Introduction
Death and dying in intensive care units (ICUs) is common [
1,
2] and increasing [
3] worldwide and even more so given COVID-19 deaths often occur in ICUs [
4]. The uncertain terminal trajectory of critical illness, the frightening nature of aggressive, life-prolonging care that threatens the patient’s bodily integrity [
5], and ultimately the loved one’s frequently unexpected death predispose family members to increased risks for post-intensive care syndrome (PICS-F) [
6] during bereavement, including symptoms of anxiety [
7‐
16], depression [
7‐
20], and post-traumatic stress disorder (PTSD) [
7,
9‐
21]. Grief reactions may reciprocally interfere with one another [
22]. Psychological distress commonly co-occurs as depression and anxiety [
23]; depression and PTSD [
24]; or anxiety, depression, and PTSD [
25] to synergistically impair physical and mental-health functioning [
23,
24,
26,
27], survival [
28], and treatment outcomes [
29]. Therefore, recognizing the potential of comorbid psychological distress and identifying modifiable risks precipitating their co-occurrence, trajectories, and future onset is paramount for improving psychological well-being of ICU decedents’ family members.
One potential mechanism for comorbid psychological distress is that preexisting psychological distress increases the risk of subsequent other psychological distress [
30]. Knowledge about the temporal relationships among anxiety, depression, and PTSD elucidates how comorbidity evolves and has important clinical implications for prevention or treatments/management of those types of psychological distress [
30,
31]. However, only one study was found to investigate the temporal reciprocal relationships on how symptoms of anxiety, depression, and PTSD influence each other over time among war veterans [
25]. Therefore, the purpose of this study was to longitudinally determine the temporal reciprocal relationships among symptoms of anxiety, depression, and PTSD of ICU decedents’ family surrogates over their first two bereavement years to examine the potential mechanism of comorbidity of these three types of psychological distress.
Discussion
We observed that examined psychological-distress levels were markedly stable over the first 2 bereavement years in line with a report for war veterans over their first two years post war [
25]. Further, we found different patterns of temporal relationships between symptoms of depression and PTSD over the first 2 bereavement years. Depressive symptoms shortly postloss predicted PTSD symptoms over the first bereavement year, whereas long-lasting PTSD symptoms predicted prolonged depressive symptoms beyond the first bereavement year. The latter result was supported by a report [
25] that PTSD at 1 year post war predicted depression at 2 years post war among veterans, but not vice versa. Furthermore, Glad and colleagues [
44] found that PTSD symptoms within the first bereavement year do not predict later grief, whereas beyond this first year, more pervasive PTSD symptoms predict grief at 30–32 months postloss. Both studies [
25,
44] confirmed that long-lasting PTSD symptoms disrupt the normal grief process and precipitate prolonged depressive (grief) symptoms beyond the first bereavement year.
The temporal relationships among psychological-distress symptoms observed in this study can be understood through the theoretical frameworks of multidimensional grief theory, stress sensitivity, and the internalizing dimension of PTSD. Multidimensional grief theory characterizes grief reactions as adaptive or maladaptive responses across three content domains: separation distress, existential/identity distress, and circumstance-related distress [
22,
45]. Separation distress is characterized by missing the deceased, sadness from persistent separation, and sorrow over the deceased’s failure to physically reunite with oneself. Losing a longstanding relationship during ICU care and starting a future without the patient trigger separation distress and predispose family surrogates to suffering depressive symptoms and separation anxiety.
Because these grief reactions typically recede over time [
46], we draw from the theory of stress sensitivity [
47] to explain why subsequent PTSD symptoms emerge among individuals with symptoms of depression. Stress sensitivity proposes that individuals vulnerable to psychopathology may be more psychologically sensitive to stress and more susceptible to perceive stressful events as highly traumatic. Therefore, the psychological struggle among depressed family surrogates shortly after the death of their loved one may sensitize them to heightened negative responses towards proximal stressors during bereavement like loss reminders (i.e., cues that evoke memories of the deceased, their eternal absence) and traumatic memories of the ICU-care experience [
48]. Sensitization to bereavement stressors may explain PTSD symptoms ensuing from depressive symptoms over the first bereavement year.
These PTSD symptoms may remain stable during bereavement due to maladaptive manifestations of separation distress [
22,
45]. Separation distress has been theorized to manifest from loss reminders and to be exacerbated by maladaptive coping with loss reminders [
45]. As an example of maladaptive coping, excessive behavioral or cognitive avoidance of loss reminders can interfere with remembering, reminiscing, and accepting the reality of the patient’s death [
48], leading to long-lasting PTSD symptoms beyond the first bereavement year.
Co-occurrence of PTSD with major depressive and anxiety disorders has been explained by the internalizing dimension of PTSD [
49]. Internalized trauma may also explain why PTSD symptoms precede subsequent anxiety and depressive symptoms in the second bereavement year. When bereaved family surrogates suffer from long-lasting PTSD symptoms, they may internalize their traumatic bereavement to manifest the existential/identity distress proposed by multidimensional grief theory [
22,
45]. In other words, they may feel excessive guilt about the death, blame themselves for the occurrence of the death (self-denigration), and may not find meaning in the patient’s death, leading to negative beliefs about themselves, hopelessness, despair, a sense of a future blighted by the death, or a disparaged worldview [
22,
45]. Therefore, when bereaved family surrogates internalize their long-lasting PTSD symptoms, they may suffer prolonged depressive symptoms subsequently in the second bereavement year. The same speculation can be applied to the associations between the prior wave of PTSD symptoms with subsequent heighted anxiety symptoms (e.g., like distress, disgust, and anger) in the second bereavement year.
We found that symptoms of anxiety in the prior wave of assessment consistently predicted symptoms of depression and PTSD at the first and second anniversary of bereavement (Fig.
2). For bereaved with anxiety symptoms [
50], loss reminders as the anniversary of the patient’s death approached may evoke circumstance-related distress, or intrusive distressing memories of the traumatic circumstances under which their loved one died [
22,
45]. Marked avoidance of such distressing intrusive memories of the loss may be elicited. Circumstance-related distress also involves feelings of helplessness, guilt, regret, or anger over not being able to save their loved one at the time of his/her death or the horrific circumstances under which the patient died. Exposure to loss reminders near the anniversary of the patient’s death also evokes bereaved caregivers’ separation distress [
22,
45] to trigger their depressive symptoms as previously discussed. Therefore, bereaved family surrogates who had higher symptoms of anxiety in the prior wave of assessment [
49] suffered higher depressive and PTSD symptoms in the subsequent anniversary of bereavement.
Symptoms of depression in the prior wave of assessment predicted symptoms of anxiety at both 3 and 6 months postloss during the first bereavement year, whereas symptoms of anxiety predicted symptoms of depression at 3 months postloss only. Therefore, a bidirectional relationship between symptoms of anxiety and depression was observed between 1 and 3 months postloss. Our results confirmed the conclusion from a meta-analysis that anxiety and depression are bidirectional risk factors for one another [
51] at early bereavement. Furthermore, symptoms of depression at 3 months postloss continuously predicted symptoms of anxiety at 6 months postloss. As previously highlighted, the psychological struggle among depressed family surrogates shortly after the death of their loved one may sensitize them [
46] to a wide range of negative emotional responses like fear, distress, disgust, and anger [
52] toward the patient’s dying process and the permanent separation from their beloved, thereby suffering heighted symptoms of anxiety.
Several limitations of our study are acknowledged. Whether our findings can be generalized to (inter)national populations beyond the sampled hospitals, family surrogates of ICU patients who died within 3 days of admission, or family surrogates who did not participate in or withdrew from bereavement surveys warrants further validation. Measuring symptoms of anxiety, depression, and PTSD by screening rather than diagnostic tools likely overestimates bereaved family surrogates’ psychological distress but avoids overlooking their need for emotional support. Factors that are associated with, mediate, or moderate the interrelationships between symptoms of anxiety, depression, and PTSD have not yet been explored. Further investigation is warranted.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.