Good grief

The loss of a loved one is well recognised as a harrowing and stressful life event. Bereavement is something that most people experience at some time in their lives and, for a minority, mental and physical health complications do occur. Our case raises a fundamental question: How does one best cope with such a loss?

For a long time, following the work of Sigmund Freud, it was thought that one had to 'do one's grief work' to come to terms with loss. One had to confront the experience of bereavement. Eric Lindemann vividly illustrated how 'working through' might be adaptive:

'This grief work has to do with the effort of reliving and working through in small quantities events which involved the now-deceased person and the survivor… Each item of this shared role has to be thought through, pained through if you want, and gradually the question is raised, how can I do that with somebody else? And gradually the collection of activities … can be torn asunder to be put to other people.' (Lindemann, 1979, p.234)

Also in the 1970s, Colin Murray Parkes identified more specific features of grief work, namely: preoccupation with thoughts of the lost person; painful repetitious recollection of the loss experience; and an attempt to make sense of the loss, to fit it into one's assumptions about the world, or to modify those assumptions if need be.

Grief work was understood, then, to be a confrontational process leading to good health outcomes. It was only toward the end of the last century that doubts were raised about the adequacy of this concept as an explanation of adaptive coping. Paul Rosenblatt was one of the first to take a critical look at such grief theories. Shortly thereafter, Camille Wortman and Roxanne Silver even coined the phrase 'myths of coping with loss', drawing attention to additional shortcomings of the grief work notion.

Our own early reservations about the grief work concept are perhaps encapsulated in a meeting at an international conference on bereavement in London in 1988 with John Bowlby, renowned attachment theorist and author of the trilogy Attachment and Loss. Being asked the difference between grief work (an integral concept in his attachment theory) and rumination, Bowlby answered, 'I don't know.' If Bowlby did not know the answer, how can we be expected to? Such concerns led to our analyses of grief work across the following years. In our view, theoretical shortcomings had to do with both the concept and how it was put into practice. These included the confounding of grief work with negative affect; the lack of understanding as to why grief work is effective, or what underlying mechanisms could account for its efficacy; failure to consider potential benefits of suppression and denial; and unexplored links between grief work and continued or relinquished ties to the deceased. These problems were compounded by lack of empirical evidence: a range of studies, including our own, failed to provide unequivocal support for the idea that working through grief furthered adjustment to bereavement. Indeed, investigations in other cultures suggested that different 'prescriptions' for grieving (such as control over expressive grieving; suppression of memories) may also be effective.

It's time to review developments in knowledge and report on current understanding of the difference between grief work and rumination, in order to shed light on the nature of (mal)adaptive coping with bereavement. This is not only of theoretical interest – for example, to establish which underlying regulatory processes account for 'getting stuck' with grief complications – it might help our 46-year-old male, too.

A closer look at rumination

Progress in understanding the difference between grief work and rumination came first through pioneering research conducted by Susan Nolen-Hoeksema within the framework of her response styles theory (RST). Originally, RST was proposed to explain how rumination may prolong depression. However, in due course Nolen-Hoeksema's work also marked a new direction in research on complicated grief (CG). Previously, the focus had been on examining how denial and suppression caused ill effects in bereavement; Susan's work changed to a focus on the 'opposite form of coping', namely, rumination, which was shown to be closely related to CG. She referred to ruminative coping as 'the polar opposite of avoidance and denial' (Nolen-Hoeksema, 1999, p.21). It is noteworthy, though, that Nolen-Hoeksema was looking to establish the causal role of rumination in poor adaptation, rather than to ascertain whether it was a confrontational or avoidant coping strategy.

The distinction of rumination from grief work remained unclear. Three points make this evident. First, there is conceptual similarity with grief work. Nolen-Hoeksema (2001) characterised rumination as persistent and repetitive, chronic and passive focus on the occurrence, causes and consequences of negative (grief-related) emotions and symptoms. There is a slightly narrower focus on confronting one's own emotions and going over related problems in rumination, while grief work encompasses a wider range of death-related concerns. The former seems closer to 'mulling over' while the latter encompasses 'meaning-making' (but then again, dwelling on a negative meaning seems close to rumination). Nevertheless, the differences hardly seem enough to explain why rumination should be systematically related to poor (bereavement) outcomes, while grief work has been postulated as adaptive.

Furthermore, there is lack of differentiation regarding underlying mechanisms. Nolen-Hoeksema described the mechanisms through which rumination could amplify and maintain distress as follows:

• Ruminating enhances effects of distressed mood on thinking, drawing attention to negative thoughts and memories.
• It may interfere with good problem-solving.
• It impairs instrumental behaviors by reducing motivation to engage in mood lifting-activities
• Social support may be reduced because ruminations violate social norms for coping.

Again, these appear similar to the grief work features described above, as identified by Parkes: similar mechanisms seem implicated in the preoccupation with one's loss and self involvement while 'doing one's grief work'.

Finally, importantly, both grief work and rumination were described as confrontational processes. So we're left without a definitive answer to the question: 'What is the critical difference between grief work and rumination?'

Unravelling the distinction

Clues as to how to understand the difference between grief work and rumination emerged from the related fields of research on generalised anxiety disorder (GAD) and post-traumatic stress disorder (PTSD). According to Thomas Borkovec and colleagues, worry is associated with a variety of disadvantages and debilities, including GAD and – like rumination – depression. Yet a difference was postulated with respect to underlying processes. Unlike the claims for rumination: '…worry partly functions as a cognitive avoidance response to threatening stimuli… Worry distinctively involves a predominance of verbal thought whose function appears to be the cognitive avoidance of threat' (Borkovec et al., 1998, p.573). Worry is conceptually similar to rumination (the former being more future-oriented than the latter). Could it be that rumination in bereavement is actually an avoidant process too? Further indications came from cognitive behavioral accounts of PTSD. For example, in 1995 Anke Ehlers and Regina Steil suggested that rumination may be one of the major strategies of cognitive avoidance in chronic PTSD, and rumination was found in several subsequent studies to be a strong predictor of persistent PTSD. These understandings seemed in stark contrast to Nolen-Hoeksema's view that ruminative coping was the polar opposite of avoidance.

It was still a few years before similar ideas to those in the GAD and PTSD areas entered the bereavement field. A breakthrough came in 2006 when Paul Boelen and his colleagues linked avoidance to complicated grief (CG). Their theory laid the foundations for a new direction of research, moving toward clarification of the grief work/rumination puzzle. While their and other researchers' focus was on avoidance and CG, shortly thereafter, in 2007, we brought together the ideas in a review article entitled 'Ruminative coping as avoidance: A reinterpretation of its function in adjustment to bereavement'. We argued that rumination is not an opposite form of coping from suppression or denial, but that it is a similar phenomenon to these processes, and very different from the types of confrontation that take place in grief work.

This reasoning laid the foundations for what we later defined as the rumination as avoidance hypothesis (RAH): Persons who ruminate go over and over events related to their loss, to avoid that which is simply too painful to confront, namely, the reality of loss. This avoidance of facing the reality of the death is the essential component of the non-adaptive, unhealthy ruminative processing. This line of argument is in accordance with an observed cognitive phenomenon among bereaved persons struggling to come to terms with their loss: to avoid acceptance of the reality of loss, the person may engage in counterfactual thinking; that is, they may generate imagined alternatives to actual events. 'If only I had made her stop smoking, then she'd still be alive.' Such ruminative thoughts act as distraction from more emotionally laden topics. There are also non-ruminative avoidance (suppression, denial, etc.) processes, which can be related to poor mental health following bereavement. Nevertheless, in our view, confrontation-avoidance processing, associated as it is with emotion regulation, is fundamentally important in adaptive versus maladaptive grieving.

Evidence for the distinction

It is remarkable that, although clinical wisdom claims it as the essence of grieving and theoretical arguments were given across decades, support for the efficacy of grief work is still based on very shaky foundations: there is hardly a single study, let alone a body of evidence, showing that working through grief enhances adaptation. Future empirical studies still need to test the grief work hypothesis, for example, to show that confrontational strategies – such as facing up to the reality and pain of loss, problem-solving, engaging in constructive instrumental behaviours and activities, making plans of action to solve difficulties, perhaps even finding meaning in the loss – lead to adaptation over time. This should not be too difficult given scientific developments (e.g. in methodological and statistical techniques; use of carefully designed, laboratory and intervention efficacy studies), as shown in the rumination area.

The domain of rumination in bereavement has provided a comparatively extensive body of relevant research, moving well beyond the state of knowledge at the time of the our 2007 review article. Initially, though, the focus was mainly on the connections between ruminative thought and health consequences. For example, the extensive work of Edward Watkins and colleagues linked ruminative thought with CG, and identified types of repetitive thought that yield more adverse versus more facilitating outcomes. More recently, examination of underlying ruminative mechanisms has taken place. Empirical studies have begun to examine rumination as an avoidant process, also providing finer-grained analyses of maladaptive ruminative processes. We continued along such lines of research, designing a programme specifically to examine RAH; we focus on this here, while recognising the related work of others.

To put the RAH ideas to adequate empirical test, several major steps and years of investigation were necessary. Our colleague Maarten Eisma developed the project, culminating in his 2015 PhD thesis containing several related lines of research.  The first essential step was to develop an assessment instrument to measure rumination in bereavement. Only with a valid, bereavement-specific instrument could we go on, for example, to distinguish high from low ruminators and test our hypothesis that the former are more avoidant of bereavement-related threats. We produced a new bereavement-specific scale, the Utrecht Grief Rumination Scale. The UGRS included subscales covering rumination about the meaning of the loss, social relationships, injustice, feelings and counterfactual thinking. We then conducted studies to validate the instrument and examine the predictive validity of its subscales. With this measure available, we could then investigate working mechanisms underlying rumination.

In a series of investigations undertaken with the cooperation of different groups of bereaved participants, we examined the relationship between ruminative coping, psychopathology (particularly CG/depression) and specific avoidance processes. We first conducted a longitudinal questionnaire study, followed by two laboratory investigations using implicit measures: an eye-tracking and an approach-avoidance task (AAT). Such tasks are by now quite standard measures of automatic behaviour tendencies. In the eye-tracking study, high ruminators showed conscious attentional avoidance of stimuli that represented the loss, while showing attentional preference for more benign negative stimuli. In the AAT, bereaved persons could use a joystick to pull stimuli closer or push them away. Our study showed that, when presented with personal loss stimuli, participants who ruminated more were faster in pushing these away and slower in pulling them closer, implying that more rumination was associated with stronger implicit loss avoidance.

Taken together, the evidence from these various studies supported RAH: Rumination during bereavement was demonstrated to increase and prolong symptomatology and this negative relationship, providing preliminary evidence for an association between rumination and loss-threat avoidance. Keen to demonstrate the potential applied value of our work, we then examined the impact of therapist-guided intervention for rumination, comparing two conditions: internet-delivered exposure and behavioral activation for CG and rumination, and a non-intervention control group. Findings supported the efficacy of the exposure intervention in reducing levels of CG and rumination at post-test and follow-up. Of course, we remain cautious about causal mechanisms and the impact of other variables. For example, rumination and avoidance behaviour have both been characterised as passive coping styles, so people who have a trait-tendency to engage in passive coping may be likely to engage more in both rumination and avoidance.

Conclusions

So, is rumination a confrontation or avoidance process? Maarten Eisma concluded that a compelling case can be made for the theory that rumination serves to avoid painful aspects of the loss and the negative emotions linked with it. Our research could be said to corroborate RAH and disconfirm the RST notion that rumination is (only) a maladaptive confrontation process.

We now believe the difference between grief work and rumination is that the former reflects a confrontational, the latter an avoidant strategy in adjustment to the loss of a loved person. The function of grief work is to gradually face up to the fact of loss of an attachment figure, accept the reality of death, and work toward finding a place for the deceased in ongoing life. By contrast, while people confront certain aspects in their ruminations about their loss, we showed that they avoid more personally threatening aspects. We think that that is why they ruminate about the (to them) lesser-threatening concerns.

It needs to be kept in mind that both grief work and rumination are part and parcel of normal grieving, a tandem of strategies reflecting complex emotion-regulation processing that occurs during bereavement. Yet, so far, grief work and rumination have been examined independently of each other. They could usefully be integrated in future research, for example, to explore their possibly interrelated functions in coming to terms with loss. Furthermore, it is important to remember that the functions of both grief work and rumination remain complex. For example, on the one hand, grief work may not always be adaptive; it cannot be done all the time (it is exhausting). On the other hand, only in extreme or unrelenting cases is ruminative coping associated with maladaptation (mulling things over 'familiarises' one with the fact of loss). So to some extent, avoidant rumination serves adaptive functions (e.g. when the reality is too hard to bear early on; or all the time).

We return once more to our 46-year-old male: What should the practitioner do, what advice can be derived from this line of research? Our results suggest that this client's persistence of intense, relentless ruminative thought over many months is likely to be a central complicating factor in his particular case: a valid indicator of maladaptive grieving (more so than it would have been during the first few months of bereavement). So application of the types of therapeutic techniques mentioned earlier could in this case be considered appropriate and may prove effective (e.g. by integrating exposure to threatening stimuli). However, we emphasise again: at this stage of research we need to be cautious in 'translating' our findings into daily practice. To mention just two reasons: complications in grieving may be unrelated to maladaptive ruminative coping – one size does not fit all; and the therapy situation itself encompasses much more, it is always more complicated than straightforward application of the small number of variables considered in research.

Despite these cautions about limited application, in our view, distinguishing healthy grief work from unhealthy rumination is a critical step forward in theoretical understanding, one that, as further knowledge accumulates, should also better inform practice. Bowlby and other earlier theorists were correct to identify the necessity of grief work in coming to terms with the loss of a loved one. Now we can better draw the fine line between this adaptive process and maladaptive coping. We can further explore the RAH idea that, in cases where complications are rumination-related, a bereaved client may need to confront rather than (only) to distract from their own personal, loss-related threats.

- Margaret Stroebe is in the Department of Clinical Psychology, Utrecht University; and the Department of Clinical Psychology & Experimental Psychopathology, University of Groningen. [email protected]

- Henk Schut is in the Department of Clinical Psychology, Utrecht University

Key sources

Boelen, P.A., van den Hout, M.A., van den Bout, J. (2006). A cognitive-behavioral conceptualization of complicated grief. Clinical Psychology: Science and Practice, 13, 109–128.
Borkovec, T., Ray, W. & Stöber, J. (1998). Worry: A cognitive phenomenon intimately linked to affective, physiological, and interpersonal behavior processes. Cognitive Therapy Research, 22, 561–576.
Ehlers, A. & Steil, R. (1995). Maintenance of intrusive memories in posttraumatic stress disorder: A cognitive approach. Behavior and Cognitive Psychotherapy, 23, 217–249.
Eisma, M. (2015). Rumination following bereavement: Assessment, working mechanisms and intervention., Ridderkerk, The Netherlands: Ridderprint BV.
Lindemann, E. (1979). Beyond grief: Studies in crisis intervention. New York, Aronson.
Nolen-Hoeksema, S., Wisco, B.E. & Lyubomirsky, S. (2008). Rethinking rumination. Perspectives on Psychological Science, 3, 400–424.
Parkes, C.M. (1972). Bereavement: Studies of grief in adult life. Harmondsworth: Penguin.
Stroebe, M.S., Boelen, P., van der Hout, M. et al. (2007). Ruminative coping as avoidance. European Archives of Psychiatry and Clinical Neuroscience, 257, 462–472.
Watkins, E.R. & Moulds, M.L. (2013). Repetitive thought: Rumination in complicated grief. In M.S. Stroebe, H. Schut & J. van den Bout (Eds.) Complicated grief: Scientific foundations for health care professionals (pp.163–175). New York: Routledge.
Wortman, C. & Silver, R. (1989). The myths of coping with loss. Journal of Consulting and Clinical Psychology, 57, 349–357.