Szabo et al., 2021_Resilience predicts posttraumatic cognitions after a trauma reminder task and subsequent positive emotion induction among veterans with PTSD

Resilience, the tendency to bounce back from difficult circumstances, is negatively associated with posttraumatic cognitions PTCs among individuals with a history of trauma, and thus it

Resilience Predicts Posttraumatic Cognitions after a Trauma Reminder Task and Subsequent Positive Emotion Induction among Veterans with PTSD

Yvette Z Szabo1*,2, Sheila B Frankfurt1,2,3, A Solomon Kurz1,2,

Austen R Anderson1,2, Adam P McGuire1*,2,4

1 VA VISN 17 Center of Excellence for Research on Returning War Veterans, Waco, TX, USA

2 Central Texas Veterans Health Care System, Temple, TX, USA

3 Texas A&M University, College of Medicine, College Station, TX, USA

4 Department of Psychology and Counseling, The University of Texas at Tyler, Tyler, TX, USA

Yvette Szabo https://orcid.org/0000-0002-5436-7081

Sheila Frankfurt https://orcid.org/0000-0002-8156-1488

Austen Anderson https://orcid.org/0000-0003-2585-8030

A Solomon Kurz https://orcid.org/0000-0001-7597-3745

Adam McGuire https://orcid.org/0000-0002-4512-1207

Author Note

This manuscript has been accepted for publication in Psychological Trauma: Theory,

Research, Practice, and Policy This postprint version is nearly identical to the accepted paper.

Corresponding: Yvette Szabo, yvette.szabo@va.gov or Adam McGuire

adam.mcguire@va.gov; 4800 Memorial Drive (151C), Waco, TX 76711

Acknowledgments

This material is the result of work with resources and the use of facilities at the VISN 17 Center

of Excellence for Research on Returning War Veterans and the Central Texas Veterans Health Care System Dr Szabo is supported by Career Development Award IK1-RX003122, Dr

Frankfurt is supported by Career Development Award IK1-RX002427, and Dr McGuire is supported by a Small Projects in Rehabilitation Research Award I21-RX003035- from the UnitedStates (U.S.) Department of Veterans Affairs, Rehabilitation Research and Development Service The views expressed herein are those of the authors and do not necessarily reflect the official policy or position of the Department of Veterans Affairs or the United States Government The authors do not have any financial conflicts of interest to disclose

Author Contributions (using CRediT categories)

Conceptualization / Investigation: YS, AM

Methodology / Data curation: AM, supporting YS

Formal analysis: YS, supporting SK

Validation: YS, supporting SK, AA

Visualization: YS, SK

Writing – Initial Draft – YS, with SF (Introduction) and AM (methods) drafting sections

Writing – Editing and Revising – YS, SF, AK, AA, AM

Abstract Objective: Posttraumatic stress disorder (PTSD) is a common problem for veterans Resilience,

the tendency to bounce back from difficult circumstances, is negatively associated with

posttraumatic cognitions (PTCs) among individuals with a history of trauma, and thus it may be important to understand responses to trauma reminders

Method: Using a quasi-experimental design, we examined the association between trait

resilience and state PTCs in veterans with PTSD (n = 47, Mage = 48.60, 91.8% male) at two

points: following a written trauma narrative exposure (Time 1 [T1]), and following a subsequent positive distraction task (i.e., brief, positive video) (T2)

Results: After controlling for PTSD symptom severity and combat exposure, resilience was

negatively associated with PTCs at T1 (ΔR2 = 19) and T2 (ΔR2 = 13) However, resilience was apoor predictor of change in PTCs from T1 to T2 We also examined the relationship between resilience and subtypes of PTCs: resilience was associated with negative views of the self (T1,

ΔR2 = 24) but not negative views of the world or self-blame (T1, ΔR2s < 07); these results were consistent at T2

Conclusions: Thus, resilience may attenuate negative trauma-related cognitions after trauma

recall; however, this study was not designed to test causal pathways Future research could examine whether resilience-building exercises reduce negative PTCs after trauma reminders among veterans Additional research is needed to generalize to other trauma-exposed

populations

Keywords: resilience, trauma, PTSD, veteran, posttraumatic cognition

Clinical Impact Statement:

PTSD is a highly prevalent and debilitating disorder for war veterans This study uses a experimental design to help further understanding about how resilience is related to one proposed mechanism of PTSD – posttraumatic cognitions – in response to reminders of the trauma Our findings suggest that resilience may buffer against negative trauma-related cognitions after trauma recall, though future research is needed to understand the long-term clinical utility

quasi-Resilience Predicts Posttraumatic Cognitions after a Trauma Reminder Task and Subsequent Positive Emotion Induction among Veterans with PTSD

Approximately 23% of post-9/11 veterans are diagnosed with posttraumatic stress

disorder (PTSD) (Fulton et al., 2015) Trauma-focused cognitive behavioral therapies (CBTs) such as Cognitive Processing Therapy (CPT; Resick et al., 2017) propose that negative or

distorted beliefs about oneself and the world generated in the aftermath of a trauma give rise to PTSD (Resick, 2001) Some prototypical negative beliefs, called posttraumatic cognitions

(PTCs), include ‘I am a bad person,’ ‘The world is a completely dangerous place,’ and ‘I am to

blame for my trauma.’ Individuals with a history of trauma may try to avoid or distract

themselves from their posttraumatic cognitions and thus miss potentially disconfirming

experiences that could counter their negative beliefs Posttraumatic cognitions are believed to maintain PTSD and so are the focus of PTSD therapies such as CPT (LoSavio et al., 2017) However, little is known about predictors of individual differences in PTCs It is possible that treatment could target factors that influence PTCs and thus enhance treatment through

individualized tailoring

Resilience, the tendency to bounce back from difficult circumstances (Smith et al., 2008),may be one such modifiable factor that accounts for variance in PTCs In some studies, resiliencewas negatively associated with PTSD symptoms and positively associated with posttraumatic growth (e.g., Bensimon, 2012) In terms of evidence of an association with PTCs, greater

resilience was associated with fewer PTCs among active-duty United States (U.S.) soldiers and veterans (Sexton et al., 2018; Zang et al., 2017) Although this suggests that resilience buffers trauma-related distress, a recent study reported that self-reported trait resilience did not predict resilience in PTSD-related impairment above the contribution of other personality factors, self-

reported psychological flexibility, and mental health symptoms among veterans (Meyer et al., 2019a) However, the latter study used a unique technique to derive their definition of resilience based on residuals in a PTSD-related factor score and thus, it is difficult to directly compare it to the larger body of research Because the specific mechanisms by which resilience might reduce PTSD are unknown, additional work is needed to further clarify how resilience is associated withPTCs

As highlighted by previous research, resilience may play an important role in

understanding individual differences in PTCs and PTSD symptom severity among exposed people However, most previous research on the role of resilience in PTCs and PTSD has relied on cross-sectional or observational assessment (e.g., Infurna & Jayawickreme, 2019; Zang et al., 2017) Naturalistic research designs limit our ability to draw conclusions about the associations between trait resilience and state-level experiences of trauma-related cognitions Forexample, cross-sectional studies typically assess associations between global ratings of resilienceand outcomes such as PTSD, failing to account for short-term fluctuations in the strength of belief in PTCs One exception to this trend was an experience sampling study that measured PTCs several times a day over 7 days and revealed both between-person and within-person variability in PTCs (Carlson et al., 2016) Further, the strength of belief in PTCs are expected to change within individual sessions of CPT After the patient identifies a trauma-related belief and rates the strength of that belief on a scale of 0-100, they complete a worksheet that promotes cognitive restructuring, followed up re-rating the strength of their belief in that thought (Resick

trauma-et al., 2017) Empirically, the degree of change in PTCs is associated with PTSD improvement during a course of treatment and often precede symptom improvement (Brown et al., 2019) Among patients receiving a trauma-focused CBT treatment, nearly a third experience sudden

gains in both trauma-related appraisals and PTSD symptoms, with the former predicting the latter (Wiedemann et al., 2020).Thus, observational research of daily life, the theory underlying effective evidence-based treatments, and clinical research findings all suggest that belief in PTCs

do fluctuate in ways that are clinically meaningful However, controlled study designs that provide a higher degree of sensitivity to changes in outcomes and internal validity are needed In line with the research demonstrating within-person variation in PTCs and the literature indicatingthat affect or mood impacts our thinking (Dolan, 2002; Forgas, 2017), we designed a controlled study to measure cognitive responses to specific stimuli to understand whether resilience can act

as a personal resource in situations that might elicit PTCs

The present study used data from a quasi-experimental study to examine the association between trait resilience and state PTCs at two points: following a written trauma narrative exposure (Time 1 [T1]), and following a positive mood-inducing video (T2) We had three hypotheses: resilience would be negatively associated with PTCs following a trauma reminder (T1; Hypothesis 1 [H1]); resilience would be negatively associated with PTCs following positivestimuli (T2; H2); and, resilience would be positively associated with decrease in PTCs from T1

to T2 assessments (H3) We assessed whether the association between resilience and PTCs at both timepoints differed by PTCs subscale and considered these analyses exploratory

Method

This study is part of a larger study on the role of positive psychology constructs in the development and maintenance of PTSD; detailed information about the parent study is reported elsewhere (McGuire & Mignogna, 2021) Briefly, the parent study recruited veterans between the ages of 18–90 with subclinical or clinical PTSD symptoms (defined as a positive screen on a symptom checklist, as described below) The sample was recruited from a southern Veterans

Affairs Hospital via mail, flyer, and provider referrals Exclusion criteria were active psychosis, active mania/hypomania, and cognitive disorders (e.g., dementia, traumatic brain injury)

Eligibility was assessed during a phone screen Veterans were asked to self-report on the domainsthat were theoretically important for completing this study Additionally, mailed recruitment letters were only sent to Veterans without documented traumatic brain injury (TBI) or dementia

in the medical record Of the 59 veterans who were screened, 48 veterans met criteria and

participated in the study One participant was excluded because of repeated disruptions to their session that compromised the fidelity of session tasks Thus, the final sample comprised 47 veterans All participants provided written informed consent and this study was approved by the local Institutional Review Board Participants were financially compensated for their

participation

Procedures

The present study was completed in a single session that lasted approximately 90

minutes First, participants completed baseline measures and wrote a trauma narrative describing their worst traumatic event for 15 minutes Next, participants completed the measure of PTCs (Time 1; T1) Then, participants were randomized to watch one of two positive emotion-inducingvideos (one condition targeted amusement or humor and the other targeted moral elevation (i.e., inspiration by virtuous acts) Following the videos, all participants completed a brief reflection exercise where they wrote about their reactions to the videos for approximately 10 minutes Lastly, participants completed the PTCs measure for a second time (T2) Participants completed T1 and T2 PTCs approximately 50 minutes apart For the purposes of the current study, the two positive emotion-inducing video conditions were collapsed into one group because video

condition did not predict PTCs at T2

The 20-item Posttraumatic Stress Disorder Checklist-5 (PCL-5; Weathers et al., 2013) assesses the respondent’s severity of PTSD symptoms over the past month Items are rated on a

Likert-style scale from 1 (not at all) to 5 (extremely) Symptom severity was calculated by

summing all items PCL-5 scores > 32 indicated subclinical or clinical PTSD symptom severity (Bovin et al., 2016) Previous research has demonstrated convergent and discriminant validity of the PCL-5 in veteran samples (Bovin et al., 2016) Mean scores on the PCL were comparable to

a sample of veterans seeking treatment for PTSD (Zalta et al., 2018) Internal consistency in the present sample was α = 89 The PCL-5 was administered during the phone screen

The 6-item Brief Resilience Scale (BRS; Smith et al., 2008) assesses one’s tendency to bounce back and recover from stressful situations The initial study demonstrated good internal consistency, factor structure, and convergent and discriminant validity in healthy and physical

rehabilitation samples Items are rated on a Likert-style scale from 1 (strongly disagree) to 5 (strongly agree), with 3 items were reverse scored Then, an item average score is calculated that

ranges from 1 to 5 (Smith et al., 2008) Scores on the BRS were slightly lower but comparable to

an online sample of veterans with mental health problems (Umucu et al., 2021).Internal

consistency of the sum score in the present study was α = 74 The BRS was administered at baseline prior to the experimental procedure

The 7-item Critical Warzone Experiences (CWE; Kimbrel et al., 2014) assesses negative warzone and military experiences, such as combat exposure The initial study demonstrated goodinternal consistency, factor structure, test-retest reliability, and validity, evidenced by correlationsranging from 51 to 71 with PTSD diagnosis and symptom severity, global functional

impairment, anxiety and depression Each item is rated based on how often it occurred using a

Likert-style 0 (not at all) to 5 (five or more times) scale Internal consistency in the present study

was α = 81 Average scores on the CWE were higher than a recent sample of OEF/OIF Veterans recruited from the same geographical area (Meyer et al., 2019b) The CWE was administered at baseline

The 36-item Posttraumatic Cognitions Inventory (PTCI; Foa et al., 1999) assesses

strength of agreement with negative cognitions and beliefs associated with an index trauma Each

item is rated from 1 (totally disagree) to 7 (totally agree) The present study modified the

measure so each question was anchored to the Veteran’s cognitions “at this moment.” In addition

to a total score, three subscales can be calculated: negative views of the self, negative views of the world, and self-blame The PTCI was administered twice, at T1 and T2 Internal consistency for the total measure was α = 94 at T1 and α = 96 at T2; see table S1 in supplementary materialsfor PTCI subscale reliability Average levels of the PTCI were higher at T1 and more comparable

at T2 to levels of the PTCI among a sample of treatment seeking Veterans (Sexton et al., 2018)

Data Analysis Plan

Primary analyses were run in R (R Core Team, 2021), using the base R stats, psych (Revelle, 2020), and effsize (Torchiano, 2000) packages We also used additional online

confidence interval calculators (Soper, 2006) There were no missing data Descriptive data are available in Table 2 All hypotheses were tested using multiple linear regression For each of the three criterion (PCTI at T1, PTCs at T2, and change in PTCs), we fit two models: a baseline model and a theoretical model Each of the baseline models contained PTSD symptom severity and combat exposure as the control covariates The three theoretical models added resilience to the two control covariates, which allowed us to assess the unique variance resilience explained above and beyond the potential effects of baseline PTSD symptoms and history of combat

exposure H1 was tested using PTCI scores at T1 as the dependent variable and H2 was tested using PTCI scores as T2 as the dependent variable The H2 model was also tested with

experimental condition as an additional predictor (β = 01, p = 94), which was not significant

and did not change the pattern of findings; therefore, we reported results from H2 models that did not include experimental condition as a predictor A paired samples t-test was used to

compare PTCs from T1 and T2 H3 was tested with change in PTCI (T2-T1) as the outcome Given some research has suggested the PTCI subscales have differential associations with PTSD outcomes (e.g., Kumpula et al., 2017; Startup et al., 2007), we conducted exploratory analyses byfitting models that included each PTCI subscale as a dependent variable As with the primary hypotheses, for each of the subscales we fit both a baseline model containing the two control variables, to which we then added resilience as the focal exploratory predictor

Effect sizes for regression analyses were calculated using the R2, which informs the

percent of variance accounted for in the model, and the change in f2 from the baseline to the theoretical models, which informs the incremental contribution of resilience in predicting

variance in PTCs after accounting for combat exposure and PTSD symptom severity Based on

(Cohen, 1988), f2 ≤ 0.02 was interpreted as a small effect; f2 ≤ .15 as a medium effect; and, f2≤

0.35 as a large effect Regression coefficients were reported in both the unstandardized (B) and

standardized metric (β)

Results

Veterans in the sample were on average 49 years old (Mage = 48.60, SD = 12.72)

Participants were predominantly male (91.8%), white (69.34%), Army veterans (79.60%) who had deployed 1.49 times and reported 13.53 years of education and an income of less than

$45,000, on average Scores on the PCL ranged from 33-80 Additional demographic informationand military history is provided in Table 1.

Table 1 Demographics and Participant Characteristics

American Indian / Alaska Native 4.3% (n = 2)

Black or African American 27.7% (n = 13)

Note N = 47, *categories were not mutually exclusive

H1: Resilience is Associated with PTCs After Recalling the Index Trauma

The baseline model with the two control covariates explained 19% of the variance in

PCTI at T1, R2 = 19, 95% CI [.00, 38] With the addition of resilience, the theoretical model

explained 38% of the variance, ΔR2 = 19, [.18, 58] The change in f2 (0.31, [0.05, 0.76])

suggested resilience made a medium- sized contribution to the model See Table 3 for the model

comparison statistics of the primary hypotheses In the theoretical model, resilience was

negatively associated with PTCs at Time 1, β = -0.51, p < 001 The coefficients for PTSD

symptom severity and combat exposure were small and not statistically significant See Table 4

for the coefficient summaries for the models of the primary hypotheses

Table 2 Correlation Matrix

Note PTCI = Posttraumatic Cognitions Inventory, T1 = Time 1 (post trauma narrative), NVS = Negative View of the Self, NVW =

Negative Views of the World, SB = self-blame, T2 = Time 2 (post positive emotion induction )

Table 3 Model comparison for primary models

Baseline model F and R2 Theoretical model F and R2 f2 95% CI Predicting PTCs at T1

(post-narrative) F(2, 44) = 5.23, p = 009, R2 = 19 F(3, 43) = 8.84, p < 001, R2 = 38 0.310 [0.05, 0.76]Predicting PTCs at T2

(post-positive video) F(2, 44) = 3.53, p = 038, R

2 = 14 F(3, 43) = 5.24, p = 004, R2 = 27 0.180 [0.01, 0.53] Predicting change in PTCs F(2, 44) = 0.22, p = 801, R2 = 01 F(3, 43) = 0.19, p = 905, R2 = 01 0.003 [0.00, 0.12]

Note Baseline model includes PTSD symptoms and combat exposure as covariates Theoretical model includes

resilience as an independent variable PTCI = Posttraumatic cognitions inventory, 95% CI presented for f2

H2: Resilience is Associated with PTCs After a Positive Mood-Inducing Video

The baseline model with the two control covariates explained 14% of the variance in

PCTI, R2 = 14, 95% CI [0, 31] With the addition of resilience, the theoretical model explained

27% of the variance, ΔR2 = 13, [.07, 47] The change in f2 (0.18, [0.01, 0.53]) suggested

resilience made a medium contribution to the model In the theoretical model, resilience was