نقش تفسیر منفی واکنش سوگ و داغداری در مشکلات عاطفی پس از سوگ

Abstract This study explored the role of negative interpretations of grief reactions in emotional problems after bereavement, with 234 individuals who had been confronted with the death of a close relative. It was found that negative interpretations of grief reactions were highly associated with the degree to which these reactions were experienced as distressing, the degree to which mourners engaged in avoidance behaviours and the severity of symptoms of traumatic grief and depression, even when controlling for the frequency of grief reactions and the influence of relevant background variables. Furthermore, behavioural and cognitive avoidance strategies were significantly related to the severity of traumatic grief and depression. Negative interpretations of grief reactions and rumination explained most variance in symptom severity, when controlling for the shared variance between the predictor variables. The results have implications for the treatment of emotional problems after bereavement.

1. Introduction In the first weeks after the death of a loved one, most people experience a wide variety of psychological symptoms such as sadness, anger, separation distress, loss of interest in usual activities, flashbacks, and intrusive thoughts. Several studies have shown that, in this period, significant proportions of bereaved individuals even meet criteria for disorders such as major depression (Zisook, Shuchter, Sledge, Paulus, & Judd, 1994), post-traumatic stress-disorder (PTSD) (Schut, De Keijser, Van den Bout, & Dijkhuis, 1991) and other anxiety disorders (Jacobs et al., 1990). In addition, a study by Prigerson et al. (1997), among conjugally bereaved elders, showed that at two months post-loss, more than half of the participants suffered from the disorder traumatic grief, encompassing symptoms of separation distress and traumatic distress. However, most of the aforementioned studies that investigated the long-term impact of bereavement have shown that as time progresses, most individuals recover from their symptoms. Only a minority of bereaved individuals continue to suffer from debilitating symptoms of traumatic grief or other emotional problems (cf. Horowitz et al., 1997; Prigerson et al., 1997). This suggests that the mere presence of intense emotions and other symptoms shortly after a loss does not indicate disturbance. Some theorists (e.g., Horowitz, 1986; Rachman, 1980) have even suggested that such symptoms are necessary for emotional processing, because they imply confrontation with the reality of the loss, and thus facilitate adaptation to the loss and the integration of the loss into pre-existing schemas. Given that the occurrence of grief reactions shortly after a loss is not abnormal, and perhaps even necessary for emotional processing, for a full understanding of the nature of persistent emotional problems after bereavement research will have to clarify what factors determine the maintenance, prolongation and intensification of these basically normal and benign grief reactions. Different theorists (e.g., Horowitz, 1986) have proposed that grief reactions persist and exacerbate if emotional processing of the loss is prevented or incomplete due to avoidance behaviour. Avoidance may prevent habituation to painful memories (Foa & Kozak, 1986) and may interfere with the integration of the loss into pre-existing schemas and the development of new schemas that take the loss into account (Horowitz, 1986; Horowitz, Bonanno, & Holen, 1993). Avoidance may manifest behaviourally in the form of staying away from places or people, or avoiding cues that are reminders of the deceased. In addition, mourners may engage in cognitive avoidance strategies, such as thought suppression and rumination. The suppression of unwanted thoughts or memories has been known to cause paradoxical increments in the frequency of these phenomena, which may lead to an increase in emotional distress (Lavy & Van den Hout, 1994; Wegner, 1989). Rumination may block emotional processing because it focuses on experiences surrounding the loss rather than the implications of the loss itself, and because it makes negative thoughts and memories more accessible and salient, thus enhancing the negative mood state (cf. Nolen-Hoeksema, 1991; Wells, 1994). What then determines who engages in avoidance behaviours? Building on the theoretical considerations of Ehlers and Steil (1995) concerning psychological maintaining factors in PTSD, it is assumed that mourners who experience their grief reactions as distressing are more likely to engage in strategies to avoid and control these reactions, than mourners who do not experience these as distressing. Thus, it is assumed that the occurrence of grief reactions does not necessarily mean that they are experienced as distressing. This assumption concords with studies on PTSD which have shown that, after a traumatic event, people differ in the extent to which they experience their intrusive symptoms as distressing and that it is the degree of distress related to these symptoms, rather than the frequency of the symptoms, that predicts avoidance behaviour (Steil & Ehlers, 2000). This leads to a new question, that is, what determines the degree of distress connected to the occurrence of grief reactions? Again following Ehlers and Steil (1995), we suggest that the variability in distress caused by grief reactions is due to cognitive factors. More specifically, we propose that negative interpretations of basically normal and benign grief reactions play an important role in persistent emotional problems after bereavement, because they determine the degree to which these reactions are experienced as distressing, and consequently, determine the degree to which the mourner engages in maladaptive avoidance strategies that may impede emotional processing and serve to prolong and exacerbate the grief reactions (cf. Boelen, Van den Bout, & Van den Hout, 2003). This notion echoes the work of theorists who have emphasised the importance of negative interpretations of particular symptoms in different forms of psychopathology. Most pertinent to the study of grief are the propositions of Ehlers and Steil (1995) that negative interpretations of particular symptoms after a trauma are important in the development of chronic PTSD, because they determine how distressing these symptoms are, and influence the extent to which the individual engages in maladaptive avoidance strategies. Evidence for these hypotheses was provided in cross-sectional studies (Clohessy & Ehlers, 1999; Steil & Ehlers, 2000) and prospective longitudinal studies (Ehlers, Mayou, & Bryant, 1998; Mayou, Ehlers, & Bryant, 2002). The purpose of the present study, performed in The Netherlands, was to gain more insight into possible psychological mechanisms involved in the development and persistence of emotional problems after bereavement, by investigating relations amongst negative interpretations of grief reactions, avoidance behaviours, and symptoms of traumatic grief and depression within a group of individuals who had suffered the loss of a first degree relative. Negative interpretations were defined as interpretations of grief reactions viewed as a sign of threat to one's mental health, as well as negative inferences concerning the self, the appropriateness of the reactions, and the loss sequelae. First, we hypothesised that negative interpretations of basically normal grief reactions would be correlated with the degree to which these reactions are experienced as distressing. Second, we expected the degree of distress related to these grief reactions to be associated with the severity of debilitating symptoms of traumatic grief and depression, and we expected these associations to be stronger than the associations between the frequency of grief reactions and these symptoms. Third, we expected negative interpretations of grief reactions to be related to the degree to which people engaged in behavioural and cognitive avoidance strategies. In addition, we expected these avoidance strategies to be related to the severity of traumatic grief and depression. Logically following from these hypotheses, it was expected that negative interpretations of grief reactions would be related to the severity of traumatic grief and depression. Finally, it was predicted that grief-related distress, the avoidance strategies, and negative interpretations of grief reactions, would explain variance in traumatic grief and depression over and above the variance explained by the frequency of grief reactions and relevant background variables and variables characterising the loss-event.

Results 3.1. Relations among background variables, loss-related variables, and symptom severity ITG scores were inversely related to the age of the respondent (r=−0.15, p<0.05) and number of years of education (r=−0.23, p<0.001). SCL depression scores were also inversely related to age (r=−0.13, p<0.05) and number of years of education (r=−0.13, p<0.05). All other background/loss variables (time from loss, gender, age of the deceased, kinship to the deceased, and cause of loss) did not influence traumatic grief and depression. 3.2. Hypothesis 1. Negative interpretation of grief reactions and grief-related distress Table 2 shows the zero-order correlations and partial correlations with respect to hypotheses 1–5. As expected, the correlation between negative meaning of grief (NIGS score) and grief-related distress was highly significant. The 11-item NIGS explained 48% of the variance in grief-related distress. The correlation between grief-related distress and grief reactions frequency was r=0.49 (p<0.001). The partial correlation between the NIGS score and grief-related distress, controlling for grief reactions frequency, remained highly significant (rp=0.61, p<0.001). Table 2. Correlations of negative meaning of grief, traumatic grief, and depression, with grief reactions frequency, grief-related distress, and avoidance strategies NIGS score ITG SCL-depression Zero-order (Partiala) Zero-order (Partiala) Zero-order (Partiala) Grief reactions frequency 0.43 0.56 0.51 Grief-related distress 0.69 (0.61) 0.68 (0.55) 0.61 (0.48) Behavioural avoidance 0.40 (0.42) 0.38 (0.42) 0.36 (0.37) Rumination 0.44 (0.36) 0.58 (0.54) 0.45 (0.37) Thought supression 0.36 (0.40) 0.23 (0.31) 0.22 (0.27) Distraction 0.31 (0.26) 0.33 (0.28) 0.24 (0.18*) ITG 0.76 (0.69) SCL-depression 0.75 (0.68) Note: Zero-order and partial correlations are significant at p<0.001, unless otherwise specified. NIGS=Negative Interpretations of Grief Scale. ITG=Inventory of Traumatic Grief. SCL=Symptom Checklist. *p<0.01. a Partial correlations controlling for grief reactions frequency. Table options 3.3. Hypothesis 2. Grief-related distress and traumatic grief and depression The degree of grief-related distress was significantly correlated with traumatic grief symptom severity (ITG) and depression (SCL depression) (Table 2). The correlation between grief-related distress and the ITG was significantly greater than the correlation between grief reactions frequency and the ITG (r=0.68 vs. r=0.56, Z=2.109, p<0.05). In addition, as a trend, the correlation between grief-related distress and depression was significantly greater than the correlation between grief reactions frequency and depression (r=0.61 vs. r=0.51, Z=1.571, p=0.06). Partial correlations of grief-related distress with the ITG and the SCL-depression subscale, after controlling for grief reactions frequency, remained highly significant. 3.4. Hypothesis 3. Negative interpretation of grief reactions and avoidance The negative meaning of grief (NIGS score) was significantly related to behavioural avoidance, rumination, thought suppression and distraction, and remained significantly correlated with all four avoidance strategies when the frequency of grief reactions was statistically controlled (see Table 2). 3.5. Hypothesis 4. Avoidance and traumatic grief and depression The four avoidance strategies were all significantly associated with traumatic grief (ITG) and depression (SCL-depression), even when controlling for the frequency of grief reactions (Table 2). The correlations of avoidance strategies with traumatic grief hardly changed when the ITG item that measures avoidance (“I go out of my way to avoid reminders that the deceased is gone”) was excluded: correlation with behavioural avoidance: r=0.37, rumination: r=0.58, thought suppression: r=0.21, distraction: r=0.32 (all ps<0.001). 3.6. Hypothesis 5. Negative interpretation of grief reactions and traumatic grief and depression The negative meaning of grief reactions (NIGS score) was significantly related to traumatic grief and depression (see Table 2). The NIGS explained 58% of the variance in traumatic grief symptoms and 55% of the variance in depressive symptoms. When controlling for the frequency of grief reactions, the negative meaning of grief reactions remained significantly related to traumatic grief (rp=0.69) and depression (rp=0.68, ps<0.001). 3.7. Multiple regression analyses Hierarchical multiple regression analyses were used to test whether grief related distress, the four avoidance strategies, and negative interpretations of grief explained variance in traumatic grief and depression over and above the frequency of grief reactions and the two background variables that were associated with symptom severity (i.e., the respondent's age and number of years of education). For both symptom measures, six multiple regression analyses were conducted. In each analysis, the frequency of grief reactions, age and years of education were simultaneously forced into the equation in block 1. Then, in block 2, one of the other six variables was entered to see if it explained variance over and above the other variables. Table 3 summarises the results of these analyses. Table 3. Variance in traumatic grief and depression explained by background/loss variables and grief-related distress, avoidance strategies and negative meaning of grief Dependent variable=ITG Dependent variable=SCL-depression ΔR2 F ΔR2 F Variables entered Step 1: grief reactions frequency educational level, and age 0.35 38.90 0.26 25.65 Step 2: grief-related distress 0.20 93.64 0.16 61.42 Step 2: behavioural avoidance 0.10 40.96 0.09 29.39 Step 2: rumination 0.17 74.99 0.09 30.48 Step 2: thought suppression 0.06 21.86 0.05 15.77 Step 2: distraction 0.05 16.92 0.02 5.47* Step 2: negative meaning of grief 0.30 187.03 0.33 177.98 Note: All F-values are significant at p<0.001 unless otherwise specified. ITG=Inventory of Traumatic Grief. SCL=Symptom Checklist. *p<0.05. Table options Grief frequency and the two background variables explained 35% of the variance in traumatic grief and 26% of the variance in depression. F-tests showed that all six variables produced a significant increase in the explained variance in traumatic grief and depression, over and above the variance explained by these variables. Next, we examined what combination of variables (i.e., negative interpretations of grief and the four avoidance strategies) explained most variance in traumatic grief and depression, after controlling for the variance explained by grief frequency and the two relevant background variables. Two multiple regression analyses were conducted in which these latter three variables were entered in block 1, and the other variables were included in a stepwise analyses in block 2. Results of these analyses are summarised in Table 4. Table 4. Summary of hierarchical regression analyses for variables predicting traumatic grief and depression Variables added (in order) ΔR2 F DF p ITG Block 1 (enter) Grief frequency/age/education 0.35 38.58 3,220 0.000 Block 2 (stepwise) Negative meaning of grief 0.30 187.03 1,219 0.000 Rumination 0.06 40.67 1,218 0.000 Behavioural avoidance 0.01 7.20 1,217 0.008 SCL depression Block 1 (enter) Grief frequency/age/education 0.26 25.48 3,219 0.000 Block 2 (stepwise) Negative meaning of grief Rumination 0.33 177.98 1,218 0.000 0.01 6.83 1,217 0.010 Note: ITG=Inventory of Traumatic Grief. SCL=Symptom Checklist. Table options In the analyses with the ITG, negative interpretations of grief reactions, rumination, and behavioural avoidance all entered the regression equation, explaining 37% of the variance in traumatic grief, over and above the 35% explained by grief frequency, age, and years of education. In the analyses with the SCL-depression scale, negative interpretations and rumination entered the equation, explaining an additional 34% of the variance in depression, over and above the 26% explained by variables entered in block 1. 3.8. Additional analyses with the shortened ITG It could be argued that confounds in content between symptoms included in the ITG and basically normal grief reactions artificially increased the correlations of traumatic grief with the frequency of normal grief reactions, the degree of distress related to these reactions, and the interpretations assigned to these reactions. To check the effects of this possible overlap, additional analyses were conducted with a shortened version of the ITG (ITG-SV), from which items that could be regarded as representing normal grief reactions, as well as other symptoms that were included in our examples of possible grief reactions, were excluded. More specifically, we excluded five items representing feelings (i.e., yearning, anger, bitterness, feeling envious, loneliness), two items representing imaginary experiences (i.e., hearing the voice of the deceased and seeing the deceased), and four items representing negative thoughts about life, the future, and the world. Thus, the ITG-SV contained 18 items (α=0.91) representing those grief symptoms that were considered to be the most debilitating and maladaptive. Results of the analyses with this shortened ITG were equivalent to the results found with the full-length version. The ITG-SV was highly correlated with grief-related distress, each of the four avoidance strategies, and negative meaning of grief reactions, even when grief reaction frequency and relevant background variables were statistically controlled. Finally, stepwise regression analyses showed that negative meaning of grief, rumination, and behavioural avoidance explained most variance in ITG-SV scores.