Is Suicide Bereavement Different? a Reassessment of the Literature

Suicide and Life-Threatening Behavior 31(1) Spring 2001 91  2001 The American Association of Suicidology

Is Suicide Bereavement Different? A Reassessment of the Literature

JOHN R. JORDAN,PHD

The question of whether suicide bereavement is different from mourning after other types of deaths has important theoretical and clinical implications. Some recent literature reviews have argued that the differences may be minimal. In contrast, this article suggests that suicide bereavement is distinct in three significant ways: the thematic content of the grief, the social processes surrounding the survivor, and the impact suicide has on family systems. In addition, problems in the methodology used to compare different bereavement experiences are addressed. Some clinical implications of these conclusions, including the need for homogeneous support groups, psychoeducational services, and family and social network interventions are also discussed.

It would seem obvious that surviving the ample, van der Wal (1989–1990) concluded death of a loved one from suicide is a difficult that “there is no empirical evidence for the experience, one that has the potential to pro- popular notion that survivors of suicide show duce a markedly different type of mourning more pathological reactions and a more com- process from other types of losses. Certainly, plicated and prolonged grief process than mourning after the suicide of a loved one is other survivor groups” (p. 167). More re- often perceived by the survivor to be a very cently, Cleiren and Diekstra (1995) have sug- different experience from the losses of other gested that “it is unlikely that the symptom- mourners (Alexander, 1991; Bolton, 1983; atology of problematic adaptation in suicide Wertheimer, 1991). A consensus of clinicians bereavement differs from that of other types and researchers also indicate that the mourn- of bereavement” (p. 31). They note that the ing process after suicide is different and more symptom patterns common in suicide be- difficult than mourning after other types of reavement are also found in other types of deaths (Clark & Goldney, 1995; Hauser, traumatic loss, even in some losses due to ill- 1987; Knieper, 1999; Rando, 1993; Range, ness. McIntosh (1993) also reached similar 1998; Sprang & McNeil, 1995; Worden, conclusions in his literature review of the 1991). Yet several researchers who have re- more methodologically rigorous investiga- cently reviewed the literature argue that tions where suicide survivors are compared there may be few, if any, empirically docu- to survivors of other types of death. He sug- mented differences between suicide bereave- gested four generalizations about survivors: ment and other types of mourning. For ex- (a) There appear to be more similarities than differences between suicide and other types of survivors (particularly sudden-death survi- JOHN R. JORDAN,PHD, is the Director, vors); (b) there may be a small number of Suicide Grief Support Program, The Trauma grief reactions that are different for survivors, Center, Allston, Massachusetts. Address correspondence to John R. Jordan but these are not yet clearly established; (c) at The Trauma Center, 14 Fordham Road, All- the course of suicide bereavement may differ ston, MA 02134; E-mail: [email protected]. over time; (d) but after the 2nd year, the re- 92 IS SUICIDE BEREAVEMENT DIFFERENT? actions observed in suicide bereavement der Wal (1989–1990) are correct in noting seem to show few differences from the that the evidence for quantitative differences mourning trajectory for other types of losses. between suicide and other types of bereave- What are we to conclude from the ap- ment is mixed, there is also considerable evi- parent contradiction between the perceptions dence that the qualitative or thematic aspects of people who are bereaved by suicide and of the grief may be different after a suicide the clinicians who work closely with them, (Clark & Goldney, 1995; Cleiren, 1993; and researchers who study survivors from a Dunn & Morrish-Vidners, 1987–1988; Ness greater distance with the tools of social sci- & Pfeffer, 1990; Range, 1998; van der Wal, ence? The question has important theoretical 1989–1990). These special themes of suicide and practical implications for caregivers who bereavement manifest themselves in three wish to help suicide survivors. For example, broad areas of grief response. First, numer- the suicide of a loved one is commonly de- ous studies have found that survivors seem scribed as a risk factor for the development to struggle more with questions of meaning of complicated mourning (Rando, 1993), and making around the death (“Why did they do special clinical services are frequently recom- it?”) (Grad & Zavasnik, 1996; Silverman, mended for survivors (Knieper, 1999). Yet if Range, & Overholser, 1994–1995; Smith, grief after suicide is not different from other Range, & Ulmer, 1991–1992; van der Wal, types of bereavement, then there may be lit- 1989–1990). Because suicide is self-inflicted tle rationale for partitioning out survivors for and violates the fundamental norms of self- additional monitoring or specialized inter- preservation, survivors often struggle to ventions. On the other hand, if we can iden- make sense of the motives and frame of mind tify what is different about suicide from other of the deceased. Second, survivors show losses, yet common to most or all suicide be- higher levels of feelings of guilt, blame, and reavement, we should be able to plan more responsibility for the death than other targeted and effective interventions for this mourners (“Why didn’t I prevent it?”) population. This article is written in response (Cleiren, 1993; Demi, 1984; Kovarsky, 1989; to the recent reviews that argue that suicide McNiel, Hatcher, & Reubin, 1988; Miles & bereavement is not fundamentally different Demi, 1991–1992; Reed & Greenwald, 1991; from other types of mourning. While ac- Silverman et al., 1994–1995). Occasionally, knowledging that it shares many elements survivors feel that they directly caused the common to other forms of loss, this article death through mistreatment or abandonment argues that bereavement after suicide is suffi- of the deceased. More frequently, they blame ciently distinct to merit additional research themselves for not anticipating and prevent- and specialized clinical services for most sui- ing the actual act of suicide (Cleiren & Diek- cide survivors. The goals are to summarize stra, 1995). Third, several studies indicate the empirical support for the themes that set that survivors experience heightened feelings suicide bereavement apart from other forms of rejection or abandonment by the loved of grief, the distinct aspects of social pro- one, along with anger toward the deceased cesses after a suicide, and the differential im- (“How could they do this to me?”) (Barrett & pact of suicide on family systems. Method- Scott, 1990; Reed, 1998; Reed & Greenwald, ological problems in suicide bereavement 1991; Silverman et al., 1994–1995; van der research and the clinical implications of the Wal, 1989–1990). Of special note is a re- differential bereavement experience after sui- cently published study that compared suicide, cide will also be addressed. accidental, and expected and unexpected natural modes of death (Bailley, Kral, & Dun- THEMATIC ASPECTS OF SUICIDE ham, 1999). This large sample study found BEREAVEMENT convincing empirical evidence for differences between suicide survivors and other types of Although reviewers such as Cleiren mourners for all three of these thematic ar- and Diekstra (1995), McIntosh (1993), van eas, including heightened feelings of respon- JORDAN 93 sibility and rejection, greater difficulty mak- titudes toward survivors may not directly re- ing sense of the death, and greater overall sult in differential treatment of survivors by grief reactions. It seems evident from these the community. Instead, it is possible that studies that there are qualitative aspects of many people genuinely wish to help the sur- the mourning process that are more intensi- vivor but also feel uncertain and uncomfort- fied and frequently more problematic for sur- able about how to provide support (Calhoun, vivors of suicide loss than for other types of Selby, & Abernathy, 1986; Dunn & Morrish- mourners. These common themes in suicide Vidners, 1987–1988). This awkwardness and bereavement may distinguish it from other hesitation may then be communicated to losses, regardless of the measured intensity of survivors, and misinterpreted as rejection the grief or psychiatric symptoms. These (Range, 1998). studies challenge the sweeping and overly Beyond the problem of their percep- simplified conclusions made by some observ- tion by others, it is equally important to ask ers that there are few differences between how suicide survivors view themselves. It suicide and other types of bereavement seems plausible that the negative attitude to- (Cleiren & Diekstra, 1995; McIntosh, 1993; ward suicide in our culture will be mirrored & van der Wal, 1989–1990). within the survivor. This is an important point, because even if others feel and demon- strate compassion for the mourner, the survi- SOCIAL PROCESSES vor may assume or fear that others are judg- SURROUNDING SUICIDE ing them negatively and therefore withdraw SURVIVORS or otherwise act in ways that inhibit social support efforts from others. Dunn and Mor- The most comprehensive review to rish-Vidners (1987) have referred to this pro- date of research on suicide survivors (McIn- cess as “self-stigmatization” (p. 177). For tosh, 1993) did not address the issue of the example, Van Dongen (1993) found that sui- impact of the social network on survivors. cide survivors worried more about what oth- Yet there is considerable evidence that survi- ers really thought of them, felt uncertain vors feel more isolated and stigmatized than about how to act and what to share with oth- other mourners and may in fact be viewed ers, and believed that community members more negatively by others in their social net- were likewise uncertain about how to behave work. Research on the social response to around them. Range and Calhoun (1990) suicide has attempted to ascertain whether found that suicide bereavement subjects felt survivors are perceived in a different and more pressure than natural death survivors to more negative fashion than mourners with a explain the cause of death and reported that different type of loss. Separate reviews of the others treated them differently after the literature by Calhoun and Allen (1991) and death. Strikingly, 76% of those bereaved by Stillion (1996), and recent individual studies accidental death reported that the changes in such as the one by Allen, Calhoun, Cann, and social interaction were positive in nature, Tedeschi (1993), have generally shown compared with only 27% of the suicide survi- that “individuals bereaved by suicide tended vors. These authors also report that survivors to be viewed as more psychologically dis- were the only group that reported lying to turbed, less likable, more blameworthy, more others about the cause of death (44% of sub- ashamed, more in need of professional men- jects). Other studies report similar findings tal health care, and more likely to remain sad (Bailley et al., 1999; McNiel et al., 1988), and depressed longer” (Calhoun & Allen, including the observation that suicide survi- 1991, p. 100). Thus there is considerable evi- vors received significantly less emotional dence that the general stigma that continues support than natural death survivors for their to be associated with suicide in our society feelings of depression and grief, and confided “spills over” to the bereaved family members. less in members of their social networks It is important to note that these negative at- (Farberow, Gallagher-Thompson, Gilewski, 94 IS SUICIDE BEREAVEMENT DIFFERENT?

& Thompson, 1992). Wagner and Calhoun cult for the family unit than death from natu- (1991–1992) and Cleiren (1993) did not find ral causes. This may hold true in several quantitative differences in the perception of ways. support, though the qualitative (i.e., interview) data of the former suggested that survi- Family Interaction Patterns vors felt pressure to recover faster and that only other survivors could actually under- stand their experience. Lastly, Se´guin, Le- The preexisting interactional patterns sage, and Kiely (1995) found survivor fami- of some families in which a suicide occurs lies to be more vulnerable and hypothesized may be different from other families, and the that survivors tended to withdraw from their suicide itself may contribute to dysfunctional social network out of shame, causing others, family dynamics. Although by no means in turn, to pull away out of feelings of frus- present in all cases, there is evidence that tration and rejection by the survivor. We can families of many suicidal people (particularly summarize these several points about suicide suicidal children and adolescents) show more bereavement and social support by noting disturbed family interactional styles and in- that there is considerable evidence that sui- creased disruptions of attachments when cide survivors are viewed more negatively by compared with families without a suicidal others and by themselves. It seems probable member (Brent, 1995; McIntosh, 1987; Mos- that both of these factors operate to interfere cicki, 1995; Samy, 1995). Adam (1990) em- with the support process after a suicide (Van phasized that a dysfunctional family environ- Dongen, 1993), depending on the personali- ment can operate as both a predisposing ties and attitudes towards suicide of the survi- element in the early psychosocial develop- vor and members of their social networks. ment of suicidal persons and as a precipitat- Taken together, these studies suggest that in- ing factor in a suicide death. Reviews by terpersonal interaction and social support is Adam (1990) and Blumenthal (1990) have frequently different and more problematic also determined that suicidal adults often after a suicide death than after most other show increased rates of childhood physical types of loss. and sexual abuse, and parental loss or depri- vation in their history. Given these consistent findings of elevated rates of family pathology prior to a suicide, it seems plausible that THE IMPACT OF SUICIDE ON some dysfunctional families might continue FAMILY SYSTEMS to be at the same, or perhaps even greater, risk after the suicide. This risk includes an The loss of an immediate member to increased chance of a subsequent suicide of death almost always has an impact on the another family member at some future point functioning of a family system. Unfortu- (see below). nately, when compared with studies of indi- Even when family functioning may viduals, there is a dearth of research on the have been within a normal range prior to the differential effects of bereavement (including suicide, there is some evidence that suicide suicide) on family functioning. Questions by itself has the potential to warp family pat- such as the impact of death on family com- terns and contribute to the development of munication patterns, conflict resolution, co- psychiatric disorder in surviving family mem- hesion and intimacy, intergenerational rela- bers. For example, in a controlled study of tions, and family developmental tasks have the impact of adolescent suicide on peers, been largely ignored (McNiel et al., 1988). siblings, and parents, Brent and his col- Nonetheless, there is considerable clinical leagues (Brent, Moritz, Bridge, Perper, & evidence, and at least some empirical data, Canobbio, 1996) found higher rates of de- suggesting that a suicide may be more diffi- pression in survivor siblings and mothers JORDAN 95 than in controls at 6 months after the death. for the family to negotiate. Jordan and his They also found continuing elevated rates of colleagues (Bradach & Jordan, 1995; Jordan, depression in survivor mothers at one year, 1991–1992) have found preliminary em- and elevated rates of grief in siblings (partic- pirical support for the negative intergenera- ularly younger siblings) at 12 and 37 months tional impact of traumatic losses on family post death. In an uncontrolled qualitative systems. Other personal and clinical accounts study, Dunn and Morrish-Vidners (1987– of the long term impact of loss on families, 1988) found that twice as many survivors in particularly suicide, have reported similar ef- their small sample reported that relationships fects (Treadway, 1996; Walsh & McGold- with family members (and friends) became rick, 1991). Sleeper effects of traumatic more distant after the suicide than reported deaths such as suicide have received very lit- an increase in closeness. In contrast, when tle empirical investigation yet may be one of they compared bereavement after suicide, ill- the most important dimensions by which sui- ness, and accident in a small sample, Nelson cide deaths differ from other types of losses and Frantz (1996) did not find statistically (Dunn & Morrish-Vidners, 1987–1988). significant differences in family variables. However, suicide survivor families did show Heightened Risk for Additional poorer scores in the expected direction on Family Suicide such variables as enmeshment, conflict, and cohesion. McIntosh (1987) noted three Suicide bereavement is an unusual themes that may be common in families with form of mourning experience, because losing child survivors after parental suicide: infor- a loved one to suicide may elevate the mation/communication distortion (hiding mourner’s own risk for suicidal behavior and the true circumstances of the death), guilt, completion (Blumenthal, 1990; Cleiren, 1993; and identification with the deceased. The Fekete & Schmidtke, 1996; Lester, 1994; creation of a powerful family secret around Moscicki, 1995; Ness & Pfeffer, 1990; Roy, the suicide may have devastating longer term 1992). There are at least two possible expla- effects on the openness of family communi- nations for this phenomenon. First, interper- cation about many emotionally charged is- sonal loss and disruption of attachments from sues, leaving the family in a vulnerable posi- any cause (including, but not limited to, be- tion should another traumatic or shame reavement) appear to elevate the risk for sui- ridden event occur later on (Walsh & Mc- cidality (Heikkinen, Aro, & Lonnqvist, 1993; Goldrick, 1991). Jordan, Kraus, and Ware Moscicki, 1995). The impact of interpersonal (1993) have also identified several aspects of loss appears to be particularly strong when a family interactions that may be affected by history of substance abuse is present in the the death of a member, including the shut- potential suicide victim (Brent, 1995; Mur- down of open communication, disruption of phy, 1995). Reviewing research on the long- role functioning of family members, develop- term impact of childhood parental loss, ment of conflict around differences in be- Adam (1990) found strong and consistent reavement coping styles, destabilization of support for the notion that early parental loss family coalitions and intergenerational is also associated with later suicidal behavior. boundaries, and disruption of relationships Loss has also been linked to increased vul- between the family and its larger social net- nerability to the psychiatric disorders that work. This group also emphasizes the long- may be highly associated with suicidality, term impact of losses, particularly traumatic such as major depression and anxiety disor- deaths such as suicide, on family develop- der in adults (Brown, 1998; Jacobs, 1999). To mental processes, communication patterns, summarize, bereavement or interpersonal and the transmission of a family world view loss in childhood or adulthood from any to future generations. These “sleeper effects” cause is a risk factor for increased suicidality, may make future separations more difficult both directly as a proximal precipitant for 96 IS SUICIDE BEREAVEMENT DIFFERENT? suicide and indirectly through the creation or use conventional methods and measures to exacerbation of psychiatric illness in survi- assess bereavement outcome. These include vors. the type of research methodology and out- Beyond the general influence of be- come criteria employed, the other types of reavement, suicide survivors may be at in- losses used for comparison with suicide be- creased risk as a result of familial factors, reavement, the possibility of relief from some both genetic and environmental, that may in- stressors after a suicide, and the longer term crease the predisposition towards suicide in a versus near-term impact of this type of death. family system. There is evidence that genetic factors can predispose people towards the de- Categories of Bereavement velopment of psychiatric disorders that are Outcome Criteria associated with suicide, particularly depression and bipolar disorders (Kety, 1990; Mos- Most studies of bereavement utilize cicki, 1995). There may also be a specific in- easily quantifiable self-report measures as heritable biological factor that increases the outcome criteria for comparing bereave- chances of suicide (Brent, 1996; Roy, 1992). ments. These measures typically assess psy- Psychological and family systems variables chiatric symptoms (depression, anxiety, or may also play a role in the familial transmis- posttraumatic stress disorder) or global mea- sion of suicide. As noted previously, suicide sures of social, medical, and occupational has been associated with family factors such functioning. Although relevant, these are not as disorganization and breakup, substance the only way to evaluate outcome for survi- abuse, intrafamily violence, and sexual abuse. vors. Simple quantitative measures of grief The dynamics of some families may also be may not detect some of the thematic or qual- “suicidogenic,” displaying scapegoating, guilt itative differences noted previously, such as induction, and hostility toward a member the heightened feelings of guilt and preoccu- that contributes to the eventual suicide pation with the question of why the death oc- (Samy, 1995). To the extent that these dys- curred. These are more likely to be observed functional patterns contributed to one sui- in qualitatively based research methodology cide, they may also increase the suicide risk that allows research participants to explain for other surviving family members. In addi- their experience to the researcher in their tion, exposure to suicide, particularly for own words (Neimeyer & Hogan, 2001). For young people, may increase the chances of example, studies by McNiel and colleagues suicidality in the exposed person (Blumen- (1988) and Wagner and Calhoun (1991– thal, 1990; Diekstra & Garnesfski, 1995; 1992) both found differences between suicide Moscicki, 1995). This modeling effect, by and other types of survivors in their interview which suicide becomes an acceptable “solu- data, but not in their quantitative data. tion” to intrapsychic and interpersonal prob- There is also growing empirical evi- lems, may have a powerful influence in some dence for a distinct form of grief that has families, particularly on children as they de- been termed traumatic grief (Jacobs,1999; velop into adults. Prigerson et al.,1999). Prigerson, Jacobs, and their colleagues have empirically demonstrated that traumatic grief is a syndrome that DIFFERENT IN WHAT WAY: is distinct from depression and anxiety (Prig- METHODOLOGICAL ISSUES IN erson et al.,1996), and that is predictive of SUICIDE BEREAVEMENT mental and physical outcome for the be- RESEARCH reaved, including suicidal ideation (Prigerson et al.,1997). If this new diagnostic entity There may be a number of ways that holds up after further empirical testing, then some of the unique effects associated with standard outcome measures used to assess suicide grief are not detected in studies that other psychiatric problems will not be ade- JORDAN 97 quate to measure this disorder. To date, no are likely to be associated with complicated studies have assessed whether suicide survi- mourning. Accordingly, our research efforts vors differ from other mourners on this im- may need to be concentrated on the common portant dimension. Nonetheless, it seems characteristics of bereavement after all trau- likely that traumatic grief is one likely se- matic death, as well as the unique character- quelae of a suicide. To summarize, studies istics of suicide bereavement. that compare suicide bereavement to other types of losses by using only quantitative (as The Relief Effect after Suicide opposed to qualitative) measures, and that assess only general aspects of functioning (as Cleiren (1993), Grad and Zavasnik opposed to suicide specific domains) may fail (1996), and Reed (1998) all found that many to detect differences that emerge with mea- of the suicidally bereaved families in their sures and research methods intended to spe- studies had a long history of problems with cifically assess suicide grief. Without these the deceased, who often exhibited chronic types of studies, we may mistakenly conclude psychiatric problems, aberrant behavior, and that there are no differences in the mourning in some cases, previous suicide attempts. process between suicide and other types of Cleiren (1993) noted that these families losses. With them, we may be able to tease would most probably show heightened stress out some of the subtle but important distinc- (and elevated levels of symptoms), even if the tions that have significant treatment implica- suicide had not occurred. He also observed tions for caregivers (Bailley et al.,1999). that relief was as common in suicide survivor families as in those where the loved one had Suicide versus Other Traumatic Loss died after a long-term illness. In the same vein, some studies have found that for a siz- Many of the controlled studies at- able number of families, the death of their tempting to ascertain whether suicide be- loved one to suicide was not completely un- reavement is different compare suicide with expected (Cleiren & Diekstra, 1995; Grad & another type of traumatic death, most com- Zavasnik, 1996). These findings indicate that monly accidental death. Several of these the families of many (though not all) suicide studies report that accidental and suicide completers have experienced a difficult and deaths produce similar types of bereavement often lengthy ordeal of living with an emo- reactions, sometimes in contrast to natural tionally disturbed and self-destructive per- death losses (Bailley et al., 1999; Barrett & son. In such cases, it seems plausible to sug- Scott, 1990; Grad & Zavasnik, 1996; McIn- gest that the death of such a member may tosh & Kelley, 1992; McNiel et al., 1988; sometimes reduce the overall stress levels in Miles & Demi, 1991–1992; Range & Cal- the family, however painful the loss may be houn, 1990; Ulmer, Range, & Smith, 1991). for the survivors. Likewise, if the death was This similarity of response between suicide to some extent anticipated (or perhaps and other traumatic deaths makes clinical feared), this may attenuate some of the shock sense and may account for some of the appar- effects associated with other types of sudden, ent “washing out” in many studies of a dis- traumatic deaths. In short, there may be a tinct effect related to suicide as a mode of “relief effect” for some survivors (Calhoun, death. It is possible that the unique features Selby, & Selby, 1982) that makes the grief a of traumatic deaths, when present in suicide mixed experience of negative emotions, such or in any other traumatic loss, account for as guilt, rejection, abandonment, and sorrow, much of the variance in bereavement out- coupled with relief at not having to cope with come in comparison to natural causes of the destructive behavior of the loved one. death. If this is true, then it may be useful to Those who experience this relief effect may conceptualize suicide as one example of the have a different course of mourning, showing more general class of traumatic deaths that a diminution of stress-related psychiatric 98 IS SUICIDE BEREAVEMENT DIFFERENT? symptoms when compared to families where than for accidental death survivors, the mea- the prior relationship with the deceased was sured grief intensity of suicidally bereaved less disturbing. Symptom levels in this group subjects stayed the same or even increased may be similar to individuals who experience over time. less traumatic losses, masking the impact of Given this conflicting data, we cannot suicide on survivors who are more severely say definitively whether the longer term tra- traumatized by the death. Again, this seems jectory of suicide bereavement is the same or particularly likely if the criteria used for be- different from that of other types of losses. reavement outcome are simply self-report Echoing this idea, Dunn and Morrish-Vid- measures of psychiatric symptoms. Nonethe- ners (1987–1988) noted that there is a dearth less, many of the thematic and qualitative as- of knowledge about the longer term, existen- pects of suicide bereavement, such as height- tial impact of suicide on survivors. For exam- ened guilt and anger at being abandoned, ple, suicide (as well as other forms of trauma) may still be present, even if the death is in may disrupt the assumptive world or cogni- some ways a relief. Whenever possible, fu- tive schemas of survivors about their sense of ture research on suicide bereavement should the safety, efficacy, and personal worthiness attempt to assess the extent to which the (Janoff-Bulman, 1992). The impact of these death may have been anticipated by the sur- profound changes in core belief systems on vivors, and the degree to which the stress lev- developmental processes in survivors has els in the family have decreased as a result of been largely ignored in empirical suicide be- the death. Suicide survivors are probably not reavement research yet may be a crucial fac- a homogeneous group (Bailley et al., 1999), tor that distinguishes this type of loss from and the relief effect may be one important more normative bereavement experiences. variable that differentiates survivors from one another. CLINICAL IMPLICATIONS Time Frame of Research on Suicide Bereavement: Sleeper Effects What are the clinical implications of these four points? Are specialized interven- Research on bereavement resulting tions warranted for suicide survivors, and, if from different modes of death has produced so, how would they differ from other types conflicting findings as to whether differences of bereavement services? First, it is probably increase or decrease with time. For example, best to make support services for survivors several studies have found that any initial dif- homogeneous with regard to mode of death. ferences due to modality of death disap- Given the special thematic aspects of suicide peared 2–4 years after the death (Barrett & bereavement, and the demonstrated stigmati- Scott, 1990; Cleiren, 1993; Demi, 1984). zation that many survivors perceive in their These studies would seem to suggest that social networks, groups limited to suicide over time the pattern of mourning from dif- survivors seem likely to cohere more quickly ferent types of losses tends to converge to a and to avoid a replication of the empathic common pathway. In contrast, Thompson, failure that too often occurs for survivors in Futterman, Farberow, Thompson, and Pe- their larger social networks. Although not al- terson (1993) found that the course of ways feasible for economic or logistical rea- mourning for suicide survivors actually di- sons, whenever possible suicide survivors verged from natural death survivors over should be offered the opportunity to interact time, such that suicide survivors took much with other suicide survivors, not just other longer for symptoms to abate and remained mourners. higher on some dimensions (anxiety) up to Second, with the elevated risk of sui- 30 months after the death. Kovarsky (1989) cidality associated with survivorship, man- found that although initially being lower agement of survivors must include not only JORDAN 99 support for their grief but also proactive latter could include psychoeducational mate- monitoring of their risk for psychiatric disor- rials and meetings designed to support and ders and suicidality. Unfortunately, most be- educate those who are directly supporting reavement support programs do not system- the mourner. atically monitor the participants’ risk for Lastly, although the case can be made development of these problems. Given the that all bereavement services should be di- demonstrable link between the suicide of a rected toward family systems, this seems par- member and the increase in risk for other ticularly true for suicide survivors. Given the family members, it is disappointing that so increased risk of additional suicides, the dam- little research or clinical attention has been aging ramifications for family communica- paid to postvention with survivors as a poten- tion and developmental processes, and the tially effective form of prevention of future special difficulties of children who lose a suicides. family member to suicide, the facilitation of Third, support services should provide adaptive family emotional functioning through- psychoeducational resources that help edu- out the mourning process is crucial. As men- cate survivors about the nature of suicide and tioned previously, effective postvention with suicide bereavement. Making sense of the suicidally bereaved families may be one of suicide of their loved one is a major recovery the most important forms of multigenera- task for survivors. Compared to other forms tional prevention available to mental health of mourning, suicide survivors typically spend professionals. much more energy trying to comprehend the Given the present state of our knowl- reasons for the death, the motivations of the edge, perhaps the fundamental question deceased, and the appropriate allocation of posed by this article, “Is suicide bereavement responsibility for the suicide. Support ser- different?” cannot be definitively answered at vices should provide many structured and in- this time. There is a need for additional in- formal opportunities for survivors to learn formation about the mourning process in more about suicide, and to put the death in a general and suicide bereavement in particu- larger perspective. Psychoeducational pre- lar, before targeted interventions for this sentations, reading materials, and discussions population can be designed with any degree with mental health professionals and other of specificity. Nonetheless, there is more survivors can all be of use in this process. than enough evidence that suicide bereave- Fourth, support services should target ment is different from other types of losses the interface between the survivor and their to justify the continuing inquiry into com- social network. Because the research suggests parative bereavement responses. Likewise, that many survivors feel stigmatized and there is a great need to develop and test in- withdraw from friends and family, survivors terventions that address the special needs of often need help in dealing with the social suicide survivors. Although some general aftermath of a suicide. Although many be- treatment techniques may be of great help to reavement support services include some dis- some survivors (Knieper, 1999), there is cussion of social problems, few programs sys- still almost a complete absence of empirically tematically target this important issue, let validated interventions that specifically ad- alone attempt to intervene directly in the dress the thematic, social, and family sys- survivor’s social network through psychoedu- tem problems noted in this article. Based on cational and network type interventions the additional knowledge and increased clini- (Provini, Everett, & Pfeffer, 2000). Discus- cal awareness that such efforts will foster, fu- sion of specific coping skills and interper- ture programs can be designed that provide sonal tactics for dealing with stigma and focused, effective, and compassionate help shame should be offered, and interventions for survivors as they travel their difficult targeted directly at the larger social network journey after the death of a loved one to sui- should be included whenever possible. The cide. 100 IS SUICIDE BEREAVEMENT DIFFERENT?

REFERENCES

ADAM,K.S. (1990). Environmental, psy- Are suicidal deaths different? Journal of Commu- chosocial, and psychoanalytic aspects of suicidal nity Psychology, 14, 213–218. behavior. In S. J. Blumentahl & D. J. Kupfer CALHOUN, L. G., SELBY, J. W., & SELBY, (Eds.), Suicide over the life cycle: Risk factors, assess- L. E. (1982). The psychological aftermath of sui- ment, and treatment of suicidal patients (pp. 39–96). cide: An analysis of current evidence. Clinical Psy- Washington, DC: American Psychiatric Press. chology Review, 2, 409–420. ALEXANDER,V. (1991). Words I never CLARK, S., & GOLDNEY,R.D. (1995). thought to speak: Stories in the wake of suicide. New Grief reactions and recovery in a support group York: Lexington Books. for people bereaved by suicide. Crisis, 16, 27–33. ALLEN, B. G., CALHOUN, L. G., CANN, CLEIREN,M. (1993). Bereavement and ad- A., & TEDESCHI,R.G. (1993–1994). The effects aptation: A comparative study of the aftermath of of cause of death on responses to the bereaved: death. Washington, DC: Hemisphere. Suicide compared to accidental and natural causes. CLEIREN, M., & DIEKSTRA,R. (1995). Omega, 28, 39–48. After the loss: Bereavement after suicide and other BAILLEY, S. E., KRAL, M. J., & DUNHAM, types of death. In B. Mishara (Ed.), The impact of K. (1999). Survivors of suicide do grieve differ- suicide (pp. 7–39). New York: Springer. ently: Empirical support for a common sense DEMI,A.S. (1984). Social adjustment of proposition. Suicide and Life-Threatening Behavior, widows after a sudden death: Suicide and non-sui- 29, 256–271. cide survivors compared. Death Education, 8, 91– BARRETT, T. W., & SCOTT,T.B. (1990). 111. Suicide bereavement and recovery patterns com- DIEKSTRA, R., & GARNEFSKI,N. (1995). pared with non-suicide bereavement patterns. Sui- On the nature, magnitude, and causality of suicide and Life-Threatening Behavior, 20, 1–15. cidal behaviors: An international perspective. Sui- BLUMENTHAL,S.J. (1990). An overview cide and Life-Threatening Behavior, 25, 36–57. and synopsis of risk factors, assessment, and treat- DUNN, R. G., & MORRISH-VIDNERS, D. ment of suicidal patients over the life cycle. In (1987–1988). The psychological and social experi- S. J. Blumentahl & D. J. Kupfer (Eds.), Suicide ence of suicide survivors. Omega, 18, 175–215. over the life cycle: Risk factors, assessment, and treat- FARBEROW, N., GALLAGHER-THOMP- ment of suicidal patients (pp. 685–734). Washing- SON, D., GILEWSKI, M., & THOMPSON, L. ton, DC: American Psychiatric Press. (1992). The role of social support in the bereave- BOLTON,I. (1983). My son . . . my son . . . : ment process of surviving spouses of suicide and A guide to healing after death, loss, or suicide. Atlanta, natural deaths. Suicide and Life-Threatening Behav- GA: Bolton Press. ior, 22, 107–124. BRADACH, K. M., & JORDAN,J.R. (1995). FEKETE, S., & SCHMIDTKE,A. (1996). Long-term effects of a family history of traumatic Suicidal models: Their frequency and role in sui- death on adolescent individuation. Death Studies, cide attempters, non-suicidal psychiatric patients, 19, 315–336. and normal control cases: A comparative German- BRENT,D.A. (1995). Risk factors for ado- Hungarian study. Omega, 33, 233–241. lescent suicide and suicidal behavior: Mental and GRAD, O., & ZAVASNIK,A. (1996). Simi- substance abuse disorders, family environmental larities and differences in the process of bereave- factors, and life stress. Suicide and Life-Threatening ment after suicide and after traffic fatalities in Behavior, 25 (Suppl.) 52–63. Slovenia. Omega, 33, 243–251. BRENT, D. A., MORITZ, G., BRIDGE, J., HAUSER,M.J. (1987). Special aspects of PERPER, J., & CANOBBIO,R. (1996). The impact grief after a suicide. In E. J. Dunne, J. L. McIn- of adolescent suicide on siblings and parents: A tosh, & K. Dunne-Maxim (Eds.), Suicide and its longitudinal follow-up. Suicide and Life-Threaten- aftermath: Understanding and counseling the survi- ing Behavior, 26, 253–259. vors (pp. 57–72). New York: Norton. BROWN,G.W. (1998). Loss and depres- HEIKKINEN, M., ARO, H., & LONNQVIST, sive disorders. In B. Dohrenwend (Ed.), Adversity, J. (1993). Life events and social support in suicide. stress, and psychopathology (pp. 358–370). New Suicide and Life-Threatening Behavior, 23, 343–358. York: Oxford University Press. JACOBS,S. (1999). Traumatic grief: Diagno- CALHOUN, L. G., & ALLEN,B.G. (1991). sis, treatment, and prevention. Philadelphia: Tay- Social reactions to the survivor of a suicide in the lor & Francis. family: A review of the literature. Omega, 32, 95– JANOFF-BULMAN,R. (1992). Shattered as- 107. sumptions: Towards a new psychology of trauma. New CALHOUN, L. G., SELBY, J. W., & ABER- York: Free Press. NATHY,C.B. (1986). The rules of bereavement: JORDAN,J.R. (1991–1992). Cumulative JORDAN 101 loss, current stress, and the family: A pilot investi- S. V., REYNOLDS, C. F., SHEAR, M. K., DAY, N., gation of individual and systemic effects. Omega, BEERY, L. C., NEWSOM, J. T., & JACOBS, S. 24, 309–332. (1997). Traumatic grief as a risk factor for mental JORDAN, J. R., KRAUS, D. R.,&WARE, E. and physical morbidity. American Journal of Psychi- S. (1993). Observations on loss and family devel- atry, 154, 616–623. opment. Family Process, 32, 425–440. PRIGERSON, H. G., SHEAR, S. C., JACOBS, KETY,S.S. (1990). Genetic factors in sui- S. C., REYNOLDS, C. F., MACIEJEWSKI, P. K., cide: family, twin, and adoption studies. In S. J. DAVIDSON, J. R. T., ROSENHECK, P. A., PIL- Blumentahl & D. J. Kupfer (Eds.), Suicide over the KONIS, P. A., WORTMAN, C. B., WILLIAMS, J. B. life cycle: Risk factors, assessment, and treatment of sui- W., WIDIGER, T. A., FRANK, E., KUPFER, D. cidal patients (pp. 127–143). Washington, DC: J., & ZISOOK,S. (1999). Consensus criteria for American Psychiatric Press. traumatic grief: A preliminary empirical test. Brit- KNIEPER,A. (1999). The suicide survivor’s ish Journal of Psychiatry, 174, 67–73. grief and recovery. Suicide and Life-Threatening Be- PRIGERSON, H. G., BIERHALS, A. J., KASL, havior, 29, 353–364. S. V., REYNOLDS, C. F., SHEAR, M. K., NEW- KOVARSKY,R.S. (1989). Loneliness and SOM, J. T., & JACOBS,S. (1996). Complicated disturbed grief: A comparison of parents who lost grief as a disorder distinct from bereavement-re- a child to suicide or accidental death. Archives of lated depression and anxiety: A replication study. Psychiatric Nursing, 3, 86–96. American Journal of Psychiatry, 153, 1484–1486. LESTER,D. (1994). Challenges in prevent- PROVINI, C., EVERETT, J. R.,&PFEFFER, ing suicide. Death Studies, 18, 623–639. C. R. (2000). Adults mourning suicide: Self-re- MCINTOSH,J.L. (1987). Survivor family ported concerns about bereavement, needs for as- relationships: Literature review. In E. J. Dunne, J. sistance, and help-seeking behavior. Death Studies, L. McIntosh, & K. Dunne-Maxim (Eds.), Suicide 24, 1–19. and its aftermath: Understanding and counseling the RANDO,T.A. (1993). Treatment of compli- survivors (pp. 73–84). New York: Norton. cated mourning. Champaign, IL: Research Press. MCINTOSH,J.L. (1993). Control group RANGE,L. (1998). When a loss is due to studies of suicide survivors: A review and critique. suicide: Unique aspects of bereavement. In J. H. Suicide and Life-Threatening Behavior, 23, 146–160. Harvey (Ed.), Perspectives on loss: A sourcebook. Phil- MCINTOSH, J. L., & KELLY,L.D. (1992). adelphia: Brunner/Mazel. Survivors’ reactions: Suicide vs. other causes. Cri- RANGE, L. M., & CALHOUN,L.G. (1990). sis, 13, 82–93. Responses following suicide and other types of MCNIEL, D. A., HATCHER, C., & REU- death: The perspective of the bereaved. Omega, BIN,R. (1988). Family consequences of suicide 21, 311–320. and accidental death: Consequences for widows. REED,M.D. (1998). Predicting grief Suicide and Life-Threatening Behavior, 18, 137–148. symptomatology among the suddenly bereaved. MILES, M. S., & DEMI,A.S. (1991–1992). Suicide and Life-Threatening Behavior, 28, 285–300. A comparison of guilt in bereaved parents whose REED, M. D., & GREENWALD, J. Y. children died by suicide, accident, or chronic dis- (1991). Survivor-victim status, attachment, and ease. Omega, 24, 203–215. sudden death bereavement. Suicide and Life- MOSCICKI,E.K. (1995). Epidemiology of Threatening Behavior, 21, 385–401. suicidal behavior. Suicide and Life-Threatening Be- ROY,A. (1992). Genetics, biology, and sui- havior, 25, 22–35. cide in the family. In R. W. Maris, A. L. Berman, MURPHY,G.E. (1995). 39 Years of suicide J. T. Maltsberger, & R. I. Yufit (Eds.), Assessment research: A personal view. Suicide and Life-Threat- and prediction of suicide (pp. 574–588). New York: ening Behavior, 25, 450–457. Guilford Press. NEIMEYER, R. A., & HOGAN,N.S. (2001). SAMY,M.H. (1995). Parental unresolved Quantitative or qualitative? Measurement issues ambivalence and adolescent suicide: A psychoana- in the study of grief. In M. S. Stroebe, R. O. Han- lytic perspective. In B. Mishara (Ed.), The impact nson, W. Stroebe, & H. Schut (Eds.), Handbook of of suicide (pp. 40–51). New York: Springer. bereavement research: Consequences, coping, and care.SE´ GUIN, M., LESAGE, A., & KIELY, M. C. Washington, DC: American Psychological Associ- (1995). Parental bereavement after suicide and ac- ation Press. cident: A comparative study. Suicide and Life- NELSON, B. J., & FRANTZ,T.T. (1996). Threatening Behavior, 25, 489–499. Family interactions of suicide survivors and survi- SILVERMAN, E., RANGE, L.,&OVER- vors of non-suicidal death. Omega, 33, 131–146. HOLSER,J. (1994–1995). Bereavement from sui- NESS, D. E., & PFEFFER,C.R. (1990). Se- cide as compared to other forms of bereavement. quelae of bereavement resulting from suicide. Omega, 30, 41–51. American Journal of Psychiatry, 147, 279–285. SMITH, P. C., RANGE, L. M., & ULMER, PRIGERSON, H. G., BIERHALS, A. J., KASL, A. (1991–1992). Belief in afterlife as a buffer in 102 IS SUICIDE BEREAVEMENT DIFFERENT? suicidal and other bereavement. Omega, 24, 217– (1991). Purpose in life: A moderator of recovery 226. from bereavement. Omega, 23, 279–289. SPRANG, G., & MCNEIL,J. (1995). The VAN DER WAL,J. (1989–1990). The after- many faces of bereavement: The nature and treatment math of suicide: A review of empirical evidence. of natural, traumatic, and stigmatized grief. New Omega, 20, 149–171. York: Brunner/Mazel. VAN DONGEN,C.J. (1993). Social context STILLION,J. (1996). Survivors of suicide. of postsuicide bereavement. Death Studies, 17, In K. Doka (Ed.), Living with grief after sudden loss: 125–141. Suicide, homicide, accident, heart attack, stroke (pp. WAGNER, K. G., & CALHOUN, L. G. 41–52). Washington, DC: Hospice Foundation of (1991–1992). Perceptions of social support by sui- America. cide survivors and their social networks. Omega, THOMPSON, D. G., FUTTERMAN, A., 24, 61–73. FARBEROW, N., THOMPSON, L. W., & PE- WALSH, F., & MCGOLDRICK,M. (1991). TERSON,J. (1993). The impact of spousal be- Living beyond loss: Death and the family. New York: reavement on older widows and widowers. In M. Norton. S. Stroebe, W. Stroebe, & R. O. Hansson (Eds.), WERTHEIMER,A. (1991). A special scar: The Handbook of bereavement: Theory, research, and in- experiences of people bereaved by suicide. London, En- tervention (pp. 227–239). New York: Cambridge gland: Tavistock/Routledge. University Press. WORDEN,J.W. (1991). Grief counseling TREADWAY,D.C. (1996). Dead reckoning: and grief therapy (2nd ed.). New York: Springer. A therapist confronts his own grief. New York: Basic Books. Manuscript Received: March 6, 1999 ULMER, A., RANGE, L., & SMITH, P. C. Manuscript Accepted: April 1, 2000