The important study by Southwick and colleagues on the consistency of memory for combat-related traumatic events in veterans of Operation Desert Storm demonstrates that there are changes in memory recall over time. It also shows that an increase in memory is associated with higher scores for posttraumatic stress disorder (PTSD). The authors correctly point out that most reports of combat exposure and its positive correlation with PTSD symptoms are based on retrospective data, thereby leaving open the possibility that the relationship is spurious and due to symptom-induced amplification of memories. They note, however, that it is also possible that "material that had been forgotten, denied, suppressed, or repressed at 1 month may have become conscious by 2 years" (p. 176). Thus, people who were more traumatized by initial combat exposure may show acute memory disturbances and be more vulnerable to later PTSD. This is indeed what we found in our prospective study of victims of the Oakland/Berkeley firestorm R401559BCEBDJCG. Dissociative symptoms, including memory disturbance, were associated with severity of exposure and also predicted the development of PTSD at 7-month follow-up. An association between peritraumatic dissociation and subsequent PTSD has been observed by other investigators as well R401559BCEDHCAJ, R401559BCECEADE.

Thus, these data, taken in concert with those in the Southwick et al. study, suggest that those individuals who suffered more combat exposure may also have more acute difficulty with traumatic memories and that dissociative symptoms are associated with higher subsequent risk of developing PTSD. In clinical terms, this would mean that individuals who can consciously work through their traumatic experiences immediately after the event are less likely to develop posttraumatic stress symptoms, while those who have difficulty accessing and working through traumatic memories are at higher risk for later PTSD. The fact that memories of trauma change with time does not establish the cause of these changes. Because we do not know which memory account is more accurate, we cannot know whether such changes reflect dissociation or repression of events that did occur or the addition of pseudomemories of events that did not occur. The findings of Southwick et al. do document that difficulties in information processing in the aftermath of trauma are associated with PTSD; therefore, presumably, correcting these difficulties in psychotherapy should be helpful in the treatment of the disorder.