Sage Journals: Discover world-class research

Abstract

Spousal bereavement is one of the most disruptive life events encountered during adulthood. However, we know little about whether and how the impacts of spousal loss on well-being have changed over the past decades. To examine such historical shifts, we applied area under the curve (AUC) metrics and latent basis growth models to multi-year within-person longitudinal change data from 2,044 participants (M_{age at event} = 65.73 years, 70% women) obtained annually across almost four decades from the German Socio-Economic Panel (1984–2020). We found that compared to the 1980s and 1990s, those experiencing spousal loss in the 2000s and 2010s showed less severe declines in well-being between 5 years before and after the loss. This improvement was driven mostly by shallower anticipatory declines and faster recovery in the adaptation phase (both by about 0.1 SD per 10 years of historical time), rather than changes in the immediate reaction phase. We found little to no evidence that the role of socio-demographic, health, and social factors as moderators of bereavement has changed across the past 40 years. We take our findings to highlight that both historical context and individual difference factors are shaping how people experience critical life events such as bereavement.

Plain language summary

The experience of losing one’s spouse has presumably changed in the 21^st century alongside, for example, the increased age of widowhood and improved healthcare as well as changing nature of social resources available. But are these historical shifts accompanied by historical differences in how well-being changes with spousal bereavement? Our data from a nationally representative German sample suggest that people who lost their spouses in the 2000s and 2010s showed overall less pronounced declines in life satisfaction than those who lost their spouse in the 1980s and 1990s. In more recent decades, the surviving spouses showed fewer anticipatory declines preceding the loss and also faster recovery following the loss, but were no different in the immediate reaction phase. We found little to no evidence that the role of socio-demographic, health, and social factors as moderators of bereavement has changed across the past 40 years. Our analysis highlights the important role historical context plays in shaping individuals’ functioning earound critical life events like spousal bereavement.

Keywords

Widowhood well-being cohort effects sociohistorical context socio-economic panel (SOEP)

Introduction

Spousal loss is one of the most disruptive life events encountered in adulthood (Clark & Georgellis, 2013; Holmes & Rahe, 1967). Empirical studies have repeatedly demonstrated that while spousal loss often undermines the health and well-being of the surviving partner (Bennett & Soulsby, 2012; Stroebe et al., 2007), it is often followed by recovery and adaptation over subsequent years (Infurna et al., 2017; Lucas et al., 2003). Less is known regarding whether and how the displacements in well-being surrounding spousal loss have changed historically over the past decades. Initial evidence indicates that spousal loss-related declines in physical health (Liu, 2012) and social and financial support (Perrig-Chiello et al., 2016), and increases in loneliness (van Tilburg & Suanet, 2019) are less pronounced now than they were in previous decades. To examine historical changes in displacement of well-being surrounding spousal loss, we analyze longitudinal panel data obtained annually across almost four decades (37 waves) by the German Socio-Economic Panel (1984–2020). Using area under the curve (AUC) and multiphase latent basis growth models, we investigate historical changes in life satisfaction trajectories in the 5 years prior to and after spousal loss and whether the role of socio-demographic, health, and social factors as moderators of bereavement has changed across the past 40 years.

Spousal bereavement and trajectories of well-being

Life satisfaction is a cognitive-evaluative indicator of well-being, indexing people’s judgement of their life circumstances in relation to their goals, norms, preferences, and needs (Diener, 1984). Contrasting empirical research chronicling the relative long-term stability of life satisfaction (e.g., Lucas & Donnellan, 2012; Schimmack & Oishi, 2005), empirical evidence suggests that life satisfaction is shaped and disrupted by major life events (Infurna et al., 2021; Krämer et al., 2025; for review, see Buhler et al., 2024; Luhman et al., 2012), and exhibits systematic age-related changes across lifespan (for review, see Buecker et al., 2023).

In the context of spousal bereavement, existing literature outlines three distinct phases of change in well-being: anticipation, reaction, and adaptation. The anticipation phase refers to declines in well-being experienced by the to-be-bereaved spouses before the actual loss (see Glaser & Strauss, 1968; Kastenbaum & Costa Jr., 1977), often characterized by increasing vulnerabilities, stressors, and burdens (e.g., caregiving, disengagement from conjoint partner goals; (Boerner & Heckhausen, 2003; Hajek & König, 2018), and lowered well-being in general (Krämer & Bleidorn, 2024). The reaction phase refers to precipitous declines in well-being around the death of one’s spouse that constitute intense feelings of grief and mourning, loss of a sense of security, and contention with the often-profound changes in everyday life routines (Brickman & Campbell, 1971). Finally, the adaptation phase refers to a period following spousal loss where the surviving partners’ well-being may return to pre-loss levels of well-being (Diener et al., 2006; Lucas, 2007a).

The existing literature even suggests a time frame for the different phases of bereavement. Two studies using nationally representative German samples reported that the anticipatory declines in life satisfaction of to-be-bereaved spouses amounted to 0.43 SD¹ over the 5 years, and to 0.37 SD² across the decade preceding the death of the spouse (Infurna et al., 2017; Wünsche et al., 2020). Lucas (2007a) found that life satisfaction takes up to 8 years to return back to average pre-loss levels. We extend previous findings by examining how these multi-year processes of anticipation, reaction, and adaptation to spousal bereavement may have changed over the last four decades.

The role of historical change

Life course theory (Elder et al., 2003) suggests that changes in macrosocial contexts and the availability of resources shape individual functioning, including how individuals experience transitions like spousal bereavement. For example, historical changes in wealth accumulation may have changed the financial consequences of widowhood. In the United States, women’s employment between ages 25 and 54 years has risen from 48% in 1970 to 75% in 2000 (England et al., 2020), and in Germany, women’s employment between ages 15 and 64 years increases from 57% in 1991 to 73% in 2019 (Statistisches Bundesamt, 2023). As such, women are less dependent today on the income and pension of their spouse (Carr, 2006). Similar historical trends, wherein older adults today are cognitively fitter (Gerstorf et al., 2023), report lower levels of loneliness (Suanet & van Tilburg, 2019), and have an increased sense of personal control over outcomes (Gerstorf et al., 2019), may also result in having more resources to draw from when dealing with challenges. Indeed, not having a partner was less predictive of loneliness among later-born cohorts of German middle-aged adults (born 1957–1974) than among those born earlier (born 1939–1956; Böger & Huxhold, 2020). Together, these historical trends suggest that today’s older adults may be better able to cope with the experience of spousal loss and thus experience less pronounced bereavement-related declines in life satisfaction than older adults living in prior decades.

Historical shifts in bereavement effects on well-being may result from demographic changes in widowhood. Bereavement now often occurs later in life (Carr, 2006; Martin-Matthews, 2011). For instance, in Germany, the average age for widowhood in 2020 increased by about 3 years in comparion to 10 years earlier (Deutsche Rentenversicherung, 2024). Timing of such life events is consequential (Elder, 2003). Lower birth rates over recent decades (United Nations, n.d.; Carr & Utz, 2001) suggest that later-bereaved spouses may have fewer adult children to provide support—although this may be partially compensated by older widows having more friends (Suanet & Huxhold, 2020). Pension reforms, such as Germany’s Riester reform (2001) that lowered the “large widow pension” from 60% to 55% of the deceased spouse’s pension, have further impacted surviving spouses, despite retained benefits for raising young children (Wilke, 2009). Together, these trends may lead to steeper declines in life satisfaction for later-bereaved spouses.

The effects of bereavement on well-being may be more pronounced today in one phase of bereavement, but less pronounced in another phase. The combination of people dying in the 2000s showing steeper terminal declines in well-being than case-matched controls dying in the 1990s (Hülür et al., 2015) alongside rising levels of caregiver responsibilities (Carr, 2006) suggest that to-be-bereaved spouses may experience steeper well-being declines today in the anticipation phase than in the past (Wünsche et al., 2020). In contrast, a number of historical changes may have contributed to less steep declines today in the reaction phase and to faster recovery. Such beneficial historical trends include preserved health and cognitive resources (Fries et al., 2011; Gerstorf et al., 2015, 2023), more diverse and larger non-kin social networks (Suanet & Huxhold, 2020), and caregivers now having better access to psychosocial support systems (Ferrell & Wittenberg, 2017).

Only a few empirical studies have examined historical shifts in how trajectories of key health and psychosocial variables are shaped by spousal bereavement. Liu (2012) found that bereavement-related health declines were less strong in the later-born 1920s birth cohort relative to the earlier-born 1910s cohort. Perrig-Chiello and colleagues (2016) found that Swiss older widowers and particularly widows in 2011 suffered from fewer social and financial difficulties than their age peers in 1979. Notably, these studies consistently found that bereavement-related declines were less pronounced among later-born cohorts. Whether historical trends in such specific domains of life translate to a more global indicator of well-being remains is not yet known. In our present analysis, we draw from and extend these earlier reports to life satisfaction, a cognitive evaluation of one’s life in general (Diener, 1984). Further, we move another step forward by using four decades of annual data that enable us to examine phase-specificity in these historical changes.

The role of individual and relationship factors

One common theme in the study of bereavement-related change is that there is substantial heterogeneity in how people are affected by the loss (Bonanno et al., 2002; Lucas et al., 2003). Stress models (Stroebe et al., 2007) and empirical research indicate that vulnerability to spousal bereavement is shaped by age (Bonanno & Kaltman, 1999), education (Nielsen et al., 2019), health (Lucas, 2007a, 2007b), caregiving (Schulz et al., 2003), and social resources (Infurna & Luthar, 2017). Little is known, however, about how the relevance of these factors has changed over time. The one study examining these changes found that bereavement-related health declines in older ages are less severe in more recent cohorts (Liu, 2012). In this study, we specifically examine historical change in how gender, length of marriage, and parenthood status moderate bereavement-related change.

Gender differences, including economic strains for women (Halleröd, 2013) and loneliness for men (Lee & DeMaris, 2007), are long-standing (Dykstra & de Jong Gierveld, 2004; Sasson & Umberson, 2014). However, as gender roles have become less traditional and more egalitarian over time (e.g., women today are less financially dependent on their husbands because of higher education and earnings: Bryant & Peck, 2009), gender differences in spousal bereavement may have narrowed. People in longer marriages often report more happiness (Hansen, 2006; Pardo et al., 2013), but divorce rates have increased (Germany: Grünheid, 2013), suggesting a shift toward marriages based on emotional attachment rather than financial dependence. It is thus possible that surviving spouses from longer marriages are more vulnerable today (Bryant & Peck, 2009). Finally, adult children often provide emotional, instrumental, and financial support to bereaved spouses (Guiaux et al., 2007; Ha, 2010). Because it becomes more and more common for older adults to not have kids and to receive support from non-family members (Suanet & Huxhold, 2020), it is possible that gaps associated with having children or not are narrowing today (see Online Supplemental material, p. 1, for a more detailed review of literature on the role of these moderator variables).

The present study

The current report capitalizes from and extends the earlier Infurna et al. (2017) study of spousal bereavement by using almost four decades of annual data obtained in 37 waves (1984–2020) of the SOEP from the bereaved spouses to examine historical shifts in (a) overall bereavement-related trajectories of life satisfaction, (b) phase-specific patterns, and (c) the role of socio-demographic, health, and social factors as moderators of bereavement. Overall, we hypothesize that bereavement-related declines have become less pronounced over historical time. We hypothesize that later-bereaved adults experience steeper declines in life satisfaction in the anticipation phase before spousal loss combined with less steep declines during the reaction and faster recovery during the adaptation phases as compared with earlier-bereaved adults. In the main text, we examine historical changes in how gender, length of marriage, and parenthood status moderate phase-specific trajectories. In particular, we expect that disparities associated with gender and having children have narrowed over historical time, whereas those associated with length of marriage have widened over time. Exploratory examination of historical changes in moderating role of the other between-person factors is in the Online Supplemental material.

Method

Transparency, openness, and data availability

The German Socio-Economic Panel Study (SOEP) is an ongoing panel study representative of adults living in private households (Kroh et al., 2008) in Germany initiated in 1984. De-identified data are freely available to registered researchers at https://www.diw.de/soep. Details about the SOEP study design, questionnaires, and measures including the sample structure, subsamples, and panel attrition have previously been reported (Goebel et al., 2019). Our hypotheses and analytic strategies were pre-registered at https://osf.io/7hce2/files/osfstorage (see also for deviations from the original pre-registration).³ The documentation of variables, the R-scripts (for data preparation and analyses), and the Mplus scripts (for analyses) are available at https://osf.io/mypzc/files/osfstorage.

Participants and procedure

The SOEP is a nationally representative yearly panel study of private households and their inhabitants of age 16 years and above. Data were collected via face-to-face interviews and self-administered mail questionnaires. Data collection, as part of the national SOEP, has been approved by the research ethics committee of the German Institute for Economic Research. For the present study, we analyzed all available data from 2,044 participants who (a) experienced spousal loss over the course of the study, (b) had provided life satisfaction data at least for one wave prior to and one wave following spousal loss, and (c) had provided data for disability (of self and partner) and social participation for one or more waves prior to spousal loss. Participants in this subsample were, on average, aged 65.73 years at spousal loss (SD = 13.05 years, range 22–100), were 70% (n = 1441) women, and had attained an average of 11.03 years of education (SD = 2.26 years, range 7–18); 9.4% of the subsample reported to have re-partnered after spousal bereavement. Participants provided a total of 17,262 longitudinal observations.

Measures

Well-being

We used life satisfaction as an indicator of participant’s well-being, measured annually with a single item, “How satisfied are you with your life, all things considered?”, on an 11-point rating scale (where 0 = completely dissatisfied and 10 = completely satisfied). This item is a measure of cognitive-evaluative (as opposed to emotional) aspects of well-being and has been used extensively in psychological research (e.g., Gerstorf et al., 2008; Lucas, 2007b). This single-item measure shows convergent validity with theoretical considerations (e.g., regarding age-related mean level changes) and empirical findings obtained with more comprehensive multi-item scales (Cheung & Lucas, 2014; Kunzmann et al., 2013). Following earlier work (Gerstorf et al., 2014), and to facilitate comparability with previous studies on well-being changes with spousal bereavement using subsamples of the SOEP (e.g., Wünsche et al., 2020), scores for life satisfaction scores were T-standardized (M = 50; SD = 0) using the 2002 SOEP sample as the point of reference (M = 6.90, SD = 1.81 on the 0 to 10 scale).

Time to/from spousal loss

Spousal loss was operationally defined with responses to the question “Has your family situation changed since the beginning of year X [e.g., 2002]?” in the category “spouse/partner has died.” We examined both married spouses and unmarried partners in the current study. Participants indicating a change in their family situation since the beginning of year X were asked a follow-up question regarding the month of occurrence of this event. Using these data, timing of spousal loss was defined as the month and year the participant reported their spouse/partner died. For the current analysis, we calculated the time (in months) until or since the spousal death, relative to the month and year of each well-being assessment. To directly compare our findings with those of Infurna et al. (2017), we followed their precedent and used data obtained from 60 months before to 60 months after the spousal death, also ensuring that the design captures any enduring changes missed by earlier studies due to their limited time frames. We further followed the approach taken by Uglanova and Staudinger (2013) and grouped the monthly data into 3-month (quarterly) intervals (e.g., −60 to −58 months, −57 months to −55 months, and so on), centered at spousal death.

Historical time

Historical time was operationally defined as year of spousal loss, specifically, the calendar year of death of the deceased spouse, as reported by the surviving spouse. In the statistical analyses, year of spousal loss was centered at the sample median, 2005, and used as a continuous variable that allows for testing linear and quadratic effects and circumvents the challenge of a priori defining cohorts categorically using cutoff points. However, for easier visualization, we present life satisfaction trajectories for two artificially created cohort groups in the figures, based on median split (a practical approach in absence of a more substantial rationale based on historical events), where the spousal death occurred either before 2005 or in/after 2005. Frequencies of observations across months to/from spousal loss separately for the two cohorts are shown in Figure 1. It can be obtained from those distributions that across the 10 years of observation, the frequency of data was by and large equally distributed across the two cohorts.

Figure 1.

Frequency of observations across months to/from spousal loss separately for the two cohorts. Note. It can be obtained that across the 10 years of observation, the frequency of data was by and large equally distributed across the two cohorts, N_{(partner died before 2005)} = 999 and N_{(partner died in or after 2005)} = 1045.

Table 1 shows the cohort-specific means of life satisfaction observations in the period spanning from 60 months prior to 60 months following the spousal loss.

Table 1.

Descriptive Statistics for T-Scores of Life Satisfaction in Relation to Spousal Loss.

Months to/from spousal loss	Year of spousal loss	No. of obs.	Mean	Standard deviation	Standard error of means
−60 to −58	1985–2004	147	52.31	10.66	0.88
	2005–2020	230	50.39	9.94	0.66
−57 to −55	1985–2004	188	51.73	10.58	0.77
	2005–2020	211	51.55	9.09	0.63
−54 to −52	1985–2004	180	51.45	9.99	0.75
	2005–2020	229	50.59	10.13	0.68
−51 to 49	1985–2004	148	50.88	10.12	0.83
	2005–2020	222	49.85	10.81	0.73
−48 to −46	1985–2004	180	50.70	10.64	0.80
	2005–2020	239	49.40	10.79	0.70
−45 to −43	1985–2004	207	50.65	11.09	0.77
	2005–2020	216	50.46	10.56	0.72
−42 to −40	1985–2004	190	51.28	10.86	0.79
	2005–2020	224	50.33	9.54	0.64
−39 to −37	1985–2004	175	51.80	10.18	0.77
	2005–2020	252	50.53	10.39	0.66
−36 to −34	1985–2004	207	51.00	10.64	0.74
	2005–2020	251	49.37	10.50	0.66
−33 to −31	1985–2004	223	50.44	10.60	0.71
	2005–2020	229	51.59	9.86	0.65
−30 to −28	1985–2004	220	50.88	11.45	0.77
	2005–2020	239	49.29	10.18	0.66
−27 to −25	1985–2004	202	50.83	10.24	0.72
	2005–2020	257	49.49	9.75	0.61
−24 to −22	1985–2004	213	49.36	11.86	0.81
	2005–2020	236	49.60	11.18	0.73
−21 to −19	1985–2004	238	50.80	11.09	0.72
	2005–2020	264	49.66	10.15	0.63
−18 to −16	1985–2004	256	50.13	11.27	0.71
	2005–2020	236	49.62	11.20	0.73
−15 to −13	1985–2004	234	49.92	10.09	0.66
	2005–2020	280	49.09	10.72	0.64
−12 to −10	1985–2004	224	48.11	11.63	0.78
	2005–2020	249	48.24	10.88	0.69
−9 to −7	1985–2004	257	49.47	11.74	0.73
	2005–2020	255	49.05	11.24	0.71
−6 to −4	1985–2004	234	48.33	11.29	0.74
	2005–2020	233	47.87	11.29	0.74
−3 to 0	1985–2004	231	47.24	11.65	0.77
	2005–2020	251	46.75	11.70	0.74
0 to 3	1985–2004	281	41.18	12.56	0.75
	2005–2020	346	41.44	12.96	0.70
4 to 6	1985–2004	266	42.42	12.37	0.76
	2005–2020	257	43.62	12.13	0.76
7 to 9	1985–2004	271	44.27	12.07	0.74
	2005–2020	262	44.56	12.39	0.77
10 to 12	1985–2004	222	45.83	12.18	0.82
	2005–2020	244	43.97	12.72	0.81
13 to 15	1985–2004	223	46.74	10.93	0.73
	2005–2020	230	46.91	10.57	0.70
16 to 18	1985–2004	264	47.13	11.76	0.72
	2005–2020	221	48.49	10.72	0.72
19 to 21	1985–2004	229	47.78	11.14	0.74
	2005–2020	206	48.07	10.73	0.75
22 to 24	1985–2004	215	47.61	10.95	0.75
	2005–2020	224	49.53	10.72	0.72
25 to 27	1985–2004	213	48.12	11.07	0.76
	2005–2020	191	48.56	10.40	0.75
28 to 30	1985–2004	254	48.54	11.63	0.73
	2005–2020	175	50.12	10.11	0.77
31 to 33	1985–2004	207	48.14	10.34	0.72
	2005–2020	186	49.15	10.10	0.74
34 to 36	1985–2004	200	49.59	10.43	0.74
	2005–2020	183	49.99	10.74	0.79
37 to 39	1985–2004	199	49.99	10.32	0.73
	2005–2020	181	48.86	10.87	0.81
40 to 42	1985–2004	235	49.31	11.02	0.72
	2005–2020	144	49.75	10.10	0.84
43 to 45	1985–2004	190	49.17	10.94	0.80
	2005–2020	161	49.76	10.76	0.85
46 to 48	1985–2004	188	48.73	11.00	0.81
	2005–2020	155	50.14	11.49	0.92
49 to 51	1985–2004	181	48.97	11.51	0.86
	2005–2020	136	49.00	10.75	0.93
52 to 54	1985–2004	221	49.80	10.52	0.71
	2005–2020	137	48.56	11.41	0.97
55 to 57	1985–2004	176	49.86	10.29	0.77
	2005–2020	136	52.17	9.12	0.78
58 to 60	1985–2004	177	50.95	9.75	0.73
	2005–2020	116	51.83	9.61	0.90

Notes: N = 2044. Number of observations = 17,262.

Individual and relationship factors

We examined how socio-demographic factors, health and social resources, and partnership factors were related to bereavement-related changes in life satisfaction (for descriptives, see Table SOM.1 in the Online Supplemental material).

Socio-demographic variables

Age at experiencing spousal bereavement was computed by subtracting birth year of a given participant from the year they reported to have experienced spousal loss, and centered at the sample mean. Figures SOM.1 and SOM.2 in the Online Supplement show the distribution of age at event for the two cohorts and the correlation between the age at event and year of spousal loss, respectively. Gender was coded as a binary variable (0 = men and 1 = women) with no information available in the survey on nonbinary gender identity. Education was measured as the number of years necessary to obtain the final school degree achieved.

Health resources

Health resources were operationally defined as disability status of both the individual (before death of spouse) and the spouse, as derived from an item asking each year whether the respondent was “officially certified as having a reduced capacity to work or being severely handicapped” (for details, see Infurna et al., 2017; Lucas, 2007b). These repeated annual assessments were used as a time-invariant variable, coded as 1 when the respondent reported to be disabled at any point in the 5 years before death (of the spouse) and 0 when not.

Social resources

Social resources, operationally defined as social participation, were indexed by the frequency of engagement in four sets of activities: (a) visit to cultural events such as concerts, theaters, or lectures; (b) active sports participation; (c) honorary activities in clubs, organizations, or social service; and (d) participation in citizen initiatives, parties, or community politics (see Headey et al., 2010; Infurna et al., 2017). Responses were reverse coded and averaged to obtain an index with higher scores indicating more social participation. Repeated observations of an individual’s social participation obtained in the available periods prior to spousal loss were averaged, in order to obtain an index of social participation before experiencing spousal loss.

Partnership factors

We examined two partnership factors. First, length of spousal relationship was defined as the number of years since the beginning of a romantic partnership (both marriage and unmarried cohabitation) with the deceased spouse until the death of the spouse. Second, having children was coded as a dichotomous variable: no children (=0) and at least one child (=1). See Online Supplemental material, p. 1–2, for a more detailed variable description of the health and social resources.

Data analyses

To examine our research questions, we made use of area under the curve (AUC) metrics and latent basis growth models (Grimm et al., 2016).

Historical change in overall well-being changes with bereavement

The overall changes in life satisfaction individuals experienced around spousal loss were operationalized using AUC negative displacement scores (i.e., negative changes from the baseline) based on T-transformed life satisfaction scores. A similar analytical approach was previously used by Kettlewell et al. (2020). Specifically, we defined the baseline level of well-being as the maximum value of life satisfaction in the pre-loss phase. Then, on the basis of this baseline value, we defined negative displacement as any changes in the negative direction from baseline and computed AUC as the area under the natural cubic spline interpolation of the life satisfaction scores. Higher AUC negative displacement scores reflect greater overall disruption of life satisfaction. To test our hypothesis that later-bereaved spouses show overall less loss-related disruption, we regressed the AUC negative displacement scores on the year of spousal loss, and in a subsequent step also the individual and relationship variables.

We first computed the AUC negative displacement directly using the annual reports of life satisfaction scores. However, these annual reports may be considered “noisy” because they are multiply-determined in that they are tracking all of life circumstances, not just the bereavement-related changes. To reduce influence of other situational factors on quantification of individuals’ bereavement-related displacement, we also computed AUC negative displacement using “filtered” life satisfaction score estimates obtained from the growth model described below. In principle, these “filtered” scores are more directly focused on the specific aspects of life satisfaction that are being influenced by the spousal loss. As well, we engaged a variety of additional models that make use of alternative operational definitions of the AUC (see Tables SOM.2a and SOM.2b), and where historical time was operationally defined as year of birth of the surviving spouse rather than as year of spousal loss (see Table SOM.3). Across these analyses, the substantive pattern of results generally mirrors the findings reported in the main text.

Historical change in phase-specific well-being changes with bereavement

We examined our research questions about phase-specific historical shifts with a set of three models. Following the approach used by Infurna et al. (2017), we used multiphase latent basis growth model (see Ram & Grimm, 2007; Singer & Willet, 2003), consisting of an intercept (i.e., the level of life satisfaction 5 years before spousal loss), and three slope parameters: anticipation, reaction, and adaptation. More specifically,

{l i f e s a t i s f a c t i o n}_{ti} = [g_{0 i}] + [g_{1 i} \cdot λ_{1 t}] + [g_{2 i} \cdot λ_{2 t}] + [g_{3 i} \cdot λ_{3 t}] + e_{ti}

(1)

where person i’s score on life satisfaction at time t, life satisfaction_ti, is modeled as a function of a person-specific intercept coefficient, g_0i; a set of person-specific slope coefficients (factor scores), g_1i to g_3i, paired with a corresponding set of time functions (factor loadings, labeled λ_1t to λ_3t in the equations) for the anticipation, reaction, and adaptation phases of bereavement, respectively; and residual error, e_ti.

In a first model, we added age as an additional predictor to the unconditional growth model. An individual-specific intercept quantifies the expected level of life satisfaction prior to spousal loss, g_0i (−60 months to −49 months). Further, we also have three distinct slope factors: anticipation, reaction, and adaptation. Anticipation (g_1i), reaction (g_2i), and adaptation factors (g_3i) indicate the extent of change in life satisfaction in the years preceding spousal loss (−48 months to −4 months), in the 6 months surrounding spousal loss (−3 months to +3 months), and in the months and years following spousal loss (+4 months to +60 months). The factor means describe the baseline level and extent of changes in life satisfaction for a prototypical widow/er, and the factor variances refer to the degree to which individuals differ in each of the phases. In a second model, we regressed year of spousal loss on each of the intercept, anticipation, reaction, and adaptation factors.

g_{0 i} = β_{00} + β_{01} {a g e}_{i} + β_{02} {y e a r o f s p o u s a l l o s s}_{i} + u_{0 i}

(2)

g_{1 i} = β_{10} + β_{11} {a g e}_{i} + β_{12} {y e a r o f s p o u s a l l o s s}_{i} + u_{1 i}

(3)

g_{2 i} = β_{20} + β_{21} {a g e}_{i} + β_{22} {y e a r o f s p o u s a l l o s s}_{i} + u_{2 i}

(4)

g_{3 i} = β_{30} + β_{31} {a g e}_{i} + β_{32} {y e a r o f s p o u s a l l o s s}_{i} + u_{3 i}

(5)

The role of individual and relationship factors

In a third model, we examined how individual and relationship factors were related to each component of the change processes by regressing the socio-demographic, health and social resources, and partnership-related factor variables on the level, anticipation, reaction, and adaptation growth factors. Specifically, the model was expanded such that

g_{0 i} = β_{00} + β_{01} {a g e}_{i} + β_{02} {y e a r o f s p o u s a l l o s s}_{i} + \dots + β_{09} {l e n g t h l o f m a r r i a g e}_{i} + β_{0201} {y e a r o f s p o u s a l l o s s}_{i} * {a g e}_{i} + \dots + β_{0209} {y e a r o f s p o u s a l l o s s}_{i} * {l e n g t h l o f m a r r i a g e}_{i} + u_{0 i}

(6)

g_{1 i} = β_{10} + β_{11} {a g e}_{i} + β_{12} {y e a r o f s p o u s a l l o s s}_{i} + \dots + β_{19} {l e n g t h l o f m a r r i a g e}_{i} + β_{1211} {y e a r o f s p o u s a l l o s s}_{i} {* a g e}_{i} + \dots + β_{1219} {y e a r o f s p o u s a l l o s s}_{i} {* l e n g t h o f m a r r i a g e}_{i} + u_{1 i}

(7)

g_{2 i} = β_{20} + β_{21} {a g e}_{i} + β_{22} {y e a r o f s p o u s a l l o s s}_{i} + \dots + β_{29} {l e n g t h l o f m a r r i a g e}_{i} + β_{2221} {y e a r o f s p o u s a l l o s s}_{i} {* a g e}_{i} + \dots + β_{2229} {y e a r o f s p o u s a l l o s s}_{i} {* l e n g t h o f m a r r i a g e}_{i} + u_{2 i}

(8)

g_{3 i} = β_{30} + β_{31} {age}_{i} + β_{32} {year of spousal loss}_{i} + \dots + β_{39} {length of marriage}_{i} + β_{3231} {y e a r o f s p o u s a l l o s s}_{i} {* a g e}_{i} + \dots + β_{3239} {y e a r o f s p o u s a l l o s s}_{i} {* l e n g t h o f m a r r i a g e}_{i} + u_{3 i}

(9)

The AUC metrics and subsequent multiple regression models were estimated using R (DescTools and lme4 packages, respectively), and the latent basis growth models were estimated using MPlus version 8.10 (Muthén & Muthén, 2017), with incomplete data accommodated under missing at random assumptions at the within-person level, and, to retain longitudinal data, missing completely at random at the between-person level (Little & Rubin, 1989; set-ups provided in Ram & Grimm, 2007). We tested all possible two-way interaction effects of the individual and relationship variables with the phase-specific changes as well as interaction effects with the year of spousal loss variable but for parsimony of presentation trimmed the final model only to the main effects as none of the interaction effects were statistically significant.

Results

Historical change in overall well-being changes with bereavement

Our analysis began by calculating the AUC negative displacement of life satisfaction scores for each individual, based on observed (T-transformed) scores surrounding spousal loss. For the purpose of illustration, we show the AUC negative displacements for two groups of 10 randomly selected individuals each who lost their spouses before 2005 (sample median of year of spousal loss) and in/after 2005 in Figure 2.

Figure 2.

AUC negative displacements plotted using observed (T-transformed) scores of life satisfaction surrounding spousal loss for random samples of 20 participants who lost their spouses before (upper panels, in red) or after (lower panels, in blue) 2005.

As reported in Model 1 of Table 2, historical time (i.e., year of spousal loss) was not significantly related to the AUC negative displacement scores computed from the T-transformed observed scores of life satisfaction (γ₀₁ = −0.63, p = .058). That is, contrary to our expectations, we did not find evidence for historical improvements in the overall bereavement-related displacement of life satisfaction of surviving spouses.

Table 2.

Regressing AUC Negative Scores (Observed) and AUC Negative Scores (Filtered) on Year of Spousal Loss and Other Individual Differences.

Parameters	Model 1 (based on observed scores of life satisfaction)		Model 2 (based on filtered scores of life satisfaction)
	N = 1672		N = 1672
	Est.	SE	Est.	SE
Intercept, γ₀₀	−88.25***	2.96	−379.93***	6.75
Year of spousal loss, γ₀₁	−0.63	0.33	3.40***	0.76
Age, γ₀₂	0.66	0.38	−0.14	0.87
Gender, γ₀₃	0.21	6.66	−23.70	15.18
Education, γ₀₄	1.36	1.47	4.05	3.35
Disability preγ₀₅	−7.36	6.75	−19.95	15.40
Disability partner, γ₀₆	4.10	6.03	41.43**	13.74
Social pre, γ₀₇	14.10*	6.39	−9.80	14.57
Parental status, γ₀₈	15.03	9.78	22.87	22.31
Relationship length, γ₀₉	−0.17	0.30	0.16	0.68
F(df)	1.58 (9, 1663)		4.47*** (9, 1663)
R²	0.01		0.02
Adj. R²	0.00		0.02

Note. AUC negative score (observed) = total amount of negative displacements from baseline levels of observed T-transformed scores of life satisfaction, i.e., the maximum observed value of life satisfaction in the pre-loss phase.

AUC negative score (filtered) = total amount of negative displacements from baseline levels of model-estimated scores of life satisfaction, i.e., the maximum estimated value of life satisfaction in the pre-loss phase.

Disability pre = disability status of individual before the spousal loss, social pre = social participation index of individual before the spousal loss.

df = degrees of freedom.

* = p < .05; ** = p < .01; *** = p < .001.

Following the latent basis growth model analysis (details described in the next section), we conducted the analysis again using the “filtered” scores of life satisfaction estimated from growth models. Specifically, we regressed the AUC negative displacement scores calculated from the smoothed trajectories on the year of spousal loss. In these analyses, as reported in Model 2 of Table 2, we found AUC negative displacement scores based on “filtered” scores were significantly related to year of spousal loss. Consistent with our prior expectations, more recently bereaved spouses showed less overall displacement of life satisfaction (γ₀₁ = 3.40, p < .001) in the period between 5 years prior to and 5 years following spousal loss. We graphically represent these two sets of findings in the two panels of Figure 3.

Figure 3.

AUC negative displacement curves plotted using average of observed T-transformed (upper Panel A) and model-estimated (lower Panel B) scores of life satisfaction surrounding spousal loss for cohorts who lost their spouses between 1984 and 2004 versus 2005 and 2020. Note. It can be obtained that losing one’s spouse in the 2000s and 2010s (in blue) was associated with overall less pronounced well-being decreases in the 5 years prior to 5 years after spousal loss than losing one’s spouse in the 1980s and 1990s (in red). Differences were statistically significant when using the AUC negative displacement scores estimated from the filtered scores (lower Panel B) but were not significant when using the original scores.

As exploratory analyses, we also regressed the AUC scores on the between-person factors and report the results in Online Supplemental material.

Historical change in phase-specific well-being changes with bereavement

Results of latent basis growth models examining our second set of research questions are reported in Model 1 of Table 3. As also seen in SOM Figure 3, the mean of life satisfaction (β₀₀) of 65-year-old participants having lost their partners in 2005 at 60 months before the death of the spouse and at the age of 65 was 51.28, roughly comparable to the average of the nation-wide SOEP sample in 2002. During the anticipation phase, life satisfaction declined by an average of −3.32 T-score units (β₀₁) over the next 45 months (3.75 years) until 3 months prior to death of the spouse. At that point, the reaction phase began and life satisfaction declined even further to −6.65 T-score units (β₀₂) over a period of 6 months (0.5 years) until 3 months after the death of the spouse. During the adaptation phase, life satisfaction recovered by an average of 8.95 T-score units (β₀₃) over the next 57 months (4.75 years) until about 5 years after the loss.

Table 3.

Fixed and Random Effects of Year of Spousal Loss and Individual Differences for Examining Historical Change in Phase-Specific Trajectories of Life Satisfaction to/from Spousal Loss.

	Model 1 (N = 2,044)		Model 2 (N = 1,672)
	Est.	SE	Est.	SE
Fixed effects
Level, β₀₀	51.28***	0.24	51.59***	0.26
Age × level, β₀₁	0.06***	0.02	0.11***	0.03
Year × level, β₀₂	−0.07**	0.02	−0.15***	0.03
Gender × level, β₀₃			−0.62	0.50
Education × level, β₀₄			−0.05	0.11
Disability pre × level, β₀₅			−3.67***	0.50
Disability partner × level, β₀₆			−0.81	0.45
Social pre × level, β₀₇			3.86***	0.48
Parental status × level, β₀₈			−1.09	0.75
Relationship length × level, β₀₉			0.01	0.02
Anticipation, β₁₀	−3.32***	0.35	−3.82***	0.40
Age × anticipation, β₁₁	−0.09***	0.02	−0.11*	0.04
Year × anticipation, β₁₂	0.11***	0.03	0.13**	0.04
Gender × anticipation, β₁₃	–	–	−0.37	0.74
Education × anticipation, β₁₄	–	–	0.04	0.16
Disability pre × anticipation β₁₅	–	–	0.52	0.73
Disability partner × anticipation β₁₆	–	–	−1.41*	0.66
Social pre × anticipation, β₁₇	–	–	0.33	0.71
Parental status × anticipation, β₁₈	–	–	0.26	1.08
Relationship length × anticipation, β₁₉	–	–	−0.00	0.03
Reaction, β₂₀	−6.65***	0.48	−6.49***	0.54
Age × reaction, β₂₁	0.18***	0.03	0.19***	0.05
Year × reaction, β₂₂	−0.08*	0.04	−0.08	0.05
Gender × reaction, β₂₃	–	–	−0.53	0.90
Education × reaction, β₂₄	–	–	−0.10	0.20
Disability pre × reaction β₂₅	–	–	1.13	0.91
Disability partner × reaction β₂₆	–	–	1.94*	0.81
Social pre × reaction, β₂₇	–	–	−2.00*	0.86
Parental status × reaction, β₂₈	–	–	2.14	1.33
Relationship length × reaction, β₂₉	–	–	−0.04	0.04
Adaptation, β₃₀	8.95***	0.55	9.22***	0.60
Age × adaptation, β₃₁	−0.17***	0.03	−0.19***	0.05
Year × adaptation, β₃₂	0.11**	0.04	0.10*	0.05
Gender × adaptation, β₃₂	–	–	0.46	0.90
Education × adaptation, β₃₄	–	–	0.30	0.20
Disability pre × adaptation, β₃₅	–	–	−2.28*	0.92
Disability partner × adaptation, β₃₆	–	–	1.38	0.81
Social pre × adaptation, β₃₇	–	–	0.92	0.86
Parental status × adaptation, β₃₈	–	–	−1.96	1.35
Relationship length × adaptation, β₃₉	–	–	0.04	0.04
Random effects
Level, σ²u₀	65.28***	2.92	56.27***	2.83
Anticipation, σ²_u1	49.99***	6.16	49.37***	6.59
Reaction, σ²_u2	128.12***	10.07	118.32***	10.85
Adaptation, σ²_u3	107.07***	12.44	104.87***	13.29
Residual variance, σ²_e	47.32***	0.69	46.00***	0.73
Fit statistics
CFI	0.94		0.94
RMSEA	0.02		0.01

Model 1 includes age and year as predictors; Model 2 includes other individual differences as additional predictors.

Est. = parameter estimates; SE = standard error; year = year of experiencing spousal bereavement, centered at 2005; age = age when experiencing bereavement, centered at sample mean = 65.73 years.

* = p < .05; ** = p < .01; *** = p < .001.

CFI: Comparative Fit Index. RMSEA: root mean square error of approximation.

Most important for our research question, and as graphically illustrated in Figure 4, we found for year of spousal loss that participants who lost their spouse in 2005 or later reported lower levels of life satisfaction at 60 months before spousal loss than those who lost their spouse earlier in historical time, with an effect size of β₀₂ = −0.07 (p = .002) per year or a little less than 0.1 SD per 10 years of historical time. We also found that those who lost their spouse later in historical time experienced shallower declines in life satisfaction in the anticipation phase (β₁₂ = 0.11, p < .001) and faster recovery (β₃₂ = 0.11, p = .006), whereas the reactive declines were steeper (β₂₂ = −0.08, p = .032). We further observed that, descriptively, earlier-bereaved spouses had slightly higher life satisfaction (by 0.12 SD) 5 years before the loss compared to later bereaved spouses but ended up with lower life satisfaction 5 years after the loss (0.20 SD lower than their initial levels). In contrast, later bereaved spouses nearly recovered to their initial levels, with only a 0.01 SD difference.

Figure 4.

Prototypical trajectories of life satisfaction in the year surrounding spousal loss for participants having lost their spouses before (red) or after (blue) 2005. Note. Based on the results of the multiphase latent basis growth model, it can be obtained that losing one’s spouse in the 2000s and 2010s was associated with fewer declines in life satisfaction in the anticipation phase and faster recovery in the adaptation phase than losing one’s spouse in the 1980s and 1990s, whereas no robust cohort differences were observed for the immediate reaction phase.

Consistent with earlier reports from a subsample of the current dataset (Infurna et al., 2017, see also Figure SOM.4 in the Online Supplemental material), older age was associated with higher baseline levels of well-being (β₀₁ = 0.06, p < .001), steeper anticipatory pre-loss declines (β₁₁ = −0.09, p < .001), less steep reactive declines (β₂₁ = 0.18, p < .001), and less steep recovery (β₃₁ = −0.17, p < .001).

The role of individual and relationship factors

In the subsequent model, we included socio-demographic factors, health and social resources, and partnership variables as well as their interaction terms with the year of spousal bereavement so as to examine whether and how the role of these factors for changes in life satisfaction with spousal bereavement have transformed historically. Results are reported in Model 2 of Table 3.

To begin with, the predictive effects of age and year of spousal loss noted in the reduced model remained statistically significant except for the reaction phase in which the effect of year was not significant (p = .094). Most relevant for our third research question, contrary to our hypotheses, we did not observe statistically significant interaction effects for any of the moderators with year of spousal loss. Thus, the moderating role of these factors has remained unchanged historically (see Online Supplemental material for the detailed report of main effects of the moderators).

In all models, the variance in the reaction (118.32–128.12, p < .001) and adaptation phases (104.87–107.07, p < .001) was consistently higher than that in level (56.27–65.28, p < .001) and anticipation (49.37–49.99, p < .001), indicating that the heterogeneity among surviving spouses was higher when coping with spousal loss than it was in the pre-loss periods.

Follow-up analyses

We ran a set of follow-up analyses to specifically test for gender differences of the effects of spousal bereavement (see Table SOM.4 in the Online Supplemental material). Both our main analyses and these follow-up analyses consistently indicated little to no differences between men and women in how adapting to spousal bereavement has changed over time. Also, in order to take into account presumably confounding effects of re-partnering, we conducted a second set of follow-up analyses, running the same models on a sub-sample excluding those who had re-partnered (see Table SOM.5 in the Online Supplemental material). Results were by and large consistent with those reported in the main analyses.

Discussion

The sociohistorical milieu of spousal bereavement has undergone considerable changes over the past decades, which might be accompanied by historical shifts in how well-being changes with the experience of spousal loss. We examined the role of historical time for life satisfaction trajectories of people who lost their spouse between 1985 and 2020 in Germany. The current study makes a novel contribution to the bereavement literature by extending the group-based approach to the heterogeneity of bereavement trajectories (Bonanno et al., 2002) to the individual level to obtain a more detailed understanding of the many different ways people respond to loss. This approach allowed for and structured the heterogeneity noted by (a) separating between different phases, (b) examining how trajectories may have changed across the past four decades, and (c) providing a new quantification of this heterogeneity with our area under the curve approach. Results of AUC analyses with original and filtered life satisfaction scores provide mixed evidence of historical differences in adjusting to spousal bereavement. Although the differences were not immediately apparent with the “noisy” life satisfaction reports, analysis of “filtered” scores suggest that, compared to their earlier-bereaved peers, later bereaved spouses showed overall less pronounced displacement of life satisfaction in the 10 years surrounding spousal loss. Going further, our study is the first to demonstrate phase-specificity in the history-graded changes. Distinguishing anticipation, reaction, and adaptation phases of bereavement revealed that historical improvements are primarily driven by shallower declines in the anticipation phase and faster recovery in the adaptation phase (both by about 0.1 SD per 10 years of historical time) rather than by changes in the immediate reaction phase.

Historical change in overall well-being changes with bereavement

We found no evidence for historical shifts in overall well-being displacements surrounding spousal bereavement when quantifying those differences using AUC negative displacement scores based on observed life satisfaction trajectories. However, such historical shifts were evident and statistically significant when quantifying extent of displacement using model-estimated smoothed life satisfaction trajectories. The observed scores likely reflect judgements about a wide variety of changes in life circumstances that manifest throughout each year. In contrast, the estimated life satisfaction scores obtained from our latent basis growth models have filtered the data to focus on changes specifically related to spousal loss, demonstrating clearer, event-specific patterns. Our findings indicated that people who have lost their spouses in more recent years were likely to experience fewer overall declines in life satisfaction during the period stretching from 5 years before to 5 years after the loss. Such a historical trend is consistent with previous empirical reports indicating less pronounced spousal loss-related decrements today compared to the past, in other domains of life, including health (Liu, 2012), social integration (Perrig-Chiello et al., 2016; van Tilburg & Suanet, 2019), and economic conditions (Perrig-Chiello et al., 2016).

One potential explanation might be that with historically reduced economic dependence of widows on their deceased husbands (Carr, 2006) as well as historically weakened associations between not being partnered and loneliness over the past decades (Böger & Huxhold, 2020), later-bereaved spouses might be less devastated with their post-bereavement life in comparison to those bereaved earlier. Further, there is growing evidence of historical improvements in various domains of well-being in late life. To illustrate, today’s older adults are cognitively fitter (Gerstorf et al., 2023), are less lonely (Suanet & van Tilburg, 2019), and have an increased sense of personal control over outcomes (Gerstorf et al., 2019). As older adults are more likely to experience spousal bereavement than their younger peers, such historical trends of improved functioning in older ages might also contribute to better resilience to widowhood nowadays.

Historical change in phase-specific well-being changes with bereavement

When analyzing phase-specific patterns, historical improvements in spousal bereavement were mainly due to shallower declines in the anticipation phase and quicker recovery in the adaptation phase, while changes in the immediate reaction phase remained inconsistent. Contrary to our hypothesis, shallower anticipatory declines in later-bereaved spouses may reflect increased support from adult children and greater use of respite care (Wolff et al., 2018), which might alleviate caregiving stress despite higher caregiver burdens today (Carr, 2006). Additionally, Hülür et al. (2015) found that SOEP participants who died later experienced steeper terminal declines, while Wünsche et al. (2020) reported that anticipatory declines for to-be-bereaved spouses were half the size of terminal declines in the dying spouse and double those of case-matched controls. Together, these findings suggest an increasing de-coupling of well-being trajectories between spouses, with steeper declines for the dying partner and shallower declines for the bereaved. Future research should address possible contributing factors such as growing individualization as well as specific factors such as more diverse and more important non-kin social networks than their peers in the past (Suanet & Huxhold, 2020), and improved health behaviors (for overview, see Crimmins & Beltrán-Sánchez, 2011).

We also note that the earlier-bereaved spouses reported significantly higher levels of well-being 5 years before spousal loss. This finding may reflect the remarkable improvements in living conditions they witnessed due to better social welfare later in life, compared to the adversities they faced earlier, which influenced their current well-being evaluations more positively than for their later bereaved peers (Michalos, 1985; see also Schilling, 2005). In contrast, later-bereaved spouses, starting from a lower level of well-being, may have had less room for decline in the anticipation phase. Recent trends of longer life expectancies, often accompanied by extended illness periods, could also mean that anticipatory declines now stretch over longer periods, as spouses live longer with chronic conditions (Crimmins et al., 2019). Consequently, while later-bereaved spouses may have already been experiencing losses 5 years before the death, those who lost their spouses in earlier historical periods may have only entered the anticipation phase closer to the time of loss. Future research could explore this hypothesis by considering changes in the health of deceased spouses over time. In order to facilitate comparability of findings to Infurna et al. (2017), we fixed the start and end points of bereavement phases in the current study.

In the reaction phase, we had expected that later-bereaved spouses would display shallower declines than earlier-bereaved spouses. However, our findings indicate an inconsistent and weak pattern. Initially, later-bereaved spouses showed steeper declines around the time of loss, but this effect was no longer significant after accounting for moderator variables. Thus, there is no evidence for historical improvements in immediate reactions to spousal loss. While two out of three phases show historical improvements, the third does not, potentially due to opposing historical changes. For instance, middle-aged adults in Germany have shown improvements in well-being, physical health, and reduced loneliness (Infurna et al., 2021, 2024). Additionally, older adults today report to have lower external control beliefs and higher internal control beliefs compared to their counterparts in the past (Drewelies, Agrigoroaei et al., 2018; Drewelies, Deeg et al., 2018), even over adverse outcomes (Gerstorf et al., 2019). Although these trends are generally beneficial, they may increase feelings of guilt or self-blame following spousal loss due to elevated internal control beliefs. Future research should therefore explore the impact of these evolving control beliefs in coping with spousal loss.

We also found evidence of historically faster adaptation to spousal loss. Those bereaved earlier showed more long-lasting effects of the loss, whereas those bereaved later returned almost to their pre-loss levels of life satisfaction 5 years after the loss. Following the HIstorical changes in DEvelopmental COntexts (HIDECO) framework (Drewelies et al., 2019), macro-level societal changes such as better access to individual resources (e.g., quantity and quality of education) and innovation in science and technology (e.g., advances in medicine and nutrition) have contributed to improved individual-level functioning in later life. We argue that such historical improvements also encompass greater availability of coping resources and adaptive strategies in later life, thereby facilitating adaptation to adverse life events. The decreasing importance of marriages in providing economic security and instrumental support, particularly for women (Coontz, 2007; Teachman et al., 2000), might have also tempered the negative financial consequences of spousal bereavement. Future research should address such mechanisms in more detail.

The role of individual and relationship factors

The final objective of the present study was to examine the role of various between-person factors in the historical change of coping with spousal bereavement. Despite an extensive literature documenting how the aftermath of losing a spouse differs for men and women (Dykstra & de Jong Gierveld, 2004; Sasson & Umberson, 2014), and a shift toward more egalitarian views on marriage in the recent years (Bryant & Peck, 2009; Family Caregiver Alliance, 2019), our findings did not reveal any evidence for (historical change in) the role of gender in the context of spousal bereavement. However, our findings align with Infurna et al. (2017), who had examined a subsample of the current data and a similar analytical approach. Possible reasons for discrepancies with past research include our broader age range, focus on life satisfaction (a cognitive rather than affective indicator of well-being), and more precise time-scale analysis (Infurna et al., 2017).

Further, in contrast to our expectations, we did not find support for the role of marital length and parenthood status—neither in general nor in context of historical change—in shaping the experience of spousal bereavement. As suggested by previous literature, evolving marital expectations and increased availability of non-family social support can potentially affect the well-being of bereaved spouses, particularly in the context of longer marriages and childlessness (Bryant & Peck, 2009; Suanet & Huxhold, 2020; van Tilburg & Suanet, 2019). However, the changing views on marriage and non-parenthood were not yet prominent among participants of the current study who were relatively old (in their mid-sixties, on average) when they lost their spouse, thereby not reflecting the expected trends. Future studies should examine these moderation effects on more recent cohorts of bereaved spouses to provide a deeper understanding of how changing societal norms shape bereavement experiences in a more modern context.

As exploratory analyses, we also tested for the historical change in the relevance of other between-person factors including age, education, health, caregiving responsibilities, and social resources in relation to spousal bereavement. Toward that end, we found little to no evidence that the relevance of these factors has been changing across the past 40 years. Whether such stability in functioning is universal is an open question to be addressed by future research using different indicators and different samples.

Most importantly, cause of death of the partner—whether expected or unexpected—requires significant attention, particularly when it involves chronic illnesses like dementia, which often entails witnessing the intense suffering of the spouse and extensive caregiving. Anticipatory declines in well-being may intensify when a spouse endures a prolonged period of terminal illness, whereas sudden, unanticipated deaths may lead to a steeper reaction phase. Carr et al. (2006) note that death has become more gradual and anticipated, typically following prolonged, debilitating illnesses. Over the study period in Germany, causes of death have changed, with dementia being the most prevalent disease among women over 70 from 2014 to 2017 (Doblhammer et al., 2022). This trend is important because dementia caregivers experience unique stressors and often show rapid improvement in well-being following the death of their spouses (Schulz et al., 2003). However, because data on causes of death were only available for 10 of the 30+ years, this factor could not be included in our analyses (see Online Supplemental material, p. 2–5, for a detailed discussion on the main effects of the moderator variables).

Limitations and outlook

In closing, we note several limitations of our study design, measures, and samples. First, we could not examine whether observed differences were due to early-life or later-life factors. We used year of spousal loss as a marker of historical time, focusing on later-life experiences. Follow-up analyses using year of birth (as a proxy for early-life experiences) revealed similar findings, suggesting both sets of factors may be involved. Second, cutoff points for historical time should be ideally based on specific events or policy changes relevant to spousal bereavement. However, in the absence of such prominent events in Germany, we treated historical time (year of spousal loss) as a continuous variable and used a practical approach of median split for graphical representation. Another potential limitation is selective attrition due to mortality or illness. We addressed this issue by including attrition-informative variables (age, sex, education, and disability), but we note that our results may not generalize to less positively selected, more vulnerable populations.

We also acknowledge measurement limitations. To begin with, we used a single-item indicator focusing on cognitive-evaluative well-being, and it would be informative to explore whether our results generalize to measures of positive/negative affect and depressive symptoms (Asselmann & Specht, 2022; Luhmann et al., 2012), which were not available for the full study period. It is possible that historical improvements in the adaptation phase observed here for life satisfaction may not necessarily generalize to happiness. Additionally, because spousal bereavement likely affects satisfaction with family life the most, future research should explore historical changes in domain-specific life satisfaction trajectories. Future research should also consider dyad-level characteristics such as spousal closeness and marital satisfaction as protective or risk factors. Further, we treated individual and relationship variables as time-invariant to prioritize analytical parsimony. It will be useful for future studies to investigate how changes in health and social resources affect well-being post-bereavement.

Concerning our study sample, our ability to test if re-partnering moderated historical changes in well-being after spousal bereavement was limited because less than 10% of the sample had re-partnered. A follow-up analysis excluding re-partnered individuals yielded similar results, with two exceptions: (a) disability of the deceased spouse no longer significantly predicted anticipatory declines in life satisfaction and (b) later bereaved spouses showed significantly worse declines around the loss. Because re-partnering has become more common in recent decades, future research should explore its evolving role in adjusting to spousal loss. We further acknowledge that cross-national and cultural differences in policies and views on death may limit the generalizability of our findings. For instance, Nakagawa and Hülür (2021) found that Japanese older adults, unlike Western samples, did not experience a recovery phase, potentially because cultural mourning rituals mitigate immediate grief but may hinder long-term recovery (Klass, 2001; Lalande & Bonanno, 2006), which in turn raise interesting questions about whether such differences have narrowed due to globalization. Additionally, reports suggest a declining importance of romantic partners for personal happiness among adolescents (Gonzalez Avilés et al., 2024; Scheling & Richter, 2021), making it an intriguing question to explore whether these findings will apply to future generations of middle-aged and older adults when they experience spousal bereavement.

Conclusion

The current study is among the firsts to examine how phase-specific changes in well-being that surround spousal loss have changed historically over the last 40 years. When compared with their earlier-bereaved peers, people losing their spouses today show (1) a little less anticipatory decline before the death, an effect that is primarily driven by their lower well-being ratings at baseline; (2) no changes in the acute reaction phase; and (3) some modest improvements on long-term post-bereavement adaptations, particularly 2 or more years after death. This lack of historical change in the immediate reaction phase might indicate that the pain of loss is a timeless and universal phenomenon, at least in the present sample. We did not find evidence for historical changes in how individual and relationship-related factors operate as moderators of bereavement. These relations have remained unchanged over the past 40 years. Our study thus highlights the role of the sociohistorical context in shaping well-being trajectories during spousal bereavement.

Supplemental Material

Supplemental Material - How spousal bereavement shapes life satisfaction: Stability and change across historical time

Supplemental Material for How spousal bereavement shapes life satisfaction: Stability and change across historical time by Urmimala Ghose, Michael D. Krämer, David Richter, Gert G. Wagner, Frank Infurna, Nilam Ram, and Denis Gerstorf in European Journal of Personality

Footnotes

Author note

The Transparency and Openness subsection affirms that the materials are available. Details about the study design, procedures, and measures are documented at https://osf.io/7hce2/files/osfstorage. Analytic code for all the data preparations and analyses are included in the OSF repository . Parts of the ideas and data appearing in the manuscript have been presented at the 25. Fachgruppentagung Entwicklungspsychologie (September 2023) of the Deutsche Gesellschaft fur Psychologie (German Psychological Society), Berlin, Germany, and at the Annual Scientific Meeting (November 2023) of the Gerontological Society of America, Tampa, United States.

Acknowledgements

This manuscript is based on part of the doctoral dissertation of Urmimala Ghose at Humboldt-Universität zu Berlin, Germany. During the work on her dissertation, she was also a pre-doctoral fellow of the International Max Planck Research School on the Life Course (LIFE, ; participating institutions: Max Planck Institute for Human Development, Freie Universität Berlin, Humboldt-Universität zu Berlin, University of Michigan, University of Virginia, and University of Zurich).

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) received no financial support for the research, authorship, and/or publication of this article.

Open science statement

De-identified data are freely available to registered researchers at https://www.diw.de/soep. Our hypotheses and analytic strategies were pre-registered at https://osf.io/739th/?view_only=da8b3239758e417c909e578747449068 https://osf.io/7hce2/files/osfstorage (see also for deviations from the original pre-registration)³. The documentation of variables, the R-scripts (for data preparation and analyses), and the Mplus scripts (for analyses) are available at https://osf.io/mypzc/?view_only=ea2ae8627c3b4719ac09fcf6f5f22a9d, .

ORCID iDs

Urmimala Ghose

Michael D. Krämer

David Richter

Frank J. Infurna

Nilàm Ram

Denis Gerstorf

Supplemental Material

Supplemental material for this article is available online.

Notes

References

Asselmann

Specht

(2022). Dramatic effects but fast adaptation: Changes in life satisfaction and different facets of affective well-being around the death of a partner and death of a child. Applied psychology. Health and well-being, 15(2), 451–465. https://doi.org/10.1111/aphw.12388

Bennett

K. M.

Soulsby

L. K.

(2012). Wellbeing in bereavement and widowhood. Illness, Crisis, and Loss, 20(4), 321–337. https://doi.org/10.2190/IL.20.4.b

Boerner

Heckhausen

(2003). To have and have not: Adaptive bereavement by transforming mental ties to the deceased. Death Studies, 27(3), 199–226. https://doi.org/10.1080/07481180302888

Böger

Huxhold

(2020). The changing relationship between partnership status and loneliness: Effects related to aging and historical time. Journals of Gerontology Series B: Psychological Sciences and Social Sciences, 75(7), 1423–1432. https://doi.org/10.1093/geronb/gby153

Bonanno

G. A.

Kaltman

(1999). Toward an integrative perspective on bereavement. Psychological Bulletin, 125(6), 760–776. https://psycnet.apa.org/doi/10.1037/0033-2909.125.6.760

Bonanno

G. A.

Wortman

C. B.

Lehman

D. R.

Tweed

R. G.

Haring

Sonnega

Carr

Nesse

R. M.

(2002). Resilience to loss and chronic grief: A prospective study from preloss to 18-months postloss. Journal of Personality and Social Psychology, 83(5), 1150–1164. https://psycnet.apa.org/doi/10.1037/0022-3514.83.5.1150

Brickman

Campbell

D. T.

(1971). Hedonic relativism and planning the good society. In Appley

M. H.

(Ed.), Adaptation level theory: A symposium (pp. 287–302). New York: Academic Press.

Bryant

C. D.

Peck

D. L.

(Eds.), (2009). Encyclopedia of death and the human experience. (1). Sage.

Buecker

Luhmann

Haehner

Bühler

J. L.

Dapp

L. C.

Luciano

E. C.

Orth

(2023). The development of subjective well-being across the life span: A meta-analytic review of longitudinal studies. Psychological Bulletin, 149(7-8), 418–446. https://doi.org/10.1037/bul0000401

10.

Bühler

J. L.

Orth

Bleidorn

Weber

Kretzschmar

Scheling

Hopwood

C. J.

(2024). Life events and personality change: A systematic review and meta-analysis. European Journal of Personality, 38(3), 544–568. https://doi.org/10.1177/08902070231190219

11.

Carr

(2006). The future of late life spousal bereavement. In Carr

Nesse

R. M.

Wortman

C. B.

(Eds.), Spousal bereavement in late life (8th ed., pp. 313–323). Springer Publishing Company.

12.

Carr

Utz

(2001). Late-life widowhood in the United States: New directions in research and theory. Ageing International, 27(1), 65–88. https://doi.org/10.1007/s12126-001-1016-3

13.

Carr

Wortman

C. B.

Wolff

(2006). How older Americans die today: Implications for surviving spouses. In Carr

Nesse

R. M.

Wortman

C. B.

(Eds.), Spousal bereavement in late life (8th ed., pp. 49–78). Springer Publishing Company.

14.

Cheung

Lucas

R. E.

(2014). Assessing the validity of single-item life satisfaction measures: Results from three large samples. Quality of Life Research: An International Journal of Quality of Life Aspects of Treatment, Care and Rehabilitation, 23(10), 2809–2818. https://doi.org/10.1007/s11136-014-0726-4

15.

Clark

A. E.

Georgellis

(2013). Back to baseline in britain: Adaptation in the British household panel survey. Economica, 80(319), 496–512. https://doi.org/10.1111/ecca.12007

16.

Coontz

(2007). The origins of modern divorce. Family Process, 46(1), 7–16. https://doi.org/10.1111/j.1545-5300.2006.00188.x

17.

Crimmins

E. M.

Beltrán-Sánchez

(2011). Mortality and morbidity trends: Is there compression of morbidity? Journals of Gerontology Series B: Psychological Sciences and Social Sciences, 66(1), 75–86. https://doi.org/10.1093/geronb/gbq088

18.

Crimmins

E. M.

Zhang

Y. S.

Kim

J. K.

Levine

M. E.

(2019). Changing disease prevalence, incidence, and mortality among older cohorts: The health and retirement study. The Journals of Gerontology. Series A, Biological Sciences and Medical Sciences, 74(Suppl_1), S21–S26. https://doi.org/10.1093/gerona/glz075

19.

Diener

(1984). Subjective well-being. Psychological Bulletin, 95(3), 542–575. https://psycnet.apa.org/doi/10.1037/0033-2909.95.3.542

20.

Diener

Lucas

R. E.

Scollon

C. N.

(2006). Beyond the hedonic treadmill: Revising the adaptation theory of well-being. American Psychologist, 61(4), 305–314. https://doi.org/10.1037/0003-066X.61.4.305

21.

Doblhammer

Fritze

Reinke

Fink

(2022). Can dementia become the most prevalent disease at the time of death in Germany? Projections up to the year 2060 for the five most important diseases at the time of death. Journal of Population Ageing, 15(2), 523–540. https://doi.org/10.1007/s12062-022-09365-7

22.

Deutsche

Rentenversicherung

. (2024). Rentenversicherung in Zeitreihen . https://www.deutsche-rentenversicherung.de/SharedDocs/Downloads/DE/Statistiken-und-Berichte/statistikpublikationen/rv_in_zeitreihen.html

23.

Drewelies

Agrigoroaei

Lachman

M. E.

Gerstorf

(2018a). Age variations in cohort differences in the United States: Older adults report fewer constraints nowadays than those 18 years ago, but mastery beliefs are diminished among younger adults. Developmental Psychology, 54(8), 1408–1425. https://doi.org/10.1037/dev0000527

24.

Drewelies

Deeg

D. J. H.

Huisman

Gerstorf

(2018b). Perceived constraints in late midlife: Cohort differences in the longitudinal aging study Amsterdam (LASA). Psychology and Aging, 33(5), 754–768. https://doi.org/10.1037/pag0000276

25.

Drewelies

Huxhold

Gerstorf

(2019). The role of historical change for adult development and aging: Towards a theoretical framework about the how and the why. Psychology and Aging, 34(8), 1021–1039. https://doi.org/10.1037/pag0000423

26.

Dykstra

P. A.

de Jong Gierveld

(2004). Gender and marital-history differences in emotional and social loneliness among Dutch older adults. Canadian Journal on Aging/La revue canadienne du vieillissement, 23(2), 141–155. https://doi.org/10.1353/cja.2004.0018

27.

Elder

G. H.

(2003). The emergence and development of life course theory. Handbook of the life course. Springer.

28.

Elder

G. H.

Johnson

M. K.

Crosnoe

(2003). The emergence and development of life course theory. In Handbook of the life course (pp. 3–19). Springer. https://doi.org/10.1007/978-0-306-48247-2_1

29.

England

Levine

Mishel

(2020). Progress toward gender equality in the United States has slowed or stalled. Proceedings of the National Academy of Sciences of the United States of America, 117(13), 6990–6997. https://doi.org/10.1073/pnas.1918891117

30.

Family Caregiver Alliance . (2019). Caregiver statistics: Demographics. https://www.caregiver.org/caregiver-statistics-demographHorowitzics (Accessed 10 April 2020).

31.

Ferrell

Wittenberg

(2017). A review of family caregiving intervention trials in oncology. CA: A Cancer Journal for Clinicians, 67(4), 318–325. https://doi.org/10.3322/caac.21396

32.

Fries

J. F.

Bruce

Chakravarty

(2011). Compression of morbidity 1980–2011: A focused review of paradigms and progress. Journal of Aging Research, 2011, 261702. https://doi.org/10.4061/2011/261702

33.

Gerstorf

Drewelies

Duezel

Smith

Wahl

H. W.

Schilling

O. K.

Kunzmann

Siebert

J. S.

Katzorreck

Eibich

Demuth

Steinhagen-Thiessen

Wagner

G. G.

Lindenberger

Heckhausen

Ram

(2019). Cohort differences in adult-life trajectories of internal and external control beliefs: A tale of more and better maintained internal control and fewer external constraints. Psychology and Aging, 34(8), 1090-1108. https://doi.org/10.1037/pag0000389

34.

Gerstorf

Heckhausen

Ram

Infurna

F. J.

Schupp

Wagner

G. G.

(2014). Perceived personal control buffers terminal decline in well-being. Psychology and Aging, 29(3), 612-625. https://doi.org/10.1037/a0037227

35.

Gerstorf

Hülür

Drewelies

Eibich

Duezel

Demuth

Ghisletta

Steinhagen-Thiessen

Wagner

G. G.

Lindenberger

(2015). Secular changes in late-life cognition and well-being: Towards a long bright future with a short brisk ending? Psychology and Aging, 30(2), 301-310. https://doi.org/10.1037/pag0000016

36.

Gerstorf

Ram

Drewelies

Duezel

Eibich

Steinhagen-Thiessen

Liebig

Goebel

Demuth

Villringer

Wagner

G. G.

Lindenberger

Ghisletta

(2023). Today’s older adults are cognitively fitter than older adults were 20 years ago, but when and how they decline is no different than in the past. Psychological Science, 34(1), 22–34. https://doi.org/10.1177/09567976221118541

37.

Gerstorf

Ram

Estabrook

Schupp

Wagner

G. G.

Lindenberger

(2008). Life satisfaction shows terminal decline in old age: Longitudinal evidence from the German socio-economic panel study (SOEP). Developmental Psychology, 44(4), 1148-1159. https://doi.org/10.1037/0012-1649.44.4.1148

38.

Glaser

B. G.

Strauss

A. L.

(1968). Time for dying. AldineTransaction.

39.

Goebel

Grabka

M. M.

Liebig

Kroh

Richter

Schröder

Schupp

(2019). The German socio-economic panel (SOEP). Jahrbucher für Nationalokonomie und Statistik, 239(2), 345–360. https://doi.org/10.1515/jbnst-2018-0022

40.

Gonzalez Avilés

Bühler

J. L.

Brandt

N. D.

Neyer

F. J.

(2024). Today’s adolescents are more satisfied with being single: Findings from a German cohort-sequential study among 14- to 40-year-olds. Personality and Social Psychology Bulletin. https://doi.org/10.1177/01461672241257139

41.

Grimm

K. J.

Ram

Estabrook

(2016). Growth modeling: Structural equation and multilevel modeling approaches. Guilford.

42.

Grünheid

(2013). Ehescheidungen in Deutschland: Entwicklungen und Hintergrunde [Divorce in Germany: Trend and Background Factors]. Bundesinstitut für Bevolkerungsforschung. BiB Working Paper 1/2003. https://nbn-resolving.org/urn:nbn:de:bib-wp-2013-011

43.

Guiaux

van Tilburg

T. G.

Van Groenou

M. B.

(2007). Changes in contact and support exchange in personal networks after widowhood. Personal Relationships, 14(3), 457–473. https://doi.org/10.1111/j.1475-6811.2007.00165.x

44.

J. H.

(2010). The effects of positive and negative support from children on widowed older adults’ psychological adjustment: A longitudinal analysis. The Gerontologist, 50(4), 471–481. https://doi.org/10.1093/geront/gnp163

45.

Hajek

König

H.-H.

(2018). The relation between personality, informal caregiving, life satisfaction and health-related quality of life: Evidence of a longitudinal study. Quality of Life Research: An International Journal of Quality of Life Aspects of Treatment, Care and Rehabilitation, 27(5), 1249–1256. https://doi.org/10.1007/s11136-018-1787-6

46.

Halleröd

(2013). Gender inequality from beyond the grave: Intra-household distribution and wellbeing after spousal loss. Ageing and Society, 33(5), 783–803. https://psycnet.apa.org/doi/10.1017/S0144686X12000268

47.

Hansen

S. R.

(2006). Courtship duration as a correlate of marital satisfaction and stability. Alliant International University.

48.

Headey

Muffels

Wagner

G. G.

(2010). Long-running German panel survey shows that personal and economic choices, not just genes, matter for happiness. Proceedings of the National Academy of Sciences of the United States of America, 107(42), 17922–17926. https://doi.org/10.1073/pnas.1008612107

49.

Holmes

T. H.

Rahe

R. H.

(1967). The social readjustment rating scale. Journal of Psychosomatic Research, 11(2), 213–218. https://doi.org/10.1016/0022-3999(67)90010-4

50.

Hülür

Ram

Gerstorf

(2015). Historical improvements in well-being do not hold in late life: Birth-and death-year cohorts in the United States and Germany. Developmental Psychology, 51(7), 998. https://doi.org/10.1037/a0039349

51.

Infurna

F. J.

Dey

N. E.

Gonzalez Avilés

Grimm

K. J.

Lachman

M. E.

Gerstorf

(2024). Loneliness in midlife: Historical increases and elevated levels in the United States compared with Europe. American Psychologist. https://psycnet.apa.org/doi/10.1037/amp0001322.

52.

Infurna

F. J.

Luthar

S. S.

(2017). The multidimensional nature of resilience to spousal loss. Journal of Personality and Social Psychology, 112(6), 926–947. https://psycnet.apa.org/doi/10.1037/pspp0000095

53.

Infurna

F. J.

Staben

O. E.

Lachman

M. E.

Gerstorf

(2021). Historical change in midlife health, well-being, and despair: Cross-cultural and socioeconomic comparisons. American Psychologist, 76(6), 870–887. https://doi.org/10.1037/amp0000817

54.

Infurna

F. J.

Wiest

Gerstorf

Ram

Schupp

Wagner

G. G.

Heckhausen

(2017). Changes in life satisfaction when losing one's spouse: Individual differences in anticipation, reaction, adaptation and longevity in the German Socio-economic Panel Study (SOEP). Ageing and Society, 37(5), 899–934. https://doi.org/10.1017/S0144686X15001543

55.

Kastenbaum

Costa Jr

P. T.

(1977). Psychological perspectives on death. Annual Review of Psychology, 28(1), 225–249. https://doi.org/10.1146/annurev.ps.28.020177.001301

56.

Kettlewell

Morris

R. W.

Cobb-Clark

D. A.

Cripps

Glozier

(2020). The differential impact of major life events on cognitive and affective wellbeing. SSM - Population Health, 10, 100533. https://doi.org/10.1016/j.ssmph.2019.100533

57.

Klass

(2001). Continuing bonds in the resolution of grief in Japan and North America. American Behavioral Scientist, 44(5), 742–763. https://doi.org/10.1177/00027640121956476

58.

Krämer

M. D.

Bleidorn

(2024). The Well-Being Costs of Informal Caregiving. Psychological Science, 35(12), 1382-1394. https://doi.org/10.1177/09567976241279203 (Original work published 2024).

59.

Krämer

M. D.

Rohrer

J. M.

Lucas

R. E.

Richter

(2025). Life events and life satisfaction: Estimating effects of multiple life events in combined models. European Journal of Personality, 39(1), 2–23. https://doi.org/10.1177/08902070241231017

60.

Kroh

Pischner

Spieß

Wagner

G. G.

(2008). On the treatment of non-original sample members in the German Household Panel Study (SOEP). Methoden, Daten, Analysen (mda), 2(2), 179–198. https://nbn-resolving.org/urn:nbn:de:0168-ssoar-126618

61.

Kunzmann

Richter

Schmukle

S. C.

(2013). Stability and change in affective experience across the adult life span: Analyses with a national sample from Germany. Emotion, 13(6), 1086–1095. https://doi.org/10.1037/a0033572

62.

Lalande

K. M.

Bonanno

G. A.

(2006). Culture and continuing bonds: A prospective comparison of bereavement in the United States and the people's Republic of China. Death Studies, 30(4), 303–324. https://doi.org/10.1080/07481180500544708

63.

Lee

G. R.

DeMaris

(2007). Widowhood, gender, and depression: A longitudinal analysis. Research on Aging, 29(1), 56–72. https://doi.org/10.1177/0164027506294098

64.

Little

R. J.

Rubin

D. B.

(1989). The analysis of social science data with missing values. Sociological Methods & Research, 18(2-3), 292–326. https://doi.org/10.1177/0049124189018002004

65.

Liu

(2012). Marital dissolution and self-rated health: Age trajectories and birth cohort variations. Social Science & Medicine, 74(7), 1107–1116. https://doi.org/10.1016/j.socscimed.2011.11.037

66.

Lucas

R. E.

(2007a). Adaptation and the set-point model of subjective well-being: Does happiness change after major life events? Current Directions in Psychological Science, 16(2), 75–79. https://doi.org/10.1111/j.1467-8721.2007.00479.x

67.

Lucas

R. E.

(2007b). Long-term disability is associated with lasting changes in subjective well-being: Evidence from two nationally representative longitudinal studies. Journal of Personality and Social Psychology, 92(4), 717–730. https://psycnet.apa.org/doi/10.1037/0022-3514.92.4.717

68.

Lucas

R. E.

Clark

A. E.

Georgellis

Diener

(2003). Reexamining adaptation and the set point model of happiness: reactions to changes in marital status. Journal of personality and social psychology, 84(3), 527–539. https://doi.org/10.1037//0022-3514.84.3.527

69.

Lucas

R. E.

Donnellan

M. B.

(2012). Estimating the reliability of single-item life satisfaction measures: Results from four national panel studies. Social Indicators Research, 105(3), 323–331. https://doi.org/10.1007/s11205-011-9783-z

70.

Luhmann

Hofmann

Eid

Lucas

R. E.

(2012). Subjective well-being and adaptation to life events: A meta-analysis. Journal of Personality and Social Psychology, 102(3), 592–615. https://psycnet.apa.org/doi/10.1037/a0025948

71.

Martin-Matthews

(2011). Revisiting widowhood in later life: Changes in patterns and profiles, advances in research and understanding. Canadian Journal on Aging, 30(3), 339–354. https://doi.org/10.1017/S0714980811000201

72.

Michalos

A. C.

(1985). Multiple discrepancies theory (MDT). Social Indicators Research Series, 16, 347–413. https://doi.org/10.1007/1-4020-3742-2_13

73.

Muthén

L. K.

Muthén

B. O.

(1998-2017). Mplus user’s guide. 8th ed. Muthén & Muthén.

74.

Nakagawa

Hülür

(2021). Life satisfaction during the transition to widowhood among Japanese older adults. Gerontology, 67(3), 338–349. https://doi.org/10.1159/000512859

75.

Nielsen

M. K.

Carlsen

A. H.

Neergaard

M. A.

Bidstrup

P. E.

Guldin

M. B.

(2019). Looking beyond the mean in grief trajectories: A prospective, population-based cohort study. Social Science & Medicine, 232, 460–469. https://doi.org/10.1016/j.socscimed.2018.10.007

76.

Pardo

Weisfeld

Hill

Slatcher

R. B.

(2013). Machismo and marital satisfaction in Mexican American couples. Journal of Cross-Cultural Psychology, 44(2), 299–315. https://doi.org/10.1177/0022022112443854

77.

Perrig-Chiello

Spahni

Höpflinger

Carr

(2016). Cohort and sex differences in psychosocial adjustment to later-life widowhood. Journals of Gerontology Series B: Psychological Sciences and Social Sciences, 71(4), 765–774. https://doi.org/10.1093/geronb/gbv004

78.

Ram

Grimm

(2007). Using simple and complex growth models to articulate developmental change: Matching theory to method. International Journal of Behavioral Development, 31(4), 303–316. https://doi.org/10.1177/0165025407077751

79.

Sasson

Umberson

D. J.

(2014). Widowhood and depression: New light on gender differences, selection, and psychological adjustment. Journals of Gerontology Series B: Psychological Sciences and Social Sciences, 69(1), 135–145. https://doi.org/10.1093/geronb/gbt058

80.

Scheling

Richter

(2021). Generation Y: Do millennials need a partner to be happy? Journal of Adolescence, 90, 23–31. https://doi.org/10.1016/j.adolescence.2021.05.006

81.

Schilling

O. K.

(2005). Cohort-and age-related decline in elder’s life satisfaction: Is there really a paradox? European Journal of Ageing, 2(4), 254–263. https://doi.org/10.1007/s10433-005-0016-7

82.

Schimmack

Oishi

(2005). The influence of chronically and temporarily accessible information on life satisfaction judgments. Journal of Personality and Social Psychology, 89(3), 395–406. https://doi.org/10.1037/0022-3514.89.3.395

83.

Schulz

Mendelsohn

A. B.

Haley

W. E.

Mahoney

Allen

R. S.

Zhang

Thompson

Belle

S. H.

Resources for Enhancing Alzheimer's Caregiver Health Investigators . (2003). End-of-life care and the effects of bereavement on family caregivers of persons with dementia. New England Journal of Medicine, 349(20), 1936–1942. https://doi.org/10.1056/NEJMsa035373

84.

Singer

J. D.

Willett

J. B.

(2003). Applied longitudinal data analysis: Modeling change and event occurrence. Oxford University Press.

85.

Statistisches Bundesamt . (2023). Destatis. https://www.destatis.de/DE/Themen/Arbeit/Arbeitsmarkt/Erwerbstaetigkeit/Tabellen/erwerbstaetigenquoten-gebietsstand-geschlecht-altergruppe-mikrozensus.html

86.

Stroebe

Schut

Stroebe

(2007). Health outcomes of bereavement. Lancet (London, England), 370(9603), 1960–1973. https://doi.org/10.1016/S0140-6736(07)61816-9

87.

Suanet

Huxhold

(2020). Cohort difference in age-related trajectories in network size in old age: Are networks expanding? Journals of Gerontology Series B: Psychological Sciences and Social Sciences, 75(1), 137–147. https://doi.org/10.1093/geronb/gbx166

88.

Suanet

van Tilburg

T. G.

(2019). Loneliness declines across birth cohorts: The impact of mastery and self-efficacy. Psychology and Aging, 34(8), 1134–1143. https://psycnet.apa.org/doi/10.1037/pag0000357

89.

Teachman

J. D.

Tedrow

L. M.

Crowder

K. D.

(2000). The changing demography of America's families. Journal of Marriage and Family, 62(4), 1234–1246. https://doi.org/10.1111/j.1741-3737.2000.01234.x

90.

Uglanova

E. A.

Staudinger

U. M.

(2013). Zooming in on life events: Is hedonic adaptation sensitive to the temporal distance from the event? Social Indicators Research, 111(1), 265–286. https://doi.org/10.1007/s11205-012-0004-1

91.

United Nations . (nd). United Nations - world population prospects. https://population.un.org/wpp/

92.

van Tilburg

T. G.

Suanet

(2019). Unmarried older people: Are they socially better off today? Journals of Gerontology Series B: Psychological Sciences and Social Sciences, 74(8), 1463–1473. https://doi.org/10.1093/geronb/gby120

93.

Wilke

C. B.

(2009). A quarter century of reforms: An evaluation of the German Pension Reform process. In German pension reform: On road towards a sustainable multi-pillar system (NED-New edition, pp. 39–87). Peter Lang AG. https://www.jstor.org/stable/j.ctv9hj6rz.5

94.

Wolff

J. L.

Mulcahy

Huang

Roth

D. L.

Covinsky

Kasper

J. D.

(2018). Family caregivers of older adults, 1999–2015: Trends in characteristics, circumstances, and role-related appraisal. The Gerontologist, 58(6), 1021–1032. https://doi.org/10.1093/geront/gnx093

95.

Wünsche

Weidmann

Grob

(2020). Until death do us part: The codevelopment of life satisfaction in couples preceding the death of one partner. Journal of Personality and Social Psychology, 119(4), 881–900. https://psycnet.apa.org/doi/10.1037/pspi0000228

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.

0.00 MB

0.80 MB