2017). PTSD is also associated with excessive activation of the sympathetic nervous system (Sherin & Nemeroff, 2011), which is a core feature of hypertension (Fisher & Paton, 2012). Excessive sympathetic activity in PTSD is observed in response to traumatic reminders, which may help explain some of the cardinal features of the disorder, such as hyperarousal (Sherin & Nemeroff, 2011). Finally, increased PTSD severity is associated with increased markers of systemic and vascular inflammation (Fonkoue et al., 2020). Such chronic stress-induced disruptions may link PTSD to hypertension through perturbation of metabolic functioning via the brain and immune system, also known as the “selfish brain/selfish immune system” theory (Peters et al., 2004; Yamagata et al., 2017). Long-term energetic imbalance could then lead to systemic toxicity, negatively affecting both the brain and other body organs, including the cardiovascular system (Yamagata et al., 2017).

Identifying PTSD as a potential modifiable risk factor has implications for the management of hypertension within the VHA and other health systems. Treating hypertension can help prevent progression of cardiovascular disease (Messerli et al., 2017), the leading cause of death in the U.S. (Centers for Disease Control and Prevention, 2021, March 1). Although effective therapies like prolonged exposure exist for the treatment of PTSD (U.S. Department of Veterans Affairs & U.S. Department of Defense, 2017), hypertension management guidelines currently fail to consider the influence of mental health (U.S. Department of Veterans Affairs & U.S. Department of Defense, 2020). Ideally, anyone experiencing PTSD would receive timely and effective treatment; however, patients and providers may not be aware of the far-reaching impact that PTSD can have on cardiovascular health, which could affect decision-making regarding assessment and treatment planning. There is evidence that treatment of PTSD may reduce hypertension incidence (Burg et al., 2017) and lower blood pressure over a standard course of psychotherapy (Schubert et al., 2019). Early treatment of PTSD, or consideration of the possible influence of PTSD during hypertension management visits, could have a substantial impact on cardiovascular health outcomes.

Interestingly, compared to PTSD diagnosis/symptom severity, depression was not as strongly associated with elevated hypertension risk, despite the compelling theoretical models and previous evidence linking the two (Nichter et al., 2019; Stevelink et al., 2020; Yamagata et al., 2017). One possible explanation for this could involve limitations with how MDD is currently conceptualized. MDD is a notoriously heterogenous diagnostic condition, with such heterogeneity impairing efforts to understand and treat the debilitating disorder (Buch & Liston, 2021). Recent research into MDD has sought to address this conundrum by identifying clinical subgroups that can guide treatment decisions (Buch & Liston, 2021). It is possible that hypertension and other cardiovascular consequences are associated with specific depressive phenotypes, such as those characterized by excessive inflammation (Adzic et al., 2018). The finding in the present study that a history of both PTSD and MDD was associated with the highest risk of hypertension suggests that MDD may play some role in hypertension pathogenesis, at least in certain veterans. Further exploration of MDD subgroups characterized by distinct genetic, neurobiological and symptom profiles may help clarify the depression-hypertension relationship.

Limitations and Future Directions

A major limitation of this study, similar to Kibler et al. (2009), is that this was not a prospective evaluation and the timing of hypertension relative to PTSD and MDD onset was not assessed. As such, hypertension may have preceded PTSD or MDD onset. However, as in Kibler et al. (2009), the median age of the present sample (47 years) tended to be younger than the median age of hypertension development in the general population (~48–57 years; Shriner et al., 2020). This increases the likelihood that PTSD preceded the development of hypertension and represents a modifiable risk factor. Ultimately, though, additional prospective research will be necessary to fully elucidate the associations between PTSD, MDD, and hypertension. Future studies should consider using both structured diagnostic interviews and repeated symptom severity measures when assessing for PTSD and MDD. Such research could also assess whether chronicity of PTSD and MDD better explains changes in hypertension risk, as chronic stress dysregulation, rather than acute stressors, would be expected to lead to cardiovascular disease (Peters et al., 2004; Yamagata et al., 2017).

Another limitation of this study is the relatively small sample size. It should be noted that the veteran sample had higher rates of MDD, PTSD, and hypertension relative to those observed in Kibler et al. (2009), meaning a smaller sample size could be sufficient for estimation. Additionally, the use of Bayesian estimation with informative priors, as was done here, is well-suited for small data sets (van de Schoot et al., 2015). However, a larger sample size can help improve the precision of parameter estimates in future studies. Other limitations include using measures of current symptom severity (i.e., BDI-II and PCL-5) as proxies for long-term symptomatology, being unable to anchor the PCL-5 to a specific trauma, and the relatively low severity of depression observed in the groups with a history of MDD. Also, other variables not included —such as family hypertension history, medication use, diabetes, and comorbid psychiatric diagnoses—may be potential confounding risk factors.

A final limitation is the use of medical records to obtain hypertension diagnostic information. Defining hypertension using medical records can underestimate hypertension prevalence compared to in-person interviews and blood pressure assessment (Peng et al., 2016). The study team attempted to minimize this concern by using VHA-specific algorithms (Klabunde et al., 2000) but the possibility remains that rates of hypertension may have been underestimated.

Conclusions

Consistent with findings by Kibler et al. (2009), this study found that PTSD was associated with hypertension independently of depression. Furthermore, severity of PTSD symptoms appears to be linked with hypertension risk. Treating PTSD may help prevent and manage hypertension, as well as attenuate progression of hypertension to more severe forms of cardiovascular disease.

Supplementary Material

Supplementary Tables
Supplementary Figures

Clinical Impact Statement.

Posttraumatic stress disorder (PTSD) and depression are both associated with hypertension, a leading risk factor for heart disease, stroke, and death. However, PTSD and depression often occur together and have similar symptoms, making it unclear if each disorder contributes to hypertension. This study of United States military veterans was conducted to explore the potentially unique relationship that each disorder may have with hypertension. Findings suggested that a history of PTSD was linked with elevated risk of hypertension. Treating PTSD may help prevent hypertension, as well as associated outcomes related to this condition.

Acknowledgments

This research was supported by the Department of Veterans Affairs Office of Academic Affiliations Advanced Fellowship Program in Mental Illness Research and Treatment, as well as the Rocky Mountain Mental Illness Research, Education, and Clinical Center. Dr. Kaizer’s efforts were supported by the National Heart, Lung, and Blood Institute of the National Institutes of Health under award number NHLBI K01HL151754. Since multiple authors are employees of the US Government and contributed to this manuscript as part of their official duties, the work is not subject to U.S. copyright.

The views expressed in this article are those of the authors and do not necessarily represent those of the US Department of Veterans Affairs, the National Institutes of Health, or the US government.

Dr. Brenner reports grants from the VA, DOD, NIH, and the State of Colorado, editorial renumeration from Wolters Kluwer, and royalties from the American Psychological Association and Oxford University Press. In addition, she consults with sports leagues via her university affiliation.

Footnotes

Data Transparency

Some of the data reported in this manuscript were collected as part of a larger data collection. Findings from the data collection have been reported in separate manuscripts. The first manuscript (published) described the study protocol and preliminary demographic findings, along with prevalence rates of psychiatric diagnoses. The second manuscript (published) examined relationships between dietary habits and gut microbiota. The third manuscript (published) compared analyses of the gut microbiome between those with and without a history of severe or moderate traumatic brain injury. The fourth manuscript (in the submission process) examined the associations between military-related exposures, health conditions, medication use, and the gut microbiome. The fifth manuscript (in the submission process) evaluated the associations between posttraumatic stress disorder, liver cirrhosis, and the gut microbiome. The sixth manuscript (in preparation) is examining the relationship between past or current homelessness and the diversity of skin, oral, and gut microbiomes. The seventh manuscript (in preparation) is examining the impact of food deserts on the gut microbiome. The eighth manuscript (the current manuscript) is focused on the association between posttraumatic stress disorder, major depressive disorder, and hypertension. Notably, hypertension data are novel and were specifically collected for this study.

References

  1. Adzic M, Brkic Z, Mitic M, Francija E, Jovicic MJ, Radulovic J, & Maric NP (2018). Therapeutic Strategies for Treatment of Inflammation-related Depression. Current Neuropharmacology, 16(2), 176–209. 10.2174/1570159x15666170828163048 [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Barbano AC, van der Mei WF, deRoon-Cassini TA, Grauer E, Lowe SR, Matsuoka YJ, O’Donnell M, Olff M, Qi W, Ratanatharathorn A, Schnyder U, Seedat S, Kessler RC, Koenen KC, & Shalev AY (2019). Differentiating PTSD from anxiety and depression: Lessons from the ICD-11 PTSD diagnostic criteria. Depression and Anxiety, 36(6), 490–498. 10.1002/da.22881 [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Beck AT, Steer RA, & Brown GK (1996). Manual for the Beck Depression Inventory—II. Psychological Corporation.
  4. Blevins CA, Weathers FW, Davis MT, Witte TK, & Domino JL (2015). The Posttraumatic Stress Disorder Checklist for DSM-5 (PCL-5): Development and Initial Psychometric Evaluation. Journal of Traumatic Stress, 28(6), 489–498. 10.1002/jts.22059 [DOI] [PubMed] [Google Scholar]
  5. Brenner LA, Hoisington AJ, Stearns-Yoder KA, Stamper CE, Heinze JD, Postolache TT, Hadidi DA, Hoffmire CA, Stanislawski MA, & Lowry CA (2018). Military-Related Exposures, Social Determinants of Health, and Dysbiosis: The United States-Veteran Microbiome Project (US-VMP). Frontiers in Cellular and Infection Microbiology, 8(400). 10.3389/fcimb.2018.00400 [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Brown M, Kaplan C, & Jason L (2012). Factor analysis of the Beck Depression Inventory-II with patients with chronic fatigue syndrome. Journal of Health Psychology, 17(6), 799–808. 10.1177/1359105311424470 [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Buch AM, & Liston C (2021). Dissecting diagnostic heterogeneity in depression by integrating neuroimaging and genetics. Neuropsychopharmacology : official publication of the American College of Neuropsychopharmacology, 46(1), 156–175. 10.1038/s41386-020-00789-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Burford NG, Webster NA, & Cruz-Topete D (2017). Hypothalamic-Pituitary-Adrenal Axis Modulation of Glucocorticoids in the Cardiovascular System. International Journal of Molecular Sciences, 18(10). 10.3390/ijms18102150 [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Burg MM, Brandt C, Buta E, Schwartz J, Bathulapalli H, Dziura J, Edmondson DE, & Haskell S (2017). Risk for Incident Hypertension Associated With Posttraumatic Stress Disorder in Military Veterans and the Effect of Posttraumatic Stress Disorder Treatment. Psychosomatic Medicine, 79(2), 181–188. 10.1097/psy.0000000000000376 [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Burg MM, & Soufer R (2016). Post-traumatic Stress Disorder and Cardiovascular Disease. Current Cardiology Reports, 18(10), 94. 10.1007/s11886-016-0770-5 [DOI] [PubMed] [Google Scholar]
  11. Bürkner P-C (2017). brms : An R Package for Bayesian Multilevel Models Using Stan. Journal of Statistical Software, 80. 10.18637/jss.v080.i01 [DOI] [Google Scholar]
  12. Burnette CE, Ka’apu K, Scarnato JM, & Liddell J (2020). Cardiovascular Health among U.S. Indigenous Peoples: A Holistic and Sex-Specific Systematic Review. Journal of Evidence-Based Social Work, 17(1), 24–48. 10.1080/26408066.2019.1617817 [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Cameron KL, Sturdivant RX, & Baker SP (2019). Trends in the incidence of physician-diagnosed posttraumatic stress disorder among active-duty U.S. military personnel between 1999 and 2008. Military Medical Research, 6(1), 8. 10.1186/s40779-019-0198-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Campbell SB, & Renshaw KD (2018). Posttraumatic stress disorder and relationship functioning: A comprehensive review and organizational framework. Clinical Psychology Review, 65, 152–162. 10.1016/j.cpr.2018.08.003 [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Centers for Disease Control and Prevention. (2021, March 1). Leading Causes of Death. Retrieved 2021, August 31 from https://www.cdc.gov/nchs/fastats/leading-causes-of-death.htm
  16. First MB, Williams JBW, Karg RS, & Spitzer RL (2015). Structured Clinical Interview for DSM-5--Research Version (SCID-5 for DSM-5, Research Version; SCID-5-RV). American Psychiatric Association. [Google Scholar]
  17. Fisher JP, & Paton JF (2012). The sympathetic nervous system and blood pressure in humans: implications for hypertension. Journal of Human Hypertension, 26(8), 463–475. 10.1038/jhh.2011.66 [DOI] [PubMed] [Google Scholar]
  18. Fonkoue IT, Marvar PJ, Norrholm S, Li Y, Kankam ML, Jones TN, Vemulapalli M, Rothbaum B, Bremner JD, Le NA, & Park J (2020). Symptom severity impacts sympathetic dysregulation and inflammation in post-traumatic stress disorder (PTSD). Brain, Behavior, and Immunity, 83, 260–269. 10.1016/j.bbi.2019.10.021 [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Holliday R, Borges LM, Stearns-Yoder KA, Hoffberg AS, Brenner LA, & Monteith LL (2020). Posttraumatic Stress Disorder, Suicidal Ideation, and Suicidal Self-Directed Violence Among U.S. Military Personnel and Veterans: A Systematic Review of the Literature From 2010 to 2018. Frontiers in Psychology, 11, 1998. 10.3389/fpsyg.2020.01998 [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Kibler JL, Joshi K, & Ma M (2009). Hypertension in relation to posttraumatic stress disorder and depression in the US National Comorbidity Survey. Behav Med, 34(4), 125–132. 10.3200/bmed.34.4.125-132 [DOI] [PubMed] [Google Scholar]
  21. Klabunde CN, Potosky AL, Legler JM, & Warren JL (2000). Development of a comorbidity index using physician claims data. Journal of Clinical Epidemiology, 53(12), 1258–1267. 10.1016/s0895-4356(00)00256-0 [DOI] [PubMed] [Google Scholar]
  22. McGinnis KA, Brandt CA, Skanderson M, Justice AC, Shahrir S, Butt AA, Brown ST, Freiberg MS, Gibert CL, Goetz MB, Kim JW, Pisani MA, Rimland D, Rodriguez-Barradas MC, Sico JJ, Tindle HA, & Crothers K (2011). Validating smoking data from the Veteran’s Affairs Health Factors dataset, an electronic data source. Nicotine & Tobacco Research, 13(12), 1233–1239. 10.1093/ntr/ntr206 [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Messerli FH, Rimoldi SF, & Bangalore S (2017). The transition from hypertension to heart failure: Contemporary update. JACC: Heart Failure, 5(8), 543–551. doi:https://doi.org/ 10.1016/j.jchf.2017.04.012 [DOI] [PubMed] [Google Scholar]
  24. Minhas S, Patel JR, Malik M, Hana D, Hassan F, & Khouzam RN (2021). Mind-Body Connection: Cardiovascular Sequelae of Psychiatric Illness. Current Problems in Cardiology, 100959. 10.1016/j.cpcardiol.2021.100959 [DOI] [PubMed] [Google Scholar]
  25. Morris WK, Vesk PA, McCarthy MA, Bunyavejchewin S, & Baker PJ (2015). The neglected tool in the Bayesian ecologist’s shed: a case study testing informative priors’ effect on model accuracy. Ecology and Evolution, 5(1), 102–108. 10.1002/ece3.1346 [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Muller CJ, & MacLehose RF (2014). Estimating predicted probabilities from logistic regression: different methods correspond to different target populations. International Journal of Epidemiology, 43(3), 962–970. 10.1093/ije/dyu029 [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Nichter B, Norman S, Haller M, & Pietrzak RH (2019). Physical health burden of PTSD, depression, and their comorbidity in the U.S. veteran population: Morbidity, functioning, and disability. Journal of Psychosomatic Research, 124, 109744. 10.1016/j.jpsychores.2019.109744 [DOI] [PubMed] [Google Scholar]
  28. Ostchega Y, Fryar CD, Nwankwo T, & Nguyen DT (2020). Hypertension prevalence among adults aged 18 and over: United States, 2017–2018. NCHS Data Brief, no 364. https://www.cdc.gov/nchs/products/databriefs/db364.htm [PubMed]
  29. Peng M, Chen G, Kaplan GG, Lix LM, Drummond N, Lucyk K, Garies S, Lowerison M, Weibe S, & Quan H (2016). Methods of defining hypertension in electronic medical records: validation against national survey data. Journal of Public Health, 38(3), e392–e399. 10.1093/pubmed/fdv155 [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Peters A, Schweiger U, Pellerin L, Hubold C, Oltmanns KM, Conrad M, Schultes B, Born J, & Fehm HL (2004). The selfish brain: competition for energy resources. Neuroscience and Biobehavioral Reviews, 28(2), 143–180. 10.1016/j.neubiorev.2004.03.002 [DOI] [PubMed] [Google Scholar]
  31. R Core Team. (2021). R: A language and environment for statistical computing [Computer software]. R Foundation for Statistical Computing. https://www.R-project.org/ [Google Scholar]
  32. Reis DJ, Namekata MS, Oehlert ME, & King N (2020). A preliminary review of the Beck Depression Inventory-II (BDI-II) in veterans: Are new norms and cut scores needed? Psychological Services, 17(3), 363–371. 10.1037/ser0000342 [DOI] [PubMed] [Google Scholar]
  33. Schubert CF, Schreckenbach M, Kirmeier T, Gall-Kleebach DJ, Wollweber B, Buell DR, Uhr M, Rosner R, & Schmidt U (2019). PTSD psychotherapy improves blood pressure but leaves HPA axis feedback sensitivity stable and unaffected: First evidence from a pre-post treatment study. Psychoneuroendocrinology, 100, 254–263. 10.1016/j.psyneuen.2018.10.013 [DOI] [PubMed] [Google Scholar]
  34. Sherin JE, & Nemeroff CB (2011). Post-traumatic stress disorder: the neurobiological impact of psychological trauma. Dialogues in Clinical Neuroscience, 13(3), 263–278. 10.31887/DCNS.2011.13.2/jsherin [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Shriner D, Bentley AR, Zhou J, Ekoru K, Doumatey AP, Chen G, Adeyemo A, & Rotimi CN (2020). Time-to-event modeling of hypertension reveals the nonexistence of true controls. Elife, 9. 10.7554/eLife.62998 [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Sivula T, Magnusson M, & Vehtari A (2020). Uncertainty in Bayesian leave-one-out cross-validation based model comparison. arXiv. [Google Scholar]
  37. Stevelink SAM, Opie E, Pernet D, Gao H, Elliott P, Wessely S, Fear NT, Hotopf M, & Greenberg N (2020). Probable PTSD, depression and anxiety in 40,299 UK police officers and staff: Prevalence, risk factors and associations with blood pressure. PLOS One, 15(11), e0240902. 10.1371/journal.pone.0240902 [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Trivedi RB, Post EP, Sun H, Pomerantz A, Saxon AJ, Piette JD, Maynard C, Arnow B, Curtis I, Fihn SD, & Nelson K (2015). Prevalence, Comorbidity, and Prognosis of Mental Health Among US Veterans. American Journal of Public Health, 105(12), 2564–2569. 10.2105/ajph.2015.302836 [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Tsai J, & Shen J (2017). Exploring the Link Between Posttraumatic Stress Disorder and inflammation-Related Medical Conditions: An Epidemiological Examination. Psychiatric Quarterly, 88(4), 909–916. 10.1007/s11126-017-9508-9 [DOI] [PubMed] [Google Scholar]
  40. U.S. Department of Veterans Affairs. (2019). Department of Veterans Affairs FY2018 – 2024 Strategic Plan. https://www.va.gov/oei/docs/va2018-2024strategicplan.pdf
  41. U.S. Department of Veterans Affairs. (2021). PTSD Checklist for DSM-5 (PCL-5). Retrieved 2021, August 31 from https://www.ptsd.va.gov/professional/assessment/adult-sr/ptsd-checklist.asp
  42. U.S. Department of Veterans Affairs, & U.S. Department of Defense. (2016). VA/DoD Clinical Practice Guideline for the Management of Major Depressive Disorder. https://www.healthquality.va.gov/guidelines/MH/mdd/ [DOI] [PubMed]
  43. U.S. Department of Veterans Affairs, & U.S. Department of Defense. (2017). VA/DoD Clinical Practice Guideline for the Management of Posttraumatic Stress Disorder and Acute Stress Disorder. https://www.healthquality.va.gov/guidelines/mh/ptsd/
  44. U.S. Department of Veterans Affairs, & U.S. Department of Defense. (2020). VA/DoD Clinical Practice Guideline for the Diagnosis and Management of Hypertension in the Primary Care Setting. https://www.healthquality.va.gov/guidelines/cd/htn/
  45. van de Schoot R, Broere JJ, Perryck KH, Zondervan-Zwijnenburg M, & van Loey NE (2015). Analyzing small data sets using Bayesian estimation: the case of posttraumatic stress symptoms following mechanical ventilation in burn survivors. European Journal of Psychotraumatology, 6, 25216. 10.3402/ejpt.v6.25216 [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Vehtari A, Gelman A, Simpson D, Carpenter B, & Bürkner P (2021). Rank-Normalization, Folding, and Localization: An Improved R^ for Assessing Convergence of MCMC (with Discussion). Bayesian Analysis, 16(2), 667–718, 652. [Google Scholar]
  47. Wagenmakers EJ, Marsman M, Jamil T, Ly A, Verhagen J, Love J, Selker R, Gronau QF, Šmíra M, Epskamp S, Matzke D, Rouder JN, & Morey RD (2018). Bayesian inference for psychology. Part I: Theoretical advantages and practical ramifications. Psychonomic Bulletin & Review, 25(1), 35–57. 10.3758/s13423-017-1343-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Yamagata AS, Mansur RB, Rizzo LB, Rosenstock T, McIntyre RS, & Brietzke E (2017). Selfish brain and selfish immune system interplay: A theoretical framework for metabolic comorbidities of mood disorders. Neuroscience and Biobehavioral Reviews, 72, 43–49. 10.1016/j.neubiorev.2016.11.010 [DOI] [PubMed] [Google Scholar]
  49. Yehuda R, Hoge CW, McFarlane AC, Vermetten E, Lanius RA, Nievergelt CM, Hobfoll SE, Koenen KC, Neylan TC, & Hyman SE (2015). Post-traumatic stress disorder. Nature Reviews Disease Primers, 1(1), 15057. 10.1038/nrdp.2015.57 [DOI] [PubMed] [Google Scholar]
  50. Yoon J, Zulman D, Scott JY, & Maciejewski ML (2014). Costs associated with multimorbidity among VA patients. Med Care, 52 Suppl 3(Suppl 3), S31–36. 10.1097/mlr.0000000000000061 [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Zulman DM, Pal Chee C, Wagner TH, Yoon J, Cohen DM, Holmes TH, Ritchie C, & Asch SM (2015). Multimorbidity and healthcare utilisation among high-cost patients in the US Veterans Affairs Health Care System. BMJ Open, 5(4), e007771. 10.1136/bmjopen-2015-007771 [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Tables
NIHMS1877170-supplement-Supplementary_Tables.docx (42.6KB, docx)
Supplementary Figures
NIHMS1877170-supplement-Supplementary_Figures.pdf (5.8MB, pdf)

RESOURCES