سوگ و داغداری حل نشده در جانبازان مبتلا به PTSD

Abstract Clinicians have documented the importance of loss of comrades during combat as a significant source of distress. However, empirical studies have not focused on unresolved grief as a possible outcome of combat experiences. Consequently, unresolved grief has often been treated “after the fact” in the context of treating PTSD and depressive symptoms. In this study, we therefore, sought to demonstrate the prominence of combat-related grief-specific symptoms in a sample of Vietnam veterans being treated for PTSD. Our results indicated that indeed this sample of veterans reported high levels of grief-specific symptoms comparable to that found in bereaved individuals whose spouse had recently died, verifying its prominence as an important component of combat-related stress. Furthermore, grief severity was uniquely associated with losses of comrades during combat whereas no such relationship was shown for trauma or depressive symptoms. The latter finding suggested that in fact higher levels of grief stemmed from interpersonal losses during the war and was not simply an artifact of current general distress level.

4. Results 4.1. Grief symptoms as distinct from PTSD and depression In order to determine whether grief symptoms could be distinguished from PTSD and depression, a principle components analysis was conducted on the grief, PTSD, and depression subscale scores. This analysis was performed on the subscale scores, as opposed to the individual items, to ensure that the criterion of having at least four to five participants per variable was met (Hair, Anderson, Tatham, & Black, 1992). The basis for selecting factors was made according to the following criteria: (a) a latent root criterion was used to select significant factors with an eigenvalue greater than one; (b) a Scree Test was performed to ascertain the maximum number of significant factors possible (Hair et al., 1992). The principal components analysis yielded a three-factor solution. A varimax rotation was performed in order to enhance the interpretability of the factor pattern. The results are summarized in Table 2. As expected, the grief-specific symptom subscales were clearly distinguished from the PTSD and depression subscales. In this final rotated solution, grief-specific symptoms accounted for 29.6% of the actual variance, PTSD accounted for 15.1%, and depression explained 14.2% of the variance. The results of the principal components analysis, therefore, demonstrate that the sample of combat veterans with a diagnosis of PTSD clearly distinguish grief-specific symptoms from PTSD and depression. Table 2. Rotated Component Matrix of Symptom Subscales Components 1 (grief) 2 (PTSD) 3 (depression) Accept 0.801 0.155 0.081 Distress 0.773 0.238 −0.052 Thoughts 0.864 0.176 0.099 Grief 0.784 0.167 0.163 Images and Thoughts 0.791 0.215 0.239 Separation 0.794 0.073 0.297 Arousal 0.184 0.762 0.177 Re-experiencing 0.368 0.599 0.453 Self-Persecution 0.368 0.540 0.385 Withdrawal 0.122 0.859 0.046 Self-Image 0.152 0.067 0.840 Motivate 0.075 0.219 0.846 Negative Attitude 0.205 0.561 0.652 Table options 4.2. Severity of grief-specific symptoms In having established that this sample of veterans were able to distinguish grief-specific symptoms from PTSD and depression on the self-report symptom measures, it was important to determine the severity of grief-specific symptoms in this sample of veterans with PTSD by comparing their scores with other samples of bereaved individuals. Indeed, the veterans’ mean CBI and TRIG scores were striking considering that approximately 30 years had passed since these combat losses were experienced. The veterans’ mean score of 48.90 on the TRIG was higher than a mean score of 47.39 in a midlife conjugal bereavement sample whose spouses had died within the previous 6 months (Field & Horowitz, 1998) and a mean score of 45.60 for an elderly conjugally bereaved sample whose spouses had died 3–6 months ago (Prigerson et al., 1995). The fact that 70% of the veterans had higher TRIG scores than the average scores for these conjugally bereaved samples attests to the degree to which these veterans continue to experience grief over interpersonal losses that occurred over 30 years ago. Because the death of a spouse is known to be one of the most stressful life events that an individual is likely to encounter (Holmes & Rahe, 1967), the results for the TRIG are indeed remarkable. A similar result was obtained for the CBI; the veterans’ mean score of 26.57 on the CBI was comparable to a normative sample who had a mean CBI score of 25.97 assessed at 1 month post loss (Burnett et al., 1997). 4.3. Symptoms predicting attachments and losses during war A multiple regression analysis was conducted to determine the unique relationship between each of the symptom measures and the extent of attachment to men in the unit during the war. This involved entering the total grief, trauma, and depression measures scores into the regression analysis simultaneously in the prediction of the attachment to men in unit measure. A significant overall model effect was found, r2 = .07 (adjusted r2 = .05), F (3, 110) = 2.80, p < .05. Grief symptoms were significantly positively associated with attachment (t = 2.79, p < .01) whereas no relationship was found for trauma symptoms (t = −0.85, ns) or depression (t = .18, ns) with attachment. The results are summarized in Table 3A. Table 3A. Regression analysis of symptom measures predicting attachment to men in unit Symptom measure B S.E. B β Grief symptoms .43 .15 .31* Trauma symptoms −.01 .01 −.12 Depression .003 .02 .02 Note. n = 114; r2 = .07 (p's < .05). * p's < .05. Table options The similar regression analysis was conducted for the relationship with a buddy. A significant overall model effect was found, r2 = .10 (adjusted r2 = .07), F (3, 110) = 3.90, p < .01. Grief symptoms were significantly positively associated with combat losses (t = 3.14, p < .001) whereas no significant association was found for trauma symptoms (t = −1.27, ns) or for depression (t = .77, ns) with combat losses. The results are summarized in Table 3B. Table 3B. Regression analysis of symptom measures predicting closeness to a buddy Symptom measure B S.E. B β Grief symptoms .45 .14 .35** Trauma symptoms −.01 .01 −.17 Depression .01 .02 .10 Note. n = 114; r2 = .10 (p's < .01). ** p's < .01. Table options A final regression analysis was conducted for the combat losses measure. A significant overall model effect was found, r2 = .23 (adjusted r2 = .21), F (3, 110) = 10.57, p < .001. Grief symptoms were significantly positively associated with combat losses (t = 4.35, p < .001) whereas no significant was found for trauma symptoms (t = .09, ns) or for depression (t = .58, ns) with combat losses. The results are summarized in Table 3C. Table 3C. Regression analysis of symptom measures predicting number of losses Symptom measure B S.E. B β Grief symptoms .54 .12 .44*** Trauma symptoms .001 .01 .01 Depression .01 .01 .07 Note. n = 114; r2 = .23 (p's < .001). *** p's < .001.