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

Jessee, Lisa; Ca , Debo ah
A icle — Published Ve sion
Sil e Spli s and Pa en –Child Disconnec edness: Men al
Heal h Consequences o Eu opean Olde Adul s
Eu opean Jou nal o Popula ion
P o ided in Coope a ion wi h:
Sp inge Na u e
Sugges ed Ci a ion: Jessee, Lisa; Ca , Debo ah (2025) : Sil e Spli s and Pa en –Child
Disconnec edness: Men al Heal h Consequences o Eu opean Olde Adul s, Eu opean Jou nal o
Popula ion, ISSN 1572-9885, Sp inge Ne he lands, Do d ech , Vol. 41, Iss. 1,
h ps://doi.o g/10.1007/s10680-025-09751-9
This Ve sion is a ailable a :
h ps://hdl.handle.ne /10419/330415
S anda d-Nu zungsbedingungen:
Die Dokumen e au EconS o dü en zu eigenen wissenscha lichen
Zwecken und zum P i a geb auch gespeiche und kopie we den.
Sie dü en die Dokumen e nich ü ö en liche ode komme zielle
Zwecke e iel äl igen, ö en lich auss ellen, ö en lich zugänglich
machen, e eiben ode ande wei ig nu zen.
So e n die Ve asse die Dokumen e un e Open-Con en -Lizenzen
(insbesonde e CC-Lizenzen) zu Ve ügung ges ell haben soll en,
gel en abweichend on diesen Nu zungsbedingungen die in de do
genann en Lizenz gewäh en Nu zungs ech e.
Te ms o use:
Documen s in EconS o may be sa ed and copied o you pe sonal
and schola ly pu poses.
You a e no o copy documen s o public o comme cial pu poses, o
exhibi he documen s publicly, o make hem publicly a ailable on he
in e ne , o o dis ibu e o o he wise use he documen s in public.
I he documen s ha e been made a ailable unde an Open Con en
Licence (especially C ea i e Commons Licences), you may exe cise
u he usage igh s as speci ied in he indica ed licence.
h ps://c ea i ecommons.o g/licenses/by/4.0/
Vol.:(0123456789)
Eu opean Jou nal o Popula ion (2025) 41:27
h ps://doi.o g/10.1007/s10680-025-09751-9
BRIEF REPORT
Sil e Spli s andPa en –Child Disconnec edness: Men al
Heal h Consequences o Eu opean Olde Adul s
LisaJessee1 · Debo ahCa 2
Recei ed: 28 Ap il 2025 / Accep ed: 10 Sep embe 2025
© The Au ho (s) 2025
Abs ac
Rising a es o “sil e spli s” in Eu ope esemble inc eases in g ay di o ce in he
U.S. Pa ne ship dissolu ions may ha m olde adul s’ men al heal h, especially o
‘disconnec ed’ pa en s who do no ecei e suppo om hei child en. Howe e ,
esea che s ha e elied p ima ily on mul ile el modeling, neglec ing unobse ed
cha ac e is ics ha may selec an indi idual in o bo h di o ce and pa en –child dis-
connec edness. This b ie epo add esses his esea ch gap by es ima ing ixed-
e ec s linea eg ession models ha con ol o ime-in a ian con ounde s. We used
da a om he Su ey o Heal h, Ageing and Re i emen in Eu ope (SHARE; 2004–
2022, N = 2216 obse a ions, 546 sil e spli s) o documen changes in dep essi e
symp oms p e- and pos -dissolu ion and e alua e whe he hese pa e ns a e mode -
a ed by pa en –child disconnec edness. Consis en wi h p e ious esea ch, we ind
ha dep essi e symp oms inc ease s eeply in he yea o dissolu ion and emain
high ou yea s pos -dissolu ion o pa en s who a e disconnec ed om hei adul
child( en). Howe e , indi iduals who main ain a ela ionship wi h all hei child( en)
show s able le els o dep essi e symp oms h oughou he dissolu ion p ocess, chal-
lenging he assump ion ha dissolu ion is uni o mly dis essing. Ou esul s e eal
ha dep essi e symp oms ajec o ies du ing he pe iod p eceding and ollowing a
majo li e e en di e ac oss socio ela ional con ex s. Social p og ams and suppo s
o di o ced olde adul s should ecognize his he e ogenei y a he han assuming
uni o mly nega i e men al heal h ou comes.
Keywo ds G ay di o ce· Dep essi e symp oms· Pa en –child ela ionships·
In e gene a ional solida i y· Union dissolu ion· Well-being
* Lisa Jessee
[email p o ec ed]oeln.de
Debo ah Ca
ca [email p o ec ed]
1 Depa men o Sociology andSocial Psychology, Uni e si y o Cologne,
Albe us-Magnus-Pla z, 50923Cologne, Ge many
2 Depa men o Sociology andCen e o Inno a ion inSocial Science, Bos on Uni e si y, 704
Commonweal h A e., Bos on, MA02215, USA
L.Jessee, D.Ca
27 Page 2 o 20
1 In oduc ion
Ra es o ‘sil e spli s’— he dissolu ion o ma iages o oman ic pa ne ships a
o a e age 50—ha e isen in Eu ope, mi o ing ‘g ay di o ce’ ends in he U.S.
(Alde o i e  al., 2022; B own & Lin, 2022; Solaz, 2021; Vignoli e  al., 2025;
Žilinčíko á & Schno , 2021). The numbe o ela ionship dissolu ions in la e li e
is expec ed o ise u he in he coming yea s, gi en popula ion aging (B own &
Lin, 2012). The loss o a co- esiden ial in ima e pa ne h ough ela ionship dis-
solu ion may dep i e adul s o a c ucial sou ce o socioemo ional and ins umen al
suppo (B own & Lin, 2022). Union dissolu ions may be especially consequen ial
o olde adul s because age- ela ed ansi ions including e i emen , onse o heal h
p oblems, and he dea hs o pee s may diminish hei le els o con ac wi h o he
po en ially suppo i e ies (Cha les & Ca s ensen, 2002).
Classic w i ings on s ess ul li e e en s emphasize ha ansi ions like di o ce
may igge men al heal h symp oms including dep ession, due o bo h he acu e
s ess o he e en and he sequelae o seconda y s esso s ha ollow, such as a
d op in household income o he loss o a helpma e and con idan e (e.g., No is &
Mu ell, 1987). The di o ce-s ess-adjus men model coun e s ha union dissolu ion
is a p ocess ha can a ec men al heal h e en be o e he e en due o p ecipi a ing
s esso s like pe sis en ela ionship con lic and he an icipa ion o u u e dis up-
ions o daily li e and e e yday ou ines (Ama o, 2000).
Empi ical assessmen s o he di o ce-s ess-adjus men amewo k d aw on
h ee dis inc i e concep ual models which o e compe ing pe spec i es ega ding
he ime cou se o men al heal h symp oms ollowing la e -li e dissolu ion (Lin &
B own, 2020). The c isis model p oposes ha union dissolu ions lead o sho - e m
men al heal h declines wi h a ela i ely quick eco e y as indi iduals manage em-
po a y s esso s like legal issues o esiden ial changes, whe eas he ch onic s ain
model p edic s pe sis en men al heal h s uggles om ongoing s esso s, such as
loneliness and he challenges o managing household cho es and inances indepen-
den ly (Lin e al., 2019).The con alescence model, speci ic o olde adul s, simila ly
p edic s long-las ing men al heal h dec emen s ye an e en ual imp o emen due
o olde adul s’ esilience and capaci y o adap in he longe e m (Lin & B own,
2020; Lin e al., 2024).
Ye , he ex en o which men al heal h ajec o ies du ing he sil e spli p ocess
align wi h he c isis, ch onic s ain, o con alescence models may be con ingen
on he a ailabili y o coping esou ces—mos no ably p o ec i e social ela ion-
ships (e.g., Pea lin e al., 2005). Fo olde adul s expe iencing union dissolu ion,
a s ong ela ionship wi h adul child en is conside ed an impo an bu e agains
emo ional dis ess (Tosi & an den B oek, 2020). Con e sely, s ained, con lic -
ual o enuous pa en –child ela ionships may in ensi y and p olong he emo ional
consequences o dissolu ion. Recen esea ch iden i ies one pa icula dimension
o weak pa en –child ies ha heigh ens he men al heal h e ec s o dissolu ion:
pa en –child disconnec edness (Jessee & Ca , 2025; Kalmijn, 2023; Lin e al.,
2024). Disconnec edness is simila o es angemen , which encompasses de icien
emo ional closeness and equency o con ac , and e e s o he pa en ’s lack o
Sil e Spli s andPa en –Child Disconnec edness: Men al Heal h… Page 3 o 20 27
con ac wi h a leas one adul child (Agllias, 2018; A ánz Becke & Hank, 2022;
Con i, 2015; Reczek e al., 2023; Scha p & Hall, 2017). An emphasis on discon-
nec edness om a leas one child is consis en wi h esea ch documen ing ha
pa en al well-being is a ec ed e en when di icul ies o s ains exis wi h only a
single child (Finge man e al., 2012; Reczek e al., 2025). In Eu ope, disconnec -
edness a ec s be ween 1 and 17% o he popula ion, wi h di o ced men (13%)
expe iencing i oughly wice as o en as di o ced women (6%; Jessee & Ca ,
2025).
Disconnec edness may exace ba e and leng hen olde adul s’ pos -dissolu ion
men al heal h symp oms because disconnec ed pa en s may be dep i ed o p ac i-
cal o emo ional suppo om adul child( en), a esou ce c i ical o hei adap a ion
(Lin e al., 2024). Mo eo e , disconnec edness may be a co-occu ing s esso , c ea -
ing a double bu den ha may u he e ode he men al heal h o olde adul s expe-
iencing a sil e spli . Disconnec edness is mo e consequen ial han s ong posi i e
pa en –child ela ions o olde adul s’ men al heal h because weak ies a e a ypical,
s igma ized, and emo ionally pain ul (Gilligan e al., 2015; Lin e al., 2024). Olde
pa en s who a e disconnec ed om hei adul child en may expe ience a mo e sig-
ni ican and p olonged inc ease in men al heal h symp oms du ing a sil e spli
compa ed o connec ed pa en s. This scena io is consis en wi h he ch onic s ain
pe spec i e, as opposed o he c isis o con alescence pe spec i es sugges ed by p e-
ious esea ch on he men al heal h e ec s o g ay di o ces (Lin e al., 2019; Tosi
& an den B oek, 2020). Thus, ou aim is o e alua e he ex en o which dep essi e
symp oms ajec o ies p e- and pos -di o ce di e on he basis o pa en –child dis-
connec edness. We ocus speci ically on dep essi e symp oms because his is one o
he mos common men al heal h p oblems among olde adul s. Un ea ed dep essi e
symp oms also unde mine physical heal h and a e p ojec ed o become a leading
cause o disabili y wo ldwide (And eas e al., 2017; Ma he s & Lonca , 2006).
To ou knowledge, Lin e al. (2024) a e he only esea che s who ha e examined
pa en –child disconnec edness as a con ex ual ac o a ec ing dep essi e symp om
ajec o ies be o e, du ing, and a e g ay di o ce. They acked a sample o US olde
adul s and ound ha pa en s epo ed heigh ened dep essi e symp oms pos -disso-
lu ion, ye he magni ude o his change was signi ican ly la ge o hose who we e
disconnec ed om ( s. connec ed o) hei adul child( en). The di e ence be ween
disconnec ed and connec ed pa en s began o diminish wo yea s a e he dissolu-
ion and con e ged six yea s he ea e . Howe e , he au ho s es ima ed mul ile el
models and did no consis en ly accoun o unobse ed ac o s ha may “selec ”
indi iduals in o la e -li e sepa a ion, dep essi e symp oms, and disconnec edness.
Mo eo e , i is unclea whe he simila pa e ns would occu beyond he U.S.,
gi en heo e ical w i ings unde sco ing ha he emo ional consequences o a pu -
po ed s esso may be condi ional upon he b oade sociocul u al con ex (Pea lin
e al., 2005). Eu ope has lowe a es o bo h di o ce and pa en –child disconnec ed-
ness ela i e o he U.S. (Jessee & Ca , 2025; Na ional Cen e o Heal h S a is-
ics, 2023; Reczek e al., 2023), so olde adul s expe iencing he “double bu den” o
hese wo s esso s may be suscep ible o s igma iza ion, lack o s uc u al suppo s,
and o he ac o s ha may ende hem pa icula ly ulne able o sus ained dep es-
si e symp oms.
L.Jessee, D.Ca
27 Page 4 o 20
Agains his backg ound, his b ie epo examines dep essi e symp oms p e-
and pos -g ay dissolu ion o wo g oups o “sil e spli e s” in Eu ope: pa en s who
a e connec ed e sus disconnec ed om hei adul child en. We d aw on longi udi-
nal da a om eigh wa es (2004–2022) o he Su ey o Heal h, Ageing and Re i e-
men in Eu ope (SHARE) and apply ixed-e ec s eg ession models. A ixed-e ec s
app oach allows us o cap u e bo h obse ed and unobse ed indi idual con ounding
ac o s by sub ac ing he indi idual mean om he obse ed alues o all a iables.
2 Me hods
2.1 Da a
Da a a e om SHARE, a longi udinal biennial su ey o adul s aged 50+ in Eu ope
and Is ael (Bö sch-Supan e  al., 2013), spanning eigh wa es (1 o 9, excluding
wa e 3) om 2004 o 2022 ac oss 26 na ions. Unlike single-na ion su eys, SHARE
allows us o ollow wi hin-pe son ajec o ies among pa en s who ecen ly expe i-
enced a sil e spli , s a i ied by pa en –child disconnec edness. We excluded Is ael
( he only non-Eu opean coun y in SHARE) and I eland (wi h only one wa e o
da a), as well as wa e 3 da a, which only included e ospec i e li e cou se in o ma-
ion. Ou baseline sample comp ises 146,868 indi iduals, wi h baseline de ined as
he i s wa e in which a esponden is obse ed in he panel.
Ou main goal is o ace wi hin-pe son changes in olde pa en s’ dep essi e
symp oms p e- and pos - ela ionship dissolu ion. Thus, we limi ou sample o
pa en s who a e a isk o and subsequen ly expe ience a sil e spli , de ined as a
ma i al o non-ma i al oman ic pa ne ship dissolu ion a e age 50 (Alde o i
e al., 2022) We exclude om ou analy ic sample indi iduals who a e: unde 50
(n = 3740); no in a pa ne ship a baseline (n = 44,008); con inuously pa ne ed
h oughou he obse a ion pe iod (n = 58,516); esponden s wi h child en unde he
age o 18 o childless (n = 6624); wi h only one wa e o da a (n = 33,433); o who
expe ienced wo o mo e sil e spli s du ing he s udy (n = 1). Responden s wi h
child en unde he age o 18 we e excluded as hey we e no a isk o expe iencing
disconnec edness om an adul child. Ou sample includes 546 pa en s who expe-
ienced a sil e spli , wi h an a e age o ou obse a ions pe pa icipan , o aling
2216 obse a ions.
2.2 Measu es
Dep essi e symp oms a e assessed a each a ailable ime poin be o e, du ing and
a e sepa a ion wi h he 12-i em EURO-D scale (P ince e al., 1999). Respond-
en s indica e whe he hey ha e expe ienced each symp om in he pas 12mon hs:
dep ession, pessimism, suicidali y, guil , sleep, in e es , i i abili y, appe i e, a igue,
concen a ion, lack o enjoymen , and ea ulness. Symp om coun s ange om 0
o 12. We used a na u al log ans o ma ion o add ess he skewed dis ibu ion o

Sil e Spli s andPa en –Child Disconnec edness: Men al Heal h… Page 5 o 20 27
symp oms (M = 2.6, SD = 2.3). Resul s we e consis en ac oss models, so we e ain
he coun o ease o in e p e a ion ( esul s a ailable om au ho s).
Ou ocal independen a iable is he ime o and since a sil e spli . Following
p e ious esea ch (Alde o i e al., 2022; Vignoli e al., 2025), we did no s a i y he
analyses by dissolu ion sub ype, bu ins ead pooled ma i al and non-ma i al ansi-
ions in o a single ca ego y o sil e spli s. This decision was made o bo h con-
cep ual and empi ical easons. Ou independen a iable encompasses pe sons who
ei he : ansi ioned om ma ied o di o ced (n = 472); ended a egis e ed pa ne -
ship (n = 20); o dissol ed a cohabi ing ela ionship (n = 53) du ing he obse a ion
pe iod. In many Eu opean coun ies, egis e ed pa ne ships a e compa able o ma -
iages, o e ing some o he same legal igh s and bene i s ( o an o e iew o how
egis e ed pa ne ships align wi h o di e om ma iage igh s ac oss coun ies,
see Sche pe & Haywa d, 2017). To ha e adequa ely powe ed mode a ion analyses,
we combined he la e wo g oups in o a single non-ma i al dissolu ion ca ego y
(n = 73). Desc ip i e s a is ics (pe son-wa e obse a ions) o he wo dissolu ion
ca ego ies a e p esen ed in Table1. We did no de ec s a is ically signi ican di e -
ences in dep essi e symp oms (2.70 s. 2.40) o a es o pa en –child disconnec ed-
ness (8 s. 12%) o ma i al e sus non-ma i al dissolu ions. Mul i a iable analy-
ses showed no signi ican di e ences in symp om ajec o ies be ween ma i al and
non-ma i al unions, and es ic ing ou analy ic sample o ma i al dissolu ions only
esul ed in simila pa e ns as o he ull sample o sil e spli e s (see Appendix
Fig.3). Based on hese conside a ions, we did no di e en ia e be ween dissolu ion
sub ypes in he analyses and ins ead p oceeded wi h a pooled analysis o all dissolu-
ion sub ypes.
SHARE da a lack p ecise in o ma ion on di o ce/sepa a ion da es. Thus, we
documen ed dep essi e symp oms h oughou he dissolu ion p ocess using i e
iming indica o s based on he cu en and las in e iew yea . While SHARE’s
biennial design ypically esul s in in e iews e e y wo yea s, some esponden s
a e obse ed one o h ee yea s apa due o i egula in e iew lags. To accoun
o his a iabili y, we g ouped yea s −1 o −3 in o a single p e-spli ca ego y and
yea s +1 o +3 in o a single pos -spli ca ego y. We c ea ed he ollowing ca ego ies:
(1) a leas ou yea s p e-spli (baseline and e e ence g oup), (2) be ween one o
h ee yea s p e-spli , (3) he yea in which he spli was i s eco ded, (4) one o
h ee yea s pos -spli , and (5) mo e han ou yea s pos -spli . In he i s (4+ yea s
p e-spli ) and las (4+ yea s pos -spli ) ca ego ies, we pooled all obse a ions ha
occu ed 4yea s o mo e be o e o a e he spli . Consequen ly, esponden s could
con ibu e mul iple obse a ions o hese ca ego ies i hey had mul iple da a poin s
in hese ime ames. To assess he obus ness o his coding s a egy, we conduc ed
supplemen al analyses in which we unca ed he sample o exclude all pe son-yea s
obse ed mo e han ou yea s a e he sil e spli , ocusing only on indi iduals
obse ed exac ly ou yea s pos -spli . The esul s emained subs an i ely unchanged
( esul s a ailable upon eques ). Gi en he inc eased s a is ical powe a o ded by
he pooled app oach, we e ained his speci ica ion in ou main analyses.
Due o panel a i ion and i egula pa icipa ion ac oss wa es, no all indi iduals
con ibu e obse a ions a e e y ime poin be o e and a e he spli . Fo example, a
esponden may pa icipa e a wa e , d op ou in + 1, and e u n in + 2, esul ing
L.Jessee, D.Ca
27 Page 6 o 20
in missing da a o in e media e pe iods. As a esul , indi iduals con ibu e a ying
numbe s o obse a ions ac oss he sil e spli s ages. Table2 displays he dis ibu-
ion o imepoin s.
Table 1 Desc ip i e s a is ics by ela ionship dissolu ion s a us
Means (and s anda d de ia ions) shown o con inuous measu es and p opo ions shown o ca ego i-
cal measu es. Values based on pe son-wa e da a. S a is ically signi ican di e ences deno ed as *p < .05,
**p < .01, ***p < .001, ns no signi ican . SHARE, wa es 1, 2, 4, 5, 6, 7, 8, 9, elease 9.0.0. unweigh ed
da a
Ma i al dissolu ion Non-ma i al dis-
solu ion
Di e ence
Dep essi e symp oms ( ange: 0–12) 2.70 2.40 ns
(2.42) (2.12)
Pa en –child disconnec edness 0.08 0.12 ns
Age a baseline 65.13 65.48 ns
(6.50) (7.23)
Sel - a ed heal h
Excellen 0.11 0.08 ns
Ve y good 0.22 0.16 ns
Good 0.37 0.29 ns
Fai 0.21 0.37 ***
Poo 0.09 0.10 ns
Pe cei ed inancial di icul ies 0.37 0.41 ns
Employed 0.32 0.37 ns
Female 0.53 0.51 ns
Eu opean egion
No he n 0.27 0.27 ns
Eas e n 0.15 0.31 ***
Sou he n 0.13 0.05 **
Wes e n 0.45 0.37 ns
Any alues impu ed 0.06 0.13 **
Pe son-wa e obse a ions 1,020 153
Table 2 Numbe o obse a ions ac oss sil e spli s ages
SHARE, wa es 1, 2, 4, 5, 6, 7, 8, 9, elease 9.0.0. unweigh ed da a
N%
Yea s be o e/a e he sil e spli
−4 o mo e 714 32.2
−1 o 3 302 13.6
0 (yea o he sil e spli ) 546 24.6
+1 o 3 204 9.2
+4 o mo e 450 20.3
Pe son-wa e obse a ions 2216 100
Sil e Spli s andPa en –Child Disconnec edness: Men al Heal h… Page 7 o 20 27
Following Lin e al. (2024), pa en –child disconnec edness in he yea o he spli
e e s o pa en s’ lack o con ac wi h a leas one adul child in he 12mon hs p e-
spli (in pe son, by phone, o mail). Pa en –child disconnec edness was included as
a ime-in a ian mode a o , measu ed in he wa e in which he sil e spli was i s
epo ed. We chose a ime-in a ian mode a o , since ecen simula ion s udies indi-
ca e ha in e ac ing wo ime- a ying a iables in a ixed-e ec s model can p oduce
spu ious esul s (Giesselmann & Schmid -Ca an, 2022). Focusing on disconnec ed-
ness om a leas one adul child (as opposed o all child en) aligns wi h esea ch
showing ha ha ing any es anged adul child en can nega i ely a ec pa en al
heal h (Reczek e al., 2025). The e e ence ca ego y includes pa en s in con ac wi h
all o hei child en. In ou analy ic sample, 10 pe cen o pe sons (49 indi iduals)
we e disconnec ed om a leas one adul child and a e hus classi ied as discon-
nec ed. While he numbe o disconnec ed pa en s who expe ience sepa a ion in ou
sample is ela i ely small, he s udy p o ides an ini ial in es iga ion o his unde -
esea ched subg oup, yielding insigh s ha can ad ance and encou age u he
esea ch on he he e ogenei y o la e -li e dissolu ion.
O he g oup o disconnec ed pa en s, 29 pe cen we e disconnec ed om all o
hei child en. Dep essi e symp oms did no di e signi ican ly o he wo sub-
g oups o disconnec ed pa en s, so we did no s a i y analyses by numbe o chil-
d en om whom he pa en is disconnec ed. In he yea o he sil e spli , 61 pe cen
o disconnec ed pa en s we e sepa a ed om a biological child, and 39 pe cen om
a s epchild. Howe e , we did no ha e su icien s a is ical powe o u he s a i y
ou mul i a iable analyses based on whe he he ie was biological. SHARE does
no p o ide in o ma ion on whe he a disconnec ed child is om he dissol ed ma -
iage o pa ne ship o om a p io ela ionship.
(Un-)obse ed ime- a ian con ounde s, such as numbe o child en, educa ion,
gene ic ac o s, pe sonali y ai s, and ela ionship du a ion, a e accoun ed o in
ou ixed-e ec s modeling app oach. Hence, we adjus ed o ime- a ying co a i-
a es only. In line wi h di o ce esea ch ha applies ixed-e ec s linea eg ession
models (Kapelle & Monden, 2024; Leopold, 2018), we included a pa simonious se
o ca ego ical co a ia es o cap u e age and pe iod e ec s, o accoun o he pos-
sibili y ha well-being sys ema ically changes wi h age o a ies ac oss his o ical
pe iods. Age was included as a ou -le el ca ego ical a iable—50–59 ( e e ence),
60–69, 70–79, and 80 and olde —and in e iew yea likewise as 2004–2007 ( e e -
ence), 2010–2013, 2015–2017, and 2019–2022, based on he yea s in which he da a
we e collec ed. Including ca ego ical con ols a oids he p oblem o pe ec collin-
ea i y. Sensi i i y analyses using al e na i e in e al leng hs and quad ic age e ms
p oduced e y simila esul s (all models a ailable om au ho s).
Fo he 14 pe cen o esponden s who had a missing alue on a leas one a i-
able used in he analysis, we impu ed missing alues using chained equa ions (See
Appendix Table 5 o i em-speci ic missingness). We impu ed 10 da ase s, pe -
o med all analyses on each impu ed da ase and combined coe icien s. Mul i a i-
able analyses included a a iable signi ying whe he any alues we e impu ed.
To a oid “o e ‐adjus ing” o a iables on he causal pa hway, we excluded
addi ional ime‐ a ying co a ia es, including heal h s a us, employmen s a us,
and inancial ulne abili y, because hese ac o s may media e he e ec o a
L.Jessee, D.Ca
27 Page 8 o 20
sil e spli on dep essi e symp oms (Ama o, 2000). Fo example, among women,
a sil e spli may in luence dep essi e symp oms h ough educed inancial secu-
i y, whe eas among men hese e ec s may ope a e ia diminishing social ne -
wo ks. I we adjus o hese media o s as simple co a ia es, we isk unde s a -
ing he ue impac o sepa a ion on dep essi e symp oms. A dedica ed media ion
analysis would be p e e able, bu ou sample lacks su icien s a is ical powe
(Kohle e al., 2024). Inco po a ing se e al ime- a ying co a ia es in ou model,
including employmen s a us (wo king s. no wo king), inancial ulne abili-
ies ( inancial di icul ies s. no inancial di icul ies), sel - a ed heal h ( ange:
1 = excellen o 5 = poo ), and epa ne ing ( epa ne ed s. s ill sepa a ed), yields
simila e ec s, bu wi h sligh ly educed magni ude as we an icipa ed. All co a i-
a es we e measu ed a each a ailable SHARE in e iew wa e ac oss he di e en
s ages o he sil e spli .
2.3 Analy ic Plan
We apply ixed-e ec s linea eg ession models wi h obus s anda d e o s o accoun
o unobse ed cha ac e is ics ha may con ound he associa ions among sil e spli s,
disconnec edness, and dep essi e symp oms. Omi ing impo an (unobse ed) a i-
ables can lead o an o e - o unde es ima ion o he ue e ec s (omi ed a iable
bias). Fixed-e ec s panel eg ession models allow esea che s o add ess omi ed a -
iable bias o unobse ed s able indi idual cha ac e is ics, such as pe sonali y ai s,
childhood expe iences, o gene ic ac o s, by sub ac ing he indi idual mean om
he obse ed alues o all a iables (Vaisey & Miles, 2017). Mo eo e , ixed-e ec s
panel eg ession models au oma ically accoun o obse ed ime-in a ian cha ac e -
is ics, such as gende , educa ion, and numbe o child en. We ind su icien wi hin-
indi idual a ia ion o jus i y a ixed-e ec s app oach (see Appendix Table6).
Fixed-e ec s eg ession models use each indi idual as hei own con ol o e
ime o ocus on changes ha occu wi hin indi iduals. We modeled a change in
dep essi e symp oms as a unc ion o ime be o e, du ing, o a e he spli among
(a) connec ed and (b) disconnec ed pa en s. To es o s a is ically signi ican di -
e ences be ween connec ed and disconnec ed pa en s, we es ima ed ully in e ac ed
models (see Appendix Table7 o comple e esul s). The basic equa ion wi hou
in e ac ions is speci ied as ollows:
whe e EUROD ep esen s he EURO-D sco e, ou ou come, o indi idual i a ime
. SILVERSPLIT, ou ocal p edic o , ca ego izes pe iods ela i e o he spli e en .
The o he ime- a ying co a ia es (age, in e iew yea , and whe he alues we e
impu ed) a e accoun ed o by a iables x2, …, xk o indi idual i a ime . The e m
μi cap u es indi idual-speci ic cha ac e is ics ha emain cons an o e ime ( ixed
e ec s) o indi idual i. Finally, Ɛ deno es he e o e m ha a ies ac oss bo h ime
and indi iduals (Replica ion iles a ailable he e h ps:// os . io/ akc7z/? iew_ only=
811 1 15e9d 7a416 890b7 0e660 239a 71).
EURODi
=𝛽
1SILVERSPLITi
+𝛽
2x2i
+⋯+𝛽
kxki
+𝜇
i
+𝜀
i
Sil e Spli s andPa en –Child Disconnec edness: Men al Heal h… Page 15 o 20 27
Second, jus 49 people in ou sample epo ed pa en –child disconnec edness,
necessi a ing a cau ious in e p e a ion o ou s udy esul s. We did no ha e su -
icien s a is ical powe o conside o he sou ces o he e ogenei y wi hin he dis-
connec ed subg oup, such as whe he a pa en was disconnec ed om only one
o mul iple child en, and o he key cha ac e is ics o he disconnec ed child,
including gende , whe he he child was om he cu en o a p e ious union, and
whe he he child was biological o non-biological. Ini ial desc ip i e indings
sugges ha disconnec edness om a biological child may be expe ienced mo e
nega i ely, as pa en s in his g oup epo ed highe a e age dep essi e symp om
sco es compa ed o hose disconnec ed om a non-biological child (3.4 s. 2.1
symp oms). The small sample size also p ecluded obus es s o he e ogenei y
by gende o egion. P elimina y analyses sugges ed no signi ican gende di -
e ences, hough disconnec ed women’s highe baseline symp oms may ansla e
in o g ea e ela i e isk o ele a ed dep ession. Regional pa e ns also poin ed
o la ge inc eases in Sou he n Eu ope, bu hese we e no s a is ically signi ican
( esul s a ailable om au ho s). La ge samples a e needed o assess hese di e -
ences mo e conclusi ely.
Finally, ou measu e o disconnec edness e e ed o con ac du ing he yea
o he spli only. We did no examine whe he he disconnec edness p eceded o
esul ed om he spli , no can we iden i y whe he he pa en o adul child ins i-
ga ed he disconnec edness. Fo ins ance, a s ained o es anged pa en –child ela-
ionship may e e be a e h oughou he amily and ul ima ely des abilize he pa -
en s’ ma iage. Al e na i ely, an adul child may ecede om amily in e ac ions
ollowing a spli o a oid con lic and ension o may cu o con ac wi h he pa en
whom hey belie e was esponsible o he ela ionship dissolu ion (A ánz Becke
& Hank, 2022; Bowen, 1992).
Despi e hese limi a ions, ou s udy makes impo an con ibu ions o he s udy
o g ay di o ce, sil e spli s, and la e-li e men al heal h mo e b oadly, e ealing ha
he e may no be a single dep essi e symp oms p o ile ha eme ges wi hin he con-
ex o s ess ul li e e en s. Ou indings sugges ha esea che s may bene i om
using ad anced longi udinal me hods, such as ixed-e ec s eg ession models, o
be e accoun o selec ion in o di o ce as well as o he po en ially s ess ul amily
ansi ions in la e li e (Kapelle & Monden, 2024; Leopold, 2018; Tosi & an den
B oek, 2020). We hope ha ou esul s o e an al e na i e o he na a i e ha la e -
li e dissolu ion is uni o mly dis essing, and ins ead highligh sou ces o he e ogene-
i y in men al heal h ou comes on he basis o one’s o he amily ies.
Appendix
See Fig.3 and Tables5, 6, and 7.

L.Jessee, D.Ca
27 Page 16 o 20
Fig. 3 Fixed e ec s linea eg ession models p edic ing changes in dep essi e symp oms o ma i al dis-
solu ions. Whiske s indica e 95% con idence in e als. Labels indica e eg ession coe icien s. Black and
g ey ma ke s indica e s a is ically signi ican di e ences be ween connec ed and disconnec ed espond-
en s (p < 0.1). SHARE, wa es 1, 2, 4, 5, 6, 7, 8, 9 elease 9.0.0. unweigh ed da a. Con olled o age,
in e iew yea and o impu ed alues.
Table 5 O e iew o numbe and sha e o missing alues
SHARE, wa es 1, 2, 4, 5, 6, 7, 8, 9 elease 9.0.0
Numbe Pe cen age
Numbe o dep essi e symp oms 34 1.53
Sil e spli 0 0.00
Pa en –child disconnec edness 198 8.90
Age a in e iew 0 0.00
In e iew yea 0 0.00
Sil e Spli s andPa en –Child Disconnec edness: Men al Heal h… Page 17 o 20 27
Acknowledgemen s The esea ch was conduc ed while he i s au ho ecei ed a schola ship om he
Table 6 Va iable composi ion o dep essi e symp oms among connec ed and disconnec ed esponden s
SHARE, wa es 1, 2, 4, 5, 6, 7, 8, 9 elease 9.0.0
Eu o-D Range Mean SD N
Connec ed 0–12 1973
O e all 2.5 2.3
Be ween 1.9
Wi hin 1.4
Disconnec ed 0–10 209
O e all 3.6 2.5
Be ween 1.8
Wi hin 1.8
Table 7 Di e ences in dep essi e symp oms ollowing a sil e spli by pa en –child disconnec edness
( ull in e ac ion model)
Con olled o age, in e iew yea and o impu ed alues. SHARE, wa es 1, 2, 4, 5, 6, 7, 8, 9, elease
9.0.0. unweigh ed da a.
† p < 0.1, *p < .05, **p < .01, ***p < .001
Dep essi e
symp oms
B/(SE)
Time be o e/since sil e spli ( e .: −4yea s o mo e) 0.00
(.)
−1/−3 0.03
(0.17)
0 (Yea o sil e spli ) 0.09
(0.17)
−1/−3 0.15
(0.22)
+4 o mo e −0.30
(0.30)
Time be o e/since sil e spli ( e .: −4yea s o mo e)×disconnec ed 0.00
(.)
−1/−3 × disconnec ed −0.13
(0.56)
0 (Yea o sil e spli ) × disconnec ed 0.93†
(0.50)
+1/+3 × disconnec ed 0.40
(0.71)
+4 o mo e × disconnec ed 1.19*
(0.53)
Pe son-wa e obse a ions 2216
Indi idual obse a ions 546
L.Jessee, D.Ca
27 Page 18 o 20
Cologne G adua e School in Managemen , Economics and Social Sciences o he Uni e si y o Cologne.
Fu he mo e, his wo k was suppo ed by a ellowship o he Ge man Academic Exchange Se ice
(DAAD). The au ho s would like o hank Ka s en Hank and Lea Ellwa d o hei aluable eedback.
Au ho Con ibu ions L. Jessee planned he s udy, pe o med all s a is ical analyses, w o e he i s d a
o he pape , and con ibu ed o e ising he pape . D. Ca helped planning he s udy and con ibu ed o
e ising he pape .
Funding No unds, g an s, o o he suppo was ecei ed.
Da a A ailabili y This pape uses da a om SHARE Wa es 1, 2, 3, 4, 5, 6, 7, 8 and 9 (DOIs: h ps:// doi.
o g/ 10. 6103/ SHARE. w1. 900, h ps:// doi. o g/ 10. 6103/ SHARE. w2. 900, h ps:// doi. o g/ 10. 6103/ SHARE.
w3. 900, h ps:// doi. o g/ 10. 6103/ SHARE. w4. 900, h ps:// doi. o g/ 10. 6103/ SHARE. w5. 900, h ps://
doi. o g/ 10. 6103/ SHARE. w6. 900, h ps:// doi. o g/ 10. 6103/ SHARE. w7. 900, h ps:// doi. o g/ 10. 6103/
SHARE. w8. 900, h ps:// doi. o g/ 10. 6103/ SHARE. w8ca. 900, h ps:// doi. o g/ 10. 6103/ SHARE. w9. 900,
h ps:// doi. o g/ 10. 6103/ SHARE. w9ca9 00) see Bö sch-Supan e  al. (2013) o me hodological de ails.
(1) The SHARE da a collec ion has been unded by he Eu opean Commission, DG RTD h ough 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 Ho izon 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 Employmen , Social A ai s & Inclusion
h ough VS 2015/0195, VS 2016/0135, VS 2018/0285, VS 2019/0332, VS 2020/0313 and SHARE-
EUCOV: GA No 101052589 and EUCOVII: GA No 101102412. Addi ional unding om he Ge man
Minis y o Educa ion and Resea ch, he Max Planck Socie y o he Ad ancemen o Science, he U.S.
Na ional Ins i u e 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 om a ious na ional unding sou ces is g a e ully acknowl-
edged (see www. sha e- e ic. eu).
Decla a ions
Compe ing in e es s The au ho s ha e no compe ing in e es s o decla e.
Consen o Publica ion All au ho s ead and app o ed he inal manusc ip .
E hics App o al and Consen o Pa icipa e The au ho s ha e no hing o decla e.
Open Access This a icle is licensed unde a C ea i e Commons A ibu ion-NonComme cial-NoDe i -
a i es 4.0 In e na ional License, which pe mi s any non-comme cial use, sha ing, dis ibu ion and ep o-
duc ion in any medium o o ma , as long as you gi e app op ia e c edi o he o iginal au ho (s) and he
sou ce, p o ide a link o he C ea i e Commons licence, and indica e i you modi ied he licensed ma e-
ial. You do no ha e pe mission unde his licence o sha e adap ed ma e ial de i ed om his a icle o
pa s o i . The images o o he hi d pa y ma e ial in his a icle a e included in he a icle’s C ea i e
Commons licence, unless indica ed o he wise in a c edi line o he ma e ial. I ma e ial is no included in
he a icle’s C ea i e Commons licence and you in ended use is no pe mi ed by s a u o y egula ion o
exceeds he pe mi ed use, you will need o ob ain pe mission di ec ly om he copy igh holde . To iew
a copy o his licence, isi h p:// c ea i eco mmons. o g/ licen ses/ by- nc- nd/4. 0/.
Re e ences
Agllias, K. (2018). Missing amily: The adul child’s expe ience o pa en al es angemen . Jou nal o
Social Wo k P ac ice, 32(1), 59–72. h ps:// doi. o g/ 10. 1080/ 02650 533. 2017. 13264 71
Sil e Spli s andPa en –Child Disconnec edness: Men al Heal h… Page 19 o 20 27
Alde o i, G., Tomassini, C., & Vignoli, D. (2022). ‘Sil e spli s’ in Eu ope: The ole o g andchild en
and o he co ela es. Demog aphic Resea ch, 46, 619–652. h ps:// doi. o g/ 10. 4054/ DemRes. 2022.
46. 21
Ama o, P. R. (2000). The consequences o di o ce o adul s and child en. Jou nal o Ma iage and Fam-
ily, 62(4), 1269–1287. h ps:// doi. o g/ 10. 1111/j. 1741- 3737. 2000. 01269.x
And eas, S., Schulz, H., Volke , J., Dehous , M., Sehne , S., Suling, A., e al. (2017). P e alence o
men al diso de s in elde ly people: The Eu opean Men Dis_ICF65+ s udy. The B i ish Jou nal o
Psychia y, 210(2), 125–131. h ps:// doi. o g/ 10. 1192/ bjp. bp. 115. 180463
A ánz Becke , O., & Hank, K. (2022). Adul child en’s es angemen om pa en s in Ge many. Jou nal
o Ma iage and Family, 84(1), 347–360. h ps:// doi. o g/ 10. 1111/ jom . 12796
Bö sch-Supan, A., B and , M., Hunkle , C., Kneip, T., Ko bmache , J., Mal e , F., e al. (2013). Da a
esou ce p o ile: The Su ey o Heal h, Ageing and Re i emen in Eu ope (SHARE). In e na ional
Jou nal o Epidemiology, 42(4), 992–1001. h ps:// doi. o g/ 10. 1093/ ije/ dy 088
Bowen, M. (1992). Family he apy in clinical p ac ice (1s ed.). Rowman & Li le ield Publishe s,
Inco po a ed.
B own, S. L., & Lin, I.-F. (2012). The g ay di o ce e olu ion: Rising di o ce among middle–aged and
olde adul s, 1990–2010. The Jou nals o Ge on ology. Se ies B, Psychological Sciences and Social
Sciences, 67(6), 731–741. h ps:// doi. o g/ 10. 1093/ ge onb/ gbs089
B own, S. L., & Lin, I.-F. (2022). The g aying o di o ce: A hal cen u y o change. The Jou nals o Ge -
on ology. Se ies B, Psychological Sciences and Social Sciences, 77(9), 1710–1720. h ps:// doi. o g/
10. 1093/ ge onb/ gbac0 57
Cha les, S. T., & Ca s ensen, L. L. (2002). Ma iage in old age. In M. Yalom, L. L. Ca s ensen, E. F eed-
man, & B. Gelpi (Eds.), Inside he Ame ican couple: New hinking, new challenges (pp. 236–254).
Be keley: Uni e si y o Cali o nia P ess.
Con i, R. P. (2015). Family es angemen : Es ablishing a p e alence a e. Jou nal o Psychology and
Beha io al Science. h ps:// doi. o g/ 10. 15640/ jpbs. 3n2a4
Finge man, K. L., Cheng, Y.-P., Bi di , K., & Za i , S. (2012). Only as happy as he leas happy child:
Mul iple g own child en’s p oblems and successes and middle-aged pa en s’ well-being. The
Jou nals o Ge on ology: Se ies B, Psychological Sciences and Social Sciences, 67(2), 184–193.
h ps:// doi. o g/ 10. 1093/ ge onb/ gb 086
Giesselmann, M., & Schmid -Ca an, A. W. (2022). In e ac ions in ixed e ec s eg ession models.
Sociological Me hods and Resea ch, 51(3), 1100–1127. h ps:// doi. o g/ 10. 1177/ 00491 24120
914934
Gilligan, M., Sui o , J. J., & Pilleme , K. (2015). Es angemen be ween mo he s and adul child en:
The ole o no ms and alues. Jou nal o Ma iage and Family, 77(4), 908–920. h ps:// doi. o g/
10. 1111/ jom . 12207
Höllinge , F., & Halle , M. (1990). Kinship and social ne wo ks in mode n socie ies: A c oss-cul u al
compa ison among se en na ions. Eu opean Sociological Re iew, 6(2), 103–124. h ps:// doi. o g/
10. 1093/ ox o djou nals. es . a0365 53
Hu, Y., Me sä-Simola, N., Malmbe g, S., & Ma ikainen, P. (2024). T ajec o ies o an idep essan use
be o e and a e union dissolu ion and e-pa ne ing in la e li e: A p ospec i e o al popula ion
egis e -based coho s udy. Jou nal o Epidemiology and Communi y Heal h, 78(5), 277–283.
h ps:// doi. o g/ 10. 1136/ jech- 2023- 221529
Jessee, L., & Ca , D. (2025). Pa en –child disconnec edness and olde Eu opean adul s’ men al
heal h: Do pa e ns di e by ma i al s a us and gende ? The Jou nals o Ge on ology: Se ies
B, Psychological Sciences and Social Sciences, 80(6), gba 024. h ps:// doi. o g/ 10. 1093/ ge onb/
gba 0 24
Kalmijn, M. (2023). Weakened pa en –child ies and he well-being o olde di o ced pa en s. Demo-
g aphic Resea ch, 48(21), 591–608. h ps:// doi. o g/ 10. 4054/ DemRes. 2023. 48. 21
Kapelle, N., & Monden, C. (2024). T ansi o y o ch onic? Gende ed loneliness ajec o ies o e wid-
owhood and sepa a ion in olde age. Jou nal o Heal h and Social Beha io , 65(2), 292–308.
h ps:// doi. o g/ 10. 1177/ 00221 46523 12237 19
Kohle , U., Class, F., & Sawe , T. (2024). Con ol a iable selec ion in applied quan i a i e sociol-
ogy: A c i ical e iew. Eu opean Sociological Re iew, 40(1), 173–186. h ps:// doi. o g/ 10. 1093/
es / jcac0 78
Leopold, T. (2018). Gende di e ences in he consequences o di o ce: A s udy o mul iple ou comes.
Demog aphy, 55(3), 769–797. h ps:// doi. o g/ 10. 1007/ s13524- 018- 0667-6
L.Jessee, D.Ca
27 Page 20 o 20
Lin, I.-F., & B own, S. L. (2020). Consequences o la e -li e di o ce and widowhood o adul well-being:
A call o he con alescence model. Jou nal o Family Theo y and Re iew, 12(2), 264–277. h ps://
doi. o g/ 10. 1111/ j . 12366
Lin, I.-F., B own, S. L., & Mellencamp, K. A. (2024). G ay di o ce and pa en –child disconnec edness:
Implica ions o dep essi e symp oms. Jou nal o Ma iage and Family, 86(1), 95–110. h ps:// doi.
o g/ 10. 1111/ jom . 12936
Lin, I.-F., B own, S. L., W igh , M. R., & Hamme smi h, A. M. (2019). Dep essi e symp oms ollowing
la e -li e ma i al dissolu ion and subsequen epa ne ing. Jou nal o Heal h and Social Beha io ,
60(2), 153–168. h ps:// doi. o g/ 10. 1177/ 00221 46519 839683
Ma he s, C. D., & Lonca , D. (2006). P ojec ions o global mo ali y and bu den o disease om 2002 o
2030. PLoS Medicine, 3(11), e442. h ps:// doi. o g/ 10. 1371/ jou n al. pmed. 00304 42
Na ional Cen e o Heal h S a is ics. (2023). Ma iage and di o ce. h ps:// www. cdc. go / nchs/ as a s/
ma i age- di o ce. h m. Accessed 4 Ma ch 2024.
No is, F. H., & Mu ell, S. A. (1987). T ansi o y impac o li e-e en s ess on psychological symp oms
in olde adul s. Jou nal o Heal h and Social Beha io , 28(2), 197–211. h ps:// doi. o g/ 10. 2307/
21371 32
Pea lin, L. I., Schieman, S., Fazio, E. M., & Mee sman, S. C. (2005). S ess, heal h, and he li e cou se:
Some concep ual pe spec i es. Jou nal o Heal h and Social Beha io , 46(2), 205–219. h ps:// doi.
o g/ 10. 1177/ 00221 46505 04600 206
P ince, M. J., Reischies, F., Beekman, A. T., Fuh e , R., Jonke , C., Ki ela, S. L., e al. (1999). De elop-
men o he EURO-D scale-a Eu opean Union ini ia i e o compa e symp oms o dep ession in 14
Eu opean cen es. The B i ish Jou nal o Psychia y: THe Jou nal o Men al Science, 174, 330–338.
h ps:// doi. o g/ 10. 1192/ bjp. 174.4. 330
Reczek, R., S acey, L., & Thomee , M. B. (2023). Pa en -adul child es angemen in he Uni ed S a es by
gende , ace/e hnici y, and sexuali y. Jou nal o Ma iage and Family, 85(2), 494–517. h ps:// doi.
o g/ 10. 1111/ jom . 12898
Reczek, R., Thomee , M. B., & Bijou, C. (2025). How in e gene a ional es angemen ma e s o ma e -
nal and adul child en’s heal h. Jou nal o Ma iage and Family, 87(1), 92–113. h ps:// doi. o g/ 10.
1111/ jom . 13013
Scha p, K. M., & Hall, E. D. (2017). Family ma ginaliza ion, aliena ion, and es angemen : Ques ioning
he non olun a y s a us o amily ela ionships. Communica ion Yea book, 41(1), 28–45. h ps:// doi.
o g/ 10. 1080/ 23808 985. 2017. 12856 80
Sche pe, J. M., & Haywa d, A. (2017). The u u e o egis e ed pa ne ships: Family ecogni ion beyond
ma iage? (Vol. 44). In e sen ia. h ps:// doi. o g/ 10. 1017/ 97817 80686 325
Sil e s ein, M., Gans, D., Lowens ein, A., Gia usso, R., & Beng son, V. L. (2010). Olde pa en –child
ela ionships in six de eloped na ions: Compa isons a he in e sec ion o a ec ion and con lic .
Jou nal o Ma iage and Family, 72(4), 1006–1021. h ps:// doi. o g/ 10. 1111/j. 1741- 3737. 2010.
00745.x
Solaz, A. (2021). Mo e equen sepa a ion and epa ne ing among people aged 50 and o e . Popula ion
and Socié és, 586(2), 1–4. h ps:// doi. o g/ 10. 3917/ popsoc. 586. 0001
Tosi, M., & an den B oek, T. (2020). G ay di o ce and men al heal h in he Uni ed Kingdom. Social Sci-
ence and Medicine, 256, 113030. h ps:// doi. o g/ 10. 1016/j. socsc imed. 2020. 113030
Vaisey, S., & Miles, A. (2017). Wha you can—and can’ —do wi h h ee-wa e panel da a. Sociological
Me hods and Resea ch, 46(1), 44–67. h ps:// doi. o g/ 10. 1177/ 00491 24114 547769
Vignoli, D., Alde o i, G., & Tomassini, C. (2025). Pa ne s’ heal h and sil e spli s in Eu ope: A gen-
de ed pa e n? Jou nal o Ma iage and Family. h ps:// doi. o g/ 10. 1111/ jom . 13077
Žilinčíko á, Z., & Schno , C. (2021). Who mo es ou and who keeps he home? Sho - e m and medium-
e m mobili y consequences o g ey di o ce in Belgium. Demog aphic Resea ch, 45, 291–328.
Publishe ’s No e Sp inge Na u e emains neu al wi h ega d o ju isdic ional claims in published maps
and ins i u ional a ilia ions.