CONTINUING BONDS AND RECONSTRUCTING
MEANING: MITIGATING COMPLICATIONS
ROBERT A. NEIMEYER, SCOTT A. BALDWIN, and
The University of Memphis, Memphis, Tennessee, USA
Drawing on attachment theory and constructivist conceptualizations of
bereavement, the authors assessed the relation between continuing bonds coping
and meaning reconstruction following the death of a loved one and complicated
grief symptomatology. Five hundred six young adults in the first two years of
bereavement from a variety of losses completed the Inventory of Complicated Grief
along with measures of the strength of their ongoing attachment to the deceased
and their capacity to make sense of the loss, find benefit in the experience, and
reconstruct a progressive sense of identity following the death. Several variables
concerning the survivor, his or her relationship to the deceased, and the
nature of the death functioned as risk factors for heightened distress, but their
role was generally moderated by meaning-making, often to the point of non-
significance. In contrast, higher levels of benefit-finding and positive identity
change were associated with lower levels of bereavement complication. Finally,
an interaction emerged between sense-making and ongoing attachment to the
deceased, suggesting that strong continuing bonds predicted greater levels of trau-
matic and especially separation distress, but only when the survivor was unable to
make sense of the loss in personal, practical, existential, or spiritual terms.
In the two decades that have elapsed since the Institute of
Medicine compiled its review of scientific research on bereave-
ment (Osterweis, Solomon, & Green, 1984), over 4,000 scientific
publications have addressed psychosocial issues arising at the
end of life and in the aftermath of loss (Neimeyer, 2004). The
robustness of this field of research is a function of several converg-
ing factors, including more extensive empirical assessment of
traditional models having relevance to bereavement, such as
attachment theory (Field, Gao, & Paderna, 2005); the emergence
Address correspondence to Robert A. Neimeyer, PhD, Department of Psychology,
The University of Memphis, Memphis, TN 38152. E-mail: email@example.com
Death Studies, 30: 715–738, 2006
Copyright # Taylor & Francis Group, LLC
ISSN: 0748-1187 print/1091-7683 online
of newer theories such as those emphasizing the role of active cop-
ing processes and meaning reconstruction in adaptation to loss
(Center for the Advancement of Health, 2004); and greatly
improved methods for assessing grief responses, especially in their
more complicated and problematic forms (Neimeyer & Hogan,
2001). The present project draws on developments in each of these
domains to evaluate the relationship between bereaved adults’
experience of a continuing bond with the deceased and their abil-
ity to attribute meaning or significance to the loss, on the one hand,
and their report of complicated grief symptomatology, on the
An Attachment Theory Perspective
Deriving originally from a broadly psychodynamic concern with
the impact of temporary separation from parents on the develop-
ing child, attachment theory ultimately evolved to include a focus
on the profound biopsychosocial transitions occasioned by the per-
manent separation from a loved one through death (Bowlby, 1980;
Parkes, 1996; Stroebe, 2002). Drawing on ethological, laboratory-
based, and social scientific research on the attachment behavioral
system, researchers in this tradition have delineated the dual func-
tions served by maintaining proximity to an attachment figure,
whether this figure is a parent in the case of a child, a partner in
the case of an adult, or any other figure who provides a safe haven
at times of threat, and a secure base from which to explore the
world (Field et al., 2005). Death of an attachment figure, of course,
represents the ultimate threat to this relationship, typically trigger-
ing a prolonged process of protest, despair, and reorganization as
the survivor attempts to adapt to the loss (Bowlby, 1980). Classi-
cally, this process was considered to require ‘‘working through’’
in the form of detachment of emotional investment in goals and
memories linked to the bond to the deceased, in order to permit
psychic and behavioral adaptation through investment in new
relationships (Freud, 1957).
In recent years, however, grief theorists and therapists have
embraced an alternative view that emphasizes the adaptive func-
tion of retaining bonds with the deceased, rather than relinquishing
them (Klass, Silverman, & Nickman, 1996). Drawing primarily
on qualitative research and cultural studies, advocates of this
716 R. A. Neimeyer et al.
continuing bonds perspective have focused attention on the fre-
quency with which bereaved persons report an ongoing engage-
ment with the memories and images of the deceased many
months or years after the loss, and the apparently salutary function
of maintaining this attachment (Attig, 2000; Klass, 1999). It is
worth noting that Bowlby (1980) himself was ambiguous in his
argument for whether detachment from or continued attachment
to the deceased was the more adaptive course for the bereaved,
ultimately acknowledging that change in the nature of the bond,
rather than its severance, per se, was the critical goal of grief
(Stroebe & Schut, 2005). In keeping with this latter view, contem-
porary theorists have emphasized how a constructive reorganiza-
tion rather than relinquishment of the bond can be achieved by
‘‘internalizing’’ the lost loved one as an extension of the self to
enhance affect regulation through maintaining psychological rather
than physical proximity to the attachment figure, taking him or her
as a role model, appreciating that individual’s unique legacy, or cul-
tivating a sense of the figure’s comforting presence at times of stress
(Field et al., 2005). Alternatively, reworking of the bond could be
construed in terms of retention of those ties to loved ones that do
not require their physical presence (e.g., continuing their life work),
coupled with developing new goals that have ‘‘substitute value’’ for
those that must be relinquished (Boerner & Heckhausen, 2003). In
either case, a refined attachment theory perspective and related
models that emphasize the postmortem relationship to the deceased
(Rubin, 1999) seem to accommodate the possibility that mainte-
nance of such bonds can play either a constructive or obstructive
role in loss adaptation, depending on their form, function, and tim-
ing (Stroebe & Schut, 2005).
Quantitative research to date underscores the complexity of
the relationship between continuing attachment to the deceased
and bereavement adaptation. For example, concrete expressions
of proximity seeking (e.g., hoping for comfort through contact
with a loved one’s possessions) has been associated with greater
grief-related distress, whereas more abstract forms of bonding
(e.g., through comforting memories) may be accompanied by less
anguish (Field, Nichols, Holen, & Horowitz, 1999). Moreover,
whereas coping through accessing the continuing bond with their
deceased husbands was predictive of overall negative mood for a
group of widows reporting their thoughts and feelings several times
Continuing Bonds 717
daily over a 2-week period, for those farther along in their bereave-
ment calling their husbands to mind was also associated with more
positive emotions. Such findings converge with research that indi-
cates that although most of the bereaved report a frequent sense of
the loved one’s continuing ‘‘presence’’ and most find this comfort-
ing, this is by no means uniformly the case (Datson & Marwit,
1997). Reviewing these and other relevant studies, Stroebe and
Schut (2005) read the evidence as inconclusive: ‘‘neither can we
categorically conclude that continuing, nor that relinquishing
bonds will be helpful to bereaved persons in coming to terms with
their grief’’ (p. 489). Instead, they suggested that ‘‘research efforts
need to be channeled toward establishing who among the
bereaved actually benefit from retaining versus relinquishing their
ties’’ (p. 490). Developments in another line of grief theory could
prove useful in pursuing this objective.
The Meaning Reconstruction Perspective
Part of the explanation for the ambiguous evidence concerning the
role of the continuing bond in bereavement adaptation might owe
less to the conceptual inadequacy of attachment theory than to its
incompleteness. That is, to the extent that other critical psychoso-
cial processes shape responses to losses—including loss of the
attachment bond—then bereavement might best be viewed
through a wider lens that brings into focus factors that interact with
ongoing attachment to predict post-loss adjustment. Recent
research on meaning reconstruction in the wake of loss suggests
that the effort to ‘‘make meaning’’ of the loss in practical, personal,
existential, or spiritual terms may be one such factor (Neimeyer,
A concern with meaning making in the aftermath of loss arises
from a cluster of related perspectives, including those associated with
assumptive world (Janoff-Bulman, 1989), cognitive (Thompson &
Janigan, 1988), coping (Folkman, 2001) and narrative (Neimeyer,
2005, 2006) approaches. A common theme in these models is that sig-
nificant loss disrupts the coherence of that matrix of personal mean-
ings by which individuals order their life experiences, calling for
activeattemptsto(a)makesenseof the loss, (b) findsomesortof‘‘silver
lining’’or benefitinthe experience,and(c) reorganizeone’sidentityas
survivor (Gillies & Neimeyer, 2006). In a constructivist view, inability
718 R. A. Neimeyer et al.
to assimilate such a psychosocial transition into the individual’s
‘‘self-narrative’’ instigates a frequently painful and protracted process
of accommodating or reconstructing central meaning structures
(Neimeyer, 2005). Successful revision of the meaning system is by
no means an assured outcome, posing the risk of clinical complica-
tions in the course of bereavement (Neimeyer, Prigerson, & Davies,
2002). Conversely, positive integration of the loss can bring about
the sort of progressive reconstruction of one’s self-narrative associated
with life-transforming ‘‘posttraumatic growth’’ (Calhoun & Tedeschi,
Evidence for the relevance of meaning making as a predictor
of bereavement outcome is accumulating as a result of several con-
verging research programs. For example, an unsuccessful struggle
to make sense of the death of a loved one predicts heightened
distress, particularly in the early months of bereavement, with
benefit-finding perhaps playing a larger role as time goes on (Davis,
Nolen-Hoeksema, & Larson, 1998). Similarly, an unresolved and
ruminative search for meaning tends to be associated with protrac-
ted, ‘‘chronic grief’’ trajectories (Bonanno, Wortman, & Nesse,
2004). Furthermore, the inability to make sense of the loss appears
to represent a near-perfect mediator of the impact of violent
bereavement (i.e., through suicide, homicide, and accident)
on complicated grief symptomatology (Currier, Holland, &
Neimeyer, 2006). Finally, identity change in bereavement has
been most explicitly studied in retrospective qualitative research,
documenting how the death of parents (Schultz, in press), siblings
(Davies, 1999) or children (Klass, 1999) can precipitate a signifi-
cant reorganization of one’s sense of self, for better or worse. Thus,
as research on perinatal bereavement suggests, a multidimensional
quest for meaning in loss appears to make an independent contri-
bution to the prediction of grief acuity beyond that associated with
security of attachment (Uren & Wastell, 2002).
The complementarity and compatibility of attachment and
meaning reconstruction perspectives is also congruent with both
clinical observations and theoretical considerations. The former
is reflected in a comparison of two young mothers whose infants
died of congenital heart problems (Neimeyer & Anderson, 2002),
both of whom reported a strong continuing attachment to their
children, who they viewed as in the hands of God. For one of
them, however, the effort to make sense of the inexplicable loss
Continuing Bonds 719
remained an ongoing challenge; as she said, ‘‘The only thought
that comes to me is that God must have really needed [my baby]
to be one of his angels in heaven, because she died so suddenly
and didn’t show any of the symptoms she suffered from that
caused her death.’’ Accordingly, she bitterly isolated herself from
her family and church community in a desperate struggle to find
some sense or hidden benefit in the tragic loss and the diminished
sense of self it introduced into her life. In contrast, the other
mother reported finding great meaning and personal growth in
her loss, noting that, ‘‘[my baby’s] purpose in life was to bring
us great joy and love in the three short days he was here on
earth .... I have greater appreciation for life and for healthy
babies. I have a greater knowledge of the heart disease that [he]
had. And I feel I can help others who have lost a child or who
will lose a child in the future. Spiritually, I know that when I
die, I will go to heaven, and we will be together again.’’ In short,
for these two mothers, whether or not the loss could be accorded
a more ample meaning in the context of their self-narratives
seemed to critically distinguish their bereavement trajectories,
despite their commonality in preserving a continuing bond with
their child. A similar struggle to simultaneously make meaning of
the loss and preserve a valued connection to an attachment figure
characterized the grief of a college student for his beloved uncle fol-
lowing his death by cancer (Neimeyer & Anderson, 2002). At a
theoretical level, the relevance of meaning reconstruction to the
reorganization of bonds following loss is implied by the way in
which attachment patterns shape ‘‘internal working models’’ of
the self and other(s) (Bowlby, 1980; Parkes, 1996), models that have
to be reworked when psychosocial transitions cannot be accommo-
dated by the person’s existing meaning system (Guidano & Liotti,
The purpose of this study therefore was to assess the possible
interaction between continuing bonds coping on the one hand,
and sense-making, benefit finding, and identity change on the
other, in predicting complicated grief symptomatology (Prigerson
& Maciejewski, 2006) in a large cohort of bereaved young
adults. A further advantage of the study was its use of an exten-
sively validated measure of complicated grief (Prigerson & Jacobs,
2001), as distinct from the psychometrically weaker or unvalidated
720 R. A. Neimeyer et al.
TABLE 1 Demographic and Background Information (n ¼ 506)
Background Demographic information
Age of the bereaved M ¼ 20.52, SD ¼ 3.83,
Range ¼ 17–53
Gender of the bereaved 119 men (23.5%)
387 women (76.5%)
Ethnicity of the bereaved 293 Caucasian (57.9%)
186 African-American (36.9%)
25 Other (5%)
Relationship of decedent 25 Parents (5%)
10 Sibling (2%)
2 Children (.4%)
4 Partners=spouses (.8%)
105 Grandparents (20.8%)
110 Aunts=uncles=cousins (21.7%)
140 Friends (27.7%)
33 Other relatives (6.5%)
Age of the deceased M ¼ 52.11 years, SD ¼ 26.43 years,
Range ¼ 4 months–102 years
Amount of contact with the deceased
in the 3 months preceding death
M ¼ 3.38, SD ¼ 1.72, Range ¼ 1–6
Closeness of the relationship with the
deceased prior to death
M ¼ 3.65, SD ¼ 0.84, Range ¼ 1–5
Cause of death 222 Natural, anticipated (43.9%)
106 Natural, sudden (20.9%)
91 Accidental (18%)
27 Suicide (5.3%
38 Homicide (7.5%)
22 Other (4.3%)
Number of months ago
M ¼ 11.82, SD ¼ 7.59,
Range ¼ 0.25–24
Other major losses prior to Yes: 300 (59.3%)
this death No: 202 (39.9%)
Prior to the death, ever received Yes: 64 (12.6%)
mental health counseling No: 426 (84.2%)
Since the death, received Yes: 51 (10.1%)
mental health counseling No: 435 (86%)
Support available Yes: 429 (84.8%)
No: 56 (11.1%)
Time spent talking about the loss M ¼ 2.25, SD ¼ 1.13, Range ¼ 1–4
Sense making M ¼ 3.25, SD ¼ 0.94, Range ¼ 1–4
Benefit finding M ¼ 2.62, SD ¼ 1.31, Range ¼ 1–5
Amount of identity change M ¼ 2.40, SD ¼ 1.30, Range ¼ 1–5
Continuing Bonds 721
measures of general grief that have characterized previous research
on smaller samples.
Following institutional review of the project, 506 participants were
recruited from undergraduate introductory psychology courses at
the University of Memphis over a 3-year period. As a criterion of
eligibility, each participant reported having experienced the loss
of a friend or loved one through death within the past 2 years, in
keeping with studies that suggest that significant bereavement
phenomena can be observed over 24 months or longer in many
cases (Prigerson & Jacobs, 2001). The sample ranged in age from
18 to 53 years old with a mean age of 21 years (SD ¼ 3.83);
76.5% were female and 23.5 % were male; 57.9% were Caucasian,
36.9% African American, and 5% were of another ethnicity,
reflecting the undergraduate distribution of ethnicities at the urban
research institution. Other demographic characteristics of the
participants and descriptions of their losses are presented in Table 1.
Each eligible participant completed a one-time battery of instru-
ments that included demographic items, the Inventory of Compli-
cated Grief (ICG), the Continuing Bonds Scale (CBS), questions
about meaning reconstruction, and questions about factors
surrounding the loss and his or her relationship to the decedent.
TABLE 1 Continued
Background Demographic information
Positive identity change M ¼ 2.18, SD ¼ 0.66, Range ¼ 1–3
Continuing bonds M ¼ 26.87, SD ¼ 10.26,
Range ¼ 11–55
ICG-Traumatic Distress M ¼ 9.65, SD ¼ 4.29, Range ¼ 5–25
ICG-Separation Distress M ¼ 18.38, SD ¼ 7.86, Range ¼ 11–51
Note. ICG ¼ Inventory of Complicated Grief.
722 R. A. Neimeyer et al.
INVENTORY OF COMPLICATED GRIEF (ICG)
(Prigerson & Jacobs, 2001; Prigerson, Maciejewski, Reynolds,
Beirhals, & al., 1995). The ICG is composed of 30 declarative state-
ments, such as ‘‘I feel like I have become numb since the death of
[the deceased]’’ and ‘‘Ever since [the deceased] died I feel like I
have lost the ability to care about other people or I feel distant
from people I care about,’’ to which responses are made on a 5-
point Likert-type scale describing the frequency of symptoms
(i.e., from 1 ¼ never to 5 ¼ always). Items assess two theoretically
distinguishable features of complicated grief, focusing on both
symptoms of separation distress (e.g., yearning, loneliness, pre-
occupation with thoughts of the deceased) and traumatic distress
(e.g., disbelief about the death; feelings of being shocked and over-
whelmed, numb and out of control; disturbances in sleep). Both
clusters of items were retained as distinct dependent variables in
the analyses that follow given the potentially greater relevance of
continuing bonds coping to separation distress symptomatology
and meaning reconstruction to traumatic symptomatology.
The first 19-item version of the ICG displayed strong
psychometric properties in a number of studies. For example,
this earlier version of the ICG (Prigerson et al., 1995) exhibited
6-month test–retest reliability of r ¼ .80 , and good concurrent
validity with the Texas Re vised Inventory of Grief (TRIG;
Faschingbauer, 1981). The expanded version of t he ICG
(Prigerson & Jacobs, 2001), which was used in the present study,
has also shown solid psychometric values. For example, the
newer version was tested in the Netherlands, where a Dutch
translatio n displayed good temporal stability (.92) over a period
ranging from 9 to 28 days (Boelen, Van den Bout, De Keijser,
& Hoitjin k, 2003). Research has a lso provided considerable sup-
port for the scale’s validity. For example, diagnoses based on the
newer ICG were found to have a sensitivity of .93 and a speci-
ficit y of .93 in the detection of intervie w-determined complicated
grief (Barry, Kasl, & Prigerson, 2002). In addition, this scale has
predicted a broa d range of serious long-term health and mental
health consequences of bereavement (Hardison, Neimeyer, &
Lichstein, 2005; Ott, 2003; Prigerson et al., 1999; Silverman,
Johnson, & Prigerson, 2001).
Continuing Bonds 723
CONTINUING BONDS SCALE (CBS)
The CBS (Field, Gal-Oz, & Bonanno, 2003) consists of 11 ques-
tions that explore to what extent the bereaved person feels the
deceased loved one remains a part of his or her life. Responses to
questions such as ‘‘I seek out things to remind me of
’’ and ‘‘I
have inner conversations with my spouse where I turn to him or
her for comfort or advice’’ are rated using a Likert-type scale that
ranges from 1 (not true at all)to5(very true). The CBS has shown
good internal consistency (a ¼ .87) and is positively related to rat-
ings of satisfaction with the relationship and negatively related to
the degree of blame in a role-played conversation with the deceased
(Field et al., 2003). In the present study we subjected the CBS to a
principal components analysis. A scree plot suggested extracting
one factor, which included all 11 items and had high internal con-
sistency (a ¼ .90). Thus, we summed the 11 items for a total score.
Meaning reconstruction was assessed by four items. Sense-
making was assessed by the question, ‘‘How much sense would
you say you have made of the loss?,’’ with anchor points of 1 (no
sense)to4(a good deal of sense); whereas benefit finding was mea-
sured by the question, ‘‘Despite the loss, have you been able to find
any benefit from your experience of the loss?,’’ with anchors of 1 (no bene-
fit)to5(great benefit). Finally, identity change was assessed by ask-
ing, ‘‘Do you feel that you are different, or that your sense of
identity has changed, as a result of this loss?,’’ with anchors of 1
(no different)to5(very different), with an additional item simply
assessing the direction of this change (1: for the better,2:mixed,or
3: for the worse). These questions correspond closely to the single-
item questions that other researchers have used to measure these
construals of meaning for quantitative analyses (e.g., Bonanno
et al., 2003; Davis et al., 1998; Lehman et al., 1987; McIntosh
et al., 1993). As with these other investigations, these single items
showed considerable utility in an earlier study in both differentiat-
ing violent and natural loss survivors and predicting complicated
grief (Currier et al., 2006).
OTHER VARIABLES MEASURED
Basic demographic information, such as age, gender, and
ethnicity, was gathered for each participant. In addition, the
724 R. A. Neimeyer et al.
participants provided information concerning their loss. The part-
icipants indicated their relationship to the deceased (e.g., father,
grandmother, cousin), how old the deceased was when he or she
died, the amount of contact they had had in the three months prior
to the death, intimacy level of the relationship, cause of death, how
many months ago the death occurred, the amount of time they had
spent talking about the loss, and whether or not they had (a)
experienced other major losses, (b) received mental health coun-
seling prior to the loss, (c) received mental health counseling since
the loss, and (d) felt that there was a supportive person with whom
they could talk.
We used hierarchical regression models to examine the relation-
ship between continuing bonds, meaning reconstruction, and grief
outcomes. The regression models for both separation and trau-
matic distress consisted of three steps. We were interested in
how well continuing bonds and meaning reconstruction predicted
grief outcomes above and beyond the demographic characteristics
of the respondent and the characteristics of the death. Thus, the first
step included only the demographic characteristics of the respon-
dent. The second step included the characteristics of the death.
Finally, the last step included the CBS, the meaning reconstruction
items, and the interaction between the two, as well as between each
of them and other statistically significant predictors of grief out-
comes from the first two steps.
The complete results of the separation distress analysis can be seen
in Table 2. The demographic characteristics of the respondent
were entered in the first step, which did not account for a statisti-
cally significant amount of variance in this aspect of complicated
¼ .023, p ¼ .061). Of the demographic variables, only
the African American variable had a statistically significant relation
to separation distress; on average, African Americans reported
more such symptomatology than Caucasians.
Continuing Bonds 725
TABLE 2 Predictors of Separation Distress
Model 1 Model 2 Model 3
Variable B SE B b BSEB b BSEB b
Intercept 1.98 .26 1.30 .34 2.32 .32
Age of bereaved .008 .01 .04 .02 .01 .10
.004 .008 .02
Gender of bereaved .04 .11 .02 .02 .09 .01 .07 .08 .04
White .009 .09 .06 .04 .08 .03 .08 .06 .06
African American .27 .09 .17
.21 .09 .13
.21 .07 .13
Age of deceased .001 .002 .03 .002 .002 .06
Immediate family .59 .14 .20
.09 .14 .03
Contact .13 .03 .25
.04 .02 .08
Murder=suicide .16 .13 .06 .08 .11 .03
Anticipated loss .07 .09 .04 .12 .07 .07
Months since death .06 .01 .05 .08 .004 .07
Intimate relationship .14 .06 .13
.02 .05 .02
Other losses .14 .08 .08 .08 .06 .04
Counseling prior to loss .10 .12 .04 .05 .10 .02
Counseling since loss .34 .15 .11
.16 .12 .05
Someone to talk to .27 .13 .10
.08 .10 .03
Time spent talking about loss .13 .04 .17
.01 .03 .02
Continuing bond .38 .04 .42
Sense making .19 .04 .21
Benefit finding .07 .03 .10
Amount of identity change .11 .03 .17
Positive identity change .12 .05 .09
CB Sense making .11 .04 .11
CB Immediate family .23 .09 .13
p < .01.
p < .05.
Characteristics of the death and support variables were
included in the second step and accounted for a significant amount
of variance in separation distress (R
¼ .266, p < .001). Six of the
twelve variables entered in Step 2 were statistically significant.
Family relationships, amount of contact with the deceased in the
months preceding death, level of intimacy with the deceased,
counseling for an emotional problem since the loss, presence of
a supportive person, and amount of time spent talking about the
loss all had a positive relationship with separation distress. Being
African American continued to have a positive relationship with
separation distress, as did being of younger age. Length of time
since the loss was noticeably unrelated to separation distress in
Step 3 was the final model and included continuing bonds
coping, the meaning reconstruction items, and interactions. These
variables accounted for a significant amount of variance in separ-
ation distress (R
¼ .276, p < .001, overall R
¼ .567). Amount of
identity reconstruction had a statistically significant positive
relationship with separation distress; the greater the identity
disruption experienced by the bereaved, the more painful the sep-
aration. In contrast, benefit finding and positive identity change
were associated with significantly less separation distress.
In addition to these simple main effects there were two statisti-
cally significant interactions. Figure 1 provides a graphic depiction
of the continuing bonds by sense-making interaction. When
amount of sense made was low, stronger continuing bonds were
associated with more separation distress. However, as the amount
of sense-making increased, the strength of the continuing bond
made less of a difference in separation symptoms. The second sig-
nificant interaction concerned continuing bonds by relationship to
the deceased. When the deceased was a member of the bereaved
individual’s immediate family, strong continuing bonds were asso-
ciated with higher levels of separation stress than weak bonds.
However, when the deceased was not a member of the immediate
family, the strength of the continued bond made less difference.
Thus sense-making and the relationship to the deceased moder-
ated the relationship between continuing bonds and separation
Once we controlled for continuing bonds, meaning recon-
struction and the interactions, only one of the variables that were
Continuing Bonds 727
predictors of separation distress in Steps 1 and 2 remained signifi-
cant. Specifically, African Americans continued to manifest greater
The complete results of the traumatic distress analysis are pre-
sented in Table 3. Demographic characteristics of the respondent
were entered in the first step but did not account for a statistically
significant amount of variance in traumatic distress (R
p ¼ .062). Of the demographic variables, only age of the respon-
dent had a statistically significant negative relationship with trau-
matic distress, suggesting that younger survivors were somewhat
more likely to respond to loss with greater subjective trauma.
Characteristics of the death and support variables were
included in the second step and accounted for a statistically signifi-
cant amount of variance in traumatic distress (R
¼ .293, p < .001).
Of the 12 new variables entered in Step 2, 9 were significant. Three
of those variables had a negative relationship with traumatic dis-
tress: age of the deceased, number of months since the death, and
availability of someone with whom the respondent was able to talk
about the loss. The other 6 significant variables had a positive
relationship with traumatic distress: loss of a member of the
FIGURE 1 Continuing bond by sense making Interaction–Separation Distress.
728 R. A. Neimeyer et al.
TABLE 3 Predictors of Traumatic Distress
Step 1 Step 2 Step 3
Variable B SE B b BSEB b BSEB b
Intercept 1.85 .20 1.62 .56 2.22 .25
Age of bereaved .02 .09 .11
.02 .01 .15
.01 .01 .06
Gender of bereaved .08 .08 .05 .02 .07 .02 .02 .06 .01
White .03 .07 .03 .05 .06 .05 .02 .05 .02
African American .11 .07 .09 .10 .07 .08 .12 .05 .10
Age of deceased .004 .001 .15
.004 .001 .14
Immediate family .309 .11 .14
.14 .11 .06
Contact .08 .02 .21
.03 .02 .09
Murder=suicide .29 .10 .14
.05 .09 .03
Anticipated loss .02 .07 .02 .02 .06 .02
Months since death .01 .004 .14
.01 .003 .14
Intimate relationship .08 .04 .09 .03 .04 .04
Other losses .11 .06 .08 .03 .05 .03
Counseling prior to loss .28 .09 .14
.21 .08 .11
Counseling since loss .27 .11 .12
.16 .09 .07
Someone to talk to .19 .10 .09
.03 .08 .01
Time spent talking about loss .07 .03 .13
.01 .03 .03
Continuing bond .19 .03 .28
Sense making .17 .04 .24
Benefit finding .06 .02 .11
Amount of identity change .07 .02 .14
TABLE 3 Continued
Step 1 Step 2 Step 3
Variable B SE B b BSEB b BSEB b
Positive identity change .10 .04 .10
CB Sense making .18 .03 .10
CB Immediate family .07 .07 .13
Sense Treatment prior to loss .18 .07 .13
Sense Age of deceased .20 .03 .08
p < .01.
p < .05.
immediate family, the amount of contact with the deceased, death
through murder or suicide, history of counseling, and the amount
of time spent talking about the loss. In summary, level of traumatic
distress increased for deaths that were premature, recent, violent,
involved immediate family or frequent interaction, or left the
bereaved unsupported, especially if he or she had a history of men-
tal health consultation. Age of the respondent continued to have a
statistically significant negative relationship with traumatic distress.
Step 3 was the final model and included continuing bonds, the
meaning reconstruction items, and the interactions. These vari-
ables accounted for a significant amount of variance in traumatic
¼ .243, p < .001, overall R
¼ .559). Amount of ident-
ity reconstruction had a statistically significant positive relationship
with traumatic distress; the greater the identity change triggered by
the loss, the greater the traumatic disruption. In contrast, benefit
finding and positive identity change had a statistically significant
negative relationship with traumatic distress, suggesting that they
functioned as protective factors against traumatic symptoms as
they did against separation distress.
In addition to these simple main effects, there were four stat-
istically significant interactions. Two of these followed the same
pattern as occurred with separation distress, such that higher levels
of sense making and a non-familial relationship to the deceased
moderated the impact of a high continuing bond. Sense-making
also interacted with prior counseling such that when sense-making
was low, counseling prior to the loss was associated with higher
levels of traumatic distress, whereas under conditions of high
sense-making, whether or the not the person had previously sought
counseling made little difference. Similarly, when sense-making
was low, survivors reported greater traumatic distress in the case
of death of younger rather than older loved ones. However, as
sense-making increased, the difference attributable to decedent
age diminished greatly.
Once we controlled for continuing bonds, meaning recon-
struction, and the interactions, many of the variables that predicted
complicated grief symptoms in Steps 1 and 2 were no longer sig-
nificant. Specifically, age of the respondent, age of the deceased,
the nature and intimacy of the relationship, death by murder or
suicide, subsequent counseling, availability of support, and the
amount of time spent talking about the loss all no longer had a
Continuing Bonds 731
statistically significant relationship with traumatic distress. How-
ever, some variables remained significant: amount of prior contact
with the deceased, number of months since the death, and prior
counseling all continued to predict traumatic impact. Interestingly,
with meaning reconstruction and bonding controlled, ethnicity
emerged as significant, with African Americans showing greater
traumatic distress than Caucasians.
The results of the study indicate that even after controlling for the
demographic characteristics of the bereaved and several factors
concerning the character of the relationship to the deceased and
the loss itself, continuing bonds and meaning reconstruction were
reliable predictors of both traumatic and separation distress.
Specifically, high levels of meaning making consistently predicted
better grief outcomes during the first two years of bereavement
for this young adult sample, whereas high levels of post-loss attach-
ment to the deceased were associated with more complicated grief
symptoms, at least under conditions of low sense-making. It is
worth noting, however, that continuing bonds coping was more
highly related to separation (b ¼ .42) than to traumatic (b ¼ .28)
distress, whereas sense-making had approximately equal relevance
to both sets of symptoms. The unique contributions of each of these
factors to the prediction of complicated grief symptomatology rein-
forces the relevance of both attachment theory and constructivist,
meaning-oriented theories to conceptualizing chronic and anguish-
ing debilitation in response to bereavement (Neimeyer, 2006a).
Moreover, these relationships were found using a measure of
grief symptoms with strong construct and predictive validity for
poor health and mental health outcomes, strengthening previous lit-
eratures on both continuing bonds and meaning making that have
relied on psychometrically weaker measures of normal grief.
The fact that strong postmortem attachment to the deceased
under conditions of low sense-making was associated with higher,
rather than lower grief-related distress challenges the assumption
that continuing bonds necessarily mitigate the stress of bereave-
ment (Attig, 2000; Klass et al., 1996). Such evidence is consistent
with at least some previous research that evidences a similar
pattern among widows in the early months of their loss, when
732 R. A. Neimeyer et al.
self-reported contact with the memory of their husbands was
related to higher levels of negative emotion (Field & Friedrichs,
2004). However, our findings leave open the possibility that with
greater time to process and integrate the loss beyond the two years
of bereavement assessed in the present study, survivors might
become able to draw upon a transformed bond to their loved
one in a comforting way, as did the widows in the Field and
Friedrichs study who were more than 2 years into their bereave-
ment. Still, it is worth emphasizing that the passage of time alone
was unrelated to the amelioration of symptoms of separation
distress in our study, cautioning against the assumption that ‘‘time
heals all wounds.’’ Instead, it seemed that it was the ability to find
some form of benefit or ‘‘silver lining’’ in the loss, to experience a
progressive rather than regressive transformation in one’s identity,
and especially to make sense of the loss in personal, practical, exis-
tential, or spiritual terms that predicted more positive grief out-
comes. Indeed, sense-making in particular moderated the effect
of continuing attachment on grief, suggesting that it is those
bereaved persons who remain closely bonded to their loved ones
but who are unable to integrate the loss into a more ample system
of personal meaning who are at greatest risk for bereavement
Results of the study also point to other factors that predict
more difficult adaptation to loss. For example, several variables
associated with the intimacy and nature of relationship to the
deceased and use of mental health services by the bereaved were
associated with greater separation distress, and these same factors
as well as others concerning recency of the loss and prematurity
and violence of the death predicted more traumatic symptoma-
tology. These findings accord with the broader literature on factors
associated with bereavement complications (Currier et al., 2006;
Gamino, Sewell, & Easterling, 1998; Hardison et al., 2005), under-
scoring their clinical relevance as observable markers of bereave-
ment risk. Importantly, however, the predictive power of the
majority of these factors was reduced by sense-making, often to
the point of non-significance. Therapeutically, this suggests that
narrative strategies that promote meaning making regarding loss
(Neimeyer, 1999, 2001a) can mitigate bereavement complications
for high-risk mourners, as evidenced by a recent randomized
controlled trial of therapy that relies on systematically telling and
Continuing Bonds 733
retelling the story of the loss in the context of reviewing and revis-
ing life goals (Shear, Frank, Houch, & Reynolds, 2005). In terms of
research, these same findings hint at the relevance of developing
more finely-tuned operational definitions of meaning-making pro-
cesses beyond the few straightforward items used in this and earlier
studies (Gillies & Neimeyer, 2006). Fortunately, the larger con-
structivist literature offers an expansive toolbox of methods for
eliciting, coding, and analyzing the features of life narratives that
are implicated in transformation of ‘‘working models’’ of self and
world triggered by substantial loss (Neimeyer, 2006b).
The one demographic factor associated with elevated risk of
complicated grief even when meaning making and continuing
bonds were taken into account was ethnicity: African Americans
were more likely than their Caucasian counterparts to suffer both
greater separation and traumatic distress. Unfortunately, very little
empirical work has examined issues of ethnicity, although a recent
qualitative study points to distinctive stressors and resources for
African American mourners as a function of a history of social
oppression and marginalization (Rosenblatt & Wallace, 2005).
The results of the present study suggest that more empirical atten-
tion needs to be given to this subject.
Finally, the strengths of the study associated with its theoreti-
cal grounding, instrumentation, large sample and diversity of losses
notwithstanding, future researchers should heed its limitations as
well. For example, the fact that participants represented a large
cohort of college students places constraints on generalization of
these findings to other groups, insofar as these participants could
be characterized by developmental issues unique to young adult-
hood, such as differentiation and individuation from family,
achievement motivation and development of a broader social net-
work (Balk & Vesta, 1998). In the context of our research, these
developmental factors could have stressed or intensified bonds
with the deceased and living support figures or influenced pro-
cesses of meaning making in ways that cannot be revealed by
Perhaps even more important are constraints on causal reason-
ing stemming from the correlational nature of the study. For
example, although it might be tempting to conclude that reliance
upon continuing bond coping or success in meaning reconstruction
influenced the course of grieving, it is also possible that more acute
734 R. A. Neimeyer et al.
separation or traumatic distress stimulated a search for solace in the
attachment to the deceased or attempts to make sense of the loss.
Alternatively, grief-related distress and attempts to seek security
and significance could influence one another in an ongoing iterative
process (Gillies & Neimeyer, 2006). Only longitudinal and experi-
mental designs that trace these several components of bereavement
adaptation can resolve these thorny questions.
Although the nature of the attachment bond between the bereaved
and the deceased has been a focus of scientific theory for nearly a
century, serious empirical attention to its role in bereavement
adaptation is of recent advent. The present study makes a contri-
bution to this evolving research by documenting the greater separ-
ation and traumatic distress experienced by bereaved young adults
who report more intense ongoing bonds with those they have lost.
Importantly, however, continuing bonds appear to interact with
meaning making in response to loss, such that those survivors
who are able to make sense of the loss in personally meaningful
terms experience fewer symptoms of complicated grief. Moreover,
the various component processes of meaning reconstruction (sense
making, benefit finding, and progressive identity change) seem to
mitigate the impact of other risk factors focusing on the character-
istics of the bereaved individual, the relationship to the deceased,
and the death itself. Thus, there appears to be ample conceptual,
empirical, and clinical warrant for further research on the
reorganization of the continuing bond, and the larger processes
of meaning reconstruction of which this is a part.
Attig, T. (2000). The heart of grief. New York: Oxford.
Balk, D. E. & Vesta, L. C. (1998). Psychological development during four years of
bereavement: A longitudinal case study. Death Studies, 22, 23–41.
Barry, L., Kasl, S., & Prigerson, H. (2002). Psychiatric disorders among bereaved
persons: The role of perceived circumstances of death and preparedness for
death. American Journal of Geriatric Psychiatry, 10, 447–457.
Boelen, P., Van den Bout, J., De Keijser, J., & Hoitjink, H. (2003). Reliability and
validity of the Dutch version of the Inventory of Traumatic Grief (ITG).
Death Studies, 27, 227–247.
Continuing Bonds 735
Boerner, K. & Heckhausen, J. (2003). To have and have not: Adaptive bereave-
ment by transforming mental ties to the deceased. Death Studies, 27, 199–226.
Bonanno, G. A., Wortman, C. B., & Nesse, R. M. (2004). Prospective patterns of
resilience and maladjustment during widowhood. Psychology and Aging, 19,
Bowlby, J. (1980). Attachment and loss: Loss, sadness and depression (Vol. 3). New
Calhoun, L. & Tedeschi, R. G. (Eds.). (2006). Handbook of posttraumatic growth.
Mahwah, NJ: Lawrence Erlbaum.
Center for the Advancement of Health. (2004). Report on bereavement and grief
research. Death Studies, 28, 489–575.
Currier, J., Holland, J., Coleman, R., & Neimeyer, R. A. (2006). Bereavement
following violent death: An assault on life and meaning. In R. Stevenson
& G. Cox (Eds.), Perspectives on violence and violent death. Amityville, NY:
Currier, J., Holland, J., & Neimeyer, R. A. (2006). Sense making, grief and the
experience of violent loss: Toward a mediational model. Death Studies, 30,
Datson, S. L. & Marwit, S. J. (1997). Personality constructs and perceived pres-
ence of deceased loved ones. Death Studies, 21, 131–146.
Davies, B. (1999). Shadows in the sun: The experiences of sibling bereavement in child-
hood. Philadelphia: Brunner=Mazel.
Davis, C. G., Nolen-Hoeksema, S., & Larson, J. (1998). Making sense of loss and
benefiting from the experience: Two construals of meaning. Journal of Person-
ality and Social Psychology, 75, 561–574.
Faschingbauer, T. R. (1981). Texas revised inventory of grief manual. Houston, TX:
Field, N. P. & Friedrichs, M. (2004). Continuing bonds in coping with the death of
a husband. Death Studies, 28, 597–620.
Field, N. P., Gal-Oz, E., & Bonanno, G. A. (2003). Continuing bonds and adjust-
ment at 5 years after the death of a spouse. Journal of Consulting and Clinical
Psychology, 71, 110–117.
Field, N. P., Gao, B., & Paderna, L. (2005). Continuing bonds in bereavement: An
attachment theory based perspective. Death Studies, 29, 277–299.
Field, N. P., Nichols, C., Holen, A., & Horowitz, M. J. (1999). The relation of
continuing attachment to adjustment in conjugal bereavement. Journal of
Consulting and Clinical Psychology, 67, 212–218.
Folkman, S. F. (2001). Revised coping theory and the process of bereavement. In
M. Stroebe, R. Hansson, W. Stroebe, & H. Schut (Eds.), Handbook of bereave-
ment research. Washington, DC: American Psychological Association.
Freud, S. (1957). Mourning and melancholia. In J. Strachey (Ed.), The complete psycho-
logical works of sigmund freud (pp. 152–170). London, England: Hogarth Press.
Gamino, L. A., Sewell, K. W., & Easterling, L. W. (1998). Scott & white grief
study: An emprical test of predictors of intensified mourning. Death Studies,
736 R. A. Neimeyer et al.
Gillies, J. & Neimeyer, R. A. (2006). Loss, grief and the search for significance:
Toward a model of meaning reconstruction in bereavement. Journal of Con-
structivist Psychology, 19, 31–65.
Guidano, V. F. & Liotti, G. (1983). Cognitive processes and emotional disorders. New
Hardison, H. G., Neimeyer, R. A., & Lichstein, K. L. (2005). Insomnia and com-
plicated grief symptoms in bereaved college students. Behavioral Sleep Medi-
cine, 3, 99–111.
Janoff-Bulman, R. (1989). Assumptive worlds and the stress of traumatic events.
Social Cognition, 7, 113–116.
Klass, D. (1999). The spiritual lives of bereaved parents . Philadelphia: Brunner Mazel.
Klass, D., Silverman, P. R., & Nickman, S. (1996). Continuing bonds: New under-
standings of grief. Washington: Taylor & Francis.
Lehman, D. R., Wortman, C. B., & Williams, A. F. (1987). Long-term effects of
losing a spouse or child in a motor vehicle crash. Journal of Personality and
Social Psychology, 52, 218–231.
McIntosh, D. N., Silver, R. C., & Wortman, C. B. (1993). Religion’s role in adjust-
ment to a negative life event: Coping with the loss of a child. Journal of
Personality and Social Psychology, 65, 812–821.
Neimeyer, R. A. (1999). Narrative strategies in grief therapy. Journal of Constructi-
vist Psychology, 12, 65–85.
Neimeyer, R. A. (2001a). Reauthoring life narratives: Grief therapy as meaning
reconstruction. Israel Journal of Psychiatry, 38, 171–183.
Neimeyer, R. A. (Ed.). (2001b). Meaning reconstruction and the experience of loss.
Washington, DC: American Psychological Association.
Neimeyer, R. A. (2004). Research on grief and bereavement: Evolution and
revolution. Death Studies, 28, 489–490.
Neimeyer, R. A. (2005). Widowhood, grief and the quest for meaning: A narrative
perspective on resilience. In D. Carr, R. M. Nesse, & C. B. Wortman (Eds.),
Late life widowhood in the united states. New York: Springer.
Neimeyer, R. A. (2006a). Complicated grief and the quest for meaning: A con-
structivist contribution. Omega, 52, 37–52.
Neimeyer, R. A. (2006b). Re-storying loss: Fostering growth in the posttraumatic
narrative. In L. Calhoun & R. G. Tedeschi (Eds.), Handbook of posttraumatic
growth: Research and practice. Mahwah, NJ: Lawrence Erlbaum.
Neimeyer, R. A. & Anderson, A. (2002). Meaning reconstruction theory. In
N. Thompson (Ed.), Loss and grief (pp. 45–64). New York: Palgrave.
Neimeyer, R. A. & Hogan, N. (2001). Quantitative or qualitative? Measurement
issues in the study of grief. In H. S. M. Stroebe, R. Hansson, & W. Stoebe
(Ed.), Handbook of bereavement research (pp. 89–118). Washington, DC:
American Psychological Association.
Neimeyer, R. A., Prigerson, H., & Davies, B. (2002). Mourning and meaning.
American Behavioral Scientist, 46, 235–251.
Osterweis, M., Solomon, F., & Green, M. (Eds.). (1984). Bereavement. Washington,
DC: National Academy Press.
Continuing Bonds 737
Ott, C. H. (2003). The impact of complicated grief on mental and physical
health at various points in the bereavement process. Death Studies, 27,
Parkes, C. M. (1996). Bereavement (2nd ed.). London & New York: Routledge.
Prigerson, H. G., Bridge, J., Maciejewski, P. K., Beery, L. C., Rosenheck, R. A., &
Jacobs, S. C., et al. (1999). Influence of traumatic grief on suicidal ideation
among young adults. American Journal of Psychiatry, 156, 1994–1995.
Prigerson, H. G. & Jacobs, S. C. (2001). Diagnostic criteria for traumatic grief. In
M. S. Stroebe, R. O. Hansson, W. Stroebe, & H. Schut (Eds.), Handbook of
bereavement research (pp. 614–646). Washington, DC: American Psychological
Prigerson, H. G. & Maciejewski, P. K. (2006). A call for sound empirical testing
and evaluation of criteria for complicated grief proposed by the DSM-V.
Omega, 52, 9–19.
Prigerson, H. G., Maciejewski, P., Reynolds, C. F., Beirhals, A. J., & al., e. (1995).
Inventory of complicated grief: A scale to measure maladptive symptoms of
loss. Psychiatry Research, 59, 65–79.
Rosenblatt, P. & Wallace, B. (2005). African American grief. New York: Routledge.
Rubin, S. (1999). The two-track model of bereavement: Overview, retrospect and
prospect. Death Studies, 23, 681–714.
Schultz, L. (in press). The influence of maternal loss on young women’s experi-
ence of identity development in emerging adulthood. Death Studies, 31.
Shear, K., Frank, E., Houch, P. R., & Reynolds, C. F. (2005). Treatment of com-
plicated grief: A randomized controlled trial. Journal of the American Medical
Association, 293, 2601–2608.
Silverman, G. K., Johnson, J. G., & Prigerson, H. G. (2001). Preliminary explora-
tions of the effects of prior trauma and loss on risk for psychiatric disorders in
recently widowed people. Israel Journal of Psychiatry, 38, 202–215.
Stroebe, M. & Schut, H. (2005). To continue or relinquish bonds: A review of con-
sequences for the bereaved. Death Studies, 29, 477–494.
Stroebe, M. S. (2002). Paving the way: From early attachment theory to contem-
porary bereavement research. Mortality, 7, 127–138.
Thompson, S. & Janigan, A. (1988). Life schemes: A framework for understanding
the search for meaning. Journal of Social and Clinical Psychology, 7, 260–280.
Uren, T. H. & Wastell, C. A. (2002). Attachment and meaning making in perina-
tal bereavement. Death Studies, 26, 279–308.
738 R. A. Neimeyer et al.