Keywords

Contribution to the LIVES Book

Intimate partner loss represents an important critical and stressful life event in later life, offering the opportunity to study vulnerability and resilience processes after what is considered one of the most important stressors that adults may face. Occurring either through separation/divorce or death of the partner, partner loss can have notable consequences in several domains of life, illustrating the interdependence of these domains and spill-over effects. Apart from the change in marital status and associated social implications, losing a long-term romantic partner can affect a person’s physical and mental well-being, social and personal identity along with more practical aspects of everyday life, such as financial resources, as well as social relations and support. The effects of partner loss vary substantially among people, which is in part due to event-related factors but also because of interindividual differences in personal resources and psychological strengths leading to variations in adaptation success. In the ‘LIVES Relationships in Later Life Study’ (LRLLS; Hutchison et al., 2013), we examined in more detail what contributes to vulnerability and resilience processes in the context of partner loss. Specifically, we investigated the extent to which individual (e.g., health) and event-related (e.g., type of partner loss: divorce vs. bereavement; time since the loss) factors contribute to vulnerability in the context of partner loss and aimed to identify personal resources (e.g., social support, psychological characteristics such as perceptions of self- and social continuity) that contribute to resilience trajectories. Furthermore, based on the project findings, an online intervention was developed to help individuals overcome partner loss. This chapter will summarise some key findings of our project.

Divorce and Bereavement in Later Life

Divorce in later life is a recent phenomenon associated with important yet understudied economic, social and psychological consequences (Brown & Lin, 2012; Dykstra & de Jong Gierveld, 2004; Perrig-Chiello et al., 2015). Between 1999 and 2019, the divorce rate of older individuals increased substantially in Switzerland: For those aged 50 to 59 years, it increased by approximately 40%; for those aged 60 years and older, it doubled (SFOS, 2021). Switzerland follows a similar trend as that observed in many countries (e.g., Japan). While the divorce rate of the general population seems to have reached an upper limit and has not increased further in more liberal northern countries (OECD, 2020; UN-DESA-PD, 2015), a continuous rise can be expected for countries with more conservative values, such as Switzerland, especially for those in old age. Consequently, divorce in later life has become a critical life event that can no longer be considered an off-time transition. Spousal loss through bereavement has also increased in old age, but it is often experienced today later in life. Because of the increase in life expectancy, a larger number of individuals are likely to lose their partner in very old age. Between 1999 and 2019, the number of widow(er)s increased by approximately 30% for those aged 80 years and older, while for the group of 65- to 79-year-old individuals, the number of widow(er)s decreased by approximately 20% (SFOS, 2021). Even though divorce and bereavement may be similar critical life events, as in both cases an individual loses their spouse, both events can have different consequences, for example, for psychological health, financial adequacy, and the family and social relations of the remaining partner.

A host of studies has shown that, for most individuals, partner loss is challenging and stressful at all ages and can have substantial negative and lasting consequences. Lucas (2005), for example, found in a large prospective longitudinal study that individuals who faced partner loss through divorce or bereavement showed substantial reductions in life satisfaction and did not recover to their pre-loss level during the following seven years. A more recent study found that widow(er)s recovered 3 to 4 years after their loss (Kettlewell et al., 2020). Neither study, however, considered the role of age, thereby neglecting the fact that speed and success of recovery may depend on the life stage at which the loss is experienced. In addition, these studies focused on levels and changes of well-being at a group level but did not consider the existence of interindividual differences in adaptation and recovery. Analysing individual trajectories with longitudinal data is crucial, as previous research has revealed different post-loss developments (e.g., Bonanno et al., 2002), with some individuals overcoming divorce and bereavement rapidly and efficiently but others feeling lost and alone. In addition to enabling a closer, more fine-grained picture of individual differences in the adaptation process, the analysis of such trajectories is critical to identifying factors that contribute to experiencing partner loss as an unbearable tragedy or as an opportunity for personal growth.

Furthermore, the specific moment in life when individuals lose their partner is important, given that personal and societal resources and challenges differ over the life course. Losing one’s spouse in later life can come with particular and/or additional risks for social, physical and mental health (Dykstra & de Jong Gierveld, 2004), stressing the importance of considering the timing of specific life events within the life course and their consequences for different life domains when investigating vulnerability (Spini et al., 2017). To begin with, older adults seem more vulnerable than younger individuals, in part, because they often face loss accumulation across many central domains of functioning when advancing towards old age. For instance, they are likely to experience age-related changes in professional life (e.g., retirement) and health (e.g., chronic illnesses, multimorbidity) that challenge identity. Social losses due to the death of network partners also represent a risk associated with advancing age (d’Epinay et al., 2010). At the same time, older individuals are often confronted with increased family needs associated with care provision: Prolonged caregiving for their even older parents can further challenge the resources of older individuals (Boerner et al., 2021; Oudijk et al., 2011). Apart from the multidimensionality of the risks associated with a precarious position after partner loss, it is of note that vulnerability in this context is also related to the limited availability of resources and support at multiple societal levels, such as close friends, organised social groups, or governmental support to family caregivers.

Partner loss often leads to a reduction in social contacts, indicating that apart from the actual loss, individuals have to cope with stressors originating from their social network instead of receiving support in times of need. Divorcees are often confronted with the loss of social ties (Widmer et al., 2012) and groups taking sides, while widow(er)s tend to associate social activities that were related to the late husband or wife with painful memories (Smith et al., 2020), which makes it difficult for them to engage in such activities. Widow(er)s also report feeling let down by their social context, as old friends diminish interactions with the remaining spouse (Ha, 2008). Replacement of resources such as social support and well-liked activities may become more difficult in older age, and, consequently, individuals may feel isolated and socially and emotionally lonely (De Jong-Gierveld et al., 2006). Thus, older individuals’ higher risk of vulnerability calls for a more systematic examination of the process of adaptation to partner loss, given that this event has consequences in multiple life domains, and the investigation of factors contributing to resilience and better adaptation over time. In the LRLLS (Hutchison et al., 2013), across three assessment waves, we therefore investigated a large sample of individuals in the second half of their lives who were confronted with partner loss over the course of four years. A married group served as the control. The goals of the study were to examine the psychological effects of partner loss on vulnerability and resilience trajectories (i.e., health, well-being) and their mediators.

Vulnerability and Resilience After Partner Loss

Adaptation trajectories to major transitions or traumatic events have been described by Bonanno and colleagues (e.g., 2002). Having identified different trajectories within the context of adaptation to bereavement first (Bonanno et al., 2002), Bonanno et al. (2011) extended their approach to different traumatic and stressful experiences, notably on war veterans, survivors of terror attacks or other disasters. Based on longitudinal data analyses, their work showed four prototypical trajectories. The first and most frequent trajectory (ranging between 35% and 65% depending on the studied event) was that of resilience, in which individuals continued to function well and showed no major physical or psychological damage after a potentially traumatic experience. Second, the recovery trajectory (ranging between 15% and 25%) was characterised by moderate to severe levels of initial physical or psychological symptoms and some difficulties meeting role obligations, from which individuals eventually recovered to their baseline levels of functioning. The third trajectory (ranging between 5% and 30%) was qualified as chronic distress, during which people continued to suffer from symptoms long after the event. Finally, the fourth trajectory was characterised by a ‘delayed’ reaction (ranging between 0% and 15%), as individuals began to show distress and malfunctioning related to the traumatic experience after some period of resilience. This typology of trajectories has been observed after various types of events, such as heart attack (Galatzer-Levy & Bonanno, 2014), spinal cord injuries (Aparicio et al., 2020; Van Leeuwen et al., 2012), and outbreaks of chronic health conditions (Debnar et al., 2020).

Inspired by this approach, we first analysed the post-bereavement profiles of Swiss older widows and widowers in the prospective LRLLS with cross-sectional and longitudinal data. Further extending prior work, we aimed to better understand the role of personal resources and contextual factors as discriminant variables among patterns of psychological (depressive symptoms, hopelessness, loneliness, life satisfaction) and subjective health. Cross-sectional analysis of first-wave data and profile analysis of the psychological and subjective health indicators revealed three different groups. The most frequent profile was the Resilients (54% of the sample). A total of 39% represented the Copers, with a similar profile but with an adjustment at a lower level of functioning. The third group, the Vulnerables (7%), showed more depressive symptoms, loneliness and hopelessness (Spahni et al., 2015). When considering how the three profile groups differed in terms of specific characteristics, we found that the most important variables for group allocation were internal resources (i.e., psychological resilience, personality) but also event and relationship variables (i.e., former relationship quality, how the loss was experienced). More specifically, the Resilients were more extraverted, more conscientious, more open, more agreeable, and less neurotic than the other two groups. The Copers were also more extraverted and less neurotic than the Vulnerables. In addition, the three groups differed in terms of psychological resilience, with the Resilients having the highest score. The latter group also had received more spousal support from their late partner than the other two groups.

Analyses of the longitudinal data (all three study waves) confirmed the three-group typology of profiles found cross-sectionally (Spahni et al., 2015): approximately 48% belonged to the Resilients group, 40% were Copers, and 12% were Vulnerables (Bennett et al., 2020). While there was a general trend of adaptation, moving from maladaptiveness to resilience, the findings also showed that a limited number of individuals shifted from higher to lower adaptation. This backwards movement was associated with closer proximity to the event and the partner’s need for care in the period preceding death, resulting in temporary relief for the survivor with a later backlash. Our results underlined the important role of internal resources (i.e., psychological resilience, personality) in facing spousal death in old age and, notably, the difficulty of maintaining self-functioning.

Extending this approach to divorce, we explored the patterns of adaptation to marital breakup after a long-term marriage—a neglected group in divorce research—and the longitudinal trajectory of these patterns using the LRLLS dataset. Latent transition analyses were used to examine the course of psychological adaptation (i.e., depressive symptoms, life satisfaction, hopelessness, mourning and subjective health) to divorce over two years. Two larger groups appeared: individuals adapting very well (Resilients: 29%) and individuals adapting quite well (Average Copers: 49%). We also observed three smaller groups characterised by major difficulties, including two groups with the most negative outcomes, the Malcontents (12%) and the Vulnerables (6%), with the latter having the largest deviation from the mean, and a third group, the Resigned (4%), which identified individuals with high depression scores (Perrig-Chiello et al., 2015). The results again showed that adaptation trajectories depended primarily on internal resources—including personality traits, notably extraversion, which was associated with a resilient trajectory—but also on contextual variables related to breakup and relationship characteristics. Specifically, a shorter time since the breakup was observed among respondents who moved from maladaptive to adaptive statuses. The length of the marriage played a somewhat counterintuitive role: Whereas for most respondents the number of years in the terminated relationship was not associated with the probability of recovery, there was a link for the resigned group, with individuals exiting marriages longer than 35 years being more likely to recover than individuals exiting shorter marriages (Knöpfli et al., 2016).

In sum, analysis of adaptation profiles and trajectories suggested that most people coped well with divorce and widowhood and were resilient when they lost their life-long romantic partner in the second half of life. However, a significant minority had difficulties and showed vulnerability. The findings further indicated that individuals possessed internal resources that help during these stressful transitions and highlighted the importance of contextual variables associated with a higher risk for vulnerability. Furthermore, our findings extended earlier work (e.g., Bonanno et al., 2002, 2011) in an important way by demonstrating that later-life partner loss is different from other stressful events through the identification of specific profiles and trajectories, by showing differences among types of partner loss, and by providing a more detailed picture of the characteristics associated with these trajectories.

Specific Internal Resources for Adaptation: Self- and Social Continuity

Extending the findings on internal resources such as personality as predictors of adaptation success, identity and self-regulatory processes were investigated in LRLLS, with a special focus on the beneficial value of perceptions of self- and social continuity. As individuals age, their sense of identity and well-being is likely to be influenced by the experience of change and/or continuity in several life domains, such as health or social ties (Atchley, 1989). However, the impact of any objective transitions, including divorce or bereavement, may depend on the extent to which an individual perceives these changes as disruptive. Critical life events in later life, such as divorce and bereavement, may destabilise the individual’s sense of personal and social identity during a time when they are most needed, as other age-related losses may also occur at the same time. Individuals who are nevertheless able to maintain a strong sense of self- and social continuity may be more successful in adapting to adversity caused by the loss of the partner.

According to Breakwell (1986), continuity is an identity end-state: Individuals strive to maintain a sense of continuity throughout their lives (Spini & Jopp, 2014). Self-continuity is defined by the extent to which individuals are able to identify similar core characteristics between their present, past and future selves. Based on this identity mechanism, people construct their own life story: At younger ages, this mechanism helps visualise how the person may be in the future (future self-continuity), while in later life, self-continuity helps evaluate one’s personal history (past self-continuity). Social continuity, in contrast, refers to persisting environmental structures, long-lasting social ties and groups, and roles in society. The two mechanisms act in parallel in constructing a unique sense of personal and social identity.

Little is known about the development of self- and social continuity across the life course, and the beneficial role of both identity mechanisms has rarely been investigated within the context of critical life events. Using LRLLS data, we aimed to close these gaps in the literature by investigating stressors and resources that shape continuity perception in later life, in line with LIVES’ dynamic view of vulnerability (Spini et al., 2017; Spini & Widmer, Chapter 2 this book), and by determining the extent to which self- and social continuity buffer the impact of partner loss on relevant outcomes, including feelings of loneliness.

Self-continuity develops early in life (Erikson, 1968). Individuals who have difficulty developing a strong sense of identity during childhood and adolescence may be more vulnerable when confronted with self-threatening life events. There is also empirical evidence that adversity in childhood and adolescence interferes with identity development, leading to psychopathology in later life (Hayward et al., 2020). Following these theoretical assumptions and findings, we used the three study waves of the LRLLS to examine the link between childhood adversity and self-continuity in later life, a relationship that had not, to our knowledge, been previously studied. We investigated this association in three different groups, namely, divorced, bereaved and married individuals, with the last serving as a control group. Our findings confirmed that individuals who had experienced childhood adversity (e.g., poverty, emotional or physical neglect) lacked self-continuity in comparison to their peers (Lampraki, Spini, & Jopp, submitted 2022a). In addition, childhood adversity was related to lower levels of self-continuity for the divorced and bereaved individuals compared to the married control group. Considering further the temporal development after the partner loss event, we found that divorcees were the most disadvantaged group in terms of self-continuity as, even six years after the divorce, they still experienced less self-continuity than the bereaved and married individuals who had, meanwhile, experienced an age-normative increase in self-continuity.

Another gap in the literature concerned the protective role of continuity perceptions in adaptation to partner loss. Results indicated that low self- and social continuity was related to increased levels of social loneliness after divorce and, especially, two years or more after marriage dissolution (Lampraki et al., 2019). In addition, positive personality traits (e.g., extraversion) seemed to protect individuals with closer proximity to divorce from feeling socially lonely.

Following up on our finding that individuals with childhood adversity (e.g., poverty, emotional or physical neglect) experienced lower self-continuity later in life, we also investigated the long-term impact of childhood events on social and emotional loneliness after divorce and bereavement and in long-term marriages while considering their perception of self- and social continuity as protective mechanisms (Lampraki, Jopp, & Spini, submitted 2022b). Although childhood adversity exerted a negative impact on well-being with the increase of social and emotional loneliness, this relationship weakened when also considering the effect of self-continuity. In other words, when individuals reported fewer negative childhood experiences, they had a stronger perception of self-continuity in later life and, therefore, had better psychological well-being outcomes after partner loss, as well as in long-term marriages. In addition, maintaining a strong feeling of social continuity in life was also related to less social and emotional loneliness.

In sum, our findings contribute to a better understanding of the development of continuity perceptions and their protective value for adaptation to partner loss in later life, showing the multilevel influence of resources and stressors (intraindividual and contextual) on well-being, the multidirectionality of individual trajectories, and the multidimensionality of consequences following a single critical life event.

An Evidence-based Online Intervention After Partner Loss: LIVIA

While the majority of individuals managed to cope successfully to losing their spouse, our findings also documented important interindividual differences in coping means, length of the adaptation process and the extent to which individuals were able to return to previous functioning and well-being. Seeking out help from mental health specialists can help coming to terms with the partner loss, yet systematic variations in professional help seeking across types of partner loss exist: Within the LRLLS study, we found that professional help seeking was much higher after separation (57%) and divorce (49%) compared to widowhood (18%; Jopp et al., 2021). Further, results indicated that higher likelihood to seek professional help was related to specific person characteristics (i.e., being a woman), social resources (e.g., having a person to count on), poorer adaptation success (i.e., more depressive symptoms; needing more time to overcome the loss) as well as event-specific characteristics (i.e., unexpected loss; needing more time to overcome the loss; Jopp et al., 2021). Informing individuals unlikely to seek help, namely those suffering from bereavement, being male, and without a confidant, about the potential benefits of professional support after partner loss in later life seems particularly important. Given potential barriers to professional support seeking, self-help online interventions have become increasing attention over the recent years.

Based on theidentified patterns of adaptation and factors of influence, the LRLLS study gave rise to the conceptualisation and completion of an online intervention aimed at actively helping individuals adapt to loss, maintain a stable identity and improve well-being. More specifically, we developed a guided online self-help intervention for prolonged grief after partner loss (LIVIA) to help individuals create new adaptative resources. Online self-help interventions (e-health) are a convenient, low-threshold, and immediate way of delivering interventions. By addressing prolonged grief after the death of a significant other, online self-help interventions complement traditional face-to-face grief counselling and have been proven to be effective (e.g., Eisma et al., 2015; Litz et al., 2014; van der Houwen et al., 2010).

LIVIA is the first online intervention that focuses explicitly on grief after spousal death in older adults and the first that addresses grief and adaptation problems after separation or divorce (Brodbeck et al., 2017). It is based on the task model by Worden (2009) and the Dual Process Model by Stroebe and Schut (1999) and encompasses loss-related interventions aimed at coping better with grief as well as restoration-oriented generic interventions for activating personal resources and improving well-being and life satisfaction. The cognitive-behavioural online self-help program consisted of ten text-based modules, which were similar for widowed and separated/divorced participants. Exceptions were two loss-related modules that had an identical structure but were adapted to the specific event: One provided information about bereavement or separation reactions, and the other focused on telling the story of the loss. The self-help intervention covered topics such as emotional reactions in the context of interpersonal loss, obstacles to successful adaptation and possibilities for overcoming them. Cognitive-behavioural intervention elements aimed at fostering positive thoughts and emotions, finding comfort and self-care. Worksheets for practising these strategies in daily life complemented the module. Writing tasks assessed changes in the participant’s life since the loss, identified unfinished business and possibilities of putting these issues at rest, and redefined the relationship to the lost partner. Restoration-oriented modules focused on creating a new life without the partner, such as finding sources of support and strength, improving existing relationships and building up new social contacts.

LIVIA proved its efficacy in a randomised controlled trial (Brodbeck et al., 2019). Inclusion criteria were the loss of a partner through death or separation/divorce more than six months ago and wanting help for better coping with grief and emotional adaptation after the loss. A total of 110 German-speaking older adults took part in LIVIA; the majority of them were divorced (77%). Although all LRLLS participants reporting difficulties coping with the loss of their partner were informed about the intervention, most of the LIVIA participants were recruited via newspaper advertisements. When comparing the intervention group with the control group, LIVIA led to significant reductions in grief (d = 0.81, p < .001), depression symptoms (d = 0.59, p < .001), psychopathological distress (d = 0.39, p = .011), embitterment (d = 0.37, p = .006), loneliness (d = 0.37, p = .025), and an increase in life satisfaction (d = 0.41, p = .007). Improvements in these areas were confirmed for both separated/divorced and widowed participants and for different grief severities at baseline. Notably, these improvements were stable at the six-month follow-up (for details, see Brodbeck et al., 2019). These findings underlined that the intervention principles for treating prolonged grief after bereavement also work for separated and divorced older adults and for participants with milder grief symptoms. This finding suggested that LIVIA may also function as a preventive measure to counteract negative developments after the loss. Further analyses addressing underlying mechanisms showed that LIVIA improved emotion regulation and loss-related coping self-efficacy, which in turn led to a reduction in grief symptoms and psychopathological distress (Brodbeck et al., 2019).

LIVIA provided the ground for further interventions. An adaptation of LIVIA is currently being evaluated in the French-speaking part of Switzerland (Debrot et al., 2017). Building on LIVIA, the ongoing LEAVES project integrated the content of the self-help intervention into an online bereavement support service for older adults in the Netherlands, Portugal and Switzerland (Van Velsen et al., 2020). The LEAVES program with further methodological refinements, including dialogues with a virtual agent, is currently being evaluated through a new randomised controlled trial. In sum, the newly designed LIVIA intervention for individuals wishing to receive help to overcome the loss of their partner had strong beneficial effects, thereby demonstrating that this online intervention is highly useful to support individuals with pathological or potential problematic trajectories.

Conclusions

Taken together, the findings from the LRLLS study offer rich and valuable insights into vulnerability processes in the context of partner loss later in life. They showed that partner loss as a critical life event has substantial consequences for multiple central life domains, that various resources and risk factors can be identified and that interventions addressing different angles are highly helpful. More specifically, our work highlighted interindividual differences in vulnerability: The profiles and trajectory results clearly demonstrated that individuals differ in adaptation progress and success and that these differences are associated with personal resources and contextual as well as event-related characteristics that promote (e.g., experiencing self-continuity) or hinder (e.g., feeling unable to overcome the loss) adaptation to this critical life event. The interdependence among vulnerability trajectories, life domains and multilevel contexts, which lies at the core of the LIVES approach to vulnerability (Spini et al., 2017), is exemplified by our research. The findings further highlighted the long arm of childhood by underlining the role of early life risk factors, such as childhood adversity, and their lasting negative influence on later life identity processes; moreover, divorce and widowhood, as additional adverse events, weigh even more on individuals who have experienced childhood adversity, with notable consequences for the adjustment to partner loss. As underlined by Spini and Widmer (Chapter 2, this book), vulnerability trajectories are shaped before, during and after a critical event such as the loss of an intimate partner from childhood to adulthood.

Although childhood is important, our results showed that programs augmenting resilience can also be implemented in later adulthood. While many individuals are well equipped to handle partner loss successfully, it is obvious that some adults do not seem to possess sufficient internal resources (e.g., protective personality traits such as extraversion, psychological strengths such as trait resilience) and external resources (e.g., having a confidante, social support; see Chap. 10) to master stressful transitions such as divorce or bereavement well. These vulnerable adults benefit from interventions that facilitate adjustment after spousal loss at an early phase of the adaptation process as well as after having developed manifest difficulty. Based on our findings, future interventions may consider more specific modules that focus on improving specific resources, such as individuals’ continuity perceptions, to strengthen these identity mechanisms, thereby going above and beyond classic coping skills.

Following a life course perspective, it is of interest to target vulnerability as early as possible. Such distal factors of vulnerability were also shown by Cullati et al. (Cullati & al., Childhood Socioeconomic Disadvantage and Health in the Second Half of Life: The Role of Gender and Welfare States in the Life Course of Europeans, 2022) and may call for early intervention to create reserves during childhood that will favour resilience during adulthood. Based on our findings, such an approach entails concentrating on preventing childhood adversity, including poverty and maltreatment, as an important governmental target and creating societal structures that effectively reduce all types of inequality. In parallel, efforts are needed to enable individuals, ideally from childhood on, not only to develop cognitive, social and functional resources but also to enhance psychological strengths (e.g., self-efficacy, internal control beliefs) and coping strategies (e.g., problem-solving skills). Building a strong identity, which will act as a resource (or reserve) in times of need in later life, should become the aim of preventions and interventions during adolescence, the life stage at which identity emerges. The development of a strong repertoire of coping skills should further target not only active, self-dependent problem solving but also the capacity to ask others for practical and emotional support, as well as strategies such as acceptance and perspective taking, as these are especially important when the individual is confronted with issues that tax personal resources or with problems that cannot be solved. In addition, providing professional support for dealing with traumatic childhood experiences early in life may exert a positive impact on psychological health and may help to create healthy and functional relationships for the years to come. Enhancing the development of psychologically strong young individuals would be a valuable societal public health goal, which should be complemented by active political measures to fight the roots of childhood adversity, as mentally healthy citizens able to shoulder personal critical life events such as partner loss are the best basis of a well-functioning and prospering society. Finally, as shown in this chapter, intervention strategies can be tailored to increase individuals’ resources and reserves at different stages of the vulnerability process, starting from childhood as a preventive strategy to adulthood as a compensatory strategy.