Gradual trauma exposure eases distress in PTSD therapy, study finds

One of the most widely prescribed treatments for posttraumatic stress disorder asks patients to do something that sounds almost unbearable: relive their worst memory, in vivid detail, over and over. Now a large study of active duty service members and veterans has found that a simple reordering of that process—working up to the worst memory through progressively less distressing ones—can dramatically lower the emotional intensity of therapy without abandoning its core mechanism. The findings, published in the Journal of Behavioral Medicine, offer a potential answer to one of trauma care’s most stubborn problems: patients who refuse to start treatment, or who quit because it simply hurts too much.

Prolonged exposure, or PE, is a trauma-focused cognitive behavioral therapy recommended by the American Psychological Association, the International Society for Traumatic Stress Studies, and the joint clinical practice guidelines of the U.S. Departments of Veterans Affairs and Defense. Over the course of roughly eight to fifteen sessions, patients repeatedly revisit the memory of their single most distressing traumatic experience—known as the index event—during imaginal exposure conducted in session, while also completing behavioral homework in which they gradually approach objectively safe reminders of the trauma in daily life. The treatment works, but emotional activation is built into its design. Under emotional processing theory, the framework that underpins PE, activating the fear structure in memory is considered necessary for corrective learning to occur. The catch is that this same emotional intensity is a leading explanation for the treatment’s elevated dropout rates. In one study cited by the research team, 60 percent of participants who dropped out said it was because the treatment was too stressful.

The new analysis, led by Kiara H. Buccellato of The University of Texas at San Antonio and colleagues with the Consortium to Alleviate PTSD, compared two ways of delivering the imaginal exposure component of PE. In the standard approach, participants focused solely on their single most distressing traumatic event from the outset—a strategy that, in the terminology of anxiety treatment, resembles flooding. In the modified, graduated approach, participants instead identified their top three most distressing events and began imaginal exposure with the third most distressing, repeating that exercise in every session until it no longer provoked strong distress before moving on to the second most distressing event, and finally the worst. This hierarchical strategy has long been standard practice in the treatment of anxiety disorders generally, and in the in vivo, between-session homework portion of PE itself, but it had never been rigorously applied to the imaginal exposure core of PTSD treatment.

The data came from a larger randomized clinical trial conducted between 2017 and 2019 at four sites in Texas, in which 234 active duty service members and veterans with PTSD were assigned to one of two intensive PE protocols. Both involved fifteen 90-minute sessions delivered over just three weeks. One group received massed PE, the standard protocol, while the other received an intensive outpatient version that added eight treatment enhancements—the most consequential being the identification of the top three traumas and graduated imaginal exposure, along with team-based treatment, clinic-based completion of homework, twice-daily therapist feedback sessions, enhanced social support involving a spouse or close friend, a brief timeline review of all lifetime traumas, and posttreatment booster sessions at one, three, and seven weeks. The present secondary analysis drew on 199 participants from three sites where distress ratings had been systematically recorded: 96 in the standard exposure group and 89 in the graduated group, covering 267 unique traumatic events.

Distress was measured with the Subjective Units of Distress Scale, or SUDS, a well-established clinical instrument in which 0 represents a state of absolute calmness and 100 represents the worst anxiety ever experienced. The researchers defined peak SUDS as the highest distress rating a participant endorsed for each traumatic memory across all treatment sessions, and compared average peak values between groups using analysis of variance with planned contrasts and Cohen’s d effect sizes.

The results were striking. Participants receiving standard exposure, confronting their worst memory from the start, reported an average peak SUDS of 88.06—just shy of the top of the scale. Participants in the graduated group reported an average peak of 75.19 for their most distressing event, a difference that was highly statistically significant with a medium-to-large effect size of 0.70. The graduated group also showed significantly lower peak distress for their second most distressing event (72.23, compared to the standard group’s 88.06, d = 0.81) and their third most distressing event (74.63, d = 0.69). Taken together, participants using graduated exposure experienced significantly lower average peak distress across all events combined—74.02 versus 88.06, d = 0.72. The overall effect of treatment group on peak distress was significant, F(3,359) = 12.46, p < .001.

The sample was largely male (79.9 percent) with a mean age of 38.5 years, and ethnically diverse: 41.2 percent identified as non-Hispanic White, 28.6 percent as African American, 25.6 percent as Hispanic, and 4.5 percent as other. The two groups were statistically indistinguishable on age, gender, rank, race and ethnicity, education, military status, and service history, with marital status the only demographic variable that differed.

Why does the ordering of traumatic memories matter so much? The researchers point to the psychology of treatment attrition. A national qualitative study of veterans cited in the paper found that treatment completers and dropouts interpreted their distress in fundamentally different ways: completers read intense distress as evidence that treatment was working, while dropouts read it as a sign their symptoms were worsening and the therapy was failing them. Crucially, 70 percent of those who dropped out did so between sessions three and six—precisely the window when early imaginal exposure sessions occur. By easing patients in with a less catastrophic memory, the graduated approach may allow them to build confidence and mastery before facing their worst experience, reducing the anticipatory anxiety that keeps many veterans from starting trauma-focused therapy at all.

The military context makes the graduated approach especially compelling. PE was originally developed with women sexual assault survivors, for whom a single clearly identifiable index trauma was often available. Combat environments are different: service members frequently accumulate multiple traumatic exposures—combat, childhood adversity, military sexual trauma—spanning several categories, making the selection of one “worst” event both difficult and arguably incomplete. Addressing the top three events allows treatment to encompass a broader range of experiences, and different trauma types map onto different PTSD symptom clusters, with combat trauma linked more strongly to intrusion and arousal symptoms and sexual trauma to avoidance and cognitive-mood symptoms. A single-event focus may leave this cumulative burden partially unaddressed.

The modified protocol has already shown clinical promise. The primary outcomes of the parent trial, published in JAMA Network Open, demonstrated significant reductions in clinician-rated and self-reported PTSD severity in both groups, with PTSD diagnostic remission rates of 48 percent for the intensive outpatient group and 61 percent for the massed group at posttreatment, and roughly 50 percent for both at six-month follow-up. Earlier work by McLay and colleagues on virtual reality graduated exposure therapy had also found greater symptom improvement in a graduated group compared to treatment-as-usual, though that study’s broad control group limited the comparison.

The authors are careful about the study’s limits. The graduated and standard groups differed in more than one way—the graduated approach was only two of eight enhancements distinguishing the intensive outpatient protocol—so other components, such as enhanced social support, could contribute to the distress differences observed. Because distress reduction was not manipulated in isolation, the specific causal effect of graduated versus traditional imaginal exposure on outcomes and dropout cannot be isolated from these data. There is also a theoretical tension to resolve: emotional processing theory holds that some degree of distress is essential for therapeutic gain, and the researchers note that some patients who interpret low distress as a sign that treatment is not working might paradoxically disengage. Future research, they argue, should hold the treatment modality constant except for the exposure ordering, and track both distress over time and attrition directly.

Even with those caveats, the findings carry real weight for clinical practice. A therapy that asks people to walk back into their worst moments will always demand courage; the question is whether the door has to open all at once. This study, the first to directly compare patient-reported distress during graduated versus traditional imaginal exposure for PTSD, suggests it does not. Average peak distress fell by roughly 13 to 16 points on a 100-point scale—a margin that could mean the difference between a veteran who walks out of session three and one who walks back in for session four. As the authors conclude, making evidence-based trauma therapy more tolerable may be one of the most practical ways to ensure that the people who need it most actually stay long enough to be helped by it.