Silver Splits and Parent–Child Disconnectedness: Mental Health Consequences for European Older Adults

Jessee, Lisa; Carr, Deborah Article — Published Version Silver Splits and Parent–Child Disconnectedness: Mental Health Consequences for European Older Adults European Journal of Population Provided in Cooperation with: Springer Nature Suggested Citation: Jessee, Lisa; Carr, Deborah (2025) : Silver Splits and Parent–Child Disconnectedness: Mental Health Consequences for European Older Adults, European Journal of Population, ISSN 1572-9885, Springer Netherlands, Dordrecht, Vol. 41, Iss. 1, https://doi.org/10.1007/s10680-025-09751-9 This Version is available at: https://hdl.handle.net/10419/330415 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by/4.0/ Vol.:(0123456789) European Journal of Population (2025) 41:27 https://doi.org/10.1007/s10680-025-09751-9 BRIEF REPORT Silver Splits andParent–Child Disconnectedness: Mental Health Consequences forEuropean Older Adults LisaJessee1 · DeborahCarr2 Received: 28 April 2025 / Accepted: 10 September 2025 © The Author(s) 2025 Abstract Rising rates of “silver splits” in Europe resemble increases in gray divorce in the U.S. Partnership dissolutions may harm older adults’ mental health, especially for ‘disconnected’ parents who do not receive support from their children. However, researchers have relied primarily on multilevel modeling, neglecting unobserved characteristics that may select an individual into both divorce and parent–child disconnectedness. This brief report addresses this research gap by estimating fixedeffects linear regression models that control for time-invariant confounders. We used data from the Survey of Health, Ageing and Retirement in Europe (SHARE; 2004– 2022, N = 2216 observations, 546 silver splits) to document changes in depressive symptoms preand post-dissolution and evaluate whether these patterns are moderated by parent–child disconnectedness. Consistent with previous research, we find that depressive symptoms increase steeply in the year of dissolution and remain high four years post-dissolution for parents who are disconnected from their adult child(ren). However, individuals who maintain a relationship with all their child(ren) show stable levels of depressive symptoms throughout the dissolution process, challenging the assumption that dissolution is uniformly distressing. Our results reveal that depressive symptoms trajectories during the period preceding and following a major life event differ across sociorelational contexts. Social programs and supports for divorced older adults should recognize this heterogeneity rather than assuming uniformly negative mental health outcomes. Keywords Gray divorce· Depressive symptoms· Parent–child relationships· Intergenerational solidarity· Union dissolution· Well-being * Lisa Jessee [email protected]oeln.de Deborah Carr car[email protected] 1 Department ofSociology andSocial Psychology, University ofCologne, Albertus-Magnus-Platz, 50923Cologne, Germany 2 Department ofSociology andCenter forInnovation inSocial Science, Boston University, 704 Commonwealth Ave., Boston, MA02215, USA L.Jessee, D.Carr 27 Page 2 of 20 1 Introduction Rates of ‘silver splits’—the dissolution of marriages or romantic partnerships at or after age 50—have risen in Europe, mirroring ‘gray divorce’ trends in the U.S. (Alderotti et al., 2022; Brown & Lin, 2022; Solaz, 2021; Vignoli et al., 2025; Žilinčíková & Schnor, 2021). The number of relationship dissolutions in later life is expected to rise further in the coming years, given population aging (Brown & Lin, 2012). The loss of a co-residential intimate partner through relationship dissolution may deprive adults of a crucial source of socioemotional and instrumental support (Brown & Lin, 2022). Union dissolutions may be especially consequential for older adults because age-related transitions including retirement, onset of health problems, and the deaths of peers may diminish their levels of contact with other potentially supportive ties (Charles & Carstensen, 2002). Classic writings on stressful life events emphasize that transitions like divorce may trigger mental health symptoms including depression, due to both the acute stress of the event and the sequelae of secondary stressors that follow, such as a drop in household income or the loss of a helpmate and confidante (e.g., Norris & Murrell, 1987). The divorce-stress-adjustment model counters that union dissolution is a process that can affect mental health even before the event due to precipitating stressors like persistent relationship conflict and the anticipation of future disruptions to daily life and everyday routines (Amato, 2000). Empirical assessments of the divorce-stress-adjustment framework draw on three distinctive conceptual models which offer competing perspectives regarding the time course of mental health symptoms following later-life dissolution (Lin & Brown, 2020). The crisis model proposes that union dissolutions lead to short-term mental health declines with a relatively quick recovery as individuals manage temporary stressors like legal issues or residential changes, whereas the chronic strain model predicts persistent mental health struggles from ongoing stressors, such as loneliness and the challenges of managing household chores and finances independently (Lin etal., 2019).The convalescence model, specific to older adults, similarly predicts long-lasting mental health decrements yet an eventual improvement due to older adults’ resilience and capacity to adapt in the longer term (Lin & Brown, 2020; Lin etal., 2024). Yet, the extent to which mental health trajectories during the silver split process align with the crisis, chronic strain, or convalescence models may be contingent on the availability of coping resources—most notably protective social relationships (e.g., Pearlin etal., 2005). For older adults experiencing union dissolution, a strong relationship with adult children is considered an important buffer against emotional distress (Tosi & van den Broek, 2020). Conversely, strained, conflictual or tenuous parent–child relationships may intensify and prolong the emotional consequences of dissolution. Recent research identifies one particular dimension of weak parent–child ties that heightens the mental health effects of dissolution: parent–child disconnectedness (Jessee & Carr, 2025; Kalmijn, 2023; Lin etal., 2024). Disconnectedness is similar to estrangement, which encompasses deficient emotional closeness and frequency of contact, and refers to the parent’s lack of Silver Splits andParent–Child Disconnectedness: Mental Health… Page 3 of 20 27 contact with at least one adult child (Agllias, 2018; Arránz Becker & Hank, 2022; Conti, 2015; Reczek etal., 2023; Scharp & Hall, 2017). An emphasis on disconnectedness from at least one child is consistent with research documenting that parental well-being is affected even when difficulties or strains exist with only a single child (Fingerman etal., 2012; Reczek etal., 2025). In Europe, disconnectedness affects between 1 and 17% of the population, with divorced men (13%) experiencing it roughly twice as often as divorced women (6%; Jessee & Carr, 2025). Disconnectedness may exacerbate and lengthen older adults’ post-dissolution mental health symptoms because disconnected parents may be deprived of practical or emotional support from adult child(ren), a resource critical to their adaptation (Lin etal., 2024). Moreover, disconnectedness may be a co-occurring stressor, creating a double burden that may further erode the mental health of older adults experiencing a silver split. Disconnectedness is more consequential than strong positive parent–child relations for older adults’ mental health because weak ties are atypical, stigmatized, and emotionally painful (Gilligan etal., 2015; Lin etal., 2024). Older parents who are disconnected from their adult children may experience a more significant and prolonged increase in mental health symptoms during a silver split compared to connected parents. This scenario is consistent with the chronic strain perspective, as opposed to the crisis or convalescence perspectives suggested by previous research on the mental health effects of gray divorces (Lin etal., 2019; Tosi & van den Broek, 2020). Thus, our aim is to evaluate the extent to which depressive symptoms trajectories preand post-divorce differ on the basis of parent–child disconnectedness. We focus specifically on depressive symptoms because this is one of the most common mental health problems among older adults. Untreated depressive symptoms also undermine physical health and are projected to become a leading cause of disability worldwide (Andreas etal., 2017; Mathers & Loncar, 2006). To our knowledge, Lin etal. (2024) are the only researchers who have examined parent–child disconnectedness as a contextual factor affecting depressive symptom trajectories before, during, and after gray divorce. They tracked a sample of US older adults and found that parents reported heightened depressive symptoms post-dissolution, yet the magnitude of this change was significantly larger for those who were disconnected from (vs. connected to) their adult child(ren). The difference between disconnected and connected parents began to diminish two years after the dissolution and converged six years thereafter. However, the authors estimated multilevel models and did not consistently account for unobserved factors that may “select” individuals into later-life separation, depressive symptoms, and disconnectedness. Moreover, it is unclear whether similar patterns would occur beyond the U.S., given theoretical writings underscoring that the emotional consequences of a purported stressor may be conditional upon the broader sociocultural context (Pearlin etal., 2005). Europe has lower rates of both divorce and parent–child disconnectedness relative to the U.S. (Jessee & Carr, 2025; National Center for Health Statistics, 2023; Reczek etal., 2023), so older adults experiencing the “double burden” of these two stressors may be susceptible to stigmatization, lack of structural supports, and other factors that may render them particularly vulnerable to sustained depressive symptoms. L.Jessee, D.Carr 27 Page 4 of 20 Against this background, this brief report examines depressive symptoms preand post-gray dissolution for two groups of “silver splitters” in Europe: parents who are connected versus disconnected from their adult children. We draw on longitudinal data from eight waves (2004–2022) of the Survey of Health, Ageing and Retirement in Europe (SHARE) and apply fixed-effects regression models. A fixed-effects approach allows us to capture both observed and unobserved individual confounding factors by subtracting the individual mean from the observed values of all variables. 2 Methods 2.1 Data Data are from SHARE, a longitudinal biennial survey of adults aged 50+ in Europe and Israel (Börsch-Supan et al., 2013), spanning eight waves (1 to 9, excluding wave 3) from 2004 to 2022 across 26 nations. Unlike single-nation surveys, SHARE allows us to follow within-person trajectories among parents who recently experienced a silver split, stratified by parent–child disconnectedness. We excluded Israel (the only non-European country in SHARE) and Ireland (with only one wave of data), as well as wave 3 data, which only included retrospective life course information. Our baseline sample comprises 146,868 individuals, with baseline defined as the first wave in which a respondent is observed in the panel. Our main goal is to trace within-person changes in older parents’ depressive symptoms preand post-relationship dissolution. Thus, we limit our sample to parents who are at risk of and subsequently experience a silver split, defined as a marital or non-marital romantic partnership dissolution after age 50 (Alderotti etal., 2022) We exclude from our analytic sample individuals who are: under 50 (n = 3740); not in a partnership at baseline (n = 44,008); continuously partnered throughout the observation period (n = 58,516); respondents with children under the age of 18 or childless (n = 6624); with only one wave of data (n = 33,433); or who experienced two or more silver splits during the study (n = 1). Respondents with children under the age of 18 were excluded as they were not at risk of experiencing disconnectedness from an adult child. Our sample includes 546 parents who experienced a silver split, with an average of four observations per participant, totaling 2216 observations. 2.2 Measures Depressive symptoms are assessed at each available time point before, during and after separation with the 12-item EURO-D scale (Prince etal., 1999). Respondents indicate whether they have experienced each symptom in the past 12months: depression, pessimism, suicidality, guilt, sleep, interest, irritability, appetite, fatigue, concentration, lack of enjoyment, and tearfulness. Symptom counts range from 0 to 12. We used a natural log transformation to address the skewed distribution of Silver Splits andParent–Child Disconnectedness: Mental Health… Page 5 of 20 27 symptoms (M = 2.6, SD = 2.3). Results were consistent across models, so we retain the count for ease of interpretation (results available from authors). Our focal independent variable is the time to and since a silver split. Following previous research (Alderotti etal., 2022; Vignoli etal., 2025), we did not stratify the analyses by dissolution subtype, but instead pooled marital and non-marital transitions into a single category of silver splits. This decision was made for both conceptual and empirical reasons. Our independent variable encompasses persons who either: transitioned from married to divorced (n = 472); ended a registered partnership (n = 20); or dissolved a cohabiting relationship (n = 53) during the observation period. In many European countries, registered partnerships are comparable to marriages, offering some of the same legal rights and benefits (for an overview of how registered partnerships align with or differ from marriage rights across countries, see Scherpe & Hayward, 2017). To have adequately powered moderation analyses, we combined the latter two groups into a single non-marital dissolution category (n = 73). Descriptive statistics (person-wave observations) for the two dissolution categories are presented in Table1. We did not detect statistically significant differences in depressive symptoms (2.70 vs. 2.40) or rates of parent–child disconnectedness (8 vs. 12%) for marital versus non-marital dissolutions. Multivariable analyses showed no significant differences in symptom trajectories between marital and non-marital unions, and restricting our analytic sample to marital dissolutions only resulted in similar patterns as for the full sample of silver splitters (see Appendix Fig.3). Based on these considerations, we did not differentiate between dissolution subtypes in the analyses and instead proceeded with a pooled analysis of all dissolution subtypes. SHARE data lack precise information on divorce/separation dates. Thus, we documented depressive symptoms throughout the dissolution process using five timing indicators based on the current and last interview year. While SHARE’s biennial design typically results in interviews every two years, some respondents are observed one or three years apart due to irregular interview lags. To account for this variability, we grouped years −1 to −3 into a single pre-split category and years +1 to +3 into a single post-split category. We created the following categories: (1) at least four years pre-split (baseline and reference group), (2) between one to three years pre-split, (3) the year in which the split was first recorded, (4) one to three years post-split, and (5) more than four years post-split. In the first (4+ years pre-split) and last (4+ years post-split) categories, we pooled all observations that occurred 4years or more before or after the split. Consequently, respondents could contribute multiple observations to these categories if they had multiple data points in these time frames. To assess the robustness of this coding strategy, we conducted supplemental analyses in which we truncated the sample to exclude all person-years observed more than four years after the silver split, focusing only on individuals observed exactly four years post-split. The results remained substantively unchanged (results available upon request). Given the increased statistical power afforded by the pooled approach, we retained this specification in our main analyses. Due to panel attrition and irregular participation across waves, not all individuals contribute observations at every time point before and after the split. For example, a respondent may participate at wave t, drop out in t + 1, and return in t + 2, resulting L.Jessee, D.Carr 27 Page 6 of 20 in missing data for intermediate periods. As a result, individuals contribute varying numbers of observations across the silver split stages. Table2 displays the distribution of timepoints. Table 1 Descriptive statistics by relationship dissolution status Means (and standard deviations) shown for continuous measures and proportions shown for categorical measures. Values based on person-wave data. Statistically significant differences denoted as *p < .05, **p < .01, ***p < .001, ns not significant. SHARE, waves 1, 2, 4, 5, 6, 7, 8, 9, release 9.0.0. unweighted data Marital dissolution Non-marital dissolution Difference Depressive symptoms (range: 0–12) 2.70 2.40 ns (2.42) (2.12) Parent–child disconnectedness 0.08 0.12 ns Age at baseline 65.13 65.48 ns (6.50) (7.23) Self-rated health Excellent 0.11 0.08 ns Very good 0.22 0.16 ns Good 0.37 0.29 ns Fair 0.21 0.37 *** Poor 0.09 0.10 ns Perceived financial difficulties 0.37 0.41 ns Employed 0.32 0.37 ns Female 0.53 0.51 ns European region Northern 0.27 0.27 ns Eastern 0.15 0.31 *** Southern 0.13 0.05 ** Western 0.45 0.37 ns Any values imputed 0.06 0.13 ** Person-wave observations 1,020 153 Table 2 Number of observations across silver split stages SHARE, waves 1, 2, 4, 5, 6, 7, 8, 9, release 9.0.0. unweighted data N% Years before/after the silver split −4 or more 714 32.2 −1 to 3 302 13.6 0 (year of the silver split) 546 24.6 +1 to 3 204 9.2 +4 or more 450 20.3 Person-wave observations 2216 100 Silver Splits andParent–Child Disconnectedness: Mental Health… Page 7 of 20 27 Following Lin etal. (2024), parent–child disconnectedness in the year of the split refers to parents’ lack of contact with at least one adult child in the 12months presplit (in person, by phone, or mail). Parent–child disconnectedness was included as a time-invariant moderator, measured in the wave in which the silver split was first reported. We chose a time-invariant moderator, since recent simulation studies indicate that interacting two time-varying variables in a fixed-effects model can produce spurious results (Giesselmann & Schmidt-Catran, 2022). Focusing on disconnectedness from at least one adult child (as opposed to all children) aligns with research showing that having any estranged adult children can negatively affect parental health (Reczek etal., 2025). The reference category includes parents in contact with all of their children. In our analytic sample, 10 percent of persons (49 individuals) were disconnected from at least one adult child and are thus classified as disconnected. While the number of disconnected parents who experience separation in our sample is relatively small, the study provides an initial investigation of this underresearched subgroup, yielding insights that can advance and encourage further research on the heterogeneity of later-life dissolution. Of the group of disconnected parents, 29 percent were disconnected from all of their children. Depressive symptoms did not differ significantly for the two subgroups of disconnected parents, so we did not stratify analyses by number of children from whom the parent is disconnected. In the year of the silver split, 61 percent of disconnected parents were separated from a biological child, and 39 percent from a stepchild. However, we did not have sufficient statistical power to further stratify our multivariable analyses based on whether the tie was biological. SHARE does not provide information on whether a disconnected child is from the dissolved marriage or partnership or from a prior relationship. (Un-)observed time-variant confounders, such as number of children, education, genetic factors, personality traits, and relationship duration, are accounted for in our fixed-effects modeling approach. Hence, we adjusted for time-varying covariates only. In line with divorce research that applies fixed-effects linear regression models (Kapelle & Monden, 2024; Leopold, 2018), we included a parsimonious set of categorical covariates to capture age and period effects, to account for the possibility that well-being systematically changes with age or varies across historical periods. Age was included as a four-level categorical variable—50–59 (reference), 60–69, 70–79, and 80 and older—and interview year likewise as 2004–2007 (reference), 2010–2013, 2015–2017, and 2019–2022, based on the years in which the data were collected. Including categorical controls avoids the problem of perfect collinearity. Sensitivity analyses using alternative interval lengths and quadric age terms produced very similar results (all models available from authors). For the 14 percent of respondents who had a missing value on at least one variable used in the analysis, we imputed missing values using chained equations (See Appendix Table 5 for item-specific missingness). We imputed 10 datasets, performed all analyses on each imputed dataset and combined coefficients. Multivariable analyses included a variable signifying whether any values were imputed. To avoid “over‐adjusting” for variables on the causal pathway, we excluded additional time‐varying covariates, including health status, employment status, and financial vulnerability, because these factors may mediate the effect of a L.Jessee, D.Carr 27 Page 8 of 20 silver split on depressive symptoms (Amato, 2000). For example, among women, a silver split may influence depressive symptoms through reduced financial security, whereas among men these effects may operate via diminishing social networks. If we adjust for these mediators as simple covariates, we risk understating the true impact of separation on depressive symptoms. A dedicated mediation analysis would be preferable, but our sample lacks sufficient statistical power (Kohler etal., 2024). Incorporating several time-varying covariates in our model, including employment status (working vs. not working), financial vulnerabilities (financial difficulties vs. no financial difficulties), self-rated health (range: 1 = excellent to 5 = poor), and repartnering (repartnered vs. still separated), yields similar effects, but with slightly reduced magnitude as we anticipated. All covariates were measured at each available SHARE interview wave across the different stages of the silver split. 2.3 Analytic Plan We apply fixed-effects linear regression models with robust standard errors to account for unobserved characteristics that may confound the associations among silver splits, disconnectedness, and depressive symptoms. Omitting important (unobserved) variables can lead to an overor underestimation of the true effects (omitted variable bias). Fixed-effects panel regression models allow researchers to address omitted variable bias for unobserved stable individual characteristics, such as personality traits, childhood experiences, or genetic factors, by subtracting the individual mean from the observed values of all variables (Vaisey & Miles, 2017). Moreover, fixed-effects panel regression models automatically account for observed time-invariant characteristics, such as gender, education, and number of children. We find sufficient withinindividual variation to justify a fixed-effects approach (see Appendix Table6). Fixed-effects regression models use each individual as their own control over time to focus on changes that occur within individuals. We modeled a change in depressive symptoms as a function of time before, during, or after the split among (a) connected and (b) disconnected parents. To test for statistically significant differences between connected and disconnected parents, we estimated fully interacted models (see Appendix Table7 for complete results). The basic equation without interactions is specified as follows: where EUROD represents the EURO-D score, our outcome, for individual i at time t. SILVERSPLIT, our focal predictor, categorizes periods relative to the split event. The other time-varying covariates (age, interview year, and whether values were imputed) are accounted for by variables x2, …, xk for individual i at time t. The term μi captures individual-specific characteristics that remain constant over time (fixed effects) for individual i. Finally, Ɛ denotes the error term that varies across both time and individuals (Replication files available here https:// osf. io/ akc7z/? view_ only= 811f1 15e9d 7a416 890b7 0e660 239af 71). EURODit =𝛽 1SILVERSPLITit +𝛽 2x2it +⋯+𝛽 kxkit +𝜇 i +𝜀 it Silver Splits andParent–Child Disconnectedness: Mental Health… Page 15 of 20 27 Second, just 49 people in our sample reported parent–child disconnectedness, necessitating a cautious interpretation of our study results. We did not have sufficient statistical power to consider other sources of heterogeneity within the disconnected subgroup, such as whether a parent was disconnected from only one or multiple children, and other key characteristics of the disconnected child, including gender, whether the child was from the current or a previous union, and whether the child was biological or non-biological. Initial descriptive findings suggest that disconnectedness from a biological child may be experienced more negatively, as parents in this group reported higher average depressive symptom scores compared to those disconnected from a non-biological child (3.4 vs. 2.1 symptoms). The small sample size also precluded robust tests of heterogeneity by gender or region. Preliminary analyses suggested no significant gender differences, though disconnected women’s higher baseline symptoms may translate into greater relative risk of elevated depression. Regional patterns also pointed to larger increases in Southern Europe, but these were not statistically significant (results available from authors). Larger samples are needed to assess these differences more conclusively. Finally, our measure of disconnectedness referred to contact during the year of the split only. We did not examine whether the disconnectedness preceded or resulted from the split, nor can we identify whether the parent or adult child instigated the disconnectedness. For instance, a strained or estranged parent–child relationship may reverberate throughout the family and ultimately destabilize the parents’ marriage. Alternatively, an adult child may recede from family interactions following a split to avoid conflict and tension or may cut off contact with the parent whom they believe was responsible for the relationship dissolution (Arránz Becker & Hank, 2022; Bowen, 1992). Despite these limitations, our study makes important contributions to the study of gray divorce, silver splits, and late-life mental health more broadly, revealing that there may not be a single depressive symptoms profile that emerges within the context of stressful life events. Our findings suggest that researchers may benefit from using advanced longitudinal methods, such as fixed-effects regression models, to better account for selection into divorce as well as other potentially stressful family transitions in later life (Kapelle & Monden, 2024; Leopold, 2018; Tosi & van den Broek, 2020). We hope that our results offer an alternative to the narrative that laterlife dissolution is uniformly distressing, and instead highlight sources of heterogeneity in mental health outcomes on the basis of one’s other family ties. Appendix See Fig.3 and Tables5, 6, and 7. L.Jessee, D.Carr 27 Page 16 of 20 Fig. 3 Fixed effects linear regression models predicting changes in depressive symptoms for marital dissolutions. Whiskers indicate 95% confidence intervals. Labels indicate regression coefficients. Black and grey markers indicate statistically significant differences between connected and disconnected respondents (p < 0.1). SHARE, waves 1, 2, 4, 5, 6, 7, 8, 9 release 9.0.0. unweighted data. Controlled for age, interview year and for imputed values. Table 5 Overview of number and share of missing values SHARE, waves 1, 2, 4, 5, 6, 7, 8, 9 release 9.0.0 Number Percentage Number of depressive symptoms 34 1.53 Silver split 0 0.00 Parent–child disconnectedness 198 8.90 Age at interview 0 0.00 Interview year 0 0.00 Silver Splits andParent–Child Disconnectedness: Mental Health… Page 17 of 20 27 Acknowledgements The research was conducted while the first author received a scholarship from the Table 6 Variable composition of depressive symptoms among connected and disconnected respondents SHARE, waves 1, 2, 4, 5, 6, 7, 8, 9 release 9.0.0 Euro-D Range Mean SD N Connected 0–12 1973 Overall 2.5 2.3 Between 1.9 Within 1.4 Disconnected 0–10 209 Overall 3.6 2.5 Between 1.8 Within 1.8 Table 7 Differences in depressive symptoms following a silver split by parent–child disconnectedness (full interaction model) Controlled for age, interview year and for imputed values. SHARE, waves 1, 2, 4, 5, 6, 7, 8, 9, release 9.0.0. unweighted data. † p < 0.1, *p < .05, **p < .01, ***p < .001 Depressive symptoms B/(SE) Time before/since silver split (ref.: −4years or more) 0.00 (.) −1/−3 0.03 (0.17) 0 (Year of silver split) 0.09 (0.17) −1/−3 0.15 (0.22) +4 or more −0.30 (0.30) Time before/since silver split (ref.: −4years or more)×disconnected 0.00 (.) −1/−3 × disconnected −0.13 (0.56) 0 (Year of silver split) × disconnected 0.93† (0.50) +1/+3 × disconnected 0.40 (0.71) +4 or more × disconnected 1.19* (0.53) Person-wave observations 2216 Individual observations 546 L.Jessee, D.Carr 27 Page 18 of 20 Cologne Graduate School in Management, Economics and Social Sciences of the University of Cologne. Furthermore, this work was supported by a fellowship of the German Academic Exchange Service (DAAD). The authors would like to thank Karsten Hank and Lea Ellwardt for their valuable feedback. Author Contributions L. Jessee planned the study, performed all statistical analyses, wrote the first draft of the paper, and contributed to revising the paper. D. Carr helped planning the study and contributed to revising the paper. Funding No funds, grants, or other support was received. Data Availability This paper uses data from SHARE Waves 1, 2, 3, 4, 5, 6, 7, 8 and 9 (DOIs: https:// doi. org/ 10. 6103/ SHARE. w1. 900, https:// doi. org/ 10. 6103/ SHARE. w2. 900, https:// doi. org/ 10. 6103/ SHARE. w3. 900, https:// doi. org/ 10. 6103/ SHARE. w4. 900, https:// doi. org/ 10. 6103/ SHARE. w5. 900, https:// doi. org/ 10. 6103/ SHARE. w6. 900, https:// doi. org/ 10. 6103/ SHARE. w7. 900, https:// doi. org/ 10. 6103/ SHARE. w8. 900, https:// doi. org/ 10. 6103/ SHARE. w8ca. 900, https:// doi. org/ 10. 6103/ SHARE. w9. 900, https:// doi. org/ 10. 6103/ SHARE. w9ca9 00) see Börsch-Supan et al. (2013) for methodological details. (1) The SHARE data collection has been funded by the European Commission, DG RTD through FP5 (QLK6-CT-2001-00360), FP6 (SHARE-I3: RII-CT-2006-062193, COMPARE: CIT5-CT-2005-028857, SHARELIFE: CIT4-CT-2006-028812), FP7 (SHARE-PREP: GA No 211909, SHARE-LEAP: GA No 227822, SHARE M4: GA No 261982, DASISH: GA No 283646) and Horizon 2020 (SHARE-DEV3: GA No 676536, SHARE-COHESION: GA No 870628, SERISS: GA No 654221, SSHOC: GA No 823782, SHARE-COVID19: GA No 101015924) and by DG Employment, Social Affairs & Inclusion through VS 2015/0195, VS 2016/0135, VS 2018/0285, VS 2019/0332, VS 2020/0313 and SHAREEUCOV: GA No 101052589 and EUCOVII: GA No 101102412. Additional funding from the German Ministry of Education and Research, the Max Planck Society for the Advancement of Science, the U.S. National Institute on Aging (U01_AG09740-13S2, P01_AG005842, P01_AG08291, P30_AG12815, R21_AG025169, Y1-AG-4553-01, IAG_BSR06-11, OGHA_04-064, BSR12-04, R01_AG052527-02, HHSN271201300071C, RAG052527A) and from various national funding sources is gratefully acknowledged (see www. shareeric. eu). Declarations Competing interests The authors have no competing interests to declare. Consent for Publication All authors read and approved the final manuscript. Ethics Approval and Consent to Participate The authors have nothing to declare. Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ byncnd/4. 0/. References Agllias, K. (2018). Missing family: The adult child’s experience of parental estrangement. Journal of Social Work Practice, 32(1), 59–72. https:// doi. org/ 10. 1080/ 02650 533. 2017. 13264 71 Silver Splits andParent–Child Disconnectedness: Mental Health… Page 19 of 20 27 Alderotti, G., Tomassini, C., & Vignoli, D. (2022). ‘Silver splits’ in Europe: The role of grandchildren and other correlates. Demographic Research, 46, 619–652. https:// doi. org/ 10. 4054/ DemRes. 2022. 46. 21 Amato, P. R. (2000). The consequences of divorce for adults and children. Journal of Marriage and Family, 62(4), 1269–1287. https:// doi. org/ 10. 1111/j. 17413737. 2000. 01269.x Andreas, S., Schulz, H., Volkert, J., Dehoust, M., Sehner, S., Suling, A., etal. (2017). Prevalence of mental disorders in elderly people: The European MentDis_ICF65+ study. The British Journal of Psychiatry, 210(2), 125–131. https:// doi. org/ 10. 1192/ bjp. bp. 115. 180463 Arránz Becker, O., & Hank, K. (2022). Adult children’s estrangement from parents in Germany. Journal of Marriage and Family, 84(1), 347–360. https:// doi. org/ 10. 1111/ jomf. 12796 Börsch-Supan, A., Brandt, M., Hunkler, C., Kneip, T., Korbmacher, J., Malter, F., etal. (2013). Data resource profile: The Survey of Health, Ageing and Retirement in Europe (SHARE). International Journal of Epidemiology, 42(4), 992–1001. https:// doi. org/ 10. 1093/ ije/ dyt088 Bowen, M. (1992). Family therapy in clinical practice (1st ed.). Rowman & Littlefield Publishers, Incorporated. Brown, S. L., & Lin, I.-F. (2012). The gray divorce revolution: Rising divorce among middle–aged and older adults, 1990–2010. The Journals of Gerontology. Series B, Psychological Sciences and Social Sciences, 67(6), 731–741. https:// doi. org/ 10. 1093/ geronb/ gbs089 Brown, S. L., & Lin, I.-F. (2022). The graying of divorce: A half century of change. The Journals of Gerontology. Series B, Psychological Sciences and Social Sciences, 77(9), 1710–1720. https:// doi. org/ 10. 1093/ geronb/ gbac0 57 Charles, S. T., & Carstensen, L. L. (2002). Marriage in old age. In M. Yalom, L. L. Carstensen, E. Freedman, & B. Gelpi (Eds.), Inside the American couple: New thinking, new challenges (pp. 236–254). Berkeley: University of California Press. Conti, R. P. (2015). Family estrangement: Establishing a prevalence rate. Journal of Psychology and Behavioral Science. https:// doi. org/ 10. 15640/ jpbs. v3n2a4 Fingerman, K. L., Cheng, Y.-P., Birditt, K., & Zarit, S. (2012). Only as happy as the least happy child: Multiple grown children’s problems and successes and middle-aged parents’ well-being. The Journals of Gerontology: Series B, Psychological Sciences and Social Sciences, 67(2), 184–193. https:// doi. org/ 10. 1093/ geronb/ gbr086 Giesselmann, M., & Schmidt-Catran, A. W. (2022). Interactions in fixed effects regression models. Sociological Methods and Research, 51(3), 1100–1127. https:// doi. org/ 10. 1177/ 00491 24120 914934 Gilligan, M., Suitor, J. J., & Pillemer, K. (2015). Estrangement between mothers and adult children: The role of norms and values. Journal of Marriage and Family, 77(4), 908–920. https:// doi. org/ 10. 1111/ jomf. 12207 Höllinger, F., & Haller, M. (1990). Kinship and social networks in modern societies: A cross-cultural comparison among seven nations. European Sociological Review, 6(2), 103–124. https:// doi. org/ 10. 1093/ oxfor djour nals. esr. a0365 53 Hu, Y., Metsä-Simola, N., Malmberg, S., & Martikainen, P. (2024). Trajectories of antidepressant use before and after union dissolution and re-partnering in later life: A prospective total population register-based cohort study. Journal of Epidemiology and Community Health, 78(5), 277–283. https:// doi. org/ 10. 1136/ jech2023221529 Jessee, L., & Carr, D. (2025). Parent–child disconnectedness and older European adults’ mental health: Do patterns differ by marital status and gender? The Journals of Gerontology: Series B, Psychological Sciences and Social Sciences, 80(6), gbaf024. https:// doi. org/ 10. 1093/ geronb/ gbaf0 24 Kalmijn, M. (2023). Weakened parent–child ties and the well-being of older divorced parents. Demographic Research, 48(21), 591–608. https:// doi. org/ 10. 4054/ DemRes. 2023. 48. 21 Kapelle, N., & Monden, C. (2024). Transitory or chronic? Gendered loneliness trajectories over widowhood and separation in older age. Journal of Health and Social Behavior, 65(2), 292–308. https:// doi. org/ 10. 1177/ 00221 46523 12237 19 Kohler, U., Class, F., & Sawert, T. (2024). Control variable selection in applied quantitative sociology: A critical review. European Sociological Review, 40(1), 173–186. https:// doi. org/ 10. 1093/ esr/ jcac0 78 Leopold, T. (2018). Gender differences in the consequences of divorce: A study of multiple outcomes. Demography, 55(3), 769–797. https:// doi. org/ 10. 1007/ s135240180667-6 L.Jessee, D.Carr 27 Page 20 of 20 Lin, I.-F., & Brown, S. L. (2020). Consequences of later-life divorce and widowhood for adult well-being: A call for the convalescence model. Journal of Family Theory and Review, 12(2), 264–277. https:// doi. org/ 10. 1111/ jftr. 12366 Lin, I.-F., Brown, S. L., & Mellencamp, K. A. (2024). Gray divorce and parent–child disconnectedness: Implications for depressive symptoms. Journal of Marriage and Family, 86(1), 95–110. https:// doi. org/ 10. 1111/ jomf. 12936 Lin, I.-F., Brown, S. L., Wright, M. R., & Hammersmith, A. M. (2019). Depressive symptoms following later-life marital dissolution and subsequent repartnering. Journal of Health and Social Behavior, 60(2), 153–168. https:// doi. org/ 10. 1177/ 00221 46519 839683 Mathers, C. D., & Loncar, D. (2006). Projections of global mortality and burden of disease from 2002 to 2030. PLoS Medicine, 3(11), e442. https:// doi. org/ 10. 1371/ journ al. pmed. 00304 42 National Center for Health Statistics. (2023). Marriage and divorce. https:// www. cdc. gov/ nchs/ fasta ts/ marri agedivor ce. htm. Accessed 4 March 2024. Norris, F. H., & Murrell, S. A. (1987). Transitory impact of life-event stress on psychological symptoms in older adults. Journal of Health and Social Behavior, 28(2), 197–211. https:// doi. org/ 10. 2307/ 21371 32 Pearlin, L. I., Schieman, S., Fazio, E. M., & Meersman, S. C. (2005). Stress, health, and the life course: Some conceptual perspectives. Journal of Health and Social Behavior, 46(2), 205–219. https:// doi. org/ 10. 1177/ 00221 46505 04600 206 Prince, M. J., Reischies, F., Beekman, A. T., Fuhrer, R., Jonker, C., Kivela, S. L., etal. (1999). Development of the EURO-D scale-a European Union initiative to compare symptoms of depression in 14 European centres. The British Journal of Psychiatry: THe Journal of Mental Science, 174, 330–338. https:// doi. org/ 10. 1192/ bjp. 174.4. 330 Reczek, R., Stacey, L., & Thomeer, M. B. (2023). Parent-adult child estrangement in the United States by gender, race/ethnicity, and sexuality. Journal of Marriage and Family, 85(2), 494–517. https:// doi. org/ 10. 1111/ jomf. 12898 Reczek, R., Thomeer, M. B., & Bijou, C. (2025). How intergenerational estrangement matters for maternal and adult children’s health. Journal of Marriage and Family, 87(1), 92–113. https:// doi. org/ 10. 1111/ jomf. 13013 Scharp, K. M., & Hall, E. D. (2017). Family marginalization, alienation, and estrangement: Questioning the nonvoluntary status of family relationships. Communication Yearbook, 41(1), 28–45. https:// doi. org/ 10. 1080/ 23808 985. 2017. 12856 80 Scherpe, J. M., & Hayward, A. (2017). The future of registered partnerships: Family recognition beyond marriage? (Vol. 44). Intersentia. https:// doi. org/ 10. 1017/ 97817 80686 325 Silverstein, M., Gans, D., Lowenstein, A., Giarrusso, R., & Bengtson, V. L. (2010). Older parent–child relationships in six developed nations: Comparisons at the intersection of affection and conflict. Journal of Marriage and Family, 72(4), 1006–1021. https:// doi. org/ 10. 1111/j. 17413737. 2010. 00745.x Solaz, A. (2021). More frequent separation and repartnering among people aged 50 and over. Population and Sociétés, 586(2), 1–4. https:// doi. org/ 10. 3917/ popsoc. 586. 0001 Tosi, M., & van den Broek, T. (2020). Gray divorce and mental health in the United Kingdom. Social Science and Medicine, 256, 113030. https:// doi. org/ 10. 1016/j. socsc imed. 2020. 113030 Vaisey, S., & Miles, A. (2017). What you can—and can’t—do with three-wave panel data. Sociological Methods and Research, 46(1), 44–67. https:// doi. org/ 10. 1177/ 00491 24114 547769 Vignoli, D., Alderotti, G., & Tomassini, C. (2025). Partners’ health and silver splits in Europe: A gendered pattern? Journal of Marriage and Family. https:// doi. org/ 10. 1111/ jomf. 13077 Žilinčíková, Z., & Schnor, C. (2021). Who moves out and who keeps the home? Short-term and mediumterm mobility consequences of grey divorce in Belgium. Demographic Research, 45, 291–328. Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.