Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2022 Apr 1.
Published in final edited form as: J Trauma Stress. 2020 Oct 31;34(2):287–297. doi: 10.1002/jts.22602

Posttraumatic Stress Disorder Symptom Cluster Structure in Prolonged Exposure Therapy and Virtual Reality Exposure

Elizabeth S Stevens 1,2, Kyle J Bourassa 1,3, Aaron M Norr 4,5, Greg M Reger 1,5
PMCID: PMC8035142  NIHMSID: NIHMS1644637  PMID: 33128806

Abstract

The emotional processing theory of posttraumatic stress disorder (PTSD) posits that avoidance is central to PTSD development and maintenance. Prolonged exposure (PE) therapy, which clinically focuses on avoidance reduction, has strong empirical support as a PTSD treatment. Virtual reality exposure (VRE) has been utilized to accelerate avoidance reduction by increasing multisensory engagement. Although some exposure therapy studies have found associations between avoidance and PTSD symptoms, others have indicated that reexperiencing or hyperarousal symptoms drive symptom trajectories. Using a cross-lagged panel design, the present secondary data analysis examined temporal associations between clinician-assessed PTSD symptom clusters during treatment with PE, VRE, or a waitlist control condition. There were no significant differences between PE and VRE regarding symptom clusters at any assessment. Compared to the waitlist condition, individuals who received VRE or PE exhibited earlier reductions in avoidance/numbing symptoms, β = −.19, 95% CI [−.33, −.05], followed by reductions in hyperarousal symptoms, β = −.21, 95% CI [−.33, −.09]. Hyperarousal symptoms predicted changes in later avoidance/numbing and reexperiencing outcomes across treatment: pretreatment to midtreatment, β = .29, 95% CI [.17, .42]; midtreatment to posttreatment, β = .23, 95% CI [.07, .39]. Reexperiencing symptoms predicted changes in hyperarousal outcomes earlier in treatment, β = .22, 95% CI [.02, .37], whereas avoidance/numbing symptoms predicted changes in hyperarousal outcomes later in treatment, β = .18, 95% CI [.04, .32]. These findings support the efficacy of exposure therapy in addressing avoidance/numbing symptoms and highlight the potential importance of hyperarousal symptoms in relation to other symptom clusters.

Posttraumatic stress disorder (PTSD) is a functionally impairing condition that affects an estimated 10%–23% of U.S. military service members, particularly those who served in recent conflicts in Iraq and Afghanistan (Hoge et al., 2014; Milliken et al., 2007). It is also a highly heterogeneous disorder, as symptoms comprise four related yet factor-analytically distinct (King et al., 1998), clusters: reexperiencing, avoidance, numbing and changes in mood and cognition, and hyperarousal (American Psychiatric Association [APA], 2013), although avoidance and numbing were previously combined into a single cluster (APA, 2000). Specifically, research has supported the centrality of avoidance symptoms in maintaining PTSD. Several investigations have shown that avoidance-oriented coping is associated with PTSD symptom severity both concurrently (Bryant & Harvey, 1995) and prospectively (Benotsch et al., 2000; Pineles et al., 2011), highlighting the importance of avoidance as a treatment target.

Exposure-based therapies, such as prolonged exposure (PE; Foa et al., 2007), have demonstrated strong empirical support for the treatment of PTSD (Powers et al., 2010), including among active duty military samples (Foa et al., 2018; Peterson et al., 2020). Prolonged exposure was developed in accord with emotional processing theory (EPT; Foa & Kozak, 1986), which hypothesizes that pathological fear structures, composed of features of the traumatic event (i.e., situation, response, and meaning), elicit distress when activated and contribute to maladaptive responses (i.e., avoidance) of trauma-related stimuli. In this framework, avoidance of trauma-related memories, emotions, or stimuli is theorized to preclude emotional processing and fear habituation and, therefore, maintains PTSD symptoms. Thus, the clinical focus of PE centers on the role and reduction of avoidance via therapeutic exposure to promote emotional processing and fear habituation and achieve symptom reduction. The PE protocol includes both in-session imaginal exposures to the trauma memory as well as in vivo exposures to situations that are perceived as dangerous but are objectively low risk. These exposure components are theorized to lead to symptom reduction by activating the pathological fear structure and incorporating incompatible information to facilitate new learning and promote habituation of the fear response to conditioned stimuli (Craske et al., 2008; Foa et al., 2007). Of these proposed mechanisms of symptom change, between-session habituation of fear responding to trauma-related stimuli (e.g., memories, external reminders) is most consistently associated with treatment outcome, although several studies have also found support for the importance of within-session habituation and fear activation (for reviews, see Brown et al., 2019, and Cooper et al., 2017). Importantly, in order to experience fear activation or habituation, it is necessary to first reduce avoidance.

One proposed method of increasing activation of the fear structure during exposure therapy is the use of virtual reality exposure (VRE), which has frequently been cited as helping to mitigate avoidance by increasing sensory and emotional engagement (e.g., Rothbaum et al., 1995), which, in turn, has been associated with improved treatment outcome in some studies (Foa et al., 1995; Jaycox et al., 1998). Engaging in VRE may be especially helpful in addressing prominent emotional numbing symptoms that are adaptive in a military context but which may lead to engaging in PE in an emotionally detached manner (Reger & Gahm, 2008), which could affect treatment outcomes (Foa et al., 2006). VRE has been implemented using a variety of methodologies, and has been shown in several studies to be effective in the treatment of PTSD (Reger et al., 2011; Rizzo et al., 2011; Rothbaum et al., 2014).

In support of the role of VRE in increasing emotional engagement and reducing avoidance during exposure (Rothbaum et al., 1995), several studies have demonstrated increased psychophysiological arousal during the use of VRE systems. For example, the findings from one study (Costanzo et al., 2014) demonstrated that engaging in standardized combat-related virtual reality (VR) events resulted in heightened arousal in veterans with PTSD relative to combat-exposed veterans without PTSD. In the study by Costanzo et al. (2014), recently deployed military service members with PTSD symptoms who did not meet the full diagnostic criteria for PTSD engaged in three combat-related simulations. Increased heart rate during VR use was associated with PTSD symptom severity. Katz and colleagues (2020) also found that compared to a waitlist control condition, engaging in VRE was associated with more physiological habituation to trauma-relevant stimuli, whereas habituation in the PE condition fell nonsignificantly between that of the waitlist control and VRE conditions; this may indicate increased emotional engagement. The evidence for VRE-facilitated emotional engagement indexed by self-reported subjective units of discomfort (SUDs) has not been as clear (Reger et al., 2018). However, given that VRE directly counters avoidance via multisensory VR stimuli and increases objective measures of arousal (i.e., psychophysiology measures), it remains possible that VRE either more directly targets or more rapidly reduces avoidance symptoms, which may lead to subsequent symptom reductions in other clusters.

Although the use of VRE in PTSD treatment is partly based on the assumption that reducing avoidance is integral to treatment outcome, as it is a prerequisite for fear habituation and the formation of new learning, and meta-analytic studies have demonstrated that PE is an effective treatment for overall PTSD symptoms (Powers et al., 2010), the process by which treatment outcomes are achieved is not fully understood. The findings from some studies have suggested that exposure therapy has unique effects on particular symptom clusters, and this information may facilitate a better understanding of treatment processes. Consistent with the theoretical links among exposure, avoidance reduction, and subsequent overall treatment outcomes, for example, Taylor and colleagues (2003) found that compared to either relaxation training or Eye Movement Desensitization and Reprocessing therapy, PE resulted in larger reductions in both avoidance and reexperiencing symptoms and faster reductions in avoidance symptoms. In another investigation (Bryant et al., 2003), trauma survivors were randomly assigned to either imaginal exposure alone, imaginal exposure plus cognitive restructuring, or a therapeutic control condition (i.e., supportive counseling). The study results indicated that imaginal exposure, with or without cognitive restructuring, led to larger reductions in both avoidance and reexperiencing symptoms compared to supportive counseling.

Despite the theoretical role of avoidance and the association between avoidant coping and PTSD symptoms, the trajectory of symptoms over time may not always be driven primarily by avoidance. The results of two studies of the natural course of PTSD after traumatic events suggested that hyperarousal symptoms, and perhaps reexperiencing symptoms, contribute to symptoms in other clusters over time (Marshall et al., 2006; Schell et al., 2004). These findings studies suggest that hyperarousal symptoms, rather than avoidance symptoms, are the best predictor of PTSD symptom severity over time. Other studies have examined changes in the trajectory of PTSD symptom clusters during treatment. One study compared differences in the trajectories of changes in symptom clusters for individuals in PE versus cognitive processing therapy (Nishith, Resick, & Griffin, 2002). Few differences emerged in the patterns of symptom change between the two therapies, and, in PE, all three symptom clusters were best described by a quadratic function. However, although it was not explicitly tested by the authors, graphical representations of findings indicate that peak levels of avoidance symptoms appeared to occur before peak levels of reexperiencing symptoms (Session 2 vs. Session 4), potentially suggesting temporal precedence of changes in avoidance symptoms.

Another study (Maples-Keller et al., 2017) examined the associations between symptom clusters across five sessions of imaginal exposure, conducted via VRE, among veterans of military conflicts in Iraq and Afghanistan. The results indicated that a model in which changes in previous reexperiencing—but not avoidance—symptoms were allowed to predict the other three symptom clusters over the course of treatment best fit the data, suggesting that changes in reexperiencing symptoms may drive treatment effects in VRE treatment for PTSD. The exposure-based intervention in the study consisted of only in-session imaginal exposure and did not include the in vivo exposures typically utilized in PE. Over time, in vivo exposures may contribute to larger reductions in avoidance symptoms than imaginal exposures (Richards et al., 1994). In vivo exposures may increase the variance that could be explained by changes in avoidance symptoms and, thus, the influence of changes in avoidance over other symptom clusters. Nevertheless, it is possible that avoidance is critical in theory and associated with symptom maintenance but that the course of symptoms during treatment is driven by changes in symptoms other than avoidance.

Although avoidance clearly plays an important role in the phenomenology and treatment of PTSD, there is no consensus as to its prominence or temporal precedence relative to other symptom clusters. Examining temporal changes in PTSD symptom clusters in relation to each other over the course of treatment may yield insight into mechanisms of change involved in PE. For example, if reductions in avoidance symptoms predict reductions in reexperiencing symptoms, this would imply that reducing avoidance by engaging in emotional material, via exposures, may allow for more emotional processing, thus reducing the reexperiencing of and reactivity to trauma-related stimuli. If reductions in avoidance predict reductions in hyperarousal symptoms, it may indicate that reducing avoidance leads to increased contextual discrimination between memories of traumatic events and present situations. If changes in avoidance symptoms do not predict changes in other symptom clusters, this would also be useful in considering strategies to improve treatment outcomes, such as interventions that explicitly targeting other symptom clusters. Furthermore, knowledge of these treatment processes may enhance providers’ abilities to recognize and respond to early signs of treatment nonresponse.

The present study was a secondary analysis of data from Reger et al. (2016), a randomized controlled trial of PE, VRE, and a waitlist control condition. The present study aimed to examine the associations between PTSD symptom clusters, per the fourth edition (text revision) of the Diagnostic and Statistical Manual of Mental Disorders (APA, 2000), over the course of treatment with PE compared to VRE or a waitlist control condition, using a cross-lagged panel design. Given limited evidence for differences in treatment outcomes between participants who engaged in PE versus VRE in the present sample (Reger et al., 2016), we hypothesized that the symptom structure would be similar across both treatment conditions. Although the literature is mixed, based on the theoretical and empirically supported role of avoidance in PTSD symptom maintenance, we expected that among the combined group of individuals who were assigned to VRE and PE, henceforth referred to as the “exposure therapy condition,” changes in avoidance and numbing symptoms would predict future changes in the other two symptom clusters across measurements.

Method

Participants

In the original study from which these data are drawn, active duty U.S. Army soldiers (N = 162) with PTSD that stemmed from deployments to Iraq or Afghanistan were enrolled in a randomized clinical trial comparing VRE, PE, and a waitlist control condition. All soldiers were diagnosed with PTSD using the Clinician-Administered PTSD Scale (CAPS; Blake et al., 1995), following criteria in the DSM-IV-TR (APA, 2000). Additional inclusion and exclusion criteria can be found in Reger et al (2016). Most participants were male (96.3%, n = 156), Caucasian (59.9%, n = 97), and had completed some college (66.0%, n = 107). All demographic information is reported in Reger et al. (2016), and relevant demographic and clinical characteristics are reported in Table 1.

Table 1.

Clinician-Administered Posttraumatic Stress Disorder (PTD) Scale (CAPS-IV) Symptom Cluster Scores for Each Treatment Group

Waitlist VRE PE Total
M SD n M SD n M SD n M SD N Cronbach’s α
Reexperiencing
 Baseline 9.81 4.37 54 9.59 4.60 54 9.94 3.89 54 9.78 4.27 162 .80
 Midtreatment 9.86 4.43 52 9.51 4.68 36 9.10 4.98 39 9.53 4.65 127 .85
 Posttreatment 8.18 5.11 47 7.48 5.91 30 5.34 5.53 32 7.16 5.55 109 .92
Avoidance/numbing
 Baseline 15.89 4.27 54 16.34 3.63 54 15.25 4.56 54 15.83 4.17 162 .72
 Midtreatment 14.40 5.52 52 12.86 5.54 36 11.92 6.32 39 13.20 5.83 127 .88
 Posttreatment 12.94 6.22 47 9.77 6.81 30 7.69 7.16 32 10.52 6.98 109 .91
Hyperarousal
 Baseline 13.75 3.08 54 14.29 2.59 54 13.94 2.69 54 13.99 2.79 162 .65
 Midtreatment 13.11 3.25 52 13.22 3.28 36 11.49 5.09 39 12.64 3.96 127 .82
 Posttreatment 12.91 3.05 47 11.28 4.88 30 9.11 5.69 32 11.35 4.71 109 .87
Open in a new tab

Note. VRE = virtual reality exposure; PE = prolonged exposure.

Procedure

The study procedures were identical to those reported in greater detail in Reger et al. (2016). The study was approved by the local institutional review board, written informed consent was obtained, and the study was registered at ClinicalTrials.gov (identifier: NCT01193725). First, participants completed informed consent, followed by a baseline assessment of inclusion and exclusion criteria, demographic information, self-report surveys, the baseline CAPS assessment, and a physiological assessment. Assessment points for the CAPS included a midtreatment assessment, which was performed after the fifth session of therapy for participants in the exposure therapy condition or after 2.5 weeks for those in the waitlist condition, and a posttreatment assessment, which was administered after Session 10 for the therapy group and after 5 weeks for individuals assigned to the waitlist condition. On average, participants completed the midtreatment assessment 31.2 days (SD = 19.5) after initiating treatment and completed the posttreatment assessment 70.9 days (SD = 36.97) after initiating treatment. Three-month and 6-month follow-up CAPS assessments were also conducted but not used in the present study.

Exposure therapy was conducted according to the published PE treatment manual (Foa et al., 2007), including standard PE imaginal and in vivo exposure homework, and consisted of 10 sessions, conducted approximately weekly, each lasting 90–120 min. Participants in the VRE condition received identical treatment to those who received PE except that imaginal exposures were conducted using VR. A head-mounted display and high-fidelity over-the-ear headphones were worn while the participant was seated or standing on a platform that vibrated with low-frequency sounds. The clinician used the VR interface to simultaneously alter the VR environment or stimuli to match the trauma narrative while clients recounted the memory aloud. In vivo exposure was delivered identically for both conditions. For more detail on the Virtual Iraq/Afghanistan system, see Rizzo and colleagues (2009). Participants assigned to the waitlist condition participated in assessments but received no active treatment during a 5-week waiting period. Following baseline assessment and study enrollment, the research coordinator used a computerized random number generator to assign treatment conditions.

Measures

The Clinician Administered PTSD Scale (CAPS; Blake et al., 1995) is a structured clinical interview that is used to assess the presence and severity of PTSD according to DSM-IV-TR criteria (i.e., five reexperiencing symptoms, seven avoidance and numbing symptoms, and five hyperarousal symptoms). The frequency and intensity of each symptom are coded on a scale ranging from 0 to 4, and for the present study, the frequency and intensity ratings were summed. Scores, therefore, ranged from 0 to 136, with higher scores indicating a higher level of PTSD symptoms. This study utilized the CAPS “last week” reference at all assessments, such that participants reported on symptoms experienced during the past week. The internal consistency for the CAPS-IV was good at baseline, Cronbach’s α = .81, and excellent at midtreatment and posttreatment, Cronbach’s αs = .94 and .96, respectively. The internal consistency for each symptom cluster and assessment point is given in Table 1.

Data Analysis

In the current secondary analyses, participants’ PTSD symptoms, as measured using the CAPS-IV, were assessed by trained clinicians at three assessment points: baseline, midtreatment, and posttreatment. Of the 162 participants with CAPS data at baseline, 127 provided data at the midpoint assessment and 109 completed either 10 sessions of exposure therapy or the waitlist period and provided data at posttreatment (see Table 1). We compared changes in PTSD symptom scores at the cluster level between PE, VRE, and the waitlist control condition over time using a cross-lagged structural equational model (SEM). To specify these models, we followed several steps of specification and model evaluation (Kaplan, 2008). Symptom cluster scores were first regressed on their previous assessment scores to create an autoregression only model. We next added within-occasion correlations between the three clusters at the midtreatment and posttreatment assessments and cross-lagged associations among the three clusters. We assessed the model fit using a number of commonly used measures of model fit, including the comparative fit index (CFI), root mean square error of approximation (RMSEA), Tucker–Lewis Index (TLI), standardized root mean square residual (SRMR), as well as the Akaike information criterion (AIC) and Bayesian information criterion (BIC) for each model. To assess goodness of fit, we used the standard cutoffs of CFI and TLI values greater than .95, RMSEA values less than .06, and SRMR values less than .08 (Hu & Bentler, 1999); in addition, lower AIC and BIC values indicated better model fit.

After specifying this baseline cross-lagged model, we tested the impact of setting the autoregressions for each symptom cluster (i.e., baseline to midtreatment; midtreatment to posttreatment) to equality in a stepwise manner. These restrictions to the model were compared against the less restrictive nested model using chi-square difference testing. We retained constraints when they did not negatively impact the model fit. Treatment condition was coded using two orthogonal contrast codes, first comparing PE coded as 1 and VRE coded as −1 (waitlist control was coded as 0), then comparing PE and VRE, as a combined exposure therapy treatment condition, coded as 1 and the control condition coded as −2. These two contrast codes were then used to predict PTSD cluster scores at midtreatment and posttreatment. This allowed us to examine differences in how the PTSD symptom clusters changed over time between conditions (i.e., directly comparing VRE to PE, then treatment to control) within the cross-lagged panel model (Cole & Maxwell, 2003) as well as the timing of direct and indirect treatment effects. We used full information maximum likelihood (FIML) to account for missing data; this method produces estimates that are unbiased when data are missing at random (Graham, 2009), and prior studies have established a lack of evidence that data are not missing at random in this sample (Reger et al., 2016). Finally, we used maximum likelihood estimation when running all regression models in Mplus (Version 7.31) and bias-corrected bootstrapping (N = 1,000) in all mediation models (Muthén & Muthén, 1998–2012).

Results

Model Specification

To specify our model, we started from a model that first included all autoregressions, then included the within-occasion associations among cluster residuals at the midtreatment and posttreatment assessments, and, finally, included cross-lagged associations between the three clusters. We then tested whether constraining autoregressions within the symptom clusters reduced the model fit. These constraints did not reduce the model fit and, as a result, were retained in the final model. We then included the direct association of the treatment contrast codes on each of the cluster scores at the midtreatment and posttreatment assessments. Table 2 provides the full model fit results for this process. Our final model fit the data well, χ2(15, N = 162) = 19.37, p = .197, CFI = .99, TLI = .98, RMSEA = .042, 90% CI [.000, .091], SRMR = .038, AIC = 6,416.10, B,IC = 6592.10. See Figure 1 for a conceptual representation of the final model.

Table 2.

Model Fit During Each Step of Model Specification

Change in model specification df χ2 p CFI TLI RMSEA SRMR AIC BIC
Full unconstrained autoregressive model only 21 181.46 < .001 .72 .56 .217 .241 6,564.82 6,666.71
Adding within-occasion correlations among symptom residuals 18 70.52 < .001 .91 .83 .134 .162 6,459.88 6,571.03
Adding cross-lags among cluster symptoms 6 10.41 .108 .99 .96 .067 .025 6,423.77 6,571.98
Constraining reexperiencing autoregressions 7 12.07 .098 .99 .96 .067 .033 6,423.43 6,568.54
Constraining hyperarousal autoregressions 8 12.82 .118 .99 .97 .071 .038 6,422.18 6,564.21
Constraining avoidance/numbing autoregressions 9 13.77 .131 .99 .97 .057 .038 6,421.13 6,560.07
Adding treatment contrast codes 15 19.37 .197 .99 .98 .042 .038 6,416.10 6,592.10
Open in a new tab

Note. Model results reflect the overall model fit at each subsequent step in the model specification process. Constraints reflect constraining the specified paths to equality. CFI = comparative fit index; RMSEA = root mean squared error of approximation; TLI = Tucker–Lewis index; SRMR = standardized root mean square residual; AIC = Akaike information criterion; BIC = Bayesian information criterion.

Figure 1.

Figure 1

Open in a new tab

Conceptual Depiction of the Cross-Lagged Panel Model

Note. PE = prolonged exposure; VRE = virtual reality exposure; tx = treatment.

Timing and Magnitude of Treatment Effects on PTSD Cluster Scores

After specifying the final model, we examined the associations of interest. We first examined the effect of PE, VRE, and the waitlist condition on PTSD cluster scores at the midtreatment and posttreatment assessments. As expected, there were no significant differences between PE and VRE at the midtreatment or posttreatment assessment with regard to any of the PTSD symptom clusters, βs = .001–.16, ps = .076–.924.

Next, we examined the effect of exposure therapy (i.e., combined treatment condition, either standard PE or PE and VRE) on PTSD cluster scores at posttreatment compared to the control condition. There was a significant effect of exposure therapy on avoidance and numbing symptoms at midtreatment, β = −.19, 95% CI [−.33, −.05], p = .006. There were not, however, significant exposure therapy effects on either the reexperiencing, p = .337, or hyperarousal, p = .086, clusters at midtreatment. We next examined the treatment effects on posttreatment PTSD cluster scores. Again, there was a significant effect of exposure therapy on avoidance and numbing symptoms, β = −.20, 95% CI [−.32, −.08], p = .001, as well as a significant effect on hyperarousal symptoms, β = −.21, 95% CI [−.33, −.09], p < .001, but not a significant effect on reexperiencing symptoms, β = −.08, 95% CI [−.20, .04], p = .202 (see Figure 2).

Figure 2.

Figure 2

Open in a new tab

Between-Group Differences in PTSD Symptom Cluster Scores for (A) Avoidance/Numbing, (B) Hyperarousal, and (C) Reexperiencing Symptoms on the CAPS-IV

Note. Error bars represent the standard error of the estimate. VRE = virtual reality exposure; PE = prolonged exposure.

Taken together, these results suggest that PE and VRE largely operate similarly regarding the timing of their impact on PTSD cluster scores. Furthermore, participants who engaged in PE and VRE evidenced decreased avoidance and numbing symptoms at midtreatment, decreased avoidance and numbing symptoms and hyperarousal symptoms at posttreatment, and no significant effects on reexperiencing symptoms when compared to those in the waitlist control condition.

Mediation Effects of Treatment on PTSD Cluster Scores

We next examined whether changes in any PTSD cluster scores at midtreatment mediated the effect of treatment on any of the other cluster scores at posttreatment. There was an indirect effect of exposure therapy on avoidance and numbing symptoms at posttreatment through midtreatment avoidance and numbing, β = −0.08, 95% CI [−0.14, −0.02], p = .012, but no other significant indirect effects of treatment (i.e., PE vs. VRE; combined PE and VRE vs. control) through midtreatment cluster scores on PTSD symptom cluster outcomes, βs = .002–.07, ps = .088–.793. These results suggest that there is not a differential effect of changes in one cluster before other cluster outcomes during PE and VRE compared to a control condition.

PTSD Cluster Associations Over Time

Finally, we examined the cross-lagged associations between the PTSD cluster scores over time across all treatment groups. The model results indicated that avoidance/numbing cluster scores significantly predicted changes in later hyperarousal cluster outcomes, but only between midtreatment and posttreatment, β = .18, 95% CI [.04, .32], p = .012. Hyperarousal cluster scores predicted changes in later avoidance and numbing cluster scores, β = .29, 95% CI [.17, .42], p < .001 for pretreatment to midtreatment, and β = .23, 95% CI [.07, .39], p = .005 for midtreatment to posttreatment. Hyperarousal cluster scores also predicted changes in later reexperiencing outcomes between pretreatment and midtreatment, β = .15, 95% CI [.002, .29], p = .046, although this association was not significant between midtreatment and posttreatment, β = .16, 95% CI [−.01, .34], p = .062. Reexperiencing cluster scores predicted changes in later hyperarousal outcomes between pretreatment and midtreatment, β = .22, 95% CI [.02, .37], p = .004, but this association was not significant between midtreatment and posttreatment. There were no associations between cluster scores for avoidance and numbing symptoms and reexperiencing symptoms across time. The full model results are presented in Table 2. In summary, there were two relevant findings. First, hyperarousal cluster scores predicted changes in later avoidance and numbing outcomes and reexperiencing outcomes across treatment. Second, reexperiencing cluster scores predicted changes in hyperarousal outcomes earlier in treatment, whereas avoidance and numbing cluster scores only predicted changes in hyperarousal outcomes later in treatment.

Discussion

The current study was a secondary data analysis that examined the effect of PE, VRE, and waitlist control conditions on PTSD cluster scores at assessments conducted midtreatment and posttreatment time points. Predicting no significant differences between the two active treatment groups, we examined the effect of exposure therapy (i.e., usual PE and PE with VRE groups combined) on PTSD cluster scores at posttreatment compared to a waitlist control condition. In the current study, we examined whether changes in any PTSD cluster scores at midtreatment mediated the effect of treatment on any other cluster scores at posttreatment and assessed the cross-lagged associations between PTSD cluster scores over time across all treatment groups.

Consistent with prior findings from this sample (Reger et al., 2016), there were no significant differences between PE and VRE when symptom outcomes were examined by PTSD symptom clusters. Both cross-sectional and mediation analyses failed to identify significant differences between the two active treatment groups (i.e., PE vs. VRE) regarding symptom cluster scores across treatment, which may suggest that PE and VRE share similar mechanisms and efficacy. When PE and VRE were combined and compared to the waitlist control condition, however, interesting results emerged. Compared to participants in the waitlist control condition, those who engaged in exposure therapy evidenced reduced avoidance and numbing symptoms at both midtreatment and posttreatment as well as reduced hyperarousal symptoms by posttreatment. Although changes in avoidance and numbing appeared to temporally precede changes in hyperarousal when compared to the waitlist control, and changes in hyperarousal symptoms predicted changes in avoidance and reexperiencing over time, changes in avoidance and numbing symptoms only predicted subsequent changes in hyperarousal symptoms after midtreatment. Furthermore, the results of mediation analyses indicated that changes in particular symptom clusters did not differentially predict posttreatment outcomes in any other cluster for exposure therapy versus waitlist control.

It is also consistent with some previous research on the beneficial effects of exposure therapy with regard to avoidance symptoms early in therapy (Taylor et al., 2003) and compared to other treatments (Bryant et al., 2003). This finding is consistent with the clinical emphasis on addressing avoidance early in the PE treatment protocol (Foa et al., 2007). Patients who participate in PE receive education on the role of avoidance during the first treatment session and this psychoeducation is reinforced throughout treatment. In vivo exposure is planned during the second session, and patients begin to reduce avoidance by intentionally approaching previously avoided situations or circumstances. Imaginal exposure begins in the third session to interfere with avoidance of the trauma memory and associated emotions. After five sessions of PE, patients and providers have dedicated a substantial effort to intentionally reducing avoidance, and it is not surprising that the midpoint assessment reflects the benefits of these efforts. By midtreatment, substantial new information has been incorporated into the fear structure regarding, for example, the difference between thinking about and experiencing traumatic events, the patient’s sense of accomplishment and competence, the relative threat of various environments and stimuli, and information learned from the experience of habituation; these together may reduce patients’ perceived need or desire to avoid these stimuli.

Exposure therapy took longer to impact hyperarousal symptoms. Hyperarousal symptom cluster scores were not significantly different for participants in the treatment group and those assigned to the waitlist condition until posttreatment. It is possible that the early, explicit emphasis the treatment places on approaching painful memories and anxiety-provoking situations rapidly reduces avoidance symptoms, whereas the interventions in exposure therapy take longer to impact arousal. Specifically, it may take time and repetition for the emotional processing associated with reducing avoidance to interfere with the strength of the conditioned fear response and improve symptoms such as insomnia, anger, concentration, hypervigilance, and startle response. It is unclear why the results of the present study did not indicate a significant impact of exposure therapy on reexperiencing symptoms relative to the waitlist control condition. It may simply be that exposure therapy conducted with this sample simply exerted a smaller effect on reexperiencing symptoms. Alternatively, reductions in avoidance and hyperarousal symptoms may increase patients’ ability and willingness to engage more fully in trauma memories; thus, reductions in reexperiencing symptoms may follow reductions in other symptom clusters. Consistent with this idea, the results of a recent meta-analysis demonstrated that reexperiencing symptoms evidenced the smallest effect size across exposure-based therapy compared to the other symptom clusters and that symptoms further decreased to a significant degree between posttreatment and follow-up assessments (Phelps et al., 2018). This finding should be explored within other samples of active duty military personnel and other trauma-exposed populations.

Avoidance and numbing symptoms did not predict later improvements in reexperiencing symptoms and only predicted hyperarousal symptoms later in treatment. Instead, hyperarousal appeared to predict both avoidance and numbing symptoms and reexperiencing symptoms, whereas only reexperiencing symptoms, then subsequently avoidance and numbing symptoms, influenced hyperarousal symptoms. This pattern of results could reflect the fact that both avoidance and numbing and hyperarousal symptoms significantly decreased by posttreatment, whereas reexperiencing symptoms were relatively more stable over time. These findings are similar to those reported in two studies of the natural course of PTSD that found hyperarousal symptoms and reexperiencing symptoms contributed to symptoms in other clusters over time in the absence of intervention (Marshall et al., 2006; Schell et al., 2004). The results of a study that included five sessions of VR exposure without in vivo exposure (Maples-Keller et al., 2017) also showed that changes in reexperiencing symptoms may drive treatment effects in exposure therapy for PTSD. The present study’s use of 10 sessions of exposure therapy and inclusion of in vivo homework may contribute to some of the observed differences.

These results should be understood within the limitations of the cross-lagged panel structural equation models used in the current study. Although cross-lagged panel models are a more sophisticated method to examine change over time when compared to examining only pre–post change, alternative model types or model specification procedures might produce different results. Cross-lagged panel models can also produce estimates that are difficult to interpret both practically and theoretically (Berry & Willoughby, 2016), particularly if the PTSD symptoms included in the study are more trait-like constructs, which results in conflating within- and between-person variance (Hamaker et al., 2015). Of note, PTSD symptoms have been shown to demonstrate change with regard to their mean level but not their factor structure over time (Elhai & Palmieri, 2011; King et al., 2009). Although PTSD symptoms, broadly measured, appear to have factorial invariance over time (King et al., 2009; McDonald et al., 2008), this assumption within the current models may bias the interpretation of our results. This said, the current model specification and interpretation reflect the goal of examining the change in PTSD symptom clusters over time and match the models used in other such investigations (e.g., Maples-Keller et al., 2017).

The present study included active duty U.S. soldiers with PTSD stemming from deployments to Iraq and Afghanistan. It is unknown whether symptom clusters would improve similarly among other trauma-exposed populations. Similarly, our sample was predominantly male, and the course of symptom improvement among women soldiers with PTSD should be examined with future research. The present study compared the effects of exposure therapy to a waitlist and did not include other non–exposure-based psychotherapies. Nisith and colleagues (2002) found similar patterns of symptom change for PE and CPT across treatment. Accordingly, we do not know if these findings are specific to exposure therapy or whether they reflect the effects of psychotherapy more generally. In addition, this study was conducted before the release of the fifth edition of the DSM (i.e., DSM-5), and our symptom clusters correspond with the DSM-IV conceptualization of PTSD. Although our study is easily interpretable within the PTSD body of literature, which predominantly reflects the DSM-IV, future studies that examine the course of symptom change during treatment following current DSM-5 criteria will be important to our understanding of treatments, mechanisms, and patients. Finally, this study only included a pretreatment, midtreatment, and posttreatment assessment. Future research could increase the frequency of symptom assessment across treatment. More frequent assessment may provide a more precise analysis of symptom changes over time. Future research could also examine differences in symptom change across a broader range of PTSD treatments to include evidence-based, non–exposure-based therapies, such as present-centered therapy.

Should these findings prove stable in future research, providers delivering exposure therapy for PTSD may wish to monitor hyperarousal and reexperiencing symptoms given the influence they may have on the improvement of other symptom clusters later in treatment. The results of our study suggest that clinicians may observe improvement in avoidance symptoms first. However, improvement in hyperarousal and reexperiencing symptoms may predict improvement in PTSD symptoms later. Future research on the clinically relevant magnitude of change and the precise timing of these changes may be helpful. Such research would likely be enhanced by including a larger number of assessments at pretreatment, during treatment, and at posttreatment time points to differentiate between mechanisms of development and maintenance versus mechanisms of treatment.

Table 3.

Results of the Cross-Lagged Panel Model Examining Associations Between Symptom Clusters Across Assessments.

Pretreatment to Midtreatment Midtreatment to Posttreatment
B 95% CI p B 95% CI p
Predicting Avoidance/Numbing
 Autoregressive 0.34 [0.20, 0.48] < .001 0.39 [0.25, 0.54] < .001
 Hyperarousal 0.29 [0.17, 0.42] < .001 0.23 [0.07, 0.39] .005
 Reexperiencing 0.07 [−0.08, 0.22] .330 0.12 [−0.02, 0.26] .098
Predicting Hyperarousal
 Avoidance/Numbing 0.01 [−0.16, 0.18] .900 0.18 [0.04, 0.32] .012
 Autoregressive 0.35 [0.23, 0.48] < .001 0.43 [0.26, 0.59] < .001
 Reexperiencing 0.22 [0.02, 0.37] .004 0.11 [−0.03, 0.24] .133
Predicting Reexperiencing
 Avoidance/Numbing 0.02 [−0.14, 0.19] .780 0.07 [−0.09, 0.23] .370
 Hyperarousal 0.15 [0.002, 0.29] .046 0.16 [−0.01, 0.34] .062
 Autoregressive 0.52 [0.40, 0.64] < .001 0.50 [0.37, 0.64] < .001
Open in a new tab

Open Practices Statement.

This secondary analysis was not formally preregistered. Neither the data nor the materials have been made available on a permanent third-party archive. Requests for the data or materials can be sent via email to the corresponding author and will be subject to review and approval by applicable oversight bodies.

Acknowledgments

This research, a secondary analysis of previously collected data, was supported by the U.S. Army Medical Research and Materiel Command Military Operational Medicine Research Program (W81XWH-08-2-0015) and supported in part by a National Institute on Aging (NIA) training grant (T32-AG000029) awarded to Kyle J. Bourassa. The opinions or assertions contained herein are the private views of the authors and are not to be construed as official or reflecting the views of the U.S. Department of Army, Department of Defense, or Department of Veterans Affairs.

References

  1. American Psychiatric Association. (2000). Diagnostic and statistical manual of mental disorders (4th ed., text rev.). Author. [Google Scholar]
  2. American Psychiatric Association. (2013). Diagnostic and statistical manual of mental disorders (5th ed.). Author. [Google Scholar]
  3. Benotsch EG, Brailey K, Vasterling JJ, Uddo M, Constans JI, & Sutker PB (2000). War zone stress, personal and environmental resources, and PTSD symptoms in gulf war veterans: A longitudinal perspective. Journal of Abnormal Psychology, 109(2), 205–213. 10.1037/0021-843X.109.2.205 [DOI] [PubMed] [Google Scholar]
  4. Berry D, & Willoughby MT (2017). On the practical interpretability of cross‐lagged panel models: Rethinking a developmental workhorse. Child Development, 88(4), 1186–1206. 10.1111/cdev.12660 [DOI] [PubMed] [Google Scholar]
  5. Blake DD, Weathers FW, Nagy LM, Kaloupek DG, Gusman FD, Charney DS, & Keane TM, 1995. The development of a clinician-administered PTSD scale. Journal of Traumatic Stress, 8(1), 75–90. 10.1007/bf02105408 [DOI] [PubMed] [Google Scholar]
  6. Brown LA, Zandberg LJ, & Foa EB (2019). Mechanisms of change in prolonged exposure therapy for PTSD: Implications for clinical practice. Journal of Psychotherapy Integration, 29(1), 6–14. 10.1037/int0000109 [DOI] [Google Scholar]
  7. Bryant RA, & Harvey AG (1995). Avoidant coping style and posttraumatic stress following motor vehicle accidents. Behaviour Research and Therapy, 33(6), 631–-635. 10.1016/0005-7967(94)00093-Y [DOI] [PubMed] [Google Scholar]
  8. Bryant RA, Moulds ML, Guthrie RM, Dang ST, & Nixon RDV (2003). Imaginal exposure alone and imaginal exposure with cognitive restructuring in treatment of posttraumatic stress disorder. Journal of Consulting and Clinical Psychology, 71(4), 706–712. 10.1037/0022-006X.71.4.706 [DOI] [PubMed] [Google Scholar]
  9. Cole DA, & Maxwell SE (2003). Testing mediational models with longitudinal data: Questions and tips in the use of structural equation modeling. Journal of Abnormal Psychology, 112(4), 558–577. 10.1037/0021-843X.112.4.558 [DOI] [PubMed] [Google Scholar]
  10. Cooper AA, Clifton EG, & Feeny NC (2017). An empirical review of potential mediators and mechanisms of prolonged exposure therapy. Clinical psychology review, 56, 106–121. 10.1016/j.cpr.2017.07.003 [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Costanzo ME, Leaman S, Jovanovic T, Norrholm SD, Rizzo AA, Taylor P, & Roy MJ (2014). Psychophysiological response to virtual reality and subthreshold posttraumatic stress disorder symptoms in recently deployed military. Psychosomatic Medicine, 76(9), 670–677. 10.1097/PSY.0000000000000109 [DOI] [PubMed] [Google Scholar]
  12. Craske MG, Kircanski K, Zelikowsky M, Mystkowski J, Chowdhury N, & Baker A (2008). Optimizing inhibitory learning during exposure therapy. Behaviour Research and Therapy, 46(1), 5–27. 10.1016/j.brat.2007.10.003 [DOI] [PubMed] [Google Scholar]
  13. Elhai JD, & Palmieri PA (2011). The factor structure of posttraumatic stress disorder: A literature update, critique of methodology, and agenda for future research. Journal of Anxiety Disorders, 25(6), 849–854. 10.1016/j.janxdis.2011.04.007 [DOI] [PubMed] [Google Scholar]
  14. Foa EB, Hembree EA, & Rothbaum BO (2007). Prolonged exposure therapy for PTSD: Emotional processing of traumatic experiences: Therapist guide. Oxford University Press. 10.1093/med:psych/9780195308501.001.0001 [DOI] [Google Scholar]
  15. Foa EB, Huppert JD, & Cahill SP (2006). Emotional processing theory: An update. In Rothbaum BO (Ed.), Pathological anxiety: Emotional processing in etiology and treatment (pp. 3–24). Guilford Press. [Google Scholar]
  16. Foa EB, & Kozak MJ (1986). Emotional processing of fear: Exposure to corrective information. Psychological Bulletin, 99(1), 20–35. 10.1037/0033-2909.99.1.20 [DOI] [PubMed] [Google Scholar]
  17. Foa EB, McLean CP, Zang Y, Rosenfield D, Yadin E, Yarvis JS, Mintz J, Young-McCaughan S, Borah EV, Dondanville KA, Fina BA, Hall-Clark BN, Lichner T, Litz BT, Roache J, Wright EC, & Peterson AL, for the STRONG STAR Consortium. (2018). Effect of prolonged exposure therapy delivered over 2 weeks vs. 8 weeks vs. present-centered therapy on PTSD symptom severity in military personnel: A randomized clinical trial. JAMA: Journal of the American Medical Association, 319(4), 354–364. 10.1001/jama.2017.21242 [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Foa EB, Riggs DS, Massie ED, & Yarczower M (1995). The impact of fear activation and anger on the efficacy of exposure treatment for posttraumatic stress disorder. Behavior Therapy, 26(3), 487–499. 10.1016/S0005-7894(05)80096-6 [DOI] [Google Scholar]
  19. Foa EB, Steketee G, & Rothbaum BO (1989). Behavioral/cognitive conceptualizations of posttreatmenttraumatic stress disorder. Behavior Therapy, 20(2), 155–176. 10.1016/S0005-7894(89)80067-X [DOI] [Google Scholar]
  20. Graham JW (2009). Missing data analysis: Making it work in the real world. Annual Review of Psychology, 60(1), 549–576. 10.1146/annurev.psych.58.110405.085530 [DOI] [PubMed] [Google Scholar]
  21. Hamaker EL, Kuiper RM, & Grasman RPPP (2015). A critique of the cross-lagged panel model. Psychological Methods, 20(1), 102–116. 10.1037/a0038889 [DOI] [PubMed] [Google Scholar]
  22. Hoge CW, Riviere LA, Wilk JE, Herrell RK, & Weathers FW (2014). The prevalence of posttraumatic stress disorder (PTSD) in U.S. combat soldiers: A head-to-head comparison of DSM-5 versus DSM-IV-TR symptom criteria with the PTSD checklist. The Lancet Psychiatry, 1(4), 269–277. 10.1016/S2215-0366(14)70235-4 [DOI] [PubMed] [Google Scholar]
  23. Hu LT, & Bentler PM (1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling, 6(1), 1–55. 10.1080/10705519909540118 [DOI] [Google Scholar]
  24. Jaycox LH, Foa EB, & Morral AR (1998). Influence of emotional engagement and habituation on exposure therapy for PTSD. Journal of Consulting and Clinical Psychology, 66(1), 185–192. 10.1037/0022-006X.66.1.185 [DOI] [PubMed] [Google Scholar]
  25. Kaplan D (2008). Structural equation modeling: Foundations and extensions (Vol. 10). Sage Publications. [Google Scholar]
  26. Katz AC, Norr AM, Buck B, Fantelli E, Edwards-Stewart A, Koenen-Woods P, Zetocha K, Smolenzki DJ, Holloway K, Rothbaum BO, Difede J, Rizzo A, Skopp N, Mishind M, Gahm G, Reger GM, & Andrasik F (2020). Changes in physiological reactivity in response to the trauma memory during prolonged exposure and virtual reality exposure therapy for posttraumatic stress disorder. Psychological Trauma: Theory, Research, Practice, and Policy. Advance online publication. 10.1037/tra0000567 [DOI] [PubMed] [Google Scholar]
  27. King DW, Leskin GA, King LA, & Weathers FW (1998). Confirmatory factor analysis of the clinician-administered PTSD Scale: Evidence for the dimensionality of what if posttraumatic stress disorder. Psychological Assessment, 10(2), 90–96. 10.1037/1040-3590.10.2.90 [DOI] [Google Scholar]
  28. King DW, Orazem RJ, Lauterbach D, King LA, Hebenstreit CL, & Shalev AY (2009). Factor structure of posttraumatic stress disorder as measured by the Impact of Event Scale–Revised: Stability across cultures and time. Psychological Trauma: Theory, Research, Practice, and Policy, 1(3), 173–187. 10.1037/a0016990 [DOI] [Google Scholar]
  29. McDonald SD, Beckham JC, Morey R, Marx C, Tupler LA, & Calhoun PS (2008). Factorial invariance of posttraumatic stress disorder symptoms across three veteran samples. Journal of Traumatic Stress, 21(3), 309–317. 10.1002/jts.20344 [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Maples-Keller J, Price M, Rauch S, Gerardi M, & Rothbaum BO (2017). Investigating relationships between PTSD symptom clusters within virtual reality exposure therapy for OEF/OIF veterans. Behavior Therapy, 48(2), 147–155. 10.1016/j.beth.2016.02.011 [DOI] [PubMed] [Google Scholar]
  31. Marshall GN, Schell TL, Glynn SM, & Shetty V (2006). The role of hyperarousal in the manifestation of posttraumatic psychological distress following injury. Journal of Abnormal Psychology, 115(3), 624–628. 10.1037/0021-843X.115.3.624 [DOI] [PubMed] [Google Scholar]
  32. Milliken CS, Auchterlonie JL, & Hoge CW (2007). Longitudinal assessment of mental health problems among active and reserve component soldiers returning from the Iraq war. Journal of the American Medical Association, 298(18), 2141–2148. 10.1001/jama.298.18.2141 [DOI] [PubMed] [Google Scholar]
  33. Muthén LK, & Muthén BO (1998—2012). Mplus user’s guide (7th ed.). Muthén & Muthén. [Google Scholar]
  34. Nishith P, Resick PA, & Griffin MG (2002). Pattern of change in prolonged exposure and cognitive-processing therapy for female rape victims with posttraumatic stress disorder. Journal of Consulting and Clinical Psychology, 70(4), 880–886. 10.1037/0022-006X.70.4.880 [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Peterson AL, Foa EB, Resick PA, Hoyt TV, Straud CL, Moore BA, Favret JV, Hale WJ, Litz BT, Rogers TE, Stone JM, Villareal R, Woodson CS, Young, McCaughan S, & Mintz J (2020). A nonrandomized trial of prolonged exposure and cognitive processing therapy for combat-related posttraumatic stress disorder in a deployed setting. Behavior Therapy. Advance online publication. 10.1016/j.beth.2020.01.003 [DOI] [PubMed] [Google Scholar]
  36. Phelps AJ, Steele Z, Cowlishaw S, Metcalf O, Alkemade N, Elliott P, O’Donnell M, Redston S, Kerr K, Howard A, Nursey J, Cooper J, Armstrong R, Fitzgerald L, & Forbes D (2018). Treatment outcomes for military veterans with posttraumatic stress disorder: Response trajectories by symptom cluster. Journal of Traumatic Stress, 31(3), 401–409. 10.1002/jts.22299 [DOI] [PubMed] [Google Scholar]
  37. Pineles SL, Mostoufi SM, Ready CB, Street AE, Griffin MG, & Resick PA (2011). Trauma reactivity, avoidant coping, and PTSD symptoms: A moderating relationship? Journal of Abnormal Psychology, 120(1), 240–246. 10.1037/a0022123 [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Powers MB, Halpern JM, Ferenschak MP, Gillihan SJ, & Foa EB (2010). A meta-analytic review of prolonged exposure for posttraumatic stress disorder. Clinical Psychology Review, 30(6), 635–641. 10.1016/j.cpr.2010.04.007 [DOI] [PubMed] [Google Scholar]
  39. Reger GM, & Gahm GA (2008). Virtual reality exposure therapy for active duty soldiers. Journal of Clinical Psychology, 64(8), 940–946. 10.1002/jclp.20512 [DOI] [PubMed] [Google Scholar]
  40. Reger GM, Holloway KM, Candy C, Rothbaum BO, Difede J, Rizzo AA, & Gahm GA (2011). Effectiveness of virtual reality exposure therapy for active duty soldiers in a military mental health clinic. Journal of Traumatic Stress, 24(1), 93–96. 10.1002/jts.20574 [DOI] [PubMed] [Google Scholar]
  41. Reger GM, Koenen-Woods P, Zetocha K, Smolenski DJ, Holloway KM, Rothbaum BO, Difede J, Rizzo AA, Edwards-Stewart A, Skopp NA, Mishkind M, Reger MA, & Gahm GA (2016). Randomized controlled trial of prolonged exposure using imaginal exposure vs. virtual reality exposure in active duty soldiers with deployment-related posttraumatic stress disorder (PTSD). Journal of Consulting and Clinical Psychology, 84(11), 946–959. 10.1037/ccp0000134 [DOI] [PubMed] [Google Scholar]
  42. Reger GM, Smolenski D, Norr A, Katz A, Buck B, & Rothbaum BO (2018). Does virtual reality increase emotional engagement during exposure for PTSD? Subjective distress during prolonged and virtual reality exposure therapy. Journal of Anxiety Disorders, 61, 75–81. 10.1016/j.janxdis.2018.06.001 [DOI] [PubMed] [Google Scholar]
  43. Richards DA, Lovell K, & Marks IM (1994). Posttraumatic stress disorder: Evaluation of a behavioral treatment program. Journal of Traumatic Stress, 7(6), 669–680. 10.1002/jts.2490070412 [DOI] [PubMed] [Google Scholar]
  44. Rizzo AA, Reger GM, Gahm GA, Difede J, & Rothbaum BO (2009). Virtual reality exposure therapy for combat related PTSD. In Shiromani P, Keane T, & LeDoux J (Eds.), Posttraumatic stress disorder: Basic science and clinical practice (pp. 375–399). Humana. 10.1007/978-1-60327-329-9_18 [DOI] [Google Scholar]
  45. Rizzo A, Reger G, Perlman K, Rothbaum B, Difede J, McLay R, Graap K, Gahm G, Johnston S, Deal R, Pair J, Parsons T, Roy M, Shilling R, & Sharkey PM (2011). Virtual reality posttraumatic stress disorder (PTSD) exposure therapy results with active duty OIF/OEF service members. International Journal on Disability and Human Development, 10(4), 301–308. 10.1515/IJDHD.2011.060 [DOI] [Google Scholar]
  46. Rothbaum BO, Hodges LF, Kooper R, Opdyke D, Williford JS, & North M (1995). Effectiveness of computer-generated (virtual reality) graded exposure in the treatment of acrophobia. American Journal of Psychiatry, 152(4), 626–628. 10.1176/ajp.152.4.626 [DOI] [PubMed] [Google Scholar]
  47. Rothbaum BO, Price M, Jovanovic T, Norrholm SD, Gerardi M, Dunlop B, Davis M, Bradley B, Duncan EJ, Rizzo AA, & Ressler KJ (2014). A randomized, double-blind evaluation of d-cycloserine or alprazolam combined with virtual reality exposure therapy for posttraumatic stress disorder in Iraq and Afghanistan war veterans. American Journal of Psychiatry, 171(6), 640–648. 10.1176/appi.ajp.2014.13121625 [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Schell TL, Marshall GN, & Jaycox LH (2004). All symptoms are not created equal: The prominent role of hyperarousal in the natural course of posttraumatic psychological distress. Journal of Abnormal Psychology, 113(2), 189–197. 10.1037/0021-843X.113.2.189 [DOI] [PubMed] [Google Scholar]
  49. Taylor S, Thordarson DS, Maxfield L, Fedoroff IC, Lovell K, & Ogrodniczuk J (2003). Comparative efficacy, speed, and adverse effects of three PTSD treatments: exposure therapy, EMDR, and relaxation training. Journal of Consulting and Clinical Psychology, 71(2), 330–338. 10.1037/0022-006X.71.2.330 [DOI] [PubMed] [Google Scholar]

RESOURCES