Predictors of involuntary and voluntary emotional episodic memories of virtual reality scenarios in Veterans with and without PTSD

ABSTRACT

This study investigated predictors of involuntary and voluntary memories of stressful virtual reality scenarios. Thirty-two veterans of the two Persian Gulf Wars completed verbal memory tests and diagnostic assessments. They were randomly assigned to a Recounting (16) or a Suppression (16) condition. After immersion in the VR scenarios, the Recounting group described the scenarios and the Suppression group suppressed thoughts of the scenarios. One week later, participants completed surprise voluntary memory tests and another thought suppression task. The best predictors of voluntary memory were verbal memory ability, dissociation, and to a lesser extent, physiological arousal before and after scenarios. Dissociation and physiological stress responses selectively affected memory for neutral elements. Higher distress during scenarios impaired voluntary memory but increased the frequency of involuntary memories. Physiological stress responses promoted more frequent involuntary memories immediately after the scenarios. More frequent initial involuntary memories, tonic physiological arousal, and stronger emotional responses to dangerous events predicted difficulty inhibiting involuntary memories at follow-up. The effects of thought suppression were transient and weaker than those of other variables. The findings suggest that posttraumatic amnesia and involuntary memories of adverse events are more related to memory ability and emotional and physiological stress responses than to post-exposure suppression.

KEYWORDS:

• PTSD
• memory
• virtual reality
• stress

Acknowledgements

The authors wish to thank the study research assistants, Amy Elitzer and Ashley Okamoto, for their invaluable work with the scoring of memory tests.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 The study was also reviewed and approved for funding by the National Institutes of Health loan repayment grant program, as well as the Weill Medical College of Cornell University’s research committee that awarded internal departmental funding.

2 We used this method to normalize variables because using z-score transformations did not sufficiently normalize kurtotic distributions. We also did not want to further reduce the small sample size by excluding outliers.

3 Two subjects completed the diagnostic assessment and the VR challenge within the same week due to scheduling issues.

4 Two subjects’ visits commenced at 5:00 and 5:15 PM due to their tardiness. For the majority, visits were scheduled within one hour (77%-90%) or two hours (91-97%) of each other.

5 We conducted a series of analyses to identify potential effects of the variability in the sampling time. A one-way ANOVA (with the outlier removed) found a longer average duration in Suppression Group compared to the Recounting Group, F(1,29) = 4.958, p = .034. However, the duration of time between cortisol samples was not significantly correlated with the post-VR challenge level of cortisol or with the pre-post challenge change in cortisol level. Nor was it correlated with the frequency of involuntary thoughts about the scenarios or making attempts to suppress thoughts during the week. It was also not correlated with any of the delayed recall and recognition memory variables. There was one significant correlation with the Immediate Free Recall of Danger stimuli. However, the regression analysis in the Results section was repeated, controlling for this variable, and the results were unchanged.

Additional information

Funding

This study was supported by NIH Loan Repayment and Weill Medical College of Cornell University Faculty Grants awarded to the first author. These agencies had no involvement in the design or execution of the research or in the writing of research manuscripts.