Psichologija ISSN 1392-0359 eISSN 2345-0061
2024, vol. 71, pp. 142–149 DOI: https://doi.org/10.15388/Psichol.2024.71.9
Giedrius Rupšys
Vilnius University
giedrius.rupsys@fsf.stud.vu.lt
https://orcid.org/0000-0001-8082-5886
https://ror.org/03nadee84
Olga Zamalijeva
Vilnius University
olga.zamalijeva@fsf.vu.lt
https://orcid.org/0000-0002-9186-8440
https://ror.org/03nadee84
Antanas Kairys
Vilnius University
antanas.kairys@fsf.vu.lt
https://orcid.org/0000-0001-8082-8016
https://ror.org/03nadee84
Abstract. This study explores the relationship between divorce and well-being among older Europeans (50+), utilizing data from eight waves of the Survey on Health, Ageing and Retirement in Europe (SHARE). A quasi-experimental design was used to compare the well-being of divorced older adults before and after divorce with a control group of married individuals. CASP-12 scale was used to assess well-being. Results from a Linear Mixed Model reveal that divorced individuals tend to have lower well-being compared to their married counterparts, likely due to negative outcomes associated with divorce or preexisting lower well-being among those who choose to divorce. However, divorce is not associated with a change in well-being in the 2–3 year period surrounding divorce, potentially due to extended intervals between measurements or the absence of immediate effect of divorce.
Keywords: divorce, well-being, older adults, ageing.
Santrauka. Šiame tyrime nagrinėjamas skyrybų ir gerovės ryšys tarp vyresnio amžiaus europiečių (50+), remiantis aštuonių Europos sveikatos, senėjimo ir išėjimo į pensiją tyrimo (SHARE) bangų duomenimis. Taikytas kvazieksperimentinis tyrimo metodas, siekiant palyginti išsiskyrusių vyresnio amžiaus asmenų gerovę prieš skyrybas ir po jų su kontrolinės susituokusių asmenų grupės gerove. Gerovei įvertinti buvo naudojama CASP-12 skalė. Tiesinio mišriojo modelio rezultatai atskleidė, kad išsiskyrusių asmenų gerovė turi tendenciją būti mažesnė nei susituokusiųjų, tikėtina, dėl neigiamų skyrybų pasekmių arba dėl to, kad skirtis yra linkę individai, kurių gerovės lygis žemesnis. Imtyje matomas gerovės pokytis per 2–3 metų po skyrybų laikotarpį, tačiau šis pokytis nėra siejamas su skyrybomis; tai gali reikšti, kad skyrybos neturi sąsajos su gerove trumpalaikiu laikotarpiu arba dėl ilgo intervalo tarp matavimų pokytis nėra matomas.
Pagrindiniai žodžiai: skyrybos, gerovė, vyresnio amžiaus žmonės, senėjimas.
Received: 2024-10-29. Accepted: 2024-12-03.
Copyright © 2024 Giedrius Rupšys, Olga Zamalijeva, Antanas Kairys. Published by Vilnius University Press. This is an Open Access article distributed under the terms of the Creative Commons Attribution Licence (CC BY), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Divorce affects many individuals. In 2022, there were 6 million divorces in Europe, affecting twice that many individuals (Eurostat, 2024). In the US, a chance for marriage to end in divorce is 43–46%.(Schoen & Canudas-Romo, 2006). For older individuals (50 years old or more), the divorce rate is even higher, with an increasing trend in the EU (Raley & Sweeney, 2020), and 1 in 3 divorcees being over 50 in the US (Brown & Lin, 2022). Given the high and growing proportion of older divorcees, it’s important to consider this demographic.
Amato (2000) argues that divorce significantly impacts various aspects of an individual’s life. Divorce is associated with increased mortality and other negative health outcomes, including mental health, reduction in social standing as well economic status (Sbarra & Whisman, 2022). Gray et al. (2011) suggest divorcees have a lower well-being for a long time after divorce. There is evidence showing that divorced individuals report to be less happy not only compared to married individuals, but also the ones who are widowed or were never married (Uğur & Aydınbakar, 2024). Finally, research shows that divorce initiators may experience fewer negative emotions, but the level of stress is the same for initiators, noninitiators and in case of a mutual decision (Frisby et al., 2012). Thus, it can be expected that divorce is likely to result in diminished well-being of both parties.
On the other hand, research on the link between divorce and well-being is mixed. There is evidence suggesting well-being improvement after a divorce (Cavapozzi et al., 2020). Amato (2000) highlights individual differences among divorcees: some experience a lasting drop in well-being, some return to baseline over time, and others may view divorce as a positive development, allowing to end an unsatisfying relationship. This highlights the need to consider individual differences.
Amato’s (2000) divorce-stress-adjustment perspective explains these differences. The model posits that well-being after divorce depends on stressors (mediators) and protective factors (moderators). Stressors include sole parenting, loss of child custody, lack of emotional support, ongoing conflict with an ex-spouse, and financial issues. Protective factors include personal resources, attitudes towards divorce, and demographic characteristics. Well-being is reflected in adjustments, such as problem severity and duration, functioning in new roles, and identity independent of the former marriage. These stressors can affect well-being long before the formal divorce process, which is often the climax of years of tension (Amato, 2000).
Despite observed increase in divorce rates among older adults (Raley & Sweeney, 2020), this demographic group remains understudied (Brown & Lin, 2022). Moreover, several additional factors make this group unique and warrant further research. It is observed, that married individuals have longer total and active life expectancies compared to unmarried (Jia & Lubetkin, 2020), highlighting the importance of partnership in later life and possible negative effects of divorce in later life for health and well-being. Furthermore, the long-term negative effects of divorce tend to diminish with remarriage (Gray et al., 2011), a process that may be more challenging for older adults. Finally, the differences in well-being between different age groups found in the studies (following a U-shaped curve in some countries and steadily decreasing in others (Steptoe et al., 2015)) raise questions about whether the findings on the relationship between well-being and divorce at younger ages can be applied to understand the situation of older people.
The purpose of this study is to examine the relationship between divorce and well-being among older Europeans.
The study was conducted using data from the Survey of Health, Ageing and Retirement in Europe (SHARE; SHARE-ERIC, 2024). SHARE is a multidisciplinary and longitudinal study focusing on individuals aged 50 and older, conducted across 27 European countries and Israel. Respondents are enrolled in the study using a stratified probability sampling and the data is collected around every 2–3 years via face-to-face interview (Bergmann et al., 2019). The data for this analysis was drawn from 8 out of 9 available data collection waves, excluding wave 3 that did not have the required data.
The study design used was quasi-experimental, specifically, the nonequivalent control group design, where well-being was compared both between the divorced and nondivorced groups (between-group comparison) and within each group over time (within-group comparison), despite the lack of random assignment. The experimental group comprised of individuals who were married or had a registered partnership in the wave before and were divorced in the wave after. Since divorcees could have divorced at any time between the two measurements, there could be up to a 3-year gap between the divorce and the well-being measurement taken after the divorce. The control group consisted of individuals who were married or had a registered partnership in both waves.
The waves were divided into 6 pairs (1–2, 4–5, etc.) for the purposes of comparing well-being between adjacent waves during which divorce occurred; then all the 6 pairs were analysed in one model. For participants in control group present in more than two waves, two adjacent waves were chosen randomly for the purposes of comparison.
The demographics of the analysed participants are presented in Table 1. It is evident that there are huge sample differences between the compared groups, however, the statistical model used is robust enough to handle unbalanced group size (Gu et al., 2022).
Table 1
Demographic characteristics of analysed participants
|
Waves |
No. participants |
Gender |
Age |
Marital status |
||
|
Male |
Female |
Divorced (experimental group) |
Married (control group) |
|||
|
1–2 |
5 157 |
2 701 (52.4%) |
2 456 (47.6%) |
63.2 ± 8.8 |
31 (0.6%) |
5 126 (99.4%) |
|
4–5 |
3 212 |
1 605 (50.0%) |
1 607 (50.0%) |
67.2 ± 8.2 |
22 (0.7%) |
3 190 (99.3%) |
|
5–6 |
7 252 |
3 649 (50.3%) |
3 603 (49.7%) |
66.0 ± 8.8 |
53 (0.7%) |
7 199 (99.3%) |
|
6–7 |
11 876 |
5 928 (49.9%) |
5 948 (50.1%) |
66.4 ± 8.5 |
89 (0.7%) |
11 787 (99.3%) |
|
7–8 |
6 769 |
3 455 (51.0%) |
3 314 (49.0%) |
66.8 ± 8.3 |
70 (1.0%) |
6 699 (99.0%) |
|
8–9 |
6 254 |
3 149 (50.4%) |
3 ١٠٥ (49.6%) |
68.3 ± 7.8 |
39 (0.6%) |
6 215 (99.4%) |
|
Total |
40 520 |
20 487 (50.7%) |
20 033 (49.5%) |
66.3 ± 8.4 |
304 (0.7%) |
40 216 (99.3%) |
Well-being was measured using Control, Autonomy, Self-realization and Pleasure (CASP)-12 scale (Knesebeck et al., 2005). The 12-item scale comprises of four afore-mentioned aspects forming well-being. The participants chose answers on a Likert scale ranging from 1 (Often) to 4 (Never). The responses were summed to form the well-being score, ranging from 12 to 48, with higher scores indicating greater levels of well-being. The psychometric properties of the CASP-12 were validated by Oliver et al. (2021) using SHARE database. Confirmatory Factor Analysis supported the second-order four-factor structure with strong fit indices, modelling well-being as a single higher-order construct with four strongly interrelated before-mentioned components. This supports the use of total score of CASP-12 as a measure of overall well-being. Reliability analysis showed excellent internal consistency for the scale with Cronbach’s α over 0.8 (Oliver et al., 2021).
The data analysis was performed using R (version 4.4.0) software package. The study was performed using a Linear Mixed Model (LMM) (using lme4; Bates et al., 2015). LMMs are preferred over traditional methods (e.g., ANOVA) for quasi-experiments because they allow for more complex and realistic data modelling (Brown, 2021). The LMM allows to examine whether and how divorced and nondivorced individuals differ in well-being, whether well-being changes across time points, and whether these changes are linked to divorce, while also assessing the role of control variables and individual differences between participants.
The dependent variable is well-being, the independent variables are divorce, time, their interaction, age, gender (as fixed-effects), the waves being compared, country (as random intercepts), and the participant (as random intercept and slope). Time is a binary variable indicating whether the observation occurred before (0) or after (1) divorce (or a reference point for the control group). This allows the model to capture changes in well-being over time and compare whether these changes differ between the groups. Yeo–Johnson transformation (using bestNormalize; Peterson, 2021) was performed on the dependent variable for normalisation. The model met the assumptions of normality, linearity, random distribution of residuals and homoscedasticity. The significance of independent variables was tested using Satterthwaite’s method.
The details of LMM are shown in Table 2. The model explains 66.2% of the variance in well-being (R2c = 0.662).
Controlling for the aforementioned variables, the results indicate that divorcees have significantly lower well-being compared to the control group (β = -0.343; t(478) = -6.355; p < 0.001). Additionally, well-being shows a slight but significant improvement over time across both groups (β = 0.013; t(45 810) = 3.161; p < 0.05). Importantly, the interaction between divorce status and time is not statistically significant (β = 0.064; t(40,520) = 1.341; p = 0.179), indicating that the rate of change in well-being is similar for both divorced and nondivorced individuals, and thus, well-being changes are not attributable to divorce. In summary, while divorcees report lower well-being, the trajectory of well-being over time is not associated with divorce.
Within the sample of individuals aged 50 and older, age was found to have a significant negative relationship with well-being (β = −0.016; t(40,720) = −34.538; p < 0.001), indicating that younger individuals reported higher well-being. Additionally, males had significantly lower well-being compared to females (β = −0.084; t(40,500) = −10.442; p < 0.001).
Regarding random effects, the variability in participants’ baseline well-being was reflected by a variance of 0.468, indicating substantial individual differences. The variance for the effect of divorce across participants was 0.066, showing smaller variability in how divorce is associated with well-being. Across countries, baseline well-being showed a variance of 0.173, suggesting moderate differences. The variance across waves was minimal (0.008), indicating negligible differences in baseline well-being across waves. Lastly, the residual variance was 0.342, capturing differences unexplained by model.
Table 2
Linear Mixed-effects Model details
|
Fixed effects |
|||||||
|
Est./ß |
SE |
t (df) |
p-value |
||||
|
Intercept |
1.210 |
0.106 |
11.434 (29) |
< 0.001 |
|||
|
Divorced |
-0.343 |
0.054 |
-6.355 (478) |
< 0.001 |
|||
|
Time |
0.013 |
0.004 |
3.161 (45 810) |
< 0.05 |
|||
|
Divorced * Time |
0.064 |
0.048 |
1.341 (40 520) |
0.179 |
|||
|
Age |
-0.016 |
0.000 |
-34.538 (40 720) |
< 0.001 |
|||
|
Gender |
-0.084 |
0.008 |
-10.442 (40 500) |
< 0.001 |
|||
|
Random effects |
|||||||
|
Variance |
Std. Deviation |
||||||
|
Participant (intercept) |
0.468 |
0.684 |
|||||
|
Participant (slope) |
0.066 |
0.256 |
|||||
|
Country (intercept) |
0.173 |
0.416 |
|||||
|
Wave (intercept) |
0.008 |
0.087 |
|||||
|
Residual |
0.342 |
0.585 |
|||||
|
Model fit |
|||||||
|
R2 |
|||||||
|
Marginal (fixed-effects only) |
0.020 |
||||||
|
Conditional (random and fixed-effects) |
0.662 |
||||||
The findings indicate that older divorced individuals tend to have lower well-being compared to their married counterparts, consistent with previous studies such as Gray et al. (2011) and Uğur & Aydınbakar (2024). This lower well-being among divorcees can be attributed to the wide-ranging negative effects of divorce on physical health, mental health, social standing, and economic outcomes (Sbarra & Whisman, 2022).
However, the study also shows that divorce is not associated with changes in well-being during 2–3 year period in which formal divorce ensues. This suggests that the decline in well-being likely occurred prior to the measurement period. Two explanations are possible: first, well-being may have deteriorated long before the formal act divorce, which often marks the culmination of years of marital tension (Amato, 2000). Second, individuals who divorce may have lower baseline well-being in the first place.
As for why divorce itself is not linked to noticeable changes in well-being, two possibilities emerge. Either well-being shortly changed but returned to baseline before the second measurement, or remained stable throughout it, in which case immediate formal divorce effect is not suggested.
As Amato (2000) noted, individual differences were observed, albeit nuanced. Well-being varied substantially within participants and across countries before divorce but became more consistent after divorce, suggesting a uniform adjustment process.
The study has several limitations. The comparability of different CASP-12 adaptations was not tested. Additionally, more frequent measurements, as well as expanding the measurement interval beyond the current 0–3 years, would be important to capture the changes over a longer period. Lastly, there is a big discrepancy between the compared group sizes. While the study provides valuable insights into the well-being of older divorcees, future research should address these limitations.
This paper uses data from SHARE Waves 1, 2, 4, 5, 6, 7, 8 and 9 (DOIs: 10.6103/SHARE.w1.900, 10.6103/SHARE.w2.900, 10.6103/SHARE.w4.900, 10.6103/SHARE.w5.900, 10.6103/SHARE.w6.900, 10.6103/SHARE.w7.900, 10.6103/SHARE.w8.900, 10.6103/SHARE.w8ca.900, 10.6103/SHARE.w9.900), see Börsch-Supan et al. (2013) for methodological details. 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 N°211909, SHARE-LEAP: GA N°227822, SHARE M4: GA N°261982, DASISH: GA N°283646) and Horizon 2020 (SHARE-DEV3: GA N°676536, SHARE-COHESION: GA N°870628, SERISS: GA N°654221, SSHOC: GA N°823782, SHARE-COVID19: GA N°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 SHARE-EUCOV: GA N°101052589 and EUCOVII: GA N°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.share-eric.eu).
Amato, P. (2000). The consequences of divorce for adults and children. Journal of Marriage and the Family, 62(4), 1269–1287.
Bates, D., Mächler, M., Bolker, B., & Walker, S. (2015). Fitting Linear Mixed-Effects Models Using lme4. Journal of Statistical Software, 67(1), 1–48. https://doi.org/10.18637/jss.v067.i01
Bergmann, M., Kneip, T., De Luca, G., & Scherpenzeel, A. (2019). Survey participation in the Survey of Health, Ageing and Retirement in Europe (SHARE), Wave 1-7. Based on Release 7.0.0. SHARE Working Paper Series 41-2019. MEA, Max Planck Institute for Social Law and Social Policy. https://share-eric.eu/fileadmin/user_upload/SHARE_Working_Paper/WP_Series_41_2019_Bergmann_et_al.pdf
Börsch-Supan, A., Brandt, M., Hunkler, C., Kneip, T., Korbmacher, J., Malter, F., Schaan, B., Stuck, S., Zuber, S., & SHARE Central Coordination Team (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
Brown, V. A. (2021). An Introduction to Linear Mixed-Effects Modeling in R. Advances in Methods and Practices in Psychological Science, 4(1). https://doi.org/10.1177/2515245920960351
Brown, S. L., & Lin, I. (2022). The graying of divorce: A half century of change. The Journals of Gerontology: Series B, 77(9), 1710–1720. https://doi.org/10.1093/geronb/gbac057
Cavapozzi, D., Fiore, S., & Pasini, G. (2020). Divorce and well-being: Disentangling the role of stress and socio-economic status. The Journal of the Economics of Ageing, 16, Article 100212. https://doi.org/10.1016/j.jeoa.2019.100212
Eurostat. (2024). Divorce indicators [Data set]. https://doi.org/10.2908/DEMO_NDIVIND
Frisby, B. N., Booth-Butterfield, M., Dillow, M. R., Martin, M. M., & Weber, K. D. (2012). Face and resilience in divorce: The impact on emotions, stress, and post-divorce relationships. Journal of Social and Personal Relationships, 29(6), 715–735. https://doi.org/10.1177/0265407512443452
Gray, M., de Vaus, D., Qu, L., & Stanton, D. (2011). Divorce and the wellbeing of older Australians. Ageing and Society, 31(3), 475–498.
Gu, X., Doetkott, C., & Magel, R. (2022). A simulation study using a mixed model framework to analyze the impact of sample size and variability on Type I error. Asian Journal of Probability and Statistics, 20(3), 74–92. https://doi.org/10.9734/ajpas/2022/v20i3428
Jia, H., & Lubetkin, E. (2020). Life expectancy and active life expectancy by marital status among older U.S. adults: Results from the U.S. Medicare Health Outcome Survey (HOS). SSM - Population Health, 12, Article 100642. https://doi.org/10.1016/j.ssmph.2020.100642
Knesebeck, O., Hyde, M., Higgs, P., Kupfer, A., & Siegrist, J. (2005). Health, ageing and retirement in Europe. First results from the survey of health, ageing and retirement in Europe. Mannheim Research Institute for the Economics of Aging (MEA).
Oliver, A., Sentandreu-Mañó, T., Tomás, J. M., Fernández, I., & Sancho, P. (2021). Quality of life in European older adults of SHARE wave 7: Comparing the old and the oldest-old. Journal of Clinical Medicine, 10(13), Article 2850. https://doi.org/10.3390/jcm10132850
Peterson, R. A. (2021). Finding Optimal Normalizing Transformations via bestNormalize. The R Journal, 13(1), 310–329. https://doi.org/10.32614/RJ-2021-041
Raley, R. K., & Sweeney, M. M. (2020). Divorce, repartnering, and stepfamilies: A decade in review. Journal of Marriage and Family, 82(1), 81–99. https://doi.org/10.1111/jomf.12651
Sbarra, D. A., & Whisman, M. A. (2022). Divorce, health, and socioeconomic status: An agenda for psychological science. Current opinion in psychology, 43, 75–78. https://doi.org/10.1016/j.copsyc.2021.06.007
Schoen, R., & Canudas‐Romo, V. (2006). Timing effects on divorce: 20th century experience in the United States. Journal of Marriage and Family, 68(3), 749–758. https://doi.org/10.1111/j.1741-3737.2006.00287.x
SHARE-ERIC. (2024). Survey of Health, Ageing and Retirement in Europe (SHARE) Wave 9 (Release version: 9.0.0) [Data set]. http://dx.doi.org/10.6103/SHARE.w9.900
Steptoe, A., Deaton, A., & Stone, A. (2015). Subjective wellbeing, health, and ageing. The Lancet, 385(9968), 640–648. https://doi.org/10.1016/S0140-6736(13)61489-0
Uğur, Z. B., & Karakaş Aydınbakar, A. (2024). When is divorce more painful? The impact of social context in Türkiye. Journal of Social and Personal Relationships, 41(11), 3503–3521. https://doi.org/10.1177/02654075241269685