Increased PTSD Risk With Combat-Related Injury: A Matched Comparison Study of Injured and Uninjured Soldiers Experiencing the Same Combat Events

The current study attempted to isolate the unique contribution of bodily injury to the development of PTSD. To accomplish this goal we used an event-based, matched design. To the best of our knowledge, this is the first study that addresses this question utilizing such a powerful design. Consistent with most of the recent literature (4, 14), our findings clearly indicate that bodily injury is a risk factor—rather than a protective one—for PTSD. Moreover, in addition to just advancing the robustness of this notion, the data also suggest that the odds of developing PTSD following traumatic injury are approximately eight times higher than following injury-free trauma. In fact, they suggest that even this rather high figure might be an underestimate. This is so because about 35% of the originally approached injured soldiers, but none of the noninjured comparison subjects, refused participation. From the explanations that these nonparticipants gave for their refusal, it was quite obvious that many of them did so for reasons that can be interpreted as avoidance.

It is interesting that while associated with higher levels of anxiety and depression symptoms, bodily injury in this study was not related to increased prevalence of comorbid psychiatric diagnoses. In contrast, psychiatric comorbidity was significantly related to presence or absence of PTSD. This finding is consistent with previous studies, including our own (7), suggesting a strong association between PTSD and psychiatric comorbidity among injured trauma survivors. Taken together, our findings indicate that PTSD moderates psychiatric comorbidity following traumatic injury.

Our findings further suggest that bodily injury does not have a differential effect on any of the PTSD symptom domains. This finding was evident by the direction and size of differences, which were generally consistent across all three PTSD cluster scores. Such a finding is consistent with our previous results in a study with injured survivors of motor vehicle accidents (7).

What other factors might explain the increased risk for PTSD associated with bodily injury? The most simple and straightforward hypothesis is that bodily injury increases the perceived threat to one’s life or physical integrity during the trauma. Indeed, according to the literature, the perceived level of danger by trauma survivors is a better predictor of PTSD than the actual severity of the traumatic event (10). However, this explanation may be too simplistic because while strongly supporting the possibility that bodily injury contributes to the appraisal of the traumatic event as more dangerous, our data also suggest that this heightened level of perceived threat is not a simple, straightforward function of the severity of injury. This finding, which is consistent with previous studies (5, 14), suggests that bodily injury exerts its effect on perceived threat via interaction with other factors, such as the effect that it has on one’s (perceived and real) ability to function and cope during the traumatic event.

Another potential mediating factor between physical injury and PTSD might be peritraumatic dissociation, which has been implicated as a risk factor for PTSD in previous studies (24). This possibility was supported in our study by injured participants, with or without PTSD, reporting significantly higher levels of peritraumatic dissociation than their noninjured fellows. Furthermore, controlling for peritraumatic dissociation remarkably decreased the differences between the injured and noninjured groups on all the PTSD as well as the other clinical measures. While dissociation is commonly conceived of as a vulnerability factor in its impairing integration and reprocessing of the trauma, an interesting alternative possibility is that the relationship between bodily injury and dissociation might be reversed. That is, people who tend to dissociate under stress might be at higher risk for getting injured. Because of the post facto, retrospective nature of this study, such a possibility cannot be ruled out.

While not directly assessed in this study, there are also noteworthy biological mechanisms that may mediate between injury and PTSD. These hypothesized mechanisms relate to the complex interactions between the immune and stress-regulating systems, in particular the interaction between the hypothalamic-pituitary-adrenal (HPA) axis and proinflammatory mediators such as cytokines. In addition to its key role in the regulation of the stress response, the HPA axis is also involved in modulation of the immune response to inflammation and injury through proinflammatory mediators such as cytokines (25). Since alterations in the HPA axis have been suggested as a vulnerability factor for PTSD (26), situations that involve both emotional and injury-related stress may create an extra burden on an already compromised HPA axis. While this hypothesis has yet to be explored, numerous preclinical studies found that cytokines have adverse effects on memory, sleep, and mood (27). Similarly, they promote sickness behavior (27). Indeed, a previous study (28) found elevated CSF concentrations of interleukin-6 in patients with PTSD versus normal comparison subjects.

Finally, the extent to which the current findings can be generalized to other populations is worth consideration. Overall, the rate of PTSD in our sample (10%) was generally in line with, although at the lower end of, previous estimates among soldiers involved in combat activity. However, prevalence of PTSD in the noninjured group was very low (2.5%). This may be explained by the nature of the participants in our study, who were soldiers in elite units who underwent extensive preinduction screening and a long and demanding course of basic training, which enables further selection and facilitates acquisition of habitual coping styles. In addition, these units are characterized by high levels of motivation and unit cohesion, which have been shown to be resilience factors against the development of PTSD (29) as well as general mental health problems (30). These factors suggest that generalizing from this study should be made with caution and call for further replication with a more heterogeneous and unscreened sample.

A few methodological issues that limit interpretation of the current findings are worth mentioning. First, PTSD symptoms were assessed relatively long after the traumatic injury. Thus, our data provide a reliable estimate of stable posttraumatic symptoms (which also minimizes concerns regarding potential effects of the slightly longer event-to-interview time difference in the injured group). However, they do not answer questions related to acute stress reactions or the temporal course of posttraumatic adjustment as it is related to recovery from the physical injury. Second, severity of injury was assessed only at the time of hospitalization but not again at the time of assessment. Thus, our data cannot clarify the degree to which current injury or physical disability status may affect reported PTSD symptoms, nor can it clarify the degree to which pace of physical recovery may be influenced by early PTSD symptoms. Third, the raters completing the SCID and the Clinician-Administered PTSD Scale were not blind to the injury status of the participants. While this could result in a potential bias in PTSD diagnosis, this possibility is less likely given that ratings on the self-report measures (i.e., Peritraumatic Dissociation Experience Scale, Beck Depression Inventory, Beck Anxiety Inventory) were remarkably consistent with the raters’ evaluations. Finally, as can be recalled, a sizable proportion (35%) of injured patients refused to participate in the study. This raises the possibility that our sampling procedure was at risk for selection bias. Because of ethical constraints (i.e., lack of consent), we could not collect the necessary data to rule this possibility out. Yet, from the little we do know about these subjects (i.e., the reasons they gave for nonparticipation), it is obvious that such bias, had it existed, could only further strengthen our findings rather than weaken them. Last, and perhaps most important, trauma in this study was defined and assessed only in terms of its objective characteristics (i.e., combat participation). Since it is possible that not all combatants perceive this experience as subjectively traumatic, the current data cannot answer questions related to the unique contribution of physical injury to PTSD only among those whose combat experience was self-identified as traumatic.

In conclusion, our study provides clear evidence that bodily injury during trauma is a major risk factor for subsequent PTSD. Considering that emotional distress is often overlooked among injured patients hospitalized in surgical and trauma units, these findings highlight the importance of paying more attention to psychological aspects of their condition in general and to the early symptoms of PTSD in particular, both during hospitalization and after discharge.

Received May 5, 2003; revision received Dec. 10, 2003, and March 31, 2004; accepted April 23, 2004. From the University of Haifa Psychology Department, Haifa, Israel; the Departments of Psychiatry and Orthopedics, Rambam Medical Center; and the Faculty of Industrial Engineering and Management, Technion, Haifa, Israel. Address correspondence and reprint requests to Dr. Koren, Psychiatry Department, Rambam Medical Center, Haifa, Israel; [email protected] (e-mail). Supported by a grant from the Chief Scientist of the Medicine Corps of the Israel Defense Forces. The authors thank O. Shelli-Vacnin, A. Tzarfati, D. Mor, Y. Goshen-Kita, and L. Ziv for their assistance in participant recruitment and interviewing. The authors also thank Y. Baruch, M.D., and E. Budner, Ph.D., from the Medicine Corps of the Israel Defense Forces for assistance in patient recruitment.