Prevalence, Correlates, and Psychiatric Burden of Prolonged Grief Disorder in U.S. Military Veterans: Results From a Nationally Representative Study

Abstract

Objectives:

To examine the point prevalence and correlates of prolonged grief disorder (PGD) in a nationally-representative sample of United States (U.S.) veterans.

Methods:

Data were analyzed from the National Health and Resilience in Veterans Study, a nationally representative study of 2,441 U.S. veterans.

Results:

A total of 158 (weighted 7.3%) veterans screened positive for PGD. The strongest correlates of PGD were adverse childhood experiences, female sex, non-natural causes of death, knowing someone who died from coronavirus disease 2019, and number of close losses. After adjusting for sociodemographic, military, and trauma variables, veterans with PGD were 5-to-9 times more likely to screen positive for post-traumatic stress disorder, major depressive disorder, and generalized anxiety disorder. After additional adjustment for current psychiatric and substance use disorders, they were 2–3 times more likely to endorse suicidal thoughts and behaviors.

Conclusions:

Results underscore the importance of targeting PGD as an independent risk factor for psychiatric disorders and suicide risk.

INTRODUCTION

Death of a loved one or close other is one of the most stressful and universal life experiences. After experiencing a loss, some individuals may develop prolonged grief disorder (PGD),¹ which was first included as a formal diagnosis in the DSM-5-TR in 2022, and is characterized by persistent loss-related emotional pain and grief that causes clinically significant distress or functional impairment.¹

Military veterans represent a population at heightened risk for PGD, given their exposure to excessive deaths due to advanced age, combat deployments, and high prevalence of suicide.² To date, however, no known population-based study has examined the prevalence and factors associated with PGD in U.S. veterans.

To address this gap, we analyzed data from a contemporary, nationally representative sample of U.S. veterans to examine the point prevalence and correlates of PGD. We were particularly interested in identifying key correlates of PGD, and quantifying the association between PGD and suicidal behaviors, above and beyond major psychiatric disorders.

METHODS

Participants

Data were analyzed from the National Health and Resilience in Veterans Study, which surveyed a nationally representative sample of 2,441 U.S. veterans. Additional information about the study is available in the Supplement.

Assessments

Prolonged Grief.

Participants were asked, “Have you ever experienced a death of someone close?” Among those who answered affirmatively, an anchoring question was then asked to think about the person who was most difficult to lose. PGD symptoms were assessed using the 5-item Brief Grief Questionnaire (BGQ).³ Higher scores indicate greater PGD symptoms (range: 0–10; Cronbach’s α = 0.85). PGD was operationalized as 1) loss at least 12 months ago; and 2) score ≥5.

Major depressive disorder (MDD).

MDD symptoms were assessed using the two depressive symptoms of the Patient Health Questionnaire (PHQ)-4. A score ≥3 was indicative of a positive screen for MDD (Cronbach’s α = 0.88).⁴

Post-traumatic stress disorder (PTSD).

PTSD symptoms were assessed with the PTSD Checklist for DSM-5.⁵ A score ≥ 33 was indicative of a positive screen for PTSD (Cronbach’s α = 0.96).

Generalized anxiety disorder (GAD).

GAD symptoms were assessed using the two generalized anxiety symptoms of the PHQ-4. A score ≥3 was indicative of a positive screen for GAD (Cronbach’s α = 0.87).⁴

Past-year suicidal ideation.

Past-year suicidal ideation was assessed using Question 2 of the Suicide Behaviors Questionnaire-Revised (SBQ-R⁶): “How often have you thought about killing yourself in the past year” Response options: Never; Rarely (1 time); Sometimes (2 times); Often (3–4 times); and Very Often (5+ times). Positive endorsement was operationalized as any frequency of suicidal ideation (1–5+ times).

Supplemental Table 1 describes additional measures used in this study.

Data Analysis

Independent-samples t tests and χ² analyses were conducted to compare characteristics of veterans with and without PGD. Multivariable logistic regression analyses were then conducted to identify factors that independently differentiated veterans with and without PGD, as well as independent associations between PGD and mental health measures. For the former analysis, a relative importance analysis⁷ was also conducted to quantify the relative contribution of each significant variable to the model-explained variance in PGD.

RESULTS

The total sample was, on average, 63.2 years old (SD = 15.3), and predominantly male (89.5%), and White, non-Hispanic (77.7%). Of the total 2,441 veterans, 2,339 (weighted 92.7%) reported loss of someone close and 158 (weighted 7.3%, 95%CI = 6.2%–8.4%) screened positive for PGD.

Among veterans who reported a loss of someone close, the point prevalence of PGD was 7.8% (95%CI = 6.7%–9.0%), and 89.2% reported that the loss occurred more than 12 months ago; 6.0% 6–12 months ago; and 4.8% <6 months ago.

Supplemental Table 2 shows characteristics of veterans by PGD status. A multivariable analysis revealed that female sex, Black race, having served in the Marine Corps, more adverse childhood experiences, losses of someone close, experiencing loss related to suicide or homicide, and a personal history of COVID-19 infection and knowing someone who died of COVID-19 complications were positively associated with PGD, while greater education and experiencing loss related to old age/natural causes was negatively associated with PGD.

Figure 1 shows results of a relative importance analysis of significant correlates of PGD. The strongest correlates of PGD were greater adverse childhood experiences (post-hoc analysis: emotional neglect; 28.0% relative variance explained [RVE]), female sex (11.8% RVE), non-natural causes of loss (10.3% RVE), knowing someone who died from COVID-19 complications (10.2% RVE), and number of close losses (9.2% RVE).

FIGURE 1.

Relative importance analysis of significant correlates of positive screen for prolonged grief disorder in U.S. veterans. Note. A relative importance analysis was conducted to determine the proportion of explained variance in a positive screen for prolonged grief disorder accounted for by each independent variable. Error bars represent 95% confidence intervals.

Table 1 shows mental health variables by PGD screening status. In multivariable analyses, PGD was independently associated with greater odds of current PTSD, MDD, and GAD; current suicidal ideation; lifetime non-suicidal self-injury, suicide planning, and suicide attempt; and current mental health treatment.

TABLE 1.

Mental Health Variables by Screening Status for Prolonged Grief Disorder in U.S. Veterans

			Bivariate Analyses			Multivariable Analysis PGD vs. No PGD OR (95%CI)
	No PGD N = 2,283 Weighted 92.7% Weighted Mean (SD) or n (Weighted %)	PGD N = 158 Weighted 7.3% Weighted Mean (SD) or n (Weighted %)	X 2	df	p
Current psychiatric disorders
Post-traumatic stress disorder	84 (4.7%)	53 (38.2%)	277.54	1	<0.001	9.71 (6.29–14.98)c
Major depressive disorder	81 (5.0%)	30 (21.5%)	77.33	1	<0.001	5.16 (3.20–8.31)c
Generalized anxiety disorder	88 (5.3%)	32 (25.6%)	105.92	1	<0.001	5.42 (3.44–8.53)c
Alcohol use disorder	178 (8.8%)	22 (12.8%)	3.08	1	0.079	1.52(0.92–2.53)
Drug use disorder	191 (10.0%)	16(11.7%)	0.52	1	0.47	0.82(0.48–1.39)
Suicidality
Lifetime suicide plan	185 (9.3%)	41 (26.0%)	48.42	1	<0.001	2.02 (1.30–3 15)b
Lifetime suicide attempt	63 (3.8%)	17 (10.7%)	18.77	1	<0.001	1.85 (1.03–3.33)a
Lifetime non-suicidal self-injury	86 (6.8%)	14 (16.9%)	23.60	1	<0.001	2.13 (1.31–3.48)b
Current suicidal ideation	162(7.6%)	31 (22.1%)	43.01	1	<0.001	2.03 (1.26–3.27)b
Current suicidal intent	31 (1.5%)	7 (2.8%)	1.97	1	0.16	1.37(0.47–3.99)
Current mental health treatment
Any mental health treatment	212 (9.9%)	45 (30.1%)	65.80	1	<0.001	2.21 (1.45–3.38)c
Psychotropic medication	185 (8.6%)	42 (28.6%)	72.27	1	<0.001	2.55 (1.66–3.92)c
Psychotherapy or counseling	103 (5.0%)	34 (22.3%)	82.75	1	<0.001	2.30 (1.39–3.83)c

Notes: Odds ratios are adjusted for all demographic, military, and trauma variables that differed by PGD screening status at the p < 0.05 level in bivariate analyses (Supplemental Table 2); analyses of suicidality and mental health treatment utilization variables are additionally adjusted for current major depressive disorder, generalized anxiety disorder, PTSD, alcohol use disorder and drug use disorder. df: degrees of freedom; OR: odds ratio; PGD: prolonged grief disorder; SD: standard deviation; 95%CI: 95% confidence interval.

Statistically significant association:

^ap < 0.05.

^bp < 0.01.

^cp < 0.001.

DISCUSSION

To our knowledge, this is the first study to examine the point prevalence, correlates, and psychiatric correlates of PGD in a nationally representative sample of U.S. veterans. Results revealed that 7.3% of veterans screened positive for PGD. Adverse childhood experiences, female sex, non-natural causes of loss, knowing someone who died from coronavirus disease 2019 (COVID-19) complications, and greater number of close losses were the strongest correlates of PGD. Further, PGD was associated with greater odds of internalizing psychiatric disorders, suicidal thoughts, planning, and attempts, and utilization of mental health treatment, even after adjusting for a broad range of potentially confounding variables.

Results of this study also revealed that more than 90% of U.S. veterans reported a loss of someone close. This may be due in part to the older age of veterans and losses endured during military experience (i.e., deployments, combat exposure). Among the total sample, 7.3% of veterans screened positive for PGD. That the vast majority of veterans have endured the loss of someone close and 1-of-14 screen positive for PGD underscores the importance of screening for PGD symptoms in this population.

Adverse childhood experiences, particularly emotional neglect, were the strongest correlate of PGD. Given the link between adverse childhood experiences⁸ and emotional neglect⁹ on anxious attachment style, which is a risk factor for PGD,¹⁰ it is possible that insecure attachment style may mediate the relation between adverse childhood experiences and PGD risk. That female sex was strongly associated with PGD aligns with previous studies,¹⁰ which found that women are more prone to be affected by certain losses, such as the loss of a child and bereavement in the reproductive phase.¹¹ The association between non-natural causes of death and PGD also aligns with previous work, which found that sudden and/or violent deaths were associated with PGD.¹⁰

It is noteworthy that knowing someone who died from COVID-19 complications was another strong correlate of PGD. One interpretation of this finding is that enduring a recent loss due to COVID-19 complications may have served as a reminder for previous loss, which in turn contributed to PGD risk. Lastly, the finding that enduring a greater number of close losses was associated with PGD accords with previous studies, which similarly found that multiple losses were associated with PGD.¹²

Results of the current study extend prior work¹⁰ showing strong associations between PGD and major psychiatric disorders. Moreover, veterans with PGD were almost two times more likely to endorse current suicidal ideation, and lifetime suicide planning, attempt, and nonsuicidal self-injury relative to veterans without PGD, even after adjusting for sociodemographic, trauma, and co-occurring psychiatric disorders. To our knowledge, this is the first study to demonstrate an increased risk of actual suicide attempt and planning in PGD. Given that clinical trials have found that PGD-specific treatment is superior to other psychiatric treatments in treating PGD and mitigating suicidal thoughts in PGD,¹³ it is likely that veterans with PGD may not be receiving evidence-based treatment. However, given the cross-sectional nature of the study, it is also possible that the co-occurrence of the psychiatric symptoms may have contributed to the development and maintenance of PGD; longitudinal studies are needed to disentangle interrelationships among these variables.

Limitations of this study include the utilization of screening instruments including the primary outcome. Further, due to the cross-sectional nature of the survey, causal associations between PGD and its correlates could not be ascertained. Another limitation is that the BGQ does not anchor questions to a particular time period. Thus, the onset of PGD symptoms cannot be ascertained. Notwithstanding these limitations, results of this study provide the first known population-based data on the point prevalence and correlates of PGD among U.S. veterans. Further research is needed to replicate these findings; evaluate interrelationships between PGD and other psychiatric outcomes; and evaluate the efficacy of interventions targeting PGD in this population.

Highlights.

What is the primary question addressed by this study?

This study examined the population-based prevalence and correlates of prolonged grief disorder (PGD) in U.S. veterans.

What is the main finding of this study?

In this nationally representative study of 2,441 veterans, 158 (weighted 7.3%) screened positive for PGD. Adverse childhood experiences, female sex, non-natural causes of death, knowing someone who died from coronavirus disease 2019, and number of losses were strongly associated with PGD. Veterans with PGD were significantly more likely to screen positive for posttraumatic stress, major depressive, and generalized anxiety disorders; endorse suicidal thoughts and attempts; and utilize mental health treatment.

What is the meaning of the finding?

PGD is prevalent in veterans, and associated with increased psychiatric burden and suicide risk.

DATA STATEMENT

The data have not been previously presented orally or by poster at scientific meetings.