Prospective Risk Factors for Intense Grief in Family Members of Veterans Who Died of Terminal Illness

Abstract

Many bereavement researchers focus on predicting and preventing complicated grief, a psychologically crippling, sometimes life-threatening response to loss that persists for lengthy periods, often with serious health consequences. Reviews of studies have identified specific risk factors (e.g., low social support, insecure attachment style) that predict high levels of complicated grief symptomatology. However, studies rarely investigate multiple risk factors in combination, and still more rarely trace factors observable during the end-of-life period and their predictive power for identifying intense grief in family members following the death. We therefore investigated several pre-loss risk factors for post-loss bereavement distress in 35 family members of Veterans who died of a terminal illness after receiving palliative care. Results revealed that being female, Caucasian, losing a spouse, and experiencing high anticipatory grief prior to the death, all predicted high levels of grief 6 to 10 weeks following the death. Moreover, psychosocial factors such as being highly dependent upon the Veteran, displaying high neuroticism, reporting low levels of social support, and being unable to make sense of the prognosis or death predicted more intense post-loss grief reactions.

Keywords

complicated grief risk factors palliative care bereavement meaning making

Complicated grief (CG; Shear et al., 2011; also known as Prolonged Grief Disorder; PGD; Prigerson et al., 2009; or Persistent Complex Bereavement Disorder¹; American Psychiatric Association, 2013) is a severe, protracted, debilitating, and sometimes life-threatening response to loss that has serious physical and mental health consequences (Lichtenthal, Cruess, & Prigerson, 2004). Predicting and preventing CG is a major focus of contemporary bereavement research. Burke and Neimeyer (2013) reviewed 43 studies of potential prospective predictors of intense grief, identifying several confirmed risk factors (e.g., marital dependence, high neuroticism, low social support) that have increased our understanding as to why some bereaved people struggle psychologically more than others. However, as only a handful of studies have investigated the possible role of multiple risk factors in the same study (e.g., Thomas, Hudson, Trauer, Remedios, & Clarke, 2014), more research is clearly called for to understand their relative priority in forecasting bereavement complications. Likewise, only a few investigations have linked pre-loss assessment with post-loss functioning of family members (e.g., Boerner, Wortman, & Bonanno, 2005) in a way that might provide guidance for end-of-life (EOL) care professionals attempting to secure appropriate care for families needing additional support after a loved one’s death.

In the current study, we attempted to bridge the above-mentioned gaps in bereavement research. Specifically, with a sample of family members of terminally ill Veterans who received palliative care, we sought to learn if any of several demographic and psychosocial predictors assessed during EOL care could predict post-death distress. In particular, we investigated whether our previous findings of predictors of family members’ pre-loss grief in this same sample (Burke et al., 2015) functioned as prospective predictors of their post-loss grief as well. Finally, we explored the relation between anticipatory grief (AG; the process of grieving the loss of loved ones in advance of their inevitable death in the course of terminal illness; Rando, 2000) and family members’ subsequent CG symptomatology 6 to 10 weeks after the death occurred.

Background

Grief is the natural, normal, and often necessary response to the loss of a loved one. For individuals experiencing the life-threatening illness of a treasured family member, grief can be experienced both before and after the death. And, when the fatal illness has run its course, it is likely that individual family members each will grieve the loss differently.

Contemporary research consistently suggests that bereaved adults commonly experience significant grief reactions following the death of a loved one, at least initially. However, at least in the case of natural-death losses, most grievers adapt to their loss within 1 to 2 years (Bonanno & Mancini, 2006). In fact, roughly 50% of bereaved individuals experience a high level of resilience in response to bereavement (Galatzer-Levy & Bonanno, 2012) and experience little in the way of psychological distress. Conversely, a small but significant number of mourners (10%–15%; Prigerson et al., 1995) become incapacitated by grief symptoms that are characterized by profound separation distress and an inability to accept the loss, which have been linked to subsequent poor physical and mental health, spiritual distress, suicidality, and early mortality (Burke & Neimeyer, 2014; Gilewski, Farberow, Gallagher, & Thompson, 1991; Hardison, Neimeyer, & Lichstein, 2005; Latham & Prigerson, 2004; Prigerson et al., 1997; Stroebe, Schut, & Stroebe, 2007). Without targeted treatment, CG sufferers often experience unabated symptoms for years, decades, or a lifetime (Prigerson et al., 2009). Having this understanding of CG symptomatology, a primary aim of this study was to understand if factors assessed during the death and dying period in a sample of family members of fatally ill Veterans could predict their level of grief distress after the death as a potential means of informing future preventive interventions.

Risk Factors for Intense Grieving

Several reviews of studies of predictors of grief-related distress (Lobb et al., 2010; Stroebe et al., 2007) have informed the field about what might exacerbate the bereavement period for some mourners but not others, at least in terms of contemporaneous factors. Burke and Neimeyer (2013), on the other hand, conducted a comprehensive, empirical review of such predictors, focusing solely on results from empirical bereavement studies (n = 43) that examined prospective predictors of grief. The following confirmed risk factors emerged from this review: (a) low levels of social support, (b) insecure attachment style, (c) seeing the body in the context of violent death or dissatisfaction with death notification, (d) being a spouse or a parent of the deceased, (e) high levels of pre-death marital dependency, and (f) high levels of neuroticism.

In addition to these primary conclusions, Burke and Neimeyer’s (2013) review also suggested that being young or non-Caucasian, having less education, little income, prior losses, or losing a child of any age to a violent, sudden death tends to predict prolonged and intense grief. A recent update of this review underscored the relevance of these predictors, and added female gender of the mourner, caregiver burden, insufficiency of institutional support (e.g., in the hospital or hospice), and inability to make sense of the loss to the list of probable risk factors (Neimeyer & Burke, 2017).

Beyond these demographic, contextual, and psychosocial factors associated with heightened bereavement distress, some research has examined whether intense pre-loss grieving in anticipation of a loved one’s impending death, commonly referred to as anticipatory grief (AG), forecasts mourners’ adaptation in the months that follow the death. However, existing evidence is mixed in terms of the possible costs and benefits of pre-death grieving (e.g., Breen, 2012). Theoretically, AG provides the mourner with opportunities to process the loss prior to the death in order to integrate loss-associated changes and emotions. On the other hand, the benefits of AG remain questionable, primarily because it is often complicated by the countless uncertainties surrounding the death itself (e.g., the patient’s unpredictable disease trajectory, treatment-choice quandaries, caregiving demands, stressful decision-making; Luhrs et al., 2005). For instance, some studies report that AG significantly increases the family member’s pre-loss tension and strain, preventing the accomplishment of necessary grief tasks prior to the death (Saldinger & Cain, 2005) and, in some cases, increases grief distress following the death (e.g., Nielsen, Neergaard, Jensen, Bro, & Guldin, 2016). In fact, in its most psychologically severe form, AG symptomatology appears to share some symptom overlap with CG (e.g., intrusive thoughts, painful yearning, intense anger, emotional dysregulation; Carr, House, Wortman, Nesse, & Kessler, 2001).

In a rare study assessing risk factors for AG with family members of Veterans receiving palliative care, Burke et al. (2015) found that high levels of AG in family members were related to having less education, discomfort with emotional closeness and intimacy, high levels of neuroticism, high pre-death dependency upon the Veteran, low levels of grief-specific support, an inability to make sense of the prognosis/looming loss, and experiencing a spiritual crisis in relation to the loved one’s impending death. However, more research is needed to determine whether such predictors observable in the EOL context as well as AG itself forecast more difficult adjustment once the death of the loved one has occurred.

The Department of Veterans Affairs’ (VA) literature suggests that being a Veteran differs in significant ways from being a civilian—with positive and negative implications that appear to be accentuated in the face of terminal illness and EOL planning/care. Although rarely explored, research indicates that many Veterans have had experiences (e.g., exposure to threatening sights, sounds, and chemicals; higher rates of physical and psychological distress; Emanuel et al., 2010; Kang et al., 2003; Zhu et al., 2009) that may, in turn, affect the Veteran’s EOL experience, and possibly even the bereavement experience of their family members following their death. This study focuses on the latter—exploring the loss and grief experiences of family members in the wake of the Veteran’s death.

Aims of This Study

Our goal in the present study was to evaluate rarely explored aspects of intense grieving in a VA setting. Specifically, we examined the pre- and post-loss bereavement experiences of survivors of Veterans who died of terminal illness to assess risk factors of post-loss grief distress. A paucity of research exists on the relation between pre- and post-loss bereavement distress, as well as what makes some individuals grieving the death of a loved one more susceptible to poor bereavement outcome. Thus, this study is an exploratory evaluation of prospective predictors of intense grieving in families of Veterans receiving palliative care at the EOL, placing emphasis on the examination of risk factors related to the family member’s background and other psychosocial factors. To do so, we conducted analyses using two assessment points: Time 1 (T1), a pre-death assessment immediately following the Veteran’s hospice eligibility determination, and Time 2 (T2), a post-death assessment at 6 to 10 weeks following the death of the Veteran to test the following hypotheses:
Based on our previous findings using data from this same sample showing that specific demographic, relational, emotional, and existential risk factors were associated with pre-loss anticipatory grief, we expected that the same set of factors would predict intensity of grief symptomatology approximately 2 months post-loss. Specifically, we hypothesized that the following factors measured at T1 would predict high levels of grief at T2: (a) low levels of education, (b) low levels of grief-specific social support, (c) high levels of dependency upon the Veteran, (d) difficulties making meaning of the prognosis/impending death, (e) insecure attachment style, (f) high levels of neuroticism, and (g) spiritual struggle.

Despite the potential benefits associated with grieving in anticipation of a loved one’s death, given the specific factors that accompany the dying process for the family member (e.g., added stress, elevated fear and concern, and high needs of both the patient and family member) coupled with findings showing that family members sometimes experience intense grief reactions in relation to imminent death, we posited that high levels of pre-loss grief distress would predict high levels of post-loss grief distress in this sample. Specifically, we hypothesize that high T1 AG scores would predict high T2 CG symptoms.

Method

Participants

Participants were 35 adult family members or friends (hereafter referred to as family members) who were grieving the death of a Veteran who had died of a terminal illness. Prior to their death, all of the deceased Veterans had either been admitted to the palliative care unit (PCU) of the Memphis Veterans Affairs Medical Center (MVAMC) or had received outpatient palliative care services at the MVAMC. Demographic characteristics of the family members are shown in Table 1.
Table 1.
Family Members’ Demographic Descriptive Statistics (N = 35).

Outcome variable Mean SD Range

Complicated Grief scores (PG-13) 8.43 8.26 0 to 29

Demographic variable n %

Age in years 58.64 13.25 29–87

Sex

Female 27 77.1

Male 8 22.9

Race/Ethnicity

Caucasian 22 62.9

African American 13 37.1

Relationship to the Veteran

Spouse 12 34.3

Daughter 6 17.1

Son 5 14.3

Sister 6 17.1

Brother 3 8.6

Other (e.g., niece, friend) 3 8.6

Education

Grade 8 or lower 1 2.9

Some high school 3 8.6

High school degree/GED 12 34.3

Some college 10 28.6

College degree 4 11.4

Beyond college degree 5 14.3

Income

<$10,000 2 5.7

$10,000–$24,999 11 31.4

$25,000–$49,000 9 25.7

$50,000–$74,999 3 8.6

$75,000–$99,999 6 17.1

>$100,000 3 8.6

DNR 1 2.8

Faith tradition

Christian 34 97.1

Secular/Nonreligious/Agnostic 1 2.9

All of the Veterans in our sample (n = 35) were male and had been diagnosed with a variety of types of terminal illnesses such as cancer, end-stage renal disease, dementia, or chronic obstructive pulmonary disease, among other conditions. They ranged in age from 49 to 93 years (M = 72.2 years, SD = 11.7 years), and 63% of them were Caucasian (n = 22), with the rest being African American (n = 13; 37%). They all served in the United States armed forces, and 37% (n = 13) had service-connected status (i.e., had developed a chronic condition associated with their military involvement). Their service was reflected in the following branches of the military: Army, n = 22 (62.9%); Navy, n = 6 (17.1%); Air Force, n = 5 (14.3%); Marines, n = 2 (5.7%).

Materials and Procedure

After receiving approval from both the MVAMC’s and the University of Memphis’ institutional review boards, participants were recruited from the MVAMC inpatient PCU and also via referrals from the Palliative Care Consult Team. The PCU is an eight-bed, acute care unit that provides services to Veterans with advanced illnesses refractory to treatment and who require inpatient care for aggressive symptom control, and are imminently dying. The PCU interdisciplinary team members (IDT; e.g., physicians, nurses, social workers, psychologists, chaplains) work together to provide services for Veterans in need of symptom palliation of a life-threatening condition. Specific services include advanced care planning, goals of care discussions, assistance with EOL issues, and pain or symptom management, among others. A subset of participants had a Veteran who received palliative care services while admitted to other units of the hospital or through outpatient services.

Only family members of Veterans who were hospice eligible were recruited to participate in this study. For the purposes of this study, to be hospice eligible the patient’s primary physician certified that the Veteran had a terminal illness and likely had less than 6 months to live. Both family members whose hospice-eligible loved one was admitted to the PCU and those who were not admitted were eligible to participate. If recruitment of a family member was not possible, we solicited the participation of an individual who visited the Veteran during his/her hospital stay (e.g., a friend).

Individuals eligible to participate in this study met the following inclusion criteria: (a) 18 years or older, (b) the family member or friend of a hospice-eligible terminally ill Veteran, and (c) able to speak English fluently. Interested participants were invited by the primary author (an IDT member) to read and sign the Informed Consent and HIPAA forms. Participants received no monetary compensation for their participation, however, as with all patients’ family members, following the death of their loved one they were offered bereavement support through the MVAMC’s chaplaincy program. Participants were asked to complete three assessments that varied in terms of time commitment. Due to attrition at T3 that prevented performing certain statistical analyses, only T1 and T2 assessment data have been analyzed for this study. At T1, participants provided data either by completing a battery of self-report, pencil and paper measures in person or by completing researcher-administered measures in person or via phone, whichever was more convenient for the participant. All T2 assessments were conducted via a phone call initiated by the lead author. Given the emotional nature of grief and our attempt to assess participants’ grief responses relatively shortly after the Veteran’s death, consistency in terms of using the same investigator (lead author) for data collection across time points was considered essential. Prior to the second assessment, a packet containing T2 questionnaires was mailed to the participant who had them in hand and followed along as the researcher read each questionnaire’s items aloud over the phone. The participant’s responses to items were recorded on copies of the questionnaires, to be used for later analysis.

Measures

As well as examining a variety of demographic factors, we assessed survivor-related characteristics using scales bearing on the following: (a) anticipatory grief, (b) complicated grief, (c) relational dependency, (d) attachment style, (e) negative religious coping, (f) grief-specific social support, (g) EOL and death-related meaning making, and (h) neuroticism.

Complicated Grief symptoms were assessed using the Prolonged Grief Disorder Scale (PG-13; Prigerson, Vanderwerker, & Maciejewski, 2008). The PG-13 is a 13-item self-report, diagnostic tool to evaluate prolonged grief disorder (also known as CG), using questions such as In the past month, how often have you had intense feelings of emotional pain, sorrow, or pangs of grief related to the lost relationship? Prigerson et al. (2008) achieved high internal validity (α = .93) when using the PG-13 in a sample of bereaved cancer caregivers. In this study, Cronbach’s alpha for the PG-13 was .92.

Anticipatory Grief was evaluated using the Anticipatory Grief Scale (Theut, Jordan, Ross, & Deutsch, 1991), a 27-item self-report tool designed to assess the bereavement experience of individuals whose spouses have been diagnosed with dementia; however, with the authors’ permission, in this study, we substituted references to “dementia” with “life-threatening illness.” Items are scored on a 5-point Likert scale, ranging from strongly disagree to strongly agree, with example items such as: I get angry when I think about my loved having a terminal diagnosis. Theut et al. (1991) found good internal consistency (α = .84) for this measure in a sample of wives grieving their husband’s diagnosis. We also found adequate internal consistency using this measure with our sample (α = .73).

Negative Religious Coping (NRC) was assessed using the NRC subscale of the Brief RCOPE (Pargament, Smith, Koenig, & Perez, 1998). The Brief RCOPE is a reliable and valid measure of religious coping, using 14 items and two subscales to assess both positive religious coping (e.g., Focused on religion to stop worrying about my problems) and NRC (e.g., Felt punished by God for my lack of devotion). The Brief RCOPE has shown adequate to high internal reliability for both subscales (α = .80 and .69, respectively) in three distinct trials of distressed individuals (Pargament, et al., 1998), and in samples of violently bereaved African Americans (Burke, Neimeyer, McDevitt-Murphy, Ippolito, & Roberts, 2011; PRC: α = .88 and NRC: α = .79). Likewise, we achieved good internal reliability using the NRC subscale in our sample (α = .84).

Grief-Specific Social Support was measured using the Inventory of Social Support (Hogan & Schmidt, 2002). The ISS is a five-item, self-report questionnaire that uses a 5-point Likert-type scale ranging from 1 = does not describe me at all to 5 = describes me very well to subjectively measure available grief-specific social support. A representative question is: I can get help for my grieving when I need it. Burke, Neimeyer, and McDevitt-Murphy (2010) examination of grief support among African American homicide survivors revealed adequate internal consistency (α = .77) using this measure. We achieved adequate internal consistency using the ISS in this sample as well (α = .80).

Attachment Style was assessed using the Adult Attachment Scale (Collins & Read, 1990). The 18-item AAS measures adult attachment style dimensions including comfort with closeness and intimacy (Close subscale), comfort with depending on others (Depend subscale), and worry about being rejected or unloved (Anxiety subscale). Each item is rated on a 5-point scale ranging from 1 = not at all characteristic to 5 = very characteristic. Collins and Read (1990) reported Cronbach’s alpha coefficients of .69 for Close, .75 for Depend, and .72 for Anxiety. Cronbach’s alpha coefficients for the AAS in our sample were .69 for Close, .82 for Depend, and .75 for Anxiety.

Meaning Made of the Prognosis/Loss was analyzed using the Integration of Stressful Life Experiences Scale (ISLES; Holland, Currier, Coleman, & Neimeyer, 2010). This scale uses a 5-point response format (1 = strongly agree to 5 = strongly disagree) and assesses the extent to which a loss makes sense and allows the bereaved individual to maintain a sense of security in the world. A representative item is: My previous goals and hopes for the future don’t make sense anymore since this event. In their two-sample study (n = 152, respectively), Burke et al. (2014) reported Cronbach’s alphas of .96 and .95 in community and college student samples, respectively. Similarly, Cronbach’s alpha for the ISLES in this study was .91.

Neuroticism (or anxiety-proneness) was measured using the Neuroticism subscale of the Big Five Inventory (John & Srivastava 1999). This eight-item subscale asks participants to indicate the extent to which they perceive themselves as being emotionally stable, by responding to such self-descriptive statements as I see myself as someone who: worries a lot, is emotionally stable, not easily upset. This subscale achieved a solid level of internal consistency (α = .85) in a recent adult bereavement study (Lee & Surething, 2013). We found sufficient internal consistency in the current sample using the Neuroticism subscale (α = .83).

Pre-loss Dependency was assessed by adapting and averaging four items from the Dyadic Adjustment Scale (Spanier, 1976). Specifically, participants were asked to endorse the following items in light of their relationship with the Veteran: The idea of losing [my loved one] is terrifying to me; No one could ever take the place of [my loved one]; If [my loved one] died, it would be the worst thing that could happen to me; I would feel completely lost if I didn’t have [my loved one]. Cronbach’s alphas revealed adequate internal consistency using these questions in an elderly sample of bereaved spouses (α = .80; Bonanno et al., 2002) and also in the current sample with Veteran family members (α = .72).

Data Analysis Plan

To assess risk factors for intense, post-loss grieving in this sample, first, CG scores were assessed in relation to demographic variables of the family members, such as gender (i.e., male vs. female) and ethnicity/race (i.e., Caucasian vs. African American), by way of an independent samples t-test, and relationship to the deceased Veteran family member (i.e., Spouse, Child, Sibling, Other—comprising friends, step-siblings, and acquaintances), household income, and level of education through the use of a one-way analysis of variance (ANOVA).

Next, because this study sought to identify both contemporaneous and prospective risk factors for intense grief observable among respondents prior to the death of their Veteran loved one (T1) and following the death (T2), associations between risk factors at T1 and T2 and CG symptomatology at T2 were assessed using Pearson’s product-moment correlations in separate analyses at each time point.

Finally, using variables having a significant association with T2 CG symptomatology, hierarchical regression analyses were conducted to examine the predictive power of identified risk factors, as well as the variance in grief scores accounted for by each predictor. Prior to analysis, it was determined that there were no violations of the assumptions of normality, multicollinearity, or homoscedasticity.

Results

Demographic Variables and Level of Grief Distress

An independent samples t-test was conducted to examine levels of grief distress relative to gender and race/ethnicity. Results indicated that there was a difference in T2 CG symptomatology between males (M = 9.25, SD = 7.46) and females (M = 16.48, SD = 11.2), such that females exhibited higher levels of grief at T2. However, due to a disproportionate number of female respondents, these results only trended toward significance; t(33) = −1.71, p = .09. When investigating differences in grief scores based on race/ethnicity, Caucasian respondents had significantly higher levels of intense grief (M = 17.91, SD = 10.76) compared with African American mourners (M = 9.62, SD = 9.06); t(33) = 2.33, p = .02.

To investigate levels of CG symptomatology among relationship categories, an ANOVA was performed. The overall effect of relationship to the deceased Veteran was significant, F(3, 31) = 4.44, p = .01. Post hoc analyses using Tukey’s criterion for significance indicated that the average level of T2 grief distress was significantly higher among spouses of Veterans (M = 22.37, SD = 10.95) compared with adult children (M = 9.27, SD = 5.80, p = .01). The difference between the spousal group and a combined “other” relationship group, trended toward significance (M = 7.33, SD = .57, p = .08). Finally, the difference between the spousal group and the sibling group was not significant (M = 14.06, SD = 11.86, p = .21). However, the lack of significance in the latter two groups may be attributable to unequal sample size comparisons. CG symptomatology did not significantly differ according to household income, F(5, 28) = 1.40, p = .26, or level of education of the survivor, F(5, 29) = 1.21, p = .33.

Relation Between Bereavement-Related Risk Factors and CG Symptomatology

Pearson’s product-moment coefficients revealed strong and significant correlations in levels of post-loss grief symptomatology with regard to pre-loss risk factors, such as (a) AG, (b) relational dependency, (c) neuroticism, (d) meaning making, and (e) loss-related social support (see Table 2). Post-loss grief outcome was not significantly associated with negative religious coping (r = .14, n = 35, p = .42) or adult attachment-style dimensions, such as the degree of comfort with depending on others (r = −.03, n = 35, p = .85) and anxiety about being rejected or unloved (r = .03, n = 35, p = .87).
Table 2.
Intercorrelations Among Prospective, Contemporaneous Risk Factors and Post-Loss Grief Symptoms.

1 2 3 4 5 6 7 8 9

1 Complicated Grief Symptoms _

2 T1 Anticipatory Grief .44** _

3 T1 Dyadic Dependence .48 .67 _

4 T1 Neuroticism .42* .32 .28 _

5 T2 Neuroticism .39* .11 .05 .78** _

6 T1 Meaning Making −.36* −.58 −.62 −.47 −.13 _

7 T2 Meaning Making −.63 −.32 −.45 −.75 −.64 .60 _

8 T1 Social Support −.40* −.32 −.15 −.34* −.15 .40 .24 _

9 T2 Social Support −.52 −.17 −.15 −.44 −.30 .47 .48 .60** ___

Note. N = 35; Pearson’s r.

*
p < .05. **p < .01.

Anticipatory grief

Our evaluation of the relation between T1 AG and T2 grief distress evidenced a significant positive relation. Family members who had high levels of AG before the death tended to have higher levels of grief after the death, r = .44, n = 35, p < .01, indicating that AG may function as a prospective risk factor of CG symptomatology.

Relational dependency

The correlation between T1 relational dependency and T2 grief indicated a positive and significant association, such that higher levels of relational dependency preceding the death of the Veteran was associated with higher levels of grief symptoms following the death (r = .48, n = 35, p < .01). Based on these results, pre-death relational dependence emerged as a prospective risk factor for CG symptomatology.

Neuroticism

A moderate, yet significant linear connection emerged between T1 neuroticism and T2 CG symptomatology, in which high levels of neuroticism at T1 were associated with high levels of grief at T2, r = .42, n = 35, p < .05. This positive correlation continued through the duration of the study, with high levels of neuroticism measured at T2 continuing to be associated with high grief intensity at T2, r = .39, n = 35, p < .05. Therefore, neuroticism was identified as a prospective and contemporaneous predictor of intense mourning.

Meaning making

Examining the relation between pre-loss meaning making and post-loss grief, we found that an inability to make meaning with regard to the patient’s terminal illness or disease prognosis was related to more severe grieving following the loss, r = −.36, n = 35, p < .05, particularly for those individuals for whom a lack of meaning making resulted in an insecure footing in the world (e.g., identity crisis), r = −.39, n = 35, p < .05. Furthermore, this inverse relation was strengthened from T1 to T2, such that low levels of meaning making at T2 were related even more strongly to levels of grief distress following the death, r = −.63, n = 35, p < .001. This was true for those who had difficulty making sense of the loss, r = −.35, n = 35, p < .05, and particularly for those who struggled to establish or maintain a secure identity in the wake of their loss, r = −.65, n = 35, p < .001. Consequently, meaning making emerged as a prospective and concurrent predictor of CG symptomatology.

Social support

A significant, moderate, inverse relation emerged between the availability of loss-related social support and post-loss grief, suggesting that lower levels of T1 social support were related to higher T2 grief symptoms (r = −.40, n = 35, p < .05). Similar to meaning making, social support’s role in the maintenance of grief symptomatology at T2 appeared to grow over time, with an even stronger inverse relation emerging between T2 bereavement-related social support and T2 grief distress, r = −.52, n = 35, p < .01. Therefore, social support was identified as a prospective and contemporaneous predictor of CG symptomatology.

Assessing the Predictive Capacity of Risk Factors on Post-Loss Grief Symptomatology

To examine the amount of variance in post-loss grief symptomatology accounted for by each of the related predictors, a hierarchical regression analysis was conducted. Prior to conducting the regression analysis, all continuous predictor variables were standardized using Z scores, as recommended by Aiken, West, and Reno (1991). Results for this analysis can be found in Table 3.
Table 3.
Hierarchical Regression Examining the Predictive Capacity of Risk Factors on Post-Loss Grief Symptoms.

Total R² ΔR² SE Beta t

Step 1: Demographic variables .20*

Ethnicity 3.54 .34* 2.13

Sex 4.07 .24 1.47

Step 2: Anticipatory grief .41 .21**

Ethnicity 3.11 .37* 2.66

Sex 3.57 .20 1.45

Anticipatory grief 1.51 .46 3.26

Step 3: Prospective predictors .55 .15

Ethnicity 3.49 .26 1.63

Sex 3.60 .31* 2.19

Anticipatory grief 2.08 .18 .96

Relational dependence 2.14 .34 1.72

T1 neuroticism 1.92 .07 .40

T1 social support 1.71 −.41* −2.58

T1 meaning making 2.15 −.21 1.08

Step 4: Contemporaneous predictors .72 .17**

Ethnicity 3.03 .22 1.63

Sex 3.31 .14 1.04

Anticipatory grief 1.83 .35* 2.06

Relational dependence 1.90 .24 1.38

T1 neuroticism 2.49 −.27 −1.19

T1 social support 1.68 −.22 −1.41

T1 meaning making 2.11 −.47* 2.43

T2 neuroticism 2.34 −.03 −.15

T2 meaning making 2.50 −.57* −2.45

T2 social support 1.70 −.34* −2.13

Note. N = 35.

*
p < .05. **p < .01.

To control for the influence of covariates, race/ethnicity and gender were entered into the model in Step 1. This initial step significantly predicted T2 grief symptomatology and accounted for 20% of the variance in CG scores, F(2, 32) = 3.89, p < .05. Step 2 controlled for the covariates noted earlier and introduced AG, which significantly predicted post-loss grief distress and accounted for an additional 21% of the variance in CG scores, F(3, 31) = 6.93, p < .01.

To examine the individual contribution of prospective risk factors (predictors measured at T1), a third iteration of the model was conducted with T1 relational dependence, neuroticism, grief-related social support, and meaning making entered in Step 3. Results indicated that the inclusion of these variables into the model accounted for an additional 15% of the variation in participants’ post-loss grief and significantly predicted T2 grief scores above and beyond the other predictors listed in the preceding steps, F(7, 27) = 4.65, p < .01. Grief-related social support emerged as the most robust prospective predictor of T2 grief in this step.

Finally, to examine contemporaneous predictors of T2 grief symptoms, neuroticism, meaning making, and grief-related social support at T2 were entered into the model at Step 4. These variables significantly predicted concurrent CG scores above and beyond all other variables listed in the previous steps, and accounted for an additional 17% of the variance in T2 grief symptoms, F(10, 24) = 6.06, p < .001. Thus, overall, high levels of AG, low levels of grief-specific social support following the loss, and an inability to make sense of either the terminal condition/prognosis or the death itself surfaced as risk factors for T2 CG symptomatology, with impaired meaning making emerging as the most robust predictor of grief distress in this sample.

Discussion

In one of only a few such studies, we evaluated demographic, relational, emotional, and existential predictors in 35 family members of Veterans who died of a fatal condition to identify risk factors associated with intense post-loss grieving. Our results revealed that Caucasian family members, and especially women who lost spouses on whom they had been especially dependent, who had experienced high levels of pre-loss grief, and had few supporters to surround them as they struggled to make sense of the impending death were more likely to experience intense, incapacitating grief after the death occurred. These findings reveal a cluster of factors associated with bereavement distress that call for early monitoring and possible treatment to avoid potential development of CG.

Demographics

Our finding in terms of the survivor’s race/ethnicity was inconsistent with our prior study with this same sample (Burke et al., 2015) which showed no difference between African Americans and Caucasians in terms of their anticipatory grief, and, likewise, incongruent with previous studies showing that African Americans generally suffer greater grief than do comparison races (e.g., Laurie & Neimeyer, 2008). In contrast, in this study of post-loss adaptation, African Americans might have experienced less intense grief as a function of reintegration into a communal context of mutual care (Rosenblatt & Wallace, 2005) that was disrupted during their preoccupying engagement with their loved one’s care during the EOL period. Although evaluating this or other explanations calls for more research, these results are also particularly important given that there are approximately 2.4 million Black military Veterans in the United States (U.S. Census Bureau, 2010), and African Americans make up 63% of the population of Memphis, TN—the site of this research project.

By contrast, it was not surprising to learn from our data that spouses in this study experienced markedly more grief distress than did other immediate family members or friends of the deceased Veteran, as other studies concur with this finding (e.g., Prigerson et al., 2002). However, the unexpected finding that Veterans’ spouses and siblings did not differ in terms of how distraught they were over their losses was inconsistent with previous results (Cleiren, 1993). Interestingly, our current findings differed from our pre-loss assessment results (Burke et al., 2015), which showed no difference in AG as a function of relationship to the Veteran.

Consistent with other samples of bereaved adults (e.g., Cohen-Mansfield, Shmotkin, Malkinson, Bartur, & Hazan, 2013), our results suggested that men and women in our sample differed in their reaction to the death of their loved one, with women having a more difficult post-loss bereavement experience. Interestingly, this finding differs from our pre-loss assessment (Burke et al., 2015) showing no gender difference in AG. Numerous factors might explain our finding that prior to the death women’s grief reactions mirrored those of men but not after, not the least of which is that in our sample heavy pre-loss reliance upon the Veteran was associated with greater pre- and post-loss psychological distress. Given that a third (34%; n = 12) of our sample were widows, it helps explain how the death might have accentuated the Veteran’s absence for females in a way that it might not have for a male family member. One benefit of a gender-based finding such as this is that it alerts EOL team members who could proactively address the potential for greater bereavement distress in women who are anticipating a loved one’s death by providing psycho-education and treatment, as needed.

Low levels of income have been inconsistently associated with poor bereavement outcome in a variety of studies (see Burke & Neimeyer, 2013) and was an insignificant factor in this sample at both assessments (pre- and post-loss). Given that nearly 63% (n = 22) of responders made less than $50,000 annually, this finding might primarily reflect the lack of range of incomes in this sample, which could prevent detection of a significant association. Having low levels of education did not predict post-death grief distress in this assessment but was significantly associated with AG in our pre-death assessment with this same sample (Burke et al., 2015), and with CG in other samples (e.g., Kersting, Brähler, Glaesmer, & Wagner, 2011). We previously hypothesized (Burke et al., 2015) that having limited education (along with other associated factors such as fewer resources and lower medical literacy) when attempting to advocate and care for a palliative care patient might exacerbate grief. Perhaps the relinquishment of such cognitively laden responsibilities brings with it a natural lessening of distress once death occurs. If so, clinicians might now be better positioned to guide less-educated grievers during the pre-loss stage when their grief is highest in a manner that provides a smooth transition into bereavement when the pain might significantly diminish.

Relational dependency

Our T1 assessment of correlates of AG in our previous study with this same sample (Burke et al., 2015) indicated that family members who were highly dependent upon their dying Veteran loved one also were more troubled at the thought of losing him than those who were not as dependent. Analyses from our T2 assessment mirror this finding and are congruent with previous studies showing pre-loss dependency predicting post-loss grief (e.g., Thomas et al., 2014). If it is excruciating to watch someone the mourner cannot bear to live without progress closer to death, it might be expected that such grief would not quickly abate and might even worsen when he or she dies. Thus, this finding could inform clinical interventions in palliative care and other EOL settings, as well as grief therapists working with survivors.

Neuroticism

Just as in our previous study with this sample (Burke et al., 2015), Veteran family members who described themselves as anxiety prone (by endorsing items on a measure of neuroticism) both prior to the death of their loved one and again afterward also endorsed intense grief symptomatology following the loss. Similarly, bereaved adults from other studies (e.g., van der Houwen et al., 2010) endorsed having similar bereavement experiences in terms of neuroticism and CG. In fact, a trait conceptualization of neuroticism implies that family members in our sample who worried excessively when the prognosis was given or who anxiously awaited the Veteran’s death likely would exhibit this same trait following his death. Studies such as this that emphasize the association between anxiety and grief suggest the importance of training grievers in the use of anxiety reduction techniques to lessen bereavement distress.

Social support

In their review, Burke and Neimeyer (2013) discovered that 86% (6 of 7) of the studies they examined reported a prospective, predictive relation between social support and post-death grief. Likewise, we found low levels of social support to be a risk factor for complicated bereavement in relation to both AG (Burke et al., 2015) and the current study of post-loss grief, which coincides with results from other studies (e.g., Burke et al., 2010). That is, individuals who provided grief-focused support to participants in our study seemed to help lower the Veteran family member’s grief-related distress both before and after the death. In fact, the relation between having someone to share one’s grief feelings with and a reduction in the intensity of those feelings seemed to strengthen over time for grievers in our sample, which speaks to the enduring power of the presence of supportive others (Bottomley, Burke, & Neimeyer, 2015).

Meaning making

An early ability to make sense of the death of a loved has been associated with better bereavement outcome in previous longitudinal research (e.g., Coleman & Neimeyer, 2010). As with our examination of AG risk factors in this same sample (Burke et al., 2015), where meaning making emerged as the strongest predictor of pre-loss grief (i.e., eclipsing other significant risk factors such as dependency on the dying patient, demographic disadvantage, and spiritual struggle), our results in terms of the prospective relation between pre-loss meaning making and post-loss grief were conclusive. Adult mourners in our study who could make sense of the approaching death of their Veteran loved one expressed less grief-related anguish when assessed 2 months after the death than did grievers who struggled with meaning making. Indeed, the association between impaired meaning making and grief-related distress was still more robust following the death of the loved one, when it functioned as the most robust predictor of adjustment. Notably, this deep emotional pain was most prevalent in those family members who experienced a crisis of identity as they sought to understand, make meaning of, and learn to negotiate life without their loved one.

Anticipatory grief

Consistent with our hypothesis and with investigations showing that high pre-loss grief can forecast more difficult post-loss bereavement (e.g., Nielsen et al., 2016), we found that participants who struggled with their grief prior to the death were more likely to continue to struggle after their Veteran family member died. However, our findings were inconsistent with occasional studies showing that AG facilitates post-death grieving (Bennett & Vidal-Hall, 2000) and those showing no relation between pre- and post-loss psychological struggle (Carr et al., 2001). Whether these inconsistencies are a function of sampling differences (i.e., certain subsets of the bereaved population might be particularly vulnerable to severe psychological distress), mixed findings such as these call for further research so that appropriate interventions can be developed for those who are vulnerable to experiencing grief distress prior to and after a death.

Post-loss grief-related risk factors

Consistent with our hypotheses, elevated levels of post-loss grief were found in families characterized by relational dependency and lower grief-specific social support who also experienced excessive worry and an inability to make sense of their loss. Although we had hypothesized, based on pre-death findings in this same sample (Burke et al., 2015), that family members with insecure attachment styles (specifically, discomfort with closeness and intimacy) and lower educational levels who had also experienced a spiritual crisis prior to the death would also have higher rates of post-loss psychological distress, no such associations were found.

However, a number of the grief-related risk factors found in this study corroborate findings in the post-loss grief literature. Burke and Neimeyer’s (2013) review confirmed consistent prospective predictors of intense grieving, including: (a) low levels of social support, (b) anxious/avoidant/insecure attachment style, (c) discovering or identifying the body (in cases of violent death), (d) being the spouse or parent of the deceased, (e) high pre-death marital dependence, and (f) high neuroticism. However, in this study, we identified several factors not found here. Specifically, rather than finding that being non-White portended susceptibility to greater grief, African Americans in our study fared better than did Caucasians following the death. Although results regarding the function of race in predicting CG are equivocal (Burke & Neimeyer, 2013), with some studies finding no difference, only one study to our knowledge shows Caucasian Americans experiencing greater grief than comparison samples. Bonanno, Papa, Lalande, Zhang, and Noll’s (2005) study with Caucasian American and Chinese bereaved parents and spouses found that Chinese grievers experienced greater grief at 4 months post-loss but at 18 months post-loss Caucasians had higher grief.

Although not included in Burke and Neimeyer’s (2013) list of confirmed risk factors, the inability to make sense or find meaning in the prognosis/death was the single strongest predictor of grief distress for Veteran family members in this sample at both time points (i.e., predicting both AG and CG symptomatology) and was also consistently associated with high dependence upon the Veteran, being prone to excessive worry, and a perceived lack of grief-related social support.

Finally, a dearth of bereavement studies with pre- and post-loss observations only serves to accentuate our finding that T1 AG predicted T2 CG symptomatology, but also calls for follow-up studies with diverse samples to determine if our results are specific to Veteran samples. Taken together, our findings help clarify risk factors associated with psychological distress in family members contemplating and experiencing the loss of a Veteran loved one to a fatal disease.

Study limitations

This bereavement study represents one of only a few that includes a pre-death assessment, allowing researchers to explore the grief experiences of respondents who are awaiting the expected death of a significant person as a result of fatal illness. In addition, it is one of only a few studies that explored the bereavement experiences of family members of dying Veterans, and fewer still that included a longitudinal assessment and a pre-loss evaluation. Using this study design and widely used validated measures yielded a prospective, predictive model to help explain the grief trajectories of family members and to highlight the role of AG in forecasting post-death bereavement outcome.

However, given the relatively small number of respondents, we were limited in terms of the statistical analyses that could be performed. A larger sample might have supported more elaborate analyses. Moreover, the inclusion of qualitative analyses of Veteran family members’ narratives of their lived experiences of loss likely would have provided a rich, descriptive understanding of their bereavement. Additionally, characteristics of our sample (e.g., all participants were either African American or Caucasian, lived in the mid-South of the United States, were a family member of a terminally ill, hospice-eligible male Veteran and, with only one exception, were Christian) makes the study of a larger bereaved population, who represent a variety of types of losses, religions, races, and regions of the world, a high priority. Even so, a notable attribute of this article was the exploration of risk factors of pre- and post-loss grief in a sample that has in the past been relatively neglected by investigators.

Our study exploring predictors of grief in family members of Veterans who received palliative care illustrated that, inasmuch as CG is conceptualized as a disordered and persistent response to loss, with symptomatology indicative of separation distress, preoccupation with the deceased, and an inability to accept the death, it is understandable that mourners whose post-loss experience (much like their pre-loss experience) is characterized by high anxiety and intense grief as they attempt to make sense of life without the person who provided them with care and security, all in the context of minimal support, would be especially susceptible to the development of CG. Early post-loss assessments (in this case, approximately two months post-death) enable researchers and clinicians to learn about factors that could portend poor long-term bereavement outcomes in the hopes of developing treatments to prevent CG, especially in the face of its known risk factors.

Future Directions

Current gaps in the literature prevent a full understanding of the bereavement experiences of family members following a Veteran’s death from terminal illness. Thus, follow-up research with this sample will be conducted and likely will include studies such as a thorough examination of factors related to the Veteran (e.g., specific life challenges, experiences, and conditions) that might affect the adjustment of family members during the death, dying, loss, and grief processes. Additionally, the role of spirituality and family members’ perception of their Veteran’s care at the end of life and their respective relation to grief are relevant topics that also deserve exploration.

Our current and future findings should be viewed as stepping stones in understanding better the bereavement experiences of this and similar samples. For instance, society’s increased reliance upon mobile devices in understanding, monitoring, and managing one’s physical and mental health means that applications likely will be developed that also provide individuals with feedback on how to interpret their grief responses and better ways to cope with the myriad reactions that people have to loss. Having a prior understanding of both pre- and post-loss experiences in a sample such as this enables new and innovative research and treatment designs to be tested more readily with a broader population and to be welcomed adjunctively alongside existing bereavement interventions.

Outcome variable	Mean	SD	Range
Complicated Grief scores (PG-13)	8.43	8.26	0 to 29
Demographic variable				n	%
Age in years	58.64	13.25	29–87
Sex
Female				27	77.1
Male				8	22.9
Race/Ethnicity
Caucasian				22	62.9
African American				13	37.1
Relationship to the Veteran
Spouse				12	34.3
Daughter				6	17.1
Son				5	14.3
Sister				6	17.1
Brother				3	8.6
Other (e.g., niece, friend)				3	8.6
Education
Grade 8 or lower				1	2.9
Some high school				3	8.6
High school degree/GED				12	34.3
Some college				10	28.6
College degree				4	11.4
Beyond college degree				5	14.3
Income
<$10,000				2	5.7
$10,000–$24,999				11	31.4
$25,000–$49,000				9	25.7
$50,000–$74,999				3	8.6
$75,000–$99,999				6	17.1
>$100,000				3	8.6
DNR				1	2.8
Faith tradition
Christian				34	97.1
Secular/Nonreligious/Agnostic				1	2.9

	1	2	3	4	5	6	7	8	9
1 Complicated Grief Symptoms	___
2 T1 Anticipatory Grief	.44**	___
3 T1 Dyadic Dependence	.48**	.67**	___
4 T1 Neuroticism	.42*	.32	.28	___
5 T2 Neuroticism	.39*	.11	.05	.78**	___
6 T1 Meaning Making	−.36*	−.58**	−.62**	−.47**	−.13	___
7 T2 Meaning Making	−.63**	−.32	−.45**	−.75**	−.64**	.60**	___
8 T1 Social Support	−.40*	−.32	−.15	−.34*	−.15	.40	.24	___
9 T2 Social Support	−.52**	−.17	−.15	−.44**	−.30	.47**	.48**	.60**	___

		Total R²	ΔR²	SE	Beta	t
Step 1: Demographic variables	.20*
Ethnicity			3.54	.34*	2.13
Sex			4.07	.24	1.47
Step 2: Anticipatory grief	.41	.21**
Ethnicity			3.11	.37*	2.66
Sex			3.57	.20	1.45
Anticipatory grief			1.51	.46**	3.26
Step 3: Prospective predictors	.55	.15**
Ethnicity			3.49	.26	1.63
Sex			3.60	.31*	2.19
Anticipatory grief			2.08	.18	.96
Relational dependence			2.14	.34	1.72
T1 neuroticism			1.92	.07	.40
T1 social support			1.71	−.41*	−2.58
T1 meaning making			2.15	−.21	1.08
Step 4: Contemporaneous predictors	.72	.17**
Ethnicity			3.03	.22	1.63
Sex			3.31	.14	1.04
Anticipatory grief			1.83	.35*	2.06
Relational dependence			1.90	.24	1.38
T1 neuroticism			2.49	−.27	−1.19
T1 social support			1.68	−.22	−1.41
T1 meaning making			2.11	−.47*	2.43
T2 neuroticism			2.34	−.03	−.15
T2 meaning making			2.50	−.57*	−2.45
T2 social support			1.70	−.34*	−2.13

Footnotes

Acknowledgments

The authors gratefully acknowledge the Veterans and the contribution of their family members who participated in this study. This research was a joint project between The University of Memphis and the Memphis Veterans Affairs Medical Center (MVAMC). This material is the result of work supported with resources and the use of facilities at the MVAMC. The contents of this study do not represent the views of the U.S. Department of Veterans Affairs or the United States Government.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Note

Author Biographies

Laurie A. Burke, PhD, is a research assistant professor at the University of Memphis, where her work focuses on aspects of death, dying, loss, and grief processes, including violent death loss, complicated grief, social support, and families bereaved by terminal illness. A major research interest of hers is complicated spiritual grief – a spiritual crisis following loss. She led the development and validation of the Inventory of Complicated Spiritual Grief (ICSG) and is currently testing its revised version, the ICSG-R. She is a licensed clinical psychologist and maintains an active private practice in Portland, OR, serving grieving individuals, with a primary focus on assisting traumatically bereaved adults (i.e., individuals grieving losses from homicide, suicide, or fatal accident).

Robert A. Neimeyer, PhD, is a Professor of Psychology, University of Memphis, where he also maintains an active clinical practice. Neimeyer has published 30 books, including Techniques of Grief Therapy: Creative Practices for Counseling the Bereaved, and Grief and the Expressive Arts: Practices for Creating Meaning, and serves as Editor of the journal Death Studies. The author of over 500 articles and book chapters and a frequent workshop presenter, he is currently working to advance a more adequate theory of grieving as a meaning-making process.

Jamison S. Bottomley, MS, is a student in the Clinical Psychology program at the University of Memphis, where he is currently training in psychotherapy. His research concentrates on the impact of both exposure to non-fatal suicidal behavior and suicide bereavement, the role of closeness and conflict to the deceased in adapting to grief, and how people who have lost a loved one to suicide make meaning of the experience as part of their effort to adapt to the loss and assist others on the same journey.

Melissa A. Smigelsky, MA, MS, is a Clinical Psychology PhD Candidate at the University of Memphis, under the mentorship of Robert Neimeyer, PhD. Melissa earned a master's degree in clinical psychology from Wheaton College Graduate School, where she developed an interest in the integration of religion/spirituality and psychology, particularly as they intersect with trauma and loss. Melissa’s current research in this area includes suicide survivorship, complicated grief, and moral injury and PTSD among combat Veterans. She will begin her predoctoral internship at the Durham VA Medical Center in the fall of 2017.

References

Aiken

L. S.

West

S. G.

Reno

R. R.

(1991) Multiple regression: Testing and interpreting interactions, Thousand Oaks, CA: Sage.

American Psychiatric Association (2013) Diagnostic and statistical manual of mental disorders, 5th ed. Washington, DC: Author.

Bennett

Vidal-Hall

(2000) Narratives of death: A qualitative study of widowhood in later life. Ageing and Society 20(4): 413–428. doi:10.1017/S0144686X99007813.

Boerner

Wortman

C. B.

Bonanno

G. A.

(2005) Resilient or at risk? A 4-year study of older adults who initially showed high or low distress following conjugal loss. The Journals of Gerontology Series B: Psychological Sciences and Social Sciences 60(2): 67–73.

Bonanno

G. A.

Mancini

A. D.

(2006) Bereavement-related depression and PTSD: Evaluating interventions. In: Barbanel

Sternberg

R. J.

(eds) Psychological interventions in times of crisis, New York, NY: Springer, pp. 37–55.

Bonanno

G. A.

Papa

Lalande

Zhang

Noll

J. G.

(2005) Grief processing and deliberate grief avoidance: A prospective comparison of bereaved spouses and parents in the United States and the Peoples Republic of China. Journal of Counseling and Clinical Psychology 73: 86–98. doi:10.1037/0022-006X.73.1.86.

Bonanno

G. A.

Wortman

C. B.

Lehman

D. R.

Tweed

R. G.

Haring

Sonnega

Nesse

R. M.

(2002) Resilience to loss and chronic grief: A prospective study from preloss to 18-months postloss. Journal of Personality and Social Psychology 83(5): 1150–1164. doi:10.1037==0022-3514.83.5.1150.

Bottomley

J. S.

Burke

L. A.

Neimeyer

R. A.

(2015) Domains of social support that predict bereavement distress following homicide loss assessing need and satisfaction. OMEGA-Journal of Death and Dying. Advance online publication, 0030222815612282.

Breen

(2012) The effect of caring on post-bereavement outcome: Research gaps and practice priorities. Progress in Palliative Care 20(1): 27–30. doi:10.1179=1743291x12y.0000000003.

10.

Burke

L. A.

Clark

K. A.

Ali

K. S.

Gibson

B. W.

Smigelsky

M. A.

Neimeyer

R. A.

(2015) Risk factors for anticipatory grief in family members of terminally ill Veterans receiving palliative care services. Journal of Social Work in End-of-Life & Palliative Care 11(3–4): 244–266. doi:10.1080/15524256.2015.1110071.

11.

Burke, L. A., & Neimeyer, R. A. (2013) Prospective risk factors for complicated grief: A review of the empirical literature. In M. S. Stroebe, H. Schut, J. van der Bout & P. Boelen. (Eds.), Complicated grief: Scientific foundations for healthcare professionals (pp. 145–161). New York: Routledge.

12.

Burke

L. A.

Neimeyer

R. A.

(2014) Complicated spiritual grief I: Relation to complicated grief symptomatology following violent death bereavement. Death Studies 38: 258–267. doi:10.1080=07481187.2013.829372.

13.

Burke

L. A.

Neimeyer

R. A.

Holland

J. M.

Dennard

Oliver

Shear

K. M.

(2014) Inventory of complicated spiritual grief scale: Development and initial validation of a new measure. Death Studies 38: 239–250. doi:10.1080/07481187.2013.810098.

14.

Burke

L. A.

Neimeyer

R. A.

McDevitt-Murphy

M. E.

(2010) African American homicide bereavement: Aspects of social support that predict complicated grief, PTSD, and depression. Omega: Journal of Death and Dying 61: 1–24. doi:10.2190/OM.61.1.a.

15.

Burke

L. A.

Neimeyer

R. A.

McDevitt-Murphy

M. E.

Ippolito

M. R.

Roberts

J. M.

(2011) In the wake of homicide: Spiritual crisis and bereavement distress in an African American sample. International Journal for Psychology of Religion 21: 1–19. doi:10.1080/10508619.2011.60741.

16.

Carr

House

J. S.

Wortman

Nesse

Kessler

R. C.

(2001) Psychological adjustment to sudden and anticipated spousal loss among older widowed persons. The Journals of Gerontology Series B: Psychological Sciences and Social Sciences 56(4): S237–S248. doi:10.1093=geronb=56.4.s237.

17.

Cleiren

(1993) Bereavement and adaptation: A comparative study of the aftermath of death, Washington, DC: Hemisphere.

18.

Cohen-Mansfield

Shmotkin

Malkinson

Bartur

Hazan

(2013) Parental bereavement increases mortality in older persons. Psychological Trauma: Theory, Research, Practice, and Policy 5(1): 84–92. doi:10.1037/a0029011.

19.

Coleman

R. A.

Neimeyer

R. A.

(2010) Measuring meaning: Searching for and making sense of spousal loss in later life. Death Studies 34: 804–834. doi:10.1080/07481181003761625.

20.

Collins

N. L.

Read

S. J.

(1990) Adult attachment, working models, and relationship quality in dating couples. Journal of Personality and Social Psychology 58(4): 644–663. doi:10.1037/0022-3514.58.4.644.

21.

Emanuel, L. L., Hauser, J. M., Bailey, F. A., Ferris, F. D., von Gunten, C. F., & Von Roenn, J. (2010). EPEC for Veterans: Education in Palliative and End-of-life Care for Veterans. Chicago & Washington, DC: Department of Veterans Affairs publication.

22.

Galatzer-Levy

I. R.

Bonanno

G. A.

(2012) Beyond normality in the study of bereavement: Heterogeneity in depression outcomes following loss in older adults. Social Science & Medicine 74: 1987–1994. doi:10.1016/j.socscimed.2012.02.022.

23.

Gilewski

M. J.

Farberow

N. L.

Gallagher

D. E.

Thompson

L. W.

(1991) Interaction of depression and bereavement on mental health in the elderly. Psychology and Aging 6: 67–75. doi:10.1037/0882-7974.6.1.67.

24.

Hardison

H. G.

Neimeyer

R. A.

Lichstein

K. L.

(2005) Insomnia and complicated grief symptoms in bereaved college students. Behavioral Sleep Medicine 3: 99–111. doi:10.1207/s15402010bsm0302_4.

25.

Hogan

N. S.

Schmidt

L. A.

(2002) Testing the grief to personal growth model using structural equation modeling. Death Studies 26(8): 615–634. doi:10.1080/07481180290088338.

26.

Holland

J. M.

Currier

J. M.

Coleman

R. A.

Neimeyer

R. A.

(2010) The integration of stressful life experiences scale (ISLES): Development and initial validation of a new measure. International Journal of Stress Management 17(4): 325–352. doi:10.1037/a0020892.

27.

John

O. P.

Srivastava

(1999) The Big Five trait taxonomy: History, measurement, and theoretical perspectives. In: Pervin

L. A.

John

O. P.

(eds) Handbook of personality: Theory and research, 2nd ed. New York, NY: Guilford Press, pp. 102–138.

28.

Kang, H. K., Natelson, B. H., Mahan, C. M., Lee, K. Y., & Murphy, F. M. (2003). Post-traumatic stress disorder and chronic fatigue syndrome-like illness among gulf war veterans: A population-based survey of 30,000 veterans. American Journal of Epidemiology, 157(2), 141–148.

29.

Kersting

Brähler

Glaesmer

Wagner

(2011) Prevalence of complicated grief in a representative population-based sample. Journal of Affective Disorders 131(1): 339–343. doi:10.1016/j.jad.2010.11.032.

30.

Latham

Prigerson

(2004) Suicidality and bereavement: Complicated grief as the psychiatric disorder presenting greatest risk for suicidality. Suicide and Life Threatening Behavior 34: 350–362. doi:10.1521/suli.34.4.350.53737.

31.

Laurie

Neimeyer

R. A.

(2008) African Americans in bereavement: Grief as a function of ethnicity. Omega: Journal of Death and Dying 57: 173–193. doi:10.2190/OM.57.2.d.

32.

Lee

S. A.

Surething

N. A.

(2013) Neuroticism and religious coping uniquely predict distress severity among bereaved pet owners. Anthrozoos 6(1): 61–76. doi:10.2752/175303713X13534238631470.

33.

Lichtenthal

W. G.

Cruess

D. G.

Prigerson

H. G.

(2004) A case for establishing complicated grief as a distinct mental disorder in DSM-V. Clinical Psychology Review 24: 637–662. doi:10.1016/j.cpr.2004.07.002.

34.

Lobb

E. A.

Kristjanson

L. J.

Aoun

S. M.

Monterosso

Halkett

G. K. B.

Davies

(2010) Predictors of complicated grief: A systematic review of empirical studies. Death Studies 34: 673–698. doi:10.1080/07481187.2010.496686.

35.

Luhrs

Meghani

Homel

Drayton

O’Toole

Paccione

Bookbinder

(2005) Pilot of a pathway to improve the care of imminently dying oncology inpatients in a Veterans Affairs Medical Center. Journal of Pain & Symptom Management 29(6): 544–551. doi:10.1016=j.jpainsymman.2005.02.010.

36.

Neimeyer

R. A.

Burke

L. A.

(2017) What makes grief complicated? Risk factors for complicated grief. In: Doka

Tucci

(eds) Living with grief: When grief is complicated, Washington, DC: Hospice Foundation of America.

37.

Nielsen

M. K.

Neergaard

M. A.

Jensen

A. B.

Bro

Guldin

M. B.

(2016) Do we need to change our understanding of anticipatory grief in caregivers? A systematic review of caregiver studies during end-of-life caregiving and bereavement. Clinical Psychology Review 44: 75–93.

38.

Pargament

Smith

Koenig

Perez

(1998) Patterns of positive and negative religious coping with major life stressors. Journal for the Scientific Study of Religion 37(4): 710–724. doi:10.2307=1388152.

39.

Prigerson

Ahmed

Silverman

G. K.

Saxena

A. K.

Maciejewski

P. K.

Jacobs

S. C.

Hamirani

(2002) Rates of risks of complicated grief among psychiatric clinic patients in Karachi, Pakistan. Death Studies 26: 781–792. doi:10.1080/07481180290106571.

40.

Prigerson

H. G.

Bierhals

A. J.

Kasl

S. V.

Reynolds

Shear

M. K.

Day

Jacobs

(1997) Traumatic grief as a risk factor for mental and physical morbidity. American Journal of Psychiatry 154: 616–623. doi:10.1176/ajp.154.5.616.

41.

Prigerson

Frank

Kasl

Reynolds

III Anderson

Zubenko

Kupfer

D. J.

(1995) Complicated grief and bereavement-related depression as distinct disorders: Preliminary empirical validation in elderly bereaved spouses. American Journal of Psychiatry 152: 22–30. doi:10.1176/ajp.152.1.22.

42.

Prigerson

H. G.

Horowitz

M. J.

Jacobs

S. C.

Parkes

C. M.

Aslan

Goodkin

Raphael

Maciejewski

P. K.

(2009) Prolonged grief disorder: Psychometric validation of criteria proposed for DSM-V and ICD-11. PLoS Medicine 6(8): 1–12. doi:10.1371/journal.pmed.1000121.

43.

Prigerson, H. G., Vanderwerker, L. C., & Maciejewski, P. K. (2008). A case for inclusion of prolonged grief disorder in DSM-V. In M. S. Stroebe, R. O. Hansson, H. Schut, W. Stroebe, M. S. Stroebe, R. O. Hansson,…W. Stroebe (Eds.), Handbook of bereavement research and practice: Advances in theory and intervention (pp. 165–186). Washington, DC: American Psychological Association doi:10.1037/14498-008.

44.

Rando

T. A.

(2000) Anticipatory mourning: A review and critique of the literature. In: Rando

T. A.

(ed.) Clinical dimensions of anticipatory mourning: Theory and practice in working with the dying, their loved ones, and their caregivers, Champaign, IL: Research Press, pp. 17–50.

45.

Rosenblatt

Wallace

B. R.

(2005) African American grief, New York, NY: Brunner-Routledge.

46.

Saldinger

Cain

A. C.

(2005) Deromanticizing anticipated death: Denial, disbelief and disconnection in bereaved spouses. Journal of Psychosocial Oncology 22(3): 69–92. doi: 10.1300=j077v22n03_04.

47.

Shear

M. K.

Simon

Wall

Zisook

Neimeyer

Duan

Keshaviah

(2011) Complicated grief and related bereavement issues for DSM-5. Depression and Anxiety 28(2): 103–117. doi:10.1002/da.20780.

48.

Spanier

G. B.

(1976) Measuring dyadic adjustment: New scales for assessing the quality of marriage and similar dyads. Journal of Marriage and the Family 38: 15–28. doi:10.2307/350547.

49.

Stroebe

Schut

Stroebe

(2007) Health outcomes in bereavement. The Lancet 370: 1960–1073. doi:10.1016/S0140-6736(07)61816-9.

50.

Theut

S. K.

Jordan

Ross

L. A.

Deutsch

S. I.

(1991) Caregiver’s anticipatory grief in dementia: A pilot study. International Journal of Aging and Human Development 33(2): 113–118. doi:10.2190/4KYG-J2E1-5KEM-LEBA.

51.

Thomas

Hudson

Trauer

Remedios

Clarke

(2014) Risk factors for developing prolonged grief during bereavement in family carers of cancer patients in palliative care: A longitudinal study. Journal of Pain and Symptom Management 47(3): 531–541. doi:10.1016/j.jpainsymman.2013.05.022.

52.

U.S. Census Bureau. (2010). Retrieved from http://www.census.gov/hhes/veterans/data/.

53.

van der Houwen

Stroebe

Schut

van den Bout

Wijngaards-de Meij

(2010) Risk factors for bereavement outcome: A multivariate approach. Death Studies 34: 195–220. doi:10.1080/0748118090355919.

54.

Zhu, K., Devesa, S. S., Wu, H., Hoar Zahm, S., Jatoi, I., … McGlynn, K. (2009). Cancer Incidence in the U.S. Military Population: Comparison with Rates from the SEER Program. Cancer Epidemiol Biomarkers Prev, 18, 1740–1745.