Grief work, disclosure and counseling: Do they help the bereaved?

Clinical Psychology Review 25 (2005) 395 – 414

Grief work, disclosure and counseling: Do they help the bereaved? Wolfgang Stroebea,T, Henk Schutb, Margaret S. Stroebeb a

Department of Social and Organizational Psychology, Utrecht University, P.O. Box 80.140, 3508 TC Utrecht, The Netherlands b Department of Clinical Psychology, Utrecht University, The Netherlands

Received 6 July 2004; received in revised form 18 November 2004; accepted 21 January 2005

Abstract Bereavement is associated with increased risk of morbidity and mortality. How to protect the bereaved against extreme suffering and lasting health impairment remains a central research issue. It has been widely accepted that to adjust, the bereaved have to confront and express intense emotions accompanying their loss. It has further been assumed that others assist in this process, and that intervention programs are effective. To assess validity of these assumptions, this article reviews research on the impact of expressing and sharing emotions across four research domains (social support; emotional disclosure; experimentally induced emotional disclosure; and grief intervention). In none of these areas is there evidence that emotional disclosure facilitates adjustment to loss in normal bereavement. Implications of these findings are discussed. D 2005 Elsevier Ltd. All rights reserved.

Suffering the loss of a loved one is a tragedy which is not only characterized by extended periods of anguish and pain, but it also increases risk of depression, physical illness and mortality (e.g., Parkes, 1972/1996; Stroebe & Stroebe, 1987). However, even though loss of a loved one is generally painful,

T Corresponding author. E-mail address: [email protected] (W. Stroebe). 0272-7358/$ - see front matter D 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.cpr.2005.01.004

396

W. Stroebe et al. / Clinical Psychology Review 25 (2005) 395–414

only a small minority of bereaved persons is likely to suffer enduring health impairment.1 The question how we can protect these bereaved from lasting detriments to their health has been a central issue in bereavement research. Since Freud’s (1917) development of the concept of bgrief workQ, it has been generally accepted by bereavement researchers and practitioners alike that a healthy process of adjustment requires that the bereaved confront and express their feelings and reactions to the death of a loved one, and that failure to do so is maladaptive (for dissenting voices, see Bonannon & Kaltman, 1999; Silver & Wortman, 1980; Stroebe & Schut, 1999; Stroebe & Stroebe, 1991; Wortman & Silver, 1989). This article questions assumptions that emotional disclosure (and bgrief workQ) facilitate coping with loss and accelerate adjustment, that support from others during bereavement ameliorates the impact of loss, and that intervention is generally efficacious in facilitating adjustment. The first part of this manuscript presents the theoretical rationale underlying the grief work hypothesis, delineates components of grief work in terms of meaning construction, and discusses the relationship between grief work and emotional disclosure. The second part reviews relevant empirical evidence on four related issues: (1) does social support facilitate adjustment? (2) Does emotional disclosure facilitate adjustment? (3) Does inducing emotional disclosure facilitate adjustment? (4) Do bereavement interventions facilitate adjustment? In reviewing research on these issues we focus mainly (but not exclusively) on marital bereavement, because the majority of studies on the health impact of loss has been conducted on the loss of a marital partner (Stroebe et al., 2001). However, since we are not assuming a qualitative difference between marital bereavement and other types of losses, we will draw on the general bereavement literature whenever relevant studies are available.

1. Grief work, emotional disclosure and adjustment to loss: theoretical rationale 1.1. Grief work and emotional detachment In his classic monograph on bMourning and MelancholiaQ Freud (1917) developed a theory of coping with bereavement, which has had a lasting impact on scientific thinking in the area of bereavement. Freud conceptualized love as the attachment (cathexis) of libidininal energy to the mental representation of the loved person. When a loved person is lost through death, the survivor’s libidinal energy remains attached to the thoughts and memories of the deceased. Since the individual has only a limited pool of energy at his or her disposal, the emotions invested in the deceased have to be detached to enable the person to form new attachments. According to Freud (1917), the psychological function of grief is to free

1

Although prevalence rates for various detriments vary not only according to the particular debility in question, but also across investigations of different bereaved samples, they are consistent in indicating that only a minority of bereaved persons are likely to suffer enduring health impairment. A review of studies of pathological grief reported estimates from different studies ranging from 5% to 33% among acutely bereaved (Middleton, Raphael, Martinek, & Misso, 1993). These prevalence rates are difficult to interpret because criteria have not been established and the definition of pathological grief remains imprecise. Major depressive syndromes have been found to occur in 24–30% of widowed persons 2 months after the death, approximately 24% 4 months after bereavement, and 16% 1 year after bereavement (Shuchter & Zisook, 1993). Research on physical ill-health has also consistently reported elevated rates among bereaved persons on measures of physical symptoms, doctor’s visits, use of medication, disability and hospitalization (Stroebe, Hanson, Stroebe, & Schut, 2001). For example, in our Tqbingen Study to be described later, 20% of the widowed (as compared to 3% of the married) scored above the cut-off point for severe physical symptomatology 4–6 months after loss (declining to 12% after 2 years) (Stroebe & Stroebe, 1993).

W. Stroebe et al. / Clinical Psychology Review 25 (2005) 395–414

397

the individual of his/her ties to the deceased, gaining gradual detachment by means of reviewing the past and dwelling on the future. Freud coined the term bgrief workQ to refer to this process. Freud’s assumptions about the necessity for the bereaved to do their grief work and to express their feelings and reactions to the death of a loved one were followed by other psychoanalytically oriented bereavement theorists (e.g., Deutsch, 1937; Lindemann, 1979; Volkan, 1981). In what has become an influential article on the bAbsence of griefQ Deutsch (1937) even argued that absence of the expression of grief was indicative and/or predictive of complicated grief. As Hagman (2001) described: bInfluenced by Deutsch’s article, analysts and nonanalysts alike have come to view the expression of grief as an essential component of mourning.. . .Many popular forms of bereavement counseling,. . .prescribe that the therapist challenge the bereaved patient’s dresistanceT to mourning, compelling them to express sadness, in the belief that the abreaction of suppressed affect is at the core of successful treatmentQ (p. 16). Similarly, Raphael and Nunn (1988) described the goals of bereavement therapy as follows: b. . .to share the grief work in uncomplicated grief reactions, and in distorted reactions to facilitate their transformation into normal grief reactions and their subsequent resolutionQ (p. 198). The role of counselors of the bereaved has been seen as one in which a person needs bto facilitate the undoing of some of these ties (to the deceased), so that the bereaved is not obsessed with and governed by bonds with the dead to the detriment of future life with the livingQ (Raphael & Nunn, 1988, p. 201). Therefore, many counseling and therapy programs for the bereaved have the goal of helping bereaved persons to adapt to life without the loved one by facilitating grief work (e.g., Worden, 1991). 1.2. Grief work and search for meaning Other bereavement theorists have endorsed the notion that grief work enables the bereaved person to gradually accept that the loss is irrevocable, but have extended their analyses to include aspects of development, for example, that changes in assumptions about the world or bmental modelsQ need to be made (e.g., Bowlby, 1980; Parkes, 1972/1996). These suggestions are consistent with more general cognitive theories of coping with traumatic events, which have proposed that an important task in coping with a traumatic event is the development of an understanding of the event: a search for meaning (e.g., Affleck & Tennen, 1996; Folkman, 2001; Frankl, 1959/1984; Janoff-Bulman, 1992; Taylor, 1983, 1989). These cognitive theories assume that prior to a traumatic event, individuals have personal theories about the world (mental models) that contain information about themselves and their world. These bassumptive worldsQ (Parkes, 1975), which enable us to set goals, plan activities and order our behavior, include beliefs concerning the predictability and controllability of the world, that the world is meaningful, that it operates according to principles of fairness, and that one is safe and secure (JanoffBulman, 1992). Traumatic experiences such as sudden, unexpected and/or violent bereavements can shatter these assumptions. As Parkes (1998) commented, bFor most people in the early stages of bereavement the world is in chaos. . .they feel as if the most central, important aspect of themselves is gone and all that is left is meaningless and irrelevant—hence the world itself has become meaningless and irrelevantQ (p. 79). According to these approaches, recovery from bereavement requires individuals to process the bereavement-related information until it can either be incorporated into their pre-existing theories or until these theories can be modified to allow integration of the new information (Lepore, Silver, Wortman, & Wayment, 1996). Attempts to integrate bereavement-related information into mental models of the world are likely to cause painful thoughts and memories. If people do not confront these painful thoughts, or if

398

W. Stroebe et al. / Clinical Psychology Review 25 (2005) 395–414

they try to avoid them, the traumatic event will not be integrated (Lepore et al., 1996). Thus, according to these cognitive processing theories, individuals engage in grief work in a search for meaning and it is the function of grief work to help them make sense of their loss. There is general agreement on the psychological importance of finding meaning for adjustment to loss and trauma. There is less consensus, however, about what constitutes meaning (Davis, Nolen-Hoeksema, & Larson, 1998; Davis, Wortman, Lehman, & Silver, 2000). There are at least three different senses in which the search for meaning has been conceptualized. In terms of theories of attribution (e.g., Weiner, 1995), a search for meaning would aim at understanding the specific causes of the negative event. Obviously, not all attributions would be equally helpful for adjustment to a loss. For example, taking personal responsibility for the loss (bI could have prevented it, if only. . .Q) would be more painful than attributing the loss to an external and uncontrollable cause (bnobody could have prevented itQ) (Fleming & Robinson, 2001). Other theorists have emphasized the ability of the bereaved to make sense of the loss within their existing fundamental worldview (e.g., Janoff-Bulman, 1992; McIntosh, Silver, & Wortman, 1993; Neimeyer, 2001; Parkes & Weiss, 1983/ 1995). For example, a religious person may make sense of the loss of a loved one by attributing it to God’s will; a patriotic person might find a beneficial interpretation of the death of a son or daughter in war in terms of bservice to the nationQ. A third sense in which theorists have interpreted finding meaning is in terms of the ability of the bereaved to develop new goals and a new, perhaps wider sense of self (Davis et al., 1998; Taylor, 1983, 1989). Thus, Taylor (1983) argued that people who have to cope with traumatic events might find meaning by considering positive implications or benefits of the event. For example, individuals who have lost a loved one sometimes report that it has taught them to place greater value on relationships (e.g., Tedeschi & Calhoun, 1996). Accordingly, the loss of a loved one may be perceived as having led to a beneficial change in one’s life priorities and personal goals. From a psychological perspective, it is reasonable to assume that the different levels on which individuals pursue meaning are interrelated. For example, the outcome of the attributional analysis of the traumatic event can prove problematic for the individual’s mental model of the world as a place which is predictable, controllable and operates according to principles of fairness. According to dissonance theory (Festinger, 1957), individuals might try to resolve such inconsistencies by construing benefits of the traumatic event. As Janoff-Bulman (1992) expressed so eloquently, bBy engaging in interpretations and evaluations that focus on the benefits and lessons learned, survivors emphasize benevolence over malevolence, meaningfulness over randomness, and self-worth over self-abasementQ (p. 133). The fact that perceived benefits can help to reduce the dissonance which a traumatic event may create in the individual’s view of the world as predictable, controllable and benevolent, does not imply that these benefits are necessarily illusory. There is empirical evidence supporting the assumption that people who are confronted with loss experiences search for meaning, particularly if death was sudden, unexpected, and/or due to violence (for a review, see Davis et al., 2000). There is less support for the assumption that finding meaning facilitates adjustment (Davis et al., 1998; McIntosh et al., 1993). We know of only one study which reported a direct facilitative effect of finding meaning. In a longitudinal study of individuals coping with the loss of a family member, Davis et al. (1998) assessed two construals of meaning, namely making sense of the death, and whether the bereaved person had found anything positive in the experience. Both these measures were associated with lower levels of distress following the loss of a loved one, but the strength of this association varied across different phases of the grieving process. Whereas making sense of the

W. Stroebe et al. / Clinical Psychology Review 25 (2005) 395–414

399

loss was most strongly related to adjustment in the first 6 months of the loss, respondents’ reports of benefits were most strongly related to adjustment 13 and 18 months after the loss. 1.3. The relationship between grief work and disclosure As conceptualized by Freud (1917), grief work is essentially an intrapsychic process which involves desensitization of emotionally loaded experiences through repeated exposure to these painful memories. Although social sharing or disclosure of emotions may not be a necessary condition of grief work, as individuals can also confront their grief and work through it in isolation, the two processes are closely linked: people have to confront their loss in order to talk about it. Furthermore, as Lepore et al. (1996) suggested, social sharing of grief may be a strategy to facilitate the process of emotional desensitization: bBy talking with supportive and emphatic others, trauma survivors may be able to contemplate and tolerate aversive trauma-related thoughts for longer periods of time than they would on their ownQ (Lepore et al., 1996, p. 271). With their emphasis on search for meaning, the cognitive processing models of coping with traumatic events assign a much more central role to social sharing of emotion. These theories consider social sharing to be an essential part of the quest by the bereaved to make sense of their loss experience. For example, Harvey, Carlson, Huff, and Green (2001) described the relationship between intrapsychic and interpersonal grief work as follows: bThe account-making and confiding model posits that not only in personal, private work on one’s story, but also in public social interaction, the individual will need to confide part of his or her story to close others over time in order to assimilate different major stressors and lossesQ (p. 235). It is interesting to note that even contemporary psychoanalytic theory has turned its back on the Freudian (1917) conception of grief work as a private, intrapsychic process. As Hagman (2001) stated in his analysis of changes in psychoanalytic thinking, bThe notion that the mind is a private, closed system that primarily functions to regulate its own inner world of energies and defenses is essentially defunct. Modern psychoanalysis has recognized that human psychological life is profoundly relational. . .A central feature of virtually all of the recent critiques of the standard model is that the intrapsychic focus does not convey the role of other people and the social milieu in facilitating or impeding recovery from bereavementQ (p. 21).

2. Grief work, emotional disclosure and adjustment: empirical evidence 2.1. Social support and adjustment to loss In order to share one’s emotions, one needs others who are willing to listen. Measures of perceived social support reflect the availability of others to whom one can disclose one’s emotions (e.g., Cohen & Wills, 1985; Stroebe & Stroebe, 1996). The assumption that support from family and friends is one of the most important moderators of bereavement outcome is widely accepted among bereavement researchers and practitioners (e.g., Lopata, 1973; Stroebe & Stroebe, 1987; Stylianos & Vachon, 1993). According to Cohen and Wills’ (1985) classic analysis, social support can play a role at two different points in the causal chain that links stress to illness: social support can influence stress appraisal and/or it can result in binhibition of maladjustive and facilitation of adjustive responsesQ (p. 313). Thus, the

400

W. Stroebe et al. / Clinical Psychology Review 25 (2005) 395–414

knowledge that one can call on the support of friends and family members, and that one does not have to face a lonely future, may help to soften the blow of the loss and buffer one against the deleterious effects of bereavement. More relevant in the context of this article is the second pathway, namely the facilitation of coping through the inhibition of maladjustive and the facilitation of adjustive counter responses. Accordingly, even if the availability of social support should fail to buffer the individual against the impact of bereavement by attenuating stress appraisal, the emotional support from family and friends should facilitate grief work by making it possible for the bereaved to express their feelings and reactions to the death of a loved one. Thus, by encouraging the expression of emotion, the provision of social support could accelerate recovery and promote long-term adjustment. Whereas buffering effects can be studied cross-sectionally, namely, by comparing symptom levels of married and recently bereaved individuals of different levels of social support (assessed through a social support scale), the study of recovery effects requires a longitudinal design. Recovery effects reflect differential rates of adjustment of bereaved samples experiencing differing levels of social support. One therefore needs to assess the impact of bereavement at least at two points in time, following the loss experience. The effect of social support on recovery would be indicated by a statistical interaction between social support and time since bereavement, with the more highly supported bereaved showing greater improvement over time. With regard to measures of social support, Cohen and Wills (1985) distinguished two types, namely structural measures which assess merely the existence of relationships and functional measures which directly assess the extent to which these relationship provide particular support functions (e.g., emotional social support; instrumental social support). Cohen and Wills (1985) argued that buffering and/or recovery effects will only occur if there is ba reasonable match between the coping requirements and the available supportQ, that is bwhen the support functions measured are those which are most relevant for the stressors faced by the personQ (p. 314). Buffering and/or recovery effects are therefore unlikely to be found when structural measures are used to assess social support, because structural measures bassess only the existence or number of relationships and do not provide sensitive measures of the functions actually provided by those relationshipsQ (p. 314). We have found only nine studies of social support as a moderator of the impact of bereavement on distress which used methodologies appropriate for assessing buffering and/or recovery effects either with regard to death of a spouse (Greene & Feld, 1989; Norris & Murrell, 1990; Okabayashi et al., 1997; Stroebe, Stroebe, Abakoumkin, & Schut, 1996; Stroebe, Zech, Stroebe, & Abakoumkin, in press) or some other types of bereavement (Krause, 1986; Murphy, 1988; Murphy, Chung, & Johnson, 2002). Four of these studies focused exclusively on the buffering hypothesis (Krause, 1986; Greene & Feld, 1989; Murphy, 1988; Norris & Murrell, 1990; Okabayashi et al., 1997)2, two assessed both buffering and recovery effects (Stroebe et al., 1996; Stroebe et al., in press), and two were limited to the study of recovery effects (Nolen-Hoeksema & Davis, 1999; Murphy et al., 2002). We focus on reviewing studies of recovery effects, because–following the line of reasoning outlined above–these are most relevant to our argument. The Tqbingen Longitudinal Study of Bereavement assessed social support and depressive symptoms in a matched sample of 60 recently widowed and 60 married men and women (Stroebe et al., 1996).

2

A study of Schwarzer (1992) was not considered appropriate, because she used the frequency of hospital visits as her measure of social support. A study by Hays, Kasl, and Jacobs (1994) assessed perceived social support but did not report appropriate statistical analyses.

W. Stroebe et al. / Clinical Psychology Review 25 (2005) 395–414

401

Depression (BDI)

16 14 12 10 8 6 4 2 0 TIME 1

TIME 2

TIME 3

Widowers Low Support Widows

TIME 1

TIME 2

TIME 3

Widows High Support Widows

Fig. 1. Depressive symptoms of widowed men and women with high or low perceived social support at three points in time.

Participants were interviewed three times. The first interview was conducted at 4 to 7 months after the loss, the second interview approximately 14 months, and third approximately 2 years after the loss. Perceived social support was measured at the first wave with a newly constructed questionnaire measure3 which was very similar to the Interpersonal Support Evaluation List (ISEL; Cohen, Mermelstein, Kamarck, & Hoberman, 1985). Depressive symptoms were assessed with the German version of the Beck Depression Inventory (BDI; Kammer, 1983) at all three points in time. Fig. 1 presents the mean BDI scores (median-split into groups of high or low social support) for the bereaved men and women at the three points of measurement. As can be seen, while there is improvement over time for both high and low support groups, there is no evidence that social support accelerated recovery from bereavement (i.e., the high and low support trajectories are parallel). This conclusion (as well as the absence of a buffering effect) is supported by statistical analyses reported in Stroebe et al. (1996).4 The second study used data that had been collected in the USA as part of a larger investigation entitled Changing Lives of Older Couples (CLOC, see, e.g., Carr et al., 2000). The CLOC study has two features which make it particularly useful for extending the Tqbingen findings: first, and most importantly, social support and depressive symptoms of the bereaved were assessed before and after the respondents lost their partner. Second, the CLOC study was conducted in a different country, thus enabling us to check the cross-national stability of our findings. The CLOC study employed a two-stage area probability sampling technique to collect information from elderly married individuals from Detroit, Michigan. Of the 319 respondents who lost a spouse during the study period, 250 participated at Wave 1 3

Our measure assessed four components of social support, namely emotional, instrumental, appraisal, and social contact support. Since the four subscales were highly correlated with each other, only the overall score was used. 4 A reviewer of the present manuscript raised the interesting point that the absence of a buffering effect in this study could have been due to the fact that only a minority of bereaved are highly distressed. Thus, even if this subgroup showed evidence of buffering, this effect might have been alleviated by the absence of buffering among the non-distressed majority. To argue against this explanation, we would like to point out that at 4 to 7 months after their loss, 42% of the widowed (as compared to 10% of the married) had scores on the BDI that indicated at least mild depression. Although 2 years after the loss, the percentage of widowed with BDI scores that indicated at least mild depression had fallen to 27%, it was still significantly higher than the 10% figure for the married. Given the high initial level of distress and the decline over time, one would have expected an interaction between social support and time on BDI scores, if social support had been more effective in alleviating the distress of bereaved with high rather than low levels of distress. More directly, there should have been a buffering effect for initially highly distressed bereaved. But, as further analysis showed, even for the highly distressed bereaved (i.e., BDI scores above the median at 4 to 7 months), there is no evidence that social support accelerated adjustment to their loss during the courses of our study.

402

W. Stroebe et al. / Clinical Psychology Review 25 (2005) 395–414

(6 months), of which 210 participated again at Wave 2 (18 months), and 106 at Wave 3 (48 months after the death). Matched controls from the original sample were interviewed at comparable times from baseline. Depression was assessed with a subset of 11 items from the 20-item Center for Epidemiologic Studies Depression (CES-D) scale (Radloff, 1977) and scores were standardized. Social support at baseline was assessed with a 4-item scale asking how much respondents felt loved and cared for by family and friends and whether family and friends were willing to listen to them talk about their worries and problems. Fig. 2 presents the mean CES-D scores for the bereaved sample (median-split into high and low support groups) for the baseline as well as three post-bereavement points in time. With the lines linking the mean CES-D-values being reasonably parallel, no buffering or recovery effects were found (Stroebe et al., in press). The third study was not on partner loss, the focus of the studies reviewed above. Rather, it assessed the role of social support in facilitating parents’ recovery from the loss of an adolescent or young child due to violence (Murphy et al., 2002). In this longitudinal study, 173 bereaved parents were assessed four times over a period of 5 years. Perceived social support was assessed at intake with a measure of bpersonal gratification resulting from the perceived effectiveness of network members’ support in reducing stress and restoring emotional and instrumental equilibriumQ (p. 429). Distress was assessed both at intake and at subsequent waves (BSI; Derogatis & Melisaratos, 1983). Using latent growth modeling, the authors examined the impact of levels of social support on the rate of change in distress over time. There was no significant effect of social support on the rate of change in distress over time and thus no evidence for a recovery effect. The fourth study tested the hypothesis that people with a ruminative coping style, who tended to focus excessively on their own emotional reaction to trauma, compared to those without a ruminative coping style, would benefit more from emotional social support (Nolen-Hoeksema & Morrow, 1991). Ruminators are individuals who focus on their depressive symptoms and ruminate on the causes and implications of these symptoms. The authors reasoned that supportive others might help ruminators to challenge negative irrational thoughts about the trauma and to help them break free of their cycle of negative thinking. These assumptions were tested in a longitudinal study of individuals, who had lost a close family member and who were interviewed three times following their loss. Rumination was measured with the Ruminative Response subscale of the RSQ (Nolen-Hoeksema & Morrow, 1991). In 0,7 0,6

Depression (CES)

0,5 0,4 0,3 0,2 0,1 0 -0,1

Baseline

6 Month

18 Month

48 Month

-0,2 -0,3 -0,4 Low Support

High Support

Fig. 2. Depressive symptoms of individuals with high or low perceived social support before, and three points after bereavement (adapted from Stroebe et al., in press).

W. Stroebe et al. / Clinical Psychology Review 25 (2005) 395–414

403

apparent support of their hypothesis, Nolen-Hoeksema and Davis (1999) found that high ruminators appeared to benefit from emotional social support but low ruminators did not. The authors concluded that when bhigh ruminators perceived that they were receiving positive emotional support, their level of depression dropped significantly. Among low ruminators, higher levels of emotional support at a particular wave were not at all associated with levels of depression at that waveQ (p. 807). However, although receiving emotional social support contributed more to the psychological well-being of high than low ruminators, there is no indication in the findings of the Nolen-Hoeksema and Davis (1999) study that social support accelerated adjustment to bereavement in the group of high ruminators: the trajectory for the depressive symptoms of high ruminators who received high levels of emotional support is parallel to that of high ruminators who received low levels of emotional support (i.e., no evidence for a rumination by social support by time interaction). Thus, results of the two studies of the impact of social support on coping with the loss of a partner (Stroebe et al., 1996, in press) as well as the two studies of the role of social support in coping with other losses (Nolen-Hoeksema & Davis, 1999; Murphy et al., 2002) provide a negative answer to our first question, whether social support facilitates adjustment to the loss of a loved one. There was no evidence that bereaved individuals who enjoyed high levels of social support adjusted more easily to their loss than individuals who had lower levels of social support. This does not imply, however, that the bereaved are not in need of social support or that social support does not help to alleviate their distress. They do benefit. Social support was associated with reduced levels of depressive symptoms in three of the four studies. However, it was equally helpful for bereaved and non-bereaved individuals alike. 2.2. Emotional disclosure and adjustment to loss Availability of individuals who are willing to listen is a necessary but not sufficient condition for the sharing of emotion. The fact that a bereaved individual is embedded in a social network of family and friends willing to serve as an audience for emotional disclosure does not necessarily imply that the bereaved will actually make use of the opportunity. If one wants to know the extent to which bereaved individuals disclose their emotions to family and friends, one does have to assess emotional disclosure directly. An early study by Pennebaker and O’Heeron (1984), which included a direct measure of emotional disclosure, appeared to provide support for the assumption that disclosure alleviates distress in bereaved individuals. They investigated the impact of emotional disclosure on health outcome in a survey of 19 individuals who had suffered the death of a spouse a year previously. Pennebaker and O’Heeron found that the more the bereaved individuals discussed their loss with others, the fewer health problems they had. However, the study had a number of methodological features, which raise doubt regarding the validity of the authors’ conclusions. The study was retrospective, in that individuals who had been bereaved for a year were asked to report the health complaints for the past year and for the year before their loss. On the basis of these reports, a change score was computed, and this change score showed a positive association with disclosure. However, it seems doubtful that individuals were able to realistically recall their health status of 2 years previously, particularly given that the 2-year period was intersected by a dramatic life event. Furthermore, a recent reanalysis of the original data, reported by Pennebaker, Zech, and Rime´ (2001), found that even though disclosure was strongly positively associated with this change score, it was mildly negatively associated with the number of complaints reported for the year after bereavement.

404

W. Stroebe et al. / Clinical Psychology Review 25 (2005) 395–414

The longitudinal study of mothers who had lost a child through SIDS by Lepore et al. (1996), which included a direct measure of emotional disclosure, also suggests that social sharing facilitates coping with loss. The aim of the study was to examine how social constraints on discussion of traumatic experiences can interfere with processing and recovery from loss. (Relationship constraint was measured with a rating scale reflecting the extent to which mothers felt they could share their grief.) This study showed that, for mothers with high levels of social constraint, having intrusive thoughts about their loss at 3 months post loss was associated with increased depressive symptoms 18 months later. In contrast, having intrusive thoughts was negatively related to depressive symptoms for mothers with low levels of social constraint. In discussing these findings, the authors argued that bit appears that unconstrained social relationships might facilitate the processing of traumatic events and emotional recoveryQ (Lepore et al., 1996, p. 279). The assumption implied here is that unconstrained social relationships facilitated emotional recovery because they allowed those mothers to talk about their loss. But if this were correct, then statistically controlling for the extent to which these mothers talked about their loss at each point in time should have reduced or eliminated the positive relationship between lack of constraint and recovery. Since the authors did not report such an analysis (even though emotional disclosure had been measured at each point in time), it remains uncertain whether these findings should be interpreted as support for a facilitative role of disclosure in coping with loss. A second reason why we are reluctant to accept this interpretation is that there is no support for the hypothesis that disclosure helps in the two longitudinal studies which related emotional disclosure directly to recovery. In the Tqbingen study, there was no evidence that talking about a loss was prospectively associated with improvements in depressive symptoms once health status, assessed at the same time as disclosure, had been controlled for (Stroebe & Stroebe, 1991). However, the fact that the Tqbingen study used only a 1-item measure of emotional disclosure tends to threaten the validity of these conclusions. Such criticism does not apply to the Utrecht Longitudinal Study of Bereavement, which used a 5-item scale to measure the extent to which the recently bereaved participants shared their grief (Schut, 1992; Stroebe, Stroebe, Schut, Zech, & van den Bout, 2002). One hundred and twenty eight participants were assessed four times during the 2 years following their loss and at each point in time emotional disclosure and psychological adjustment were assessed. Linear structural analysis was used to test the central hypothesis that disclosure at each point in time was associated with reduced distress at the subsequent point of measurement. None of the links between disclosure and subsequent distress even approached significance. Thus, there was no support for the assumption that sharing one’s thoughts and feelings with others after the loss of a loved one alleviated distress. Should we conclude from these findings that social sharing of emotions and confronting a loss does not facilitate coping? Not necessarily. These studies only provide information about the quantity but not the quality of emotional disclosure among the bereaved in our samples. Grief work implies a process of confronting a loss—an active ongoing effortful attempt to come to terms emotionally with the loss. Beneficial disclosure outcomes are only likely to occur if people recognize and acknowledge personally stressful experiences, access and activate emotional memories of those experiences, identify and put their emotions into words, and eventually come to think differently about the experience (Lumley, Tojek, & Macklem, 2002). Many of the bereaved in our samples may have engaged in rumination, that is, in thoughts and behaviors that maintain one’s focus on one’s negative emotions. There is a great deal of evidence that rumination aggravates depressive symptoms (Nolen-Hoeksema, 1991).

W. Stroebe et al. / Clinical Psychology Review 25 (2005) 395–414

405

2.3. Induced emotional disclosure and adjustment The experimental work of Pennebaker et al. (e.g., Pennebaker & Beall, 1986; Pennebaker, Colder, & Sharp, 1990; Pennebaker, Kiecolt-Glaser, & Glaser, 1988) suggests a procedure which should enable researchers to experimentally induce grief work in bereaved research participants. Participants in studies employing the bPennebaker paradigmQ are randomly assigned to either a bTrauma conditionQ or a bControl ConditionQ. They typically have to come to the laboratory for several consecutive days to write for 15 to 30 min either about the trauma or about a trivial topic. Those who write about some trauma are asked to bwrite about their very deepest thoughts and feelingsQ about this traumatic event. Pennebaker suggested that being induced to write about a traumatic event would result in health improvement for two reasons, which correspond to the two interpretations of the grief work hypothesis discussed earlier. The first reason is that the writing task might break down an individual’s inhibition to confront threatening thoughts. The assumption implicit in this explanation is that inhibition is unhealthy. More recently, Pennebaker suggested that disclosure might be effective because b. . .it helps people to gain meaning about their experiences, reframe these experiences as non-threatening, assimilate them into the self, and in some cases, engage in dramatic reconstruction of the self-systemQ (Pennebaker & Keough, 1999, pp. 109–110). Until recently, reviews of the effectiveness of the Pennebaker paradigm were extremely positive. For example, a meta-analysis of 13 studies using this paradigm concluded that bThis writing task was found to lead to significantly improved health outcomes in healthy participants. Health was enhanced in 4 outcome types—reported physical health, psychological well-being, physiological functioning, and general functioning. . .Q (Smyth, 1998, p. 174). Smyth (1998) further concluded that this writing task resulted in (medium-sized) effects which are similar to that of other psychological treatment. Subsequent to this meta-analysis, numerous relevant studies have been conducted, and a more recent meta-analysis of 61 randomized, controlled trials by Meads and Nouwen (in press) has raised some doubt about the effectiveness of the writing procedure in improving the health of healthy participants.5 No significant effects of emotional disclosure could be found on either psychological or physical health outcomes. There were no effects of interventions on measures of anxiety, depression, the impact of events or health center visits (whether objectively measured or self-reported). However, some improvements could be found on some self-report measures (pain, sleep quality, physical dysfunction, physical symptoms, perceived somatic symptoms), at least in people with preexisting morbidity. Although Meads and Nouwen only included randomized controlled trials in their review, they could be criticized for not weighing these studies by other methodological features (e.g., in terms of how close the manipulation in a given study emulated the Pennebaker paradigm). However, unless one would argue that emotional disclosure manipulations which deviate from the Pennebaker paradigm could have a negative health impact (rather than no effect), it is surprising that the presence of such studies in a metaanalysis could have completely wiped out the positive effects of studies which closely followed the paradigm (Smyth had computed a fail-safe N of 117). Therefore, the failure of the meta-analysis of Meads and Nouwen to find significant overall effects for the impact of emotional disclosure, even on measures of health center visits, does raise some doubt regarding the general effectiveness of this

5

A more extensive version of this review is available on the following website http://www.pcpoh.bham.ac.uk/publichealth/wmhtac/pdf/ emotionaldisclosure.htm.

406

W. Stroebe et al. / Clinical Psychology Review 25 (2005) 395–414

procedure. More specifically, it suggests that the emotional disclosure manipulation may only be beneficial for specific populations under some specific situation. Even if the Pennebaker paradigm were to be proven ineffective for healthy adults, this would not preclude its efficacy in cases of bereavement, which after all is one of the most stressful life events. However, the few studies that have investigated the efficacy of the disclosure paradigm within the bereavement domain have yielded generally negative results. Segal, Bogaards, Becker, and Chatman (1999; see also, e.g., Segal, Chatman, Bogaards, & Becker, 2001) assigned 30 older widowed persons to either a treatment or a delayed treatment group. Respondents in the treatment group were instructed to talk about the loss of their spouse and to express their deepest emotions in four 20 min vocal expressions sessions over a 2 week period. Experimental comparison data were only available for the first posttest, because the delayed intervention group was given the essay-writing task shortly after the posttest had been administered to the immediate intervention group. Segal et al. claimed some effects of their disclosure manipulation at follow-up, but it is difficult to attribute effects specifically to the intervention procedure. As the authors themselves noted, bThe weakness [of this study] is the lack of a control group against which to measure the follow-up effects. Without this experimental control, it is impossible to attribute the changes to the interventionQ (Segal et al., 1999, p. 307). The reported improvements could have been due to the passage of time, or demand characteristics, rather than to the intervention. A study by Stroebe et al. (2002) also failed to find any positive effects of disclosure on bereaved adults. Analyses were based on data from 119 recently bereaved men and women who had been randomly assigned to either the Pennebaker writing task, or to a no-essay control condition. The writing task did not result in a reduction of distress or of doctors’ visits either immediately after, or at a 6-month follow-up. Stroebe et al. (2002) also failed to find support for the hypothesis advanced by Pennebaker et al. (2001) that individuals would benefit more from disclosure following a sudden, unexpected than an expected loss. Beneficial effects could not even be demonstrated for those bereaved who had suffered an unexpected loss, nor for those, who at the time of the study had expressed a high need for emotional disclosure. Studies using the writing task in non-spousal bereavement also failed to find any beneficial effects. For example, a study conducted among 64 undergraduates who had experienced the accidental or homicidal death of a significant person also failed to find effects of written disclosure (Range, Kovac, & Marion, 2000). Participants were asked to write on 4 consecutive days for 15 min about their deepest emotions and thoughts surrounding the death of their loved one, or to write about some trivial topic. There was only a time effect, but no significant interactions by writing condition. There were negative findings not only on psychological measures, but also on reported number of visits to doctors. Another non-spousal bereavement study by Kovac and Range (2000) found at least partial effects of a written disclosure manipulation. They investigated the impact of this task on 40 individuals who had lost a loved one to suicide in the previous 2 years. They found no indication of general, overall improvement in the group that wrote about their deepest feelings rather than a trivial topic, but there was evidence of a beneficial effect between posttest and follow-up for the manipulation compared with the control condition on a measure that assessed specific aspects of reactions to suicidal bereavement. Further null-effects of writing about a past bereavement were reported recently in a disclosure intervention study by Bower, Kemeny, Taylor, and Fahey (2003). These investigators examined changes in meaning-related goals as well as in immune function following diary writing. Participants were 40 women who had lost a close relative to cancer, and who were themselves at risk of this disease. Written disclosure about the death, as compared with writing about non-emotional topics, did not induce changes in these parameters (although a positive association was found between reporting positive changes in

W. Stroebe et al. / Clinical Psychology Review 25 (2005) 395–414

407

meaning-related goals and improvement in immune function for both bereavement and control conditions). It is important to note, however, that some of the participants in this study had suffered the bereavement a long time ago. Bereavement research by Bonanno et al. also included a manipulation of disclosure, although it did not strictly follow the Pennebaker paradigm (e.g., Bonanno, Keltner, Holen, & Horowitz, 1995; Bonanno, Notarius, Gunzerath, Keltner, & Horowitz, 1998). Adults who had lost their spouses within the previous 6 months were asked into a laboratory to talk about their deceased partner for 6–10 min. The more emotional and the more they talked about their spouse’s death, the more poorly they coped at a later point in time. However, it remains unclear from these results whether disclosure reflected grief or whether it affected it (Pennebaker et al., 2001). Many studies report high symptomatology and expression of grief early in bereavement as being highly predictive of grief later on (e.g., Stroebe & Stroebe, 1991). All in all, there is little evidence that induced disclosure of emotions is effective in coping with bereavement. Of the five studies that used the Pennebaker paradigm to assess the impact of essay writing on coping with loss, only one study (Kovac & Range, 2000) found evidence of a statistically significant beneficial effect, but this effect was limited to one of several health measures included in that study. While methodological weaknesses could account for some failures to find an effect, even studies which followed the paradigm procedure closely failed to observe an impact of induced disclosure on recovery. 2.4. Grief counseling and adjustment to loss Offering social support, encouraging emotional disclosure, and helping the bereaved to reframe or reinterpret the meaning of their loss experience are the main functions of bereavement counseling and therapy (Raphael & Nunn, 1988). Since disclosure in these situations is guided by trained counselors or therapists, inducing emotional disclosure in grief counseling or therapy might be effective, even if essay writing is not. It is therefore interesting to note that the failure of inducing emotional disclosure to show consistent beneficial effects in the Pennebaker paradigm–noted above–is paralleled by a similar failure in the effectiveness of bereavement interventions. A number of recent reviews of controlled studies of bereavement interventions have concluded that these interventions are rather ineffective (Jordan & Neimeyer, 2003; Kato & Mann, 1999; Schut, Stroebe, van den Bout, & Terheggen, 2001). Next, we examine these conclusions in more detail. Kato and Mann (1999) reported both a narrative and a meta-analytic review of bereavement intervention studies. They restricted their review to controlled intervention studies which required random assignment to treatment and control groups and similar recruitment procedures for both groups. Their meta-analysis, based on 13 studies, reported average effect sizes (Cohen’s d) for measures of depression and grief of 0.052, of 0.272 on physical symptoms, and of 0.095 for other psychological symptoms. Based on 37 effect size measures, the global average effect size across all types of variables was 0.114. The authors concluded that bThe effect sizes for these studies suggest that psychological interventions for bereavement are not effective interventions for the depression experienced by the bereavedQ (p. 293).6 6 A somewhat greater effect size of 0.43 was reported by Allumbaugh and Hoyt (1999). However, since this meta-analysis included studies that used single-group (pre-posttest) designs, it is not really relevant in the context of a discussion about whether the improvement due to bereavement interventions exceeds that of untreated controls.

408

W. Stroebe et al. / Clinical Psychology Review 25 (2005) 395–414

The global average effect size of the meta-analysis of Kato and Mann (1999) is similar in size to the global effect size of an unpublished meta-analysis by Fortner and Neimeyer (reported in Jordan & Neimeyer, 2003). This meta-analysis was based on controlled intervention studies published between 1975 and 1998. Fortner and Neimeyer found an overall effect size of 0.13 across the sample of 23 studies. To further clarify their findings, Jordan and Neimeyer investigated the differential effectiveness of variables such as length of therapy, credentials of the therapist (professional vs. non-professional), modality of treatment (individual vs. group), and theoretical approach used by the therapist. They found no association between these variables and effect size. However, they did find that the effect size increased with greater length of time since death and higher levels of risk (i.e., sudden violent death or chronic grief). A narrative review of controlled studies of bereavement interventions has been published by Schut et al. (2001). This review is particularly relevant here, because these authors categorized interventions into three groups, namely (1) primary preventive interventions that were open to all bereaved, (2) secondary preventive interventions that target bereaved individuals who, through screening of risk factors, can be considered vulnerable, and (3) tertiary preventive interventions directed at bereaved who suffered from complicated grief. Based on 16 controlled studies of primary interventions, these authors drew the following conclusions: bbased on the evidence to date, outreaching primary preventive interventions for bereaved people cannot be regarded as being beneficial in terms of diminishing grief-related symptoms, with a possible exception of interventions being offered to bereaved childrenQ (p. 731). It is important to note that the effectiveness of primary preventive intervention has only been studied in an outreaching mode. Outreach accesses bereaved persons irrespective of their grieving status, thereby including the majority for whom there are no complications, and for whom grief must take its course across the passage of time. Only a minority of those approached could actually be expected to need and benefit from intervention. Thus, the above findings may not apply to bereaved individuals who themselves feel a strong enough need for counseling to take the initiative to join intervention programs. This is a critical distinction requiring further research (see below). Findings with regard to secondary preventive interventions were somewhat more positive. Seven studies could be included in the category that focused on bereaved persons who were at high risk for developing bereavement-related problems (e.g., traumatic death; loss of a child; high levels of distress). Schut et al. concluded that these secondary preventive interventions resulted in some, albeit rather modest, effects. They also found that studies which specifically screened for high levels of distress (rather than selecting on the basis of membership of a high risk category) tended to show better results for the interventions. Tertiary interventions which focused on bereaved individuals who had already developed a complicated grief reaction were most effective. Interventions with individuals who suffer from complicated grief usually take place longer after bereavement, primarily because most forms of complicated or traumatic grief take time to develop. Finally, studies of the effectiveness of tertiary preventive intervention usually take place through processes of help-seeking rather than recruitment. Based on a review of seven studies, the authors concluded that bmost studies. . .(of interventions with individuals, who suffer from complicated grief). . .find positive and lasting resultsQ (p. 729). The general pattern to emerge from these reviews of the effectiveness of grief counseling and therapy is that the chances of such interventions being effective appear to increase the more complicated the grief process appears to be or to become. As Parkes (1998), the doyen of grief research and one of the world’s most experienced grief therapists stated, bThere is no evidence that all bereaved people will benefit from

W. Stroebe et al. / Clinical Psychology Review 25 (2005) 395–414

409

counseling and research has shown no benefit to arise from the routine referral of people to counseling for no other reason than that they have suffered bereavementQ (p. 18).

3. Conclusions The findings reviewed in this article challenge beliefs about grief work, emotional disclosure, and beneficial interventions that were considered truisms and as practically self-evident by bereavement researchers only a decade ago. The pattern of findings which emerges from our review is very consistent. Although the support of family and friends is positively associated with well-being among the bereaved, there is little evidence that social support moderates the impact of bereavement on psychological health and/or accelerates adjustment to the loss and this despite the fact that distress levels are rather high among the recently bereaved. Since we had assumed that the facilitation of emotional disclosure was one of the ways through which social support would facilitate coping with loss, the failure to find evidence that disclosure of emotions (whether natural or induced) facilitates adjustment to loss goes some way toward explaining why social support failed to enhance recovery from bereavement. We have argued that offering social support, encouraging emotional disclosure, and helping the bereaved to reframe or reinterpret the meaning of their loss experience are the main functions of bereavement counseling and therapy. The fact that grief counseling and therapy have not proved effective for individuals who were referred to counseling for no other reason than that they had suffered bereavement is also consistent with the general pattern of findings that emerged from our review. Primary preventive interventions are likely to reach the same type of bereaved individuals who are usually being recruited into bereavement research. Why do bereaved individuals with uncomplicated courses of grief fail to derive benefits from assistance with their grief work? According to findings from the Tqbingen study (Stroebe et al., 1996), the most common emotional difficulty suffered by the bereaved is emotional loneliness: the missing of the deceased and the feeling of being utterly alone, even when in the company of friends and family. This finding is consistent with the theoretical reasoning of Weiss (1975) that the loneliness of emotional isolation appears in the absence of a close emotional attachment and can only be remedied by the integration of another emotional attachment or the reintegration of the one who has been lost. Obviously, the latter solution is not feasible in the case of bereavement. It is possible that this type of loneliness only abates with time and that nothing can be done to further the recovery process. There is evidence that losses which are sudden, unexpected and/or due to violence increase the likelihood of grief complications (Stroebe & Schut, 2001). By contrast, bereaved individuals with uncomplicated courses of grief will typically have suffered losses which were less traumatic. Thus, persons in the latter category may have had fewer difficulties in making sense of their loss, because their experience may somehow bfitQ their mental models (e.g., following the expected, appropriate death of an elderly person following a life well-lived). Even losses that were unexpected may fit mental models if, for example, the life-style of the deceased increased the risk of an early death. Thus, even though the spouse of a deceased person who had been overweight and a heavy smoker may in retrospect be angry at him or her for not taking better care, the bereaved partner may not need any assistance in making sense of the loss. There are also individual differences in the vulnerability to bereavement complications. Attachment theory provides a useful framework for understanding different patterns of adjustment to loss (Parkes, 2001; Shaver & Tancredy, 2001; Stroebe et al., in press). According to Bowlby (1980), how individuals

410

W. Stroebe et al. / Clinical Psychology Review 25 (2005) 395–414

react to loss depends on certain childhood experiences, in particular, on the pattern of parental attachment behavior. A secure base has to be provided for the infant by the attachment figure to enable development of a secure pattern of attachment. The child builds implicit mental models of relationships between self and others and develops schemas about how these relationships work. These working models are later used as guides for interpreting experiences and as ways of viewing the world. As Shaver and Tancredy (2001) explained, securely attached individuals can access attachment-related emotional memories without difficulty and discuss them coherently. In line with this pattern, they react emotionally to the loss of an important relationship partner but will not feel overwhelmed by grief. Securely attached individuals are therefore unlikely to suffer the complications associated with bereavement, unless they have undergone an unusually traumatic bereavement. Since they already confront and express their emotions to others, and are able to give coherent accounts of their experiences, they are less likely than insecurely attached individuals to be in need of, or to benefit from, grief intervention (Stroebe, Schut, & Stroebe, in press). Nevertheless, there are bereaved individuals who need help and who derive benefits from grief counseling and therapy. Most typically, these are individuals who have been unable to cope with their loss and for whom the grief reaction has in some way bgone wrongQ. The term complicated grief is used to refer to grief reactions which show a marked deviation from the normal pattern and which are associated with maladjustment and psychiatric problems (Prigerson & Jacobs, 2001; Stroebe et al., 2000). Grief counselors and therapists should focus their efforts on this subgroup of bereaved. Obviously, it would be important to identify these individuals early in the grieving process, before complications have developed. Risk factors for poor bereavement outcome are factors such as the traumatic nature of the death of the loved one, concurrent other stresses, previous psychiatric problems, insecure style of attachment, and initial high level of distress (Stroebe & Schut, 2001). Since there is evidence that initial high level of distress is one of the best predictors of the distress experience later (e.g., Stroebe & Stroebe, 1991), and since high levels of distress may also be a factor which motivates individuals to seek counseling or grief therapy, the need to seek out counseling may be another valid indicator that a person is at high risk for poor bereavement outcome. Although no single risk factor will ever allow us to predict grief complications with any degree of certainty, it would still increase the efficacy of grief counseling and therapy if potential interveners would reserve their services for those individuals who both request this type of help and show an accumulation of (risk-)factors that bereavement research has shown to be associated with an increased likelihood of grief complications (for a review, see Stroebe & Schut, 2001).

References Affleck, G., & Tennen, H. (1996). Construing benefits from adversity: Adaptational significance and dispositional underpinnings. Journal of Personality, 64, 899 – 922. Allumbaugh, D. L., & Hoyt, W. T. (1999). Effectiveness of grief therapy: A meta-analysis. Journal of Counseling Psychology, 46, 370 – 380. Bonanno, G. A., Keltner, D., Holen, A., & Horowitz, M. J. (1995). When avoiding unpleasant emotions might not be such a bad thing: Verbal-autonomic response dissociation and midlife conjugal bereavement. Journal of Personality and Social Psychology, 46, 975 – 989. Bonanno, G. A., Notarius, C. I., Gunzerath, L., Keltner, D., & Horowitz, M. J. (1998). Interpersonal ambivalence, perceived relationship adjustment, and conjugal loss. Journal of Consulting and Clinical Psychology, 66, 1012 – 1022.

W. Stroebe et al. / Clinical Psychology Review 25 (2005) 395–414

411

Bonannon, G. A., & Kaltman, D. (1999). Toward and integrative perspective on bereavement. Psychological Bulletin, 125, 760 – 776. Bower, J. E., Kemeny, M. E., Taylor, S. E., & Fahey, J. L. (2003). Finding positive meaning and its association with natural killer cell cytotoxicity among participants in a bereavement-related disclosure intervention. Annals of Behavioral Medicine, 25, 146 – 155. Bowlby, J. (1980). Attachment and Loss: Vol. 3. Loss: Sadness and Depression. New York7 Basic Books. Carr, D., House, J. S., Kessler, R. C., Nesse, R. M., Sonnega, J., & Wortman, C. B. (2000). Marital quality and psychological adjustment to widowhood among older adults: A longitudinal analysis. Journal of Gerontology: Social Sciences, 56B, S197 – S207. Cohen, S., Mermelstein, R., Kamarck, T., & Hoberman, H. M. (1985). Measuring the functional components of social support. In I. G. Sarason, & B. Sarason (Eds.), Social Support: Theory, Research and Applications (pp. 73 – 94). Dordrecht7 Martinus Nijhoff. Cohen, S., & Wills, T. A. (1985). Stress, social support, and the buffering hypothesis. Psychological Bulletin, 98, 310 – 357. Davis, C. G., Nolen-Hoeksema, S., & Larson, J. (1998). Making sense of loss and benefiting from the experience: Two construals of meaning. Journal of Personality and Social Psychology, 75, 561 – 574. Davis, C. G., Wortman, C. B., Lehman, D. R., & Silver, R. C. (2000). Searching for meaning in loss: Are clinical assumptions correct? Death Studies, 24, 497 – 540. Derogatis, L., & Melisaratos, N. (1983). The Brief Symptom Inventory: An introductory report. Psychological Medicine, 13, 595 – 605. Deutsch, H. (1937). Absence of grief. Psychoanalytic Quarterly, 6, 12 – 22. Festinger, L. (1957). A Theory of Cognitive Dissonance. Stanford, CA7 Stanford University Press. Fleming, S., & Robinson, P. (2001). Grief and cognitive-behavioral therapy: The reconstruction of meaning. In M. Stroebe, R. O. Hansson, W. Stroebe, & H. Schut (Eds.), Handbook of Bereavement Research: Consequences, Coping, and Care (pp. 647 – 669). Washington, DC7 American Psychological Association Press. Folkman, S. (2001). Revised coping theory and the process of bereavement. In M. Stroebe, R. O. Hansson, W. Stroebe, & H. Schut (Eds.), Handbook of Bereavement Research: Consequences, Coping, and Care (pp. 563 – 584). Washington, DC7 American Psychological Association Press. Frankl, V. (1959/1984). Man’s Search for Meaning: An Introduction to Logotherapy (1st/3rd ed.). New York7 Simon & Schuster. Freud, S. (1917). Mourning and melancholia. In J. Strachey (Ed. and Trans.). Standard Edition of the Complete Psychological Works of Sigmund Freud (pp.152–170). London: Hogarth Press. Greene, R. W., & Feld, S. (1989). Social support coverage and the well-being of elderly widows and married women. Journal of Family Issues, 10, 333 – 351. Hagman, G. (2001). Beyond decathexis: Toward a new psychoanalytic understanding and treatment of mourning. In R. A. Neimeyer (Ed.), Meaning Reconstruction and the Experience of Loss (pp. 13 – 32). Washington7 American Psychological Association. Harvey, J. H., Carlson, H. R., Huff, T. M., & Green, M. (2001). Embracing their memory: The construction of accounts of loss and hope. In R. A. Neimeyer (Ed.), Meaning Reconstruction and the Experience of Loss (pp. 231 – 244). Washington7 American Psychological Association. Hays, J., Kasl, S., & Jacobs, S. (1994). Past personal history of dysphoria, social support, and psychological distress following conjugal bereavement. Journal of the American Geriatrics Society, 42, 712 – 718. Janoff-Bulman, R. (1992). Shattered Assumptions: Towards a New Psychology of Trauma. New York7 Free Press. Jordan, J. R., & Neimeyer, R. A. (2003). Does grief counselling work? Death Studies, 27, 765 – 786. Kammer, D. (1983). Eine Untersuchung der psychometrischen Eigenschaften des deutschen Beck-Depressionsinventars (BDI). Diagnostica, 28, 48 – 60. Kato, P. M., & Mann, T. (1999). A synthesis of psychological interventions for the bereaved. Clinical Psychology Review, 19, 275 – 296. Kovac, S. H., & Range, L. M. (2000). Writing projects: Lessening undergraduates’ unique suicidal bereavement. Suicide and Life-Threatening Behavior, 30, 50 – 60. Krause, N. (1986). Social support, stress, and well-being among older adults. Gerontology, 41, 512 – 519. Lepore, S. J., Silver, R. C., Wortman, C. B., & Wayment, H. A. (1996). Social constraints, intrusive thoughts, and depressive symptoms among bereaved mothers. Journal of Personality and Social Psychology, 70, 271 – 282.

412

W. Stroebe et al. / Clinical Psychology Review 25 (2005) 395–414

Lindemann, E. (1979). Beyond Grief: Studies in Crisis Intervention. New York7 Aronson. Lopata, H. (1973). Widowhood in an American City. Morristown, N.J.7 General Learning Press. Lumley, M. A., Tojek, T. M., & Macklem, D. J. (2002). Effects of written emotional disclosure among repressive and alexithymic people. In S. J. Lepore, & J. M. Smyth (Eds.), The Writing Cure: How Expressive Writing Promotes Health and Emotional Well-Being (pp. 75 – 95). Washington, DC7 American Psychological Association Press. McIntosh, D. N., Silver, R. C., & Wortman, C. B. (1993). Religion’s role in adjustment to a negative life event: Coping with the loss of a child. Journal of Personality and Social Psychology, 65, 812 – 821. Meads, C., & Nouwen, A. (in press). Does emotional disclosure have any effects? A systematic review of the literature with meta-analyses. International Journal of Technology Assessment in Health Care. Middleton, W., Raphael, B., Martinek, N., & Misso, V. (1993). Pathological grief reactions. In M. Stroebe, W. Stroebe, & R. Hansson (Eds.), Handbook of Bereavement: Theory, Research, and Intervention (pp. 44 – 61). New York7 Cambridge University Press. Murphy, S. (1988). Mental distress and recovery in a high-risk bereavement sample three years after untimely death. Nursing Research, 37, 30 – 35. Murphy, S., Chung, I., & Johnson, L. (2002). Patterns of mental distress following the violent death of a child and predictors of change over time. Research in Nursing and Health, 25, 425 – 437. Neimeyer, R. A. (Ed.). (2001). Meaning Reconstruction and the Experience of Loss. Washington7 American Psychological Association Press. Nolen-Hoeksema, S. (1991). Responses to depression and their effects on the duration of depressive episodes. Journal of Abnormal Psychology, 100, 282 – 569. Nolen-Hoeksema, S., & Morrow, J. (1991). A prospective study of depression and posttraumatic stress symptoms after a natural disaster: The 1989 Loma Prieta earthquake. Journal of Personality and Social Psychology, 61, 115 – 121. Nolen-Hoeksema, S., & Davis, C. D. (1999). Thanks for sharing that: Ruminators and their social support networks. Journal of Personality and Social Psychology, 77, 801 – 814. Norris, F. H., & Murrell, S. A. (1990). Social support, life events, and stress and modifiers of adjustment to bereavement by older adults. Psychology and Aging, 5, 429 – 436. Okabayashi, H., Sugisawa, H., Yatomi, N., Nakatani, Y., Takanashi, K., Fukaya, T., et al. (1997). [Title of the article is missing]. Japanese Journal of Psychology, 68, 147 – 154. Parkes, C. M. (1972/1996). Bereavement: Studies of Grief in Adult Life (1st/3rd ed.). London7 Routledge. Parkes, C. M. (1975). Psychosocial transition: Comparison between reactions to loss of a limb and loss of a spouse. British Journal of Psychiatry, 127, 204 – 210. Parkes, C. M. (1998). Editorial. Bereavement Care, 17, 18. Parkes, C. M. (2001). A historical overview of the scientific study of bereavement. In M. Stroebe, R. O. Hansson, W. Stroebe, & H. Schut (Eds.), Handbook of Bereavement Research: Consequences, Coping, and Care (pp. 25 – 45). Washington, DC7 American Psychological Association Press. Parkes, C. M., & Weiss, R. (1983/1995). Recovery from Bereavement (1st/2nd ed.). Northvale, NJ7 Aronson. Pennebaker, J. W., & Beall, S. K. (1986). Confronting a traumatic event: Toward an understanding of inhibition and disease. Journal of Abnormal Psychology, 95, 274 – 281. Pennebaker, J. W., Colder, M., & Sharp, L. K. (1990). Accelerating the coping process. Journal of Personality and Social Psychology, 58, 528 – 537. Pennebaker, J. W., & Keough, K. A. (1999). Revealing, organizing, and reorganizing the self in response to stress an emotion. In R. J. Contrada, & R. D. Ashmore (Eds.), Self, Social Identity, and Physical Health: Interdisciplinary Explorations (pp. 101 – 121). Oxford7 Oxford University Press. Pennebaker, J. W., Kiecolt-Glaser, J. K., & Glaser, R. (1988). Disclosure of traumas and immune function: Health implications for psychotherapy. Journal of Consulting and Clinical Psychology, 56, 239 – 245. Pennebaker, J. W., & O’Heeron, R. C. (1984). Confiding in others and illness rates among spouses of suicide and accidental death. Journal of Abnormal Psychology, 93, 473 – 476. Pennebaker, J. W., Zech, E., & Rime´, B. (2001). Disclosing and sharing emotion: Psychological, social, and health consequences. In M. Stroebe, R. O. Hansson, W. Stroebe, & H. Schut (Eds.), Handbook of Bereavement Research: Consequences, Coping, and Care (pp. 517 – 543). Washington, DC7 American Psychological Association Press. Prigerson, H., & Jacobs, S. C. (2001). Traumatic grief as a distinct disorder: A rationale, consensus criteria, and a preliminary empirical test. In M. Stroebe, R. O. Hansson, W. Stroebe, & H. Schut (Eds.), Handbook of Bereavement

W. Stroebe et al. / Clinical Psychology Review 25 (2005) 395–414

413

Research: Consequences, Coping, and Care (pp. 613 – 645). Washington, DC7 American Psychological Association Press. Radloff, L. (1977). The CES-D Scale: A self-report depression scale for research in the general population. Applied Psychological Measurement, 1, 385 – 401. Range, L. M., Kovac, S. H., & Marion, M. S. (2000). Does writing about the bereavement lessen grief following sudden, unintentional death? Death Studies, 24, 115 – 134. Raphael, B., & Nunn, K. (1988). Counseling the bereaved. Journal of Social Issues, 44, 191 – 206. Shuchter, & Zisook (1993). The course of normal grief. In M. Stroebe, W. Stroebe, & R. Hansson (Eds.), Handbook of bereavement: Theory, research, and intervention (pp. 23 – 43). New York7 Cambridge University Press. Schut, H., Stroebe, M., van den Bout, J., & Terheggen, M. (2001). The efficacy of bereavement interventions: Determining who benefits. In M. Stroebe, R. O. Hansson, W. Stroebe, & H. Schut (Eds.), Handbook of Bereavement Research: Consequences, Coping, and Care (pp. 705 – 737). Washington, DC7 American Psychological Association Press. Schut, H. A. W. (1992). Omgaan Met De Dood Van De Partner: Effecten Op Gezondheid; Effecten Van Rouwbegleiding [Coping with Conjugal Bereavement: Effects on Psychological Functioning and Effects of Grief Counselling]. Amsterdam7 Thesis Publishers. Schwarzer, C. (1992). Bereavement, received social support, and anxiety in the elderly: A longitudinal analysis. Anxiety Research, 4, 287 – 298. Segal, D. L., Bogaards, J. A., Becker, L. A., & Chatman, C. (1999). Effects of emotional expression on adjustment to spousal loss among older adults. Journal of Mental Health and Aging, 5, 297 – 310. Segal, D. L., Chapman, C., Bogaards, J. A., & Becker, L. A. (2001). One year follow-up of an emotional expression intervention for bereaved older adults. Journal of Mental Health and Aging, 7, 465 – 472. Shaver, P. R., & Tancredy, C. M. (2001). Emotion, attachment, and bereavement: A conceptual commentary. In M. Stroebe, R. O. Hansson, W. Stroebe, & H. A. W. Schut (Eds.), Handbook of Bereavement Research: Consequences, Coping and Care (pp. 63 – 88). Washington DC7 American Psychological Association Press. Silver, R., & Wortman, C. (1980). Coping with undesirable life events. In J. Garber, & M. E. P. Seligman (Eds.), Human Helplessness: Theory and Applications (pp. 279 – 340). New York7 Academic Press. Smyth, J. M. (1998). Written emotional expression: Effect sizes, outcome types, and moderating variables. Journal of Consulting and Clinical Psychology, 66, 174 – 184. Stroebe, M., Hansson, R. O., Stroebe, W., & Schut, H. (2001). Handbook of Bereavement Research: Consequences, Coping and Care. Washington DC7 American Psychological Association Press. Stroebe, M., & Schut, H. (1999). The Dual Process Model of coping with bereavement: Rationale and description. Death Studies, 23, 197 – 224. Stroebe, M., Schut, H., & Stroebe, W. (in press). Attachment in coping with bereavement: A theoretical integration. Review of General Psychology. Stroebe, M., & Stroebe, W. (1991). Does grief work work? Journal of Consulting and Clinical Psychology, 59, 479 – 482. Stroebe, M., Stroebe, W., Schut, H. A. W., Zech, E., & van den Bout, J. (2002). Does disclosure of emotions facilitate recovery from bereavement? Evidence from two prospective studies. Journal of Consulting and Clinical Psychology, 70, 169 – 178. Stroebe, M., van Son, M., Stroebe, W., Kleber, R., Schut, H., & van den Bout, J. (2000). On the classification and diagnosis of pathological grief. Clinical Psychology Review, 20, 57 – 75. Stroebe, W., & Schut, H. (2001). Risk factors in bereavement outcome: A methodological and empirical review. In M. Stroebe, R. O. Hansson, W. Stroebe, & H. Schut (Eds.), Handbook of Bereavement Research: Consequences, Coping and Care (pp. 349 – 372). Washington DC7 American Psychological Association Press. Stroebe, W., & Stroebe, M. (1987). Bereavement and Health: The Psychological and Physical Consequences of Partner Loss. New York7 Cambridge University Press. Stroebe, W., & Stroebe, M. (1993). Determinants of adjustment to bereavement in younger widows and widowers. In M. Stroebe, R. O. Hansson, W. Stroebe, & H. Schut (Eds.), Handbook of bereavement: Theory, research, and intervention. New York7 Cambridge University Press. Stroebe, W., & Stroebe, M. (1996). The social psychology of social support. In E. T. Higgins, & A. W. Kruglanski (Eds.), Social Psychology: Handbook of Basic Principles (pp. 597 – 621). New York7 Guilford. Stroebe, W., Stroebe, M., Abakoumkin, G., & Schut, H. (1996). The role of loneliness and social support in adjustment to loss: A test of attachment versus stress theory. Journal of Personality and Social Psychology, 70, 1241 – 1249.

414

W. Stroebe et al. / Clinical Psychology Review 25 (2005) 395–414

Stroebe, W., Zech, E., Stroebe, M., & Abakoumkin, G. (in press). Does social support help in bereavement? Its impact on vulnerability and recovery. Journal of Social and Clinical Psychology. Stylianos, S. K., & Vachon, M. L. S. (1993). The role of social support in bereavement. In M. Stroebe, W. Stroebe, & R. O. Hansson (Eds.), Handbook of Bereavement: Theory, Research, and Intervention (pp. 397–410). Taylor, S. E. (1983). Adjusting to threatening events: A theory of cognitive adaptation. American Psychologist, 38, 1161 – 1173. Taylor, S. E. (1989). Positive Illusions: Creative Self-Deception and the Healthy Mind. New York7 Basic Books. Tedeschi, R. G., & Calhoun, L. G. (1996). Trauma and Transformation: Growing in the Aftermath of Suffering. Thousand Oaks, CA7 Sage. Volkan, V. (1981). Linking Objects and Linking Phenomena. New York7 International Universities Press. Weiner, B. (1995). Judgments of Responsibility. New York7 Guilford. Weiss, R. S. (1975). Loneliness: The Experience of Emotional and Social Isolation. Cambridge, MA7 MIT Press. Worden, W. (1991). Grief Counseling and Grief Therapy: A Handbook for the Mental Health Practitioner (2nd ed.). New York7 Springer. Wortman, C., & Silver, R. (1989). The myths of coping with loss. Journal of Consulting and Clinical Psychology, 57, 349 – 357.