غم و اندوه طولانی مدت متمایز از استرس پس از سانحه مربوط به سوگ و دغداری؟

Abstract Prolonged grief disorder (PGD) (previously called complicated grief (CG)) is proposed as a distinct post loss syndrome, with its own core symptoms. A key issue concerning the diagnostic validity of PGD is whether it can reliably be distinguished from related psychiatric outcomes following bereavement. This study therefore sought to determine whether the core symptoms of PGD could be distinguished from those of bereavement-related anxiety, depression and posttraumatic stress disorder (PTSD). Data were derived from a community sample of 223 bereaved adults in Croatia. PGD symptomatology was measured using the Revised Inventory of Complicated Grief. Depression and anxiety symptoms were measured using the Beck Depression and Anxiety Inventories, respectively. The intrusion and avoidance symptoms of PTSD were assessed using the Revised Impact of Event Scale. The distinctiveness of the five symptom clusters was examined using principal component analysis (PCA). Symptoms of prolonged grief, depression, anxiety, PTSD-intrusion, and PTSD-avoidance clustered together into five distinct factors. These results support the phenomenological distinctiveness of prolonged grief symptoms, from those of bereavement-related anxiety, depression and, for the first time, PTSD.

Introduction Following the loss of a significant other, a period of acute distress is common. However, a notable minority of individuals so bereaved develop clinically disabling grief symptoms. At present, the Diagnostic and Statistical Manual of Mental Disorders (DSM) does not recognise such grief reactions as a separate disorder but lists bereavement as a “V” code condition that “may be a focus of clinical attention” (DSM-IV-The Current Version (TR)). However, there is a growing consensus that pathological grief reactions — variously called “complicated grief” (Boelen and van den Bout, 2008, Boelen et al., 2007, Boelen et al., 2006, Prigerson et al., 1995a and Prigerson et al., 1996a), “complicated grief disorder” (Horowitz, et al., 2003), “Traumatic Grief” (Boelen and van den Bout, 2002, Boelen et al., 2003, Prigerson et al., 1997a, Prigerson et al., 1997b, Prigerson et al., 1999a and Prigerson et al., 1999b) or, most recently, “prolonged grief disorder (PGD)” (Boelen and Prigerson, 2007, Goldsmith et al., 2008 and Prigerson et al., 2007) merit inclusion in the DSM as a distinct mental disorder that induces clinically significant distress associated with substantive disability. The current consensus criteria for PGD are presented in the Appendix. One of several key criteria for establishing the validity of a putative psychiatric diagnosis is the distinctiveness of the clinical phenomenology (see Robins and Guze, 1970, for a fuller discussion). Distinctiveness is important because one wants to know that the disorder in question is not simply a different but overlapping expression of another condition. In the case of PGD, efforts to demonstrate distinctiveness have focussed on delineating its core symptoms from those of anxiety, and depression surrounding the bereavement (Boelen and van den Bout, 2005, Boelen et al., 2003, Ogrodniczuk et al., 2003, Prigerson et al., 1995a, Prigerson et al., 1995b and Prigerson et al., 1996a). The most persuasive empirical data in support of the distinctiveness of PGD phenomenology come from a series of innovative factor analytic (FA) studies on bereaved samples revealing that the core symptoms of PGD load on separate factors to the symptoms of anxiety and depression related to the bereavement using principal component analysis (PCA) and principal axis factoring (Boelen et al., 2003, Chen et al., 1999, Prigerson et al., 1995a and Prigerson et al., 1996a); using exploratory factor analysis (EFA) (Ogrodniczuk et al., 2003 and Prigerson et al., 1995b) and using confirmatory factor analysis (CFA) (Boelen and van den Bout, 2005 and Prigerson et al., 1996b). For example, in the prototypical study (Prigerson, et al., 1995a) on elderly, bereaved spouses who completed various measures of pathological grief and depression symptomatology, PCA revealed a two-factor solution including a prolonged grief factor that constituted symptoms of yearning, searching, preoccupation with thoughts of the deceased, crying, disbelief regarding the death, feeling stunned by the death and lack of acceptance of the death. Critically, this grief factor was distinct from a bereavement-depression symptom factor (e.g., apathy, depressed mood). This work was extended to include an examination of bereavement-related anxiety symptoms revealing separable factors reflecting prolonged grief, anxiety and depression symptoms, associated with the bereavement, in bereaved elders (Prigerson, et al., 1996a) and a Dutch community sample (Boelen et al., 2003), and using both exploratory (Boelen et al., 2003) and confirmatory factor analysis (Boelen and van den Bout, 2005). Taken together, these studies provide compelling evidence that, despite overlap in the symptom criteria of PGD and anxiety and depressive reactions following bereavement, the different syndromes are not isomorphic and the core symptom clusters are empirically dissociable. Notably absent from these PCA/EFA/CFA studies, however, is the examination of posttraumatic stress symptoms in the wake of bereavement. There is considerable debate as to whether PGD itself can be conceptualized as a form of posttraumatic stress disorder (PTSD) with the setting event as the bereavement, as opposed to a severe threat-related trauma (Bonanno, 2006, Bonanno et al., 2007, Dalgleish and Power, 2004, Prigerson et al., 1997a and Prigerson et al., 1997b). There is currently gathering evidence in favour of distinguishing PTSD and PGD, with the emergence of data showing that the two conditions are not 100% concordant (Barry et al., 2002, Melhem et al., 2001 and Simon et al., 2007), and that they have differential predictive validity as regards later psychosocial functioning (Bonanno et al., 2007, Prigerson et al., 1999b and Silverman et al., 2000). Nevertheless, demonstrating the distinctiveness of the phenomenology of PGD and PTSD symptomatology using techniques such as PCA remains a key challenge in establishing the diagnostic validity of the syndrome (Robins and Guze, 1970). To date, only one study has used the FA approach described above to examine the distinctiveness of PGD symptoms from the hallmark PTSD symptom dimensions of intrusion and avoidance (Ogrodniczuk, et al., 2003). However, unlike earlier studies which have used samples selected on the basis of bereavement itself rather than on the basis of diagnosis or symptom profiles, this study was restricted to a cohort of psychiatric patients almost all of whom were classified as meeting the criteria for “complicated grief” (i.e., not the current PGD criteria) and almost 90% of whom met the criteria for an existing DSM Axis I disorder. The results showed that for this psychiatric sample a number of the core symptoms of CG/PGD (e.g., yearning, searching, disbelief) and the intrusive symptoms of posttraumatic stress (e.g., involuntary intrusive thoughts, images and feelings associated with the traumatic experience of the bereavement) loaded on the same factor, while other core symptoms of CG/PGD loaded onto a separate factor. However, these two factors were distinguishable from bereavement-related avoidance and bereavement-related depression. These findings provide some support for a distinction between PGD symptoms and bereavement-related avoidance — one of the principal symptom dimensions of PTSD — and thus potentially strengthen the case for PGD as a distinct syndrome. However, some caution is warranted. The results from this study revealed two separable PGD symptom factors, as opposed to the single factor in the previous studies (Boelen and van den Bout, 2005, Boelen et al., 2003, Prigerson et al., 1995a and Prigerson et al., 1996a). These two factors comprised PGD symptom items from two different grief measures and interestingly all of the items from one measure loaded onto one of the factors, while the items from the second measure loaded onto the other factor, despite considerable overlap in the range of symptoms assessed by the two measures. Closer inspection reveals that the two measures actually assessed grief symptoms in different ways: by eliciting symptom reports over the previous week, on the one hand; and by asking how the person felt at that moment, on the other hand. Given the differential factor loadings for the two measures, it seems likely that this key psychometric difference between the measures significantly accounts for the factor structure in this particular study. Other potential issues are the restricted range of scores on the various measures and the skewing of the symptom profile in favour of grief symptoms, given that this was a psychiatric patient sample selected on the basis of a primary CG diagnosis. Finally, the study failed to include items assessing bereavement-related anxiety. Given these concerns it seems imperative to examine the distinctiveness of PGD symptoms from anxiety, depression and posttraumatic stress symptoms surrounding the bereavement in a community sample, selected on the basis of bereavement rather than psychiatric diagnosis, akin to those used in the earlier PCA/EFA/CFA studies (Boelen and van den Bout, 2005, Boelen et al., 2003, Prigerson et al., 1995a and Prigerson et al., 1996a). Such community samples include a broader and more balanced spread of symptoms than a psychiatric sample by a single diagnosis. This breadth will provide a less skewed context within which the relevant factors can be identified. It is also important to ensure that all of the measures index symptomatology in psychometrically similar ways. This approach was the aim of the present research. In sum, the key aim of the current study was to examine the distinctiveness of PGD symptoms from symptoms of anxiety, depression and posttraumatic stress following a loss of a loved one in a community sample from Croatia.

3. Results 3.1. Characteristics of the sample The mean scores for the full-scale versions of each of the instruments are presented in Table 1. On the ICG, 17.5% of the sample scored above the cut off score of 25 indicative of probable PGD (see Prigerson, et al., 1995b for arguments pertaining to CG). The data for these 17.5% are presented separately in Table 1 for additional information. On the BDI, 9.4% of the sample scores above the cut-off score of 21 indicative of severe depression (Beck et al., 1988). On the BAI, 29.1% of the sample scored above the cut off of 18 indicative of significant anxiety (Beck, 1988). Finally, on the IES-R, 23.8% of the sample scored above the cut off of 17 for the avoidance subscale and 20.6% of the sample scored above the cut of 17 off for the intrusion subscale, indicative of significant posttraumatic stress (Weiss and Marmar, 1997). Table 1. Descriptive characteristics of the 223 and 44 bereaved individuals on the full versions of the measures used in the study. Sample Total sample N = 223 ICG score >25 n = 39 Variable Mean S.D. Mean S.D. ICG 12.02 12.01 34.74 9.04 BDI 8.43 8.57 15.41 11.00 BAI 13.16 10.21 17.23 11.41 IES-R-avoidance 10.94 6.89 14.95 5.90 IES-R-intrusion 9.62 6.72 13.87 5.00 IES-R-total 21.42 12.43 28.82 9.05 Note: ICG = Inventory of Complicated Grief; BDI = Beck Depression Inventory; BAI = Beck Anxiety Inventory; IES-R = Impact of Event Scale-Revised. Table options 3.2. Principal components analysis (PCA) The principal axis factoring resulted in the emergence of 8 factors with eigenvalues greater than 1.00 after rotation (8.57, 4.98, 4.65, 3.41, 3.29, 1.70, 1.44 and 1.14). However, the scree plot showed a break between the steep slope of the fifth factor and a gradual trailing of the other factors (Cattell, 1966). The items sorted themselves into five distinct and meaningful factors that accounted for 58.13% of the variance (Table 2). This compares favourably with the amount of explained variance in the previous PCA studies (Boelen and van den Bout, 2005, Boelen et al., 2003, Prigerson et al., 1995a and Prigerson et al., 1996a). The first factor accounted for 29.19% of the variance. Symptoms of prolonged grief from the ICG loaded highly on this factor, with factor loadings ranging from 0.51 to 0.87 and only one loading was above 0.20 on the other four factors (this was the ICG ‘avoidance of reminders item’ loading onto Factor 5 (avoidance factor) with a factor loading of 0.46. The second factor accounted for 12.31% of the variance. Symptoms of anxiety from the BAI clustered together in this factor, with factor loadings from 0.42 to 0.77 and no loadings above 0.24 on the other four factors. The third factor accounted for 7.54% of the variance. Symptoms of depression from the BDI loaded highly on this factor, with factor loadings ranging from 0.49 to 0.71 and no loadings above 0.34 on the other four factors. The fourth factor accounted for 4.85% of the variance. Symptoms of posttraumatic intrusion from the IES-R intrusion subscale loaded highly on this factor, with factor loadings ranging from 0.66 to 0.77 and no loadings above 0.31 on the other four factors. The fifth factor accounted for 4.24% of the variance. Symptoms of posttraumatic avoidance from the IES-R avoidance subscale loaded highly on this factor, with factor loadings ranging from 0.74 to 0.84 and no loadings above 0.25 on the other four factors. Table 2. Rotated factor loadings for symptoms of prolonged grief, depression, anxiety, intrusion and avoidance among 223 bereaved individuals (significant factor loadings > 0.40 are in bold). Symptom Loading on factor Loading on factor Loading on factor Loading on factor Loading on factor 1 2 3 4 5 (Prolonged grief) (Anxiety) (Depression) (Intrusion) (Avoidance) Inventory of Complicated Grief Preoccupation with thoughts of the loss 0.59 0.13 0.00 0.14 0.05 Drawn to places and things associated with the loss 0.57 0.09 0.04 0.11 − 0.01 Avoidance of reminders of the loss 0.51 0.16 − 0.05 − 0.19 0.46 Feel disbelief over [deceased person] 's death 0.75 − 0.00 − 0.02 0.06 0.17 Part of self died as a result of the loss 0.81 0.04 0.12 0.15 0.11 Feel pain in the same area or assuming behaviours from that which was lost 0.62 0.07 0.22 0.17 − 0.13 Distressingly strong yearnings for that which was lost 0.74 0.04 − 0.04 0.17 0.04 Feel lonely since the loss 0.78 0.19 0.04 0.21 − 0.01 Feeling that future holds no meaning or purpose without that which was lost 0.84 0.11 0.15 − 0.01 0.02 Pervasive numbness or detached since the loss 0.73 0.07 0.27 0.05 0.09 Feeling stunned, dazed, or shocked by the loss 0.76 0.12 0.12 0.04 0.08 Feeling that life is empty or meaningless since the loss 0.87 0.06 0.22 0.06 0.05 Extreme difficulty imagining fulfilling without that which was lost 0.86 0.02 0.20 0.12 0.07 Feel that the death has changed their view of the world 0.61 0.09 0.09 0.19 − 0.01 Extreme bitterness related to the loss 0.70 0.01 0.28 0.08 0.20 Beck Anxiety Inventory Unable to relax 0.05 0.67 0.15 0.23 0.11 Fear of the worse happening 0.15 0.64 0.23 − 0.04 0.20 Heart pounding or racing 0.08 0.65 0.10 0.04 0.07 Unsteady 0.03 0.71 0.24 0.13 0.16 Terrified 0.01 0.77 0.20 0.18 0.05 Nervous 0.05 0.68 0.17 0.19 0.02 Hands trembling 0.11 0.42 0.07 0.16 0.13 Feeling shaky 0.08 0.48 0.10 0.16 0.10 Fear of losing control 0.12 0.58 0.28 0.07 0.22 Feeling scared 0.14 0.75 0.10 0.16 0.05 Beck Depression Inventory Feeling sad and unhappy 0.21 0.34 0.55 0.23 0.01 Feeling like a failure 0.20 0.32 0.63 − 0.02 0.00 Feeling dissatisfied or bored 0.33 0.18 0.63 − 0.01 0.09 Feeling guilty 0.27 0.17 0.68 0.01 0.01 Feeling punished 0.20 0.04 0.49 0.21 − 0.05 Blaming myself 0.12 0.17 0.62 0.05 − 0.02 Lost interest in people 0.07 0.19 0.66 − 0.06 0.16 Problems with decision making 0.14 0.09 0.71 0.11 0.04 Problems working − 0.04 0.18 0.67 0.09 0.07 Feeling too tired to do anything − 0.04 0.13 0.59 0.21 0.09 Impact of Event Scale — Revised (Intrusion subscale) Any reminder brought back feelings about it 0.24 0.08 0.02 0.77 0.11 Other things kept making me think about it 0.19 0.31 0.01 0.74 0.12 I thought about it when I didn't mean to 0.15 0.18 0.15 0.72 0.25 Pictures about it popped into my mind 0.15 0.15 0.15 0.70 0.28 I had waves of strong feelings about it 0.14 0.28 0.09 0.66 0.19 Impact of Event Scale — Revised (Avoidance subscale) I stayed away from reminders about it 0.25 0.02 0.09 0.16 0.74 I tried not to think about it − 0.02 0.18 0.07 0.25 0.79 I tried to remove it from my memory 0.08 0.15 0.08 0.22 0.81 I tried not to talk about it 0.05 0.17 0.02 0.17 0.84 Table options Cronbach's alphas for the present sample for each of the above 5 factors were: ‘prolonged grief’ = 0.94; ‘anxiety’ = 0.90; ‘depression’ = 0.87; ‘posttraumatic intrusion’ = 0.87; and ‘posttraumatic avoidance’ = 0.87, attesting to the robust reliability of the identified dimensions. Dimension scores were calculated using the sum of items for each measure that were entered into the PCA analysis followed by z-scoring. Intercorrelations among these z-scored dimensions are presented in Table 3, with most falling in the medium to large range ( Cohen, 1988). Table 3. Correlations among the prolonged grief, depression, anxiety, avoidance and intrusion dimensions. Prolonged grief Intrusion Anxiety Depression Intrusion 0.40 – – – Anxiety 0.28 0.48 – – Depression 0.39 0.34 0.52 – Avoidance 0.27 0.50 0.37 0.28 All intercorrelations are significant at P < 0.001.