Abstract
Objective:
Recognizing important bereavement-related needs among sudden loss survivors (e.g., suicide, overdose)—a population that is burgeoning and at risk for deleterious outcomes—is a critical task as needs may reflect modifiable grief-related variables that can assist with post-loss adjustment.
Methods:
Latent profile analysis was used among 347 sudden loss survivors to (a) identify distinct patterns of needs among survivors of sudden loss, (b) assess predictors of such profiles, and (c) investigate differences in profiles in terms of bereavement outcomes.
Results:
Four classes of bereavement-related needs were identified: a low needs class, a moderate needs-spiritual class, a moderate needs-relational class, and a high needs (HN) class. Clear differences emerged between need classes with the HN class evidencing the greatest level of grief and mental health sequelae.
Conclusion:
Tending to bereavement-related needs is critical, as they indicate the degree of distress and reflect modifiable therapeutic variables.
Keywords: bereavement-related needs, mood and anxiety disorders, overdose loss, posttraumatic stress disorder (PTSD), prolonged grief disorder (PGD), sudden loss, suicide loss
1 |. INTRODUCTION
The loss of a loved one through death is a universal experience. The sudden death (e.g., intracranial hemorrhage, suicide, and fatal overdose) of a loved one or close other, although not universal, is still fairly common in the United States. Population-based studies suggest that over half of American adults will experience at least one unexpected death of a close other in their lifetime, with nearly one-third rating this loss as the most traumatic event they have experienced (e.g., Keyes et al., 2014; Kilpatrick et al., 2013). Regarding the annual prevalence of sudden death, sudden cardiac deaths are responsible for approximately half of all heart disease deaths in the United States, resulting in nearly a third of a million deaths each year due to this cause alone (e.g., Chugh et al., 2008; Curtin, 2019). Other unexpected forms of death occur at alarmingly high rates and appear to be on the rise. In 2018, 67,367 deaths occurred due to fatal overdose in the United States. However, between May 2019 and 2020, over 81,000 people died due to a fatal overdose (Stephenson, 2020). Suicide, the 10th leading cause of death annually, has been steadily increasing each year and accounted for over 48,000 deaths in 2018 (Hedegaard et al., 2020). With the United States currently grappling with the COVID-19 pandemic in myriad ways, the number of deaths due to overdose and suicide are likely to burgeon (e.g., Reger et al., 2020; Wakeman et al., 2020), contributing substantially to the proportion of individuals in the United States who will experience the tragic and unexpected death of a close other.
1.1 |. Mental health consequences of sudden loss
The majority of people who experience the loss of a loved one will adjust to the death without complication or impairment (e.g., Stroebe et al., 2007). However, when the death is unexpected, initial shock may prevent the mourner from grasping the reality of the loss, as individuals are unable to share a comforting last message with the decedent or carry out wishes before their death. It is unsurprising, therefore, that the phenomenology of sudden loss may give rise to grief-related pathology. Complicated grief (CG), often referred to as prolonged grief disorder in the psychiatric nomenclature, and poised to be included in the forthcoming text revision for the DSM-5 (Moran, 2020), includes symptoms of intense yearning for the deceased, identity disruption, disbelief, avoidance of bereavement-related stimuli, protracted emotional pain, difficulties moving on, numbness, loss of meaning, and loneliness following the death of someone close that exceeds cultural and contextual norms (e.g., Prigerson et al., 2009). While it has been estimated that 10%–15% of bereft individuals will meet the criteria for CG following the death of a close other due to natural causes (e.g., Prigerson, 2004), far higher rates of CG diagnoses have been found among individuals who lose a loved one or close other to sudden causes (e.g., Lobb et al., 2010).
Beyond the grief-specific pathology of CG, other deleterious mental health outcomes, such as posttraumatic stress disorder (PTSD), appear to be associated with sudden loss. For example, population-based studies have found that sudden or unexpected loss may represent a unique risk factor for PTSD (e.g., Van Ameringen et al., 2008) with the degree of unexpectedness being strongly associated with PTSD symptoms (Boelen, 2015). Similarly, the odds of being diagnosed with major depressive disorder (MDD) are significantly greater for those who had experienced an unexpected death in their lifetime compared to those who did not, while older individuals appear more vulnerable to anxiety-related disorders following sudden loss (Keyes et al., 2014). In a prospective study of 193 older adults, an unexpected loss was associated with marked increases in depression 6 and 12 months following the death compared to losses that were expected (Burton et al., 2006).
1.2 |. Additional vulnerability factors in sudden death bereavement
Although sudden losses pose unique challenges compared to expected losses, a body of evidence suggests that the degree of violence or volition associated with the loss, rather than the death’s suddenness, may account for elevated risk of adverse mental health outcomes (e.g., Kaltman & Bonanno, 2003). In a study that assessed CG symptoms among a large sample of young adults who had experienced the sudden and violent (e.g., homicide, suicide, and accident) or natural loss of a close other, Currier et al. (2006) found that the violence of the loss, but not the loss’ unexpectedness, predicted risk for CG symptomatology. The violent or volitional nature of the loss may also account for markedly elevated risk for PTSD above and beyond the suddenness of the loss (e.g., Kaltman & Bonanno, 2003). A study of 350 newly bereaved widows and widowers found little difference in the rates of PTSD diagnoses between individuals whose spouses died following a chronic illness or unexpected cause, with 10% and 9% of individuals in the sample meeting criteria for PTSD, respectively, whereas 36% of participants who lost a spouse due to suicide met criteria for PTSD (Zisook et al., 1998). Other studies have demonstrated that when the death is violent or has volitional qualities, the bereft are more vulnerable to negative mental health outcomes, such as MDD and anxiety-related disorders (e.g., Brent et al., 2009). In a systematic review, Pitman et al. (2014) found evidence that parents bereaved by the suicide of their offspring were at greater risk for mood-related disorders compared to parents who lost a child due to other causes, and in particular, mothers whose child died by suicide were at an elevated risk for psychiatric hospitalization. Perhaps the most severe adverse outcome that appears to be particularly associated with violent or volitional death losses is the risk for suicide (e.g., Agerbo, 2005). Studying a sample of treatment-seeking adults following violent (i.e., suicide, accident, and homicide) and natural death of a close other, the former group reported elevated active suicidal thoughts compared to the latter (Tal et al., 2017). Similar results were confirmed by a recent systematic review of the literature that found that rates of suicide ideation were extremely high among individuals bereaved by a violent death (Molina et al., 2019).
Other aspects of the loss, such as the acuteness of the loss, are widely understood as a factor that substantially affects one’s bereavement course and experience and this is especially true in the case for sudden, violent, or volitional losses (Feigelman et al., 2009). Using a person-centered approach with 245 individuals who experienced a sudden loss, Boelen et al. (2016) identified three classes of individuals based on bereavement-related outcomes that largely differed in terms of time since the loss, with individuals in the class reflecting greater resiliency having a loss that occurred longer ago compared to the two other classes that included high levels of CG and depressive symptoms. Relationship factors, such as the kinship of the person who died, and more particularly, the quality of the pre-death relationship, also appear to predict levels of post-loss morbidity in the context of sudden or violent and volitional loss (e.g., Smigelsky et al., 2019). Hardison et al. (2005) found that, among a large sample of heterogeneous bereft college students, the perceived closeness to the decedent, but not necessarily the kinship category, predicted grief-related pathology, and likewise, Cerel et al. (2016) found that increased perceptions of closeness greatly increased the odds for a depressive or anxiety-related disorder, and nearly quadrupled the odds for PTSD among a sample of suicide loss survivors.
1.3 |. Bereavement-related needs and sudden loss: A promising path to adaptation
Given the unique and difficult course of bereavement that is often generated by sudden losses, particularly when violent, recognizing important bereavement-related needs among this population is a critical task as needs may reflect modifiable grief-related variables. Indeed, scholars have attempted to identify the needs of survivors of sudden losses to understand ways in which meeting needs may attenuate risk for mental health symptomatology or, ideally, promote adaptive outcomes (e.g., Dyregrov, 2002; Provini et al., 2000). However, these investigations fail to demarcate specific needs, fail to directly assess needs among the bereft, or fail to examine needs in relation to mental health and grief-related outcomes. For example, Wisten and Zingmark (2007) examined the needs of individuals who lost a close other due to sudden cardiac arrest and identified salient needs such as a need for an unambiguous notification and information about the death (e.g., to make sense of the death), as well as emotional sensitivity from others. Though precise, these needs were interpreted by the authors in the context of a medical care environment (i.e., emergency department), and thus were narrow in focus. With the goal of informing future postvention efforts, a recent qualitative study (Ross et al., 2019) acknowledged a need for proactive and practical support among survivors of suicide (a mode of death that is often violent), particularly early in the course of bereavement, such as providing information on the grief process or case management support (Ross et al., 2019). Although descriptive and illuminating, qualitative examinations preclude an investigation of the association between such needs and mental health sequelae. To illuminate prospective pathways to adaptation following sudden loss, it is imperative that the bereavement-related needs of survivors of sudden loss, as well as the mental health and grief-related outcomes associated with the importance of these needs, be examined.
1.4 |. The current study
The current study seeks to address this substantial gap by examining the bereavement-related needs among a heterogeneous sample of individuals bereaved by a sudden death. Thus, using a person-centered approach, the current study had three aims: (1) to identify patterns of needs among sudden loss survivors, (2) to identify the individual, relational, and death-related factors that predict patterns of need, and (3) to examine the association between the importance of bereavement-related needs and mental health and grief-related outcomes. Although this is the first study to examine bereavement-related needs using a person-centered approach, based on the variability of grief responses evidenced in the literature, it was hypothesized that (1) the data would best fit into a multi-class solution consisting of at least three salient needs profiles, (2) closeness to the decedent, recency of the loss, and deaths that may have violent or volitional qualities (e.g., suicide and overdose) would be similarly grouped and would endorse greater levels of bereavement-related needs and (3) high-need classes would endorse more adverse mental health and grief-related challenges, and rate meaning making and pragmatic needs as most important.
2 |. METHODS
2.1 |. Participants and procedures
The sample included 347 adults, aged 18–79 (Mage = 40.83 years, SD = 17.25) who had lost a close other due to sudden, violent, or volitional means, including deaths due to suicide (n = 151), opioid overdose (n = 95), and natural but sudden causes (e.g., acute myocardial infarction; n = 101). The decedent represented a broad range of kinship categories, but the majority of participants lost a child (n = 121; 34.9%), sibling (n = 56; 16.1%), parent (n = 36; 10.4%), or grandparent (n = 38; 11%), with the remaining participants experiencing the death of a distant family member (e.g., cousin, aunt/uncle; n = 30; 10.1%), spouse/partner (n = 29; 8.4%), or friend (n = 22; 6.3%). A small number of other relationships (e.g., coworkers; n = 15; 4.4%) were also represented in the sample. Time since the loss (TSL) varied, with a range between two months and 5 years (Mtsl = 25.27 months, SD = 17.49 months). Most participants identified as female (n = 284; 81.8%) and White (n = 252; 72.6%). More than a quarter of participants represented ethnic minority populations, with 48 participants identifying as African American/Black (13.8%), 23 as Hispanic/Latino (6.6%), 18 as Asian (5.2%), and six as Native American (1.4%).
Data collection occurred via online surveys using Qualtrics, a secure system that meets established standards for Internet security, research, and IRB policy (“Qualtrics Security Statement,” 2016). Upon IRB approval, participants who were at least 18 years of age and experienced the death of a close other due to sudden causes within the previous 5 years—a timeframe widely employed in thanatology research—were recruited through virtual flyers to social media groups, word-of-mouth referrals, newsletter advertisements, an online subject pool system at a large university, and survivor listservs.
2.2 |. MEASURES
2.2.1 |. Demographic, relational, and loss variables
Select demographic variables were obtained, including gender identity, age, ethnicity, cause of death, kinship category of the decedent, and time since the loss. Pre-death levels of closeness with the decedent were assessed using the Quality of Relationships Inventory—Bereavement Version (QRI-B; Bottomley et al., 2019), a 13-item self-report measure that assesses closeness and conflict between the mourner and deceased before the death. The Closeness scale was utilized for the purpose of the current study, and items on this scale assess the degree to which the relationship was supportive and intimate before the death (e.g., “To what extent could you count on this person to help you if a family member very close to you died?”). Items are rated on a 4-point Likert scale ranging from 1 (not at all) to 4 (very much). In the current study, the Closeness factor of the QRI-B demonstrated high internal consistency, α = 0.88.
2.2.2 |. Bereavement-related needs
To identify needs among individuals bereaved by a sudden loss, the Sudden Bereavement Needs Inventory (SBNI; Bottomley & Smigelsky, 2021) was used. The SBNI is comprised of 21 items that comprehensively assess bereavement-related needs. The SBNI was originally co-developed by bereft community partners who had lost a loved one to sudden causes and psychologists who specialize in traumatic bereavement. As an initial procedure of the SBNI, respondents first selected needs that were salient to them at the time of the assessment in a dichotomous fashion (yes/no). Next, respondents were asked to rate, using a 1–5 Likert-type response, how important the selected need was at the time of assessment (1 = not at all important; 5 = extremely important). Based on a recent study (Bottomley & Smigelsky, 2021), the SBNI consisted of six factors reflecting distinct bereavement-related need categories: Meaning Making Needs (e.g., to understand who I am after the loss; to make sense of the loss), Informational Needs (e.g., to better understand the grief journey following this type of loss), Emotional Needs (e.g., to express my thoughts and feelings about the loss with those I love), Pragmatic Needs (e.g., to eat well; to sleep well; to successfully complete daily tasks), Spiritual Needs (e.g., to have an ongoing connection with God; to have an ongoing connection with my spiritual self), and Relational Needs (e.g., to be with those who experienced a similar loss). Total scores within each factor are summed to create a composite for that need category. Psychometric properties of the SBNI were determined to be strong, including adequate convergent and discriminant validity with acceptable internal reliability within each need factor (α = 0.70–0.91; Bottomley & Smigelsky, 2021). In the current study, inter-item reliability was also determined to be acceptable, with alpha values ranging from 0.68 to 0.90.
2.2.3 |. Complicated grief
The Inventory of Complicated Grief-Revised (ICG-R; Prigerson et al., 1995), consists of 19 items that measure severity of CG symptoms, such as yearning or longing for the deceased, numbness, meaninglessness, difficulty with acceptance, and identity confusion. Items of the ICG-R are rated on a 5-point Likert-type scale that primarily assesses frequency (1 = never to 5 = always). Strong psychometric properties of the ICG-R have been demonstrated in a number of studies, including support for its validity in determining a likely CG diagnosis (Barry et al., 2002), and prediction of a range of adverse physical and mental health consequences (e.g., Latham & Prigerson, 2004). The ICG-R demonstrated excellent internal consistency in the current sample, α = 0.91.
2.2.4 |. Posttraumatic stress disorder
The PTSD Checklist for DSM-5 (PCL-5; Weathers et al., 2013) is a 20-item self-report measure of past-month PTSD symptom criteria. Items correspond with the DSM-5 symptom clusters (i.e., Cluster B [re-experiencing], items 1–5; Cluster C [avoidance], items 6–7; Cluster D [negative alterations in cognition/mood], items 8–14; Cluster E [hypervigilance], items 15–20). Sample items include, “In the past month, how much were you bothered by: ‘repeated disturbing dreams of the stressful experience’ and ‘feeling jumpy or easily startled.’” Items are summed and rated on a scale from (0) Not at all to (4) Extremely (range 0–80), with composite scores of 33 or above indicating likely PTSD diagnosis. In the current sample, α = 0.93.
2.2.5 |. Depression and anxiety
The Patient Health Questionnaire-8 (PHQ-8; Kroenke et al., 2009) is an 8-item self-report measure of depressive symptomatology per the DSM-IV (e.g., “Over the past 2 weeks, how often have you been bothered by: ‘little interest or pleasure in doing things’ and ‘Poor appetite or overeating’”). Items are rated on a scale from (0) Not at all to (3) Nearly every day (range 0–24). Items were summed to produce a total score, with scores of 5, 10, 15, and 20 represent mild, moderate, moderately severe, and severe depression, respectively. To assess anxiety symptomatology, the Generalized Anxiety Disorder seven-item scale (GAD-7; Spitzer et al., 2006) was included in the survey battery. The GAD-7 assesses how often participants experienced anxiety symptoms over the previous two weeks (e.g., “Over the past 2 weeks, how often have you been bothered by: ‘feeling nervous, anxious, or on edge’ or ‘trouble relaxing’”). Items are rated on a scale from (0) Not at all to (3) Nearly every day (range 0–21) and summed to produce a total severity score. Both the PHQ-8 and GAD-7 had strong alpha coefficients within the current sample, with α = 0.90 and 0.91 for the PHQ-8 and GAD-7, respectively.
2.2.6 |. Suicide risk
Suicide risk was measured using the 4-item Suicidal Behaviors Questionnaire-Revised (SBQ-R; Osman et al., 2001), a brief self-report that assesses previous suicide attempts, frequency of suicidal ideation, suicidal communication, and the subjective likelihood of a future suicide attempt. Construct validity for the SBQ-R is strong based on its ability to reliably differentiate between suicidal and non-suicidal subgroups in both clinical and nonclinical contexts. Total scores for the SBQ-R range from 3 to 18, with a cutoff of 7 or higher indicating elevated suicide risk (Osman et al., 2001). For the current study, we transformed these values by subtracting 3 from each participant’s total score such that “no suicide risk” was reflected by a score of 0. In this study, internal consistency was moderately high (α = 0.84).
2.2.7 |. Analytic plan
The current study used a person-centered approach to (a) identify distinct latent classes of bereavement-related needs among survivors of sudden loss, (b) assess individual, relational, and loss-related predictors of needs class membership, and (c) investigate whether survivors in these need profiles differ in terms of bereavement outcomes (i.e., CG, PTSD, depression and anxiety, suicidality, meaning making, and posttraumatic growth), after controlling for important covariates (e.g., TSL, pre-death closeness, mode of loss, age, and sex).
2.2.8 |. Statistical analysis
Latent profile analysis (LPA) was used to estimate varying numbers of latent class solutions based on the responses to the six bereavement-related need indicators. All analyses were conducted using Mplus Version 8.0 (Muthén & Muthén, 1998–2017) using full-information maximum likelihood estimation (FIML) with robust standard errors to account for missing and non-normally distributed data. Models specifying correlated indicators were estimated using 1000 sets of random start values with 100 iterations to ensure the reproduction of global maxima and to avoid misidentification of a false local solution (Hipp & Bauer, 2006). Means and variances of the bereavement-related need indicators were freely estimated in all profiles and significant differences in indicators across profiles were assessed using Wald chi-square tests.
Both statistical and substantive criteria were used to determine the optimal class solution. The Bayesian Information Criterion (BIC; Schwarz, 1978), the Lo-Mendell-Rubin test (LMR; Lo et al., 2001), and entropy were compared across solutions. Lower BIC values suggest better model fit (Rose et al., 2007), and differences equivalent to 10 or greater are indicative of evidence supporting one model compared to another (Raftery, 1995). The LMR test assesses relative improvement in model fit by comparing a model with k latent classes with a model with k−1 classes, with significant p-values indicative of better fit when including an additional class (Nylund et al., 2007). Entropy, which ranges from 0.00 to 1.00, represents an index used to quantify the likelihood that participants are accurately classified into the appropriate class (Magidson & Vermunt, 2002). Greater classification accuracy is suggested by high entropy values (>0.80; Berlin et al., 2014). A scree-plot was generated to aid in determining the optimal class solution, in which log-likelihood values associated with each model were plotted (Nylund et al., 2007). Finally, size and interpretability of differing latent profile solutions were assessed, as classes containing less than 5% of the total sample may suggest over-extraction of the data (Berlin et al., 2014). In addition, the theoretical meaning of each solution was considered in determining the optimal number of profiles.
Upon identification of the optimal class solution, covariates (e.g., age, sex, race, pre-death closeness, time since loss, and mode of loss) were mean-centered and simultaneously assessed as predictors of latent profile membership and grief-related outcomes using multinomial logistic regression and path analysis, respectively. The R3STEP method (Asparouhov & Muthén, 2014a) was used to conduct logistic regression, which adjusts for measurement error in profile classification in estimating associations between predictors and latent class membership. To address missing data on covariates (0%–4% missing), multiple imputation (MI) with 100 imputed data sets was used (Graham et al., 2007; Little et al., 2013).
Significant predictors of class membership and grief-related outcomes were then retained and controlled for in auxiliary models examining differences in grief and mental health-related outcomes across profiles using the manual three-step BCH method (Asparouhov & Muthén, 2014b). Wald tests were used to assess significant differences in weighted intercepts across profile-specific grief-related outcomes using a secondary auxiliary model adjusting for age, sex, race, pre-death closeness, time since loss, and mode of loss. Hedge’s g was used as a bias-corrected effect size and statistical significance was set at 0.01 to control for Type-I errors.
3 |. RESULTS
3.1 |. Class enumeration
Models with increasing numbers of latent classes were estimated specifying alternative variance and covariance assumptions. Models specifying classes with freely estimated means and variances fit significantly better compared to those with class varying means and class invariant variances. In addition, models estimated with class varying covariances tended to converge upon non-identified solutions in which the highest log-likelihood was not replicated. Thus, covariances were constrained to equality across classes to assist in model convergence. Fit statistics for models estimated with one through six latent classes were examined. Models failed to converge beyond six classes despite increasing the amount of random start values to 4000.
Examination of the information criteria revealed a superior fit for the four-class solution as evidenced by the decline in BIC values from the one-class solution through the four-class solution. BIC values began to increase with additional classes, suggesting that the estimation of a fifth and sixth class failed to improve upon the fit of a four-class model. The LMR test statistic was significant through the four-class solution, suggesting that a four-class model fit significantly better than a three-class model. The LMR test statistic then reached non-significance (p > 0.05) when the model was expanded to a five-class solution, suggesting that the inclusion of an additional fifth class did not provide significant improvement over a four-class solution. Based on the overall pattern of information criteria, LMR test results, and meaningful interpretation, the four-class model was retained as the optimal solution, and revealed good classification accuracy as indicated by entropy (0.89) and posterior probabilities for most likely class membership ranging from 0.91 to 0.99.
The four-class model probabilistically assigned participants to a low needs class (LN; n = 71, 20.5%), a moderate needs-spiritual class (MNS; n = 36, 10.4%), a moderate needs-relational class (MNR; n = 98, 28.2%), and a high needs class (HN; n = 142, 40.9%). Figure 1 shows patterns of standardized bereavement-related need indicators across the four profiles. Table 1 displays unstandardized means of each need indicator by profile as well as tests of between-profile differences.
FIGURE 1.
Four class solution bereavement-related needs profiles. HN, high needs; LN, low needs; MNR, moderate needs-relational; MNS, moderate needs-spiritual
TABLE 1.
Means (SE) for bereavement-related need indicators for four-class solution
Latent bereavement-related need profiles |
||||
---|---|---|---|---|
LN (n = 71) | MNS (n = 36) | MNR (n = 98) | HN (n = 142) | |
Meaning making needs | 2.65 (1.25)abc | 8.68 (1.23)ae | 10.91 (1.03)bf | 13.49 (0.58)cef |
Informational needs | 0.55 (0.26)abc | 6.18 (0.84)ae | 7.62 (1.40)bf | 12.71 (0.53)cef |
Emotional needs | 2.22 (0.83)abc | 8.10 (1.06)ae | 8.39 (0.86)bf | 12.61 (0.22)cef |
Pragmatic needs | 8.49 (1.52)abc | 12.86 (1.53)ae | 11.31 (0.87)bf | 16.05 (0.57)cef |
Relational needs | 0.49 (0.54)bc | 0.37 (0.17)de | 6.75 (1.04)bdf | 11.55 (0.31)cef |
Spiritual needs | 2.86 (0.65)ac | 9.36 (0.16)ade | 1.29 (0.43)df | 6.16 (0.41)cef |
Note: SE = standard error; LN = low needs; MNS = moderate-spiritual needs; MNR = moderate-relational needs; HN = High Needs.
Columns that are significantly different from one another share a superscript such that:
= LN versus MNS,
= LN versus MNR,
= LN versus HN,
= MNS versus MNR,
= MNS versus HN, and
= MNR versus HN.
Significant differences were tested using Wald's test.
3.2 |. Class descriptions
3.2.1 |. Low needs (n = 71)
Participants in this profile reported the lowest levels of meaning making, informational, emotional, and pragmatic needs of the four profiles. This profile also endorsed relatively low levels of relational needs that were similar to the MNS profile as well as low spiritual needs that were similar to the MNR profile.
3.2.2 |. Moderate needs-spiritual (n = 36)
Participants in this profile reported moderate levels of meaning making, informational, emotional, and pragmatic needs which were significantly greater than those endorsed by the LN profile but similar to those of the MNR profile. This profile reported the highest levels of spiritual needs which were significantly greater than those endorsed by the other three profiles. Individuals in this profile also displayed the lowest levels of relational needs in the sample, similar to those of the LN profile.
3.2.3 |. Moderate needs-relational (n = 98)
Similar to the MNS profile, participants in this profile reported moderate levels of meaning making, informational, emotional, and pragmatic needs which were significantly greater than those of the LN profile but significantly less than the HN profile. Participants in this profile also endorsed moderate levels of relational needs which were significantly greater than those of both the LN and MNS profiles and significantly less than the HN profile. Participants in this profile also endorsed the lowest levels of spiritual needs in the sample, which were significantly less than those of both the MNS and HN profiles but similar to those of the LN profile.
3.2.4 |. High needs (n = 142)
Participants in this profile reported the highest levels of meaning making, informational, emotional, pragmatic, and relational needs, all of which were significantly greater than those endorsed by the other three profiles. Participants in this profile also reported moderate levels of spiritual needs which were significantly greater than those endorsed by both the LN and MNR profiles but significantly less than those of the MNS profile.
3.3 |. Predictors of latent profile membership
Unadjusted baseline differences on covariates between profiles are presented in Table S1. Multinomial logistic regression was used to assess predictors of latent profile membership (see Table 2). Adjusted odds ratios (aORs) were calculated holding all other predictors at their average. Older age was significantly associated with an increased probability for expected classification in the HN class relative to all other classes (aORs ranging from 2.01 to 3.16). Compared to the LN profile, females were significantly more likely to belong to the MNS (aOR = 1.79, 95% confidence interval [CI]: [1.12, 2.88]), MNR (aOR = 1.41, 95% CI: [1.04, 1.93]), and HN (aOR = 2.21, 95% CI: [1.48, 3.28]) profiles. In addition, females were significantly more likely to belong to the HN profile (aOR = 1.56, 95% CI: [1.05, 2.32]) relative to the MNR profile. Relative to the MNS profile, identifying with the BIPOC community was significantly associated with a decreased probability for expected classification in both the MNR (aOR = 0.58, 95% CI: [0.39, 0.88]) and HN (aOR = 0.64, 95% CI: [0.43, 0.94]) profiles.
TABLE 2.
Multinomial logistic regression analysis examining predictors of class membership
MNS versus LN |
MNR versus LN |
HN versus LN |
MNR versus MNS |
HN versus MNS |
HN versus MNR |
|
---|---|---|---|---|---|---|
OR (95% CI) | OR (95% CI) | OR (95% CI) | OR (95% CI) | OR (95% CI) | OR (95% CI) | |
Age | 1.51 [0.84, 2.70] | 1.58 [0.99, 2.50] | 3.16 [1.97, 5.07]*** | 1.05 [0.64, 1.70] | 2.10 [1.28, 3.42]** | 2.01 [1.38, 2.91]** |
Female | 1.79 [1.12, 2.88]* | 1.41 [1.04, 1.93]* | 2.21 [1.48, 3.28]*** | 0.79 [0.49, 1.27] | 1.23 [0.73, 2.07] | 1.56 [1.05, 2.32]* |
BIPOC identity | 1.26 [0.81, 1.95] | 0.73 [0.51, 1.06] | 0.80 [0.54, 1.18] | 0.58 [0.39, 0.88]** | 0.64 [0.43, 0.94]* | 1.09 [0.74, 1.61] |
Pre-death closeness | 2.09 [1.21, 3.63]** | 1.46 [1.04, 2.05]* | 2.35 [1.57, 3.53]*** | 0.70 [0.42, 1.16] | 1.13 [0.68, 1.86] | 1.61 [1.10, 2.36]* |
Time since loss | 0.99 [0.63, 1.57] | 0.88 [0.62, 1.25] | 0.63 [0.45, 0.90]** | 0.88 [0.57, 1.37] | 0.64 [0.41, 0.99]* | 0.72 [0.51, 1.02] |
Suicide | 1.53 [0.50, 2.98] | 7.42 [2.73, 16.11]*** | 6.63 [2.52, 14.96]*** | 4.85 [1.64, 12.07]** | 4.34 [1.53, 12.29]** | 0.89 [0.32, 2.49] |
Overdose | 3.16 [1.76, 13.22]*** | 7.04 [1.84, 21.76]** | 10.34 [2.92, 23.71]*** | 2.23 [0.64, 7.76] | 3.27 [1.25, 10.27]* | 1.47 [0.46, 4.67] |
Note: OR = odds ratio; Cl = Confidence interval. The second latent class of each contrast is the reference category. All measures were mean-centered, thus results reflect the probability of being classified into a particular profile holding all other variables at their average. Bold effects are significant at p < 0.05.
p < 0.05;
p < 0.01;
p < 0.001.
Compared to the LN profile, participants reporting greater levels of pre-death closeness were significantly more likely to belong to the MNS (aOR = 2.09, 95% CI: [1.21, 3.63]), MNR (aOR = 1.46, 95% CI: [1.04, 2.05]), and HN (aOR = 2.35, 95% CI: [1.57, 3.53]) profiles. Participants reporting greater levels of pre-death closeness were significantly more likely to belong to the HN profile (aOR = 1.61, 95% CI: [1.10, 2.36]) relative to the MNR profile. Longer time since loss was significantly associated with a decreased probability for expected classification in the HN profile relative to the LN (aOR = 0.63, 95% CI: [0.45, 0.90]), and MNS (aOR = 0.64, 95% CI: [0.41, 0.99]) profiles.
Participants reporting suicide loss were significantly more likely to belong to the MNR profile relative to both the LN (aOR = 7.42, 95% CI: [2.73, 16.11]) and MNS (aOR = 4.85, 95% CI: [1.64, 12.07]) profiles. In addition, those reporting suicide-related loss were also significantly more likely to belong to the HN profile relative to both the LN (aOR = 6.63, 95% CI: [2.52, 14.96]) and MNS (aOR = 4.34, 95% CI: [1.53, 12.29]) profiles.
Compared to the LN profile, participants reporting overdose-related loss were significantly more likely to belong to the MNS (aOR = 3.16, 95% CI: [1.76, 13.22]), MNR (aOR = 7.04, 95% CI: [1.84, 21.76]), and HN (aOR = 10.34, 95% CI: [2.92, 23.71]) profiles. In addition, those reporting overdose-related loss were also significantly more likely to belong to the HN profile (aOR = 3.37, 95% CI: [1.25, 10.27]) relative to the MNS profile. Overall, these results identifying time since the death, pre-death closeness to the decedent, and violent or volitional losses as predictive of class membership, and associated with the HN and MRN classes in particular, are aligned with the second hypothesis of the current study.
3.4 |. Grief-related outcomes across bereavement-related need profiles
To assess significant between-profile differences in grief-related outcomes, Wald chi-square tests were conducted using covariate-adjusted means, controlling for covariates. Table 3 presents covariate-adjusted means while significant between-profile differences are presented in Table S2. Several significant and large between-profile differences in grief-related outcomes emerged in a manner consistent with our final hypothesis and are summarized below.
TABLE 3.
Intercepts (SE) and comparisons across grief-related outcome measures for each latent class
Measures | LN Estimate (SE) | MNS Estimate (SE) | MNR Estimate (SE) | HN Estimate (SE) | Class comparisons |
---|---|---|---|---|---|
PCL | 15.91 (2.67) | 28.90 (3.10) | 31.18 (2.04) | 37.88 (1.88) | HN > LN & MNS & MNR; MNS = MNR; MNS & MNR > LN |
Suicide Risk | 6.66 (0.71) | 6.30 (0.51) | 7.69 (0.41) | 9.17 (0.57) | HN > LN & MNS & MNR; MNR > MNS; MNS & MNR = LN |
ICG | 23.08 (2.37) | 26.37 (2.75) | 35.21 (1.47) | 37.48 (1.33) | HN = MNR; HN & MNR > MNS & LN; MNS = LN |
PHQ-8 | 5.95 (1.41) | 8.77 (1.14) | 8.65 (0.67) | 11.73 (0.63) | HN > LN & MNS & MNR; MNS = MNR = LN |
GAD-7 | 5.41 (1.20) | 6.57 (0.89) | 7.87 (0.60) | 9.64 (0.62) | HN > LN & MNS; HN = MNR; MNS = MNR = LN |
Note: SE = standard error; LN = low needs; MNS = moderate-spiritual needs; MNR = moderate-relational needs; HN = High Needs; All measures conducted using covariate-adjusted means, controlling for age, sex, race, pre-death closeness, time since loss, and mode of loss.
PTSD symptoms reported by the HN profile were significantly greater than those reported by the LN profile (g = 1.38), the MNS profile (g = 0.56), and the MNR profile (g = 0.41). In addition, both the MNS and MNR profiles reported significantly more PTSD symptoms compared to the LN profile (g = 0.77; g = 0.90, respectively).
Suicide risk of the HN profile was significantly greater than that of the LN profile (g = 0.54), the MNS profile (g = 0.62), and the MNR profile (g = 0.33). The MNR profile reported significantly greater suicide risk compared to the MNS profile (g = 0.37).
Severity of CG symptoms of the HN and MNR profiles were significantly greater than that of both the LN (g = 1.20; g = 0.90, respectively) and MNS profiles (g = 0.91; g = 0.63, respectively).
Regarding mood and anxiety symptoms, depression scores of the HN profile were significantly greater than those of the LN profile (g = 0.98), the MNS profile (g = 0.51), and the MNR profile (g = 0.52). Anxiety symptoms of the HN profile were significantly greater than that of the LN (g = 0.72) and MNS profile (g = 0.52).
4 |. DISCUSSION
Although there is strong consensus regarding the importance of identifying the needs of sudden loss survivors (e.g., Feigelman et al., 2018; Neimeyer et al., 2017), methodological limitations have produced gaps in the extant literature. The present study is the first to identify specific bereavement-related needs and their association with mental health and grief-related sequelae among sudden loss survivors. Using LPA among a heterogeneous sample of survivors, four subgroups of participants were identified with regard to patterns of bereavement-related needs: an LN group, an MNS group, an MNR group, and an HN group.
4.1 |. Need categories in each class
Based on our results, approximately one-fifth (20.4%; n = 71) of participants comprised the LN group and reported the lowest levels of importance of meaning making, informational, emotional, and pragmatic needs. The MNS group had the lowest number of participants (10.3%; n = 36) and reported that these same needs were moderately important, whereas spiritual needs were rated as highly important, to a level significantly greater than all other groups. Perhaps leaning on their faith orientations more readily than other mourners, individuals in the MNS group rated relational needs as being of low importance, consistent with responses of the LN group. More than one-fourth (28.2%; n = 98) of the sample comprised the MNR group, which like the MNS class indicated that meaning making, informational, emotional, and pragmatic needs were moderately important, though they differed from the latter group in endorsing the lowest level of spiritual needs compared to any other group. However, they viewed being around individuals with a similar experience of loss as important, differentiating them from the LN and MNS groups in terms of relational needs. Finally, and of greatest concern, the largest proportion of survivors in the sample were placed in the HN group (40.1%; n = 142), endorsing markedly high levels of need to make sense of the loss, receive information about grief, process grief-related emotions, engage in practical and life-enhancing activities, and affiliate with others who have faced a similar loss.
With specific regard to meaning making and informational needs, results presented here are consistent with qualitative studies that examined the needs of individuals following the sudden fatal cardiac event of a loved one (Wisten & Zingmark, 2007), as well as those following a suicide loss (e.g., Bottomley et al., 2018) in which the bereaved identified needing more information about why the loved one died to make greater sense of the loss. Emotional needs appeared to be equally important for the moderate needs groups (MNS and MNR), though of greater importance to the HN group and lower in the LN group, respectively. The finding that the latter group views meaning making, relational, and emotional needs as being of little importance suggests that the death occurred in a manner that was consistent with the survivor’s broader narrative of loss or understanding of the world (Neimeyer, 2019).
Consistent with the concept that loss has the potential to negatively impact various domains of life-sustaining behaviors (e.g., sleep, exercise; Hardison et al., 2005; Lancel et al., 2020), pragmatic needs were rated as being of moderate to high importance across all need classes. Participants in our sample nearly unanimously endorsed the importance of eating well, exercising regularly, and improving sleep, among others. These pragmatic issues are reflected in Rubin’s (1999) Two-Track Model of Bereavement, which conceptualizes the impact of loss along two distinct but interactive axes, with one “track” being devoted to the relationship to the deceased while the other addresses the biopsychosocial functioning of the mourner, including the self-care needs identified by the survivors in the present study.
4.2 |. Predictors of needs classes
Regarding the composition of each needs class, a number of notable patterns emerged. For example, participants in the HN group were significantly older than all other groups and more likely to be female, while the LN group had the lowest proportion of females compared to all other groups. This finding is unsurprising given that research consistently indicates that being in an older age group and identifying as female are risk factors for complicated grieving (Shear et al., 2013), and are thus likely to endorse a higher degree of needs. Individuals in the LN group reported the least pre-death closeness to the decedent relative to all other groups, with the HN group reporting the greatest closeness to the decedent, in keeping with prior research (e.g., Cerel et al., 2017; Servaty-Seib & Pistole, 2007). Additionally, individuals reporting a more recent loss were more likely to belong to the HN group compared to the LN group, which supports previous findings that grief difficulties are most acute in the early months following a loss for a majority of mourners (e.g., Feigelman et al., 2009; Schwartz et al., 2018).
Regarding the mode of sudden loss, the HN and MNR groups were predominately comprised of suicide loss survivors, with over half of individuals in these groups losing a close other in this tragic fashion. Moreover, after adjusting for numerous covariates, results indicated that suicide loss survivors were nearly seven times more likely to be in the HN or MNR group than the LN group. Research suggests that suicide is distinguished from other forms of sudden loss by its tendency to shatter fundamental assumptions of life, often launching mourners into a protracted quest to make sense of the actions of the decedent (e.g., Bottomley et al., 2018). As such, the fact that the high incidence of suicide loss survivors in the HN group, where meaning making needs were rated as highly important, is unsurprising.
Like suicide loss survivors, individuals who lost a loved one to opioid-related overdose primarily fell into the MNR or HN group. Over a quarter of the MNR group, and greater than a third of the HN group, was comprised of opioid-loss survivors, while only a tenth of the LN group included mourners of this form of loss, suggesting a moderate to high endorsement of needs among individuals bereaved by this form of sudden loss. In fact, overdose loss survivors in the sample were 10 times more likely to be in the HN group, and seven times more likely to be in the MNR group, than in the LN group, revealing an overall pattern of high need in this population. This finding is consistent with literature that suggests that grieving an unexpected overdose calls for additional resources and support in the wake of a loss that is both traumatic and stigmatizing (e.g., Kheibari et al., 2021; Templeton et al., 2016). It has been suggested that the grief of overdose loss survivors is often “devalued” as a result of the decedent engaging in behavior that is morally and socially objectionable (Valentine et al., 2016). This broad societal stance may then stifle the public response and limit resources, making needs among this group of sudden loss survivors especially glaring.
One reason for the seeming overlap of needs between suicide and overdose loss survivors in the current study may be due to the violent and volitional quality of the death. Bereavement following these modes of loss is associated with social stigmatization and “disenfranchisement” (Doka, 2002), and begets existential challenges as the mourner often seeks to comprehend why and how the death occurred (e.g., Bottomley et al., 2018; Guy & Holloway, 2007). Survivors of these potentially volitional deaths may view making sense of the death as a chief task (e.g., Bottomley et al., 2018), including understanding the circumstances of the death, the nature of their relationship to the decedent, and implications of the death for aspects of their own identity. All of these meaning making needs were prominent in both the MNR and HN groups. Furthermore, the perceived and actual stigmatization of these two groups of suicide and overdose loss survivors (Feigelman et al., 2009) often prompts them to seek others with shared experience through participation in peer support groups (Feigelman & Feigelman, 2008). This comports with the placement of the vast majority of suicide and opioid-loss survivors in the present study in either the MNR or HN classes, which rated relational needs as being much higher than the other two classes.
4.3 |. Mental health outcomes among classes
Regarding CG symptomatology, results indicated that the HN group had significantly greater levels of CG symptoms and grief-related impairment compared to the LN and MNS groups. However, no significant differences emerged between the HN and MNR groups. These findings reinforce research suggesting that violent or volitional loss, such as suicide, places mourners at greater risk for CG symptoms compared to individuals grieving the death by non-violent or volitional means (Jordan, 2008), even when those losses are also sudden (Shear et al., 2011). For example, Mitchell et al. (2004) reported that 43% of those bereaved by suicide met the criteria for CG, which is four times the rate found in the general population. Furthermore, meaning making needs were highest among the HN and MNR groups, and research has consistently shown that an inability to make sense of a violent or volitional death fully mediates the relationship between the cause of death and CG symptomatology (Currier et al., 2006; Milman et al., 2019).
Likewise, significant differences in PTSD symptoms were identified, with the HN group evidencing markedly high levels compared to all others. With approximately 88% of this group being comprised of suicide or overdose loss survivors, this finding is consistent with the literature that has identified symptoms of traumatic stress associated with the sudden violent or volitional death of a close other (Keyes et al. 2014). For example, in an epidemiological study conducted by Erlangsen et al. (2017), survivors of a spousal suicide had a sixfold risk for PTSD compared to others who had lost a spouse due to other causes. Further highlighting the role of violent or volitional loss, Zisook et al. (1998) found comparable rates of PTSD among spouses grieving a suicide or accident death (36% meeting criteria for PTSD) while rates of PTSD in other modes of natural unexpected loss were between 9% and 10%. Our findings reinforce these conclusions and extend this literature to mourners bereaved by an overdose death, a group largely ignored by previous studies.
An exceptionally unfortunate outcome that is associated with bereavement is suicidal behavior. Results from the current study revealed that individuals in the HN group were significantly more at risk of suicide compared to the other needs groups, and with CG severity being highest in this class, this finding supports the association between CG and suicide risk in the extant literature (e.g., Latham & Prigerson, 2004). Furthermore, suicide loss survivors accounted for over half of the HN group, reinforcing findings from other studies that identify sudden loss, and suicide bereavement in particular (e.g., Hamdan et al., 2019), as potent risk factors for subsequent suicide. Regarding the association between specific need categories and suicide risk, the HN group rated meaning making needs as being of great importance, and an anguishing search for meaning and suicide risk have been associated in previous studies (e.g., Milkin et al., 2019). Elevations in suicide risk in the HN group can also be explained by the importance of relational needs in this group. According to Wisten and Zingmark (2007), sudden natural loss survivors endorsed the need to garner consistent emotional support from others in an effort to avoid isolation—a powerful predictor of suicide risk (e.g., Calati et al., 2019). Relatedly, given the stigma associated with suicide and drug use, mourners of these types of sudden death often find themselves isolated (e.g., Feigelman et al., 2009; Valentine et al., 2016), whether prompted by the mourner or by society (Kheibari et al., 2021). Taken together, the importance of making meaning of the loss and social connectedness is highlighted in the context of the propagation of suicide, in keeping with our findings of the importance of meaning making and relational needs among individuals who were at greatest risk for suicide in the current study.
Depressive and anxiety symptoms in the HN class were found to be significantly greater than nearly all other groups. When examining the composition of the HN and MNR groups in terms of the mode of loss, a trend emerges suggesting that individuals confronted with sudden violent or volitional loss may be vulnerable to depressive and anxiety symptomatology, which both aligns with and departs from the extant literature that largely focuses on suicide losses. For example, in a large digit-dial study, individuals who were exposed to a suicide death had an increased prevalence of depressive and anxiety disorders compared to those who were not exposed (Cerel et al., 2016). However, research reviews have suggested that differences in mood and anxiety symptomatology between sudden and violent or volitional loss survivors are equivocal, and likely are influenced by the nature of the relationship (Pitman et al., 2014; Sveen & Walby, 2008). With regard to bereavement-related needs, our findings appear to be consistent with the literature that acknowledges the importance of the ability to integrate stressful events into one’s understanding of self and the world, the role of emotional disclosure, and the significance of pragmatic support in the wake of loss and other life adversities in combatting mood and anxiety-related outcomes. For instance, in a study examining meaning making as a possible mechanism of change in mood and anxiety-related symptoms during the provision of cognitive-behavioral therapy, meaning in life and meaning made of life transitions was found to be a partial mediator of reductions in anxiety and depressive symptoms (Marco et al., 2021). With regard to pragmatic support, in a longitudinal study of survivors of homicide loss, satisfaction with instrumental support received (e.g., assistance with child care, providing meals, or running errands) prospectively predicted reductions in depression and anxiety, above and beyond all other forms of support (Bottomley et al., 2017).
4.4 |. Limitations of the current study
Results should be interpreted in the context of several limitations. First, the current study did not include all possible forms of sudden death. As such, generalization of these findings to individuals bereaved due to other sudden forms of loss, such as accident, homicide or mass violence events, and natural disasters should be undertaken with caution. Further, the current study did not include information regarding the perceived intentionality of the death, which may have influenced the current results. Furthermore, although we endeavored to reduce sampling bias by recruiting participants from a variety of sources, convenience sampling was utilized, and, accordingly, the current sample may experience greater levels of bereavement-related distress relative to individuals who did not seek the assistance of organizations used for recruitment. In addition, these data are cross-sectional, and as such, temporal interpretations of the relationship between needs and outcomes should be attempted with caution. Future studies using longitudinal designs will help to clarify the relationship between bereavement-related needs and outcomes over time. Relatedly, the current study did not account for previous mental health diagnoses, history of treatment engagement of the participant, or exposure to adverse life events. Differences in outcomes presented here could have been caused by variations in mental health symptomatology or the cumulative impact of trauma and loss that could have preceded the death.
5 |. CONCLUSION
The current study addresses a significant void in the literature by examining factors that predict patterns of needs among survivors of sudden loss as well as their associated and mental health outcomes. Using a person-centered approach, we identified four homogenous classes of survivors: one group that reported that bereavement-related needs were not of great importance, and three that indicated that meaning making, informational, emotional, and pragmatic needs were of moderate or high importance, respectively. The vast majority of individuals bereaved by a violent or volitional death comprised the moderate or high needs classes, the latter of which displayed the most adverse bereavement-related outcomes after adjusting for time since loss, age, gender, and interpersonal closeness between the mourner and decedent. Furthermore, the current study identified needs that are important, tangible, and reflect modifiable aspects of bereavement that may be amenable to change through psychological intervention (e.g., meaning reconstruction, pragmatic support, emotional disclosure, companionship of mourners of a similar death). Identifying needs in this manner is critical to develop or deliver directed interventions for individuals contending with these tragic forms of loss.
Supplementary Material
Funding information
American Foundation for Suicide Prevention, Grant/Award Number: PRG-1–020–17
Footnotes
PEER REVIEW
The peer review history for this article is available at https://publons.com/publon/10.1002/jclp.23261
SUPPORTING INFORMATION
Additional supporting information may be found in the online version of the article at the publisher’s website.
DATA AVAILABILITY STATEMENT
The data that support the findings of this study are available from the corresponding author upon reasonable request.
REFERENCES
- Agerbo E (2005). Midlife suicide risk, partner’s psychiatric illness, spouse and child bereavement by suicide or other modes of death: A gender specific study. Journal of Epidemiology & Community Health, 59(5), 407–412. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Asparouhov T, & Muthén B (2014a). Auxiliary variables in mixture modeling: Three-step approaches using M plus. Structural Equation Modeling: A Multidisciplinary Journal, 21(3), 329–341. [Google Scholar]
- Asparouhov T, & Muthén B (2014b). Auxiliary variables in mixture modeling: Using the BCH method in Mplus to estimate a distal outcome model and an arbitrary secondary model. Mplus Web Notes, 21(2), 1–22. [Google Scholar]
- Barry LC, Kasl SV, & Prigerson HG (2002). Psychiatric disorders among bereaved persons: The role of perceived circumstances of death and preparedness for death. The American Journal of Geriatric Psychiatry, 10(4), 447–457. [PubMed] [Google Scholar]
- Berlin KS, Williams NA, & Parra GR (2014). An introduction to latent variable mixture modeling (part 1): Overview and cross-sectional latent class and latent profile analyses. Journal of Pediatric Psychology, 39(2), 174–187. [DOI] [PubMed] [Google Scholar]
- Boelen PA (2015). Peritraumatic distress and dissociation in prolonged grief and posttraumatic stress following violent and unexpected deaths. Journal of Trauma & Dissociation, 16(5), 541–550. [DOI] [PubMed] [Google Scholar]
- Boelen PA, Reijntjes A, Djelantik AMJ, & Smid GE (2016). Prolonged grief and depression after unnatural loss: Latent class analyses and cognitive correlates. Psychiatry Research, 240, 358–363. [DOI] [PubMed] [Google Scholar]
- Bottomley JS, Burke LA, & Neimeyer RA (2017). Domains of social support that predict bereavement distress following homicide loss: Assessing need and satisfaction. OMEGA - Journal of Death and Dying, 75(1), 3–25. [DOI] [PubMed] [Google Scholar]
- Bottomley JS, & Smigelsky MA (2021). Bereavement in the aftermath of suicide, overdose, and sudden-natural death: Evaluating a new measure of needs [Manuscript submitted for publication]. Department of Psychiatry and Behavioral Sciences, Medical University of South Carolina. [Google Scholar]
- Bottomley JS, Smigelsky MA, Bellet BW, Flynn L, Price J, & Neimeyer RA (2018). Distinguishing the meaning making processes of survivors of suicide loss: An expansion of the meaning of loss codebook. Death Studies, 43, 1–11. 10.1080/07481187.2018.1456011 [DOI] [PubMed] [Google Scholar]
- Bottomley JS, Smigelsky MA, Floyd RG, & Neimeyer RA (2019). Closeness and conflict with the deceased: Exploring the factor structure of the quality of relationships inventory in a bereaved student sample. OMEGA-Journal of Death and Dying, 79(4), 377–393. [DOI] [PubMed] [Google Scholar]
- Brent D, Melhem N, Donohoe MB, & Walker M (2009). The incidence and course of depression in bereaved youth 21 months after the loss of a parent to suicide, accident, or sudden natural death. American Journal of Psychiatry, 166(7), 786–794. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Burton AM, Haley WE, & Small BJ (2006). Bereavement after caregiving or unexpected death: Effects on elderly spouses. Aging and Mental Health, 10(3), 319–326. [DOI] [PubMed] [Google Scholar]
- Calati R, Ferrari C, Brittner M, Oasi O, Olié E, Carvalho AF, & Courtet P (2019). Suicidal thoughts and behaviors and social isolation: A narrative review of the literature. Journal of Affective Disorders, 245, 653–667. [DOI] [PubMed] [Google Scholar]
- Cerel J, Maple M, van de Venne J, Moore M, Flaherty C, Brown M, (2016). Exposure to Suicide in the Community: Prevalence and Correlates in One U.S. State. Public Health Reports, 131, (1), 100–107. 10.1177/003335491613100116 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Cerel J, Maple M, van de Venne J, Brown M, Moore M, & Flaherty C (2017). Suicide exposure in the population: Perceptions of impact and closeness. Suicide and Life-Threatening Behavior, 47(6), 696–708. [DOI] [PubMed] [Google Scholar]
- Chugh SS, Reinier K, Teodorescu C, Evanado A, Kehr E, Al Samara M, Mariani R, & Jui J (2008). Epidemiology of sudden cardiac death: Clinical and research implications. Progress in Cardiovascular Diseases, 51(3), 213–228. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Currier JM, Holland JM, & Neimeyer RA (2006). Sense-making, grief, and the experience of violent loss: Toward a mediational model. Death Studies, 30(5), 403–428. 10.1080/07481180600614351 [DOI] [PubMed] [Google Scholar]
- Curtin SC (2019). Trends in cancer and heart disease death rates among adults aged 45–64: United States, 1999–2017. National Vital Statistics Reports 2019, 68(5), 1. https://www.cdc.gov/nchs/data/nvsr/nvsr68/nvsr68_05-508.pdf [PubMed] [Google Scholar]
- Doka KJ. (Ed.). (2002). Disenfranchised grief: New directions, challenges, and strategies for practice. Research Press. [Google Scholar]
- Dyregrov K (2002). Assistance from local authorities versus survivors’ needs for support after suicide. Death Studies, 26(8), 647–668. 10.1080/07481180290088356 [DOI] [PubMed] [Google Scholar]
- Erlangsen A, Runeson B, Bolton JM, Wilcox HC, Forman JL, Krogh J, Shear MK, Nordentoft M, & Conwell Y (2017). Association between spousal suicide and mental, physical, and social health outcomes: A longitudinal and nationwide register-based study. JAMA Psychiatry, 74(5), 456–464. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Feigelman B, & Feigelman W (2008). Surviving after suicide loss: The healing potential of suicide survivor support groups. Illness, Crisis, & Loss, 16(4), 285–304. 10.2190/IL.16.4.b [DOI] [Google Scholar]
- Feigelman W, Feigelman B, & Range LM (2018). Grief and healing trajectories of drug-death-bereaved parents. OMEGA - Journal of Death and Dying, 80, 003022281875466. 10.1177/0030222818754669 [DOI] [PubMed] [Google Scholar]
- Feigelman W, Gorman BS, & Jordan JR (2009). Stigmatization and suicide bereavement. Death Studies, 33(7), 591–608. 10.1080/07481180902979973 [DOI] [PubMed] [Google Scholar]
- Graham JW, Olchowski AE, & Gilreath TD (2007). How many imputations are really needed? Some practical clarifications of multiple imputation theory. Prevention Science, 8(3), 206–213. [DOI] [PubMed] [Google Scholar]
- Guy P, & Holloway M (2007). Drug-related deaths and the “Special Deaths” of late modernity. Sociology, 41(1), 83–96. 10.1177/0038038507074717 [DOI] [Google Scholar]
- Hamdan S, Berkman N, Lavi N, Levy S, & Brent D (2019). The effect of sudden death bereavement on the risk for suicide: The role of suicide bereavement. Crisis: The Journal of Crisis Intervention and Suicide Prevention, 41, 214–224. [DOI] [PubMed] [Google Scholar]
- Hardison HG, Neimeyer RA, & Lichstein KL (2005). Insomnia and complicated grief symptoms in bereaved college students. Behavioral Sleep Medicine, 3(2), 99–111. [DOI] [PubMed] [Google Scholar]
- Hedegaard H, Curtin SC, & Warner M (2020). Increase in suicide mortality in the United States, 1999–2018. National Center for Health Statistics; 2020. NCHS Data Brief No 362. https://www.cdc.gov/nchs/data/databriefs/db362-h.pdf [PubMed]
- Hipp JR, & Bauer DJ (2006). Local solutions in the estimation of growth mixture models. Psychological Methods, 11(1), 36–53. [DOI] [PubMed] [Google Scholar]
- Jordan JR (2008). Bereavement after suicide. Psychiatric Annals, 38(10). [DOI] [PMC free article] [PubMed] [Google Scholar]
- Kaltman S, & Bonanno GA (2003). Trauma and bereavement: Examining the impact of sudden and violent deaths. Journal of Anxiety Disorders, 17(2), 131–147. [DOI] [PubMed] [Google Scholar]
- Keyes KM, Pratt C, Galea S, McLaughlin KA, Koenen KC, & Shear MK (2014). The burden of loss: Unexpected death of a loved one and psychiatric disorders across the life course in a national study. American Journal of Psychiatry, 171(8), 864–871. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Kheibari A, Cerel J, & Victor G (2021). Comparing attitudes toward stigmatized deaths: Suicide and opioid overdose deaths. International Journal of Mental Health and Addiction, 1–15. [Google Scholar]
- Kilpatrick DG, Resnick HS, Milanak ME, Miller MW, Keyes KM, & Friedman MJ (2013). National estimates of exposure to traumatic events and PTSD prevalence using DSM-IV and DSM-5 criteria. Journal of Traumatic Stress, 26(5), 537–547. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Kroenke K, Strine TW, Spitzer RL, Williams JB, Berry JT, & Mokdad AH (2009). The PHQ-8 as a measure of current depression in the general population. Journal of Affective Disorders, 114(1), 163–173. [DOI] [PubMed] [Google Scholar]
- Latham AE, & Prigerson HG (2004). Suicidality and bereavement: Complicated grief as psychiatric disorder presenting greatest risk for suicidality. Suicide and Life-Threatening Behavior, 34(4), 350–362. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Lancel M, Stroebe M, Eisma MC (2020). Sleep disturbances in bereavement: A systematic review. Sleep Medicine Reviews, 53, 101331. 10.1016/j.smrv.2020.101331 [DOI] [PubMed] [Google Scholar]
- Little TD, Jorgensen TD, Lang KM, & Moore EWG (2013). On the joys of missing data. Journal of Pediatric Psychology, 39(2), 151–162. [DOI] [PubMed] [Google Scholar]
- Lo Y, Mendell NR, & Rubin DB (2001). Testing the number of components in a normal mixture. Biometrika, 88(3), 767–778. [Google Scholar]
- Lobb EA, Kristjanson LJ, Aoun SM, Monterosso L, Halkett GK, & Davies A (2010). Predictors of complicated grief: A systematic review of empirical studies. Death Studies, 34(8), 673–698. [DOI] [PubMed] [Google Scholar]
- Marco JH, Alonso S, Baños R (2021). Meaning-making as a mediator of anxiety and depression reduction during cognitive behavioral therapy intervention in participants with adjustment disorders. Clinical Psychology & Psychotherapy, 28(2), 325–333. 10.1002/cpp.2506 [DOI] [PubMed] [Google Scholar]
- Magidson J, & Vermunt J (2002). Latent class models for clustering: A comparison with K-means. Canadian Journal of Marketing Research, 20(1), 36–43. [Google Scholar]
- Miklin S, Mueller AS, Abrutyn S, & Ordonez K (2019). What does it mean to be exposed to suicide?: Suicide exposure, suicide risk, and the importance of meaning-making. Social Science & Medicine, 233, 21–27. 10.1016/j.socscimed.2019.05.019 [DOI] [PubMed] [Google Scholar]
- Milman E, Neimeyer RA, Fitzpatrick M, MacKinnon CJ, Muis KR, & Cohen SR (2019). Prolonged grief and the disruption of meaning: Establishing a mediation model. Journal of Counseling Psychology, 66(6), 714–725. [DOI] [PubMed] [Google Scholar]
- Mitchell AM, Kim Y, Prigerson HG, & Mortimer-Stephens M (2004). Complicated grief in survivors of suicide. Crisis, 25(1), 12–18. [DOI] [PubMed] [Google Scholar]
- Molina N, Viola M, Rogers M, Ouyang D, Gang J, Derry H, & Prigerson HG (2019). Suicidal ideation in bereavement: A systematic review. Behavioral Sciences, 9(5), 53. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Moran M (2020). Board approves new prolonged grief disorder for DSM. Psychiatric News, 9, 221–6. 10.1176/appi.pn.2020.11a12 [DOI] [Google Scholar]
- Muthén LK, & Muthén BO (1998–2017). Mplus user’s guide (8th ed.). Muthén & Muthén. [Google Scholar]
- Neimeyer RA (2019). Meaning reconstruction in bereavement: Development of a research program. Death Studies, 43(2), 79–91. [DOI] [PubMed] [Google Scholar]
- Neimeyer RA, Cerel J, & Maple M (2017). Recommendations for research on suicide loss: A commentary. Death Studies, 1187, 673–679. 10.1080/07481187.2017.1335555 [DOI] [PubMed] [Google Scholar]
- Nylund KL, Asparouhov T, & Muthén BO (2007). Deciding on the number of classes in latent class analysis and growth mixture modeling: A Monte Carlo simulation study. Structural Equation Modeling, 14(4), 535–569. [Google Scholar]
- Osman A, Bagge CL, Gutierrez PM, Konick LC, Kopper BA, & Barrios FX (2001). The suicidal behaviors questionnaire-revised (SBQ-R): Validation with clinical and nonclinical samples. Assessment, 8(4), 443–454. [DOI] [PubMed] [Google Scholar]
- Pitman A, Osborn D, King M, & Erlangsen A (2014). Effects of suicide bereavement on mental health and suicide risk. The Lancet Psychiatry, 1(1), 86–94. 10.1016/S2215-0366(14)70224-X [DOI] [PubMed] [Google Scholar]
- Prigerson H (2004). Complicated grief: When the path of adjustment leads to a dead-end. Bereavement Care, 23(3), 38–40. [Google Scholar]
- Prigerson HG, Horowitz MJ, Jacobs SC, Parkes CM, Aslan M, Goodkin K, Raphael B, Wortman C, Neimeyer RA, Bonanno GA, Block SD, Kissane D, Boelen P, Maercker A, Litz BT, Johnson JG, First MB, & Maciejewski PK (2009). Prolonged grief disorder: Psychometric validation of criteria proposed for DSM-V and ICD-11. PLOS Medicine, 6(8):e1000121. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Prigerson HG, Maciejewski PK, Reynolds CF III., Bierhals AJ., Newsom JT., Fasiczka A., Frank E., & Miller M. (1995). Inventory of Complicated Grief: A scale to measure maladaptive symptoms of loss. Psychiatry Research, 59(1–2), 65–79. [DOI] [PubMed] [Google Scholar]
- Provini JR, Everett CR, & Pfeffer C (2000). Adults mourning suicide: Self-reported concerns about bereavement, needs for assistance, and help-seeking behavior. Death Studies, 24(1), 1–19. [DOI] [PubMed] [Google Scholar]
- Qualtrics LLC (2016). Security statement. [Google Scholar]
- Raftery AE (1995). Bayesian model selection in social research. Sociological Methodology, 25, 111–164. [Google Scholar]
- Reger MA, Stanley IH, & Joiner TE (2020). Suicide mortality and coronavirus disease 2019—A perfect storm? JAMA Psychiatry, 77, 1093–1094. [DOI] [PubMed] [Google Scholar]
- Rose JS, Chassin L, Presson C, Sherman SJ, Stein MD, & Col N (2007). A latent class typology of young women smokers. Addiction, 102(8), 1310–1319. [DOI] [PubMed] [Google Scholar]
- Ross V, Kõlves K, & De Leo D (2019). Exploring the support needs of people bereaved by suicide: A qualitative study. OMEGA - Journal of Death and Dying, 82, 632–645. 10.1177/0030222819825775 [DOI] [PubMed] [Google Scholar]
- Rubin SS (1999). The two-track model of bereavement: Overview, retrospect, and prospect. Death Studies, 23(8), 681–714. [DOI] [PubMed] [Google Scholar]
- Schwartz LE, Howell KH, & Jamison LE (2018). Effect of time since loss on grief, resilience, and depression among bereaved emerging adults. Death Studies, 42(9), 537–547. [DOI] [PubMed] [Google Scholar]
- Schwarz G (1978). Estimating the dimension of a model. The Annals of Statistics, 6(2), 461–464. [Google Scholar]
- Servaty-Seib HL, & Pistole MC (2007). Adolescent grief: Relationship category and emotional closeness. OMEGA - Journal of Death and Dying, 54(2), 147–167. [DOI] [PubMed] [Google Scholar]
- Shear MK, Ghesquiere A, & Glickman K (2013). Bereavement and complicated grief. Current Psychiatry Reports, 15(11), 406. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Shear MK, Simon N, Wall M, Zisook S, Neimeyer R, Duan N, Reynolds C, Sung S, Ghesquiere A, Gorscak B, Clayton P, Ito M, Nakajima S, Konishi T, Melhem N, Meert K, Schiff M, O’Connor MF, First M, … Keshaviah A (2011). Complicated grief and related bereavement issues for DSM-5. Depression and Anxiety, 28(2), 103–117. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Smigelsky MA, Bottomley JS, Relyea G, & Neimeyer RA (2019). Investigating risk for grief severity: Attachment to the deceased and relationship quality. Death Studies, 44, 402–411. [DOI] [PubMed] [Google Scholar]
- Spitzer RL, Kroenke K, Williams JB, & Löwe B (2006). A brief measure for assessing generalized anxiety disorder: The GAD-7. Archives of Internal Medicine, 166(10), 1092–1097. [DOI] [PubMed] [Google Scholar]
- Stephenson J (2020). Drug overdose deaths head toward record number in 2020, CDC warns. JAMA Health Forum, 1(10):e201318. [DOI] [PubMed] [Google Scholar]
- Stroebe M, Schut H, & Stroebe W (2007). Health outcomes of bereavement. The Lancet, 370(9603), 1960–1973. [DOI] [PubMed] [Google Scholar]
- Sveen C -A., & Walby, F. A. (2008). Suicide survivors’ mental health and grief reactions: A systematic review of controlled studies. Suicide and Life-Threatening Behavior, 38(1), 13–29. 10.1521/suli.2008.38.1.13 [DOI] [PubMed] [Google Scholar]
- Tal I, Mauro C, Reynolds CF, Shear MK, Simon N, Lebowitz B, Skritskaya N, Wang Y, Qiu X, Iglewicz A, Glorioso D, Avanzino J, Wetherell JL, Karp JF, Robinaugh D, Zisook S (2017). Complicated grief after suicide bereavement and other causes of death. Death Studies, 41(5), 267–275. 10.1080/07481187.2016.1265028 [DOI] [PubMed] [Google Scholar]
- Templeton L, Ford A, McKell J, Valentine C, Walter T, Velleman R, Bauld L, & Hollywood J (2016). Bereavement through substance use: Findings from an interview study with adults in England and Scotland. Addiction Research and Theory, 24(5), 341–354. 10.3109/16066359.2016.1153632 [DOI] [Google Scholar]
- Valentine C, Bauld L, & Walter T (2016). Bereavement following substance misuse. OMEGA - Journal of Death and Dying, 72(4), 283–301. 10.1177/0030222815625174 [DOI] [Google Scholar]
- Van Ameringen M, Mancini C, Patterson B, & Boyle MH (2008). Post-traumatic stress disorder in Canada. CNS Neuroscience & Therapeutics, 14(3), 171–181. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Wakeman SE, Green TC, & Rich J (2020). An overdose surge will compound the COVID-19 pandemic if urgent action is not taken. Nature Medicine, 1–2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Weathers FW, Litz BT, Keane TM, Palmieri PA, Marx BP, & Schnurr PP (2013). The ptsd checklist for dsm-5 (pcl-5). Scale available from the National Center for PTSD at http://www.ptsd.va.gov,10. [Google Scholar]
- Wisten A, & Zingmark K (2007). Supportive needs of parents confronted with sudden cardiac death—A qualitative study. Resuscitation, 74(1), 68–74. 10.1016/j.resuscitation.2006.11.014 [DOI] [PubMed] [Google Scholar]
- Zisook S, Chentsova-Dutton Y, & Shuchter SR (1998). PTSD following bereavement. Annals of Clinical Psychiatry, 10(4), 157–163. 10.1023/A:1022342028750 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The data that support the findings of this study are available from the corresponding author upon reasonable request.