Introduction
During the end-of-life care phase, family caregivers often suffer from their inner sadness resulting from the impending death. Grief severity in family caregivers before the death of a loved one is referred to as pre-loss grief (Lindauer and Harvath Reference Lindauer and Harvath2014; Nielsen et al. Reference Nielsen, Neergaard and Jensen2017; Singer et al. Reference Singer, Roberts and McLean2022) or pre-death grief (Holm et al. Reference Holm, Årestedt and Alvariza2019). A recent systematic review (Treml et al. Reference Treml, Schmidt and Nagl2021) concluded that caregivers with high levels of pre-loss grief and low levels of preparedness for impending death were associated with poor bereavement adjustment. While the underlying mechanism remains unclear, it is essential to examine possible psychological factors or processes to explain the relationship between pre-loss grief and post-loss adjustment.
The experience of caring for terminally ill cancer patients may involve both challenges and personal growth (Tang Reference Tang2019). Therefore, assessing the positive aspects of care can help caregivers identify the meaning of the experience and better adjust throughout palliative care. A systematic review showed that positive aspects of caregiving include an enhanced relationship with the patient, feeling rewarded, a sense of personal growth, and a perception of personal satisfaction (Li and Loke Reference Li and Loke2013). A recent study also revealed that identifying meaning in the caregiving experience could be a protective factor favoring adaptation (Palacio and Limonero Reference Palacio and Limonero2020). These studies support a potential relationship between the caring experience, the caregiver–patient relationship, and the personal growth of caregivers.
Caregivers’ relationships with terminally ill patients play a crucial role in their subsequent bereavement adjustment. Attachment, the sense of psychological security in an individual’s relationships with others, is fundamental for developing healthy interpersonal relationships (Bowlby Reference Bowlby1969, Reference Bowlby1973, Reference Bowlby1980, Reference Bowlby1982; Wayment and Vierthaler Reference Wayment and Vierthaler2002). In adult attachment research, attachment insecurities have been conceptualized in terms of 2 dimensions: anxiety and avoidance (Cohen and Katz Reference Cohen and Katz2015; Fraley et al. Reference Fraley, Heffernan and Vicary2011; Tsilika et al. Reference Tsilika, Parpa and Zygogianni2015). Evidence indicates that attachment insecurities can affect patients’ and caregivers’ caring quality and well-being (e.g., Nicholls et al. Reference Nicholls, Hulbert-Williams and Bramwell2014; Tsilika et al. Reference Tsilika, Parpa and Zygogianni2015). One recent study (Sękowski et al. Reference Sękowski, Ludwikowska-Świeboda and Prigerson2024) showed that attachment anxiety has a positive relationship with posttraumatic growth via intrusive and deliberate rumination. Research has also explored the possibility of insecure attachment style being a potential risk factor for caregivers to develop prolonged grief disorder (Liljeroos et al. Reference Liljeroos, Krevers and Milberg2024). While a meta-analysis concluded that longitudinal analyses showed neither anxious nor avoidant attachment styles increase prolonged grief symptoms (Eisma et al. Reference Eisma, Bernemann and Aehlig2023), researchers suggested that disorganized attachment style might moderate the effect between avoidant attachment style and prolonged grief symptom severity (Sekowski and Prigerson Reference Sekowski and Prigerson2022b). In sum, caregivers’ attachment insecurities may affect their caring experience, likely extending to their bereavement adjustment.
During the last decade, continuing bonds (CBs), which highlights the importance of continuing relationships with the deceased person during bereavement adjustment, has initiated a revolutionary change in grief theory (Klass et al. Reference Klass, Silverman and Nickman1996). Field et al. (Reference Field, Gao and Paderna2005) proposed an attachment-based CB theory to explain how attachment might interact with the formation of CBs. Individuals with insecure attachment styles tend to establish maladaptive CBs with the deceased person. To elaborate on the theory, Field and Filanosky (Reference Field and Filanosky2009) distinguished CBs into 2 types: internalized and externalized. Internalized CBs, which function as a safe and stable inner resource, result from successful internalization of the deceased. In contrast, externalized CBs are expressed as illusions and hallucinations regarding the physical presence of the deceased person and are a maladaptive means of coping with grief.
To the best of our knowledge, studies investigating the longitudinal effects of pre-loss grief, pre-loss growth, and attachment insecurities on caregivers’ post-loss adjustment have been limited. Prior studies focused primarily on the association between pre-loss grief and prolonged grief symptomatology, and aimed to identify the potential risk factors for prolonged grief disorder in the pre-loss phase (Nielsen et al. Reference Nielsen, Neergaard and Jensen2017; Stroebe et al. Reference Stroebe, Schut and Boerner2010; Zordan et al. Reference Zordan, Bell and Price2019). Limited research has focused on the association between normal pre-loss grief and post-loss grief (Holm et al. Reference Holm, Årestedt and Alvariza2019). Moreover, empirical findings on the association between attachment insecurities and types of CBs have been inconsistent (Ho et al. Reference Ho, Chan and Ma2013; Root and Exline Reference Root and Exline2014; Yu et al. Reference Yu, He and Xu2016). One reason for this inconsistency may involve the cross-sectional nature of most CB studies (Root and Exline Reference Root and Exline2014; Yu et al. Reference Yu, He and Xu2016). Given such, a longitudinal study is needed to elucidate how caregivers’ attachment insecurities with a patient in the caregiving phase may affect caregivers’ grief and CB expressions when they face the death of patients.
Previous studies suggested that relational closeness might be a risk factor for prolonged grief disorder (Harrison et al. Reference Harrison, Windmann and Rosner2022; Sekowski and Prigerson Reference Sekowski and Prigerson2022). However, Taiwanese society values relational harmony over individual fulfillment (Kim et al. Reference Kim, Yang and Hwang2006). Thus, caregivers may benefit more from close relationships during the caring phase than caregivers from non-Confucian cultures. Generally speaking, caring experiences can improve relational closeness between the caregiver and the patient by resolving unfinished business or previous conflicts. Considering this, the present study also measured the relational closeness between the caregiver and the patient during the palliative care phase.
The present study explored the longitudinal effects of pre-loss grief, pre-loss growth, relational closeness, and attachment insecurities on caregivers’ post-loss adjustment. We also assessed caregivers’ CB expressions 6–12 months after the patient’s death with the aim of exploring the association between pre-loss attachment insecurities and post-loss CB expressions. The present study sought to answer the following research questions: (1) What are the longitudinal effects of pre-loss grief, pre-loss growth, relational closeness, and attachment insecurities on post-loss grief and post-loss growth? (2) Are attachment insecurities more associated with externalized CBs, as predicted by CBs theory?
This study measures pre- and post-loss grief to assess the severity of grief rather than prolonged grief symptoms (Prigerson et al. Reference Prigerson, Horowitz and Jacobs2009; Sękowski et al. Reference Sękowski, Ludwikowska-Świeboda and Prigerson2024). Research has shown a strong correlation between grief severity and prolonged grief symptoms (Nielsen et al. Reference Nielsen, Neergaard and Jensen2017). Given such, we suggest that lower grief severity might indicate better adjustment to bereavement.
Methods
Participants
A total of 66 bereaved caregivers (43 females and 23 males) participated in the study. Data collection took place mainly between January 2017 and December 2019 (n = 59), with several additional data collected between July and October 2020 (n = 7). The age of the participants ranged from 18 to 69 years, with a mean age of 46.82 years (SD = 11.67). The majority of the sample was well-educated, and most of the patients being cared for were the parents or spouses of the participants (see Table 1 for the summary of the demographic information).
Demographic information of participants (N = 66)
Measures
Hogan Grief Reaction Checklist
The Hogan Grief Reaction Checklist (HGRC), designed to measure the multiple dimensions of the bereavement process, is a 61-item questionnaire rated on a self-reported 5-point Likert scale (Hogan and Schmidt Reference Hogan and Schmidt2015). In our previous study (Lee et al., Reference Lee, Yu and Lin2024), we had obtained permission from Dr. Hogan to translate the HGRC into the Traditional Chinese Version (TC-HGRC). The TC-HGRC comprises 6 factors corresponding to the items in the original version of the HGRC. The personal growth subscale is scored independently to represent respondents’ growth after the experience of loss. The total grief score is calculated by summing 5 subscales, excluding the personal growth subscale. The test–retest reliability of the TC-HGRC subscales after 1 month were acceptable, (Person’s correlation coefficients: despair, .70; panic behavior, .73; personal growth, .56; blame and anger, .77; detachment, .83; and disorganization, .82, Author et al., 2022). In the present study, the internal consistencies (Cronbach’s alphas) for the 6 factors in the TC-HGRC are as follows: For the pre-loss version: despair, .84; panic behavior, .90; blame and anger, .77; detachment, .80; and disorganization, .83; personal growth, .70. For the total grief score: Cronbach’s α = .95; For the post-loss version: despair, .93; panic behavior, .92; blame and anger, .80; detachment, .89; and disorganization, .88; personal growth, .87. For the total grief score: Cronbach’s α = .97.
To assess the pre-loss state of the caregivers, we slightly modified the wording of 2 items in the TC-HGRC to indicate that the patients were still alive. Caregivers’ pre-loss growth was measured by the subscale of personal growth, which included the following items: “I have learned to cope better with life,” “I feel as though I am a better person,” “I have a better outlook on life,” and “I have more compassion for others.” The scale was used to assess participants’ pre-loss grief (pre-loss phase) and post-loss grief (post-loss phase).
The Experiences in Close Relationships–Relationship Structures Questionnaire
The Experiences in Close Relationships–Relationship Structures Questionnaire (ECR-RS) is a 7-point self-report scale assessing attachment insecurities (i.e., attachment-related avoidance and anxiety) toward significant attachment figures in young adulthood (Fraley et al. Reference Fraley, Heffernan and Vicary2011). The scale comprises 9 items: 6 items measure avoidance, and 3 items measure anxiety. The Traditional Chinese version of the ECR-RS has been found to have satisfactory internal reliabilities (Cronbach’s α = .86–.90 for avoidance; Cronbach’s α = .90 – .91 for anxiety) and test–retest reliabilities (r = .73 for avoidance; r = .70 for anxiety) (Lin Reference Lin2016). In the present study, Cronbach’s α = .82 for avoidance; Cronbach’s α = .66 for anxiety. The scale was used in the pre-loss phase.
The Inclusion of Other in the Self Scale
The Inclusion of Other in the Self scale (IOS) (Aron et al. Reference Aron, Aron and Smollan1992) was used to assess caregivers’ relational closeness with the patients during the pre-loss phase. As a diagram-like measure, the scale contains 7 pairs of overlapping circles, with each pair overlapping slightly more than the preceding pair. Respondents should select 1 out of 7 Venn-like diagrams that best depict their relational closeness with another person. In the present study, we adapted IOS to ask caregivers which best depicts their relationship with the patient during palliative care. The IOS has been shown to possess good test–retest reliability and convergent and discriminant validity (Aron et al. Reference Aron, Aron and Smollan1992). The Chinese version also has acceptable test-retest reliability (r = 0.739, p < 0.01; Ke 2021).
The Continuing Bond Scale
The original Continuing Bond Scale (CBS) (Field and Filanosky Reference Field and Filanosky2009) is a 16-item measure used to identify the ongoing relationships with the deceased patient, and was used in the post-loss phase. The CBS comprises 2 subscales: internalized and externalized CB. In the present study, we adopted the Traditional Chinese version developed by Ho et al. (Reference Ho, Chan and Ma2013). Ho et al. (Reference Ho, Chan and Ma2013) reported that the Cronbach’s alpha of the Traditional Chinese version of the Continuing Bond Scale (C-CBS) was .92, and that the Cronbach’s alphas of the internalized and externalized CB subscales were .92 and .84, respectively. In the present study, Cronbach’s alphas of the internalized and externalized CB subscales were .89 and .79, respectively.
Procedure and ethical considerations
Participants were recruited through referrals from a palliative care team in the cancer center of a hospital. Informed consent was obtained in the end-of-life care phase. The inclusion criteria were as follows: Caretakers who were at least 18 years of age, with no psychiatric diagnosis during the caring phase. The duration of hospice care for the patients cared for ranged from days to months before the patients died, with a median duration of approximately 2 weeks. If there was more than 1 caregiver met the inclusion criteria in a family, only the primary caregiver would be invited to participate in the research. However, the whole family in the palliative and hospice department would be approached and cared for by psychologists during the first few days of the patients’ admission, whether they decided to participate in the current research or not. Individuals who agreed to participate in the study were asked to complete the demographic information and the TC-HGRC (pre-loss version), ECR-RS, and the IOS.
After the patient passed away at least 6 months, psychologists started to contact the caregivers by telephone to provide them with psychosocial support and psychological information on bereavement care. This phone call was conducted between 6 and 12 months after the patient’s death. The psychologists then confirmed with the caregivers that they wished to participate in the follow-up study while they were grieving. They were asked to complete the TC-HGRC and TC-CBS. Among the 96 caregivers, 10 participants declined to participate the follow-up studies, and 20 participants did not respond. The overall response rate was 68.8%.
The Institutional Review Board approval had been obtained for the present study (KFSYSCC-IRB: 20180504A). Moreover, the present study was a part of a multiyear research project sponsored by the Ministry of Science and Technology in Taiwan. The entire project had received approval from a cancer center institutional ethics review board in 2015 (protocol number KFSYSCC-IRB-20150402A).
Results
We used IBM SPSS 25 to conduct the following analyses. First, we conducted correlation and differential analyses to examine the impact of sociodemographic variables (age, gender, educational levels, and relationship types) on caregivers’ post-loss grief and growth. Given that only 1 respondent (n = 1) has a certain level of education, education levels have been transformed into a binary variable, higher education versus less than university education, in the following analyses. The results indicate that there was a marginal but not significant negative correlation between age and post-loss grief. (r = − .23, p = .061), while none of the sociodemographic variables significantly correlated or differed in post-loss grief scores. For personal growth, only gender had a marginal but not significant effect on HGRC-personal growth subscale; that is, women had slightly higher scores on the HGRC-personal growth subscale (M = 51.57, SD = 21.14, F (1, 64) = 3.31, p = .074) in comparison to men (M = 43.13, SD = 9.25).
Then, bivariate correlation analyses were conducted to examine potential correlations among the caregivers’ pre-loss and post-loss variables. Table 2 presents the descriptive statistics and bivariate correlations between the study variables.
Descriptive statistics and bivariate correlations between study variables
* p < 0.05.
** p < 0.01.
Relational closeness: Inclusion of Other in the Self Scale; Pre-loss growth: HGRC-personal growth subscale during end-of-life care; Pre-loss grief: pre-loss version of HGRC; Internalized CB: CBS internalized subscale; Externalized CB: CBS externalized subscale; growth-post: HGRC personal growth subscale after loss; post-loss grief: HGRC scores after loss.
As shown in Table 2, relational closeness during end-of-life care (IOS) was positively correlated with both pre-loss growth (r = .33, p = .007) and post-loss growth (r = .26, p = .037). IOS was negatively correlated with pre-loss grief (r = − .35, p = .004), which was in line with our research hypothesis. Not surprisingly, pre-loss grief was positively correlated with post-loss grief (r = .35, p = .004), and externalized CB was also positively correlated with post-loss grief (r = .37, p = .002). Contrary to our prediction, neither attachment anxiety nor attachment avoidance had a significant correlation with both pre-loss and post-loss grief. For the relationship between attachment insecurities and CB expression, both attachment avoidance and attachment anxiety were not significantly correlated with externalized CBs. Attachment anxiety was negatively correlated with internalized CBs, although this correlation was only marginally significant (r = − .23, p = .077).
We conducted partial correlation analyses to control for the effects of gender and age. This allowed us to examine the possible impact of pre-loss grief, pre-loss growth, attachment insecurities, relational closeness, and CB expressions on post-loss grief and post-loss growth. Results are shown in Table 3. Not surprisingly, pre-loss grief was positively correlated with post-loss grief (r = .30, p = .015). Attachment avoidance (r = − .25, p = .045), but not attachment anxiety (r = − .03, p = .82), was negatively correlated with post-loss grief. Externalized CBs (r = .39, p < .001) but not internalized CBs (r= .20, p = .119) strongly correlated with post-loss grief. There was a significant positive correlation between both pre-loss growth (r = .46, p < .001) and relational closeness during the caring phase (r = .27, p = .035) with post-loss growth. Additionally, pre-loss grief negatively correlated with post-loss growth (r = − .28, p = .024).
Partial correlations between study variables after controlling for demographic variables
* p < .05.
*** p < .001.
Partial correlation control variables: gender and age.
Pre-loss growth: pre-loss version of the Chinese version of the Hogan Grief Reaction Checklist (C-HGRC)-personal growth subscale; Pre-loss grief: pre-loss version of the C-HGRC; Attachment avoidance: the Chinese version of the Experiences in Close Relationships–Relationship Structures questionnaire (ECR-RS)-avoidance subscale; Attachment anxiety: Chinese version of the ECR-RS-anxiety subscale; Relational closeness: the Inclusion of Other in the Self scale; Internalized CB: the Chinese Continuing Bond Scale-internalized subscale; Externalized CB: the Chinese Continuing Bond Scale- externalized subscale; Post-loss growth: HGRC personal growth subscale after loss; Post-loss grief: HGRC scores after loss.
Two hierarchical multiple regressions were conducted. Predicting variables significantly correlated with post-loss grief, namely pre-loss grief, externalized CBs, and attachment avoidance, were included in the first hierarchical multiple regression model. In step 1, the caregiver’s age, which was significantly correlated with post-loss grief, was entered. In step 2, pre-loss grief and attachment avoidance were entered. In step 3, externalized CBs were entered. The analyses revealed that the caregiver’s age had no significant impact on the level of post-loss grief [R 2 = .06 (adjusted R 2 = .04), F (1,63) = 3.66, p = .061]. The addition of pre-loss grief and attachment avoidance significantly improved the model of post-loss grief [R 2 = .21 (adjusted R 2 = .17), F (2,61) = 5.95, p = .004]. Pre-loss grief had a significant impact on post-loss grief, whereas attachment avoidance was negatively significantly associated with post-loss grief. The results of the third step of the analysis revealed that the addition of externalized CBs significantly improved the model of post-loss grief [R 2 = .31 (adjusted R 2 = .26), F(1,60) = 8.36, p = .005]. Age, pre-loss grief, and externalized CBs were significant predictions for a higher level of post-loss grief, whereas attachment avoidance was not. The result indicated that model 3 has provided the best explanation for the variance of post-loss grief. The results of hierarchical multiple regression analyses are summarized in Table 4.
Hierarchical multiple regression analysis of post-loss grief (N = 66)
* p < 0.05.
** p < 0.01.
The second hierarchical multiple regression model included predicting variables significantly correlated with post-loss growth: pre-loss grief, pre-loss growth, and relational closeness. In step 1, the caregiver’s gender, which was significantly correlated with post-loss growth, was entered. In step 2, pre-loss grief and pre-loss growth were entered. In step 3, relational closeness was entered. The analyses revealed that the caregiver’s gender had no significant impact on the level of post-loss growth (R 2 = .05 [adjusted R 2 = .03], F [1,63] = 3.29, p = .075). The addition of pre-loss grief and pre-loss growth significantly improved the model of post-loss growth (R 2 = .28 [adjusted R 2 = .25], F (2,61) = 9.97, p < .001]. Pre-loss growth significantly impacted post-loss growth, whereas pre-loss grief did not (p = .094). The results of the third step of the analysis revealed that the addition of relational closeness did not significantly improve the model of post-loss growth [R2 = .29 (adjusted R 2 = .24), F(1,60) = .35, p = .556], indicating that model 2 has provided the best explanation for the variance of post-loss growth. The results of hierarchical multiple regression analyses are summarized in Table 5.
Hierarchical multiple regression analysis of post-loss growth (N = 66)
** p < 0.01.
*** p < 0.001.
Discussion
The present study investigated the longitudinal effects of caregivers’ pre-loss grief, relational closeness, and attachment insecurities with the patients they cared for during the end-of-life care phase. We assessed caregivers’ pre-loss grief, pre-loss growth, attachment insecurities, and relational closeness during the palliative care phase. We also measured the caregiver’s post-loss grief, post-loss growth, and CB expressions 6–12 months after the patient’s death. The main findings are as folows: (1) Correlational analysis showed that caregivers’ level of attachment insecurities during the palliative care phase did not significantly correlate with CB expressions. Only attachment avoidance shows a negative correlation with post-loss grief, controlling for the caregiver’s age. (2) Hierarchical multiple regression analyses revealed that the caregiver’s age, externalized CBs, and pre-loss grief were significant predictors of post-loss grief. In contrast, attachment avoidance had no significant impact on post-loss grief. (3) For the hierarchical multiple regression model on post-loss growth, only pre-loss growth had a significant impact on post-loss growth.
First, attachment insecurities were not significantly associated with CB expressions and post-loss grief. Although attachment avoidance had a mild negative correlation with post-loss grief, there was no significant association between attachment avoidance and post-loss grief in the final regression model. The findings of this study are consistent with those of Eisma et al. (Reference Eisma, Bernemann and Aehlig2023), which suggested that longitudinal analyses showed inconsistent evidence that insecure attachment styles lead to more grief symptoms. This calls for a reconsideration of the role of adult attachment in modern grief theories. The study also suggests that in a culture with deeply rooted Confucian values, conflicted family relationships (Kissane et al. Reference Kissane, McKenzie and Bloch2006), as opposed to dyadic attachment relationships, may constitute stronger risk factors. We propose that future studies consider including the measurement of family relationships in the palliative care phase to examine the potential influence of family dynamics on the caregiver’s post-loss grief adjustment, thereby guiding future research and practice in this area.
Second, the results of our study support a significant relationship between externalized CBs and post-loss grief. Externalized CBs may serve as a strong predictor of post-loss grief because of the following 2 reasons: (1) Externalized CBs is an indicator of unresolved loss and complicated relationship with the deceased, as defined in the CBs theory and adult attachment literature (Stroebe et al. Reference Stroebe, Schut and Boerner2010). Because bereaved individuals may become dependent on their hallucinations and illusions of the deceased patient, externalized CBs can be viewed as a maladaptive coping strategy and a potential risk factor in the development of prolonged grief disorder in bereaved caregivers. (2) Externalized CBs may be a comforting behavior triggered by intense grief reactions. In such cases, grief is the cause of externalized CB expressions. It is worth noting that 2 new CB models have been proposed since 2020. One captures the overt behavior to continue bonds with the deceased person (Eisma & Nguyen, Reference Eisma and Nguyen2023), and another distinguishes symbolic and concrete CB to indicate normal and pathological grief, respectively (Sekowski Reference Sekowski2021). More studies are needed to explore the relationship between these new CB concepts and grief adaptation.
Finally, our results support a significant relationship between pre- and post-loss growth. While relational closeness did not significantly impact post-loss growth in our final regression model, a previous study (Lee et al. Reference Lee, Cheng and Hou2023) found that the caregiver’s relational closeness with the patient during palliative care significantly predicts the caregiver’s pre-loss growth. A possible explanation might be that the close relationship between the caregiver and the patient results in greater pre-loss growth for the caregiver. This, in turn, could lead to increased post-loss growth for the caregiver after the patient has passed away. Nevertheless, more research evidence is needed to support this explanation. Future studies could adopt a longitudinal design to examine whether the caregiver’s pre-loss growth could mediate the relationship between relational closeness and post-loss growth.
Study limitations
The first limitation of this study is its small sample size. This might be the reason why attachment insecurities were not significantly correlated with pre-loss and post-loss grief severity. Second, we did not gather data on the duration of patients’ hospice care and the specific timing of post-loss variable measurements. Third, the sample is dominated by university-educated females who have experienced the loss of a parent. This severely limits the generalizability of the results to other populations. Finally, we measured caregivers’ grief at only 1 follow-up time point. Future follow-up, longitudinal studies with grief assessments at multiple time points should be conducted to elucidate the mechanism of the transformation of caregivers’ relationships with the deceased patient as well as the potential mechanism underlying the association between CB expressions and grief adaptation.
Clinical implications
Despite the limitations, this longitudinal study proves that the relationship quality between the caregiver and the patient is critical to caregivers’ post-loss adjustment. The strong predictability of externalized CB on the severity of post-loss grief also indicates a possible risk factor for prolonged grief disorder. Breen et al. (Reference Breen, Aoun, O’Connor, Johnson and Howting2020) revealed that it takes 9–10 months for caregivers to adapt to the impact of caregiving and bereavement, highlighting the need for palliative care services to support family caregivers in the caring phase and bereavement. Echoing with Breen et al. (Reference Breen, Aoun, O’Connor, Johnson and Howting2020), we suggest that psychosocial intervention routinely assesses the caregiver’s pre-loss grief and relationship quality with the patient. Besides, follow-up support might also be needed between 6 and 12 months after the death. An early intervention and follow-up psychosocial support for at-risk caregivers could help ameliorate the severity of post-loss grief and facilitate the bereavement adjustment.
Conclusions
Caregivers’ pre-loss grief and continuing relationship with the patient/deceased are associated with their following post-loss adjustment. Our study demonstrated that the caregiver’s pre-loss grief and externalized CB expression could significantly predict the caregiver’s post-loss grief severity; pre-loss growth was significantly associated with the caregiver’s personal growth after the patient had passed away. Our findings have highlighted the importance of caregivers’ relational quality with the patient regarding caregivers’ bereavement adjustment.
Data availability statement
The data that support the findings of this study are openly available in OPENICPSR at 10.3886/E193181V1.
Acknowledgments
This study is supported by the National Science and Technology Council [111-2410-H-030-084 and 112-2410-H-030-082-MY2].
Open access
