Mortality of young offenders: a national register-based follow-up study of 15-to 19-year-old Finnish delinquents referred for forensic psychiatric examination between 1980 and 2010

Lindbe g e al. Child Adolesc Psychia y Men Heal h (2017) 11:37 DOI 10.1186/s13034-017-0174-3 RESEARCH ARTICLE Mo ali y o young o ende s: a na ional egis e -based ollow-up s udy o 15- o19-yea -old Finnish delinquen s e e ed o  o ensic psychia ic examina ion be ween1980 and2010 Nina Lindbe g1*, Jouko Mie unen2,3, Anni Heiskala2 and Rii ake u Kal iala‑Heino4,5,6 Abs ac Backg ound: The mo ali y a e o young o ende s is high. Fu he mo e, mo ali y in young o ende s is associa ed wi h psychia ic and subs ance use diso de s. The p ima y aim o his na ional egis e ‑based ollow‑up s udy was o in es iga e he mo ali y a e o Finnish delinquen s who unde wen a o ensic psychia ic examina ion be ween 1980 and 2010. As delinquency is no a solid en i y, we u he aimed o compa e he isk o p ema u e dea h among di ‑ e en subg oups o he delinquen s; iolen e sus non‑ iolen o ende s, o ende s wi h alcohol use diso de s e sus hose wi h no such diagnoses, o ende s wi h schizoph enia spec um diso de s e sus conduc ‑ and pe sonali y‑ diso de ed o ende s, unde ‑aged e sus young adul o ende s, and, inally, boys e sus gi ls. Me hods: We collec ed he o ensic psychia ic examina ion epo s o all 15‑ o 19‑yea ‑old o ende s who we e bo n in Finland and had unde gone he examina ion be ween 1.1.1980 and 31.12.2010 (n = 606) om he a chi es o he Na ional Ins i u e o Heal h and Wel a e and e ospec i ely e iewed hem. Fo each delinquen , ou age‑, gen‑ de ‑ and place o bi h‑ma ched con ols we e andomly selec ed om he Cen al Popula ion Regis e (n = 2424). The delinquen s and hei con ols we e ollowed un il he end o 2015. The median ollow‑up ime was 23.9 yea s (in e qua ile ange 15.3–29.5). We ob ained he mo ali y da a om he causes o dea h egis e . Dea hs a ibu able o a disease o an occupa ional disease we e conside ed na u al, and hose a ibu able o an acciden , suicide o homicide we e conside ed unna u al. Resul s: By he end o he ollow‑up pe iod, 22.1% (n = 134) o he delinquen s and 3.4% (n = 82) o hei con ols had died (OR 8.11, 95% CI 6.05–10.86, p < 0.001). Among boys, 22.0% (n = 121) o he delinquen s and 3.7% (n = 81) o he con ols had died (OR 7.38, 95% CI 5.46–9.95, p < 0.001). Male delinquen s’ isk o unna u al dea h was almos 11‑ old, o na u al dea h mo e han wo old, and o unclea dea h mo e han ou old compa ed o ha o hei con ols. No gi ls had na u al o unclea dea hs, bu 23.6% (n = 13) o he delinquen s and 0.5% (n = 1) o he con ols had died due o unna u al causes (OR 67.79, 95% CI 8.63–532.00, p < 0.001). The iolen delinquen s’ isk o p ema u e dea h was wice ha o he non‑ iolen delinquen s. The o he compa isons demons a ed no s a is ically signi ican di e ences be ween subg oups. Conclusions: E en hough he Finnish co ec ion sys em p e e s psychia ic ea men and ehabili a ion o e c imi‑ nal sanc ions, and he na ional heal h ca e sys em o e s de elopmen al‑phase‑speci ic psychia ic ca e, he mo ali y © The Au ho (s) 2017. 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 ecommons.o g/licenses/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. The C ea i e Commons Public Domain Dedica ion wai e (h p://c ea i ecommons.o g/ publicdomain/ze o/1.0/) applies o he da a made a ailable in his a icle, unless o he wise s a ed. Open Access Child and Adolescen Psychia y and Men al Heal h *Co espondence: nina.lindbe g@hus. i 1 Fo ensic Psychia y, Helsinki Uni e si y and Helsinki Uni e si y Hospi al, PO Box 590, 00029 HUS Helsinki, Finland Full lis o au ho in o ma ion is a ailable a he end o he a icle Page 2 o 9 Lindbe g e al. Child Adolesc Psychia y Men Heal h (2017) 11:37 Backg ound The p ema u e mo ali y o young o ende s is high. Fo example, a ollow-up s udy by Co ey e al. [1] epo ed an o e all s anda dized mo ali y a io (SMR) o 9.4 [95% con idence in e al (CI) 7.4–11.9] o male and 41.3 (95% CI 20.2–84.7) o emale delinquen s aged 10–20. The link be ween c iminal beha io and excess p ema- u e mo ali y emains h oughou li e: a me a-analysis by Zlod e and Fazel [2] showed SMR anges be ween 1.0 and 9.4 o men and be ween 2.6 and 41.3 o women ollowing elease om p ison. P ema u e dea h is o en a ibu able o acciden s, suicides, homicides [3–5], in ec ions [4], and subs ance abuse [4], and is commonly associa ed wi h iolence [3–5]. Acco ding o a 50-yea ollow-up s udy by Laub and Vaillan [4], along wi h aging, he a o emen ioned causes o dea h a e accompa- nied by medical condi ions and diseases such as cance s and ca dio ascula diseases. The mo ali y a e a ies depending on he se e i y o jus ice sys em in ol emen : a es ed you h has he lowes a e o mo ali y, ollowed by de ained and inca ce a ed you h [6]. Mo eo e , high delinquency se e i y shows ea lie and highe mo ali y isk [7]. Ju enile delinquency is s ongly linked o psychia ic mo bidi y. Be ween abou hal and wo hi ds o young o ende s mee he diagnos ic c i e ia o a conduc dis- o de , and be ween abou a hi d and a hal o a sub- s ance-use diso de [8–10]. Mo eo e , bo h bo de line [11, 12] and an isocial [13] pe sonali y diso de s associ- a e highly wi h c iminali y. Findings on he p e alence o psychoses seem con adic o y, bu acco ding o a me a-analysis o 25 su eys ocusing on de ained ado- lescen s, i seems o be abou h ee pe cen [10]. People wi h men al heal h diso de s show excessi e mo ali y [14]. Se e e conduc -diso de ed beha io is associa ed wi h an inc eased isk o p ema u e dea h, mos ly due o acciden s, suicide and homicide [15]. Recen ly, a commu- ni y-based ollow-up s udy highligh ed he associa ion be ween adolescen conduc p oblems and p ema u e dea h a ibu able o ca dio ascula diseases and cance s [16]. Bo h bo de line and an isocial pe sonali y diso de s ela e s ongly o impulsi e and sel -dis up i e beha io [17–20], leading o an ele a ed isk o unna u al dea h. An isocial pe sonali y diso de is o en linked o Clon- inge ’s ype 2 alcoholism [21] leading o high mo ali y om alcohol- ela ed diso de s [22]. Acco ding o a s udy by Hoye e al. [23], men wi h pe sonali y diso de s also su e highe mo ali y due o soma ic diso de s o he han hose ela ed o alcohol han he gene al popula ion, mos p obably due o unde epo ing o soma ic symp- oms and less help-seeking. Wi h ega d o people wi h schizoph enia spec um diso de s, coho s udies ha e e ealed ha he mo ali y a e o diseases and medical condi ions is mo e han double han ha o he gene al popula ion [24, 25]. Among delinquen s, hospi aliza ion o a psychia ic diso de o subs ance abuse associa es signi ican ly wi h he isk o p ema u e dea h [3]. The numbe o ju enile p isone s in Finland has a- di ionally been low by in e na ional s anda ds because o he widesp ead use o condi ional p ison sen ences and special you h punishmen s, including social eha- bili a ion and occupa ional ac i i ies [26]. Unde -aged delinquen s wi h se e e o ending a e ypically e e ed o child-wel a e ins i u ions, which ocus on ehabili a- ion a he han co ec ion. Fo example, he e we e only 11 unde -aged indi iduals in Finnish p isons in 2015, (C iminal Sanc ions Agency, pe sonal communica ion). Fu he , acco ding o Finnish law, cou s can send an indi idual o a p e ial o ensic psychia ic examina ion o de e mine whe he he/she is su e ing om a men- al diso de ha migh a ec his/he c iminal esponsi- bili y. The inal o ensic psychia ic epo includes an opinion on he o ende ’s le el o c iminal esponsibili y, as well as possible psychia ic diagnoses and an assess- men as o whe he o no he o ende ul ils he c i e ia o in olun a y psychia ic ca e. Fo example, o end- e s wi h schizoph enia spec um diso de s a e ypically e e ed o in olun a y psychia ic ea men . Finland was he i s coun y in Eu ope o acknowledge adoles- cen psychia y as an independen special y in he ield o medicine wi h i s own se ice egime: I was a dis inc subspecial y in psychia y be ween 1979 and 1994, un il i became an independen medical special y. Since hen, bo h educa ion and esea ch on adolescen men al heal h diso de s and hei ea men ha e gained s eng h in Finland. In his na ional egis e -based ollow-up s udy, we col- lec ed a consecu i e sample o Finnish delinquen s who we e sen o a p e ial o ensic psychia ic examina ion be ween 1980 and 2010, and ollowed hem un il he end o 2015. Ou p ima y aim was o in es iga e hei mo - ali y a e. We assumed ha due o he abo e-men ioned a e o delinquen s, especially o hose wi h a his o y o iolen o ences, is high. The excess mo ali y o o ende s can be ega ded as a speci ic public‑heal h inequi y ha calls o mo e e ec i e in e en ion p ocedu es han hose used hus a . Keywo ds: Dea h, Delinquency, Mo ali y, O ending, Psychopa hology Page 3 o 9 Lindbe g e al. Child Adolesc Psychia y Men Heal h (2017) 11:37 special ea u es ela ed o Finnish judicial p ac ices and men al heal h sys em, we would obse e lowe mo ali y a es han hose epo ed by Co ey e al. [1]. Secondly, as delinquency is no a solid en i y, we wan ed o compa e he isk o p ema u e dea h among di e en subg oups o delinquen s; iolen e sus non- iolen o ende s, o end- e s wi h alcohol use diso de s e sus hose wi h no such diagnoses, o ende s wi h schizoph enia spec um diso - de s e sus conduc - and pe sonali y-diso de ed o end- e s, unde -aged e sus young adul o ende s, and inally, boys e sus gi ls. On he basis o ea lie esea ch indings, we hypo hesized ha delinquen s wi h iolen o end- ing would show signi ican ly highe p ema u e mo ali y han hose wi h non- iolen o ending, and ha delin- quen s wi h ea ly onse alcohol abuse/dependence would ha e a signi ican ly highe p ema u e mo ali y a e han hose wi hou hese diso de s. We also assumed ha , as delinquen s wi h schizoph enia spec um diso de s a e e e ed o in olun a y psychia ic inpa ien ca e, hei mo ali y a e would be signi ican ly lowe han ha o delinquen s wi h conduc and pe sonali y diso de s. We u he assumed ha , as unde -aged delinquen s wi h se e e o ending a e e e ed o child-wel a e ins i u ions, hei mo ali y a e would be signi ican ly lowe han ha o hei adul coun e pa s. The p ema u e mo ali y a e o Finnish males is known o be app oxima ely wice ha o emales [27]. We hypo hesized ha we would obse e he same kind o gende di e ence in ou o ende popula ion. Sample andme hods Sample andp ocedu e The minimum age o c iminal liabili y in Finland is 15 yea s. The Na ional Ins i u e o Heal h and Wel- a e o ganizes o ensic psychia ic examina ions. These examina ions a e inpa ien e alua ions ha las app oxi- ma ely 2mon hs, and include da a ga he ed om a i- ous sou ces ( he ju enile him/he sel , amily membe s, ela i es; and medical, c iminal, school, child-wel a e and mili a y eco ds), psychia ic e alua ions, s anda dized psychological es s, in e iews conduc ed by a mul i- p o essional eam, an e alua ion o he o ende ’s physi- cal condi ion, and con inuous obse a ion by he hospi al s a . The o e all high quali y and eliabili y o Finnish o ensic psychia ic examina ions a e acknowledged in he cou s and among scien is s [28]. The In e na ional Classi ica ion o Diseases—Eigh h e ision (ICD-8) [29] was used o he pu poses o psychia ic classi ica- ion in Finnish clinical p ac ice be ween 1968 and 1986. I was eplaced by he Diagnos ic and S a is ical Manual o Men al Diso de s—Thi d e ised edi ion (DSM-III-R) [30] which was used be ween 1987 and 1995. Howe e , he diagnoses we e con e ed in o ICD-9 [31] diagnoses, e.g. when epo ing hem o he ca e egis e o heal h ca e. The ICD-10 [32] has been used since 1996. We collec ed psychia ic examina ion epo s on all 15- o 19-yea -old o ende s who we e bo n in Finland and unde wen he examina ion be ween 1.1.1980 and 31.12.2010 om he a chi es o he Na ional Ins i u e o Heal h and Wel a e. The E hics Commi ee o Hel- sinki Uni e si y Hospi al and he pe inen ins i u ional au ho i ies app o ed he s udy p o ocol. Pa icipan s The s udy pa icipan s comp ised 606 delinquen s (gi ls: 9.1%; n=55) o Finnish o igin. The c iminal his o ies o he delinquen s we e e iewed, and, in line wi h Fazel and G ann [33], mu de , a emp ed mu de , manslaugh- e , a emp ed manslaugh e , a son, agg a a ed assaul , assaul , obbe y, kidnapping, ape, a emp ed ape, and child moles a ion we e all ega ded as iolen o ences; and he , bu gla y, d i ing while in oxica ed, dange - ous d i ing, d i ing wi hou a d i ing license, as well as d ug smuggling and selling d ugs we e ega ded as non- iolen o ences. I he pe son had commi ed mo e han one o ence, we ook he mos se e e one. O all delin- quen s, 83.5% (n=506) had o ended iolen ly and 16.5% (n=100) non- iolen ly. Fu he , 42.7% (n=259) o he delinquen s had commi ed hei index o ence as mino s. Acco ding o he p incipal psychia ic diagnoses gi en in he o ensic psychia ic examina ion, 1.3% (n=8) o he delinquen s we e men ally disabled (IQ<70), 9.4% (n = 57) we e diagnosed wi h a schizoph enia- ela ed diso de , and 79.2% (n=479) had a conduc diso de o a pe sonali y diso de (mainly an isocial, bo de line o mixed). Fu he , 7.6% (n=46) o he delinquen s had some o he psychia ic diso de (mainly dep essi e and anxie y diso de s) and 2.6% (n=16) o hem showed no psychia ic diso de . O he 606 delinquen s, 45.4% (n=275) we e diagnosed wi h como bid alcohol abuse o dependence. Con ols Fo each delinquen , ou con ols we e andomly selec ed om he Cen al Popula ion Regis e and ma ched o age, gende and place o bi h. The e we e no exclusion c i e ia o his gene al popula ion da a. Follow‑up The ollow-up pe iod s a ed when he delinquen ’s o ensic psychia ic examina ion was comple e (= he index day). The delinquen s and hei con ols we e ol- lowed un il 31.12.2015. The median ollow-up ime was 23.9yea s (in e qua ile ange=IQR 15.3–29.5). Page 4 o 9 Lindbe g e al. Child Adolesc Psychia y Men Heal h (2017) 11:37 Mo ali y The pe sonal iden i ica ion numbe ha is assigned o all esiden s o Finland by he Cen al Popula ion Regis e was used o link he da a wi h S a is ics Finland’s cause o dea h egis e . Dea h ce i ica es con ain in o ma ion on he p ima y cause o dea h, which is always epo ed, and in some cases in o ma ion on seconda y causes. The class o dea h is epo ed as ollows: a ibu able o a dis- ease, an occupa ional disease, an acciden , medical ca e, suicide, homicide, wa , o a eason ha emains unclea . In his s udy, dea hs a ibu able o a disease o an occu- pa ional disease we e conside ed na u al, and hose a ibu able o an acciden , suicide o homicide we e conside ed unna u al. I was assumed ha no dea hs a ibu able o a wa would be epo ed in his sample, gi en ha he Finnish–So ie Con inua ion Wa , which was he las wa in which Finland was engaged, ended in he au umn o 1944. All dea h ce i ica es a e e i ied by a physician a he egional le el, and by medical expe s a S a is ics Finland. The egis e includes he dea hs o all ci izens and pe manen esiden s o Finland, as well as hose o Finnish ci izens ab oad. Du ing he ollow- up pe iod, 53 people (15 delinquen s and 38 con ols) mo ed ab oad—mos commonly o o he No dic coun- ies. They we e included in he analysis on he assump- ion ha he S a is ics Finland is in o med o dea hs ha occu ab oad. S a is ics We used he likelihood a io Chi squa e- es o com- pa e he delinquen s and hei gende -, age- and place o bi h-ma ched con ols. Su i al was analyzed by means o Kaplan–Meie su i al plo s, and g oup compa isons we e made using he log ank es : Findings we e consid- e ed signi ican when wo- ailed p<0.05. The odds a io (OR) and he haza d a io (HR) wi h 95% CIs we e used as measu es o he e ec size. R e sion 3.3.1 (h p:// www.R-p ojec .o g) was used o he s a is ical analyses. Resul s Table1 p esen s he mo ali y o delinquen s and hei age-, gende -, and place o bi h-ma ched con ols. Wi h ega d o ORs, delinquen s’ isk o all-cause mo ali y was mo e han eigh imes ha o hei con ols. Male delinquen s’ isk o all-cause mo ali y was mo e han se en imes ha o male con ols. Among gi ls, delinquen s’ isk o all-cause mo ali y was mo e han 67 imes ha o hei con ols. Fig- u e1 p esen s he cumula i e su i al o delinquen s and hei con ols du ing he 36-yea ollow-up pe iod. Figu e2 ocuses on he cumula i e su i al o boys, and Fig.3 on ha o gi ls. Wi h ega d o di e en classes o dea h, male delinquen s’ isk o unna u al dea h was almos 11- old, isk o na u al dea h mo e han wo old, and isk o unclea dea h mo e han ou old compa ed o ha o male con- ols. Among gi ls, he e we e no na u al o unclea dea hs, Table 1 Mo ali y o delinquen s and ha o  hei age-, gende -, andplace o bi h-ma ched con ols du ing he ollow- up OR odds a io, CI con idence in e al, Na u al dea hs dea hs a ibu able o a disease o an occupa ional disease, Unna u al dea hs acciden s, suicides, homicides Delinquen s Con ols OR CI 95% p n=606 n=2424 All Dea hs o al; n (%) 134 (22.1) 82 (3.4) 8.11 6.05–10.86 <0.001 Na u al 23 (3.8) 36 (1.5) 2.62 1.54–4.45 <0.001 Unna u al 107 (17.6) 42 (1.7) 12.16 8.40–17.60 <0.001 Unclea 4 (0.7) 4 (0.2) 4.02 1.00–16.12 0.03 n=551 n=2204 Boys Dea hs o al 121 (22.0) 81 (3.7) 7.38 5.46–9.95 <0.001 Na u al 23 (4.2) 36 (1.6) 2.62 1.54–4.46 <0.001 Unna u al 94 (17.1) 41 (1.9) 10.85 7.42–15.87 <0.001 Unclea 4 (0.7) 4 (0.2) 4.02 1.00–16.13 0.03 n=55 n=220 Gi ls Dea hs o al 13 (23.6) 1 (0.5) 67.79 8.63–532.00 <0.001 Na u al 0 (0.0) 0 (0.0) – – – Unna u al 13 (23.6) 1 (0.5) 67.79 8.63–532.00 <0.001 Unclea 0 (0.0) 0 (0.0) – – – Page 5 o 9 Lindbe g e al. Child Adolesc Psychia y Men Heal h (2017) 11:37 bu he delinquen s’ isk o unna u al dea h was mo e han 67- old compa ed o ha o hei con ols. O he 94 unna u al dea hs among he male delin- quen s, 45.7% (n=43) we e classi ied as suicides, 20.2% (n=19) as homicides, and 34.0% (n=32) as conse- quences o a ious acciden s. Among he emale delin- quen s, 69.2% (n=9) o hei 13 dea hs we e a ibu able o suicide, 15.4% (n=2) o homicide, and 15.4% (n=2) o an acciden . The causes o he 36 na u al dea hs among delinquen males included ci hosis o he li e , gas ic bleeding, pneumonia, ca diac disease, ce eb al hemo - hage, and malignancies. Table2 p esen s he compa isons be ween he di e en subg oups o delinquen s. The di e ences be ween he Fig. 1 The cumula i e su i al o 606 delinquen s and 2424 con ol subjec s du ing he 36‑yea ollow‑up pe iod Fig. 2 The cumula i e su i al o 551 male delinquen s and 2204 con ol subjec s du ing he 36‑yea ollow‑up pe iod Fig. 3 The cumula i e su i al o 55 emale delinquen s and 220 con ol subjec s du ing he 36‑yea ollow‑up pe iod Table 2 Compa isons be weendi e en subg oups o 606 delinquen s HR haza d a io, CI con idence in e al, CD conduc diso de , PD pe sonali y diso de Dea hs n (%) HR CI 95% p Gende Boys 121/551 (22.0) 0.85 0.48–1.51 NS Gi ls 13/55 (23.6) Re Age Mino s 50/259 (19.3) 0.76 0.54–1.08 NS Young adul s 84/347 (24.2) Re C iminal his o y Violen 121/506 (23.9) 2.01 1.14–3.57 0.009 Non‑ iolen 13/100 (13.0) Re Diagnosis Schizoph enia spec um diso de s 8/57 (14.0) 0.62 0.31–1.28 NS CD/PD 116/479 (24.2) Re Como bid alcohol abuse/dependence Yes 59/275 (21.5) 1.10 0.78–1.55 NS No 75/331 (22.7) Re Page 6 o 9 Lindbe g e al. Child Adolesc Psychia y Men Heal h (2017) 11:37 isk o p ema u e dea h among delinquen s wi h alcohol use diso de s and hose wi hou hese diso de s, among delinquen s wi h schizoph enia spec um diso de s and hose wi h conduc /pe sonali y diso de s, among unde - aged and young adul o ende s, o among boys and gi ls we e no s a is ically signi ican . Howe e , when iolen delinquen s we e compa ed o non- iolen ones, hei isk o p ema u e dea h was wo old. Discussion Ou sample appea s o be highly dis inc om hose p e- sen ed in ea lie s udies [8–10]: <3% o delinquen s we e no gi en a psychia ic diagnosis in he o ensic psychi- a ic examina ion, and psycho ic diso de s we e app oxi- ma ely h ee imes as common as in a la ge me a-analysis by Fazel e al. [10]. Ne e heless, as in p e ious delin- quency s udies [8, 9], conduc and pe sonali y diso de s u ned ou o be he mos common psychia ic diagnoses. In line wi h he esul s o ea lie mo ali y s udies [1, 4], ou delinquen s aced a high mo ali y isk, and he inci- dence o unna u al causes o dea h was p onounced. The isk o p ema u e dea h was compa able o ha epo ed by Co ey e al. [1] and by Zlod e and Fazel [2], bu was e en highe among emale delinquen s. The numbe o emale delinquen s was small (<10% o he o al sample), bu all dea hs among gi ls we e a ibu ed o unna u al causes. Delinquen s wi h a his o y o iolen o ending aced a wo old isk o p ema u e dea h compa ed o delin- quen s wi h no such his o y. This inding is in line wi h he esul s o a p e ious ollow-up s udy by T umbe a e al. [7] among 14- o 15-yea -old nin h g ade s, which indica ed ha he se e i y o delinquency is ela ed o he isk o p ema u e dea h. Mo eo e , a ollow-up s udy among almos 50,000 young men ec ui ed om he mili a y se ice ound ha iolen o ending is associ- a ed wi h excess mo ali y and he isk o dying om an alcohol o d ug- ela ed cause o suicide [34]. The highes p opo ion o cases o dea h was obse ed among men who had commi ed a leas one iolen c ime bu we e also ic ims o iolence hemsel es [34]. In empe amen s udies, iolen ly beha ing o ende s ha e sco ed high on no el y-seeking, impulsi eness and i i abili y [35]. These ai s make hem p one o acciden s, and hei isk o becoming ic ims o iolence is high, which bo h lead o an inc eased mo ali y isk [36]. Schizoph enia [37], conduc diso de [15, 16] as well as an isocial [38] and bo de line [39] pe sonali y diso de s a e all associa ed wi h excess mo ali y. Con a y o ou hypo hesis, we we e unable o ind a subs an ial di e - ence be ween he isk o p ema u e dea h among delin- quen s diagnosed wi h schizoph enia spec um diso de s and among hose diagnosed wi h conduc o pe sonali y diso de s. This inding is in e es ing, gi en ha he Finn- ish co ec ion sys em singles ou hose who su e om a psycho ic diso de and e e s hem o psychia ic hospi- als, whe e he ea men pe iods equen ly las o se - e al yea s. This was also he case among ou delinquen s wi h schizoph enia spec um diso de s: hey we e all e alua ed as lacking c iminal esponsibili y and sen o in olun a y psychia ic inpa ien ea men . This p oce- du e did no seem o help in e ms o educing hei p e- ma u e mo ali y, howe e , when hey we e compa ed o an isocially beha ing delinquen s who we e no simila ly ea ed. Indeed, low socioeconomic s a us, un a o able heal h beha io s such as smoking and d ug and alcohol abuse, as well as sel -des uc i e beha io s and suicide endencies a e all linked o o ende popula ions in gen- e al, ega dless o indi idual psychia ic diagnoses. Some au ho s ha e iden i ied an ipsycho ic medica ion as a cause o excess mo ali y [40], bu o he s ha e epo ed he opposi e inding [41]. An isocial pe sonali y diso de is linked o Cloninge ’s ype 2 alcoholism, which is cha ac e ized by s ong physical dependence ha ypically al eady de elops in adolescence [21]. Almos hal o ou delinquen s we e diagnosed wi h alcohol abuse o dependence, which mos p obably mi o s his well-known ela ionship. Con a y o ou hypo hesis, we obse ed no signi ican di e ence be ween he all-cause mo ali y o hose who we e diag- nosed wi h alcohol use diso de s and hose who we e no . Un o una ely, we had no in o ma ion on o ende s’ la e alcohol use, bu we assume ha many o hose wi h no alcohol use diso de a he ime o he o ensic psy- chia ic examina ion migh ha e been diagnosed acco d- ingly la e in li e, which migh explain ou inding. The majo i y o Finnish unde -aged delinquen s wi h se e e o ending a e e e ed o child-wel a e ins i u ions such as esiden ial schools, which ocus on ehabili a- ion a he han punishmen . Educa ion, o example, has a high p io i y in hese ins i u ions. The esiden ial school placemen ends a he age o 18, when he adoles- cen s a e o e ed a olun a y 5-yea a e -ca e p og am [42]. Con a y o ou hypo hesis, he isk o p ema u e dea h did no signi ican ly di e be ween delinquen s who had commi ed hei index c imes unde -aged and hose who had commi ed hem as young adul s (aged 18 o 19). Howe e , many o ou young adul o ende s had also been clien s o child-wel a e du ing hei childhood o adolescence; hence we canno d aw any inal conclu- sions ega ding he ole o child-wel a e in educing la e mo ali y isk. Howe e , Manninen e  al. [43] ecen ly epo ed ha Finnish adolescen s placed in esiden ial schools show a subs an ially ele a ed mo ali y isk in young adul hood compa ed o ha o he gene al popu- la ion. The au ho s concluded ha o educe mo ali y, Page 7 o 9 Lindbe g e al. Child Adolesc Psychia y Men Heal h (2017) 11:37 mo e e ec i e ea men o men al and subs ance-use- ela ed p oblems du ing and a e he esiden ial school placemen is needed. Con a y o ou hypo hesis, he e was no subs an ial gende di e ence be ween he all-cause mo ali y o male and emale delinquen s. We ha e p e iously shown ha when male and emale delinquen s ma ched o age and ype o c ime a e compa ed o each o he , he emales esemble hei male coun e pa s in many psychoso- cial ad e si ies [44]. The p esen esul adds o his wi h ega d o mo ali y. S eng hs andlimi a ions The main s eng h o his s udy lies in i s na ionwide and comp ehensi e na u e. The Finnish adi ion o ho ough o ensic psychia ic examina ion and eliable s a is ics con- s i u e a solid basis o egis e -based s udies. The clea - ance a e o iolen c imes in Finland is high [45]. The p opo ion o emales was low, bu co esponds qui e well o he p opo ion o emales who commi iolen c imes, as epo ed by na ional s a is ics on police-in es iga ed c ime and sel - epo ed ju enile delinquency [46]. The es ima e o emale delinquen s’ mo ali y isk mus be ega ded as un eliable, as he e was only one dea h among he emale con ols. In addi ion, some 95% CIs we e qui e la ge. The median ollow-up pe iod was subs an ial, almos 24yea s. As in many egis e -based ollow-up s udies, he con ols we e ma ched acco ding o age, gende and place o bi h, bu no o socio-economic s a us o auma his o y, bo h o which a e known o inc ease mo ali y isk [47–49]. I should also be bo ne in mind ha he cou s decide who is sen o a p e ial o ensic psychia ic examina ion, and ha hese decisions could be desc ibed as somewha a bi- a y. We know ha he majo i y o homicidal adolescen s unde go his examina ion [50]. Wi h ega d o o he c ime ypes, an adolescen is ypically sen o an examina ion, i he cou expec s ha he/she migh be su e ing om a men al heal h diso de , bu his decision is o en decided wi hou he bene i o medical expe ise. E en hough we collec ed delinquen s’ p ima y psychia ic diagnoses as well as alcohol use diso de diagnoses, we did no assess he se e i y o psychopa hological symp oms. Ou sample comp ised only 16 delinquen s wi h no psychia ic diso - de ; hus, we did no compa e psychia ic and non-psychi- a ic delinquen s. To shed ligh on his issue, new s udies wi h su icien ly la ge sample sizes a e needed. All he p e- ial o ende s we e Finns by o igin, hus he esul s canno be gene alized o immig an s. Conclusions E en hough he Finnish co ec ion sys em p e e s psy- chia ic ea men and ehabili a ion o e c iminal sanc ions, and he na ional heal h ca e sys em o e s de elopmen al-phase-speci ic psychia ic ca e, he mo - ali y a e o delinquen s, especially o hose wi h a his- o y o iolen o ending, is high. The excess mo ali y o o ende s can be ega ded as a speci ic public-heal h inequi y ha calls o mo e e ec i e in e en ion p oce- du es han hose used hus a . Abb e ia ions CI: con idence in e al; DSM: he Diagnos ic and S a is ical Manual o Men al Diso de s; HR: haza d a io; ICD: In e na ional Classi ica ion o Diseases; OR: odds a io. Au ho s’ con ibu ions NL collec ed he da a and wo ked as he i s au ho , AH and JM conduc ed he s a is ical analyses, RK‑H planned he s udy p o ocol wi h NL, and all he au ho s pa icipa ed in he w i ing p ocess. All au ho s ead and app o ed he inal manusc ip . Au ho de ails 1 Fo ensic Psychia y, Helsinki Uni e si y and Helsinki Uni e si y Hospi al, PO Box 590, 00029 HUS Helsinki, Finland. 2 Cen e o Li e Cou se Heal h Resea ch, Uni e si y o Oulu, Oulu, Finland. 3 Medical Resea ch Cen e Oulu, Oulu Uni e ‑ si y Hospi al and Uni e si y o Oulu, Oulu, Finland. 4 School o Medicine, Tampe e Uni e si y, Tampe e, Finland. 5 Depa men o Adolescen Psychia y, Tampe e Uni e si y Hospi al, Tampe e, Finland. 6 Vanha Vaasa Hospi al, Vaasa, Finland. Acknowledgemen s We would like o acknowledge he au ho i ies a he Depa men o Psychia‑ y, Helsinki Uni e si y Hospi al, and Kellokoski Hospi al. Compe ing in e es s The au ho s decla e ha hey ha e no compe ing in e es s. A ailabili y o da a and ma e ials The da a con ibu ing o hese analyses a e s o ed in a secu e da abase a Helsinki Uni e si y, Depa men o Psychia y, in acco dance wi h Eu opean da a‑p o ec ion legisla ion. Resea che s and clinicians seeking access o hese da a o academic non‑comme cial pu poses a e welcome o submi a eques o he co esponding au ho (NL). All such eques s will be g an ed whene e possible. E hics app o al and consen o pa icipa e The E hics Commi ee o he Helsinki and Uusimaa Hospi al Dis ic , Finland e alua ed he s udy plan. Pe mission o conduc he s udy was g an ed by he adminis a ion o he Helsinki and Uusimaa Hospi al Dis ic , Finland, and he Na ional Ins i u e o Heal h and Wel a e, Finland. The s udy was ca ied ou in acco dance wi h he Decla a ion o Helsinki. The p esen s udy is egis e ‑ based and he pa icipan s we e no con ac ed. Funding This s udy was unded by he Helsinki and Uusimaa Hospi al Dis ic . Howe e , he unde had no ole in he s udy design, he da a collec ion and analysis, he decision o publish, o in he p epa a ion o he manusc ip . Publishe ’s No e Sp inge Na u e emains neu al wi h ega d o ju isdic ional claims in pub‑ lished maps and ins i u ional a ilia ions. Recei ed: 27 Decembe 2016 Accep ed: 27 June 2017 Re e ences 1. Co ey C, Vei F, Wol e R, Cini E, Pa on GC. Mo ali y in young o ende s: e ospec i e coho s udy. BMJ. 2003;326:1064. Page 8 o 9 Lindbe g e al. Child Adolesc Psychia y Men Heal h (2017) 11:37 2. Zlod e J, Fazel S. All‑cause and ex e nal mo ali y in eleased p isone s: sys ema ic e iew and me a‑analysis. Am J Public Heal h. 2012;102:67–75. 3. Sailas ES, Feodo o B, Lindbe g NC, Vi kkunen ME, Sund R, Wahlbeck K. The mo ali y o young o ende s sen enced o p ison and i s associa‑ ion wi h psychia ic diso de s: a egis e s udy. Eu J Public Heal h. 2005;16:193–7. 4. Laub JH, Vaillan GE. Delinquency and mo ali y: a 50‑yea ollow‑up s udy o 1000 delinquen and nondelinquen boys. Am J Psychia y. 2000;157:96–102. 5. Ezell ME, Tanne ‑Smi h EE. Examining he ole o li es yle and c iminal his o y a iables on he isk o homicide ic imiza ion. Homicide S ud. 2009;13:144–73. 6. Aalsma MC, Lau KSL, Pe kins AJ, Schwa z K, Tu W, Wiehe SE, Monahan P. Mo ali y o you h o ende s along a con inuum o jus ice sys em in ol e‑ men . Am J P e Med. 2016;50:303–10. 7. T umbe a SL, Sel ze BK, Go esman II, McIn y e KM. Mo ali y p edic‑ o s in a 60‑yea ollow‑up o adolescen males: explo ing delinquency, socioeconomic s a us, IQ, high‑school d op‑ou s a us, and pe sonali y. Psychosom Med. 2010;72:46–52. 8. Ve mei en R, Jespe s I, Mo i T. Men al heal h p oblems in ju enile jus ice popula ions. Child Adolesc Psychia Clin N Am. 2006;15:333–51. 9. Colins O, Ve mei en R, V eugdenhil C, an den B ink W, Do eleje s T, B oe‑ kae E. Psychia ic diso de s in de ained male adolescen s: a sys ema ic li e a u e e iew. Can J Psychia y. 2010;55:255–63. 10. Fazel S, Doll H, Långs öm N. Men al diso de s among adolescen s in ju enile de en ion and co ec ional acili ies: a sys ema ic e iew and me a eg ession analysis o 25 su eys. J Am Acad Child Adolesc Psychia‑ y. 2008;47:1010–9. 11. A ola R, An ila H, Riipinen P, Hakko H, Riala K, Kan ojä i L. Bo de line pe sonali y diso de associa es wi h iolen c iminali y in women: a popula ion based ollow‑up s udy o adolescen psychia ic inpa ien s in No he n Finland. Fo ensic Sci In . 2016;266:389–95. 12. We e bo g D, Långs öm N, Ande sson G, Eneb ink P. Bo de line pe sonali y diso de : p e alence and psychia ic como bidi y among male o ende s on p oba ion in Sweden. Comp Psychia y. 2015;62:63–70. 13. Ame ican Psychia ic Associa ion. Diagnos ic and s a is ical manual o men al diso de s. 5 h ed. Washing on, DC: Ame ican Psychia ic P ess; 2013. 14. Ha is EC, Ba aclough B. Excess mo ali y o men al diso de . B J Psychia‑ y. 1998;173:11–53. 15. Pique o AR, Fa ing on DP, Shephe d JP, Au y K. O ending and ea ly dea h in he Camb idge S udy in delinquen de elopmen . Jus ice Q. 2014;31:445–72. 16. Maughan B, S a o d M, Shah I, Kuh D. Adolescen conduc p oblems and p ema u e mo ali y: ollow‑up o age 65 yea s in a na ional bi h coho . Psychol Med. 2014;44:1077–86. 17. E onen M, Hakola P, Tiihonen J. Men al diso de s and homicidal beha io in Finland. A ch Gen Psychia y. 1996;53:497–501. 18. Vi kkunen M, Egge M, Rawlings R, Linnoila M. A p ospec i e ollow‑up s udy o alcoholic iolen o ende s and i e se e s. A ch Gen Psychia y. 1996;53:523–9. 19. Goodman M, New A. Impulsi e agg ession in bo de line pe sonali y diso de . Cu Psychia y Rep. 2000;2:56–61. 20. Skodol AE, Sie e LJ, Li esley WJ, Gunde son JG, P ohl B, Widige TA. The bo de line diagnosis II: biology, gene ics, and clinical cou se. Biol Psychia‑ y. 2002;51:951–63. 21. Cloninge CR, Bohman M, Sig a dsson S. Inhe i ance o alcohol abuse. C oss‑ os e ing analysis o adop ed men. A ch Gen Psychia y. 1981;38:861–8. 22. Timko C, DeBenede i A, Moos BS, Moos RH. P edic o s o 16‑yea mo ali y among indi iduals ini ia ing help‑seeking o an alcoholic use diso de . Alcohol Clin Exp Res. 2006;30:1711–20. 23. Hoye A, Jacobsen BK, Hansen V. Sex di e ences in mo ali y o admi ed pa ien s wi h pe sonali y diso de s in No h No way—a p ospec i e egis e s udy. BMC Psychia y. 2013;133:17. 24. Osby U, Co eia N, B and L, Ekbom A, Spa en P. Mo ali y and causes o dea h in schizoph enia in S ockholm coun y, Sweden. Schizoph Res. 2000;45:21–8. 25. Fo s BM, Isacson D, Binge o s K, Wide lo B. Mo ali y among pe sons wi h schizoph enia in Sweden: an epidemiological s udy. No d J Psychia y. 2007;61:252–9. 26. Lappi‑Seppälä T. Regula ing he p ison popula ion. Expe iences om a long‑ e m policy in Finland. Resea ch Communica ions 28. Helsinki: Na ional Resea ch Ins i u e o Legal Policy; 1998. 27. Vohlonen I, Mackiewicz K, Kois inen V. In e na ional compa ison o di ‑ e ences in p ema u e dea hs and hei causes in he selec ed NDPHS pa ne coun ies and Bela us measu ed by po en ial yea s o li e los indica o : ND – PYLL S udy 2016/Final esul s. Abs ac p esen ed in he 12 h no he n dimension pa ne ship in public heal h and social wellbe‑ ing (NDPHS) pa ne ship annual con e ence, 27 h o Oc obe , 2016, Sopo , Poland. 28. E onen M, Repo E, Va iainen H, Tiihonen J. Fo ensic psychia ic o ganiza‑ ion in Finland. In J Law Psychia y. 2000;23:541–6. 29. Wo ld Heal h O ganiza ion. Manual o he in e na ional s a is ical clas‑ si ica ion o diseases, inju ies, and causes o dea h. 8 h Re ision. Gene a: Wo ld Heal h O ganiza ion; 1965. 30. Ame ican Psychia ic Associa ion. Diagnos ic and s a is ical manual o men al diso de s (3 d ed., e ised). Washing on, DC: Ame ican psychia ic P ess; 1987. 31. Wo ld Heal h O ganiza ion. Manual o he in e na ional s a is ical clas‑ si ica ion o diseases, inju ies, and causes o dea h. 9 h Re ision. Gene a: Wo ld Heal h O ganiza ion; 1977. 32. Wo ld Heal h O ganiza ion. The ICD‑10 classi ica ion o men al and beha iou al diso de s: clinical desc ip ions and diagnos ic guidelines. Gene a: Wo ld Heal h O ganiza ion; 1992. 33. Fazel S, G ann M. The popula ion impac o se e e men al illness on iolen c ime. Am J Psychia y. 2006;163:1397–403. 34. S enbacka M, Mobe g T, Romelsjö A, Jokinen J. Mo ali y and causes o dea h among iolen o ende s and ic ims—a Swedish popula ion based longi udinal s udy. BMC Public Heal h. 2012;12:38. 35. Tikkanen R, Holi M, Lindbe g N, Vi kkunen M. Ques ionnai e da a on alco‑ holic iolen o ende s: speci ic connec ions o se e e impulsi e clus e B pe sonali y diso de s and iolen c iminali y. BMC Psychia y. 2007;7:36. 36. Tikkanen R, Holi M, Lindbe g N, Tiihonen J, Vi kkunen M. Recidi is ic o ending and mo ali y in alcoholic iolen o ende s: a p ospec i e ollow‑up s udy. Psychia y Res. 2009;168:18–25. 37. Saha S, Chan D, McG a h J. A sys ema ic e iew o mo ali y in schizo‑ ph enia: is he di e en ial mo ali y gap wo sening o e ime? A ch Gen Psychia y. 2007;64:1123–31. 38. Ea on WW, Ro h KB, B uce M, Co le L, Wu L, Nes ad G, Fo d D, Bien ‑ enu OJ, C um RM, Rebok G, An hony JC, Munoz A. The ela ionship o men al and beha io al diso de s o all‑cause mo ali y in a 27‑yea ollow‑up o 4 epidemiologic ca chmen a ea samples. Am J Epidemiol. 2013;178:1366–77. 39. Chesney E, Goodwin GM, Fazel S. Risks o all‑cause and suicide mo ali y in men al diso de s: a me a‑ e iew. Wo ld Psychia y. 2014;13:153–60. 40. Su isaa i J, Pa i K, Pe älä J, Vie iö S, Saa ni SE, Lönnq is J, Saa ni SI, Hä känen T. Mo ali y and i s de e minan s in people wi h psycho ic diso de . Psychosom Med. 2013;75:60–7. 41. Tiihonen J, Lönnq is J, Wahlbeck K, Klaukka T, Niskanen L, Tanskanen A, Haukka J. 11‑yea ollow‑up o mo ali y in pa ien s wi h schizoph enia: a popula ion‑based coho s udy (FIN11 s udy). Lance . 2009;374:620–7. 42. Ki inoja M. A he end o he oad, a e o m school. A s udy o he child wel a e clien ing and school his o y o child en placed in e o m schools. Helsinki: Na ional Resea ch and De elopmen Cen e o Wel a e and Heal h (STAKES); 2005. 43. Manninen M, Pankakoski M, Gissle M, Su isaa i J. Adolescen s in a esiden ial school o beha io diso de s ha e an ele a ed mo ali y isk in young adul hood. Child Adolesc Psychia y Men Heal h. 2015;9:46. 44. Lindbe g N, Oshuko a S, Mie unen J, Kal iala‑Heino R. Do se iously o ending gi ls di e om hei age‑ and o ence ype‑ma ched male coun e pa s on psychopa hic ai s o psychopa hy‑ ela ed backg ound a iables? Child Adolesc Psychia y Men Heal h. 2016;10:38. 45. Na ional S a is ics. 2007. h p://www.s a is ics.go .uk/cci/nugge . asp?ID=6. 46. Rikollisuus ilanne 2014. Ka sauksia 4/2015. Helsinki: Helsingin yliopis o, K iminologian ja oikeuspoli iikan ins i uu i; 2014. Page 9 o 9 Lindbe g e al. Child Adolesc Psychia y Men Heal h (2017) 11:37 • We accep p e-submission inqui ies • Ou selec o ool helps you o ind he mos ele an jou nal • We p o ide ound he clock cus ome suppo • Con enien online submission • Tho ough pee e iew • Inclusion in PubMed and all majo indexing se ices • Maximum isibili y o you esea ch Submi you manusc ip a www.biomedcen al.com/submi Submi you nex manusc ip o BioMed Cen al and we will help you a e e y s ep: 47. Muennig P, F anks P, Jia H, Lube kin E, Gold MR. The income‑associa ed bu den o disease in he Uni ed S a es. Soc Sci Med. 2005;61:2018–26. 48. A i i TO, Enns MW, Cox BJ, DeG aa R, TenHa e M, Sa een J. Child abuse and heal h‑ ela ed quali y o li e in adul hood. J Ne Men Dis. 2007;195:797–804. 49. Wegman HL, S e le C. A me a‑analy ic e iew o he e ec s o child‑ hood abuse on medical ou comes in adul hood. Psychosom Med. 2009;71:805–12. 50. Hagels am C, Häkkänen H. Adolescen homicides in Finland: o ence and o ende cha ac e is ics. Fo ensic Sci In . 2006;164:110–5.