Coping Flexibility and Complicated Grief: A Comparison of American and Chinese Samples

Abstract

Background

The ability to process a death, and the ability to remain optimistic and look beyond the loss, are both thought to be effective means of coping with loss and other aversive events. Recently, these seemingly contrary dimensions have been integrated into the idea of coping flexibility.

Method

In this study we assessed the ability of married and bereaved individuals in the US and Hong Kong to use both coping approaches as operationalized by the trauma-focused and forward-focused coping scales of a previously validated questionnaire. We also calculated a single flexibility score.

Results

Bereaved participants reported greater trauma-focused coping ability than did married participants. However, bereaved participants meeting criteria for complicated grief (CG) reported less forward-focused coping than both asymptomatic bereaved and married participants. The CG group also showed less overall coping flexibility than the asymptomatic bereaved and married groups. Country was not a factor.

Conclusions

Findings suggest that deficits in coping flexibility are indicative of pathology in bereaved individuals, and that this relationship extends across cultures. Limitations of the study and directions for future research are discussed.

Theories about how people cope with potentially traumatic life events (PTE) emphasize two seemingly contradictory approaches. The dominant perspective historically has viewed loss and potential trauma as requiring an effortful focus of the thoughts, images and memories associated with the event.¹ More recently, however, a growing body of research has pointed to the salutary importance of behaviors that appear to minimize the focus on the stress event, such as optimism² distraction or emotional avoidance.³ Even more recently, a third perspective has emerged that emphasizes both approaches⁴ and integrates these streams in the idea of coping flexibility. ^5–7 According to this perspective, focusing on the PTE or focusing beyond the event may be adaptive coping strategies at different points in time or under different circumstances. However, the exclusive reliance on one of these coping strategies over the other will not be as effective overall as will be the ability to flexibly engage in either type of coping response as dictated by the changing demands of the situation. Using a recently developed questionnaire measure,⁸ the current investigation examined levels of coping flexibility among bereaved and married adults in both the United States and Hong Kong.

Coping Flexibility

Initial research on coping flexibility documented the effectiveness of employing multiple coping and emotion regulation behaviors among undergraduate samples. ^6,7,9,10 More recently, an experimental measurement of flexibility has proved effective in distinguishing pathological and non-pathological responses to aversive life events, including the death of a spouse.¹¹ In practice, however, the experimental assessment of coping flexibility among samples exposed to loss or potential trauma is impractical. Accordingly, Bonanno and colleagues developed a questionnaire measure of trauma-related coping flexibility, the Perceived Ability to Cope with Trauma scale (PACT).⁸ Consistent with the hypothesized salubrious nature of flexibility, both focusing on trauma and focusing forward beyond the trauma were found to independently contribute to overall adjustment, especially when trauma exposure was high, and this was also true of the combination of these abilities into a single coping flexibility index.⁸ In the current investigation we sought to extend this research by comparing coping flexibility on the PACT among matched samples of bereaved individuals with complicated grief (CG), bereaved individuals without complicated grief (asymptomatic bereaved), and non-bereaved (married) individuals.

Coping Flexibility and Bereavement

Historically, the successful adjustment to the pain of bereavement had been linked almost exclusively with the intense processing of personal meanings and emotions associated with the loss.^12,13 Freud in fact referred to this process as the “work” of mourning.¹² By contrast, bereaved individuals who avoided or failed to evidence grief work were assumed to be pathological and likely to suffer delayed grief syndromes. Interestingly, recent reviews of the bereavement literature revealed little empirical evidence to support the idea that grief work as essential to healthy coping or that the avoidance of negative emotions is necessarily pathological.^4,14,15 On the contrary, emerging research has explored some of the potential benefits of focusing attention away from the loss, and evidence suggests that avoidant responses to grief are associated with favorable rather than unfavorable outcomes. ^3,16 Beyond bereavement, avoidant coping has been associated with both adaptive^5,17,18 and maladaptive responses to PTEs. ^19,20

The concept of flexibility accommodates these seemingly contradictory findings by viewing avoidance as a subset of coping behaviors that might comprise a more elaborate coping repertoire. Research on context sensitivity suggests for example that the experience and expression of emotion is associated with a better or worse grief course, depending upon the context in which the emotion occurs.²¹ Similarly, research on non-bereaved samples has indicated that both emotional suppression and emotional expression can be adaptive but also that the contribution of these separate regulatory responses is superseded by the overall ability to flexibly use either response in accord with situational demands. ^6,22

The contrasting orientation of focusing on the impact of a loss and on focusing beyond the loss is captured in Stroebe and Schut’s dual-process model of bereavement.⁴. Consistent with the general notion of flexibility, the dual-process model distinguishes loss-oriented processes, such as processing of the loss itself, and restoration-oriented processes, such as focusing on secondary stressors that are also consequences of bereavement. An important distinction from the flexibility concept, however, is that the dual-process model assumes that a grieving person will necessarily oscillate between these two orientations and that both loss-oriented and restoration-oriented processes are essential for successful adaptation to loss. In contrast, the broader notion of coping flexibility posits that both types of processes are of potential importance, and that whether either or both processes are necessary depends more explicitly upon contextual demands. In the current study we measured the perceived ability to use each of these two coping orientations in groups that differed in their exposure to loss.

Cultural Perspectives on Grief

While a variety of culturally defined mourning reactions have been observed by researchers from diverse disciplines, ^17,23–25 relatively little empirical research on the potential influence of cultural factors on the development of CG reactions exists. Some attention has been drawn to the cultural difference found among the Chinese, particularly in their structured mourning rituals and relationship with the deceased. ^17,26 However, the existing literature is inconclusive as to the possible role played by trauma-focused and forward-focused coping in Chinese samples, or whether these abilities might hold the same or different patterns of association with adjustment among Chinese bereaved. In the current study we compared responses to these measures in three different bereavement status groups (bereaved with CG, asymptomatic bereaved, and married individuals) obtained from samples collected in two culturally distinct regions (the United States and Hong Kong).

Methods

Participants

Conjugally bereaved individuals in both samples were recruited through advertisements, referrals and mailings. Inclusion criteria specified the death of a spouse 1.5–3.0 years previously and absence of Axis I psychopathology prior to the death-event. Married individuals were recruited through fliers and Internet advertisements. No more than one individual from a married couple was enrolled in the study. Within each country, there were no significant differences across bereavement status groups for any of the demographic variables (see Table 1). Participants from both countries signed consent forms approved by respective institutional review boards.

Table 1.

Demographic Characteristics

American Sample	Married (n = 37)	Asymptomatic Bereaved (n = 35)	Complicated Grief (n = 23)	Test Statistic
Demographics
Age	45.97 (6.74)	49.62 (10.21)	47.86 (8.55)	F(2,93) = 1.63
Female	24	27	14	χ2(2, N = 95) = 2.06
White	19	23	13	χ2(2, N = 95) = 1.55
Family Income	$84,486 (47,857)	$104,984 (47,857)	$53,954 (45,926)	F(2, 93) = 1.46
Years Married	15.79 (10.16)	16.24 (10.43)	17.80 (10.55)	F(2, 93) = .28
GAF Score	78.00	76.09	62.65	F(2,93) = 16.03**
Hong Kong Sample	Married (n = 45)	Asymptomatic Bereaved (n = 55)	Complicated Grief (n = 16)	Test Statistic
Demographics
Age	45.87 (7.57)	44.95 (7.33)	47.63 (7.62)	F(2,116) = .82
Female	35	48	14	χ2(2, N=116) = 1.83
Family Income	HK$33,145 (35,619)	HK$17,338 (17,659)	HK$22,528 (27,567)	F(2, 113) = 4.11*
Years Married	19.34 (8.99)	16.62 (8.76)	17.00 (8.76)	F(2, 113) = 1.20

^*indicates p < .05

^**indicates p < .001

Defining chronic grief

Using a Structured Clinical Interview for DSM-IV-TR, all participants were asked questions regarding symptoms of Major Depressive Disorder (MDD).²⁷ Bereaved participants were questioned about symptoms of CG. Symptom criteria specified frequency and intensity (e.g., “significant difficulties more days than not”):.^1,28–30 strong yearning for the deceased; recurrent and intrusive recollections of the death event; intense distress over symbolic reminders of the loss; preoccupation with thoughts about the loss; recurrent regrets or self-blame about behavior toward the deceased; difficulty accepting the finality of the loss; marked loneliness; pervasive sense that life is meaningless; unusual difficulty developing new relationships; efforts to avoid thoughts, feelings, or conversations associated with the loss; and efforts to avoid activities, places, or people that arouse recollections of the loss (11 items, American sample, α = .84; Chinese sample, α = .83).

Bereaved participants were categorized in the CG group (American sample, n = 23; Chinese sample, n = 16) if they had at least 3 of 8 grief symptoms, and at least one of 3 separation distress symptoms.³¹ Consistent with previous research,³² participants in this group had significantly lower levels of functioning than married or asymptomatic bereaved (Table 1).

Coping Flexiblity

The 20-item PACT scale asks participants to endorse their ability to use different coping strategies on a 7-point scale (1=not true, 7=extremely true).⁸ Previous research and factor analysis indicated two subscales: Forward-focus (12 items, US, α = .83; HK, α = .92) assesses coping abilities related to thinking optimistically, attending to the needs of others, maintaining plans and goals, remaining calm, reducing painful emotion, and being able to laugh; Trauma-focus (8 items, US, α = .73; HK, α = .67ⁱ) gauges the ability to remain focused on and fully experience the emotional and cognitive significance of a PTE, temporarily withdraw from social interactions, revise goals and plans, and think realistically. A single coping flexiblity score is computed using the averages of the two coping scales in a Negative Acceleration Model ^33,34 (Priester & Petty, 1996; Scott, 1966), F = [(2S + 1)/(S + L +2)], where S is the smaller mean of 2 means and L is the larger (e.g. if a participant has a mean forward focus score of 5 and trauma focus score of 7, then F = 0.79). This method produces a single score, ranging from 0 to 1, where larger scores indicate relatively equal and greater use of both abilities.

Data Analysis

We initially conducted a repeated-measures ANOVA for the within-subjects PACT scales (Forward-Focused, Trauma-Focused) and between-subject variables of bereavement status (married, asymptomatic bereaved, CG) and country (US, HK). Simple effects were examined to determine directionality and magnitude of observed differences. We then repeated these analyses controlling for depression, using an alternative diagnostic criteria for CG, and using the summary coping flexibility score in place of the two coping ability scales.

Results

Coping Ability Scales

The initial repeated-measures ANOVA revealed main effects for coping ability type, F(1, 205) = 9.31, p < .005, Eta² = .043, country, F(1, 205) = 11.62, p < .005, Eta² = .054, and bereavement status, F(2, 205) = 3.62, p < .05, Eta² = .034. These effects were qualified by a significant interaction between coping ability type and bereavement status, F(2, 205) = 11.40, p < .001, Eta² = .100. The three-way interaction did not approach significance, F(2, 205) = .79, p = .456, Eta² = .008.

Follow-up simple effects for coping ability type indicated significant group differences in forward-focus F(2,210) = 6.54, p < .005, and trauma-focus, F(2,210) = 3.17, p < .05. As shown in Figure 1, the asymptomatic bereaved and CG participants reported greater ability to engage in trauma-focused coping than did the married group. However, the CG group reported less forward-focus ability than both the asymptomatic bereaved and married groups. Simple effects for the bereavement status groups indicated significant within-group differences for the married (F(1,81) = 5.10, p < .05), asymptomatic bereaved (F(1,89) = 7.52, p < .01), and CG (F(1,38) = 9.44, p < .01). We also calculated the absolute value of the difference between mean ability scores and found that the difference was significantly greater in the CG group than asymptomatic bereaved (F(1,128) = 7.33, p < .01) and married individuals (F(1,120) = 6.85, p < .05). Married and asymptomatic bereaved did not differ (F(1,171) = .003, p = .96).

Figure 1.

Forward-Focus and trauma-focus means for the married, asymptomatic bereaved, and CG groups. Scores were collapsed across countries.

Alternative Analyses

To more precisely examine whether the effects were specific to the diagnosis of complicated grief, we repeated the analysis described above while controlling for total number of depression symptoms and, in a separate analysis, the presence-absence of MDD. In each case the interaction remained significant. In addition, to examine whether the findings were dependent on the separation distress criteria, we repeated these analyses using a simplified CG diagnostic criteria that collapsed all grief symptoms including separation distress symptoms and defined CG as 4 or more grief symptoms.¹¹ The relevant interaction again remained significant.

Summary Flexibility Score

To test flexibility more comprehensively, we next conducted group-level analyses using the summary flexibility score. An ANOVA revealed a main effect for bereavement status, F(2,210) = 5.61, p < .005, Eta² = .052, but not country, F(1,210) = .456, p = .50, Eta² = .002 (Figure 2). Follow-up pairwise comparisons indicated that CG participants had less flexibility (M = .79, SD = .11) than both the married (M = .84, SD = .07; p < .005) and the asymptomatic bereaved (M = .84, SD = .08; p < .01) groups. The asymptomatic bereaved and married groups did not differ (p = .93). As previously, we repeated this analysis covarying for total depression symptoms, absence/presence of MDD, and using the simplified CG criterion. In each case, the main effect of bereavement group remained significant.

Figure 2.

Mean coping flexibility scores (see Methods) for the married, asymptomatic bereaved, and CG groups. Scores were collapsed across countries.

Discussion

Comparing coping ability in American and Chinese samples revealed an interaction between coping type and bereavement status. Asymptomatic bereaved and CG groups both reported higher trauma-focused coping than married individuals but the CG group reported less forward-focused coping than both the asymptomatic bereaved and married groups. These results were found to be similar in each country. When we collapsed the scales into a single flexibility score, the results were even clearer. The CG group was less flexible than either the married or asymptomatic bereaved groups. Again, country of origin did not influence the results.

These findings suggest that the ability to flexibly employ different coping strategies is more important in distinguishing pathological reactions to grief than simply evaluating an individual’s ability to exercise a particular coping strategy. An alternative explanation for these findings, however, might dismiss the relevance of flexibility, and focus solely on the deficits in forward-focused coping shown by people with CG. That is, the ability to alternate between the two coping orientations may not be as important as merely being capable of looking beyond emotionally salient experiences. In support of the flexibility hypothesis, however, we note that both bereaved groups had higher trauma-focused coping than married individuals, which demonstrates the relevance of this form of coping during bereavement. We emphasize nonetheless that further research is needed more conclusively rule out the possibility that excessive forward-focused coping characterizes pathological outcomes.

An intriguing finding in the current study was that coping flexibility showed similar patterns in two different cultures. The flexibility construct distinguishes pathological responses to grief while taking into account the diversity of reactions to loss. Although myriad aspects of Chinese culture have been implicated in producing dramatically distinct bereavement experiences, we still found the relationship between lower coping flexibility and complicated grief reactions to be similar in American and Hong Kong samples. It was also of interest that the preferred coping style was identical for groups across countries, with the married groups tending toward forward-focused coping and the bereaved groups tending toward trauma-focused coping. The negligible effect size of the 3-way interaction between bereavement group, coping style and country suggests that in this case nationality did not influence the degree one can utilize a particular coping style prior to and following the death of a spouse. This finding contrasts with earlier research comparing Eastern and Western coping strategies. ^35,36

Limitations

The PACT was developed to measure coping abilities in reaction to a broad range of PTEs. In the current study, we used the scale to measure coping abilities in married persons and individuals who had experienced the specific event of loss. Our findings support the utility of this scale as a meaningful predictor of bereavement stress, but further research is clearly needed on this question. By definition, the concept of coping flexibility applies to the varied demands of different stressors. However, delimiting a singular reference stressor to subjects prior to their completing the PACT may yield different patterns of results. Previous studies argue against this possibility. For example, PACT scores were unrelated to PTE exposure.⁸ Moreover, the more inclusive form of the PACT used in the current study was consistent with both the general concept of coping flexibility⁵ and the more bereavement-specific dual-process model.⁴ Nonetheless, future studies might better tease apart these models and illuminate their similarities and differences, as well as the unique situational constraints on their measurement.

A major limitation of the current study was its cross-sectional design. Although this and a previous study⁸ provide clear evidence for group differences in coping flexibility, it is unclear if the construct forms a stable trait, or if the differences develop following bereavement. Previous research on expressive flexibility, measured with an experimental paradigm, demonstrated surprisingly stable test-retest reliability.²² Implicit in the PACT is the assumption that coping abilities are also relatively stable. However, data relevant to this question are not yet available and future studies with repeated assessments are needed, along with experimental paradigms that can discern if PACT scores are predictive of actual coping behaviors in response to a specific PTE.

A further important step in research on the PACT will be to obtain prospective data prior to the advent of a PTE. The paucity in prospective data is a problem that plagues many of the personality-based measures assumed to influence trauma and loss outcomes.³⁷ Nonetheless, we underscore the importance of such data because only when a scale is measured prior to the occurrence of a PTE is it possible to accurately and independently assess the scale’s influence on post-event outcomes.

Another concern raised by this study, as well as in other coping research, is the limitations inherent in self-report ratings. Most of our team’s research on flexibility has used experimental measures to assess the construct.^6,11,22 We developed the PACT⁸ for the explicit purpose of providing a simple, easy-to-administer self-report measure for trauma field studies. Nonetheless, the limitations of this approach need to be acknowledged, as does the fact that participant self-report is not the only non-experimental method for assessing coping flexibility. For example, broader measures that provide vignettes or problems scenarios¹⁰ or observational diary studies are viable alternatives.

Finally, the cultural comparisons made in this study should be interpreted with caution, as little research has been performed in this area. The two coping abilities we measured were broad in nature, and do not take into account the more subtle approaches of confronting extremely distressing situations. The heterogeneity of the Hong Kong sample, combined with the consideration that Hong Kong is more Western in nature than other areas of Southeast Asia, further cautions against making strong cultural claims.

Clinical implications and future research

Within the context of these limitations, our findings suggest several implications for future research and clinical intervention. First, our findings point to new questions about specific aspects of coping with loss. We noted that the findings are consistent with both the flexibility model and dual-process model of bereavement.⁴ However, the models are not identical. The dual-process model assumes that healthy coping requires both orientations and that neglecting either results in psychopathology.^4,38 For example, individuals too focused on the loss are expected to suffer CG while individuals too focused on restoration are expected to suffer from “absent grieving.”^24,38 By contrast, the flexibility model is less prescriptive about the proportions of trauma-focused and forward-focused coping. Rather these types of coping ability may be utilized to differing degree depending on the demands presented by the stressor.

Although our data cannot resolve this difference, we note that CG individuals did not display greater trauma-focused coping than asymptomatic bereavement, but to the contrary a deficit in forward-focused coping. Moreover, although we could not directly test the hypothesized association between restoration and absent grief, previous research has called into question the legitimacy of the concept of absent grief as a pathological form of mourning.^39–41

Second, our findings contain implications for intervention. Traditionally, grief treatments have focused almost exclusively on the impact and meaning of the loss. However, as our results show, CG participants appear to have no trouble engaging in these types of behaviors. More to the point, our findings indicate that individuals with CG have clear deficits in engaging in forward-focused or more restoration-oriented behaviors. Consistent with this finding, recent interventions for CG that specifically targeted the ability to move beyond a loss have reported promising results. ^2,42 It will be important for future research using these interventions to dismantle the extent that forward-focused techniques are necessary for successful post-treatment recovery.

It should be noted that forward-focused and trauma-focused coping each encompass sets of more specific coping behaviors that when combined constitute a person’s coping repertoire.^5,43 However, the pool of coping behaviors one can choose from is but one dimension of adapting to life stressors. As evidenced by other researchers,^10,21,28 correctly discerning situational demands is also an important component of flexibility. Although the current study does not take the context of quotidian stressors and selected coping strategies into account, clinical interventions in particular would benefit from investigations of deficits in context sensitivity and how they might be improved.

To conclude, our results provide evidence that individuals with CG evidenced a deficit of coping flexibility, and in particular an inability to engage in processes aimed at moving forward beyond the stressor event. Further, this deficit was evident in bereaved samples both in the US and in Hong Kong. Though the PACT was developed for use in researching traumatized populations, the current findings suggest it is an effective tool for investigating coping abilities in other forms of psychopathology. Hopefully, future studies will elucidate if and how forward-focused coping, trauma-focused coping, and the flexibility to use both relates to pathological responses to grief and other aversive experiences.