Abstract
This study identified the trajectories of instrumental and emotional perceived social support (PSS) from 1996 to 2003 and investigated the associated risk of mortality in bereaved older adults in Taiwan. The study analyzed 1,188 bereaved older adults who had experienced loss of a spouse, a child, or both before 1996 from the Taiwan Longitudinal Study on Aging. More favorable trajectories of PSS are associated with lower mortality risk in bereaved older adults. Compared with the spousal or dual bereavement, the parental bereavement benefited more from consistently high instrumental PSS. The present study revealed that consistently high emotional PSS had a stronger protective effect on mortality risk for a widow or widower than did consistently high instrumental support. The findings of this study can serve as an empirical reference to inform social policies and clinical practices for bereaved older adults in culturally similar societies.
The death of a spouse or child is a stressful life event that not only necessitates psychosocial adjustment throughout the remainder of a person’s life but also is associated with increased risk of mortality partially because of grief responses and the disruption of the person’s social network and social support system (Ennis & Majid, 2021; Moon et al., 2011; Stroebe et al., 2007). Heightened risk of mortality has been reported not only immediately after such deaths (Moon et al., 2011) but also as long as 20 years later (Shor et al., 2012; Song et al., 2019). Research has suggested that meaningful social support from relatives and close friends can be beneficial during bereavement (Burke & Neimeyer, 2013; Scott et al., 2020). Thus, a stable, persistent, and adequate social support tends to favor the adaptation process required to cope with the distress of losing a loved one and therefore to reduce the related mortality risk among older people (Hill et al., 2016; Powers et al., 2014). However, little is known about the relationship between a variety of social support trajectories following different types of bereavement and the risk of mortality. Because many countries worldwide are facing a rapidly aging population and because aging is associated with an increased risk of experiencing multiple stressful life events (Hardy et al., 2002), greater attention needs to be paid to the need of support for older adults, particularly those who have experienced bereavement in early old age.
The literature suggests that the elevated risk of mortality associated with bereavement exhibits two pathogenetic mechanisms: (1) physiological changes associated with distress and grief caused by the loss of a loved one and (2) behavioral changes that necessitate health maintenance or chronic illness management (Jacobs & Ostfeld, 1977; Moon et al., 2011; Seiler, et al., 2020). The physiological change mechanism indicates that the toxic stress of grief directly impairs physiological functions, whereas the behavioral change mechanism indicates that the breakdown of an individual’s social support network following a loss of a family member might result in insufficient instrumental and emotional support and therefore a lack of encouragement to engage in healthy behaviors. Therefore, changes in the compositions of social relationships due to the death of a family member in older adults lead such adults to face various challenges after the bereavement (Arbuckle & de Vries, 1995; Osterweis et al., 1984; Weiss, 2001). The adjustment mechanisms for coping with the loss of a spouse might differ from those for coping with the loss of a child (Lieberman, 1989). For example, a widow or widower must emotionally and cognitively adapt to their new singlehood, which involves not only developing a new routine but also the challenge of investing in new social relationships and restoring meaning in their lives (Vähäkangas et al., 2021). Prolonged emotional difficulty was the most common feature after spousal bereavement (Szabó et al., 2020). By contrast, from the framework of intergenerational solidarity, parents have expectation about filial responsibilities (Bengtson & Robert, 1991). That is, older adults expect that their adult children are obligated to provide support if they need it. Therefore, the experience of losing a child could be perceived as the worst event of a person’s life because it conflicts with their lifecycle expectations and is compounded by a multitude of other losses, such as the losses of caregivers and financial security (Floyd et al., 2013; Rostila et al., 2012; Van Humbeeck et al., 2013; Wang & Hu, 2021). Moreover, studies have increasingly noted that repeated episodes of bereavement can amplify the adverse effects of bereavement and that older adults are the societal most vulnerable group to multiple bereavements (Infurna & Mayer, 2019; Luhmann & Eid, 2009). Accordingly, an understanding of the adaptation to different types of bereavement can serve as a crucial reference to inform social policies and clinical practices.
Among the factors that can promote resilience in the face of loss, emphasis should be placed on the role of social support. Social support is often conceptualized based on its structure (i.e., social networks) and function (i.e., exchange activities) (Holt-Lunstad et al., 2010; Magrin et al., 2014; Leahy-Warren, 2014; Mondesir et al., 2018). During or after stressful events, functional social support plays a more crucial role than does its structural counterpart (Barth et al., 2010; Charuvastra & Cloitre, 2008; Cohen, 2004; Li et al., 2022). Functional support can be conceptualized as the perception of an exchange involving a social network (i.e., perceived social support [PSS]) and comprises instrumental and emotional support (Cohen, 2004). Instrumental support refers to assistance from family members or friends in the form of handling external tangible demands, such as providing access to health care, buying groceries, cooking, cleaning, and bill paying. By contrast, emotional support is related to the perception of love, caring, sympathy, thoughtfulness, and gratitude in relation to managing emotional demands (e.g., those of depression and anxiety); such emotional support can promote self-esteem and self-efficacy in an individual affected by bereavement and alleviate their bereavement-related depression.
Evidence suggests that higher PSS is associated with higher physical and psychological wellbeing and longevity, both directly and as a stress buffer (Cohen, 2004). However, according to the stress-support matching hypothesis (Cohen & McKay, 1984; Cutrona & Russell, 1990), specific types of stress determine what forms of social support are needed. Some potential mechanisms can be used to understand how PSS can buffer stress-related responses (Berkman & Krishna, 2014; Procidano & Smith, 1997). First, when people become old and need support, it is generally their children who provide support, the death of a child represents the loss of a primary caregiver when they approach old age or become ill. Therefore, perceived instrumental social support from the remaining family members and close friends could encourage the bereaved parents to sustain healthy behaviors and daily activities, especially when the bereaved parents become old and day-to-day functioning prove extremely difficult (Cao et al., 2020; Wang et al., 2021). Instrumental support should be given a top priority for bereaved parents (Cao et al., 2020; Wang et al., 2021). Second, perceived emotional support helps improve the affective and psychological states of a person who has experienced bereavement including their self-esteem, self-efficacy, level of social engagement, and adaptive coping skills to mitigate the deleterious influences of the stressful life event. Previous research suggested that emotional social support is more critical than instrumental social support for bereaved spouses (Bisconti et al., 2006; Powers et al., 2014). Togetherness is considered the best medicine for older bereaved spouses even if it does not solve their problem(s) (Holm et al., 2019). Third, PSS might directly alleviate the harmful physiological functions associated with bereavement through physiological mechanisms, namely the biological regulation of the sympathetic nervous system and hypothalamic–pituitary–adrenal axis activity (Uchino, 2006; Buckley et al., 2012). According to the aforementioned findings, we hypothesized that different types of social support might serve as different channels for meeting the varying needs of bereaved adults (Bottomley et al., 2017; Cacciatore et al., 2021; Holmgren, 2022; Jacobson et al., 2017). Thus, the present study focused on the effects of perceived functional social support on mortality risk associated with different types of bereavement. The results of this study contribute to the literature on the development of theory and interventions related to bereavement policies.
Moreover, research interest in the longitudinal patterns of social support following loss is increasing (Bennett et al., 2020; Bisconti et al., 2006; Stelle & Uchida, 2004). From the resilience perspective, studies have suggested that stable and persistent social support tends to be associated with favorable adaptation among those experiencing bereavement (Hill et al., 2016; Powers et al., 2014). According to the life course approach, social support trajectories present the temporal process of social support to progress and facilitate scrutiny regarding the duration and stages of poor social support, which can diminish an individual’s resilience after bereavement. Because the achievement of resilience in the face of bereavement is a long-term, complex, and dynamic process, instrumental support and emotional support during three time periods were considered in the trajectory analysis in this study to obtain information regarding how the social support trajectory buffers the effects of bereavement.
How individuals adjust to bereavement varies considerably across cultures and societies (Stroebe & Schut, 1998; Wright, 2021), particularly between individualistic and collectivistic cultures. A collectivitistic culture is defined as one where individuals are closely connected with and interdependent on one another, and prioritize group needs and desires over individual ones. By contrast, people in individualistic cultures tend to prioritize their own needs over those of the collective (Hui & Triandis, 1986). Moreover, people in collectivistic cultures tend to seek out implicit and problem-focused social support to derive psychological benefits, whereas those in individualistic cultures often seek out more explicit and emotion-focused support to obtain such benefits (Chen et al., 2015). In general, Western countries are regarded as more individualistic, with their people being more independent, whereas East Asian countries are regarded as more collectivistic, with their people being more interdependent (Markus & Kitayama, 1991). A study revealed that higher social relationship quality was associated with fewer depressive symptoms in the United States (an individualistic culture) but not in Japan (a collectivistic culture) (Kaveladze et al., 2022). Taiwan is generally regarded as a collectivistic culture based on traditional Confucian beliefs and concepts. In Taiwan, older people tend to be reluctant to discuss death and have difficulty expressing their personal needs and seeking support, leaving many of them unprepared for bereavement and with a diminished capacity to adapt and perceive existing social support as inadequate (Hsu et al., 2009; Lai et al., 2021). Traditional Confucian culture values family ties and filial piety. Consequently, in Taiwan, an individual’s spouse or adult child often becomes their primary caregiver when they approach old age or become ill (Yeh et al., 2013). Furthermore, many older adults expect to live with their family members rather than reside in a care institution during old age. In other words, family and community have a crucial influence on the quality of life of bereaved older adults in Taiwan. Thus, cultural differences in reactions to bereavement must be studied in greater detail to inform relevant social policies.
To fill this research gap, the present study (1) investigated the trajectories of perceived instrumental and emotional social support among older adults in Taiwan who had experienced the death of a spouse, a child, or both; (2) employed PSS trajectories as predictors of mortality among bereaved older adults; and (3) evaluated the roles of these PSS trajectories as effect modifiers following various types of loss. The findings of this study can serve as an empirical reference for the design of culturally specific interventions for bereaved older adults in societies culturally similar to Taiwan. Four major hypotheses drove the present study: 1. Because of fluctuations in PSS after loss, multiple trajectory groups exist for perceived instrumental and emotional social support. 2. More favorable trajectories of PSS are associated with lower mortality risk among those grieving the loss of a spouse, a child, or both. If Hypothesis 2 is supported: 3. Trajectories of perceived instrumental social support are strongly associated with mortality risk among those grieving the loss of a child. 4. Trajectories of perceived emotional social support are strongly associated with mortality risk among those grieving the loss of a spouse.
Methods
Data Sources
Data were collected from the Taiwan Longitudinal Study on Aging (TLSA), which is a nationally representative longitudinal panel survey of individuals aged 50 and older that was initiated in 1989, with follow-up interviews being held every 4 years thereafter (Zimmer et al., 2005). Because the questions regarding social support in the TLSA have been revised since 1996, we measured social support from 1996 onward. Consequently, this study contained two cohorts: the participants aged 60 years or older who entered the survey in 1989 and the participants aged 50–66-year-olds who entered the survey in 1996. The present study sample included everyone in both cohorts had completed the 1996 survey, was a widow or widower in 1996 or had experienced the loss of a child before 1996, and was alive in 2003. Because a study suggested that remarriage after widowhood leads to lower mortality because of health advantages and adequate resources from partners (Berntsen & Kravdal, 2012), to simplify the research, we excluded widows and widowers who had remarried after 1996 (n = 17) as well as those living in residential care institutions (n = 189). Consequently, the final analysis comprised 1,188 older bereaved adults living in the community, who were classified into three distinct bereavement groups: spousal (loss of a spouse, n = 603), parental (loss of a child, n = 344), and dual (loss of both a spouse and a child, n = 241) bereavements. This study was approved by the Human Research Ethics Committee at National Chung Cheng University (no. CCUREC109110301).
Measures
Mortality
The primary outcome of interest was 7-year all-cause mortality, which is a time (years)-to-death (or censoring) parameter indicating the survival years between 2003 and 2010. Information regarding deaths occurring during the study period was extracted from the death registry of Taiwan’s Ministry of Health and Welfare (MOHW). In total, 333 total deaths (28.0%), namely 162 (26.9%) in the spousal group, 80 (23.3%) in the parental group, and 91 (37.8%) in the dual group, were recorded.
PSS
Two types of PSS were assessed through self-reporting to examine their effects on mortality risk. First, instrumental PSS was assessed based on participants’ answers to the following item: “When you are sick or in need of care, what do you think of support you perceive from persons close to you?” This item was assessed using a 5-point scale ranging from 1 for poor to 5 for excellent. Emotional PSS was examined using one 5-point item on the willingness of close people to listen to the personal problems and feelings of the participants. Similar measurement instruments have been employed in previous studies (Procidano & Smith, 1997; Lue et al., 2010; Ogle et al., 2014).
Covariates
A set of demographic, socioeconomic, health behavior, and health variables known to be associated with mortality among people who have experienced bereavement were incorporated into the survival analyses. All covariates were measured in 2003. The demographic variables were age (in years) and sex. The socioeconomic variables were education (in years) and self-perceived income adequacy (1 = adequate vs. 0 = inadequate). The health behavior variable was smoking status (smoker or nonsmoker). Finally, health variables were number of chronic health conditions (hypertension, diabetes, heart disease, and stroke) and the presence or absence of physical activity limitations.
Analytic Plan
We examined the descriptive characteristics of all the study participants and subgroups stratified by bereavement type as well as differences in mortality risk by using a series of likelihood-ratio chi-square tests. We employed group-based trajectory modeling (GBTM) to identify a latent cluster of older adults following similar PSS trajectories (Johnson-Lawrence et al., 2013; Nagin, 2005). The GBTM involved clustering the participants into meaningful subgroups that exhibited statistically similar PSS trajectories over time that had been unidentifiable prior to the analysis. In contrast to the ex-ante determination of PSS trajectories on the basis of individual traits, the aforementioned method enabled trajectories to emerge from the collected data. The GBTM indicated the probability of group membership for each older adult, and each adult was assigned to a group for which they had the highest probability (above 0.7). A combination of criteria, such as the Akaike information criterion, Bayesian information criterion (BIC), and sample-size-adjusted BIC, was used to guide the decision on how many latent trajectory groups should be included. As per the suggestion of Nylund et al. (2007), we adopted the BIC as the optimal empirical criterion for selecting the number of trajectory groups. The model with the lowest BIC value (closest to 0) was considered to best fit the data. The final number of trajectory groups was determined when sequential comparisons of BIC values among models yielded no further substantial reductions in these BIC values. Model selection also involved the evaluation of substantive usefulness, theoretical meaningfulness, and policy implications (Nagin & Odgers, 2010). Next, we fitted the instrumental and emotional support trajectory models to the censored normal distribution. In accordance with suggestions from previous studies (Enders & Bandalos, 2001; Raghunathan, 2004), the present study compensated for missing values in estimations of PSS trajectory groups by using the full-information maximum likelihood method. The trajectory groups were then employed as a predictor in Cox proportional hazards regression models. The results are presented in this paper as hazard ratios (HRs) with 95% confidence intervals (CIs). All statistical analyses were conducted using the Stata/SE 16.1 statistical analysis software (StataCorp 2019, College Station, TX, USA).
Results
Descriptive Statistics
Characteristics and PSS Trajectories of the Study Participants by Bereavement Type.
aS.D.: Standard Deviation.
bPhysical activity limitations were assessed by asking the participants if they had encountered difficulty performing seven physical activities, namely standing continuously for 15 min, squatting, raising both hands above their head, picking up or manipulating items by using the fingers, lifting or carrying an item weighing 12 kg, running a short distance (20–30 m), and climbing stairs to ascend 2–3 floors. If a participant reported difficulty performing any one of these seven activities, they were regarded as having a physical activity limitation.
Note. †p < .10; *p < .05; **p < .01; ***p < .001.
The fit statistics of GBTM for the instrumental and emotional PSS data are presented in SupplementaryTable S1. According to the BIC, the four-group trajectory model exhibited the best fit for instrumental PSS, whereas the three-group trajectory model exhibited the best fit for emotional PSS. Figure 1 illustrates the instrumental and emotional PSS trajectory patterns from 1996 to 2003. The identified groups for instrumental PSS trajectories were as follows (Figure 1A): consistently high instrumental PSS (n = 726, 61.1%), late decreasing instrumental PSS (n = 212, 17.9%), early decreasing instrumental PSS (n = 164, 13.8%), and increasing instrumental PSS (n = 86, 7.2%). The emotional PSS trajectory groups (Figure 1B) were as follows: consistently high emotional PSS (n = 861, 72.5%), decreasing emotional PSS (n = 277, 23.3%), and consistently low emotional PSS (n = 50, 4.2%). PSS trajectory groups of the older Taiwanese adults experiencing bereavement (1996–2003). (A) Perceived instrumental support trajectory. (B) Perceived emotional support trajectory.
Associations Between PSS Trajectory and Mortality
Figure 2 illustrates the HRs and 95% CIs estimated for mortality risk among all the participants according to the various PSS trajectories after adjustment for all covariates. Of all the bereaved participants, those with consistently high and increasing instrumental PSS trajectories had a lower mortality risk than did those with early decreasing instrumental PSS trajectory (HR = 0.41, 95% CI = 0.30–0.55 and HR = 0.61, 95% CI = 0.39–0.97, respectively). Moreover, membership in the late decreasing instrumental PSS trajectory group was significantly associated with an increased risk of mortality (HR = 1.34, 95% CI = 0.99–1.82). Similarly, regarding the emotional PSS trajectory, those with consistently high emotional PSS exhibited lower mortality risk (HR = 0.30, 95% CI = 0.24–0.38) than did those with decreasing and low emotional PSS. HRs and 95% CIs for mortality risk in relation to the PSS trajectories of all the bereaved participants.
Regarding the covariates (SupplementaryTable S2), the results remained consistent among all the models and indicated that the participants who were men, were smokers, had self-reported inadequate income, and had physical limitations exhibited a higher mortality risk than their counterparts. In addition, mortality risk increased with age and number of chronic health conditions.
Figure 3 illustrates the adjusted HRs for mortality for the distinct PSS trajectories according to bereavement type. After adjustment for all covariates, compared with those with early decreasing instrumental PSS trajectory, those in the consistently high instrumental PSS trajectory group had a 54% lower mortality risk for spousal bereavement (HR = 0.46, 95% CI = 0.30–0.70), a 71% lower risk for parental bereavement (HR = 0.29, 95% CI = 0.16–0.54), and a 59% lower risk for dual bereavement (HR = 0.41, 95% CI = 0.23–0.73). In addition, the participants who experienced spousal bereavement and exhibited late decreasing instrumental PSS trajectory had a higher risk of mortality than did those who experienced spousal bereavement and exhibited early decreasing instrumental PSS trajectory (HR = 1.58, 95% CI = 1.03–2.43). HRs and 95% CIs for mortality risk by bereavement type.
Moreover, after adjustment for other covariates, the participants with consistently high emotional PSS trajectory had a decreased risk of mortality than did those with decreasing emotional PSS trajectory. The degree of the reduction in mortality risk for the participants with a high emotional PSS trajectory was higher in the spousal bereavement group (72%, HR = 0.28, 95% CI = 0.20–0.39) and parental bereavement group (74%, HR = 0.26, 95% CI = 0.16–0.43) than in the dual bereavement group (58%, HR = 0.42, 95% CI = 0.26–0.68).
Discussion
This large population-based study yielded four major findings. First, a comparison of the sizes of the instrumental and emotional PSS trajectory groups revealed that the trajectory patterns of these groups differed. Our Hypothesis 1 is supported. Most of the participants had perceived stable and favorable instrumental (61%) and emotional (73%) PSS; only 4% of the participants reported perceived low emotional PSS. However, 30% and 23% of the participants exhibited decreasing instrumental and emotional PSS trajectory, respectively. Instrumental PSS fluctuated more than emotional PSS did. This result is consistent with the findings of previous studies (Hill et al., 2016; Powers et al., 2014; Stelle & Uchida, 2004). Thus, the present empirical results suggested a distinction between the instrumental and emotional PSS trajectories of older bereaved adults and indicated that neither trajectory type should be overlooked when investigating the role of PSS in the bereavement process.
Second, we observed that bereaved older adults with stable high PSS had a lower risk of mortaltiy, whereas those with low or decreasing PSS had a higher risk of mortaltiy. Hypothesis 2 which hypothesizes that more favorable trajectories of PSS are associated with lower mortality risk among those experiencing the loss of a spouse, a child, or both, is supported. Three mechanisms explained the increased risk of mortality in older bereaved adults with low and decreasing PSS. First, the risk accumulation mechanism posits that exposure to low PSS accumulates over a long period and might compromise a person’s health in later life (Kuh et al., 2003; Ogle et al., 2014). Next, under the critical period effect, contemporaneous PSS might be considerably more influential later in life. Conversely, bolstering a person’s social support when they are old can counteract the previous occurrence of inadeuqate social support at an earlier age. Moreover, in line with socioemotional selective theory (Carstensen, 1991), older adults gradually and deliberately minize their peripheral social networks and numbers of social contacts (friends, work relationships) with increasing age, because of their poor health status. These findings suggest that identifying those experiencing bereavement with low and declining PSS as a group at high risk of developing prolonged grief is a crucial cross-system and cross-cultural task.
Third, we observed different effects of instrumental and emotional PSS trajectories on mortality risk. Consistent with prior research (Lyyra & Heikkinen, 2006), the beneficial effect of an emotional PSS trajectory on the survival of older participants was stronger than that of an instrumental PSS trajectory. This outcome might have been caused by the mechanism through which instrumental PSS and emotional PSS yield different coping resources. Specifically, instrumental PSS relies on the adequacy of tangible support, which is primarily associated with biological functions and physical health, as a coping resource (Uchino, 2006). By contrast, emotional PSS constitutes companionship, which not only mitigates loneliness and isolation but also promotes psychosocial resources, including self-esteem, self-efficacy, and a sense of meaning in life, among people experiencing bereavement (Cacciatore et al., 2021; Holt-Lunstad, et al., 2015; Thoits, 2011; Vähäkangas et al., 2021; Wright, 2021). Older bereaved people might benefit more from psychosocial resources (i.e., emotional PSS) than from other resources, because according to Erikson’s theory of psychosocial development, older adults are triggered to develop ego integrity to feel a sense of fulfillment and life success, as their functional abilities decline (Perry et al., 2015). However, the literature suggests that emotional PSS and instrumental PSS can interact to predict health and well-being and that instrumental PSS combined with emotional meaning (e.g., caring, esteem, empathy) is more effective than assistance-only support in stressful situations (Morelli et al., 2015; Semmer et al., 2008). These findings provide insight regarding the practical implications of delivering support following bereavement; specially, the emotional needs of the bereaved individual should be considered in addition to any formal or professional support.
Fourth, we discovered that the benefits of consistently high support varied depending on bereavement type. This finding implies that appropriate adjustments for people experiencing spousal bereavement might differ from those for people experiencing parental bereavement (Arbuckle & de Vries, 1995). Compared to those experiencing spousal or dual bereavement, those experiencing parental bereavement benefited more from consistently high instrumental PSS. For instrumental social support, our Hypothesis 3 is supported. This reality is particularly true in Taiwanese society. In traditional Confucian cultures children are the main source of support for parents in their old age, the death of a child represents the loss of a primary caregiver when bereaved parents become old and day-to-day functioning was difficult (Wang & Hu, 2021). Moreover, in the Taiwanese society, where the traditional notion of filial piety leads older adults to expect to live with one of their adult children so that they might receive appropriate tangible support rather than reside in a care institution. A nationwide survey conducted in 2017 revealed that more than 54% of older people in Taiwan expected to live with their children (MOHW, 2018). In a previous study, the consistent perception of adequate instrumental support indicated that the bereaved older parents felt they had received adequate and high-quality of care when required; such support can prevent the decline of biological functioning and thus contribute to a reduced risk of mortality (Uchino, 2006). Instrumental support from friends and family members could help bereaved parents cope with the difficulty in everyday activities (Cao et al., 2020; Wang et al., 2021). Our findings have practical implications that instrumental social support is the top priority for those experiencing the loss of a child (Rostila et al., 2012; Rubin, 1993).
Next, the present study revealed that consistently high emotional PSS trajectory had a considerably stronger protective effect on mortality risk for a widow or widower than did consistently high instrumental PSS trajectory. In collectivistic societies, such as Taiwanese society (Insights, 2022), people tend to consider their spouse as an integral part of their social network and therefore develop emotional dependence on their spouse; thus, an older individual with a deceased spouse may be faced with the challenge of restoring meaning in their life, which might require extensive psychosocial resources (Anusic & Lucas, 2014; Vähäkangas et al., 2021). Therefore, persistent emotional social support for widows and widowers can promote a robust sense of social connection and safety and a perceived attachment to society among them, which reduces their mortality risk. The findings imply that the most crucial aspects of social support in meeting the needs of bereaved individuals should be understood and not overlooked.
However, this study found that consistently high emotional PSS trajectory also contributed to lower mortality risk for the parental bereavement group. Particularly, the magnitudes of reduced risk associated with high emotional PSS trajectory were equivalent among the spousal and parental bereavement groups. Thus, in terms of emotional social support, hypothesis 4 is partially supported. A possible explanation might be that parental bereavement seems to lose not only caregivers for their late life but also a social role as a parent. Previous research provide evidence that negative emotions such as pervasive feelings of guilt, great difficulty in coming to term with the death, and lack of meaning of life from experiencing a loss of children can proliferate and penetrate through their life course (Arbuckle & de Vries, 1995; Rostila et al., 2012). Furthermore, according to Confucianism, it matters for elderly parents to enjoy the harmonious intergenerational relationship of their adult children (Wang & Hu, 2021). Thus, consistently high emotional PSS trajectory is more associated with empathy, affection, and emotional commitment and favorable views of life that helps older bereaved parents to diminish feelings of distress or even enhance moments of happiness (Procidano & Smith, 1997). In addition, the provision of instrument support might simultaneously bring emotional meaning such as caring, understanding, and esteem that can be helpful to the stress-coping process (Semmer et al., 2008). Our finding on both consistently high emotional and instrumental PSS trajectory contribute to reduced risks of mortality among elderly bereaved parents in Taiwan will encourage future social policy to invest both emotional and instrumental support resources for individuals experiencing a loss of children.
Finally, this study provides evidence of the beneficial effects of high instrumental PSS and emotional PSS for the dual bereavement group; however, the effect sizes of PSS for this group were lower than those for the spousal and parental bereavement groups. A study indicated that repeated instances of bereavement had cumulative harmful effects on the life satisfaction of a sample of older people (Infurna & Mayer, 2019). Thus, people who have experienced dual bereavement might require extensive social support; consistently high PSS can mitigate a part of the negative effects of dual bereavement on survival. This finding underscores the importance of identifying individuals suffering from dual bereavement because such individuals might be at high risk of greater and prolong grief. In summary, the findings of this study highlight the importance the analyzing multiple dimensions of social support in relation to the coping processes of people, especially older adults, who experienced distinct types of bereavement.
Limitations
The present study had certain limitations. First, the analyzed older adults differed in terms of the years, timings, durations, and circumstances of their bereavement experiences. Moreover, differences were found in the quality of relationship between pairs of deceased and surviving individuals. However, such differences were not considered in the analysis because related information was not available from the TLSA. Therefore, further studies are recommended to collect data related to the factors and include them in their analyses (Stroebe et al., 2007). Second, although we categorized the participants according to three types of bereavement, we did not investigate the additive or multiplicative effects of each distinct type of bereavement because of a lack of data regarding the frequency of the participants’ loss experiences. Third, because the data used in this study were limited to those related to the participants who were alive in 2003, our analysis might have yielded conservative estimates regarding the association between PSS trajectory and mortality because of health selection. Finally, the use of self-reported instrumental and emotional PSS measurement might have led to recall bias as well as bias due to the participants’ feelings at the time of the survey.
Conclusion
This study makes a unique contribution to the extant research. To the best of our knowledge, this study is the first large-scale, population-based study of older Taiwanese people to demonstrate the roles of longitudinal patterns of PSS in mortality risk in relation to different types of bereavement, namely spousal, parental, and dual bereavement. By using GBTM to conduct a trajectory analysis of PSS, we elucidated how instrumental PSS and emotional PSS affected older adults in Taiwan who had experienced bereavement. We also observed a strong association between a reduced risk of mortality and persistently high levels of PSS over time among the analyzed bereaved older adults. Furthermore, the effect sizes of the instrumental PSS and emotional PSS trajectories differed markedly depending on bereavement type. These findings shed light on the roles of varied, persistent, and adequate PSS following the loss of a spouse, a child, or both as well as the associated risk of mortality.
Supplemental Material
Supplemental Material - Research on Aging Trajectories of Instrumental and Emotional Social Support and the Associated Risk of Mortality in Bereaved Older Adults in Taiwan
Supplemental Material for Research on Aging Trajectories of Instrumental and Emotional Social Support and the Associated Risk of Mortality in Bereaved Older Adults in Taiwan by Huei-Jia Tzeng, Chiachi Bonnie Lee, Cheng-Tsung Chen and Miaw-Chwen Lee in Journal of Research on Aging
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the Ministry of Science and Technology, Taiwan (Grant Number MOST 108-2410-H-194-084-MY2), the Center for Innovative Research on Aging Society from the Featured Areas Research Center Program within the framework of the Higher Education Sprout Project of Taiwan’s Ministry of Education, and China Medical University (Grant Number CMU109-MF-125).
Supplemental Material
Supplemental material for this article is available online.
Author Biographies
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
