Risk for Complicated Grief after the COVID-19 Death of a Marital Partner in Late Life

Abstract

Objectives:

To identify (1) complicated grief symptom clusters among acutely-bereaved older adults who have lost a spouse to COVID-19: and (2) if spousal death due to COVID-19 increased risk of developing probable PGD.

Methods:

Eighty adults participating in a randomized controlled trial for depression prevention (mean age [± SD] = 70.4 [6.6]) completed the Inventory of Complicated Grief, every three months over a maximum of 15 months. Twenty-four percent (n=19) of participants lost a spouse to COVID-19; 76% (n=61) lost a spouse to other causes of death. Adjusted linear regression examined the associations between COVID-19 bereavement and six symptom clusters: yearning and preoccupation, anger and bitterness, shock and disbelief, estrangement from others, hallucinations, and behavior change.

Results:

Compared to the non-COVID-19 group, the COVID-19 bereaved group reported greater shock and disbelief, hallucinations of the deceased, and estrangement from others. COVID-19 death was also associated with higher risk for probable prolonged grief disorder (PGD) at 12 months (odds ratio = 4.38, p = 0.027).

Conclusions:

Older adults who have lost a spouse to COVID-19 present with specific symptoms of distress and may eventually require clinical care for PGD.

INTRODUCTION

Approximately 290,000 older adults in the U.S. have lost a spouse or life partner to a COVID-19-related death¹. The stress and social isolation associated with COVID-19 have made it more difficult than usual to accept the finality of the loss.² Oftentimes the death was unexpected and traumatic, and the rituals intended to facilitate mourning such as funerals or memorial services were greatly limited due to pandemic-related restrictions. Consequently, COVID-19 deaths were projected to cause an increase in Prolonged Grief Disorder (PGD) among surviving spouses.¹ PGD is characterized by persistent longing or yearning and/or preoccupation with the deceased and diagnosed at 12+ months after the death.³ The complications of PGD in older adults are substantial, including declines in physical health, increases in anxiety and depression, significant sleep problems, increased alcohol use, and suicidal thoughts and behavior.⁴

The aim of this report is to compare rates of prevalent complicated grief symptoms among older adults bereaved either by COVID-19 or by other causes. We further explored the association between COVID-19 death and risk for probable incident PGD over 12 months. This exploratory study will help us understand how older spouses are grieving because of COVID-19 and whether COVID-19 bereavement increases the risk for probable incident PGD. Recent work by Denckla and colleagues (2023)⁵ found that bereavement during COVID-19 was associated with elevations in depression and anxiety among older women. The present study adds to our knowledge of the adverse mental health outcomes after bereavement by examining rates of probable PGD in acutely bereaved men and women.

METHODS

Study Design Overview

This study is a secondary analysis embedded in a parent randomized controlled trial (RCT). We recruited older spousally-bereaved adults using virtual methods (NCT04016896), as previously published.⁶ In brief, the primary objective of the parent Widowed Elders’ Lifestyle after Loss (“WELL”) study is to determine if a digital health intervention reduces depressive symptoms by stabilizing the 24-hr rhythm of sleep, meals, and activity. The control group received enhanced usual care consisting of weekly telephone calls and similar assessment schedules. Intervention and observation were delivered over a 3-month period. We collected psychosocial assessments at baseline and at 3-, 6-, 9-, and 15-months later. This study was approved by the University of Pittsburgh Institutional Review Board, and all participants provided electronic consent via REDCap.

Participants

Participants were 60 years and older who had experienced the death of a spouse or life partner within the previous 12 months. We enrolled participants at risk for major depression due to subthreshold symptoms of depression (defined as a Hamilton Rating Scale for Depression [HRSD] score ≥ 9).⁷ We excluded participants with current DSM-5 diagnosis of syndromal mood or psychosis within the last 12 months; dementia (defined by a Telephone Interview for Cognitive Status score < 19); and acute suicide risk. We also excluded participants taking new antidepressants and benzodiazepines after spousal death. The focus of this report is the study group of 80 adults enrolled into the WELL study between January 2020 and December 2022.

Measures

Participants provided basic demographic information including age, sex, race/ethnicity, and time since loss. They also completed a battery of assessments that measured medical illness burden, depression symptoms, anxiety symptoms, and loneliness. These associated features were not used as covariates because doing so may have over-corrected modeling and obscured potential mediators of risk for PGD.

Complicated grief, the outcome variable in this analysis, was measured using the Inventory of Complicated Grief (ICG).⁸ Nineteen items are scored on a Likert-type scale from 0 (not at all) to 4 (always) with higher scores indicating more pathological grief (range = 0 – 76). We report the six symptom clusters (based on the psychometric characteristics originally reported for the ICG)⁹ for yearning and preoccupation with the deceased, anger and bitterness, shock and disbelief, estrangement from others, hallucinations of the deceased, and behavior change including avoidance and proximity seeking. Probable PGD was defined as ICG scores ≥ 30, interpreted at 12 months following each death when PGD is considered diagnosable by DSM5-TR criteria.

Analyses

Analyses used linear regression models to evaluate the association between COVID-19 death (versus non-COVID-19 deaths) and each complicated grief symptom cluster. To examine patterns of symptoms clusters by time since death, we ran regression models at the assessment date closest to 6-, 12-, and 15 months post-death. Time since death was calculated at each assessment and defined as the number of days between the deceased spouse’s date of death and the surviving spouse’s assessment date. We used the ICG collected during the closest interview visit 6-, 12-, and 15 months after the date of spousal death. We ran separate regression models at each time point because the samples were not identical due to time-varying enrollment (i.e., bereaved spouses were eligible for the WELL RCT up to 12 months after the death). We used logistic regression to examine the association between COVID-19 death and probable PGD at 12 months post-death. We controlled for randomized assignment (intervention or control) in our regression models. We used SPSS (ver. 27) to analyze the data.

RESULTS

The mean age of participants was 70.4 years (SD = 6.6 years); 80% (n=64) were women. Twenty-four percent (19 out of 80) were bereaved by COVID-19 with a mean (SD) time since death of 6.1 (3.4) months at enrollment. The COVID-19 bereaved group had higher ICG scores compared the non-COVID-19 group (30.4 versus 21.8; F_{(1, 75)} = 9.10, p = .004). We observed no differences in mean age, proportion of females, medical comorbidity, loneliness, depression, or anxiety between the two groups (see Supplemental Table 1).

In terms of individual symptom clusters, COVID-19 death was significantly associated with shock and disbelief at 6- and 12 months post-death; hallucinations of the deceased at 6-, 12- and 15 months post-death; and estrangement from others at 15 months post-death (see Table 1 and Supplemental Figures 1 – 3). In the COVID-19 death group 44% met criteria for probable PGD compared with 13% in the non-COVID-19 bereaved group (X² = 6.887, p = 0.014). In multivariable regression, COVID-19 death was strongly associated with probable PGD at 12 months post-death (OR=4.38 [95% CI = 1.19 – 16.14] p = 0.027).

Table 1.

Complicated grief symptom clusters associated with COVID-19 death at 6-, 12-, and 15-months post-loss.

Model	Unstandardized		Standardized	t	p-value	95% Confidence Interval for B
	B	Std. Error	Beta			Lower Bound	Upper Bound
Yearning and preoccupation
6 months	1.535	1.240	0.173	1.238	0.222	−0.955	4.025
12 months	1.405	1.176	0.160	1.195	0.237	−0.949	3.760
15 months	1.322	1.260	0.150	1.049	0.299	−1.208	3.851
Anger and bitterness
6 months	1.027	0.610	0.222	1.685	0.098	−0.197	2.251
12 months	0.541	0.55	0.130	0.983	0.330	−0.560	1.641
15 months	0.814	0.659	0.200	1.236	0.224	−0.519	2.147
Shock and disbelief
6 months	2.644	1.092	0.318	2.422	0.019	0.452	4.835
12 months	2.387	0.854	0.350	2.794	0.007	0.676	4.098
15 months	1.658	0.841	0.260	1.97	0.054	−0.030	3.346
Estrangement from others
6 months	1.122	0.634	0.243	1.770	0.083	−0.151	2.395
12 months	0.892	0.508	0.220	1.754	0.085	−0.126	1.910
15 months	1.593	0.580	0.36	2.747	0.008	0.429	2.757
Hallucinations of the deceased
6 months	0.974	0.289	0.422	3.371	0.001	0.394	1.554
12 months	0.982	0.387	0.320	2.535	0.014	0.206	1.758
15 months	1.304	0.376	0.440	3.467	0.001	0.549	2.059
Behavior change
6 months	0.262	0.659	0.056	0.397	0.693	−1.062	1.586
12 months	0.631	0.517	0.160	1.219	0.228	−0.405	1.667
15 months	0.305	0.510	0.080	0.598	0.552	−0.718	1.329

Notes. Bolded values are significant at the p < .05 level. All models controlled for group assignment (intervention or control).

DISCUSSION

We examined complicated grief symptom clusters in older adults who had lost their spouse to COVID-19. We demonstrated that after adjusting for treatment assignment, older adults bereaved by COVID-19 were more likely to report three symptom clusters: shock and disbelief, hallucinations of the deceased, and estrangement from others. These findings are consistent with several other studies showing more severe pathological grief reactions among adults bereaved by COVID-19 compared to adults bereaved by other causes of death.^10–12

It is not surprising that feelings of shock and disbelief over the death were prevalent among the COVID-bereaved. COVID-19 deaths were sudden, unexpected, and random.¹ Hallucinatory symptoms are less commonly reported among the bereaved, but highly specific and tend to map onto the highest level of PGD severity.⁹ Hallucinations were persistent at 6, 12, and 15 months after the loss. Future research should examine whether persistent hallucinations in the COVID-19 bereaved are a reliable risk for the development of PGD. The third symptom cluster – estrangement from others – may be a result of the cultural stigma of death due to COVID-19. Others may have questioned whether the deceased person was to blame for his or her own death due to pre-existing conditions, being overweight, and/or refusing to get vaccinated. This stigma may have forced surviving spouses to cut ties with family and friends who were not willing to acknowledge the reality of COVID-19. The COVID-bereaved may have also been COVID-positive themselves and consequently social distancing from family and friends. Estrangement from others could have also been due to social isolation and changes in social connectedness during the COVID-19 pandemic.¹³

The “grieving mind/brain” must learn many new tasks, while also accepting the finality of loss and enacting a new life in the absence of the loved one.¹⁴ These three symptom clusters may be obstacles to such learning and to integrated grief, rather than prolonged (maladaptive) grief. It is worth noting that yearning for the deceased was similar between the two groups. This symptom cluster is a gateway for criterion B of DSM-5 PGD. If criterion B is not met, then a diagnosis of PGD is not indicated. Persistent longing for the deceased has been reported in cases of sudden violent or accidental deaths. While COVID-19 death was sudden, persistent longing for the deceased (that becomes maladaptive) may only be prevalent among individuals who experienced traumatic distress. Our results suggest that shock/disbelief, hallucinations, and estrangement from others reflect the challenges to be met in leaning to accept the finality of a COVID-19 loss.

We also found that older adults bereaved by COVID-19 were over 4 times more likely to screen positive for probable PGD at 12 months from loss. Clinical attention should therefore be devoted to screening and providing care options for COVID-bereaved older adults. Given the frequent co-occurrence of PGD, depression, and post-traumatic stress, care needs to be tailored to potentially address any of these conditions. Prolonged Grief Disorder Therapy is an evidence-based psychotherapy that addresses all three conditions. Improving social connections with family/friends may be another safeguard to reduce symptoms of PGD.¹⁵ Additionally, providers should be made aware that hallucinations in PGD are not a form of psychotic illness that would require antipsychotic pharmacotherapy. Providers should inquire about it in individuals who are COVID-bereaved as patients may be less likely to disclose hallucinations and/or be alarmed and ashamed by them. Given the lag in time required to diagnose PGD and to clearly distinguish it from normal grief, providers may need to re-screen for symptoms at several time intervals.

We acknowledge several limitations. First, our sample was predominantly white, female, and from Pennsylvania. Our results may not generalize to other racial/ethnic groups. Given the well-characterized disparities in COVID-related deaths in minoritized race and ethnic groups^16,17, more research is needed on racial disparities in bereavement due to COVID-19 and implications for mental health. Second, the number of statistical tests performed results in an increased risk for Type 1 error. Third, the wide confidence interval reported in the logistic regression analysis could be a result of the small sample size and/or variation in the population. Given the exploratory nature of the study, it suggests that further information and future research is needed to better understand the effect. We were also underpowered for treatment group analysis which may have affected the observations. Fourth, we did not include other types of relationship loss. Finally, we used the ICG as a screening tool for PGD and did not use the DSM-5-TR to screen and diagnosis PGD. The ICG has largely been supplanted by the PG-13-R - a symptom inventory aligned with DSM-5TR diagnostic criteria.¹⁸ The PG-13R- is a more up-to-date screening tool than the ICG. Finally, Given the exploratory nature of this study, our findings should be interpreted with caution.

In conclusion, losing a spouse to COVID-19 during late-life was associated with specific complicated grief symptom clusters not seen in non-COVID-19 bereaved. Our sample was solely from the United States; therefore, observations from an international or more diverse sample may have been different. This symptom profile may signal risk for a more severe course, possibly resulting in PGD. PGD is highly comorbid with post-traumatic stress, anxiety disorders, and depression. Future research is needed to better understand the association between COVID-19 bereavement and these psychiatric disorders. Because effective treatment exists, it is clinically appropriate to screen for PGD in COVID-bereaved persons and to make appropriate referral for clinical care in those meeting diagnostic criteria for PGD.

Highlights.

• What is the primary question addressed by this study? We compared complicated grief symptom clusters in acutely-bereaved older adults who either had or had not lost a spouse to a COVID-19- related death.

• What is the main finding of this study? Older adults bereaved by COVID-19 reported more shock and disbelief, more hallucinations of the deceased, and more estrangement from others compared to older adults bereaved by non-COVID-19 deaths. COVID-19 bereavement was also associated with a high risk for probable prolonged grief disorder (PGD) by 12 months from the loss.

• What is the meaning of the findings? Older adults who have lost a spouse to COVID-19 have specific symptoms of distress and may eventually require clinical care for PGD.