Early quality-of-life and psychological predictors of disease-free time and survival in localized prostate cancer

Vol.:(0123456789) 1 3 Quali y o Li e Resea ch (2019) 28:677–686 h ps://doi.o g/10.1007/s11136-018-2069-z Ea ly quali y-o -li e andpsychological p edic o s o disease- ee ime andsu i al inlocalized p os a e cance Ulla‑SiskoLeh o1,4,5 · Ma kkuOjanen2· AnnaVäke ä1,5· TadeuszDyba3,6· A poA omaa4· Pi kkoKellokumpu‑Leh inen1,5 Accep ed: 22 No embe 2018 / Published online: 3 Decembe 2018 © The Au ho (s) 2018 Abs ac Pu pose The cons uc s e alua ed in in es iga ing associa ion be ween psychosocial ac o s and cance su i al has a ied be ween s udies, and ac o s ela ed o quali y o li e (QOL) ha e shown con adic o y esul s. We in es iga ed he e ec o socioeconomic and ea ly QOL and psychological ac o s on disease- ee ime and su i al in localized p os a e cance . Me hods A consecu i e sample o pa ien s wi h localized p os a e cance (T1–3, N0, M0) ea ed wi h ex e nal beam adio he apy comple ed alida ed ques ionnai es on coping wi h cance ( he Ways o Coping Ques ionnai e WOC-CA), ange exp ession ( he Ange Exp ession Scale), li e e en s ( he Li e Expe ience Su ey), and a ious aspec s o QOL ( he Ro e dam Symp om Checklis , he Dep ession Scale DEPS, he EORTC QLQ-C30, he LENT-SOMA ou come measu e) app oxima ely 4.5mon hs a e diagnosis. Cox eg ession analyses we e used o de e mine he p edic o s o he disease- ee and o e all su i al imes measu ed om he da e o diagnosis o he da e o a PSA- elapse and da e o dea h. Resul s A e con olling o biological p ognos ic ac o s, age, and adju an ho monal he apies, mode a e and high socio- economic s a us and an inc eased le el o pain p edic ed longe su i al, whe eas an inc eased le el o p os a e-a ea symp- oms and a igue and, especially, epo s o no/ ew physical symp oms we e p edic o s o a sho e su i al ime. A longe PSA- elapse- ee ime was p edic ed by Cogni i e A oidance/Denial coping, whe eas p oblems in social unc ioning, hope- lessness, and an excellen sel - epo ed QOL p edic ed a sho e PSA- elapse- ee ime. Conclusions Highe socioeconomic s a us was p ognos ic o longe su i al, as p e iously epo ed. Pa ien s wi h a seem- ingly good QOL ( ew physical complain s, excellen sel - epo ed QOL) had poo e p ognoses. This associa ion may due o he su i al dec easing e ec o emo ional non-exp ession; pa ien s wi h high emo ional non-exp ession may o e - epo hei wellbeing in simple measu es, and hus ac ually be in need o ex a a en ion and ca e. Keywo ds Disease- ee ime· Emo ional non-exp ession· Pa ien - epo ed quali y o li e· P os a e cance · Socioeconomic s a us· S ess, Psychological· Su i al In oduc ion The associa ion be ween psychosocial ac o s and cance su i al has been con incingly demons a ed only du ing ecen decades [1–12]. The psychological p ocesses and cons uc s in es iga ed ha e a ied among s udies [4, 8]. Many o he ea lie s udies we e e ospec i e, biological p ognos ic ac o s we e no su icien ly con olled o , and only ce ain ac o (s) we e in es iga ed. In many o he p o- spec i e s udies, he su i al ollow-up ime was e y sho *Ulla-Sisko Leh o ulla-sisk[email p o ec ed] 1 Medical School, Oncology, Uni e si y o Tampe e, 33014Tampe e, Finland 2 Depa men o Psychology, Uni e si y o Tampe e, 33014Tampe e, Finland 3 Finnish Cance Regis y, Pieni Roobe inka u 9, 00130Helsinki, Finland 4 Public Heal h E alua ion andP ojec ion, Na ional Ins i u e o Heal h andWel a e THL, Manne heimin ie 166, P.O. Box30, 00271Helsinki, Finland 5 Depa men o Oncology, Tampe e Uni e si y Hospi al, PL 2000, 33521Tampe e, Finland 6 P esen Add ess: Join Resea ch Cen e (JRC), Eu opean Commission, Building 58A/006, Via En ico Fe mi 2749, TP 581, 21027Isp a, I aly 678 Quali y o Li e Resea ch (2019) 28:677–686 1 3 [4, 8–10, 13, 14]. In addi ion, socioeconomic ac o s we e no o en aken in o accoun [4, 15, 16]. S udies sugges ha some psychosocial ac o s, such as emo ional exp ession and psychological s ess, a e mo e likely o ha e an e ec on su i al compa ed wi h o he ac o s [4]. Speci ically, emo ional non-exp ession (“ ep es- sion”) [6, 17–19], hopelessness [7, 20, 21], dep ession [1, 10, 11, 20], and s ess- ela ed psychological ac o s [8, 12, 20] ha e been sugges ed o p edic a sho e su i al, whe eas social suppo [22] and denial o minimizing coping esponse o he cance o p edic a mo e a o able p ognosis [2, 5–8, 12, 21]. Fac o s ela ed o well-being and quali y o li e (QOL) ha e shown con adic o y esul s [2, 5–7, 9, 23]. In some s udies, a good epo ed QOL has shown an associa- ion wi h a poo e p ognosis [6, 7, 9]. Among he numbe o s udies in es iga ing he impac o psychosocial ac o s on cance p og ession, e y ew ha e deal wi h p os a e cance . Howe e , p os a e cance is he mos common cance among men in de eloped coun ies, including Finnish men (h p://www.cance egi s y. i). In mos coun ies, he a e age age a p os a e cance diagnosis is g ea e han 70yea s. The incidence is inc easing wo ld- wide in pa due o he aging popula ions and he widesp ead es ing o p os a e-speci ic an igen (PSA) [24–27]. Today, mos p os a e cance s a e de ec ed when hey a e localized o a mos locally in asi e, and he p ognosis is gene ally good [24–28]. A majo limi a ion o many p e ious s udies was ha he e ec o psychosocial ac o s has been in es iga ed in isola ion om o he ela ed ac o s o p ocesses [4, 13, 29–31]; howe e , he e ec o hese ac o s does no occu in isola ion. In es iga ing hei impac s sepa a ely esul s in o e -simpli ica ion and misleading esul s. In pa icula , connec ions be ween coping and pe sonali y p ocesses a e impo an [18, 32, 33]. The e o e, he e ec o psychologi- cal and psychosocial ac o s and hei ela i e impac can be iden i ied only when se e al ac o s a e in es iga ed and analyzed join ly, adjus ing hei e ec s on each o he . The psychobiological mechanisms ha a e hough o ha e an impac on cance su i al a e ela ed o psycho- logical s ess [4, 8, 34, 35]. Acco ding o psychological s ess heo ies, he e ec s o s ess on heal h ou comes depend on how a pe son can cope wi h he s ess, cop- ing being he main media o in he p ocess om s ess ul e en s o ou comes such as psychological symp oms and soma ic illness [36, 37]. Social suppo and pe sonali y ac o s modi y he coping p ocess [38]. In he p ocess-o i- en ed iew [36, 37, 39], coping is de ined as “ he pe son’s cons an ly changing cogni i e and beha io al e o s o manage speci ic ex e nal and/o in e nal demands ha a e app aised as axing o exceeding he pe son’s esou ces.” The e a e no in buil assump ions on good o bad coping. The ac ual coping p ocess e e s o he pe son’s e o s o educe, minimize, mas e o ole a e he pe son–en i on- men ansac ion ha has been app aised o be demanding. Psychological s ess p ocesses ha e been linked wi h bio- logical immune down- egula ion in cance [34]. Acco ding o a model p oposed o guide esea ch in psycho-oncology [33], he impac o cance disease and ea men e ec s a e media ed by pe sonal (including pe sonali y and coping s yle, adjus men , and sociode- mog aphic ac o s), medical, social-en i onmen al, and li e s esso ac o s [3, 12, 29, 33]. The ou comes o hese pa hways a e QOL and su i al. In pa icula , a pe sonal- i y- ela ed coping esponse called non-exp ession o nega- i e emo ion is po en ially associa ed wi h poo e cance p og ession [4, 7, 17–19, 40]. Non-exp ession o nega- i e emo ion e e s o a endency o supp ess (i.e., no o exp ess) nega i e eelings [17, 32], such as ange and ea , and leads o less e ec i e coping in s ess ul si ua ions. I has been desc ibed as being emo ionally con ained and “main aining a acade o con en men ” in social ela ion- ships. Rela ed concep s, such as emo ional de ensi eness (an iemo ionali y, i.e., supp ession and con ol o emo- ions) has also been ound o p edic educed su i al in cance [6, 17]. Fu he mo e, he e is a sugges ed cance - p one Type C esponse s yle [18, 19], which is a mul- idimensional cons uc ha includes non-exp ession o nega i e emo ions as a co e elemen , whe e helplessness/ hopelessness in s ess ul si ua ions, and sel -sac i icing, o e -coope a i e, and appeasing beha io a e added. Type C esponse s yle has been ound o be associa ed wi h educed su i al in cance . We p e iously p esen ed a concep ual model on he psy- chosocial p edic o s o wellbeing o cance pa ien s [35]. In he model, we assume ha bo h QOL and su i al a e ou - comes in a s ess p ocess, which is media ed by psychologi- cal s ess p ocesses and ini ia ed by cance and ea men . Coping wi h cance is he main media o in he p ocess, and social suppo and pe sonali y ac o s modi y he coping p ocess. Fu he mo e, we assume ha hese p ocesses a e in e e ed by non-cance li e s esses. We p e iously applied he model o s udying baseline psychosocial p edic o s o su i al ou comes in localized melanoma [7, 12] and b eas cance [6] pa ien s unde 72yea s o age. We iden i ied some gende di e ences, indica ing ha men may espond o di - e en ypes o psychosocial ac o s compa ed wi h women [12]. In he p esen s udy, on he basis o he model and expe i- ences om ou p e ious s udies, we in es iga ed he base- line and ea ly p edic o s o disease- ee and o e all su i al imes in p os a e cance pa ien s o all ages and ea ed wi h ex e nal beam adio he apy in Finland. We hypo hesized ha speci ic ac o s in he psychological s ess p ocesses (pa e ns o coping, ange exp ession ai s, non-cance li e e en s), and componen s o pa ien - epo ed QOL e alua ed 679Quali y o Li e Resea ch (2019) 28:677–686 1 3 a he ime o p ima y ea men a e associa ed wi h disease- ee ime and su i al in localized p os a e cance . Me hods Pa ien s andp ocedu e Newly diagnosed T1–3, N0, M0 p os a e cance pa ien s who we e admi ed o ea men o he Depa men o Oncology, Tampe e Uni e si y Hospi al (Finland) in 2002 and we e ea ed wi h cu a i e ex e nal beam adio he apy we e consecu i ely included as eligible pa ien s o he s udy (n = 104). Pa ien s who we e men ally and physically able o pa icipa e in a s udy in e iew and comple e ques ionnai es (no ac i e se e e men al o neu ological disease) and wi h no p e ious cance disease1 we e in i ed o pa icipa e; 6 pa ien s we e no in i ed because hey could no be in e - iewed (3 demen ia, 2 consequences o s oke, and 1 acu e s a e o ano he physical illness), and 3 pa ien s based on a p e ious cance . Six pa ien s e used o pa icipa e. In he end, 89 pa ien s (86%) we e in e iewed. An addi ional 3 pa ien s we e excluded om he su i al analyses because hey had a T4 disease, and 5 pa ien s because hey had unde wen adical p os a ec omy p io o he adio he apy. Thus, he inal s udy g oup comp ised 81 pa ien s (Table1). In Finland, hospi al dis ic s o ganize specialized medical ca e. Some specialized medical ca e se ices, e.g., oncology, a e o ganized on he basis o special esponsibili y a eas o he i e uni e si y hospi als o Finland. Because o his cen alized cance ca e, nea ly all pa ien s diagnosed in he egion and ea ed wi h ex e nal beam adio he apy we e included in he eligible pa ien s. The pa ien s we e in e iewed a app oxima ely 4.5mon hs ( ange 2–9) a e diagnosis by he same in e - iewe (a medical s uden , he hi d au ho ) du ing a isi o he Depa men o Oncology o ex e nal beam adio he apy. A pa ien was no in e iewed un il wo mon hs a e he diagnosis o a oid dis u bing he pa ien s in an acu e can- ce c isis and o achie e a mo e eliable measu emen , and each pa ien was cu en ly ecei ing ex e nal adio he apy (65% wi h LHRH analog as neo-adju an ho monal he - apy). Each in e iew ook app oxima ely 1.5–2.5h. Du - ing he in e iew he pa ien s comple ed se e al alida ed ques ionnai es add essing a ious psychological ac o s and aspec s o pa ien - epo ed well-being and QOL (see, Meas- u es, below). The pa ien s we e also asked o epo hei demog aphics, oca ional educa ion, and amily income. Mos o he measu es we e he same as used in ou p e ious Table 1 Demog aphic, socioeconomic, disease, and ea men a i- ables in pa ien s Va iable N = 81 (%) Age, mean 66.5yea s (median 68, ange 51–82yea s) 50–59 10 (12) 60–69 48 (59) 70–79 22 (27) > 80 1 (1) Ma i al s a us Single 1 (1) Ma ied o cohabi ing 68 (84) Di o ced 7 (9) Widowed 5 (6) Child en, mean 2.3 Ha e child en 75 (93) Voca ional educa ion None 33 (41) Voca ional cou ses 7 (9) Voca ional school 17 (21) College 17 (21) Uni e si y educa ion (any) 7 (9) Yea ly amily income (EUR) < 17,000 19 (23) 17,000–25,000 22 (27) 25,000–34,000 19 (23) 34,000–42,000 10 (12) > 42,000 11 (13) Socioeconomic s a us (SES) Low 12 (15) Mode a e 57 (70) High 12 (15) Gleason classi ica ion (x + x) 3–4 9 (11) 5 9 (11) 6 28 (35) 7 21 (26) 8 7 (9) 9–10 4 (5) Uniden i ied 3 (4) Tumo classi ica ion T1 28 (35) T2 29 (36) T3 24 (30) Biological iska Low 28 (35) Mode a e 31 (38) High 22 (27) Radia ion To al dose Mean 69.9, ange 68.0–70.4 Radia ion o pel ic lymph nodes 8 (10) 1 A p e ious cance was an icipa ed o in luence he psychological s ess p ocesses ha we e he ini ial esea ch a ge . 680 Quali y o Li e Resea ch (2019) 28:677–686 1 3 melanoma [7, 12] and b eas cance [6] su i al s udies, bu wo heal h- ela ed QOL measu es ( he EORTC QLQ-C30 [41] and he LENT-SOMA ou come measu e [42]) we e added in he p esen sample. Measu es Cance and ea men da a (s aging, Gleason classi ica ion, neo-adju an ho monal he apy, and adju an ho monal he - apy; Table1) and su i al ime and disease- ee ime da a (da e o diagnosis, da e o dea h, da e o PSA- e i ica ion o ecu ence o he cance ) we e ob ained om hospi al medical eco ds. Socioeconomic s a us was ope a ionalized as a combi- na ion o oca ional educa ion and yea ly amily income ( o al income o he household). I was classi ied as ‘low᾽ i he pa ien had no occupa ional educa ion and he yea ly income was < 17,000EUR, ‘mode a e᾽ i he income was 17,000–34,000EUR, and ‘high᾽ i he pa ien had a col- lege o uni e si y educa ion and/o he yea ly income was ≥ 34,000EUR (Table1). The measu ed psychological/psychosocial and QOL domains a e lis ed below: 1. Coping wi h cance was measu ed using he Ways o Coping Ques ionnai e (WOC), a 50-i em sel - epo ques ionnai e (scale 0–3 in e e y i em) de eloped “ o iden i y he hough s and ac ions an indi idual has used o cope wi h a speci ic s ess ul encoun e ”—he e any aspec o he p os a e cance since he diagnosis. We used an i em s uc u e p oposed o o m a cance - speci ic WOC-CA measu e [43, 44], including he pa - e ns Focusing on he Posi i e, Dis ancing, Seeking and Using Social Suppo , Cogni i e A oidance, and Escape A oidance. 2. The ai s o ange exp ession we e e alua ed using he Ange Exp ession Scale (AX Scale) [45]. I e e s o “ he ex en ha an indi idual engages in agg essi e beha - io s when mo i a ed by ang y eelings” and aps h ee dimensions: Ange -in (“indi idual di e ences in he equency ha ang y eelings a e expe ienced bu held in”), Ange -ou (“... eelings o ange a e exp essed in agg essi e beha io ”), and Ange con ol (“...an indi id- ual a emp s o con ol he ou wa d exp ession o ang y eelings”). 3. S ess ul li e e en s we e eco ded om he p eceding yea using he Li e Expe ience Su ey (LES) [46], a lis o 50 e en s add essing bo h he numbe o li e e en s and hei pe cei ed impac . 4. To measu e pa ien s’ symp oms and hei in ensi y, we used he Ro e dam Symp om Checklis (RSCL) [47], which has been de eloped o measu e symp oms epo ed by cance pa ien s. The RSCL includes 30 symp oms (8 psychological and 22 physical, scale 0–3 in e e y symp om) ha he pa ien s may ha e expe i- enced du ing he pas week and a sepa a e single i em index on o e all global quali y-o -li e wi h scale om 1 (ex emely poo ) o 7 (excellen ). The le el o dep essi e symp oms was measu ed using he Dep ession Scale DEPS [48], which has been de eloped o sc eening o dep ession in p ima y heal h ca e se ings and e alua es 10 eelings and dep essi e symp oms (scale 0–3) wi h a co e age-pe iod o he p e ious mon h. One o he DEPS symp oms is hopelessness. 5. Heal h- ela ed QOL o he pa ien s was measu ed using he Eu opean O ganiza ion o Resea ch and T ea - men o Cance Quali y o Li e Ques ionnai e (EORTC QLQ-C30)2 e sion 3.0, Finnish ansla ion [41]. The measu e is di ided in o a global heal h s a us scale, i e unc ional scales (Physical/Role/Emo ional/Cogni i e/ Social unc ioning) and se e al symp om scales/i ems, e.g., a igue, nausea, and pain; all e alua ed om he p e ious week. Local p os a e-a ea symp oms we e e al- ua ed using a modi ica ion o he LENT-SOMA ou come measu e (see Foo no e 2) [42], a clinical ool de eloped o eco d and sco e no mal issue e ec s o adio he apy (u ina y, bowel, and sexual symp oms and dys unc ion). The s udy p o ocol was app o ed by he e hical commi - ee o Tampe e Uni e si y Hospi al, and in o med consen was ob ained om each pa icipan . The i s au ho is bound by na ional (The Union o Finnish Psychologis s) and in e - na ional (Ame ican Psychological Associa ion) e hical codes o psychology. The hospi al dis ic was e hnically homogeneous ega ding he age g oups in ques ion a Acco ding o nomog am o p os a e cance ecu ence [50]: clas- si ied as ‘Low isk’ i Gleason ≤ 6 and PSA < 10, ‘Mode a e isk’ i Gleason = 7 o PSA 10–20, and ‘High isk’ i Gleason ≥ 7 o PSA > 20 Table 1 (con inued) Va iable N = 81 (%) Neo-adju an ho monal he apy None 28 (35) LHRH analog 53 (65) Adju an ho monal he apy Yes 34 (42) O he ch onic disease/condi ion Yes 66 (81) 2 The EORTC QLQ-C30 and he LENT-SOMA ou come measu e we e no used in ou p e ious su i al s udies. 681Quali y o Li e Resea ch (2019) 28:677–686 1 3 S a is ical analyses Desc ip i e s a is ics, ANOVA, es , and Pea son’s co - ela ion ( ), we e used o explo e he sample. Reg ession analysis was used o in es iga e he associa ion be ween backg ound ac o s, psychological/psychosocial ac o s, and he QOL indica o s. The Cox p opo ional haza ds eg es- sion model [49] was used o de e mine he simul aneous and ela i e con ibu ion o he socioeconomic, psychologi- cal, and QOL p edic o s on disease- ee and o e all su i al imes, con olled o age, biological p ognos ic ac o s, and ho monal ea men . The p edic o a iables we e sepa a ely es ed o bo h ou comes. P edic o s o he su i al imes in he Cox models we e conside ed signi ican i he co - esponding p- alue was < 0.05. O e all su i al was measu ed om he da e o diagnosis (da e o PAD) o he da e o dea h and disease- ee ime om he da e o diagnosis o he da e o a PSA- e i ied elapse, o he su i al was censo ed a he da e o las ollow-up (Ma ch 31, 2011). A biological isk classi ica ion in o mild, mode a e, and se e e o T1–2 umo s [50] (Table1, oo - no e a) and sepa a ely T3 umo s was used as a biological p ognos ic ac o . Implemen a ion o neo-adju an ho monal he apy and adju an ho monal he apy we e used as p og- nos ic ea men ac o s. In he Cox models, age, biological p ognos ic ac o s, and ea men ( he ho monal he apies) we e always included in he models. Second, we added he socioeconomic s a us a iable(s) o he models o in es iga e hei impac , and adjus o hei e ec when in es iga ing he e ec s o he o he a iables. Thi d, he psychological and QOL indica- o s (see Measu es; selec ed on he basis o he heo e ical model [35]) we e added in o he models indi idually and in a ious combina ions, and ollowing expe iences om ou p e ious co esponding s udies. To a oid es ing o ela ed a iables, we in es iga ed hei psychome ic p ope ies and mu ual associa ions, and selec ed he a iables ha we e he bes (high C onbach’s Alpha, symme ical dis ibu ion) and we e leas co ela ed wi h each o he . In some a iables, we di ided he answe op ions in o wo ca ego ies (no/yes). The p opo ional haza d assump ion was es ed [51] o speci ic a iables and globally. S a is ical analyses we e pe o med using SPSS o Windows 18 and 20, and S a a 11 (S a a- Co p. 2013. S a a S a is ical So wa e: Release 13. College S a ion, TX: S a aCo p LP, USA). Resul s By Ap il 2011, eigh een (22%) o he 81 pa ien s had died, and he ollow-up ime was ≥ 8yea s 4mon hs (median 8.54 yea s, 95% CI 31.01–31.25 days; no pa ien s we e los om he o e all su i al ollow-up). In o ma ion on p og ession o he p os a e cance , i.e., PSA ollow-up was a ailable o 77 pa ien s, and 18 (23%) o hem had been diagnosed wi h PSA e i ica ion o p os a e cance elapse. Fi e pa ien s wi h elapse had died. Associa ion be ween hepsychological andQOL a iables Acco ding o ou heo e ical model [35] (see, In oduc ion), we in es iga ed i s he associa ion be ween he po en ial p edic o a iables (backg ound and disease/ ea men a i- ables, ange exp ession, and coping) and he wellbeing/QOL indica o s. All he QOL indica o s we e s ongly in e co e- la ed (p- alues gene ally < 0.001). The e o e, we sepa a ely in es iga ed p edic o s o he RSCL and he EORTC QLQ- C30 subscales, and he DEPS. On he whole, he disease and ea men a iables did no associa e wi h QOL, whe eas he coping and ange exp ession a iables we e associa ed wi h se e al wellbeing and QOL indica o s. Highe income o he amily p edic ed mo e psychological symp oms and p oblems in Role unc ioning, Emo ional unc ioning, and Cogni i e unc ioning. Escape coping was associa ed wi h a poo e and Dis ancing coping wi h a be e psychologi- cal heal h and sel -pe cei ed QOL, as measu ed by all he indica o s. The Ange -in ai was associa ed wi h poo e wellbeing, whe eas Ange con ol was associa ed wi h be e wellbeing when e alua ed by any o he measu es despi e social unc ioning. P edic o s o o e all su i al anddisease‑ ee ime T3 umo (HR 5.51, 95% CI 0.88–34.45, p = 0.07) and neo- adju an ho monal ea men (HR 0.26, 95% CI 0.05–1.31, p = 0.1) we e only weakly associa ed wi h o e all su i al. Disease- ee ime was no signi ican ly p edic ed by any o he disease and ea men a iables. Indi idual associa ions be ween hep edic o a iables andsu i al ou comes Lack o oca ional educa ion showed an indi idual asso- cia ion wi h bo h sho e su i al and disease- ee imes, and low amily income was associa ed wi h sho e su i al (Table2). Some indica o s o a good QOL we e indi idu- ally associa ed wi h poo e ou comes: low le el o epo ed physical symp oms ( he lowes 18.5% s. he emaining) was associa ed wi h bo h sho e su i al and disease- ee imes, and he highes sco e in he single-i em quali y-o -li e index (‛excellen ’, epo ed by 18.2%) was associa ed wi h a sho e disease- ee ime. Fu he mo e, inc eased pain and epo ed numbe o li e e en s we e indi idually associa ed wi h longe su i al. 682 Quali y o Li e Resea ch (2019) 28:677–686 1 3 Simul aneous impac s o  hep edic o a iables ono e all su i al Pa ien s’ mode a e and high socioeconomic s a us p edic ed longe o e all su i al when adjus ed o he biological and ea men a iables, wi h mul i a ia e haza d a ios o 0.16 (95% CI 0.05–0.46, p = 0.001) and 0.11 (95% CI 0.02–0.6, p = 0.015), espec i ely. When he psychological and QOL a iables we e added o he model, we ound ha a complex combina ion o a i- ables p edic ed o e all su i al (Table3). Di e en QOL measu es exhibi ed ei he a a o able o an un a o able impac , i.e., an inc eased le el o pain (HR 0.05; 95% CI 0.01–0.32) p edic ed longe su i al, whe eas p os a e-a ea symp oms (HR 1.18; 95% CI 1.03–1.36), inc eased a igue (HR 7.08; 95% CI 1.77–28.32), and epo s o no o ew physical symp oms (HR 9.90; 95% CI 1.48–66.30) we e sig- ni ican p edic o s o sho e su i al ime. Howe e , when he o e all quali y-o -li e index ( o al scale 1–7) was es ed ins ead o he p os a e-a ea symp om scale, i p edic ed a longe su i al (HR 0.51, 95% CI 0.27–0.95, p = 0.033); when bo h scales we e included, he e ec o he o e all QOL was weake (HR 0.56; 95% CI 0.27–1.15, p = 0.113). Simul aneous impac s o  hep edic o a iables onPSA‑ elapse‑ ee ime The PSA- elapse- ee ime was no signi ican ly p edic ed by any o he biological p ognos ic ac o s, ho monal ea - men , o socioeconomic s a us. A longe disease- ee ime was p edic ed by Cogni i e A oidance/Denial coping (HR 0.76; 95% CI 0.59–0.97), whe eas a sho e elapse- ee ime was seen o pa ien s wi h p oblems in social unc ioning (HR 3.32; 95% CI 1.45–7.56), hopelessness (HR 8.90; 95% CI 1.62–48.87), and an excellen sel - epo ed QOL (HR 47.31; 95% CI 6.35–352.33) (Table4). Table 2 Measu es wi h s a is ically signi ican indi idual associa ions wi h o e all su i al ime and disease- ee ime HR haza d a ios in Cox models: impac o each p edic o analyzed indi idually and adjus ed o age, bio- logical p ognos ic ac o s and adju an ho monal he apies a Physical symp oms scale (RSCL) low = sco e < 7 (18.5%) e sus he emaining sco es. When he scale was di ided in o wo le els based on he median, he highe le el was associa ed wi h a longe o e all su - i al wi h HR 0.43, p = 0.1 b In he RSCL quali y-o -li e index (1–7) he highes sco e 7 ‘excellen ’ (18.2%) e sus he emaining sco es c A scale in he EORTC QLQ-C30 d By he LES Indi idual p edic o s Uni a ia e analyses Associa ion wi h o e all su - i al ime Associa ion wi h PSA- elapse- ee ime HR (95% CI) pHR (95% CI) p No oca ional educa ion 2.78 (1.02–7.59) 0.046 2.40 (0.89–6.46) 0.083 Family income low (< 17,000) 4.85 (1.69–13.90) 0.003 n.s Le el o physical symp oms low, no/yesa5.99 (2.02–17.77) 0.001 3.21 (0.96–10.72) 0.058 Sel - epo ed QOL ‘excellen ’, no/yesbn.s 5.21 (1.81–14.95) 0.002 Painc0.20 (0.05–0.83) 0.026 n.s No. o non-cance li e e en sd0.71 (0.51–1.00) 0.048 n.s Table 3 Psychosocial and QOL measu es a he ime o cu a i e adi- o he apy p edic ing o e all su i al HR haza d a ios in he Cox model; all a iables we e adjus ed o each o he a Based on he RSCL Physical symp oms scale: he lowes 18.5% e - sus he emaining sco es b A scale in he EORTC QLQ-C30 c The LENT-SOMA scale. When he RSCL o e all quali y-o -li e index (1–7) was es ed ins ead o he LENT-SOMA p os a e-a ea symp oms, i s ela i e HR was 0.51 (95% CI 0.27–0.95, p = 0.033) P edic o Mul i a ia e analysis HR (95% CI) p Age, yea s 0.98 (0. 89–1.07) 0.599 Biological isk Mode a e isk in Tumo 1–2 0.15 (0.02–0.97) 0.047 High isk in Tumo 1–2 0.52 (0.06–4.64) 0.562 T3 Tumo 4.98 (0.63–39.29) 0.127 Neo-adju an ho monal he apy, no/yes 0.17 (0.02–1.21) 0.076 Adju an ho monal he apy, no/yes 1.54 (0.24–9.77) 0.649 Socioeconomic s a us Mode a e 0.20 (0.04–0.91) 0.037 High 0.04 (0.01–0.34) 0.003 Le el o physical symp oms low, no/yesa9.90 (1.48–66.30) 0.018 Painb0.05 (0.01–0.32) 0.002 U ina y, bowel, and sexual symp oms c1.18 (1.03–1.36) 0.017 Fa igueb7.08 (1.77–28.32) 0.006 683Quali y o Li e Resea ch (2019) 28:677–686 1 3 The p opo ional haza d assump ion was success ully es ed o he inal models locally and globally using Sch- oen eld esiduals [51], indica ing ha hese impac s emain s able o e he ollow-up pe iod. Discussion O e all su i al and disease- ee ime in pa ien s wi h local- ized p os a e cance we e p edic ed by socioeconomic s a- us, psychological ac o s, and pa ien - epo ed QOL. Di - e en QOL measu es demons a ed con adic o y esul s as ei he a a o able o an un a o able impac . Su p isingly, we ound ha he pa ien s’ epo s o no o ew physical complain s p edic ed sho e su i al, whe eas epo ed pain was p ognos ic o longe su i al. In addi ion, a longe PSA- elapse- ee ime was p edic ed by esponding o he cance diagnosis wi h a Cogni i e A oidance/Denial coping pa e n, whe eas p oblems in social unc ioning, hopeless- ness, and—con adic o ily— epo ing an excellen QOL, p edic ed a sho e PSA- elapse- ee ime. We sugges an explana ion o hese su p ising indings, i.e., ha he obse ed su i al dec easing e ec o high sco es in ce ain simple QOL measu es may e lec a ai o non-exp ession o nega i e emo ions (a endency no o exp ess nega i e eelings and o keep ‘ acade o con en men ’), ha has been ound po en ially associa e wi h cance p og ession. We p o- pose ha he pa ien s who sco ed excep ionally low/high in simple wellbeing/QOL measu es may ha e been hose wi h high non-exp ession o nega i e emo ions and ha hey hus o e - epo ed hei wellbeing. Few ea lie s udies on psychosocial ac o s and cance su i al ha e add essed p os a e cance [8, 52, 53] o male pa ien samples ha also include he oldes age-g oups. We we e able o collec a ela i ely small bu egionally ep esen a i e sample o newly diagnosed p os a e cance pa ien s. A he ime o he da a collec ion, he p os a e cance he apy was di e en han oday (e.g., ac i e su - eillance was ha dly used) and he ac i e ea men modes ha e imp o ed e e since. The e we e no popula ion PSA- sc eening p og ams going on in Finland. The s udy sample was limi ed o pa ien s who unde wen cu a i e adio he apy, and he o he majo de ini e he apy mode, adical p os a ec- omy, was no ep esen ed. Howe e , younge and heal hie pa ien s mo e o en unde go ope a ion, whe eas ex e nal adia ion is ecei ed by a wide pa ien coho [54]. Fo p ac- ical easons, he iming o he measu emen since diagnosis a ied (2–9mon hs); he onse o he adio he apy di e ed, and he pa ien s we e willing o ake pa o he in e iew a di e en ime poin s du ing hei adio he apy. We in es iga ed bo h o e all su i al and disease- ee imes because hey a e e y di e en ou comes in cance ypes wi h good p ognoses, such as p os a e cance , and may hus be p edic ed by di e en mechanisms; we ound ha hey we e indeed p edic ed by di e en ac o s. The po en ial p edic o s we e simul aneously in es iga ed o accoun o hei ela i e impac , and we e ei he de i ed om a heo e i- cal model [35] o we e p e iously demons a ed o exhibi an associa ion wi h cance su i al [4, 20, 33]. Mos o he p edic o a iables we e subscales o es ablished alida ed ques ionnai es. The adio he apy was cu en ly going on in all pa ien s. The analyses we e ca e ully adjus ed o age, known biological p ognos ic ac o s, and ea men .3 We ound ha a highe socioeconomic s a us p edic ed a longe su i al bo h independen ly and when e alua ed wi h he psychological and QOL ac o s. The su i al p olonging e ec o socioeconomic s a us in cance is well known [15, 16, 55], and has been p e iously epo ed in p os a e cance pa ien s ecei ing cu a i e ea men [53]. Howe e , p e i- ous s udies o en conside ed i s e ec alone wi hou o he po en ial p edic o s [52, 53, 56]. We concu wi h p e ious s udies sugges ing ha socioeconomic s a us should be aken in o accoun o ob ain alid in o ma ion on he impac o QOL on su i al in p os a e cance [57]. The measu es associa ed wi h poo e ou comes appea ed o be he simple indica o s o sel - epo ed wellbeing, i.e., low sco es in pe cei ed physical symp oms and an excel- len sco e in an o e all QOL index. We did no ind nega- i e associa ion be ween es ablished QOL measu es and Table 4 Psychosocial and QOL measu es a he ime o cu a i e adi- o he apy p edic ing PSA- elapse- ee ime HR haza d a ios in he Cox model; all a iables we e adjus ed o age, biological isk, ho monal he apies, and SES (which we e all s a- is ically non-signi ican ), and we e adjus ed o each o he a Cogni i e a oidance coping was s a is ically signi ican only when all he o he a iables we e included in he model b Based on he RSCL quali y-o -li e index (1–7): he sco e ‘excellen ’ (= 7) e sus he emaining sco es c The EORTC QLQ-C30. When Ange -in pe sonali y ai was es ed ins ead o he social unc ioning, i showed HR o 1.15 (95% CI 0.98– 1.37, p = 0.1) Mul i a ia e analysis HR (95% CI) p Cogni i e a oidance/denial copinga0.76 (0.59–0.97) 0.031 Sel - epo ed QOL ‘excellen ’, no/ yesb 47.31 (6.35–352.33) < 0.001 Hopelessness, no/yes 8.90 (1.62–48.87) 0.012 P oblems in social unc ioningc3.32 (1.45–7.56) 0.004 3 As expec ed, ha ing a locally in asi e umo (T3) p edic ed a sho e su i al when compa ed wi h ully localized disease, and neo- adju an ho monal he apy was e ec i e. 684 Quali y o Li e Resea ch (2019) 28:677–686 1 3 he su i al ou comes, o i he e was an associa ion, i was posi i e. We sugges ha hese esul s may e lec he ha m- ul e ec o non-exp ession o nega i e emo ions, which has been de ined o e e a endency no o exp ess nega i e eelings and o keep ‘ acade o con en men ’[17]. Acco d- ing o his heo y, indi iduals wi h high non-exp ession o nega i e emo ions po en ially unde - epo hei symp oms as hey a e p one o a ep essi e coping s yle, i.e., end o a oid a nega i e a ec , and consequen ly answe wellbeing measu es o e ly posi i ely [17, 32]. This idea was suppo ed by he inding ha he Ange con ol pe sonali y ai was associa ed wi h less epo ed symp oms and a be e QOL. One suppo ing ac o o ou a gumen is ha sco ing low in symp om measu es and e y high in o e all QOL a he ime o ecen cance diagnosis and ongoing adia- ion ea men may no e lec he ue objec i e si ua ion. Ins ead, hese sco es may e lec a pe sonali y ai o coping beha io a a s ess ul si ua ion ( ecen cance diagnosis, ongoing adio he apy). Fu he mo e, when e.g., he quali y- o -li e index was es ed as a o al scale ( e y poo –excellen ) i p edic ed a longe su i al. So, ou esul s did no indica e ha epo ing a good QOL was haza dous, bu ha epo - ing an ‘excellen ’ QOL was. Indi iduals wi h a ep essi e coping s yle ha e also p e iously been ound o sco e low in sel - epo ed physical symp om scales [58]. Up o 20% o he popula ion has been ound o answe sel - epo scales in an o e ly posi i e ashion [32], which co esponds wi h ou inding ha he haza dous e ec on su i al was iden i ied when he sco es we e di ided in o wo ca ego ies: no/ e y ew symp oms o an ‛excellen ’ QOL e sus o he (bo h we e p esen in app oxima ely 18% o he pa ien s). The esul s ob ained on simple wellbeing/QOL indica- o s suppo he conclusions we ha e made in ou p e ious s udies [6, 7], i.e., ha simple global- a ing me hods a e no sui able o elici ing in o ma ion on QOL o all cance pa ien s. Ce ain subg oups (“ ep esso s”) seem o unde - s a e hei condi ion and o e - epo hei wellbeing. The e is a isk ha he simple global- a ing measu es ac ually p o ide mo e in o ma ion on non-exp ession o nega i e emo ions beha io han QOL. I has been sugges ed [32] ha o elici in o ma ion om ep esso s measu es equi ing speci ic answe s should be used ins ead o single-i em indexes o symp om lis s. P os a e-a ea symp oms (u ina y, sexual, bowel) and a igue—as e alua ed wi h well-es ablished alida ed meas- u es comp ising speci ic ques ions—we e p ognos ic o sho e su i al, as expec ed. These a e common ad e se e ec s in ex e nal adia ion and a e likely o be well eco ded a he oncology clinic, and pa ien s may be less likely o unde - epo hem. This measu emen was, howe e , no cong uen because he iming o he in e iew du ing he adio he apy a ied, which po en ially a ec s he le el o side e ec s. Howe e , he su i al model also held when he o e all QOL index ( o al sco e 1–7) was es ed ins ead o he p os a e-a ea symp oms. These ac o s bo h indica ed he wellbeing/ill-being o he pa ien du ing he p ima y ea - men . Sel -e alua ion o he wellbeing s a us is con inuously p esen ed o be a eliable indica o o an indi idual’s cu en and u u e heal h. The ob ained esul s on p oblems in social unc ioning and hopelessness, and a sho e elapse- ee ime may also ela e o concep s close o non-exp ession o nega i e emo- ions, such as an iemo ionali y (“emo ional de ensi eness”) [17] and Type C esponse s yle [18, 19]. An iemo ionali y (supp ession and con ol o emo ions) is mani es ed in sel - sac i icing beha io , which may e en ually lead o p oblems in social unc ioning. Helplessness/hopelessness in s ess ul si ua ions is a componen o he Type C s yle. We p e iously ound ha an iemo ionali y p edic ed sho e su i al in b eas cance [6] and hopelessness in melanoma [7]. Hope a cance diagnosis is essen ial [59] and i may p edic su i al [1, 11]. Hopelessness has been ound o a ec cance mo al- i y a he popula ion le el [60], especially in olde popula- ions [61], o which mos p os a e cance pa ien s belong. Mo eo e , he obse ed a o able e ec s o pain and num- be o s ess ul li e e en s may be ela ed o non-exis ence o non-exp ession o nega i e emo ions: hose wi hou may ha e epo ed mo e symp oms and e en s. In acco dance wi h ou p e ious s udies [6, 7, 12], we iden i ied only one p o ec i e psychological ac o : he Cog- ni i e A oidance/Denial coping pa e n, and i p edic ed only he disease- ee ime. As no ed abo e, o e all su i al and PSA- elapse- ee ime a e e y di e en ou comes in p os a e cance ; he o me is less a ec ed by he cance because o he good p ognosis, whe eas he la e is cance - speci ic. This esul is in line wi h cumula ing esea ch sug- ges ing ha a a o able p ognosis in cance is p edic ed by esponding o he cance diagnosis by a coping pa e n o denying [21] o , nowadays mo e o en called, ‘minimizing’ he ac o ha ing cance [1, 2, 4–7, 17, 20, 29]. This concep e e s, acco ding o de ini ion o coping, he pa ien s’ e o s o manage demands caused by he cance by minimizing impac o he disease, no he consequen nega i e a ec . Howe e , denying/minimizing may a a la e poin o ime ha e a di e en e ec [59]. Ou esul s p o ide an explana ion o why good sel - epo ed QOL in some p e ious s udies appea ed o ha e had an un a o able impac on cance su i al ou comes. Pa ien epo s o a good wellbeing in simple measu es may e lec non-exp ession o nega i e emo ions. Thus, e y high epo s o wellbeing in newly diagnosed cance pa ien s may indica e a psychological isk ac o o educed su i al ou - comes. I may be good o he pa ien s o ha e complain s a he ime o p ima y ea men . This inding may ha e clinical ele ance gi en ha pa ien s who epo no o ew p oblems (no symp oms, excellen QOL) may ac ually be 685Quali y o Li e Resea ch (2019) 28:677–686 1 3 ulne able and in need o ex a a en ion and ca e. How- e e , mo e esea ch is needed, e.g., o in es iga e he issue in pa ien s wi h di e en cance ypes/di e en diseases, in bo h gende s, in di e en phases o ca e and ehabili a ion, and in p ospec i e designs. Clinicians a e o in e p e sco es o alida ed ques ion- nai es on wellbeing and QOL as usual. A clinician who in e p e s he sco es should know he pa ien ’s si ua ion, e.g., phase o ea men and ime since diagnosis. Conclusions o he pa ien s’ wellbeing should no be based on a single measu emen . Valida ed ques ionnai es a e good ools o sc eening o s a e o QOL, which may equi e u he clini- cal examina ion. Ou c i icism is a ge ed o simple QOL indica o s which a e easily in luenced by o he ac o s, e.g., non-exp ession o nega i e emo ions. Acknowledgemen s Open access unding p o ided by Na ional Ins i- u e o Heal h and Wel a e (THL). This wo k was suppo ed by he Medical Resea ch Fund o Tampe e Uni e si y Hospi al, he Compe i- i e S a e Resea ch Financing o he Expe Responsibili y a ea o Tam- pe e Uni e si y Hospi al, 9U020; he Finnish Cance Founda ion; and he Signe and Ane Gyllenbe g Founda ion. We hank P o esso Teu o Tammela, he head o Depa men o U ology o Tampe e Uni e si y Hospi al, o his help in p epa ing he da a o he su i al analyses. Funding This s udy was unded by he Medical Resea ch Fund o Tampe e Uni e si y Hospi al, he Compe i i e S a e Resea ch Financ- ing o he Expe Responsibili y a ea o Tampe e Uni e si y Hospi al, he Finnish Cance Founda ion, and he Signe and Ane Gyllenbe g Founda ion. Compliance wi h e hical s anda ds Con lic o in e es The au ho s decla e ha hey ha e no con lic o in e es . E hical app o al All p ocedu es pe o med in s udies in ol ing human pa icipan s we e in acco dance wi h he e hical s anda ds o he ins i- u ional and/o na ional esea ch commi ee and wi h he 1964 Helsinki decla a ion and i s la e amendmen s o compa able e hical s anda ds. In o med consen In o med consen was ob ained om all indi idual pa icipan s included in he s udy. Open Access This a icle is dis ibu ed unde he e ms o he C ea- i e Commons A ibu ion 4.0 In e na ional License (h p://c ea i eco mmons .o g/licen ses/by/4.0/), which pe mi s un es ic ed use, dis ibu- ion, and ep oduc ion in any medium, p o ided 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 license, and indica e i changes we e made. Re e ences 1. Wa son, M., Ha iland, J., G ee , S., e al. (1999). In luence o psychosocial esponse on su i al in b eas cance : A popula ion based coho s udy. Lance , 354, 1331–1336. 2. Bu ow, P. N., Coa es, A. S., & Dunn, S. (2000). Psychosocial p e- dic o s o su i al: Me as a ic b eas cance . Annals o Oncology, 11, 469–474. 3. Reynolds, P., Hu ley, S., To es, M., e al. (2000). Use o cop- ing s a egies and b eas cance su i al: Resul s om he Black/ Whi e Cance Su i al S udy. Ame ican Jou nal o Epidemiology, 152, 940–948. 4. Ga ssen, B. (2004). Psychological ac o s and cance de elop- men : E idence a e 30 yea s o esea ch. Clinical Psychology Re iew, 24, 315–338. 5. Bu ow, P. N., Coa es, A. S., & Dunn, S. M. (1999). Psychosocial p edic o s o su i al in me as a ic melanoma. Jou nal o Clinical Oncology, 17, 2256–2263. 6. Leh o, U.-S., Ojanen, M., Dyba, T., e al. (2006). Baseline psy- chosocial p edic o s o su i al in localized b eas cance . B i ish Jou nal o Cance , 94, 1245–1252. 7. Leh o, U.-S., Ojanen, M., Dyba, T., e al. (2007). Baseline psy- chosocial p edic o s o su i al in localized melanoma. J Psycho- soma ics Resea ch, 63, 9–15. 8. Chida, Y., Hame , M., Wa dle, J., e al. (2008). Do s ess- ela ed psychosocial ac o s con ibu e o cance incidence and su i al? Na u e Re iews Clinical Oncology, 5(8), 466–475. 9. B own, J. E., Bu ow, P. N., Culjak, G., e al. (2000). Psychosocial p edic o s o ou come: ime o elapse and su i al in pa ien s wi h ea ly s age melanoma. B i ish Jou nal o Cance , 83, 1448–1453. 10. Sa in, J., Linden, W., & Phillips, M. (2009). Dep ession as a p e- dic o o disease p og ession and mo ali y in cance pa ien s. A me a-analysis. Cance , 115, 5349–5361. 11. Pinqua , M., & Dube s ein, P. (2010). Dep ession and cance mo ali y: A me a-analysis. Psychological Medicine, 40(11), 1797–1810. 12. Leh o, U.-S., Ojanen, M., Dyba, T., e al. (2012). Impac o li e e en s on su i al o pa ien s wi h localized melanoma. Psycho- he apy Psychosoma ics, 81, 191–193. 13. Ba aclough, J., Pinde , P., C uddas, M., e al. (1992). Li e e en s and b eas cance p ognosis. BMJ, 304, 1078–1081. 14. Tie sma, E. S. M., an de Lee, M. L., Ga ssen, B., e al. (2005). Psychosocial ac o s and he cou se o ce ical in a-epi helial neoplasia: A p ospec i e s udy. Gynecol Oncol, 97(3), 879–886. 15. Woods, L. M., Rache , B., & Coleman, M. P. (2006). O igins o socio-economic inequali ies in cance su i al: A e iew. Annals o Oncology, 17, 5–19. 16. K is enson, M., E iksen, H. R., Slui e , J. K., e al. (2004). Psy- chobiological mechanisms o socioeconomic di e ences in heal h. Social Science & Medicine, 58, 1511–1522. 17. Ga ssen, B., & Remie, M. (2004). Di e en concep s o di e en wo ds? Concep s ela ed o non-exp ession o nega i e emo ions. In In I. Nyclicek, L. Temoshok, & A. Vinge hoe s (Eds.) Emo- ional exp ession and heal h. (pp.117–136) Ho e and New Yo k: B unne -Rou ledge. 18. Temoshok, L. (1987). Pe sonali y, coping s yle, emo ion and can- ce : Towa ds an in eg a i e model. Cance Su ial, 6, 545–567. 19. G oss, J. (1989). Emo ional exp ession in cance onse and p o- g ession. Social Science & Medicine, 28, 1239–1248. 20. Ga ssen, B., & Goodkin, K. (1999). On he ole o immunologi- cal ac o s as media o s be ween psychosocial ac o s and cance p og ession. Psychia y Resea ch, 85, 51–61. 21. G ee , S., Mo is, T., & Pe ingale, K. W. (1979). Psychosocial esponse o b eas cance ; e ec on ou come. Lance , ii, 785–787. 22. Nausheen, B., Gid on, Y., Pe ele , R., e al. (2009). Social suppo and cance p og ession: A sys ema ic e iew. J Psychosoma ic Resea ch, 67, 403–415. 23. Cousson-Gélie, F., B uchon-Schwei ze , M., Dilhuydy, J., e al. (2007). Do anxie y, body image, social suppo and coping