Dr. Southwick and Colleagues Reply

TO THE EDITOR: We read with interest the letter by Dr. Roemer and colleagues that described systematic inconsistency in retrospective accounts of combat-related events among Somalia veterans that resembled our findings among Desert Storm veterans. We agree with Roemer et al.'s observation that the findings in both studies do not directly address the issue of recovered memories. As we noted in our article, there were multiple possible explanations for the observed inconsistencies in memory. Further, because the study involved retrospective accounts, it was not possible to know whether events recalled at 1 month were more or less accurate than events recalled at 2 years. However, in both our study and in the study by Roemer et al., it is clear that memory changed or was inconsistent over time. Such inconsistencies suggest that recall at any one time point may be inaccurate.

Dr. French's comments also were of interest. In several different veteran populations we too have found confusion, shame, isolation, and guilt to be major concerns for combat veterans with PTSD. However, we do not agree that PTSD arises in response to these concerns “rather than” in response to trauma, pain, and abuse. Further, we do not believe that memories of a traumatic incident are immaterial to the development of PTSD. Instead, we agree with Roemer et al. who note that a simple dose-response model of the relationship between trauma experiences and symptoms is insufficient. It appears that multiple factors, including degree of traumatic exposure, are related to level of PTSD symptoms.