Está en la página 1de 120

1

European Journal of Psychotraumatology Co-Action Publishing

Predicting criminality from child maltreatment typologies and posttraumatic stress symptoms
Ask Elklit, Karen-Inge Karstoft, [... , an! "ogens Christoffersen A!!itional article information

Abstract
Background
#he associations bet$een chil!hoo! abuse an! subse%uent criminality an! posttraumatic stress !isor!er &P#'() are $ell kno$n. *o$e+er, a ma,or limitation of research relate! to chil!hoo! abuse an! its effects is the focus on one particular type of abuse at the e-pense of others. .ecent $ork has establishe! that chil!hoo! abuse rarely occurs as a uni!imensional phenomenon. #herefore, a number of stu!ies ha+e in+estigate! the e-istence of abuse typologies.

Methods
#he stu!y is base! on a (anish stratifie! ran!om probability sur+ey inclu!ing /012 inter+ie$s of /3-year-ol! people. #he sample $as constructe! to inclu!e an o+ersampling of chil! protection cases. 4uil!ing on a pre+ious latent class analysis of four types of chil!hoo! maltreatment, three maltreatment typologies $ere use! in the current analyses. A criminality scale $as constructe! base! on se+en types of criminal beha+ior. P#'( symptoms $ere assesse! by the PC-P#'( 'creen.

Results
'ignificant !ifferences $ere foun! bet$een the t$o gen!ers $ith males reporting heightene! rates of criminality. 5urthermore, all three maltreatment typologies $ere associate! $ith criminal beha+ior $ith o!!s ratios &6.s) from /.02 to 7.8/. 5emale gen!er ha! an 6. of 2.78 an! possible P#'( an 6. of 9.13.

Conclusion

#he in!epen!ent association of participants at risk for P#'( an! three types of maltreatment $ith criminality shoul! be stu!ie! to !etermine if it can be replicate!, an! consi!ere! in social policy an! pre+ention an! rehabilitation inter+entions. Keywords: Chil!hoo! maltreatment, se-ual abuse, emotional abuse, posttraumatic stress !isor!er, latent classes, criminal beha+ior, national representati+e stu!y Criminal beha+ior in a!olescence is of great societal concern. #herefore, the i!entification of risk factors relate! to criminality has been of great interest. 6ne chil!hoo! factor that has been consistently associate! $ith a!olescent criminal beha+ior is chil!hoo! abuse &'te$art, :i+ingston, ; (ennison, /221< #hornberry, Irelan!, ; 'mith, /229). =otably, a $ell-establishe! outcome of chil!hoo! abuse is psychopathology inclu!ing posttraumatic stress !isor!er &P#'(< Claussen ; Critten!en, 9009< (eblinger, "c:eer, Atkins, .alphe, ; 5oa, 9010< 5ergusson, *or$oo!, ; :ynskey, 900>). In a!!ition, albeit contro+ersial, stu!ies ha+e reporte! that psychiatric !isor!ers in general are associate! $ith criminal con+ictions &*o!gins, "e!nick, 4rennan, 'chulsinger, ; Engberg, 900>). #he e-tant literature has consistently sho$n significant associations bet$een chil!hoo! maltreatment an! !ifferent forms of criminal an! a!+ersi+e beha+ior in a!olescence. 5or e-ample, o+erall !elin%uency &*eck ; ?alsh, /222), +iolent !elin%uency &'mith ; #hornberry, 9007), ,u+enile offen!ing &'te$art et al. /221< #hornberry et al. /229), an! antisocial, aggressi+e, an! +iolent criminal beha+iors &*aapasalo ; Pokela, 9000) ha+e all been associate! $ith chil!hoo! maltreatment. 'ome researchers ha+e reporte! that specific maltreatment types are relate! to subse%uent antisocial !e+elopment. 5or e-ample, physical abuse has been foun! to be in!epen!ently pre!icti+e of +iolent beha+ior &*errenkohl, *uang, #a,ima, ; ?hitney, /228< "aas, *errenkohl, ; 'ousa, /221< ?i!om, 9010) an! more specifically +iolent se-ual offen!ing &?i!om ; Ames, 9003). '$antson et al. &/228) reporte! that chil!ren $ho ha! e-perience! se-ual abuse $ere o+er t$ice as likely &o!!s ratio [6. @/./0) to report in+ol+ement in criminal acti+ity compare! to non-abuse! controls. :ike$ise, 'iegel an! ?illiams &/228) reporte! that chil!ren $ho ha! e-perience! se-ual abuse $ere o+er t$ice as likely to be arreste! as ,u+eniles &6.@/.3) an! a!ults &6.@/.2) compare! to non-abuse! controls. #he e-perience of chil!hoo! abuse has also long been associate! $ith psychopathology. In particular, it is estimate! that bet$een /7 an! >7A of +ictims of chil!hoo! abuse $ill go on to !e+elop P#'( &Albach ; E+eraer!, 9008< Chu ; (ill, 9002< Kiser et al. 9009< Palmer et al. 900/). #o !ate, research has pre!ominately focuse! on the e-perience of se-ual abuse an! physical abuse in relation to subse%uent psychopathology &cf. 4ro$n ; An!erson, 9009< 5ergusson, 4o!en, ; *or$oo!, /221< Jumper, 9007< Ken!ler et al. /222). Albeit contro+ersial, some stu!ies ha+e reporte! that psychopathology may also pre!ict criminality. =otably, the pre+alence of psychopathology $ithin the prison system far e-cee!s that of the general population &4rin!e! et al. /229). In!ee!, 'ingleton et al. &9001) reporte! that almost 02A of the prison population has a mental health issue.

5urthermore, criminal offen!ers ha+e been foun! to possess higher rates of traumatic e-periences than non-offen!ers &Bol!enson, Beffner, 5oster, ; Clipson, /22C) an! community samples &Abram, #eplin, Charles, :ong$orth, "cClellan!, ; (ulcan /223< =eller, (enney, PietD, ; #homlinson, /22>). An!, the !e+elopment of +iolent beha+ior has been argue! to be strongly associate! $ith posttraumatic stress &4egic ; Jokic-4egic, /22/). 6ne notable limitation of research concerne! $ith the effect of chil!hoo! abuse is the focus on particular types of abuse at the e-pense of others &*iggins ; "cCabe, /229< Pears, Kim, ; 5isher, /221). In!ee!, chil!hoo! abuse is commonly !i+i!e! across !omains of physical abuse, emotional abuse, se-ual abuse, an! chil!hoo! neglect &'e!lak et al., /221). ?hile these categories of chil!hoo! abuse are generally agree! upon, no agreement is foun! $hen it comes to the relati+e pre+alence of the abuse types. ?hereas some stu!ies ha+e foun! chil!hoo! neglect to be the most pre+alent type of abuse &E.'. (epartment of *ealth an! *uman 'er+ices, /292), other stu!ies fin! that emotional abuse is the most pre+alent &Armour, Elklit, ; Christoffersen, in press< 4rooker, Ca$son, Kelly, ; ?attam, /229). #his might reflect !isagreement on the !efinition an! operationaliDation of chil! maltreatment $ithin the literature< $ith !ifferent abuse types being !ifficult to separate. *o$e+er, it may also reflect a high co-occurrence bet$een the !ifferent forms of chil!hoo! abuse &Kantor ; :ittle, /228< 'tanley ; Bo!!ar!, /223), in that researchers ask %uestions $hich !o not co+er the full spectrum of abuse types or in a reporting bias $hereby in!i+i!uals only report e-periencing the particular abuse type $hich they percei+e as their $orst, or in!ee! their most recent e-perience. "ore recently, multiple abuse e-periences has been regularly reporte! $ithin the literature &e.g., Armour et al., in press< 5inkelhor et al. /22C< *iggins ; "cCabe, /229). *o$e+er, a less stu!ie! area is one $hich in+estigates abuse typologies. Armour et al. &in press) con!ucte! a latent class analysis stu!y &using the same nationally representati+e !atabase as use! in this stu!y) $hich unco+ere! four abuse typologies, base! on ho$ respon!ents ans$ere! to /2 in!i+i!ual in!icators of abuse across four !ifferent abuse !omains. #he abuse typologies $ere characteriDe! as a non-abuse! group, a pre!ominately psychologically maltreate! group, a pre!ominately se-ually abuse! group, an! a group e-periencing multiple abuse e-periences. "embership of abuse typologies $as pre!icte! by chil! protection status &i.e., ha+ing been pre+iously kno$n to the (anish social ser+ices), female gen!er $as pre!icti+e of membership in the pre!ominately se-ually abuse! class an! chil! protection status $as pre!icti+e of membership in all abuse classes compare! to the non-abuse! class. #he Armour et al. &in press) stu!y !i! not, ho$e+er, in+estigate the association bet$een the abuse typologies an! criminal beha+ior. =or !i! the Armour et al. &in press) stu!y in+estigate the relationship bet$een the abuse typologies an! P#'(. #herefore, this stu!y aims to buil! on the Armour et al. &in press) stu!y by in+estigating the relationship bet$een chil!hoo! abuse typologies, gen!er, P#'(, an! criminal beha+ior in a nationally representati+e youth sample.

Methods

A stratifie! ran!om probability sur+ey $as con!ucte! in (enmark by the =ational Centre for ?elfare bet$een /221 an! /220. 'tatistics (enmark ran!omly selecte! 3,C91 participants, age! /3, from the total birth cohort of (enmark in 9013. 'tructure! inter+ie$s $ere con!ucte! by traine! inter+ie$ers either in the home or +ia the telephone. 5urther !etails pertaining to the proce!ure are a+ailable in Christoffersen, Armour, :asgaar!, An!ersen, ; Elklit &in press). A total of /,012 inter+ie$s $ere successfully con!ucte!, e%uating to a response rate of >CA. #o increase the number of participants $ho ha! e-perience! chil!hoo! abuse an! neglect, chil!ren $ho ha! been in chil! protection $ere o+er-sample! by stratifying the number of Fchil! protection casesG +ersus Fnon-chil! protection casesG &9H8I/H8). A chil! protection case $as !efine! as a case $here the council &accor!ing to the files of local social $orkers) ha! pro+i!e! support for the chil! an! the family or placement $ith a foster family !ue to concerns about the $ell-being an! !e+elopment of the chil!. A total of 17/ inter+ie$s $ere con!ucte! $ith in!i+i!uals $ho ha! been pre+iously i!entifie! by the (anish authorities as chil! protection cases. #he most common reason for non-participation $as refusal &/9A). 6ther reasons inclu!e! illness, !isability, an! being un-contactable. #he sample consiste! of 9,7C0 males an! 9,329 females. #he ma,ority of the sample either o$ne! or rente! their o$n pri+ate accommo!ation &08.CA) an! almost half $ere marrie! or cohabiting &3>.2A). All !emographics $ere analyDe! employing a $eight +ariable to account for the o+ersampling of chil! protection cases so that fin!ings are representati+e of the total (anish population of young people age! /3 years. Chil! protection status &$eighte!) $as gi+en as >.8A of the total sample.

Measures
#he inter+ie$ a!ministere! a series of %uestions pertaining to se+eral psychological an! physical !omains in a!!ition to %uerying about se+eral !emographics. .espon!ents ans$ere! se+eral specific %uestions across four !omains of chil!hoo! maltreatmentI physical abuse, emotional abuse, neglect, an! se-ual abuse. 5or an elaborate! !escription of the measures use! to assess chil! maltreatment, see Christoffersen et al. &in press). As pre+iously note!, the abuse typologies $ere !eri+e! by Armour et al. &in press) $ho implemente! the statistical techni%ue of latent class analysis. 5our abuse typologies $ere unco+ere! from /2 separate in!icators of abuse spanning four abuse !omainsI the pre!ominately psychologically maltreate! group &1.1A), the pre!ominately se-ually abuse! group &/.2A), the o+erall abuse group &physical abuseJneglectJemotional abuse< /.9A), an! the non-abuse! group &1C.9A). Criminal beha+ior $as assesse! by se+en items %uestioning $hether an in!i+i!ual ha! e+er shoplifte!, stole a bicycle, stole a car, committe! burglary, +an!alism, or +iolence, or been con+icte! of a crime &Christoffersen, 900>< Ky+sgaar!, 900/). .espon!ents ans$ere! FyesG or FnoG to these se+en items &#able 9). A scale $as compute! $ith these se+en items &CronbachKs alpha@2.C7). Criminality $as en!orse! if scores $ere L8, $hich $as the case for 92.0A of the sample. 5our items &stealing a car, burglary, +iolence, an!

con+iction) $ere en!orse! by >0M0>A of the L8 criminality group< the three other items $ere en!orse! by 81M37A of the same group. #his in!icates that the cut-off presents $ell as a criterion for more serious crime.

#able 9 Pre+alence of the criminal beha+ior by gen!er 'creening for possible P#'( $as assesse! using the Primary Care P#'( 'creen &PCP#'(< Prins et al., /228) $ith four !ichotomous items representing the main symptom groups, inclu!ing intrusi+e memories, a+oi!ance, increase! arousal, an! emotional numbing. A score L8 $as !etermine! to be the most efficient cut-off in E.'. a!ult primary care samples an! is use! to i!entify participants $ith likely P#'( &Prins et al., /228).

Data analysis
#he relationship bet$een chil!hoo! abuse an! criminality $as teste! by a logistic regression mo!el using criminality as the !epen!ent, categorical +ariable an! e-ploring the pre!icti+e ability of the chil!hoo! abuse typologies, gen!er, an! P#'(. #he force! entry metho! $as chosen as all pre!ictor +ariables are teste! in one block to assess their pre!icti+e ability, $hile controlling for the effects of other pre!ictors in the mo!el.

Results
(escripti+e statistics on the pre+alence of criminal beha+ior by gen!er can be seen in #able 9. All !ifferences $ere significant &all FsN/7< all psO2.2227) by a one-$ay A=6PA test. #able / sho$s criminality scores by abuse classes an! by gen!er en!orsement. 'e-ual abuse has an en!orsement a little less than the non-abuse! class. #he o+erall abuse an! the emotional abuse classes ha+e comparati+ely large 6.s< the same is the case for se-ual abuse but one has to be a$are of the +ery small n in this particular class.

#able / Abuse classes by gen!er en!orsement A logistic regression $as performe! for the three abuse classes, gen!er, P#'(, an! their relation to criminal beha+ior to estimate the pre!icti+e +alue of these +ariables &see #able 8). #he 6mnibus #ests of "o!el Coefficients an! the *omer an! :emesho$ test both supporte! the mo!el an! in!icate! a goo!ness of fit. All +ariables contribute! significantly to the pre!icti+e ability. All three abuse classes $ere significantly associate! $ith criminal beha+ior $ith o!!s ratios &6.s) +arying from /.02 to 7.8/, se-ual abuse ha+ing the lo$est, follo$e! by emotional abuse an! o+erall abuse ha+ing the highest 6.s. Ben!er &being female) $as a pre!icti+e factor $ith an 6. of 2.78. Controlling for all other factors, P#'( a!!e! in!epen!ently to the pre!iction of criminal beha+ior $ith an 6. of 9.13. #he regression mo!el accounte! for bet$een 9/A &Co- ; 'nell) an! /3A &=agelkerke) of the +ariation in criminal beha+ior.

#able 8 :ogistic regression analysis pre!icting criminal beha+ior from three classes of chil!hoo! maltreatment, gen!er, an! P#'(

Discussion
A significant positi+e relationship bet$een chil! abuse an! criminal beha+ior has been foun! rather consistently in the literature. In this stu!y of a large, representati+e (anish youth sample, $e foun! se+eral significant associations bet$een all three classes of chil!hoo! maltreatment an! criminal beha+ior. 5urthermore, $e foun! substantial in!epen!ent effects of both gen!er an! P#'( symptomatology on criminal beha+ior, controlling for the effects of abuse classes. #he relationships obser+e! in this stu!y $ere %uite strong. #he stu!y e-ten!s the e-tant research by using empirical base! abuse classes an! not isolate!, constructe! abuse! types. In line $ith the e-tant research in the fiel! &'iegel ; ?illiams, /228< 'te$art et al. /221< '$antson et al. /228< cf. #hornberry et al. /229), $e foun! that all abuse typologies $ere associate! $ith criminal beha+ior to +arious !egrees. In!ee!, the associations bet$een classes an! criminal beha+ior appear to be theoretically meaningful. #he o+erall abuse! class, $ho strongly en!orse! e-periences of physical abuse, emotional abuse, an! physical neglect, ha! the strongest risk for criminal beha+ior later in life. #his is supporte! by literature relate! to the cumulati+e effects of trauma. In!ee!, stu!ies ha+e sho$n that cumulati+e trauma can e-acerbate psychopathology &'he+lin,

*ouston, (orahy, ; A!amson, /221)< thus, $e can speculate that cumulati+e trauma in the form of multiple types of chil!hoo! abuse e-periences also has the ability to e-acerbate alternati+e negati+e outcomes such as criminality. #hose $ho $ere pre!ominately emotionally maltreate! in chil!hoo! also ha! a high risk of later criminal beha+ior an! the same $as the case, although to a lesser !egree for the group $ho $ere se-ually abuse!. In!epen!ent of abuse class, female gen!er appeare! to be a protecti+e factor, an! this $as e-pecte! base! on the abun!ance of research $hich in!icates that males are more likely to participate in criminal acts or !isplay criminal beha+iors compare! to their female counterparts &'teffensmeier ; Allan, 900>). #hus, it stan!s to reason that being female $ill !ecrease the likelihoo! $ith $hich an in!i+i!ual is criminal. Criminogenic theories suggest that females are more likely not to partake in criminal acti+ity because gen!er norms, social control, lack of physical strength, an! moral an! relational concerns restrict female access to criminal opportunity. #he same factors also limit female $illingness to participate in crime at the motivational le+el. In a!!ition, at a contextual le+el there are typical gen!er !ifferences, females often being in+ol+e! in simple forms of !elin%uency, an! are unlikely to use $eapons or inten! serious in,ury to their +ictims &'teffensmeier ; Allan, 900>). #he interesting aspect of the current stu!y is that $e ha+e sho$n that this association hol!s e+en $hen controlling for the e-perience of abuse an! P#'(. #hose $ho suffere! from P#'( !ue to chil!hoo! abuse or !ue to other traumas later in life ha! almost a !ouble risk of criminality. #his is supporte! by :oeber, 5arrington, 'touthamer-:oeber, "offitt, an! Caspi &9001) $ho lai! the groun! for a !e+elopmental criminology that suggests that offen!ing has both long-term an! imme!iate antece!ents. #hey +ie$ serious offen!ing as cause! by the accumulation of !e+iance processes from chil!hoo! through to a!ulthoo!. #he basic premise of the theory is that offen!ing results from forces $ithin the in!i+i!ual &impulsi+ity, lack of guilt) an! forces in the social en+ironment &parental rearing practices, re,ection by peers) in !ifferent conte-ts &$ork place, home, neighborhoo!). All three abuse classes are e-amples fitting $ell to this theory as forces in the social en+ironment an! the chronic P#'( state is an e-ample of forces from $ithin the in!i+i!ual, $hich represents a +ulnerability that is often turne! in$ar!s &particularly in females) but can also be turne! out$ar!s in acting-out beha+ior &especially in males< "iller, Kaloupek, (illon, ; Keane, /223). #his stu!y ha! se+eral limitations. 5irst, the stu!y $as cross-sectional an! relie! on selfreport. Bi+en that the abuse e+ents ha+e occurre! many years earlier, this might cause a certain bias, potentially lea!ing to o+er- or un!er-reporting. Also, current psychosocial functioning coul! bias the recall of prior abuse. #he PC-P#'( 'creen is pro-y for P#'( an! !oes not co+er all symptoms of the !iagnosis. :ike$ise, the inclusion of specific abuse e+ents before the latent class analysis $as performe! !oes not rely on an establishe! consensus. #he same can be sai! of the constructing of the criminality measure an! the cut-off points chosen. #he small N for se-ual abuse in the multi+ariate regression analysis also !eman!s caution against conclu!ing that se-ual abuse e-posure is associate! $ith criminal beha+ior. #he foun! association shoul! be teste! in other stu!ies $ith larger samples.

#he fin!ings on the long-term effects of the three abuse classes an! P#'( !eser+e further in+estigation. #he relationship bet$een chil!hoo! abuse an! criminal beha+ior shoul! be further in+estigate! $ith other measures to fin! more me!iating or mo!erating current +ariables like alcohol an! !rug abuse, cogniti+e mechanisms, peer an! parent relationships, attachment style, personal resources, an! others. In a!!ition, longitu!inal stu!ies coul! pro+i!e important historical information about critical time perio!s, parental functioning, re+ictimiDation, etc. In conclusion, this stu!y !emonstrates the in!epen!ent association of possible P#'( $ith self-reporte! criminal beha+ior, controlling for the effects of chil!hoo! abuse an! gen!er base! on a large representati+e national sample. Combining chil!hoo! abuse assessment $ith screening for P#'( strengthens the possibilities for an informe! early inter+ention $hich coul! ai! in the pre+ention of young people entering a life of criminality.

otes
5or the abstract or full te-t in other languages, please see 'upplementary files un!er Article #ools online

Conflict of interest and funding


#here is no conflict of interest in the present stu!y for any of the authors.

Article information
Eur J Psychotraumatol. /298< 3I 92.832/He,pt.+3i2.901/7. Publishe! online /298 April /3. !oiI 92.832/He,pt.+3i2.901/7 P"CI(I P"C8>8>39C Ask Elklit,9,Q Karen-Inge Karstoft,9 Cherie Armour,9 (agmar 5e!!ern,/ an! "ogens Christoffersen8 9 =ational Centre for Psychotraumatology, Eni+ersity of 'outhern (enmark, (enmark / (anish Center for .esearch in Alcohol an! (rug Abuse 'tu!ies, Aarhus Eni+ersity, (enmark 8 #he =ational .esearch Centre for ?elfare, Copenhagen, (enmark Q Correspon!ence toI Ask Elklit, =ational Centre for Psychotraumatology, Eni+ersity of 'outhern (enmark, Campus+e, 77, (K-7/82 6!ense ", (enmark, EmailI aelklitHatHhealth.s!u.!k .ecei+e! =o+ember 9>, /29/< .e+ise! "arch >, /298< Accepte! "arch >, /298. Copyright R /298 Ask Elklit et al. #his is an open-access article !istribute! un!er the terms of the Creati+e Commons Attribution :icense, $hich permits unrestricte! use, !istribution, an! repro!uction in any me!ium, pro+i!e! the original $ork is properly cite!. Articles from European Journal of Psychotraumatology are pro+i!e! here courtesy of Co! Action Publishing

References
1. Abram K. ", #eplin :. A, Charles (. ., :ong$orth '. :, "cClellan! B. ",

(ulcan ". K. Posttraumatic stress !isor!er an! trauma in youth in ,u+enile !etention. Archi+es of Beneral Psychiatry. /223<>9I328M392. [P"C free article [Pub"e! 2. Albach 5, E+eraer! ?. Posttraumatic stress symptoms in +ictims of chil!hoo! incest. Psychotherapy an! Psychosomatics. 9008<7CI938M979. [Pub"e! 8. Armour C, Elklit A, Christoffersen ". Co-occurring chil!hoo! maltreatment in (enmarkI A latent class analysis. Journal of :oss an! #rauma. in press 3. 4egic (, Jokic-4egic =. Piolent beha+iour an! posttraumatic stress !isor!er. Current 6pinion in Psychiatry. /22/<97I>/8M>/>. 5. 4rin!e! P. ". J, 'impson A. I. 5, :ai!la$ #. ", 5airley =, "alcolm 5. Pre+alence of psychiatric !isor!ers in =e$ Sealan! prisonsI A national stu!y. Australian an! =e$ Sealan! Journal of Psychiatry. /229<87I9>>M9C8. [Pub"e! >. 4rooker ', Ca$son P, Kelly B, ?attam C. #he pre+alence of chil! abuse an! neglectI A sur+ey of young people. International Journal of "arket .esearch. /229<38I/30M/10. 7. 4ro$n B. ., An!erson 4. Psychiatric morbi!ity in a!ult inpatients $ith chil!hoo! histories of se-ual an! physical abuse. American Journal of Psychiatry. 9009<931I77M>9. [Pub"e! 1. Christoffersen ". =. A follo$-up stu!y of out-of-home care in (enmarkI :ongterm effects on self-esteem among abuse! an! neglecte! chil!ren. International Journal of Chil! ; 5amily ?elfare. 900><9I/7M80. 0. Christoffersen ", Armour C, :asgaar! ", An!ersen #. E, Elklit A. #he pre+alence of four types of chil!hoo! maltreatment in (enmark. Chil! Abuse ; =eglect. in press 10. Chu J. A, (ill (. :. (issociati+e symptoms in relation to chil!hoo! physical an! se-ual abuse. #he American Journal of Psychiatry. 9002<93CI11CM10/. [Pub"e! 11. Claussen A, Critten!en P. Physical an! psychological maltreatmentI .elations among types of maltreatment. Chil! Abuse an! =eglect. 9009<97I7M91. [Pub"e! 12. (eblinger E, "c:eer ', Atkins ", .alphe ", 5oa E. Post-traumatic stress in se-ually abuse!, physically abuse!, an! nonabuse! chil!ren. Chil! Abuse an! =eglect. 9010<98I328M321. [Pub"e! 13. 5ergusson (. ", 4o!en J. ", *or$oo! :. J. E-posure to chil!hoo! se-ual an! physical abuse an! a!,ustment in early a!ulthoo!. Chil! Abuse ; =eglect. /221<8/I>2CM>90. [Pub"e! 14. 5ergusson (. ", *or$oo! :. J, :ynskey ". #. Chil!hoo! se-ual abuse an! psychiatric !isor!er in young a!ulthoo!, III Psychiatric outcomes of chil!hoo! se-ual abuse. Journal of the American Aca!emy of Chil! an! A!olescent Psychiatry. 900><87I98>7M98C3. [Pub"e! 15. 5inkelhor (, 6rmro! .. K, #urner *. A. Poly-+ictimiDationI A neglecte! component in chil! +ictimiDation trauma. Chil! Abuse an! =eglect. /22C<89ICM />. [Pub"e!

16. Bol!enson J, Beffner ., 5oster '. :, Clipson C. .. 5emale !omestic +iolence

offen!ersI #heir attachment security, trauma symptoms, an! personality organiDation. Piolence an! Pictims. /22C<//I78/M737. [Pub"e! 9C. *aapasalo J, Pokela E. Chil!-rearing an! chil! abuse antece!ents of criminality. Aggression an! Piolent 4eha+ior. 9000<3I92CM9/C. 91. *eck C, ?alsh A. #he effects of maltreatment an! family structure on minor an! serious !elin%uency. International Journal of 6ffen!er #herapy an! Comparati+e Criminology. /222<33I9C1M908. 19. *errenkohl #. I, *uang 4, #a,ima E. A, ?hitney '. (. E-amining the link bet$een chil! abuse an! youth +iolenceI An analysis of me!iating mechanisms. Journal of Interpersonal Piolence. /228<91I9910M9/21. [Pub"e! /2. *iggins (. J, "cCabe ". P. "ultiple forms of chil! abuse an! neglectI A!ult retrospecti+e reports. Aggression an! Piolent 4eha+iour. /229<>I73CM7C1. 21. *o!gins ', "e!nick '. A, 4rennan P. A, 'chulsinger 5, Engberg ". "ental !isor!ers an! crimeI E+i!ence from a (anish birth cohort. Archi+es of Beneral Psychiatry. 900><78I310M30>. [Pub"e! 22. Jumper '. A. A meta-analysis of the relationship of chil! se-ual abuse to a!ult psychological a!,ustment. Chil! Abuse ; =eglect. 9007<90IC97MC/1. [Pub"e! /8. Kantor B. K, :ittle :. (efining the boun!aries of chil! neglectI ?hen !oes !omestic +iolence e%uate $ith parental failure to protectT Journal of Interpersonal Piolence. /228<91I881M877. 24. Ken!ler K. ', 4ulik C. ", 'ilberg J, *ettema J. ", "yers J, Prescott C. A. Chil!hoo! se-ual abuse an! a!ult psychiatric an! substance use !isor!ers in $omenI An epi!emiological an! cot$in control analysis. Archi+es of Beneral Psychiatry. /222<7CI078M070. [Pub"e! /7. Kiser :. J, *eston J, "illsap P. A, Pruitt (. 4. Physical an! se-ual abuse in chil!hoo!I .elationship $ith post-traumatic stress !isor!er. Journal of the American Aca!emy of Chil! ; A!olescent Psychiatry. 9009<82ICC>MC18. [Pub"e! />. Ky+sgaar! 4. =y ung!omT 6m familie, skole, friti!, lo+ly!ighe! og kriminalitet. CopenhagenI Jurist- og Ukonomforbun!ets 5orlag [=e$ VouthT 6n 5amily, 'chool, :eisure #ime, .espect for the :a$, an! Criminality < 900/. /C. :oeber ., 5arrington (. P, 'touthamer-:oeber ", "offitt #. E, Caspi A. #he !e+elopment of male offen!ingI Key fin!ings from the first !eca!e of the Pittsburgh Vouth 'tu!y. 'tu!ies in Crime an! Crime Pre+ention. 9001<CI939M9C/. 28. "aas C, *errenkohl #. I, 'ousa C. .e+ie$ of research on chil! maltreatment an! +iolence in youth. #rauma, Piolence, ; Abuse. /221<0I7>M>C. [Pub"e! 29. "iller ". ?, Kaloupek (. B, (illon A. :, Keane #. ". E-ternaliDing an! internaliDing subtypes of combat-relate! P#'(I A replication an! e-tension using the P'V-7 scales. Journal of Abnormal Psychology. /223<998I>8>M37. [Pub"e! 30. =eller (. J, (enney .. :, PietD C. A, #homlinson .. P. #he relationship bet$een trauma an! +iolence in a ,ail inmate sample. Journal of Interpersonal Piolence. /22></9I9/83M9/39. [Pub"e! 31. Palmer .. :, Chaloner (. A, 6ppenheimer .. Chil!hoo! se-ual e-periences $ith a!ults reporte! by female psychiatric patients. #he 4ritish Journal of Psychiatry. 900/<9>2I/>9M/>7. [Pub"e!

32. Pears K. C, Kim *. K, 5isher P. A. Psychosocial an! cogniti+e functioning of

chil!ren $ith specific profiles of maltreatment. Chil! Abuse ; =eglect. /221<8/I071M0C9. [P"C free article [Pub"e! 88. Prins A, 6iumette P, Kimerling ., Cameron .. P, *ugelshofer (. ', 'ha$*eg$er J, W'heikh J. I. #he primary care P#'( screen &PC-P#'()I (e+elopment an! operating characteristics. Primary Care Psychiatry. /228<0I0M 93. 34. 'e!lak A, Bragg 5, "ettenburg J, Ciarico J, ?inglee ", 'hapiro B, *artge J, "cPherson K. 5ourth =ational Inci!ence 'tu!y of Chil! Abuse an! =eglect &=I'-3) /221. Aug 9, .etrie+e! from httpIHH$$$.nis3.orgH(6C'H=is3(esignX"etho!X'ummary.p!f. 87. 'ingleton =, "eltDer *, Bat$ar! .. Psychiatric "orbi!ity among Prisoners in Englan! an! ?ales &6ffice for =ational 'tatistics) :on!onI 'tationery 6ffice< 9001. 36. 'he+lin ", *ouston J. E, (orahy ". J, A!amson B. Cumulati+e traumas an! psychosisI An analysis of the =ational Comorbi!ity 'ur+ey an! 4ritish Psychiatric "orbi!ity 'ur+ey. 'chiDophrenia 4ulletin. /221<83I908M900. [P"C free article [Pub"e! 8C. 'iegel J. A, ?illiams :. ". #he relationship bet$een chil! se-ual abuse an! female !elin%uency. Journal of .esearch in Crime an! (elin%uency. /228<32IC9M 03. 81. 'mith C. A, #hornberry #. P. #he relationship bet$een chil!hoo! maltreatment an! a!olescent in+ol+ement in !elin%uency. ?orking Paper =o. 9C. Criminology. 9007<88I379M319. 80. 'tanley J, Bo!!ar! C. "ultiple forms of +iolence an! other criminal acti+ities as an in!icator of se+ere chil! maltreatment. Chil! Abuse .e+ie$. /223<98I/3>M />/. 32. 'teffensmeier (, Allan E. Ben!er an! crime. #o$ar! a gen!ere! theory of female offen!ing. Annual .e+ie$ of 'ociology. 900><//I370M31C. 41. 'te$art A, :i+ingston ", (ennison '. #ransitions an! turning pointsI E-amining the links bet$een chil! maltreatment an! ,u+enile offen!ing. Chil! Abuse ; =eglect. /221<8/I79M>>. [Pub"e! 3/. '$antson V. *, Parkinson P. =, 6K#oole 4. I, Plunkett A. ", 'hrimpton ', 6ates .. K. Ju+enile crime, aggression an! !elin%uency after se-ual abuseI A longitu!inal stu!y. 4ritish Journal of Criminology. /228<38IC/0MC30. 43. #hornberry #, Irelan! #, 'mith C. #he importance of timingI #he +arying impact of chil!hoo! an! a!olescent maltreatment on multiple problem outcomes. (e+elopment an! Psychopathology. /229<98I07CM0C0. [Pub"e! 33. E.'. (epartment of *ealth an! *uman 'er+ices. ?ashington, (CI E' Bo+ernment Printing 6ffice< /292. Chil! maltreatment /221. A!ministration on chil!ren, youth an! families. 45. ?i!om C. '. #he cycle of +iolence. 'cience. 9010</33&3029)I9>2M9>>. [Pub"e! 46. ?i!om C. ', Ames ". A. Criminal conse%uences of chil!hoo! se-ual +ictimiDation. Chil! Abuse ; =eglect. 9003<91I828M891. [Pub"e!

2 PLoS ONE Public Library of Science

"mpact on "nfants# Cogniti$e De$elopment of Antenatal %&posure to "ron Deficiency Disorder and Common Mental Disorders
#hach (uc #ran, 4e+erley-Ann 4iggs, [... , an! Jane 5isher A!!itional article information

Abstract
'b(ecti$es
#he aim of this stu!y $as to e-amine the effects of antenatal e-posure to iron !eficiency anemia &I(A) an! common mental !isor!ers &C"() on cogniti+e !e+elopment of > months ol! infants in a !e+eloping country.

Methods
A prospecti+e population-base! stu!y in a rural pro+ince in Pietnam, $hich enrolle! pregnant $omen at 9/M/2 $eeks gestation an! follo$e! them up $ith their infants until si- months postpartum. Criteria for I(A $ere *b O99 gH!: an! serum ferritin O97 ngHm:. C"( symptoms $ere assesse! by the E!inburgh Postnatal (epression 'calePietnam +ali!ation. Infant cogniti+e !e+elopment $as assesse! by 4ayley 'cales of Infant an! #o!!ler (e+elopment, 8r! E!. Path analyses $ere performe! to !etermine the !irect an! in!irect, partly or fully me!iate!, causal effects of the antenatal e-posures.

Results
A total of 30C pregnant $omen $ere recruite!, of those 8C1 $omen pro+i!e! complete !ata $hich $ere inclu!e! in the analyses. 'tatistically significant !irect a!+erse effects of persistent antenatal I(A &estimate! !ifference of Y99.>/ points< 07A CI Y/8.29 to Y2.//) an! antenatal C"( &Y3.12 points< 07A CII Y0.32 to Y2./2) on infant 4ayley cogniti+e scores at si- months $ere foun!. *igher birth$eight, househol! $ealth, an! self-rate! sufficient supply of breastmilk $ere associate! $ith higher cogniti+e scores. "aternal age N82 years an! primiparity ha! an in!irect a!+erse effect on infantsZ 4ayley cogniti+e scores.

Conclusions
#hese fin!ings suggest that antenatal I(A an! C"( both ha+e a!+erse effects on chil! cogniti+e !e+elopment, $hich if unrecogniDe! an! una!!resse! are likely to be lasting. It is crucial that both these risks are consi!ere! by policy makers, clinicians, an! researchers seeking to impro+e chil! cogniti+e function in !e+eloping countries.

"ntroduction
It is estimate! that /22 million chil!ren age! less than 7 years in the $orl! !o not reach their !e+elopmental potential an! most of them are li+ing in lo$- an! lo$er-mi!!leincome countries [9 . #he intrauterine en+ironment appears to be a critical !eterminant of the !e+elopment of the baby in later life through Fprogramming mechanismsG [/ . 5etal programming postulates that fetal a!,ustments to a!+erse con!itions in utero may cause long-lasting changes in physiological functions that ren!er the brain or bo!y +ulnerable to !e+elopmental !elay an!Hor illnesses later in life [/ . #here is con+incing e+i!ence that fetal programming contributes to chronic a!ult con!itions inclu!ing obesity an! car!io+ascular !iseases [8 , . .ecently attention has turne! to e-amining fetal programming hypotheses in relation to neurobeha+ioral outcomes. Iron !eficiency, $hich at its most se+ere is terme! iron !eficiency anemia &I(A), is the most common micronutrient !eficiency $orl!$i!e [7 . It is estimate! that 3/A of pregnant $omen are anemic in lo$- an! lo$er-mi!!le-income countries [> . 'e+eral longitu!inal stu!ies in humans ha+e conclu!e! that foetal or neonatal I(A is associate! $ith beha+ioral !ifficulties in infants an! chil!ren. #hese inclu!e !iminishe! general autonomic response, motor maturity an! self-regulation [C , higher le+els of negati+e emotionality an! lo$er le+els of alertness an! soothability in infants [1 , slo$er neuronal con!uction [0 , $orse learning ability an! memory at 8 to 3 years [92 , an! poorer performance on mental an! psychomotor e+aluations at preschool age [99 . 6n the other han!, .iou- et al. [9/ in a stu!y of >8 mother infant pairs in Cana!a foun! no significant relationship bet$een maternal antenatal iron le+el an! infantZs cogniti+e performance at > months. 'e+eral trials of uni+ersal antenatal iron supplementation ha+e also conclu!e! that it ha! no consistent effect on the intelligence %uotient of chil!ren $hen they $ere four years ol! [98 an! chil! beha+ior at early school age [93 . All of the human stu!ies $ere con!ucte! in $ell-nourishe! populations in high-income countries. 4esi!e iron !eficiency, maternal mental health status may play a crucial role in fetal an! infant !e+elopment. #he common mental !isor!ers &C"() of !epression an! an-iety are estimate! to occur on a+erage in 92A of pregnant $omen in high-income countries an! 9>A &range from 7./ to 8/.0A) in lo$- an! lo$er-mi!!le-income countries [97 . 'tu!ies from high-income countries suggest that antenatal an-iety increases the risk of chil! beha+ioralHemotional problems at four years of age [9> < an! has a negati+e association $ith cogniti+e ability [9C , [91 . *o$e+er, (iPietro et al. [90 foun! that antenatal an-iety an! !epression $ere associate! $ith higher 4ayley II "ental (e+elopment In!escores in chil!ren age! /3 months. #here ha+e been +ery fe$ attempts to e-amine this relationship in !e+eloping countries $here common mental !isor!ers are more pre+alent

an! can interact $ith other !eterminants such as micronutrient !eficiencies $hich are common in these settings but not in !e+elope! countries. ?e coul! fin! only one publishe! stu!y, $hich reporte! no association bet$een symptoms of common mental !isor!ers in mothers in the thir! trimester of pregnancy an! infant cogniti+e !e+elopment at 9/ months of age [/2 . Pietnam is ranke! by the ?orl! 4ank as a lo$er-mi!!le-income country ha+ing only recently been re-classifie! from being a lo$-income country. About 82A of $omen li+ing in rural Pietnam e-perience C"( !uring pregnancy or after gi+ing birth [/9 . Iron !eficiency is the most common micronutrient !eficiency an! anemia affects about 8>.7A pregnant $omen in this setting [// . #he aim of this stu!y $as to test the hypothesis that maternal I(A an! clinically significant symptoms of C"( !uring pregnancy $oul! ha+e a!+erse effects on the cogniti+e !e+elopment of infants at > months of age in rural Pietnam. #he hypothesiDe! mo!el &5igure 9) $as constructe! from international an! Pietnamese e+i!ence. In the mo!el, $e postulate! that maternal antenatal I(A an!Hor C"( $oul! affect infant cogniti+e !e+elopment at si- months +ia both !irect an! in!irect path$ays. (emographic characteristics an! other psychosocial factors are potential confoun!ers of both the e-posures an! the main outcome an! therefore ha! to be inclu!e! in e+ery aspect of the hypothesiDe! mo!el.

Fi ure 1 Conceptual model of the direct and indirect effects of antenatal iron deficiency anemia (IDA) and common mental disorders (CMD) on infant cognitive development.

#he hypothesiDe! !irect path$ay $as that antenatal I(A an!Hor C"( symptoms coul! cause a!+erse con!itions in utero $hich affect fetal neuro!e+elopment an! lea! to lasting changes in cogniti+e function. #he hypothesiDe! in!irect path$ays $ere first that maternal I(A an!Hor C"( !uring pregnancy ha+e a!+erse effects on birth$eight an! infant $eight an! secon! that maternal postpartum C"(, $hich is pre!icte! by antenatal C"(, compromises caregi+ing an! that these me!iate effects on infant cogniti+e !e+elopment. 4irth$eight an! postnatal C"( $ere inclu!e! in this mo!el as the main me!iators !ue to their associations $ith the e-posures of interest [/8 , [/3 an! the outcome [/7 in lo$ an! lo$-mi!!le-income countries.

Methods
)tudy )etting

#his stu!y $as con!ucte! in *a =am, a typical .e! .i+er !elta rural pro+ince in the north of Pietnam locate! appro-imately 72 km south of *anoi. *a =am Pro+ince has a population of 2.1 million inhabitants. In /299 the a+erage annual per capita income $as appro-imately E'(122 an! C.7A of people li+e! belo$ the international po+erty line of E'(9./7 per !ay. "ost births in *a =am are at either commune health centers, or at !istrict an! pro+incial hospitals an! more than 00A of $omen atten! for antenatal care. "ental health is not consi!ere! either in antenatal care or in primary healthcare. #he national iron supplementation program $as implemente! in 900C an! recommen!s that $omen take !aily iron an! folic aci!. At the time of this stu!y neither iron supplements nor io!iDe! salt $as a+ailable for free an! $omen li+ing in the stu!y site $ere e-pecte! to purchase them o+er-the-counter.

)tudy Design
#he stu!y $as a prospecti+e population-base! stu!y in $hich $omen $ere enrolle! !uring pregnancy an! follo$e!, $ith their babies until si- months postpartum. (ata $ere collecte! in four sur+eys bet$een (ecember /220 an! "arch /299. #he first &?a+e 6ne, ?9) $as con!ucte! $hen the $omen $ere recruite! before mi!-pregnancy an! the secon! &?a+e #$o, ?/) $hen participants $ere at least /1 $eeks gestation. After chil!birth t$o assessments of mothers an! infants $ere con!ucte! $hen the babies $ere 1 $eeks &?a+e #hree, ?8) an! > months &?a+e 5our, ?3) ol!.

Participants
Participants $ere recruite! by a t$o-stage sampling proce!ure in $hich first, 72 of the 99> communes in the pro+ince $ere selecte! ran!omly by an in!epen!ent statistician an! secon! all $omen $ho $ere pregnant $ith a single fetus an! bet$een 9/ an! /2 $eeks of gestation in the selecte! communes !uring the enrollment perio! &(ecember /220 to January /292) $ere eligible an! in+ite! to participate [/> .

Data )ources

"nfant cogniti$e de$elopment


Infant cogniti+e !e+elopment &collecte! only at ?3) $as assesse! by !irectly testing the infant using the 4ayley 'cales of Infant an! #o!!ler (e+elopment 8r! E!, Cogniti+e 'cale &4'I() [/C . In the original +ali!ation stu!ies con!ucte! in E', the reliability coefficient of 4'I( for infants at > months $as 2.1C an! the stability coefficient $as 2.C0 [/1 . #here is current no !ata a+ailable on the +ali!ation of the 4'I( for use in Pietnam.

Psycho!social data
"aternal common mental !isor!ers symptoms &C"(, all 3 ?s) $ere assesse! by the E!inburgh Postnatal (epression 'cale-Pietnam Pali!ation &EP('-P) [/0 , [82 . #he EP('-P inclu!es ten fi-e! choice items score! from 2M8 $hich assess !imensions of lo$

moo! o+er the pre+ious se+en !ays an! yiel! a total score from 2 to 82. #he EP('-P $as translate! from the English +ersion, culturally +erifie! an! +ali!ate! against psychiatrista!ministere! 'tructural Clinical Inter+ie$s for ('" IP !iagnoses to establish local cut off scores for pregnant $omen an! those $ho ha+e recently gi+en birth. In this setting, $here emotional literacy is lo$, the cut-off point to !etect clinically significant symptoms $ith a sensiti+ity of C2A an! a specificity of C8A is L3 [82 . Intimate partner physical, se-ual, an! emotional abuse an! history of chil! abuse &?9, ?/, ?3) $ere assesse! by the pregnancy section of the ?*6 "ulti-country stu!y on (omestic Piolence sur+ey [89 . #his %uestionnaire has been !emonstrate! to be appropriate for use in e%ui+alent settings [8/ . 'ocial circumstances an! repro!ucti+e health &?9, ?/, ?8, an! ?3) inclu!e! current non-economic coinci!ental life a!+ersity< %uality of relationship $ith her mother an! mother-in-la$< family support in caring for the baby !uring the !ay an!, at night, an! $ith house$ork< repro!ucti+e health history inclu!ing gra+i!ity, parity, history of spontaneous abortions, fetal or neonatal !eaths< an! $hether or not the pregnancy $as $elcome $ere collecte! by stu!y-specific %uestions $hich ha+e been use! in pre+ious stu!ies in the same settings [/9 . #hese %uestions ha+e been foun! to be comprehensible to participants an! to yiel! interpretable !ata. *ousehol! economic status &?9) $as assesse! by the ?orl! 4ank metho! $hich calculates a *ousehol! ?ealth In!e- from information about 9C househol! characteristics, ser+ices an! !urable assets [88 . "aternal age, marital status, e!ucational le+el, an! occupational status an! security of income $ere collecte! by stu!y-specific %uestions [/9 . 'tu!y-specific structure! %uestions &?8 an! ?3) $ere use! to assess $hether or not a $oman $as offering any breastfee!s to the infant an! her appraisal of $hether the amount of milk $as sufficient to meet her infantZs nee!s.

Biological data
"aternal hemoglobin &?9, ?/) $as e+aluate! in the fiel! from finger stick bloo!, using a hemoglobinometer &*emoCue A4, Angelholm '$e!en). A 8 m: sample of +enous bloo! &?9, ?/) $as taken from a pregnant $oman an! centrifuge! to har+est serum, froDen at Y12[C. 'erum ferritin $as e+aluate! by Chemiluminescent "icroparticle Immuno Assay performe! on the Archicentre ci>/222 instrument &Abbott, Illinois, E'A) at Alfre! Pathology 'er+ices, Alfre! *ealth, Australia. A spot urine sample &?/) $as obtaine!, froDen in a fiel! freeDer an! transporte! in a col! chain to the laboratory of the =ational *ospital of En!ocrinology in *anoi. At the laboratory, urine io!ine concentration $as !etermine! by means of the 'an!ell-Kolthoff reaction, as recommen!e! by ?*6, E=ICE5 an! the ICCI(( [83 .

Infant birth$eights $ere collecte! by maternal reports at ?8. If the mother !i! not remember or $as not sure, the infant birth$eight $as collecte! from or +erifie! against the birth recor!s at the health facility. Infant $eight &?8, ?3) $as measure! by the 'eca 1C> 'cale $hich first measures maternal $eight an! secon! measures the $eight of the infant $hen hel! in her arms.

Procedure
#he 4'I( $ere translate! from English into Pietnamese an! back-translate! to English for +erification [87 . 4'I( a!ministrators $ere psychologists practicing at the TuNa Clinic $hich is a ser+ice associate! $ith the .esearch an! #raining Centre for Community (e+elopment, $ho $ere e-perience! in early chil! !e+elopment assessment. #he a!ministrators $ere traine! by a local e-pert in 4'I( an! successfully complete! a post-training test. #his stu!y use! the original toolkits an! follo$e! the gui!elines of the A!ministration "anual an! the training (P(s strictly [8> . A pilot stu!y $as con!ucte! to check the language an! to stan!ar!iDe the test proce!ure of 4'I( an! psychosocial %uestionnaires before con!ucting this stu!y. Psychosocial !ata $ere collecte! by professional inter+ie$ers of the !ata collection team of the .esearch an! #raining Centre for Community (e+elopment. #he inter+ie$ers complete! a three!ay training course, $hich inclu!e! an intro!uction to the stu!y, !etaile! !iscussion on the %uestionnaires, an! a practice in the real-$orl! situation. As Pietnamese $omen li+ing in rural areas are unfamiliar $ith the completion of selfreport %uestionnaires, all !ata $ere collecte! by face-to-face an! paper-base! inter+ie$s. All inter+ie$s of mother inclu!ing a!ministration of 4'I( $ere con!ucte! in pri+ate rooms at commune health centers.

Data Management and Data Analysis


(ata management an! !escripti+e analyses $ere performe! in '#A#A +ersion 99 &'tataCorp :P, College 'tation, #e-as, E'A). #he total ra$ scores of 4'I( $ere con+erte! to composite scores base! on the gui!elines of 4ayley 8r! E!ition "anual [8> . In the reference population, composite scores ha+e a range of 77 to 937, a mean of 922, an! a stan!ar! !e+iation of 97. #his transformation allo$s for a stan!ar!iDe! measurement, $hich represents in stan!ar! !e+iation units ho$ far an infantZs score is from the mean or a+erage score for infants of that precise age. Composite scores $ere use! in all analyses. InfantZs $eight-for-age S scores &?AS) $ere calculate! from infantZs age, se-, an! $eight by ?*6 Anthro +ersion 8././ &?orl! *ealth 6rganiDation, January /299). #he t$o e-posures of interest, I(A an! C"(, $ere !ichotomiDe! into those $ith an! $ithout clinically significant symptoms. Criteria for I(A $ere *b O99 gH!: an! serum ferritin O97 ngHm: as recommen!e! by the ?*6 an! the Enite! =ations Chil!renZs 5un! [8C . Persistent antenatal I(A $as !efine! as meeting the criteria for I(A in both ?9 an! ?/. #he EP('-P $as +ali!ate! as a screening tool $hich coul! only classify perinatal $omen into t$o groups, those $ith

normal moo! an! those $ith probable C"( &cut-off L3, [82 ). In this setting higher EP('-P scores ha+e not been sho$n !efiniti+ely to be in!icati+e of more se+ere C"(. A family support +ariable $as !eri+e! by summing three +ariablesI ha+ing support in caring for the infant !uring the !ay time &9I yes< 2< no), !uring the night &9I yes< 2< no), an! assistance $ith house$ork &9I yes< 2< no), to yiel! scores ranging from 2 &no support) to 8 &ha+ing all of the three kin!s of support). #he househol!Zs economic situation $as assesse! by a $ealth in!e-, $hich is a pro-y measure constructe! from 9C househol! characteristics, ser+ices an! !urable assets [88 . #he lo$est income group $as the lo$est %uartile of the in!e- an! the highest income $as the highest %uartile. (ata $ere !escribe! by fre%uencies an! percentages &for categorical +ariables)< means an! stan!ar! !e+iations &for normally !istribute! continuous +ariables), an! me!ians an! inter%uartile ranges &for ske$e! !istributions). #he !irect an! in!irect effects on infant 4'I( scores $ere e-amine! simultaneously by Path Analysis performe! in "plus +ersion > &"uth\n ; "uth\n, :os Angeles, Enite! 'tates of America). 6ther potential contributing factors to the outcome an! the me!iators $ere a!!e! into the mo!el base! on the results of uni+ariate analyses an! prior e+i!ence to control for possible confoun!ing. #he comman! F"6(I=(ICE'G in "plus $as use! after the mo!el estimation to check if there $as any significant missing path in the mo!el. #he mo!el $as estimate! using $eighte! least-s%uares an! the probit link function that are robust to non-normality. 6nly one mo!el $as fitte! $hich inclu!e! // manifest +ariables. #here $ere no latent +ariables in the mo!el. All of the +ariables $ere retaine! in the mo!el e+en if non-significant. (ata of all four $a+es $ere use! in this analysis. Pariables $hich $ere collecte! in multiple $a+es such as C"( $ere treate! as separate +ariables an! put into the mo!el in the or!er in $hich they ha! been collecte!. =o multile+el techni%ue $as use! in this analysis. In or!er to e+aluate mo!el fit, $e use! the Chi-'%uare #est of "o!el 5it $ith p +alues greater than 2.27 in!icating a goo! fit, .oot "ean '%uare Error 6f Appro-imation &."'EA) $ith +alues less than 2.27 in!icating a goo! fit, an! #ucker-:e$is In!e- &#:I) an! Comparati+e 5it In!e- &C5I) $ith +alues greater than 2.07 in!icating a goo! fit [81 .

%thics Appro$als
Appro+als to con!uct the stu!y $ere pro+i!e! by the *a =am Pro+incial *ealth (epartment Ethics Committee, the Pietnam "e!ical Association Ethics an! 'cientific Committee an! the Eni+ersity of "elbourneZs *ealth 'ciences *uman Ethics Committee. All participants $ere pro+i!e! $ith an oral an! $ritten plain language !escription of the stu!y an! either signe! a consent form, or those $ho coul! not $rite pro+i!e! a thumbprint or +erbal consent $itnesse! by an in!epen!ent obser+er. #he consent proce!ure $as appro+e! by the ethics committees. Permission to transport biological samples from Pietnam to Australia for analyses $as pro+i!e! by the "inistry of *ealth.

Results

)ample
A total of 7/8 $omen $ere eligible to participate, of $hom 30C &0CA) $ere recruite! an! pro+i!e! !ata at ?9. Complete !ata $ere a+ailable for 8C1 mother an! infant pairs an! inclu!e! in the analyses &see 5igure /). #he !emographic characteristics of the $omen an! anthropometric in!ices of the infants are sho$n in #ables 9 an! an!/./. #here $ere 910H8C1 &72A) $omen $ith clinical symptoms of C"(s at any point of early &?9) or late pregnancy &?/) an! >3H8C1 &9>.0A) $ith persistent clinical symptoms of C"(s at both ?9 an! ?/ pregnancy assessments. 6+erall, 70H8C1 &97.>A) $omen $ere e-perience! I(A at any antenatal stu!y point an! >H8C1 &9.>A) at both points.

Fi ure 2 Numbers of participants and attrition by waves of data collection.

!able 1 eneral characteristics of the !"# pregnant women who attended four waves.

!able 2 $irth outcomes% health and development of !"# infants.

Among the 8C1 $omen, only one participant !i! not initiate breastfee!ing. At about 1 $eeks postpartum, most mothers &891H8C1, 13.9A) percei+e! that they ha! sufficient breast milk to meet the infantZs nee!s. At > months postpartum, 87/ &08.9A) $omen $ere still pro+i!ing their infants $ith at least some breastmilk. #he number of mothers $ho reporte! at ?8 that they !i! not recei+e any support $ith infant care !uring the night $as C7H8C1 &/2.7A), !uring the !ay $as />H8C1 &C.2A), an! $ith househol! $ork $as 97H8C1 &3.9A). *o$e+er at si- months postpartum, these ha! increase! to /23H8C1 &73.9A) at night, 19H8C1 &/9.7A) !uring the !ay, an! 9/8H8C1 &8/.>A) $ith house$ork.

Path Analysis
6+erall, there are // +ariables inclu!e! in the path analysis. (etails of the path mo!el are sho$n in #able 8. #he statistically significant paths are presente! by a graphic mo!el in 5igure 8. All of the fit in!ices are $ithin the range in!icating that the mo!el fits the !ata $ell.

Fi ure 3 &ignificant pathways of the factors of infant $ayley Cognitive scores at ' months of age.

!able 3 (ull )ath analysis model of infant $ayley Cognitive scores at ' months of age.

5igure 8 sho$s that persistent antenatal I(A, antenatal C"( in either early or late pregnancy, an! infant ?AS at 1 $eeks of age ha! !irect effects on infant 4'I( composite scores at > months of age. Infants of mothers $ho ha! persistent antenatal I(A ha! 4'I( scores CC.7A of a normati+e stan!ar! !e+iation lo$er than the other infants &Y99.>/ points< 07A CII Y/8.29 to Y2.//) an! of those $ho ha! e-perience! C"( at any point !uring pregnancy 8/A of a stan!ar! !e+iation cogniti+e score lo$er &Y3.12 points, 07A CII Y0.32 to Y2./2). #ogether these t$o risk factors contribute! to more than one stan!ar! !e+iation re!uction in infant cogniti+e scores at > months of age. 5urther, infants ha+ing one stan!ar! !e+iation of $eight-for-age higher at 1 $eeks ol! ha! 4'I( cogniti+e scores that $ere /.9> points higher at > months of age &07A CII 2./8 to 3.21). #he in!irect path$ays from antenatal I(A an! antenatal C"( to infant cogniti+e !e+elopment +ia infant birth$eight, ?AS score at 1 $eeks, an! maternal postpartum C"( $ere teste! simultaneously in the mo!el. *o$e+er, none of the path$ays $ere statistically significant. #here $ere no associations bet$een the antenatal risk factors an! infant birth$eight or infant ?AS at 1 $eeks of age. #he link bet$een maternal C"( !uring pregnancy an! maternal C"( after chil!birth $as statistically significant. *o$e+er there $as no association bet$een postpartum C"( an! infant cogniti+e !e+elopment.

'e+eral +ariables ha! in!irect associations $ith cogniti+e scores at > monthsI higher infant birth$eight, higher househol! $ealth, an! maternal reports of a sufficient supply of breastmilk $ere in!irectly positi+ely relate! to the infant cogniti+e score. "others age! N82 years an! $ho $ere primiparous ha! infants $ith lo$er cogniti+e scores than the others.

Discussion
#o our kno$le!ge this is the first stu!y to e-amine the effects of both iron status an! common mental !isor!ers among pregnant $omen li+ing in a lo$-income setting on the cogniti+e !e+elopment of their > month ol! infants. #he sample $as systematically recruite! $ith high recruitment an! retention fractions an! all reasons for non-pro+ision of complete !ata $ere kno$n. #he path analysis techni%ues use! in this stu!y !etermine! !irect an! in!irect effects of the probable factors on the main outcome simultaneously [80 . 6ur !ata in!icate that there are !irect a!+erse effects of persistent antenatal I(A an! C"( on the cogniti+e !e+elopment of si- month-ol! infants. *o$e+er, in!irect effects of the risk factors on the outcome +ia infant birth$eight, ?AS an! maternal postpartum C"( $ere not foun!. #here is a bo!y of e+i!ence about the effect of maternal antenatal iron !eficiency $ith or $ithout anemia on a!+erse pregnancy outcomes inclu!ing lo$ birth$eight an! prematurity [32 M[33 . *o$e+er, fe$ stu!ies ha+e in+estigate! the long-term effects of maternal iron !eficiency !uring pregnancy on the neuro!e+elopment of the baby. #o !ate the strongest e+i!ence $as !eri+e! from a longitu!inal stu!y of /C1 American chil!ren [99 , $hich foun! that lo$ umbilical cor! serum ferritin concentrations in term ne$born infants correlate! $ith impaire! cogniti+e function at 7 year of age. *o$e+er, the relationship reporte! in that stu!y might actually reflect a correlation bet$een poor placental functioning an! neuro!e+elopmental outcome rather than a causal link bet$een maternal I(A an! chil! cogniti+e !efects [37 . 6ur stu!y pro+i!es e+i!ence that maternal iron !eficiency !uring pregnancy relates to lo$er infant cogniti+e !e+elopment, through both !irect an! in!irect potential path$ays. 5irst, lo$ maternal iron storage coul! result !irectly in lo$ iron transferre! through the placenta to fulfill the fetusZ nee!s, lea!ing to a!+erse conse%uences for chil! neuro!e+elopment [3> . 'econ!, anemia contributes to maternal fatigue $hich may interfere $ith self-care inclu!ing foo! preparation to ensure a sufficient !ietary intake or participation in health care [/9 . A limitation of our stu!y is that it coul! not !istinguish bet$een these t$o path$ays in the analyses. Antenatal persistent I(A $as not pre+alent in our sample &9.>A). Although point pre+alence of anemia of 78A ha! been reporte! in one prior stu!y in rural Pietnam, there $ere no e-isting !ata about persistent anemiaHI(A on $hich to base our estimates, $hich $ere ne+ertheless higher than $ere foun! in this stu!y [3C . ?e suggest that the fin!ing of this stu!y about the association bet$een e-posure to persistent I(A an! the infant outcome shoul! be interprete! $ith care. 5urther in+estigation to confirm the fin!ing in other settings is $arrante!. #hese !ata strengthen the results of pre+ious stu!ies about the links bet$een maternal antenatal mental health problems an! infant neuro!e+elopment. In the =etherlan!s,

pregnancy-specific an-iety in mi! gestation pre!icte! lo$ infant 4ayley II "ental (e+elopment In!e- &4ayley II "(I) &less than percentile /7) at 1 months ol! &6. 9.9, 07A CI 9.2/M9.91) [31 , but they !i! not state $hat potential confoun!ing factors $ere controlle! for. (ea+e et al [30 reporte! fin!ings from the EK A+on :ongitu!inal 'tu!y of Parents an! Chil!ren in 0/33 $omen an! their chil!ren $hich sho$e! an association bet$een ha+ing EP(' L92 &usual cut-off score in high-income Anglophone countries to !etect clinically significant symptoms in community samples) !uring pregnancy an! risk of !e+elopmental !elay in their infants at 91 months of age &6. 9.83, 07A CI 9.99M9.>/). *o$e+er, they use! the (en+er (e+elopmental 'creening #est to e+aluate chil! !e+elopment, $hich !oes not permit !ifferent !e+elopmental !omains to be assesse!. 4ergman et al. [9C foun! prenatal stressful life e+ents $ere associate! $ith infant 4ayley II "(I at 93 to 90 months ol! in EK &regression coefficient Y8.23 &07A CII Y3.9C to Y9.02). In that stu!y, $hile maternal postpartum !epression $as controlle! for, they !i! not control for infant nutritional status. A limite!-sample-siDe stu!y in 71 Cana!ian $omen $ho $ere pregnant !uring the ice storm obser+e! prenatal stress e-posure accounte! for 99.3A of the +ariation in the to!!lersZ 4ayley II "(I at t$o years of age [72 . =one of the e-isting stu!ies in high-income countries controlle! for the effect of maternal antenatal anemia or other micronutrient !eficiencies. #hese factors may not ho$e+er limit their fin!ings because these con!itions are not pre+alent in the high-income settings. 6ur fin!ings are inconsistent $ith the conclusions of 'er+ili et al [/2 $ho ha+e publishe! the only comparable !ata to !ate in a lo$- or lo$er-mi!!le-income country. #heir Ethiopian stu!y !i! not fin! an association bet$een maternal antenatal mental health problems an! infant cogniti+e functioning. #here $ere se+eral !ifferences of stu!y !esign that coul! ha+e le! to the !ifference $ith our fin!ings. 5irst, the stu!y in Ethiopia !i! not control for I(A $hich $e foun! $as strongly associate! $ith the 4'I( scores. #he pre+alence of anemia in pregnant $omen in Ethiopia has been reporte! to be as high as /0A [79 . 'econ!, 'er+ili et al. use! ra$ 4'I( scores in the analyses, so the scores $ere not a!,uste! to account for the infantZs age. #he norms of 4'I( change for each t$o $eeks of infant age, therefore using ra$ scores intro!uces biases !ue to the age of infants. :astly, the timing of maternal antenatal mental health assessments $as only !one at one time, in late pregnancy in the Ethiopian stu!y, $hile our stu!y collecte! these !ata t$iceI in both early an! late pregnancy. (ata at t$o points of time can sho$ a persistent con!ition !uring pregnancy that coul! ha+e a larger effect siDe than cross-sectional !ata that might !etect an acute con!ition. #he fin!ings of our stu!y $ere also !ifferent from those of other researchers in lo$- an! lo$er-mi!!le-income countries [7/ M[77 $ho foun! that postpartum C"( ha! an a!+erse effect on infant cogniti+e function. #he reason for the !ifference coul! be that our stu!y is the only one $hich takes into account antenatal C"(, a significant confoun!er, $hen e-amining the relationship bet$een postpartum C"( an! infant !e+elopment. #he mechanisms of the influence of maternal antenatal an! postpartum psychological problems on chil! !e+elopment are probably !ifferent. (uring pregnancy, cortisol transferring through the placenta might be the path$ay [7> $hile parenting practices are suggeste! as a me!iator in the postpartum perio! [7C . #he timing of infant

cogniti+e assessment in our stu!y $as > months after birth, $hich might not be a sufficient inter+al to reflect the impact of parental caregi+ing on the cogniti+e functioning of the infant. A pre+ious stu!y sho$s that the effect of parenting %uality on chil! cogniti+e !e+elopment coul! not be !etecte! until chil!ren reache! t$o years of age [71 . "echanisms un!erlying the effects of I(A an!Hor C"( on infant !e+elopment outcome ha+e been !iscusse! $here else [7> , [70 , [>2 . 4oth of those risk factors share the most important path$ay +ia the ele+ation of maternal stress hormones &cortisol) an! placental corticotropin-releasing hormone &C.*). I(A increases norepinephrine concentrations that in turn ele+ates the release of maternal cortisol an! placental C.* [>9 , [>/ $hile C"( trigger ele+ate! maternal cortisol le+els an! placental C.* +ia the acti+ation of maternal hypothalamic-pituitary-a!renal &*PA) a-is [>8 . #he fetus may be e-pose! to maternal cortisol through the maternal-fetal bloo! e-change e+en though the placenta may protect the fetus from the effects of maternal cortisol through increasing a placental hormone, 99]-hy!ro-ysteroi! !ehy!rogenase type /, $hich o-i!iDes cortisol to its inacti+e form, cortisone, [>3 , [>7 . Placental C.*, $hich is i!entical to hypothalamic C.* in structure an! function, is release! into both mother an! fetus an! can act to release cortisol in the fetus [>> . 5etal a!,ustments to this a!+erse con!ition in utero may cause long-lasting changes in physiologic functions that ren!er the brain +ulnerable to cogniti+e !e+elopmental !elay later in life [/ M[3 , [7> . #he fin!ings of this stu!y coul! not pro+e this mechanism, but may support it.

Conclusions
?e ackno$le!ge se+eral limitations. 5irst, the loss to follo$ up $as relati+ely high in this stu!y &990H30C, /8.0A). *o$e+er, high rates of loss to follo$-up $ere common in similar pre+ious longitu!inal stu!ies inclu!ing in 'er+illi et al. >3H/71 &/3.1A), 4uitelaar et al. >2H/82 &/>.9A), 4ergman et al. /32H8>7 &>7.1A), an! :aplante et al. 9C/H//3 &C>.1A). #here $ere no significant !ifferences in the socio-!emographic, psychological, or biological characteristics bet$een $omen $ho pro+i!e! complete !ata an! those $ho $ere not retaine! at the si--month postpartum follo$ up. #he secon! limitation $as that $e use! the EP('-P to !etect maternal C"(. #he EP(' is a screening tool, $hich !oes not yiel! !iagnoses an! !oes not !istinguish bet$een !epression an! an-iety. *o$e+er, in this setting the EP(' clinical cut-off score that $e use! has a high le+el of sensiti+ity &C2A) an! specificity &C8A) $hen +ali!ate! against a !iagnostic psychiatric inter+ie$ to !etect C"( inclu!ing !epression an! an-iety in perinatal $omen in the same setting [82 . ?e ackno$le!ge that the pre+alence of antenatal persistent I(A in our sample $as lo$. In or!er to mitigate the effect of that limitation, $e ha+e use! a robust metho! &?:'"P) $hich recommen!e! for non-normality !ata to estimate the parameters of the mo!el. 5inally, e+en though Path Analysis is an a!+ance! mo!ern statistical techni%ue it makes assumptions about linearity, a!!iti+ity, normality, homoske!asticity, lack of multicollinearity, an! i!entification. #he assumptions $ere all checke! carefully an! there $ere +ery fe$ !i+ergences, but $e ackno$le!ge that any of these can re!uce the +ali!ity of the fin!ings. Potential confoun!ers that $ere not assesse! or inclu!e! in the mo!el might affect the fin!ings [>C .

6+erall, $e belie+e that the strengths of this stu!y &a systematically recruite! an! representati+e sample an! assessment of both biological an! psychosocial factors) out$eigh potential limitations an! that the !ata are reliable an! can be generaliDe! $ith some confi!ence to rural Pietnam an! other !e+eloping countries. #hey in!icate that persistent I(A an! C"( in pregnant $omen may increase the risk of lo$er infant cogniti+e !e+elopment in si--month-ol! infants in these settings. 6ur stu!y contributes to scarce literature on the relationship bet$een maternal antenatal health an! infant !e+elopment in !e+eloping countries to suggest further attentions to this critical perio! in or!er to impro+e cogniti+e function of chil!ren in these settings.

Acknowledgments
#he in+estigators are +ery grateful to the *a =am Pro+incial *ealth (epartment $ho permitte! the stu!y to be un!ertaken in the pro+ince, generously allo$e! !ata collection to occur in the commune health stations an! enable! recruitment of participants. ?e are also grateful to the research staff at the .esearch an! #raining Centre for Community (e+elopment in *anoi an! (epartment of "e!icine &."*H?*), #he Eni+ersity of "elbourne $ho contribute! to stu!y !esign an! un!ertook the !ata collection an! management highly professionally. ?e are grateful to the in!epen!ent statistician (r 6bioha Ekoumunne $ho un!ertook the ran!om selection of communes. ?e appreciate the collaborati+e approach of the =ational Institute of "alariology an! Entomology in *anoi in assisting $ith collection an! storage of bloo! samples for this research. ?e appreciate an! ackno$le!ge especially, the generous contributions of time an! personal information gi+en by the stu!y participants.

*unding )tatement
!"e #$u%y &a# fun%e% by 'u#$ralian (e#earc" )ouncil *i#co+ery Pro,ec$ -ran$ *P0986594. !*! i# #u..or$e% by a /ni+er#i$y of 0elbourne 1n$erna$ional (e#earc" Sc"olar#"i.. !"e fun%er# "a% no role in #$u%y %e#i n2 %a$a collec$ion an% analy#i#2 %eci#ion $o .ubli#"2 or .re.ara$ion of $"e 3anu#cri.$.

Article information
PLoS One. 20134 85967 e74876. Publi#"e% online 2013 Se.$e3ber 23. %oi7 10.13718,ournal..one.0074876 P0)1*7 P0)3781140 !"ac" *uc !ran21222329 :e+erley;'nn :i #24 !uan !ran21 <ulie 'nne Si3.#on25 Sara" =anie"24 !erence *&yer26 an% <ane Fi#"er223 >a3aru%%in Ni?a3i2 E%i$or 1 (e#earc" an% !rainin )en$re for )o33uni$y *e+elo.3en$2 =anoi2 @ie$na3 2 )en$re for Ao3enB# =eal$" -en%er an% Socie$y2 0elbourne Sc"ool of Po.ula$ion an% -lobal =eal$"2 $"e /ni+er#i$y of 0elbourne2 0elbourne2 'u#$ralia 3 <ean =aile# (e#earc" /ni$2 Sc"ool of Public =eal$" an% Pre+en$i+e 0e%icine2 0ona#" /ni+er#i$y2 0elbourne2 'u#$ralia

*e.ar$3en$ of 0e%icine 5(0=8A=62 $"e /ni+er#i$y of 0elbourne2 $"e (oyal 0elbourne =o#.i$al2 0elbourne2 'u#$ralia 5 )en$re for 0olecular2 En+iron3en$al2 -ene$ic C 'naly$ic E.i%e3iolo y2 0elbourne Sc"ool of Po.ula$ion an% -lobal =eal$"2 $"e /ni+er#i$y of 0elbourne2 0elbourne2 'u#$ralia 6 0ur%oc" )"il%renB# (e#earc" 1n#$i$u$e2 (oyal )"il%renB# =o#.i$al2 0elbourne2 'u#$ralia ' a D"an /ni+er#i$y2 PaEi#$an 9 E;3ail7 in%$"ac"8a$8ya"oo.co3 Competing Interests* !"e au$"or# "a+e %eclare% $"a$ no co3.e$in in$ere#$# eFi#$.

Concei+e! an! !esigne! the e-perimentsI #(# 44 ## #( J5. Performe! the e-perimentsI #(# 44 ## '* J5. AnalyDe! the !ataI #(# JA' J5. Contribute! reagentsHmaterialsHanalysis toolsI #(# J5. ?rote the paperI #(# J5. 'ecuring the competiti+e grantI #(# 44 ## #( J5.
(ecei+e% <une 22 20134 'cce.$e% 'u u#$ 62 2013. )o.yri "$ no$ice !"i# i# an o.en;acce## ar$icle %i#$ribu$e% un%er $"e $er3# of $"e )rea$i+e )o33on# '$$ribu$ion Licen#e2 &"ic" .er3i$# unre#$ric$e% u#e2 %i#$ribu$ion2 an% re.ro%uc$ion in any 3e%iu32 .ro+i%e% $"e ori inal au$"or an% #ource are cre%i$e%. 'r$icle# fro3 PLoS ONE are .ro+i%e% "ere cour$e#y of )ublic +ibrary of &cience

References
1. -ran$"a3;0c-re or S2 )"eun G2 )ue$o S2 -le&&e P2 (ic"$er L2 e$ al. 520076 *e+elo.3en$al .o$en$ial in $"e fir#$ 5 year# for c"il%ren in %e+elo.in coun$rie#. !"e Lance$ 3697 60H70. IP0) free ar$icleJ IPub0e%J 2. %e(e nier (;'2 *e#ai S 520106 Fe$al *e+elo.3en$. 1n7 :re3ner <-2 Aac"# !*2 e%i$or#. !"e Ailey;:lacE&ell "an%booE of infan$ %e+elo.3en$. 2n% e%. )"ic"e#$er2 Ae#$ Su##eF7 Ailey;:lacE&ell. 3. :arEer *< 520026 Fe$al .ro ra33in of coronary "ear$ %i#ea#e. !ren%# En%ocrinol 0e$ab 137 364H368. IPub0e%J 4. *e#ai 02 -ayle *2 :abu <2 (o## 0- 520056 Pro ra33e% obe#i$y in in$rau$erine ro&$";re#$ric$e% ne&born#7 3o%ula$ion by ne&born nu$ri$ion. '3erican <ournal of P"y#iolo y (e ula$ory2 1n$e ra$i+e an% )o3.ara$i+e P"y#iolo y 2887 (91H96. IPub0e%J 5. Ki33er3ann 0:2 =urrell (F 520076 Nu$ri$ional iron %eficiency. Lance$ 3707 511H520. IPub0e%J 6. 0cLean E2 )o #&ell 02 E li 12 Ao,%yla *2 %e :enoi#$ : 520096 Aorl%&i%e .re+alence of anae3ia2 A=O @i$a3in an% 0ineral Nu$ri$ion 1nfor3a$ion Sy#$e32 1993H2005. Public =eal$" Nu$ri$ion 127 444H454. IPub0e%J 7. =ernLn%e?;0ar$Mne? )2 )anal# <2 'ran%a N2 (ibo$ :2 E#cribano <2 e$ al. 520116 Effec$# of iron %eficiency on neona$al be"a+ior a$ %ifferen$ #$a e# of .re nancy. Early =u3an *e+elo.3en$ 877 165H169. IPub0e%J

8. Aac"# !*2 Polli$$ E2 )ue$o S2 <acoby E2 )ree%;Dana#"iro = 520056 (ela$ion of neona$al iron #$a$u# $o in%i+i%ual +ariabili$y in neona$al $e3.era3en$. *e+elo.3en$al P#yc"obiolo y 467 141H153. IPub0e%J 9. '3in S:2 Orlan%o 02 E%%in# '2 0ac*onal% 02 0onc?yn#Ei )2 e$ al. 520106 1n u$ero iron #$a$u# an% au%i$ory neural 3a$ura$ion in .re3a$ure infan$# a# e+alua$e% by au%i$ory brain#$e3 re#.on#e. <ournal of Pe%ia$ric# 1567 377H 381. IP0) free ar$icleJ IPub0e%J 10. (i in# !2 0iller N)2 :auer P<2 -eor ieff 0D2 Nel#on )' 520096 Elec$ro."y#iolo ical in%ice# of 3e3ory for $e3.oral or%er in early c"il%"oo%7 i3.lica$ion# for $"e %e+elo.3en$ of recollec$ion. *e+ Sci 127 209H219. IP0) free ar$icleJ IPub0e%J 11. !a3ura !2 -ol%enber (L2 =ou <2 <o"n#$on DE2 )li+er SP2 e$ al. 520026 )or% #eru3 ferri$in concen$ra$ion# an% 3en$al an% .#yc"o3o$or %e+elo.3en$ of c"il%ren a$ fi+e year# of a e. <ournal of Pe%ia$ric# 1407 165H170. IPub0e%J 12. (iouF F02 :elan er;Plour%e <2 Leblanc )P2 @i neau F 520116 (ela$ion#"i. be$&een 3a$ernal *=' an% iron #$a$u# an% infan$#B co ni$i+e .erfor3ance. )ana%ian <ournal of *ie$e$ic Prac$ice an% (e#earc" 727 76. IPub0e%J 13. K"ou S<2 -ib#on ('2 )ro&$"er )'2 :a "ur#$ P2 0aEri%e# 0 520066 Effec$ of iron #u..le3en$a$ion %urin .re nancy on $"e in$elli ence Nuo$ien$ an% be"a+ior of c"il%ren a$ 4 y of a e7 lon ;$er3 follo&;u. of a ran%o3i?e% con$rolle% $rial. '3erican <ournal of )linical Nu$ri$ion 837 1112H1117. IPub0e%J 14. Par#on# '-2 K"ou S<2 S.urrier N<2 0aEri%e# 0 520086 Effec$ of iron #u..le3en$a$ion %urin .re nancy on $"e be"a+iour of c"il%ren a$ early #c"ool a e7 lon ;$er3 follo&;u. of a ran%o3i#e% con$rolle% $rial. :ri$i#" <ournal of Nu$ri$ion 997 1133H1139. IPub0e%J 15. Fi#"er <2 %e 0ello 0)2 Pa$el @2 (a"3an '2 !ran !2 e$ al. 520126 Pre+alence an% %e$er3inan$# of co33on .erina$al 3en$al %i#or%er# in &o3en in lo&; an% lo&er;3i%%le;inco3e coun$rie#7 a #y#$e3a$ic re+ie& :ull Aorl% =eal$" Or an. 907 139H149-. IP0) free ar$icleJ IPub0e%J 16. OB)onnor !-2 =eron <2 -lo+er @ 520026 'n$ena$al anFie$y .re%ic$# c"il% be"a+ioral8e3o$ional .roble3# in%e.en%en$ly of .o#$na$al %e.re##ion. <ournal of $"e '3erican 'ca%e3y of )"il% an% '%ole#cen$ P#yc"ia$ry 417 1470H1477. IPub0e%J 17. :er 3an D2 SarEar P2 OB)onnor !-2 0o%i N2 -lo+er @ 520076 0a$ernal #$re## %urin .re nancy .re%ic$# co ni$i+e abili$y an% fearfulne## in infancy. <ournal of $"e '3erican 'ca%e3y of )"il% an% '%ole#cen$ P#yc"ia$ry 467 1454H1463. IPub0e%J 18. :er 3an D2 SarEar P2 -lo+er @2 OB)onnor !- 520106 0a$ernal .rena$al cor$i#ol an% infan$ co ni$i+e %e+elo.3en$7 3o%era$ion by infan$;3o$"er a$$ac"3en$. :iolo ical P#yc"ia$ry 677 1026H1032. IP0) free ar$icleJ IPub0e%J 19. *iPie$ro <'2 No+aE 0F2 )o#$i an D'2 '$ella L*2 (eu#in SP 520066 0a$ernal .#yc"olo ical %i#$re## %urin .re nancy in rela$ion $o c"il% %e+elo.3en$ a$ a e $&o. )"il% *e+elo.3en$ 777 573H587. IPub0e%J 20. Ser+ili )2 0e%"in -2 =anlon )2 !o3lin#on 02 AorEu :2 e$ al. 520106 0a$ernal co33on 3en$al %i#or%er# an% infan$ %e+elo.3en$ in E$"io.ia7 $"e P;0a0iE :ir$" )o"or$. :0) Public =eal$" 107 693. IP0) free ar$icleJ IPub0e%J 21. Fi#"er <2 !ran !2 La :!2 Drii$3aa D2 (o#en$"al *2 e$ al. 520106 )o33on .erina$al 3en$al %i#or%er# in nor$"ern @ie$ Na37 co33uni$y .re+alence an%

"eal$" care u#e. :ull Aorl% =eal$" Or an 887 737H745. IP0) free ar$icleJ IPub0e%J 22. Na$ional 1n#$i$u$e of Nu$ri$ion8/N1)EF 520116 ' re+ie& of $"e nu$ri$ion #i$ua$ion in @ie$ Na3 2009H2010. =anoi7 0e%ical Public =ou#e. 23. 'llen L= 520006 'ne3ia an% iron %eficiency7 effec$# on .re nancy ou$co3e. '3erican <ournal of )linical Nu$ri$ion 717 1280SH1284S. IPub0e%J 24. -ro$e ND2 :ri% e <'2 -a+in '(2 0el+ille <L2 1yen ar S2 e$ al. 520106 ' 3e$a; analy#i# of %e.re##ion %urin .re nancy an% $"e ri#E of .re$er3 bir$"2 lo& bir$" &ei "$2 an% in$rau$erine ro&$" re#$ric$ion. 'rc"i+e# of -eneral P#yc"ia$ry 677 1012H1024. IP0) free ar$icleJ IPub0e%J 25. AalEer SP2 Aac"# !2 0eeE# -ar%ner <2 Lo?off :2 Aa##er3an -2 e$ al. 520076 )"il% %e+elo.3en$7 ri#E fac$or# for a%+er#e ou$co3e# in %e+elo.in coun$rie#. !"e Lance$ 3697 145H157. IPub0e%J 26. Fi#"er <2 !ran !2 :i # :2 !ran !2 *&yer !2 e$ al. 520116 1o%ine #$a$u# in la$e .re nancy an% .#yc"o#ocial %e$er3inan$# of io%i?e% #al$ u#e in rural nor$"ern @ie$ Na3. :ulle$in of Aorl% =eal$" Or ani?a$ion 897 813H820. IP0) free ar$icleJ IPub0e%J 27. :ayley N 520066 :ayley Scale# of 1nfan$ an% !o%%ler *e+elo.3en$H!"ir% E%i$ion. San 'n$onio7 !O7 =arcour$ '##e##3en$. 28. :ayley N 520066 :ayley Scale# of 1nfan$ an% !o%%ler *e+elo.3en$H!"ir% E%i$ion7 !ec"nical 0anual. San 'n$onio7 !O7 =arcour$ '##e##3en$. 29. )oF <2 =ol%en <2 Sa o+#Ey ( 519876 *e$ec$ion of .o#$na$al %e.re##ion. *e+elo.3en$ of $"e 10;i$e3 E%inbur " Po#$na$al *e.re##ion Scale. :ri$i#" <ournal of P#yc"ia$ry 1507 782H786. IPub0e%J 30. !ran !*2 !ran !2 La :2 Lee *2 (o#en$"al *2 e$ al. 520116 Screenin for .erina$al co33on 3en$al %i#or%er# in &o3en in $"e nor$" of @ie$na37 a co3.ari#on of $"ree .#yc"o3e$ric in#$ru3en$#. <ournal of 'ffec$i+e *i#or%er# 1337 281H293. IPub0e%J 31. -arcia;0oreno )2 <an#en =2 Ell#ber 02 =ei#e L2 Aa$$# ) 520056 A=O 0ul$i;)oun$ry S$u%y on Ao3enB# =eal$" an% *o3e#$ic @iolence ' ain#$ Ao3en7 1ni$ial (e#ul$# on Pre+alence2 =eal$" Ou$co3e# an% Ao3enB# (e.on#e#. -ene+a7 Aorl% =eal$" Or ani?a$ion. 32. Ell#ber 02 <an#en ='2 =ei#e L2 Aa$$# )=2 -arcia;0oreno ) 520086 1n$i3a$e .ar$ner +iolence an% &o3enB# ."y#ical an% 3en$al "eal$" in $"e A=O 3ul$i;coun$ry #$u%y on &o3enB# "eal$" an% %o3e#$ic +iolence7 an ob#er+a$ional #$u%y. Lance$ 3717 1165H1172. IPub0e%J 33. !ran ! 520046 )o33uni$y ba#e% e+i%ence abou$ $"e "eal$" care #y#$e3 in rural @ie$ Na3. Ne&ca#$le'u#$ralia7 /ni+er#i$y of Ne&ca#$le. 297 .. 34. A=O8/N1)EF81))1** 520016 '##e##3en$ of 1o%ine *eficiency *i#or%er# an% 0oni$orin $"eir Eli3ina$ion7 ' -ui%e for Pro ra33e 0ana er#2 A=O8N=*801.1 2n% e%. -ene+a7 Aorl% =eal$" Or ani?a$ion. 35. Laun ani P 520006 Po#$na$al %e.re##ion acro## cul$ure#7conce.$ual an% 3e$"o%olo ical con#i%era$ion#. 1n$erna$ional <ournal of =eal$" Pro3o$ion an% E%uca$ion 387 86H94. 36. :ayley N 520066 :ayley Scale# of 1nfan$ an% !o%%ler *e+elo.3en$H!"ir% E%i$ion7 '%3ini#$ra$ion 0anual. San 'n$onio7 !O7 =arcour$ '##e##3en$. 37. A=O8/N1)EF8/N/ 520016 1ron %eficiency anae3ia2 a##e##3en$2 .re+en$ion an% con$rol7 a ui%e for .ro ra33e 3ana er#. -ene+a7 Aorl% =eal$" Or ani?a$ion.

38. Dline (:2 EbooE# )or.ora$ion. 520116 Princi.le# an% .rac$ice of #$ruc$ural eNua$ion 3o%elin . 0e$"o%olo y in $"e #ocial #cience#. 3r% e%. Ne& GorE7 -uilfor% Pre##. ... F+i2 427 .. 39. Pearl < 520096 )au#ali$y7 3o%el#2 rea#onin 2 an% inference4 2n%2 e%i$or. Ne& GorE7 )a3bri% e /ni+ Pre##. 40. Di%an$o =L2 0o ren 12 Lin%3arE -2 0a##a&e S2 Ny#$ro3 L 520096 (i#E# for .re$er3 %eli+ery an% lo& bir$" &ei "$ are in%e.en%en$ly increa#e% by #e+eri$y of 3a$ernal anae3ia. Sou$" 'frican 0e%ical <ournal 997 98H102. IPub0e%J 41. <aleel (2 D"an ' 520086 Se+ere anae3ia an% a%+er#e .re nancy ou$co3e. <ournal of Sur ery PaEi#$an 51n$erna$ional6 137 147H150. 42. 0ara"a$$a ( 520076 S$u%y of anae3ia in .re nancy an% i$# ou$co3e in Ne.al 0e%ical )olle e !eac"in =o#.i$al2 Da$"3an%u2 Ne.al. Ne.al 0e% )oll < 97 270H274. IPub0e%J 43. -eel"oe% *2 ' a%?i F2 @i##er L2 'blor%e..ey E2 '#are D2 e$ al. 520066 0a$ernal an% fe$al ou$co3e af$er #e+ere ane3ia in .re nancy in rural -"ana. 'c$a Ob#$e$ricia e$ -ynecolo ica Scan%ina+ica 857 49H55. IPub0e%J 44. Lone FA2 >ure#"i (N2 E3anuel F 520046 0a$ernal anae3ia an% i$# i3.ac$ on .erina$al ou$co3e. !ro.ical 0e%icine an% 1n$erna$ional =eal$" 97 486H490. IPub0e%J 45. Fle3in (E 520026 )or% #eru3 ferri$in le+el#2 fe$al iron #$a$u#2 an% neuro%e+elo.3en$al ou$co3e#7 correla$ion# an% confoun%in +ariable#. <ournal of Pe%ia$ric# 1407 145H148. IPub0e%J 46. Lo?off :2 :ear% <2 )onnor <2 :arbara F2 -eor ieff 02 e$ al. . 520066 Lon ; la#$in neural an% be"a+ioral effec$# of iron %eficiency in infancy. Nu$ri$ion (e+ie&# 647 S34H434 %i#cu##ion S72H91. IP0) free ar$icleJ IPub0e%J 47. !rin" L2 *ibley 02 :yle# < 520076 'n$ena$al care .roce%ure# an% infor3a$ion re.or$e% by &o3en in $"ree rural area# of @ie$na3. Sou$"ea#$ '#ian <ournal of !ro.ical 0e%icine an% Public =eal$" 387 927H935. IPub0e%J 48. :ui$elaar <D2 =ui?inE ')2 0ul%er E<2 %e 0e%ina P-2 @i##er -= 520036 Prena$al #$re## an% co ni$i+e %e+elo.3en$ an% $e3.era3en$ in infan$#. Neurobiolo y of ' in 24 Su..l 17 S53H604 %i#cu##ion S67H58. IPub0e%J 49. *ea+e !2 =eron <2 E+an# <2 E3on% ' 520086 !"e i3.ac$ of 3a$ernal %e.re##ion in .re nancy on early c"il% %e+elo.3en$. :<O- 1157 1043H1051. IPub0e%J 50. La.lan$e *P2 :arr (-2 :rune$ '2 -albau% %u For$ -2 0eaney 0L2 e$ al. 520046 S$re## %urin .re nancy affec$# eneral in$ellec$ual an% lan ua e func$ionin in "u3an $o%%ler#. Pe%ia$ric (e#earc" 567 400H410. IPub0e%J 51. -ib#on (S2 'bebe G2 S$abler S2 'llen (=2 Ae#$co$$ <E2 e$ al. 520086 Kinc2 ra+i%a2 infec$ion2 an% iron2 bu$ no$ +i$a3in :;12 or fola$e #$a$u#2 .re%ic$ "e3o lobin %urin .re nancy in Sou$"ern E$"io.ia. !"e <ournal of nu$ri$ion 1387 581H586. IP0) free ar$icleJ IPub0e%J 52. Pa$el @2 *e#ou?a N2 (o%ri ue# 0 520036 Po#$na$al %e.re##ion an% infan$ ro&$" an% %e+elo.3en$ in lo& inco3e coun$rie#7 ' co"or$ #$u%y fro3 -oa2 1n%ia. 'rc"i+e# of *i#ea#e in )"il%"oo% 887 34H37. IP0) free ar$icleJ IPub0e%J 53. :lacE 002 :aNui '=2 Ka3an D2 0cNary SA2 Le D2 e$ al. 520076 *e.re##i+e #y3.$o3# a3on rural :an la%e#"i 3o$"er#7 i3.lica$ion# for infan$ %e+elo.3en$. <ournal of )"il% P#yc"olo y an% P#yc"ia$ry an% 'llie% *i#ci.line# 487 764H772. IPub0e%J

54. =a%ley )2 !e e n '2 !e##e3a F2 '#efa 02 -alea S 520086 Paren$al #y3.$o3# of co33on 3en$al %i#or%er# an% c"il%renB# #ocial2 3o$or2 an% lan ua e %e+elo.3en$ in #ub;Sa"aran 'frica. 'nnal# of =u3an :iolo y 357 259H275. IPub0e%J 55. -aller <2 =arri#on (2 (a3#&y F2 For%e @2 :u$ler S 520006 0a$ernal %e.re##i+e #y3.$o3# affec$ infan$ co ni$i+e %e+elo.3en$ in :arba%o#. <ournal )"il% P#yc"olo y an% P#yc"ia$ry 417 747H757. IPub0e%J 56. !al e N02 Neal )2 -lo+er @ 520076 'n$ena$al 3a$ernal #$re## an% lon ; $er3 effec$# on c"il% neuro%e+elo.3en$7 "o& an% &"yP <ournal of )"il% P#yc"olo y an% P#yc"ia$ry an% 'llie% *i#ci.line# 487 245H261. IPub0e%J 57. 0urray L2 =alli an S2 )oo.er P 520106 Effec$# of .o#$na$al %e.re##ion on 3o$"er;infan$ in$erac$ion# an% c"il% %e+elo.3en$. 1n7 :re3ner <-2 Aac"# !*2 e%i$or#. Ailey;:lacE&ell "an%booE# of %e+elo.3en$al .#yc"olo y. 2n% e%. )"ic"e#$er2 Ae#$ Su##eF7 Ailey;:lacE&ell. 58. Lu o;-il <2 !a3i#;Le0on%a )S 520086 Fa3ily re#ource# an% .aren$in Nuali$y7 LinE# $o c"il%renB# co ni$i+e %e+elo.3en$ acro## $"e fir#$ 3 year#. )"il% %e+elo.3en$ 797 1065H1085. IPub0e%J 59. Fiel% ! 520116 Prena$al %e.re##ion effec$# on early %e+elo.3en$7 a re+ie&. 1nfan$ :e"a+ *e+ 347 1H14. IPub0e%J 60. 'llen L= 520016 :iolo ical 3ec"ani#3# $"a$ 3i "$ un%erlie ironB# effec$# on fe$al ro&$" an% .re$er3 bir$". < Nu$r 1317 581SH589S. IPub0e%J 61. *all3an P( 519866 :ioc"e3ical ba#i# for $"e 3anife#$a$ion# of iron %eficiency. 'nnu (e+ Nu$r 67 13H40. IPub0e%J 62. -ul3e?o lu '02 0a"o3e% D2 =of3eyr -<2 NiEo%e3 @)2 Dra3er ! 519966 Fe$al an% 3a$ernal ca$ec"ola3ine le+el# a$ %eli+ery. < Perina$ 0e% 247 687H 691. IPub0e%J 63. Aa%"&a P* 520056 P#yc"oneuroen%ocrine .roce##e# in "u3an .re nancy influence fe$al %e+elo.3en$ an% "eal$". P#yc"oneuroen%ocrinolo y 307 724H 743. IPub0e%J 64. 0ul%er E<2 (oble# %e 0e%ina P-2 =ui?inE ')2 @an %en :er " :(2 :ui$elaar <D2 e$ al. 520026 Prena$al 3a$ernal #$re##7 effec$# on .re nancy an% $"e 5unborn6 c"il%. Early =u3 *e+ 707 3H14. IPub0e%J 65. :ro&n (A2 *ia? (2 (ob#on ')2 Do$ele+$#e+ G@2 0ullin# <<2 e$ al. 519966 !"e on$o eny of 11 be$a;"y%roFy#$eroi% %e"y%ro ena#e $y.e 2 an% 3ineralocor$icoi% rece.$or ene eF.re##ion re+eal in$rica$e con$rol of lucocor$icoi% ac$ion in %e+elo.3en$. En%ocrinolo y 1377 794H797. IPub0e%J 66. Pe$ra lia F2 Florio P2 Na..i )2 -ena??ani '( 519966 Pe.$i%e #i nalin in "u3an .lacen$a an% 3e3brane#7 au$ocrine2 .aracrine2 an% en%ocrine 3ec"ani#3#. En%ocrine (e+ie&# 177 156H186. IPub0e%J 67. @an%erAeele !< 520126 1n+i$e% co33en$ary7 #$ruc$ural eNua$ion 3o%el# an% e.i%e3iolo ic analy#i#. '3erican <ournal of E.i%e3iolo y 1767 608H612. IP0) free ar$icleJ IPub0e%J

3 PLoS ONE Public Library of Science

Risky Beha$ior in +ambling ,asks in "ndi$iduals with AD-D . A )ystematic /iterature Re$iew
V+onne Broen, Beral!ina 5. Baastra, [... , an! 6li+er #ucha A!!itional article information

Abstract
'b(ecti$e
#he aim of this re+ie$ $as to gain insight into the relationship bet$een Attention !eficit hyperacti+ity !isor!er &A(*() an! risky performance in gambling tasks an! to i!entify any potential alternate e-planatory factors.

Methods
PsycI=56, Pub"e!, an! ?eb of Kno$le!ge $ere searche! for rele+ant literature comparing in!i+i!uals $ith A(*( to normal controls &=Cs) in relation to their risky performance on a gambling task. In total, fourteen stu!ies in chil!renHa!olescents an! ele+en stu!ies in a!ults $ere inclu!e! in the re+ie$.

Results
*alf of the stu!ies looking at chil!renHa!olescents $ith A(*( foun! e+i!ence that they run more risks on gambling tasks $hen compare! to =Cs. 6nly a minority of the stu!ies on a!ults $ith A(*( reporte! aberrant risky beha+ior. #he effect siDes range! from small to large for both age groups an! the outcome pattern !i! not !iffer bet$een stu!ies that applie! an implicit or e-plicit gambling task. #$o stu!ies !emonstrate! that comorbi! oppositional !efiant !isor!er &6(() an! con!uct !isor!er &C() increase! risky beha+ior in A(*(. :imite! an!Hor inconsistent e+i!ence $as foun! that comorbi! internaliDing !isor!ers &I(s), A(*( subtype, methylpheni!ate use, an! !ifferent forms of re$ar! influence! the outcomes.

Conclusion
#he e+i!ence for increase! risky performance of in!i+i!uals $ith A(*( on gambling tasks is mi-e!, but is stronger for chil!renHa!olescents $ith A(*( than for a!ults $ith A(*(, $hich may point to !e+elopmental changes in re$ar! an!Hor penalty sensiti+ity or a publication bias for positi+e fin!ings in chil!renHa!olescents. #he literature suggests that comorbi! 6((HC( is a risk factor in A(*( for increase! risky beha+ior. Comorbi!

I(s, A(*( subtype, methylpheni!ate use, an! the form of re$ar! recei+e! may affect risky performance in gambling tasks< ho$e+er, these factors nee! further e-amination. 5inally, the implications of the fin!ings for A(*( mo!els an! the ecological +ali!ity of gambling tasks are !iscusse!.

"ntroduction
Attention !eficit hyperacti+ity !isor!er &A(*() is characteriDe! by attentional problems, hyperacti+ity, an! impulsi+ity [9 . 4ase! on these symptoms, three A(*( subtypes can be !istinguishe!I the A(*( combine! type &A(*(-C), the A(*( inattenti+e type &A(*(-I), an! the A(*( hyperacti+e-impulsi+e type &A(*(-*). #he pre+alence of A(*( in the general population has been estimate! at 7.8A for in!i+i!uals belo$ 91 years of age an! at 3.3A for a!ults [/ , [8 . A(*( symptoms often !ecline !uring a!olescence &remittent A(*(), therefore, only a portion of chil!ren $ith A(*( still meet all the ('"-IP criteria for A(*( $hen they reach a!ulthoo! &persistent A(*() [9 , [3 . In!i+i!uals $ith A(*( ha+e often been foun! to suffer from comorbi! con!itions, inclu!ing oppositional !efiant !isor!er &6((), con!uct !isor!er &C(), an! internaliDing !isor!ers &I(s) such as an-iety an! moo! !isor!ers [7 . In general, in!i+i!uals $ith A(*( ten! to be in+ol+e! in a greater proportion of risky situations an! beha+iors in e+ery!ay life than in!i+i!uals $ithout A(*(. 'pecifically, those $ith A(*( ten! to !emonstrate more !angerous !ri+ing beha+ior [> M[1 , increase! in+ol+ement in traffic acci!ents [0 , [92 , increase! criminality [99 , [9/ , more risky se-ual beha+ior [98 , [93 , an! increase! !rug abuse [97 . In a!!ition, in their metaanalytic re+ie$, :ee, *umphreys, 5lory, :iu, ; Blass [9> conclu!e! that chil!hoo! A(*( $as a risk factor for the !epen!ence on, an! abuse of, nicotine, alcohol, marihuana, an! cocaine later in life. In!i+i!uals $ith A(*( also ha+e an increase! chance to !e+elop problem or pathological gambling, especially in!i+i!uals $ith A(*(C [9C , in!i+i!uals $ith se+ere A(*( symptoms [91 , or in!i+i!uals $ith persistent A(*( [90 . #he relationship bet$een A(*( an! risky beha+ior may be e-plaine! by e-ecuti+e !ysfunctioning, in particular inhibition !eficits, that for many years ha+e been the focus of A(*( mo!els [/2 , [/9 . In these mo!els, it is assume! that risky beha+ior in A(*( is cause! by impaire! impulse control !ue to !eficiencies in inhibition of prepotent responses, interference control, an! the stopping of ongoing responses after fee!back on errors. "ore recently, some mo!els of A(*( ha+e also incorporate! moti+ational !eficits as the core problem in A(*( [// M[/3 , $hich are characteriDe! by an aberrant le+el of sensiti+ity to re$ar!s an! penalties [/7 . 4oth beha+ioral stu!ies an! animalmo!els ha+e suggeste! that chil!ren $ith A(*( ha+e a greater preference for imme!iate o+er !elaye! re$ar!s compare! to normal controls. #his increase! orientation to$ar!s imme!iate re$ar!s is pre!icte! by mo!els such as the (ual Path$ay "o!el &(P") [// , [/> , the (ynamic (e+elopmental #heory &((#) [/3 , an! the (opamine #ransfer (eficit #heory &(#() [/8 . #he (P" proposes that !isturbances in at least t$o in!epen!ent neural circuits can lea! to A(*(. 'pecifically the +entrolateral an! !orsolateral cortico-striatal circuitry, $hich subser+es e-ecuti+e processes, an! the

mesolimbic &me!ial-prefrontal an! orbitofrontal) +entral striatal circuitry, $hich subser+es moti+ational processes. (isturbances in the former circuitry gi+e rise to cogniti+e an! beha+ioral !ysregulation, $hereas !isturbances in the latter gi+e rise to !elay a+ersion, resulting in relati+ely strong preferences for smaller imme!iate re$ar!s o+er larger !elaye! re$ar!s. #he ((# an! (#( are both base! on the assumption that A(*( is associate! $ith a !ysfunction of the mi!brain !opamine system &although the e-act mechanism propose! !iffers bet$een the mo!els) an! not only pre!ict an increase! preference for imme!iate o+er !elaye! re$ar!s, but also pre!icts that chil!ren $ith A(*( nee! fre%uent reinforcement to learn optimally, sho$ impaire! learning in response to reinforcement, an! sho$ an impaire! integration of earlier e-periences of reinforcement $hen planning an! carrying out beha+iors. 'e+eral other mo!els pre!ict that chil!ren $ith A(*( also suffer from a re!uce! sensiti+ity to punishment or nonre$ar!, $hich makes them more focuse! on re$ar!ing stimuli than chil!ren $ithout A(*( [/C M[/0 . *o$e+er, there is also e+i!ence for re!uce! psychophysiological sensiti+ity to re$ar!s an! penalty in in!i+i!uals $ith A(*( [82 M[8/ , but accor!ing to the literature re+ie$ by :uman an! colleagues these results are inconsistent [/7 . #his inconsistency in research fin!ings is presumably cause! by the many factors that influence !ecision-making in A(*(, such as characteristics of the in!i+i!uals an! the a!opte! task para!igm. #o gain more insight into the relationship bet$een A(*( an! risky beha+ior, cogniti+e tasks $ith a gambling component can be use! to in+estigate the risky beha+ior of in!i+i!uals $ith A(*(. In gambling tasks, participants can usually choose bet$een se+eral options that !iffer in the chance for a re$ar! or penalty. #he e-act probability !istribution of the outcome can be e+i!ent for the participant &e-plicit) or not &implicit). E-amples of implicit gambling tasks are the 4alloon Analogue .isk #ask &4A.#) [88 , the Car! Playing #ask &C#) [83 , the (oor 6pening #ask &(6#) [87 , an! the Io$a Bambling #ask &IB#) [8> &see "etho!s section for a more !etaile! !escription of implicit gambling tasks). ?ith regar! to the IB#, $hich is one of the most often use! para!igms, t$o phases of !ecision-making can be !istinguishe! [8C . In the initial phase, the conse%uences of the !ecision are completely un!efine! an! participants !o not ha+e any information about ho$ likely positi+e or negati+e conse%uences $ill appear an!, therefore, !ecision-making in this phase is calle! ambiguous. In the secon! phase, ho$e+er, participants ha+e some abstract kno$le!ge of the conse%uences an! the associate! probabilities of their choices. (ecisions in this phase are commonly referre! to as ^!ecisions un!er riskZ. In e-plicit gambling tasks, the e-act probability of recei+ing a re$ar! or penalty is ma!e e-plicit or can easily be !e!uce!, an! !ecisions on these types of tasks are also consi!ere! to be un!er risk. E-amples of e-plicit gambling tasks are the Cambri!ge Bambling #ask &CB#) [81 , the Bame of (ice #ask &B(#) [80 , the "ake-a"atch Bame &""B) [32 , an! the Probabilistic (iscounting #ask &P() [39 &see "etho!s section for a more !etaile! !escription of e-plicit gambling tasks). Implicit an! e-plicit gambling tasks aim to measure !ifferent types of !ecision-making. Implicit gambling tasks are thought to !epen! on ^hotZ !ecision-making in+ol+ing emotional an! affecti+e responses to the options of choice as $ell as on ^col!Z !ecisionmaking in+ol+ing the rational an! cogniti+e !eterminations of risks an! benefits

associate! $ith the options in the later stages of the task [3/ , [38 . E-plicit gambling tasks, ho$e+er, are more focuse! on ^col!Z !ecision-making strategies because the kno$le!ge of the probability !istributions can be use! to rationally !etermine the risks an! benefits of the options right from the start of the task. Accor!ing to the (ual-'ystem E-planation risky beha+ior is the result of a competition bet$een ^hotZ an! ^col!Z !ecision-making processes that are subser+e! by, respecti+ely, a phylogenetically ol!er affecti+e-moti+ational system &comprise! of subcortical an! cortical mi!brain !opamine systems) an! a phylogenetically younger !eliberati+e cogniti+e control system &comprise! of the !orsal an! +entral lateral prefrontal corte- an! the posterior parietal corte-) [33 M[3> . "aking a !istinction bet$een implicit an! e-plicit gambling tasks may allo$ for conclusions on the type of !ecision-making that is impaire! in A(*( an! the un!erlying systems that are affecte!. 'tu!ies on the gambling task performance of in!i+i!uals $ith A(*( sho$ mi-e! results, $hich may be cause! by the use of !ifferent task para!igms an!Hor by sample characteristics. #herefore, the aim of this re+ie$ is to gain more insight into the relationship bet$een A(*( an! risky !ecision-making on gambling tasks in e-isting research, an! to i!entify any alternate e-planatory factors that coul! ha+e influence! the outcomes presente! in the literature. 4ase! on the increase! sensiti+ity to imme!iate re$ar!s an! !ecrease! sensiti+ity to penalties pre!icte! by moti+ational mo!els of A(*(, it is hypothesiDe! that in!i+i!uals $ith A(*( $ill !isplay more risky beha+ior in gambling tasks than in!i+i!uals $ithout A(*(. 'pecifically, purely moti+ational mo!els [/8 , [/3 , [/C M[/0 $oul! pre!ict that risky beha+ior is increase! on especially implicit gambling tasks because these tasks strongly !epen! on both ^hotZ an! ^col!Z !ecision-making, $hich are, respecti+ely, un!erpinne! by affecti+e-moti+ational an! cogniti+e control systems. .isky beha+ior in e-plicit gambling tasks, $hich mostly !epen! on ^col!Z !ecision-making, $oul!, ho$e+er, be less e+i!ent !ue to the assume! reliance on mostly cogniti+e control in e-plicit tasks. "oreo+er, as reinforcement learning is an important component of implicit gambling tasks, the ((# an! (#( $oul! pre!ict re!uce! performance in in!i+i!uals $ith A(*( on specifically this type of task. Purely cogniti+e mo!els [/2 , [/9 an! combine! cogniti+e-moti+ational mo!els [// , [/> on the other han! $oul! pre!ict increase! risky beha+ior on both implicit an! e-plicit gambling tasks, because both types of tasks rely on the cogniti+e control systems that are pre!icte! to be impaire! in A(*( by these mo!els. #he literature $as searche! for stu!ies that compare! in!i+i!uals $ith A(*( to normal controls &=Cs) concerning their risky performance on a gambling task. A neuropsychological approach $as taken by only inclu!ing stu!ies using stan!ar!iDe! tasks an! e-perimentally controlle! metho!s. 5urthermore, non-e-perimental stu!ies that e-amine! !ecision-making in e+ery!ay life $ere outsi!e the scope of this re+ie$. #he stu!ies inclu!e! $ere searche! for the follo$ing alternate e-planatory +ariablesI the type of gambling task, comorbi!ity &6((HC( an! I(s), methylpheni!ate &"P*) use, the form of re$ar! use!, an! the !emographic characteristics of the participants &age, se-, an! intelligence an!Hor e!ucation le+el).

Methods

)tudy )election Procedure


#his systematic literature re+ie$ $as carrie! out accor!ing to the gui!elines of Preferre! .eporting Items for 'ystematic .e+ie$s an! "eta-Analyses &P.I'"A) &see #able '9 for a complete! checklist of the P.I'"A gui!elines for this stu!y). =o protocol e-ists for this re+ie$. #he stu!y selection process is summariDe! in 5igure 9. #he literature $as searche! in PsycI=56, Pub"e!, an! ?eb of Kno$le!ge inclu!ing all of the a+ailable literature up until the !ate of June 9, /29/. #he key$or!s ^A(*(Z or ^attention !eficit hyperacti+ity !isor!erZ $ere combine! $ith key$or!s relate! to gambling, such as ^riskZ, ^gamblingZ, ^re$ar!Z, ^punishmentZ, ^!ecision-makingZ an! ^probabilistic !iscountingZ. #he follo$ing selection criteria $ere use! for the inclusion of stu!iesI &a) the stu!y $as $ritten in English< &b) the inclusion of both an A(*( sample an! a sample of =Cs< &c) a cogniti+e task $ith a gambling component $as use!< &!) the performance on the applie! gambling task $as measure! in terms of risky performance. Criterion &!) means that stu!ies that only reporte! reaction times or biologicalHphysiological measures $ere e-clu!e! from this re+ie$. #he reference lists of the initial stu!ies $ere then use! to trace other rele+ant stu!ies. After the completion of the search /7 stu!ies publishe! bet$een 9009 an! /29/ $ere inclu!e! in the re+ie$ &see #able '/ for an o+er+ie$ of these stu!ies).

Fi ure 1 (low diagram depicting the selection of studies according to the guidelines of )referred ,eporting Items for &ystematic ,eviews and Meta-Analyses (),I&MA).

"dentified +ambling ,asks and 'utcome Measures


#he stu!ies in this re+ie$ all use! one or more of the follo$ing gambling tasksI the IB# or a +ariant of the IB#, the C#H(6#, or the 4A.# as implicit gambling tasks, an! the CB#, B(#, ""B, or P( as e-plicit gambling tasks. #he i!entifie! gambling tasks are !escribe! belo$.

"mplicit gambling tasks and outcome measures


#he Balloon Analogue Risk Task (BART) [88 $as !e+elope! to simulate risky beha+ior in e+ery!ay life. .isky beha+ior is reinforce! until an implicit point in time, at $hich further riskiness results in poorer outcomes. In the 4A.#, the sub,ect is instructe! to pump up a series of 02 balloons. ?ith e+ery pump the siDe &magnitu!e) of the balloon +isibly increases an! a fi-e! amount of money is !eposite! in a temporary bank, $hich is

in+isible to the sub,ect. *o$e+er, the balloons $ill e-plo!e after an unkno$n an! +ariable number of pumps. After an e-plosion the money in the temporary bank $ill be lost an! the ne-t empty balloon $ill be presente!. #he sub,ect can pre+ent an e-plosion by stopping the pump in time. #he money in the temporary bank $ill then be transferre! to the permanent bank, $hich is +isible for to sub,ect. #he goal is to earn as much money as possible in the permanent bank. E-amples of outcome measures are the total number of pumps, the number of pumps on the non-e-plo!e! trials &a!,uste! number of pumps), an! the number of e-plo!e! balloons. #he punishment sensiti+ity can be measure! by subtracting the number of pumps on the trial follo$ing an e-plo!e! balloon from the number of pumps prece!ing an e-plo!e! balloon &post e-plosion reacti+ity). #he Card Playing Task (CPT) [ ! and "oor #pening Task ("#T) [87 $ere originally !e+elope! as a response perse+eration task, but also contain a gambling component. In the C#, car!s are se%uentially presente! on a screen &ma-imally 922 car!s), $ith a pre!efine! or!er of face car!s an! number car!s. #he face car!s sho$ a fi-e! re$ar! an! number car!s sho$ a fi-e! penalty. Enbekno$nst the sub,ect, the chance for recei+ing a penalty &number car!) increases by 92A after each block of 92 car!s, starting at 92A an! then rising by 92A e+ery 92 car!s until it reaches 922A. #he sub,ect starts $ith a specifie! stake an! may !eci!e on each trial to play the car! or to %uit the $hole game. 4oth %uitting too soon an! playing too long $ill result in a suboptimal outcome. #he C# has se+eral outcome measures. #he total number of playe! car!s or the number of playe! car!s after the optimal inter+al &number of responses) are regar!e! as a measure for response perse+eration, but may also be use! as a measure of risky performance. #he financial outcome reflects suboptimal !ecision-making !ue to early %uitting or perse+eration. #he (6# uses the same principle as the C#. *o$e+er, !oors instea! of car!s are presente! that hi!e a happy face &re$ar!) or a sa! face &penalty). #he $o%a &am'ling Task ($#(A) [8> $as !e+elope! to simulate real-life !ecisionmaking un!er uncertainty. #he sub,ects are instructe! to ma-imiDe their gain by making 922 choices &i.e. selections of car!s) from four !ifferent !ecks of car!s. #hey are allo$e! to s$itch !ecks after each selection. #he sub,ect recei+es a starting amount of, usually, ficti+e money an! recei+es a re$ar! for each car! that is pulle!, $ith the e-ception of some car!s $hich penaliDe the sub,ect. ?hile a re$ar! results in a gain of money, penalties take money a$ay from the sub,ect. 6n each trial, the amount of money gaine! or lost is presente! on the screen. #he four !ecks !iffer in the magnitu!e of the re$ar! an! in the magnitu!e an! fre%uency of the penalty. Enbekno$nst to the participant, the re$ar!Hpenalty sche!ule of the car!s is pre!efine! &see #able 9). (ecks A an! 4 are regar!e! as the risky !isa!+antageous !ecks, because consistent car! selection from these !ecks $ill lea! to a net loss. (ecks C an! ( are regar!e! as the safe an! a!+antageous !ecks, because consistent car! selection from these !ecks lea!s to a net gain. (ecks A an! C !eli+er fre%uent small penalties, $hereas !ecks 4 an! ( !eli+er infre%uent large penalties. 'e+eral outcome measures can be compute! for the IB#, such as the number of choices for each separate !eck, the number of safe choices, the number of risky choices, an! the financial outcome. #he outcome measure that is most use! an! reflects risky performance is the ^net scoreZ, $hich is !efine! as the number of selecte! car!s from the a!+antageous !ecks minus those from the !isa!+antageous !ecks [&CJ() M &AJ4) . In

or!er to chart the sub,ectsZ learning effects or strategies, the outcome measures are often compute! for each block of /2 trials. 'e+eral alternate +ariants of the IB# ha+e also been !e+elope!.

!able 1 .he classic reward/penalty schedule of the Iowa ambling .as0 1!'2 for 34 successive card selections from the ris0y/disadvantageous dec0s A and $% and the safe/advantageous dec0s C and D.

%&plicit gambling tasks and outcome measures


In the Cam'ridge &am'ling Task (C&T) [81 a line of ten re! an! blue bo-es is presente! on a screen, in $hich the number of re! or blue bo-es is !iffers each trial &$ith ratios of 0 9, 1 /, C 8, an! > 3). #he aim is to guess $hich color of bo- hi!es the re$ar!. #he sub,ects start $ith a stake of points an! may, on each of C/ trials, bet on one color by selecting a proportion of their stake &$hich is also presente! on the screen). #he right color choice is re$ar!e! $ith the number of points bet, $hereas the $rong color choice is penaliDe! $ith the same number of points bet. 'e+eral outcome measures can be compute!. #he %uality of the performance is assesse! by the proportion of trials $here the ma,ority color is chosen &rational choices). .isky beha+ior is represente! by the o+erall proportion bet &amount bet) an! risk a!,ustment is the rate at $hich sub,ects increase the bet proportion in response to more fa+orable ratios of re!Iblue bo-es, $ith lo$er scores being !isa!+antageous. #he &ame o) "ice Task (&"T) [80 is a computeriDe! task in $hich a +irtual !ie is thro$n 91 times. #he aim of the task is to ma-imiDe your money by betting on the !ie outcome. 'ub,ects can bet on one single !ie outcome $ith a possible re$ar! of 9222 &9 > chance), or on a combination of t$o, three or four !ifferent !ie outcomes $ith the respecti+e re$ar!s of 722 &/ > chance), /22 &8 > chance) an! 922 &3 > chance). ?rong bets lea! to a penalty of the same magnitu!e as the possible re$ar! &i.e. 9222, 722, /22, or 922). #he options $ith three an! four !ice are regar!e! as the safe options, $hereas the options $ith one or t$o !ice are regar!e! as risky. 'e+eral outcome measures can be compute! for the B(#, such as the number of choices for each separate option, the number of safe choices, the number of risky choices, an! the financial outcome. #he most often use! outcome measure is the net score, $hich is !efine! as the number of safe choices minus the number of risky choices. #he "ake-a-"atch Bame &""B) [32 is a probabilistic !iscounting task that can be easily un!erstoo! by chil!ren. #he aim of this computeriDe! task is to fin! the copy of a target car! in a line of car!s $ith their faces hi!!en &similar to the game memory). 6n

each of the 9/ trials, the sub,ects may choose from a set of t$o, three or four car!s $ith the respecti+e re$ar!s of one &9 / chance), t$o &9 8 chance), or three &9 3 chance) can!ies $hen the correct car! is chosen. Choosing the $rong car! lea!s to a re$ar! omission, but not to a !irect penalty. #he outcome measure is the number of choices for the three separate options or the number of can!ies recei+e!. #he Pro'a'ilistic "iscounting Task (P") [39 aims to measure the !egree to $hich the sub,ecti+e +alue of a large re$ar! !ecreases $hen the probability of obtaining it !ecreases. :ess !iscounting of the +alue of lo$ probable &uncertain) re$ar!s is relate! to risky choices. In the P(, sub,ects may choose on each of 9/2 trials bet$een a small certain an! a large uncertain re$ar!. #he magnitu!e of the certain re$ar! +aries from 2 to 92 cents &2, /, 3, >, 1, an! 92), $hile the chance to recei+e it is constantly 922A. #he magnitu!e of the uncertain re$ar! is constant &92 cents), an! +aries in probability from 2 to 9 &2,./7,.72,.C7, an! 9). 5or e+ery trial, the options are !epicte! by t$o piggy banks each containing a %uantity of money. #he probability of obtaining the re$ar! is represente! by the thickness of the piggy bankZs shell, an! by a colore! bar, in $hich re! in!icates the thickness of the shell. Pushing the button of the preferre! piggy bank acti+ates a hammer that hits it. If the piggy bank breaks, the sub,ect recei+es the %uantity of money in the piggy bank. #he sub,ecti+e +alues of the probabilistic re$ar!s &$hich is al$ays 92 cents) can be calculate! for e+ery probabilistic le+el. #he sub,ecti+e +alue of the probabilistic re$ar! is !efine! as the magnitu!e of the small certain re$ar! for $hich the participant sho$s in!ifference in a choice against the large probabilistic re$ar!. #he area un!er the cur+e &AEC) for the probabilistic !iscounting function can be use! as the outcome measure [39 . In general, a smaller AEC reflects a steeper !iscounting function an! more risky performance.

)tudy Analysis
#he results in #able '/ !escribe outcome measures of risky performance an! the use of fee!back in in!i+i!uals $ith A(*( an! =Cs. A significance le+el of pO.27 $as a!opte!. Effect siDes &CohenZs !) $ere reporte! for those stu!ies that pro+i!e! the re%uire! information to compute them. *o$e+er !ue to insufficient reporting of statistics in some of the papers, lea!ing to missing effect siDes, an! the large +ariation in output measures it $as not possible to calculate reliable a+erage effect siDes across stu!ies in chil!renHa!olescents an! a!ults. #herefore, in or!er to gi+e an in!ication of the magnitu!e of the effect siDe the range of the effect siDes has been pro+i!e! for chil!renHa!olescents an! a!ults separately. #he re+ie$ $as structure! accor!ing to the age of in!i+i!uals inclu!e! in the stu!ies &chil!renHa!olescents +ersus a!ults) an! accor!ing to the type of gambling task applie! &implicit +ersus e-plicit). ?e i!entifie! se+eral potential alternate e-planatory +ariables in the literature, $hich $ere age, se-, intelligence an!Hor e!ucation le+el, 6((HC(, I(s, A(*( subtype, "P* use, an! the form of re$ar! use!. #he potential influences of these alternate e-planatory +ariables are a!!resse! in a separate section, in $hich a comparison $as ma!e bet$een stu!ies in $hich !ifferences in risky beha+ior $ere foun! bet$een in!i+i!uals $ith A(*( an! =Cs &positi+e fin!ings) an! stu!ies that !i! not fin! any group !ifferences

&null fin!ings). A rather conser+ati+e strategy $as a!opte! for allocating stu!ies to the positi+e fin!ings category in or!er to ma-imiDe generaliDability. As such, stu!ies that only foun! group !ifferences for specific aspects or parts of a gambling task $ere allocate! to the null fin!ings category. In cases $here more than one A(*( group $as compare! to =Cs only the results of the A(*( group $ith the least comorbi!ity $ere use! for the classification. A potential alternate e-planatory +ariable $as regar!e! as controlle! for $hen the A(*( an! =C samples $ere matche! or !i! not !iffer on this +ariable, $hen statistics sho$e! that this +ariable !i! not correlate $ith the performance on a particular gambling task, or $hen an appropriate statistical correction $as carrie! out for the +ariable in %uestion.

Results
"mplicit +ambling ,asks in Children0Adolescents with AD-D
#en stu!ies in+estigate! the performance of chil!renHa!olescents $ith A(*( on an implicit gambling task. 'i- stu!ies use! the IB# or a IB# +ariant [3C M[7/ , three stu!ies use! the (6# [87 , [78 , [73 , an! one stu!y use! the 4A.# [77 . An o+er+ie$ of these stu!ies an! their results is gi+en in #able /.

!able 2 ,is0y performance outcomes on implicit gambling tas0s in children/adolescents with AD5D.

,he "owa +ambling ,ask 1"+,2


'i- stu!ies in+estigate! the performance of chil!renHa!olescents $ith A(*( on the IB# or a +ariant of the IB# [3C M[7/ , of $hich t$o stu!ies reporte! that chil!renHa!olescents clearly !isplaye! more risky beha+ior than =Cs [3C , [30 . Baron et al. [3C use! a chil! +ersion of the IB# an! foun! that chil!ren $ith A(*( &$ithout a comorbi! I() less often chose the a!+antageous !ecks than =Cs &CohenZs ! @ 9.93). #he =Cs also ma!e more a!+antageous !ecisions as the task progresse!, $hereas the chil!ren $ith A(*( &$ithout a comorbi! I() !i! not sho$ this pattern, an! !i! not choose the a!+antageous !ecks more often than pre!icte! by chance. *obson et al. [30 e-amine! the performance of a!olescents $ith A(*( on the secon! phase of the IB#, i.e. the risky !ecision-making phase in $hich the participants ha+e some abstract kno$le!ge of the riskiness of their choices &see the !escription in the intro!uction), an! also foun! that in!i+i!uals $ith A(*( ma!e more risky choices than the =Cs &CohenZs ! @ 2.>0).

:uman et al. [72 use! a +ariant of the IB# $ith three options, one a!+antageous option &small re$ar!sHsmall punishments) an! t$o !isa!+antageous options &large re$ar!sHlarge penalties an! small re$ar!sHlarge penalties). #he participants performe! the task in t$o con!itions< in the ^magnitu!e con!itionZ the magnitu!e of the penalty of the !isa!+antageous !ecks increase! $ith task progression, $hereas in the ^fre%uency con!itionZ the fre%uency of the penalty of the !isa!+antageous !ecks increase! $ith task progression. #he results !emonstrate! that in the fre%uency con!ition, both the chil!ren $ith A(*( an! the =Cs sho$e! a preference for the a!+antageous !eck. *o$e+er, in the magnitu!e con!ition, only the =Cs ha! a preference for the a!+antageous !eck, $hereas the chil!ren $ith A(*( !i! not. #he authors, therefore, presume! that chil!ren $ith A(*( are sensiti+e to the fre%uency, but blin! to the magnitu!e of a punishment. Contrary to e-pectations, the chil!ren $ith A(*( !i! not sho$ a particular specific preference for the !isa!+antageous !eck $ith large re$ar!s. Also, the group effect !uring the secon! task session $as re!uce!, suggesting that chil!ren $ith A(*( !o learn from pre+ious e-periences. #hree stu!ies foun! no abnormalities in the !egree of risk-taking on the IB# in chil!renHa!olescents $ith A(*( [31 , [79 , [7/ . Beurts et al. [31 use! a chil!renZs +ariant of the IB# [7> $ith t$o con!itionsI the ^stan!ar! con!itionZ &$hich is the !efault IB#) an! the ^re+erse! con!itionZ. In the stan!ar! con!ition, the re$ar!s are constant an! the penalties are unpre!ictable, $hereas in the re+erse! con!ition the penalties are constant an! the re$ar!s are unpre!ictable. #he stu!y re+eale! no !ifferences bet$een chil!ren $ith A(*( an! =Cs in net score &CohenZs ! @ 2.23). 4oth groups more often chose the a!+antageous !ecks as the task progresse! $ith this pattern emerging sooner in the re+erse! con!ition. #he t$o groups also !i! not !iffer in the use of fee!back from the pre+ious trial, as they both change! !eck more often after recei+ing a penalty than after a re$ar!. "asunami et al. [79 e-amine! !ecision-making patterns an! sensiti+ity to re$ar!s an! penalties on the IB# in chil!ren $ith A(*(. #he authors !i! not fin! abnormalities in the number of a!+antageous choices. *o$e+er, they foun! !ifferences bet$een chil!ren $ith A(*( an! =Cs in the so-calle! #-patterns that are relate! to the sensiti+ity to re$ar!s an! penalties. #-patterns are pairs of e+ents, in this case the outcomes an! choices of chil!ren, $hich are repeate! in the same or!er $ith a fi-e! time inter+al. An e-ample of a returning #-pattern is if a chil! recei+es a penalty from !eck !isa!+antageous !eck A, then selects from safe !eck C but the penalty appears in !isa!+antageous !eck 4, an! the chil! then selects !isa!+antageous !eck 4. #he results sho$e! that there $ere significantly less #-patterns inclu!ing penalties in chil!ren $ith A(*( compare! to =Cs, $hich in!icates that chil!ren $ith A(*( pai! less attention to penalties than the =Cs. #oplak et al. [7/ in+estigate! the performance of a!olescents $ith A(*( on the IB#. =o group !ifferences $ere foun! in the net score an! financial outcome of the A(*( group compare! to the =Cs. Pisual inspection !emonstrate! that car! selections $ere ran!om in the first ambiguous phase &O72 trials) in both groups. *o$e+er, in the secon!, risky, phase a!olescents $ith A(*( chose the !isa!+antageous !eck $ith infre%uent penalties more often an! chose the a!+antageous !eck $ith infre%uent penalties less often $hen compare! to =Cs. #here $ere no group !ifferences in the choices for the t$o !ecks $ith fre%uent penalties, $ith both A(*( in!i+i!uals an! =CZs more often selecting the a!+antageous !eck in this case. #his supports the i!ea that

in!i+i!uals $ith A(*( are more sensiti+e for the fre%uency than the magnitu!e of penalties. A!!itionally, t$o A(*( subtypes &A(*(-C an! A(*(-I) $ere compare!. =o !ifference $as foun! in net score bet$een these t$o subtypes of A(*(. *o$e+er, the a!olescents $ith A(*(-C chose the !ecks $ith infre%uent penalties more often an! the !ecks $ith fre%uent penalties less often compare! to those $ith A(*(-I. In!i+i!uals $ith A(*(-C appear therefore to be more sensiti+e to the fre%uency an! less sensiti+e for the magnitu!e of penalties in comparison to in!i+i!uals $ith A(*(-I. As mentione! abo+e, Baron et al. [3C reporte! that chil!ren $ith A(*( $ithout an I( ma!e less a!+antageous choices on a chil! +ersion of the IB# than =Cs. #his stu!y also inclu!e! a group of chil!ren $ith A(*( an! an-ietyH!epression, $ho ma!e significantly more a!+antageous choices than the A(*( group $ithout an-ietyH!epression &CohenZs ! @ 9.22). #he chil!ren $ith A(*( an! an-ietyH!epression also !i! not !iffer from the =Cs &CohenZs !O2.81) an! as the task progresse! they ma!e more a!+antageous choices. #he authors, therefore, assume! that in chil!ren $ith A(*( an I( has a protecti+e effect on reinforcement learning. Another possibility they suggeste! is that fear, $hich is often increase! in those $ith an-ietyH!epression, lea!s to an increase! a$areness of $hich !ecks are better or $orse. 5inally, as mentione! abo+e, *obson et al. [30 foun! that a!olescents $ith A(*( !isplaye! more risky beha+ior in the IB# than =Cs. A!!itionally !imensional analyses &multiple regression analyses) re+eale! that 6((HC( but not A(*( symptoms $ere associate! $ith risky beha+ior on the IB#.

,he Door 'pening ,ask 1D',2


#hree stu!ies in+estigate! the performance of chil!renHa!olescents $ith A(*( on the (6# [87 , [78 , [73 . #$o out of the three stu!ies reporte! that chil!ren $ith A(*( an! comorbi! 6((HC( playe! the task longer an! therefore ran more risks than the =Cs [87 , [78 &CohenZs ! respecti+ely @ 2.0C an! 9.8/). Con+ersely, ?iers et al. [73 foun! no !ifference in the number of playe! !oors bet$een chil!ren $ith A(*( &$ithout comorbi! 6((HC() an! =Cs &CohenZs ! @ 2.91).

,he Balloon Analogue Risk ,ask 1BAR,2


*umphreys ; :ee [77 e-amine! risky beha+ior an! sensiti+ity to punishment on the 4A.# in chil!ren $ith A(*( $ith an! $ithout comorbi! 6(( an! =Cs. #he stu!y sho$e! that the A(*( group %it* comorbi! 6(( ran more risks by pumping up the balloons more than the A(*( group %it*out comorbi! 6((, $ho !i!, ho$e+er, still pump the balloons more than the =Cs. Contrary to e-pectations, the chil!ren $ith A(*( an! comorbi! 6(( $ere most sensiti+e to punishment, in that they pumpe! the balloon less in trials after ha+ing ,ust been penaliDe! $ith a balloon pop, follo$e! by the =Cs, an! then the chil!ren $ith A(*( $ithout 6((, $ho $ere the least sensiti+e to punishment. #he authors therefore assume! that chil!ren $ith A(*( an! comorbi! 6(( are characteriDe! by poor affect regulation, $hich makes them too reacti+e an!Hor unable to cope a!e%uately $ith punishment. #he authors further hypothesiDe! that this cause! chil!ren $ith A(*( an! comorbi! 6(( to perform inconsistently on the

gambling task, thereby !emonstrating an increase in risky beha+ior an! an increase in the fre%uency of impulsi+e a!,ustments of beha+ior after recei+ing penalties.

%&plicit +ambling ,asks in Children0Adolescents with AD-D


5our stu!ies in+estigate! the performance of chil!renHa!olescents $ith A(*( in e-plicit gambling tasks. All of $hich ma!e use of a !ifferent task para!igm &CB#, B(#, ""B, an! P() [32 , [39 , [7C , [71 . An o+er+ie$ of these stu!ies an! their results is gi+en in #able 8.

!able 3 ,is0y performance outcomes on e6plicit gambling tas0s in children/adolescents with AD5D.

,he Cambridge +ambling ,ask 1C+,2


(ePito et al. [7C in+estigate! the performance of chil!ren $ith A(*( on the CB# in a !ouble-blin! placebo-controlle! $ithin-sub,ects trial of "P*. In the placebo con!ition, the chil!ren $ith A(*( !i! not !iffer from the =Cs on the mean betting proportion &risk-taking< CohenZs ! @ 2./C). *o$e+er, chil!ren $ith A(*( ma!e less rational choices an! score! lo$er on risk a!,ustment than the =Cs. In the "P* con!ition, the chil!ren $ith A(*( bet fe$er points, thereby lo$ering their risk, but !i! not !iffer from the placebo con!ition in their number of rational choices or risk a!,ustment.

,he +ame of Dice ,ask 1+D,2


(rechsler et al. [71 in+estigate! risky beha+ior on the B(# in chil!ren $ith A(*(. #he chil!ren playe! the B(# t$ice. =o !ifferences bet$een chil!ren $ith A(*( an! =Cs $ere foun! in the first game &CohenZs ! @ 2.27), but chil!ren $ith A(*( !isplaye! more risky beha+ior than =Cs !uring the secon! game &CohenZs ! @ 2.18). 'pecifically, in the secon! game, chil!ren $ith A(*( chose the most risky alternati+e &one !ie) more often than !uring the first trial. #his poorer performance on the secon! trial means that if the o+erall performance on the first an! secon! game is e-amine! then the chil!ren $ith A(*( performe! $orse o+erall than the =Cs, 4ase! on these fin!ings the authors suggeste! that chil!ren $ith A(*( respon! to fee!back in a similar fashion as =Cs $hen confronte! $ith something ne$, but sho$ aberrant beha+ior $hen they become more use! to the task.

,he Make!a!Match +ame 1MM+2


(rechsler et al. [32 !e+elope! the ""B an! !emonstrate! that chil!ren $ith A(*( ha! a greater preference for con!itions $ith a lo$ probability large re$ar!, than =Cs &four-car! selections< CohenZs ! @ 9./2). 4oth groups !i! not change their strategy !uring task progression an! s$itche! set e%ually often follo$ing positi+e or negati+e fee!back. #he authors e-plain this lack of learning effects by the absence of e-plicit punishments for incorrect choices in the ""B, an! the fact that in this stu!y there $as no !ifference in the final re$ar! that $as obtaine! for a cautious or more risky strategy. #he authors suggeste! that the !isplaye! preference for larger but less probable re$ar!s in chil!ren $ith A(*( points to an a!!itional aspect of a !ysfunctional re$ar! system. #he authors argue that the fin!ings cannot be solely e-plaine! by !elay a+ersion or o+ersensiti+ity to imme!iate re$ar!s.

,he Probabilistic Discounting ,ask 1PD2


'cheres et al. [39 in+estigate! $hether age an! A(*( symptoms affecte! choice preferences in chil!ren &> to 99 years) an! a!olescents &9/ to 9C years) on the P(. =o !ifferences bet$een chil!ren an! a!olescents $ith A(*( an! =Cs $ere foun! in the area un!er the cur+e &AEC) of the probabilistic !iscounting function &see "etho!s section for an e-planation of this outcome measure), in!icating that both groups ran similar risks in this task &CohenZs ! @ 2./C). Also, there $as neither an age effect nor an interaction effect of age an! !iagnosis, an! all groups ma!e choices that ma-imiDe! the total gain. #he authors ascribe! these null fin!ings among other things to the use of e-plicit chances in the task !esign an! hypothesiDe that in!i+i!uals $ith A(*( ha+e poor learning of risks, $hich is best measure! $ith gambling tasks in $hich the chances are implicit an! ha+e to be learne!.

)ummary of the Results for Children0Adolescents with AD-D


5ourteen stu!ies in+estigate! the performance of chil!renHa!olescents $ith A(*( on +arious gambling tasks. #he effect siDes of the group !ifferences in these stu!ies range! from a CohenZs ! of 2.23 to a ! of 9.8/. #en stu!ies use! an implicit gam'ling task, of $hich fi+e stu!ies &7H92 @ 72A) foun! clear e+i!ence that chil!renHa!olescents $ith A(*( !isplaye! more risky beha+ior than =Cs [87 , [3C , [30 , [78 , [77 . An a!!itional stu!y only reporte! aberrantly risky beha+ior in chil!ren $ith A(*( in one con!ition &the magnitu!e con!ition) on a +ariant of the IB#, but not in the other con!ition &the fre%uency con!ition) [72 . #$o of the fourteen stu!ies in+estigate! the effects of comorbi! con!itions, an! foun! that chil!ren $ith A(*( an! comorbi! 6((HC( performe! in a more risky fashion than chil!ren $ith A(*( $ithout comorbi!ity [77 . *o$e+er, chil!ren $ith A(*( an! a comorbi! I( &an-ietyH!epression) performe! in a less risky fashion than the chil!ren $ith A(*( $ithout comorbi!ity, $ho coul! not be !ifferentiate! from the =Cs [3C . Another stu!y compare! !ifferent subtypes of A(*( an! reporte! no !ifferences in risky beha+ior bet$een a!olescents $ith A(*(-C an! A(*(-I [7/ . *o$e+er, the a!olescents $ith A(*(-C !i! choose !ecks $ith infre%uent penalties in the IB# more often an! the !ecks $ith fre%uent penalties less often than

those $ith A(*(-I. 5our of the fourteen stu!ies $ith chil!renHa!olescents use! an explicit gam'ling task an! t$o stu!ies &/H3 @ 72A) foun! that chil!renHa!olescents $ith A(*( performe! in a more risky fashion than =Cs [32 , [71 . 5inally, another stu!y !emonstrate! that "P* re!uce! the number of points bet in the CB#, $hich in!icates that fe$er risks $ere run by chil!renHa!olescents $ith A(*( $ho $ere treate! $ith "P* [7C . In summary, half of the stu!ies $ith chil!renHa!olescents &CH93 @ 72A) foun! e+i!ence for more risky beha+ior on gambling tasks in chil!renHa!olescents $ith A(*( compare! to =Cs, in!epen!ently from the type of gambling task use! &implicit or e-plicit). ?ith regar! to the sensiti+ity to re$ar!s an! penalties &fee!back use) in chil!renHa!olescents, one stu!y foun! significantly less #-patterns that inclu!e! penalties in the IB# in chil!ren $ith A(*( compare! to =Cs [79 . Another stu!y reporte! that chil!ren $ith A(*( score! lo$er on post e-plosion reacti+ity on the 4A.# than =Cs, $hereas chil!ren $ith A(*( $ith comorbi! 6(( score! higher on this measure than the =Cs [77 . :astly, t$o other stu!ies foun! no !ifferences in the number of s$itches after negati+e or positi+e fee!back in the ""B bet$een chil!ren $ith A(*( an! =Cs [32 , [31 .

"mplicit +ambling ,asks in Adults with AD-D


Eight stu!ies in+estigate! the performance of a!ults $ith A(*( on implicit gambling tasks. 'i- of these stu!ies use! the IB# or a +ariant of the IB# [70 M[>3 , t$o stu!ies use! the 4A.# [>3 , [>7 , an! one stu!y use! the C# [>> . An o+er+ie$ of these stu!ies an! their results is gi+en in #able 3.

!able 4 ,is0y performance outcomes on implicit gambling tas0s in adults with AD5D.

"owa +ambling ,ask 1"+,2


6f the si- stu!ies in+estigating the performance of a!ults $ith A(*( on the IB# [70 M [>3 , t$o stu!ies reporte! that a!ults $ith A(*( clearly performe! in a more risky manner than the =Cs [>/ , [>8 . 'pecifically, "alloy-(iniD et al. [>/ , [>8 e-amine! t$o !ifferent samples of a!ults $ith A(*( an! foun! that in comparison to =Cs, a!ults $ith A(*( obtaine! a lo$er net score on the stan!ar! IB# &CohenZs ! respecti+ely @ 2.C0 an! 2.C2). #he authors suggeste! that this $as because in!i+i!uals $ith A(*( ha+e !ifficulties learning from pre+ious e-periences.

Con+ersely, Agay et al. [70 foun! no aberrant performance of a!ults $ith A(*( on the stan!ar! IB#. *o$e+er, they !i! obser+e an increase in the risky beha+ior of their participants in the IB# +ariant calle! the ^5oregone Payoff Bambling #askZ &5PB#). #he 5PB# is !ifferent from the classic form of the IB# in that not only is the outcome of the chosen car! presente! in the 5PB# but also the outcomes of the unselecte! car!s of the other three !ecks. #his pro+i!es the participant $ith e-tra information, but may also !istract attention of the participants. In the 5PB#, a!ults $ith A(*( chose the !isa!+antageous !ecks more often than =Cs. #he authors suggeste! that the suboptimal performance of a!ults $ith A(*( on the 5PB# is !ue to higher !istractibility, an! problems $ith !i+i!e! an! selecti+e attention in the A(*( participants. #he Agay et al. [70 stu!y also e-amine! the effects of "P* by applying a placebo-controlle! ^bet$eensub,ectsZ !esign in $hich both a!ults $ith A(*( an! =Cs recei+e! either "P* or a placebo. =o effects of "P* $ere foun! on the performance of a!ults $ith A(*( or in =Cs on the stan!ar! IB# or the 5PB#. "uch like Agay et al. [70 , t$o other stu!ies ha+e also reporte! no greater le+els of risky performance in a!ults $ith A(*( on the stan!ar! IB# $hen compare! to =Cs [>9 , [>3 , an! one other stu!y only re+eale! aberrant performance on the stan!ar! IB# in a subgroup of a!ults $ith A(*( $ith both har! !rug !epen!ence an! $orking memory problems [>2 . 5urthermore, a stu!y by Ernst et al. [>9 also !i! not fin! any !ifferences in the net score on the stan!ar! IB# bet$een a!ults $ith A(*( an! =Cs &CohenZs ! @ 2.21). *o$e+er, Positron Emission #omography &PE#) analyses of the participants in the Ernst et al. [>9 stu!y !i! re+eal the in+ol+ement of !ifferent neural net$orks &in particular the anterior cingulate, hippocampus, an! insula) subser+ing emotion an! memory processing in a!ults $ith A(*( as compare! to the =Cs !uring the performance of the IB#. #he stu!y of "_ntyl_ et al. [>3 initially appears !ifferent from those !escribe! abo+e in that they foun! that a!ults $ith A(*( earne! less money on a stan!ar! IB# than =Cs &CohenZs ! @ 2.7>). #his group effect, ho$e+er, appears to ha+e been me!iate! by the e!ucational le+el of the participants. 5inally, (uarte et al. [>2 in+estigate! the IB# performance of a!ults $ith A(*( an! a comorbi! methamphetamine !epen!ence &"A). #he results in!icate! that only a!ults $ith A(*(J"A $ho also ha! $orking memory problems selecte! the !isa!+antageous !ecks more often than both a!ults $ith A(*(J"A %it*out $orking memory problems an! =Cs both $ith an! $ithout $orking memory problems &9.03OCohenZs !O/.23).

Balloon Analogue Risk ,ask 1BAR,2


#$o stu!ies in+estigate! the performance of a!ults $ith A(*( on the 4A.# an! neither stu!y re+eale! significant !ifferences in risky performance bet$een a!ults $ith A(*( an! =Cs [>3 , [>7 . 'pecifically, $hile in the stu!y of "_ntyla et al. [>3 the a!ults $ith A(*( pumpe! the balloons more up than =Cs !uring the first of 92 trials, there $ere no group !ifferences in the remaining 72 trials, resulting in no o+erall group !ifference on this task. 'imilarly, ?eafer et al. [>7 !i! not fin! any group !ifferences in the total number of pumps on the 4A.# bet$een a!ults $ith A(*( an! =Cs &CohenZs ! @ 2.93).

Card Playing ,ask 1C,2


6nly one stu!y assesse! the performance of a!ults $ith A(*( on the C# [>> . In this stu!y, a!ults $ith persistent A(*(, a!ults $ith remittent A(*( &only A(*( in chil!hoo!), an! the =Cs, !i! not !iffer from each other in the number of playe! car!s &2.29OCohenZs !O2./3). A!ults $ith persistent or remittent A(*( %it* a comorbi! C(, ho$e+er, playe! longer &perse+ere!) compare! to a!ults $ith persistent or remittent A(*( %it*out a comorbi! C( &CohenZs ! @ 2.38).

%&plicit +ambling ,asks in Adults with AD-D


#hree stu!ies in+estigate! the performance of a!ults $ith A(*( using an e-plicit gambling task, the B(# [>C , [>1 &reference >C !escribes / separate stu!ies). An o+er+ie$ of these stu!ies an! their results is gi+en in #able 7.

!able 5 ,is0y performance outcomes on e6plicit gambling tas0s in adults with AD5D.

+ame of Dice ,ask 1+D,2


6f the three stu!ies assessing the B(# in a!ults $ith A(*( [>C , [>1 , only one stu!y foun! that a!ults $ith A(*( performe! in a more risky fashion than the =Cs [>C &stu!y 9). 'pecifically, the a!ults $ith A(*( gaine! a lo$er net score than the =Cs &CohenZs ! @ 2.08) by ten!ing to choose the option $ith t$o !ice more often an! the option $ith four !ice less often than the =Cs. #he authors of this stu!y [>C also reporte! !ifferences in the $ay a!ults $ith A(*( an! =Cs ma!e use of fee!back. In that the a!ults $ith A(*( staye! less often $ith a safe option after positi+e fee!back, an! staye! more often $ith a risky option after negati+e fee!back in comparison to the =Cs. In the secon! stu!y, reporte! in the same paper by "atthies et al. [>C &stu!y /), a A(*( an! =C sample $as assesse! in $hich bore!om $as elicite! by forcing the participant to $ait for 7 minutes in front of a black screen before the B(# starte!. In contrast to the first stu!y [>C &stu!y 9), this stu!y !i! not fin! any !ifference bet$een a!ults $ith A(*( an! the =Cs in net score &CohenZs ! @ 2.C2) or fee!back use. A !irect comparison bet$een these t$o stu!ies re+eals that the =Cs in the bore!om con!ition more often selecte! the risky options than the =Cs in the con!ition $ithout bore!om, $hile the A(*( groups performe! in a similar fashion in both con!itions. #he authors, therefore, suggeste! that a!ults $ith an! $ithout A(*( !iffer in the $ay they regulate bore!om. #his interpretation shoul! be +ie$e! $ith caution since the samples of the t$o stu!ies !iffere! &e.g. the sample of a!ults $ith A(*( in the secon! stu!y score! higher on the 4eha+ioral Inhibition 'cale than the patient sample in the first stu!y) an! bore!om $as not assesse! in the first stu!y.

In a separate stu!y, ?ilbertD an! colleagues [>1 !i! not fin! any !ifferences bet$een a!ults $ith A(*( an! =Cs $ith regar! to their performance on the B(# &CohenZs ! block 9 ; / respecti+ely @ 2./8 an! 2.23). *o$e+er, the A(*( an! =C group !i! !iffer in their f".I an! electro!ermal responsi+eness to re$ar! +alue. ?hereas in the =Cs the re$ar! +alue &high +ersus lo$ incenti+e) $as !ifferentially co!e! in the me!ial orbitofrontal corte-, this $as not the case in the A(*( group. #his !ysfunctional co!ing in patients correlate! $ith risky performance in the B(# an! $as parallele! by physiological arousal.

)ummary of the Results for Adults with AD-D


In total, ele+en stu!ies e-amine! the performance of a!ults $ith A(*( on +arious gambling tasks. #he effect siDes of the group !ifferences bet$een a!ults $ith A(*( an! =Cs in these stu!ies range! from CohenZs ! 2.29 to /.23. Eight stu!ies use! implicit gam'ling tasks, of $hich t$o stu!ies &/H1 @ /7A) pro+i!e! clear e+i!ence that a!ults $ith A(*( performe! in a more risky fashion than =Cs [>/ , [>8 . Another stu!y pro+i!e! mi-e! e+i!ence for !ifferences bet$een the performance of A(*( in!i+i!uals an! =Cs, in that only performance on the 5PB# but not on the stan!ar! IB# $as aberrant in a!ults $ith A(*( [70 . In a!!ition to looking at the !ifferences bet$een A(*( a!ults an! =Cs, one stu!y in+estigate! the effects of 6((HC( comorbi!ity an! foun! that a!ults $ith A(*( an! comorbi! C( pro!uce! more risky choices than a!ults $ith A(*( $ithout comorbi!ity [>> . 6nly one stu!y out of these eight that e-amine! implicit gambling tasks in+estigate! the effects of "P* on gambling task performance in a!ults $ith A(*( an! foun! no e+i!ence for an effect of "P* [70 . 5inally, the performance of a!ults $ith A(*( on an explicit gam'ling task &B(#) $as e-amine! in three stu!ies. ?hile one stu!y obser+e! &9H8 @ 88A) increase! risky beha+ior in a!ults $ith A(*( [>C &stu!y 9), the other t$o stu!ies faile! to fin! any o+erall group !ifferences bet$een a!ults $ith A(*( an! =Cs. #herefore, summariDing the fin!ings of stu!ies in a!ults it appears that only a minority &8H99 @ /CA) of the stu!ies foun! e+i!ence that a!ults $ith A(*( compare! to =Cs perform in a more risky fashion on gambling tasks, a fin!ing that appears to be in!epen!ent of the type of gambling task use! &implicit or e-plicit). In terms of the impact of fee!back, only one t$o-part stu!y in+estigate! fee!back use in a!ults $ith A(*(. In their first stu!y, "atthies et al. [>C) &stu!y 9 foun! that a!ults $ith A(*( compare! to =Cs staye! less often $ith a safe option after positi+e fee!back an! staye! more often $ith a risky option after negati+e fee!back. *o$e+er, in their secon! stu!y, in $hich bore!om $as in!uce! before performing the gambling task, there $ere no !ifferences in fee!back use bet$een a!ults $ith A(*( an! =Cs [>C &stu!y /).

Potential Alternate %&planatory *actors


As sho$n abo+e, the fin!ings on risky !ecision-making in gambling tasks in in!i+i!uals $ith A(*( are inconsistent. 'e+eral alternate factors that might influence the outcomes of the gambling tasks ha+e alrea!y been a!!resse! in this re+ie$ that may e-plain this inconsistency, inclu!ing type of gambling task, comorbi!ity &6((HC( in

chil!renHa!olescents an! I(s), A(*( subtype, "P* use, the form of the re$ar! recei+e!, an! !emographic factors &age, se-, an! intelligence or e!ucational le+el). In this section, stu!ies that pro+i!e e+i!ence for more risky performance of in!i+i!uals $ith A(*( in comparison to =Cs, i.e. stu!ies $ith positi+e fin!ings, $ill be contraste! $ith regar! to the aforementione! alternate e-planatory +ariables, $ith stu!ies that faile! to fin! such e+i!ence, i.e. stu!ies $ith null fin!ings. ?e ma!e a broa! categoriDation of stu!ies $ith positi+e fin!ings [87 , [32 , [3C , [30 , [78 , [77 , [71 , [>/ , [>8 , [>C &reference >C refers to stu!y 9) an! stu!ies $ith null fin!ings [39 , [31 , [72 M[7/ , [73 , [7C , [70 M[>9 , [>3 M[>> , [>C , [>1 &reference >C refers to stu!y /). #his approach $as aime! at gaining insight into the factors that may cause the inconsistency of fin!ings that appear to !ominate this area. A graphical !epiction of the comparison of the stu!ies $ith positi+e an! null fin!ings for each of the alternate e-planatory factors is gi+en in 5igure /.

Fi ure 2 &tac0ed bar charts depicting the presence and absence of alternate e6planatory factors split for studies with positive findings (7% i.e. increased ris0-ta0ing performance in AD5D compared to NCs) and studies with null findings (4% i.e. no AD5D-NC difference ...

#he first potential alternate e-planatory factor is the age o) t*e participants, see 5igure /A. 5ourteen stu!ies $ere con!ucte! in chil!renHa!olescents. *alf of these stu!ies re+eale! a group effect [87 , [32 , [3C , [30 , [78 , [77 , [71 , $hereas the other half reporte! null fin!ings [39 , [31 , [72 M[7/ , [73 , [7C . Con+ersely, of the ele+en stu!ies in a!ults, only three stu!ies foun! a group effect [>/ , [>8 , [>C &reference >C refers to stu!y 9) an! eight stu!ies ha! null fin!ings [70 M[>9 , [>3 M[>> , [>C , [>1 &reference >C refers to stu!y /). #herefore, the e+i!ence for more risky performance in gambling tasks in in!i+i!uals $ith A(*( compare! to =Cs appears to be stronger for chil!renHa!olescents &CH93 @ 72A of stu!ies in chil!renHa!olescents ha! positi+e fin!ings) than for a!ults &8H99 @ /CA of stu!ies in a!ults ha! positi+e fin!ings). In terms of the effect of age $ithin the in!i+i!uals stu!ies, it is unlikely that age contribute! meaningfully to the outcomes of the stu!ies, because all stu!ies e-cept one [>7 controlle! for age by group matching or statistical correction &/3H/7 @ 0>A of stu!ies controlle! for age as an alternate e-planatory factor). A secon! potential alternate e-planatory factor is the sex of the participants. #$enty-one out of the /7 stu!ies e-amine! matche! their samples on se- or ha! e%ual se- ratios in the A(*( an! control groups. 6f the remaining stu!ies, one stu!y inclu!e! se- as a co+ariate [7/ an! t$o stu!ies mentione! that se- !i! not correlate $ith the performance

on the gambling task [79 , [77 . 6nly one stu!y [>8 , $hich ha! positi+e fin!ings, inclu!e! samples that !iffere! in se- ratio but !i! not control for this +ariable. 6+erall, se- can therefore be regar!e! as $ell-controlle! for in the ma,ority of the stu!ies &/3H/7 @ 0>A of stu!ies controlle! for se- as an alternate e-planatory factor). A thir! potential alternate e-planatory factor is the intelligence+education level of the participants. Ele+en out of /7 stu!ies reporte! no !ifferences in I` scores bet$een the A(*( an! control groups. 6f the remaining stu!ies, three stu!ies entere! I` as a co+ariate in the statistical analyses [30 , [7/ , [73 , t$o stu!ies reporte! that I` !i! not correlate $ith performance on the gambling task [72 , [>> , one stu!y !i! not report the influence of I` [>7 , an! the last eight stu!ies !i! not measure I` [87 , [79 , [77 , [7C , [70 , [>3 , [>C . "ost of the stu!ies that !i! not control for I`, ho$e+er, controlle! for e!ucation le+el [87 , [7C , [70 , [>3 , [>7 , [>C &CH0 @ C1A of the stu!ies not controlling for I` controlle! for e!ucation le+el). #o summariDe, the ma,ority of the stu!ies in this re+ie$ matche! their participants for I` or e!ucation le+el, correcte! statistically the effect of I`, or checke! for the influence of group !ifferences in I` &/8H/7 @ 0/A of the stu!ies performe! controlle! for intelligenceHe!ucation as an alternate e-planatory +ariable). A fourth potential alternate e-planatory factor is the type o) gam'ling task use!, see 5igure /4. At the implicit +ersus e-plicit gambling task le+el it seems that there is similar e+i!ence that in the ma,ority of stu!ies, no matter the task type, risky performance in in!i+i!uals $ith A(*( is not foun! to be higher than =Cs. 5rom the ten stu!ies that foun! a group effect, three stu!ies applie! an e-plicit gambling task [32 , [71 , [>C &8H92 @ 82A of the stu!ies $ith positi+e fin!ings< reference >C refers to stu!y 9) an! se+en stu!ies applie! an implicit gambling task [87 , [3C , [30 , [78 , [77 , [>/ , [>8 &CH92 @ C2A of the stu!ies $ith positi+e fin!ings). #his !istribution $as appro-imately the same in the fifteen stu!ies that $ere categoriDe! as ha+ing null fin!ings, in that four stu!ies applie! an e-plicit gambling task [39 , [7C , [>C , [>1 &3H97 @ /CA of the stu!ies $ith null fin!ings< reference >C refers to stu!y /) an! ele+en stu!ies applie! an implicit gambling task [31 , [72 M[7/ , [73 , [70 M[>9 , [>3 M[>> &99H97 @ C8A of the stu!ies $ith null fin!ings). Inspection of the gambling tasks applie! $ithin the implicit an! e-plicit categories re+eale! that the same tests $ere performe! &IB#, C#H(6#, 4A.#, an! B(#) in stu!ies that foun! a group effect as $ell as in stu!ies reporting null fin!ings. #herefore, the inconsistencies of fin!ings cannot be attribute! !irectly to the !ifferent types of gambling tasks or the specific tasks use! for the assessment of risky !ecisionmaking. Although, possible +ariations bet$een stu!ies in ho$ these tasks $ere !eli+ere! an! presente! cannot be rule! out. A fifth potential alternate e-planatory factor is psyc*iatric comor'idity, such as 6((HC( &see 5igure /C) an! I(s &see 5igure /(). #$o stu!ies in this re+ie$ compare! chil!ren $ith A(*( an! comorbi! 6((HC( $ith in!i+i!uals $ith A(*( $ithout 6((HC( in their performance on the (6#, an! foun! that the groups $ith comorbi!ity performe! in a more risky fashion than the groups $ithout comorbi!ity [77 , [>> . 6f the ten stu!ies that !emonstrate! group !ifferences in risky performance, t$o stu!ies inclu!e! in!i+i!uals $ith A(*( $ho all ha! comorbi! 6((HC( [87 , [78 &/H92 @ /2A of the

stu!ies $ith positi+e fin!ings). Broup !ifferences bet$een A(*( groups an! =Cs $ere also foun! in stu!ies $ith mi-e! A(*( samples consisting of both in!i+i!uals $ith an! $ithout comorbi! 6((HC( [3C , [30 , [71 &8H92 @ 82A of stu!ies $ith positi+e fin!ings), for stu!ies that inclu!e! in!i+i!uals $ith A(*( $ithout 6((HC( comorbi!ity [32 , [77 &/H92 @ /2A of stu!ies $ith positi+e fin!ings), an! for stu!ies in $hich 6((HC( comorbi!ity $as not reporte! [>/ , [>8 , [>C &8H92 @ 82A of stu!ies $ith positi+e fin!ings< reference >C refers to stu!y 9). A comparable proportion of stu!ies $ith null fin!ings inclu!e! mi-e! A(*( samples [39 , [31 , [72 , [7/ , [7C , [>> &>H97 @ 32A of stu!ies $ith null fin!ings), A(*( samples $ithout 6((HC( comorbi!ity [73 , [70 , [>9 &8H97 @ /2A of stu!ies $ith null fin!ings), or !i! not report 6((HC( comorbi!ity [79 , [>2 , [>3 , [>7 , [>C , [>1 &>H97 @ 32A of stu!ies $ith null fin!ings< reference >C refers to stu!y /). In summary, the four stu!ies that inclu!e! participants $ith both A(*( an! 6((HC( consistently sho$e! increase! risky performance compare! $ith =Cs, $hereas the stu!ies inclu!ing mi-e! A(*( samples an! A(*( samples $ithout 6((HC( comorbi!ity $ere not consistent in their fin!ings. 6ne stu!y in this re+ie$ !irectly compare! chil!ren $ith A(*( $ith an! $ithout a comorbi! I( an! foun! that I(s in chil!ren $ith A(*( lea! to less risky beha+ior on a gambling task [3C . *o$e+er, comparing stu!ies $ith group effects an! stu!ies $ith null fin!ings $ith regar! to the inclusion of in!i+i!uals $ith I(s resulte! in an inconsistent pattern. #hree of the eight stu!ies &8H1 @ 81A) that recruite! only in!i+i!uals $ithout I(s foun! group !ifferences [87 , [3C , [78 , $hereas the other fi+e stu!ies &7H1 @ >8A) !i! not [72 , [73 , [7C , [70 , [>9 . 'imilarly, only t$o out of the se+en stu!ies &/HC @ /0A) inclu!ing a mi-e! sample $ith in!i+i!uals $ith A(*( $ith an! $ithout comorbi! I(s foun! group !ifferences [>/ , [>C &reference >C refers to stu!y 9). #he remaining fi+e stu!ies &7H1 @ C9A) !i! not fin! any !ifferences [39 , [7/ , [>2 , [>C , [>1 &reference >C refers to stu!y /). #en stu!ies !i! not mention the presence of comorbi! I(s in their sample, of $hich one half ha! positi+e [32 , [30 , [77 , [71 , [>8 an! the other half ha! null fin!ings [31 , [79 , [>3 M[>> . In summary, one stu!y $hich inclu!e! chil!ren $ith A(*( an! comorbi! I(s reporte! less risky performance, $hereas the stu!ies that inclu!e! mi-e! samples an! A(*( samples $ithout comorbi!ity $ere inconsistent in their results. A si-th potential alternate e-planatory factor might be the su'type o) A"," o) t*e participants, see 5igure /E. 6nly one stu!y in this re+ie$ !irectly e-amine! $hether in!i+i!uals $ith A(*( of !ifferent subtypes !iffere! $ith regar! to their performance on gambling tasks [7/ . Comparisons $ithin this stu!y [7/ in!icate! that a!olescents $ith A(*(-C an! A(*(-I $ere similar on the IB# in terms of their risky performance. *o$e+er, the a!olescents $ith A(*(-C more often chose the options $ith infre%uent penalties an! less often chose the options $ith fre%uent penalties $hen compare! to the a!olescents $ith A(*(-I. If fin!ings are compare! across stu!ies, higher le+els of risky performance $ere foun! in mi-e! samples of in!i+i!uals $ith A(*( compare! to =Cs, $hich inclu!e! stu!ies that e-amine! participants of all three A(*( subtypes [32 , [71 &/H92 @ /2A of stu!ies $ith positi+e fin!ings) or participants $ith A(*(-C an! A(*(-I [>/ &9H92 @ 92A of stu!ies $ith positi+e fin!ings). 5urthermore, three stu!ies foun! that in!i+i!uals $ith A(*(-C ha! performe! in a more risky fashion than the =Cs [87 , [3C ,

[>8 &8H92 @ 82A of stu!ies $ith positi+e fin!ings). *o$e+er, there $as a comparable number of stu!ies that faile! to !emonstrate any !ifference in !ecision-making from =Cs in mi-e! samples of participants $ith A(*(. #hese mi-e! samples inclu!e! stu!ies e-amining participants of all three A(*( subtypes [39 , [31 , [72 , [>2 &3H97 @ /CA of stu!ies $ith null fin!ings), participants $ith A(*(-C an! A(*(-* [7/ , [73 &/H97 @ 98A of stu!ies $ith null fin!ings) as $ell as participants $ith A(*(-C an! A(*(-I [79 , [>9 , [>1 &8H97 @ /2A of stu!ies $ith null fin!ings). #he fin!ings of ten a!!itional stu!ies coul! not be taken into consi!eration in this !iscussion, since no information $as pro+i!e! about the participantsZ subtypes of A(*( &[30 , [78 , [77 , [7C , [70 , [>3 M [>> , [>C stu!y 9;/). In conclusion, $hereas one stu!y reporte! subtle !ifferences in risky performance bet$een A(*(-C an! A(*(-I, no consistent pattern emerge! in the outcomes of stu!ies that inclu!e! samples $ith !ifferent subtypes an! many stu!ies !i! not pro+i!e enough !etails to be able to control for this factor. A se+enth potential alternate e-planatory factor is the treatment of A(*( symptoms by using stimulant drug treatment &e.g. "P*), see 5igure /5. #$o stu!ies !irectly in+estigate! the effects of "P* by using a placebo-controlle! !esign. ?hile one stu!y foun! that chil!ren $ith A(*( on "P* performe! in a less risky fashion on a gambling task compare! to $hen on a placebo [7C , the other stu!y re+eale! no effects of "P* on the performance on a gambling task in a!ults $ith an! $ithout A(*( [70 . An inconsistent pattern also emerge! $hen comparing stu!ies $ith positi+e fin!ings an! null fin!ings concerning the use of "P* !uring the assessment. Broup !ifferences bet$een A(*( participants an! =Cs $ere present in stu!ies in $hich participants !iscontinue! me!ication treatment $ith "P* 1 to 31 hours before the e-periment [32 , [30 , [78 , [71 , [>C &7H92 @ 72A of stu!ies $ith positi+e fin!ings< reference >C refers to stu!y 9), in $hich participants $ere "P*-naa+e [>/ , [>8 &/H92 @ /2A of stu!ies $ith positi+e fin!ings), an! in $hich participants $ere on "P* treatment at the time of assessment [77 &9H92 @ 92A of stu!ies $ith positi+e fin!ings). *o$e+er, null fin!ings $ere reporte! in a comparable proportion of stu!ies &!iscontinuation of "P* treatmentI 92H97 @ >CA of stu!ies $ith null fin!ings [39 , [31 , [72 M[7/ , [7C , [70 , [>9 , [>7 , [>C &reference >C refers to stu!y /< "P*-naa+e patientsI 3H97 @ /CA of stu!ies $ith null fin!ings [73 , [>2 , [>> , [>1 < an! "P* treatment !uring assessmentI 9H97 @ CA of stu!ies $ith null fin!ings [>3 ). #he t$o remaining stu!ies $ith positi+e fin!ings unfortunately !i! not specify $hether participants $ere on me!ication at the time of assessment [87 , [3C . #o summariDe, no consistent effects of "P* on participantZs performances in gambling tasks $ere obser+e! in the re+ie$e! stu!ies. An eighth potential alternate e-planatory factor is the )orm o) re%ard received 'y t*e participants, see 5igure /B. #hirteen stu!ies in this re+ie$ e-plicitly mentione! that the sub,ects coul! $in tangible re$ar!s &such as presents an! real money), of $hich sistu!ies foun! a group effect [87 , [32 , [3C , [30 , [78 , [77 an! se+en stu!ies ha! null fin!ings [39 , [72 , [7/ , [70 , [>9 , [>7 , [>1 &>H98 @ 3>A of the stu!ies using tangible re$ar!s ha! positi+e fin!ings). #he other t$el+e stu!ies use! ficti+e re$ar!s &such as points or ficti+e money), or !i! not e-plicitly mentione! that they use! real re$ar!s [31 , [79 , [73 , [7C , [71 , [>2 , [>/ M[>3 , [>> , [>C , #he results of these, ficti+e re$ar!,

stu!ies $ere also inconsistent &3H9/ @ 88A of the stu!ies using ficti+e re$ar!s ha! positi+e fin!ings).

Discussion
#he aim of this re+ie$ $as to gain more insight into the relationship bet$een A(*( an! risky performance in gambling tasks an! to i!entify potential alternate e-planatory factors that may ha+e also affecte! the outcome. In total, /7 stu!ies $ere re+ie$e! that e-amine! the performance of chil!renHa!olescents &93 stu!ies) an! a!ults &99 stu!ies) $ith A(*( on a gambling task. #en of the /7 inclu!e! stu!ies, i.e. 32A, reporte! that in!i+i!uals $ith A(*( !isplaye! more risky beha+ior in gambling tasks, as in!icate! by significantly higher scores than =Cs on outcome measures relate! to risk-taking. In terms of potential alternate e-planatory factors, age appeare! to play an important role in the relationship bet$een A(*( an! risky !ecision-making, as half of the stu!ies in chil!renHa!olescents &72A), but only a minority of stu!ies in a!ults &/CA) reporte! greater risky performance in in!i+i!uals $ith A(*( $hen compare! to =Cs. Across the stu!ies $ith chil!renHa!olescents an! a!ults effect siDes range! from small to large, $ith no clear pattern relate! to age. #he results of the stu!ies e-amine! !i! not !iffer bet$een stu!ies applying e-plicit gambling tasks, in $hich the e-act probability !istribution is e+i!ent for the participants &such as the CB#, B(#, ""B an! P(#), an! implicit gambling tasks in $hich the e-act probability !istribution is not e+i!ent &such as the 4A.#, C#, (6#, or the IB#). #his re+ie$ therefore pro+i!es e+i!ence that chil!renHa!olescents $ith A(*( appear to be likely to perform in a risky fashion in gambling tasks than =Cs &although, only 72A of the stu!ies foun! this result), $hereas a!ults $ith A(*( are less likely to perform !ifferently from =Cs on gambling tasks. #his fin!ing hol!s for both age groups irrespecti+e of the use of an implicit or e-plicit gambling task. *o$e+er, as the results +arie! bet$een stu!ies it $as in+estigate! $hether other alternate e-planatory factors coul! also help e-plain these inconsistencies. #$o stu!ies sho$e! consistently that the presence of comorbi! 6(( or C( increases the risky performance on implicit gambling tasks in chil!ren an! a!ults $ith A(*(. 5urthermore, t$o a!!itional stu!ies inclu!ing only chil!ren $ith A(*( an! a comorbi! 6((HC( ha! positi+e fin!ings. Chil!ren an! a!ults $ith A(*( an! comorbi! 6((HC( coul!, therefore, be more prone to risky performance in implicit gambling tasks. 'e+eral other alternate e-planatory factors ha+e been reporte! in the literature, such as comorbi! I(s, A(*( subtype, "P* use, an! the form of re$ar! use!. #he e+i!ence for a substantial influence of these +ariables $as limite! an!Hor inconsistent. 5uture stu!ies on risky beha+ior shoul! therefore take these +ariables into account in the stu!y !esign. #he outcomes of this re+ie$ are, ho$e+er, not likely to be !istorte! by !emographical !ifferences bet$een in!i+i!uals $ith A(*( an! =Cs $ithin the stu!ies concerning age, se-, an! intelligenceHe!ucation, because a +ast ma,ority of stu!ies controlle! for these factors by means of group matching or statistical correction. It is unclear ho$ limite! po$er of the inclu!e! stu!ies may ha+e contribute! to the pre+alence of null fin!ings. #he +ariability in effect siDes reporte! limits the possibility to compute the re%uire! sample siDe for reaching a!e%uate po$er. =e+ertheless, 9C of the /7 stu!ies &>1A)

inclu!e! A(*( samples that $ere smaller than n @ 82, $hich are in general small sample siDes. 5uture stu!ies shoul! therefore inclu!e larger sample siDes to assure a!e%uate po$er. #he result that chil!renHa!olescents $ith A(*( may be more likely to perform in a more risky fashion on gambling tasks than =Cs, irrespecti+e of the use of e-plicit or implicit tasks, implies both altere! ^col!Z an! ^hotZ !ecision-making strategies. 6n the one han!, more risky beha+ior on e-plicit gambling tasks implies an impaire! ^col!Z !ecisionmaking strategy, $hich may be !ue to !eficiencies in the cogniti+e control system that is comprise! of the !orsal an! +entral lateral prefrontal corte- an! posterior parietal corte[33 M[3> . Important capabilities in this strategy are the un!erstan!ing of probabilities, the ability to up!ate this kno$le!ge in $orking memory an! store it in long-term memory, an! to be able to inhibit responses to occasional fee!back [>0 . 6n the other han!, more risky beha+ior on implicit gambling tasks implies impaire! ^hotZ an! ^col!Z !ecision-making strategies, $hich may be !ue to !eficiencies in the cogniti+e control system as $ell as the affecti+e-moti+ational system that is comprise! of the subcortical an! cortical mi!brain !opamine systems. 5unctions that are important for ^hotZ !ecisionmaking are the processing of re$ar! an! punishment &$hich is also linke! to inhibitory control) an! the +isceral responses to these moti+ational cues [3/ , [38 . #here $as no e+i!ence that !ecision-making in A(*( $as especially impaire! on implicit gambling tasks, as $oul! be pre!icte! by purely moti+ational mo!els like the ((# [/3 an! (#( [/8 . #he stu!ies $ith positi+e results reporte! in this re+ie$ are, ho$e+er, more in line $ith A(*( mo!els that pre!ict cogniti+e !eficits [/2 , [/9 an! $ith A(*( mo!els that pre!ict combine! cogniti+e-moti+ational !eficits, i.e. the (P" [// , [/> . Purely moti+ational mo!els also !o not e-plain the impaire! performance on the stu!ies $hich reporte! positi+e fin!ings for e-plicit gambling tasks, tapping primarily ^col!Z !ecisionmaking strategies. Although, it shoul! be note! that the ma,ority of stu!ies o+erall !i! not report any impairment in A(*( participants, $hich is a challenge to both the cogniti+e an! moti+ational mo!els. ?ith regar! to moti+ational !eficiencies, the moti+ational mo!els an! combine! moti+ational an! cogniti+e mo!els ha+e primarily focuse! on the stronger !iscounting of future o+er imme!iate re$ar!s &!elay a+ersion) in in!i+i!uals $ith A(*(. #he increase! likelihoo! of chil!ren $ith A(*( to perform more risky on implicit gambling tasks foun! in the literature, ho$e+er, coul! point to an a!!itional aspect of a !ysfunctional re$ar! &or punishment) system, $hich is that chil!renHa!olescents $ith A(*( fa+or less probable large re$ar!s o+er more probable smaller re$ar!s, an! risk higher penalties for those re$ar!s. #his is in some$hat in line $ith the pre!iction of moti+ational mo!els that participants $ith A(*( $oul! perform poorer un!er partial or !iscontinuous reinforcement sche!ules, because reinforcement !uring gambling tasks is by !efinition !iscontinuous, i.e. a beha+ioral response may lea! to !ifferent outcomes &although some outcomes are more probable than others). An important ability in the face of !iscontinuous reinforcement is the use of outcome fee!back in or!er to subse%uently a!apt beha+ior or to change strategy. 6nly a fe$ stu!ies in+estigate! the use of fee!back an! the fin!ings $ere mi-e!. 'ome stu!ies !emonstrate! aberrant fee!back use in chil!renHa!olescents $ith A(*(, such as a re!uce! use of negati+e fee!back [79 an! a

re!uce! a!,ustment of strategy after a punishment [77 , [>C &reference >C refers to stu!y /). 6ther stu!ies faile! to fin! any !e+iations in fee!back use in in!i+i!uals $ith A(*( [32 , [31 , [>C . Interestingly, t$o stu!ies pro+i!e! e+i!ence that chil!renHa!olescents $ith A(*( !i! not !iffer from =Cs in the number of risky !ecisions in con!itions $ith relati+ely )re-uent punishment [72 , [7/ . #his fin!ing suggests that $hile in cases $here fee!back is fre%uent in!i+i!uals $ith A(*( react in a similar $ay to their =C peers, but in cases $here fee!back is infre%uent there are likely to be more problems. #he outcomes of this literature re+ie$ therefore pro+i!e some e+i!ence that chil!renHa!olescents $ith A(*( !o not only more strongly prefer imme!iate o+er future re$ar!s than =Cs, but also ha+e a greater preference for less probable large re$ar!s o+er more probable smaller re$ar!s, an! risk higher penalties for these larger re$ar!s. *o$e+er, there is some e+i!ence that $ith more fre%uent penalties chil!renHa!olescents are better able to !e+elop an a!+antageous strategy. #he e+i!ence for aberrant risk-taking performance on gambling tasks is stronger for chil!renHa!olescents $ith A(*( than for a!ults $ith A(*(, although is it still only present in aroun! fifty percent of the re+ie$e! stu!ies in chil!renHa!olescents. A possible e-planation for the higher proportion of positi+e stu!ies in chil!renHa!olescents is the !e+elopmental tra,ectory of A(*(, $hich is characteriDe! by a re!uction of symptoms from chil!hoo! to a!ulthoo! &often accompanie! $ith remission of A(*(). #he prefrontal reco+ery hypothesis [C8 postulates that the re!uction of A(*( symptoms !uring a!olescence is relate! to the !egree in $hich prefrontal cogniti+e control functions &^col!Z !ecision-making) compensate for primary an! persistent subcortical !eficits &^hotZ !ecision-making). #he $eaker e+i!ence for increase! risky performance in gambling tasks for a!ults compare! to chil!ren $ith A(*( may therefore be !ue to !e+elopmental impro+ements in cogniti+e control functions. Another e-planation for the !ifferent outcomes in chil!renHa!olescents an! a!ults may, ho$e+er, be that the stu!y results in this re+ie$ ha+e been influence! by publication bias. .isky beha+ior on gambling tasks in chil!renHa!olescents $ith A(*( has been stu!ie! for roughly t$enty years at the time of this re+ie$, an! in those years fin!ings in!icating group !ifferences may ha+e been gi+en preference for publication. #herefore, attempts to replicate the !ifferences obser+e! bet$een chil!renHa!olescents $ith A(*( an! =Cs in the a!ult population, may ha+e resulte! in more publications reporting null fin!ings $ithin the past ten years. #here is, therefore, a nee! for longitu!inal or cross-sectional stu!ies in or!er to !irectly test the hypothesis that the !e+elopment from chil!hoo! to a!ulthoo!, an! the relate! persistence or remittance of A(*( symptoms, influences the performance in gambling tasks in A(*(. E+en though "P* is the most prescribe! pharmacological treatment for A(*(, only t$o stu!ies in+estigate! the effects of "P* on gambling task performance. 6ne placebocontrolle! stu!y foun! that chil!ren $ith A(*( taking "P* bet fe$er points in the CB#, in!icating more conser+ati+e play. #he other placebo-controlle! stu!y $as carrie! out on a!ults an! foun! no effects of "P* on performance on the IB# an! 5PB# in a!ults $ith A(*( or in =Cs. Bi+en the small number of stu!ies an! the inconsistent fin!ings, no conclusions can be !ra$n about the effecti+eness of "P* in re!ucing risky beha+ior in gambling tasks. :iterature on the effects of "P* on risky beha+ior in the

real-$orl!, ho$e+er, suggests that "P* has a beneficial effect. "P* has for e-ample been !emonstrate! to re!uce the risk for !rug abuse [C3 an! risky !ri+ing beha+ior [C7 in in!i+i!uals $ith A(*(. #he mechanisms un!erlying these effects are unclear, so further controlle! stu!ies on this sub,ect are nee! to gain more insight into these mechanisms. Although there might be some association bet$een risky performance on gambling tasks an! chil!hoo! A(*(, little is kno$n about the relationship bet$een the performance on such tasks an! beha+ior in real life. 'ome of the re+ie$e! stu!ies in chil!ren $ith A(*( suggest that there is an association bet$een risky performance on gambling tasks an! the se+erity of A(*( symptoms [30 , [7/ , [71 . *o$e+er, such associations ha+e also been foun! for 6((HC( symptoms in these chil!ren [31 , [30 $hich is in line $ith the conclusion of this re+ie$ that the presence of comorbi! 6((HC( in A(*( increases risky beha+ior in gambling tasks. #o the best of our kno$le!ge no stu!ies are a+ailable on the ecological +ali!ity of gambling task performance in chil!renHa!olescents. *o$e+er, stu!ies on a!ults ha+e re+eale! a link bet$een risky performance in gambling tasks an! clinically rele+ant risky beha+iors, e.g. bet$een the performance in the IB# an! substance use !isor!ers, pathological gambling an! psychopathic beha+ior [C2 , bet$een performance in the P( an! pathological gambling as $ell as alcohol !epen!ence [C9 , [C/ , an! bet$een performance on the 4A.# an! self-reporte! occurrence of a!!icti+e, health, an! safety risk beha+iors [88 . #hese stu!ies on a!ults clearly suggest that there is a relationship bet$een risky performance in gambling tasks an! real-life risky beha+ior, but more research is nee!e! to firmly establish this, especially in chil!renHa!olescents.

Conclusion
#his systematic literature re+ie$ on performance in gambling tasks of in!i+i!uals $ith A(*( foun! mi-e! e+i!ence for increase! risky beha+ior. 'pecifically, in chil!ren $ith A(*( half of the stu!ies sho$e! increase! risky performance $hen compare! to =Cs. In a!ults, the e+i!ence $as $eaker, $ith only a minority of stu!ies &/CA of the stu!ies in a!ults) fin!ing any increase in risky beha+ior on gambling tasks in a!ults $ith A(*(. #he effect siDes in these stu!ies range! from small to large for both age groups. Bi+en this +ariability in effect siDes an! the generally small sample siDes &nO82 in >1A of the inclu!e! stu!ies), it is unclear $hether limite! po$er has contribute! to the mi-e! fin!ings. It is possible that the age relate! !ifference is !ue to !e+elopmental changes occurring !uring the transition from chil!hoo! to a!ulthoo!. *o$e+er, this age relate! pattern might also reflect a publication bias for positi+e fin!ings in chil!renHa!olescents in the past t$enty years of research. Concerning the gambling tasks themsel+es, the outcome !i! not !iffer bet$een stu!ies applying implicit or e-plicit gambling tasks, $hich implies that, in the cases $here risky performance $as obser+e!, both ^col!Z an! ^hotZ !ecision-making strategies may ha+e been altere! in chil!renHa!olescents $ith A(*(. #his fin!ing cannot solely be e-plaine! by moti+ational mo!els, because also ^col!Z !ecision-making appears to be !eficient, but is in line $ith the pre!ictions of cogniti+e an! combine! moti+ational an! cogniti+e mo!els of A(*(, such as the beha+ioral inhibition mo!el [/2 , e-ecuti+e functioning

mo!el [/9 , an! the (ual Path$ay "o!el [// , [/> . Although, these mo!els $oul! still struggle to e-plain $hy there are many null fin!ings in the literature. *o$e+er, gi+en the age-relate! pattern, future stu!ies shoul! aim to eluci!ate the tenability of these mo!els for a!ults $ith A(*(. ?ith regar! to potential alternate e-planatory factors, the literature in!icates that the presence of 6((HC( is a risk factor in A(*( that can result in increases in risky beha+ior in gambling tasks. 'e+eral other potential alternate e-planatory factors ha+e been reporte! in the literature, inclu!ing comorbi! I(s, A(*( subtype, use of "P*, an! the form of re$ar! use!. #he e+i!ence for a substantial contribution of these +ariables to the relationship bet$een A(*( an! risky !ecision-making $as limite! an!Hor inconsistent, especially gi+en the pre+alence of null fin!ings in the literature. *o$e+er, the outcomes of this re+ie$ are not likely !istorte! by age, se- or intelligenceHe!ucational !ifferences bet$een participants $ith A(*( an! =Cs, because the ma,ority of stu!ies controlle! for these +ariables. #he increase! risky performance in some chil!renHa!olescents $ith A(*( in implicit gambling tasks pro+i!es some e+i!ence that chil!renHa!olescents $ith A(*( !o not only prefer imme!iate o+er future re$ar!s, but also prefer less probable large re$ar!s o+er more probable smaller re$ar!s, an! risk higher penalties for these larger re$ar!s. *o$e+er, there is also some e+i!ence that $ith more fre%uent punishment that both chil!renHa!olescents $ith A(*( are better able to !e+elop an a!+antageous strategy. It remains unclear, ho$e+er, ho$ increase! risky beha+ior in gambling tasks relates to reallife !ecision-making, firstly because of the mi-e! fin!ings in the area, but also because e+i!ence for the ecological +ali!ity of the a+ailable gambling tasks is limite!, especially in chil!renHa!olescents.

)upporting "nformation
,able )3
Checklist of the gui!elines of Preferre! .eporting Items for 'ystematic .e+ie$s an! "eta-Analyses &P.I'"A). &(6C)
)licE "ere for a%%i$ional %a$a file.564D2 %oc6

,able )4
(escripti+es an! results of stu!ies consi!ere! in the re+ie$. &(6C)
)licE "ere for a%%i$ional %a$a file.591D2 %oc6

*unding )tatement
!"i# #$u%y &a# no$ fun%e% by an eF$ernal ran$in in#$i$u$ion. !"e con%uc$ of $"i# #$u%y &a# facili$a$e% by $"e *e.ar$3en$ of )linical an% *e+elo.3en$al Neuro.#yc"olo y2 /ni+er#i$y of -ronin en. !"e fun%er# "a% no role in #$u%y %e#i n2 %a$a collec$ion an% analy#i#2 %eci#ion $o .ubli#"2 or .re.ara$ion of $"e 3anu#cri.$.

Article information
PLoS One. 20134 85967 e74909. Publi#"e% online 2013 Se.$e3ber 13. %oi7 10.13718,ournal..one.0074909 P0)1*7 P0)3772864 G+onne -roen2Q9 -eral%ina F. -aa#$ra2Q :en Le&i#;E+an#2 an% Oli+er !uc"a 0a$"ia# Pe##i lione2 E%i$or *e.ar$3en$ of )linical an% *e+elo.3en$al Neuro.#yc"olo y2 /ni+er#i$y of -ronin en2 -ronin en2 !"e Ne$"erlan%# 1n#er32 France Q )on$ribu$e% eNually. 9 E;3ail7 y. roen8a$8ru .nl Competing Interests* !"e au$"or# "a+e %eclare% $"a$ no co3.e$in in$ere#$# eFi#$.

Concei+e! an! !esigne! the e-perimentsI VB 6#. Performe! the e-perimentsI BB. AnalyDe! the !ataI BB VB. Contribute! reagentsHmaterialsHanalysis toolsI 6#. ?rote the paperI VB BB 4:E.
(ecei+e% 0arc" 142 20134 'cce.$e% 'u u#$ 72 2013. )o.yri "$ no$ice !"i# i# an o.en;acce## ar$icle %i#$ribu$e% un%er $"e $er3# of $"e )rea$i+e )o33on# '$$ribu$ion Licen#e2 &"ic" .er3i$# unre#$ric$e% u#e2 %i#$ribu$ion2 an% re.ro%uc$ion in any 3e%iu32 .ro+i%e% $"e ori inal au$"or an% #ource are cre%i$e%. 'r$icle# fro3 PLoS ONE are .ro+i%e% "ere cour$e#y of )ublic +ibrary of &cience

References
1. '3erican P#yc"ia$ric '##ocia$ion 520006 *ia no#$ic an% #$a$i#$ical 3anual of 3en$al %i#or%er#. !eF$ re+i#ion 5*S0;1@;!(6. Aa#"in $on2 *)7 '3erican P#yc"ia$ric '##ocia$ion. 2. Polanc?yE -2 %e Li3a 0S2 =or$a :L2 :ie%er3an <2 (o"%e L' 520076 !"e &orl%&i%e .re+alence of '*=*7 ' #y#$e3a$ic re+ie& an% 3e$are re##ion analy#i#. '3 < P#yc"ia$ry 1647 942H948. IPub0e%J 3. Polanc?yE -2 (o"%e L' 520076 E.i%e3iolo y of a$$en$ion; %efici$8"y.erac$i+i$y %i#or%er acro## $"e life#.an. )urr O.in P#yc"ia$ry 207 386H392. IPub0e%J

4. Ailen# !E2 *o%#on A 520046 ' clinical .er#.ec$i+e of a$$en$ion; *efici$8=y.erac$i+i$y %i#or%er in$o a%ul$"oo%. < )lin P#yc"ia$ry 657 1301H1313. IPub0e%J 5. :ie%er3an <2 Ne&corn <2 S.ric" S 519916 )o3orbi%i$y of a$$en$ion %efici$ "y.erac$i+i$y %i#or%er &i$" con%uc$2 %e.re##i+e2 anFie$y2 an% o$"er %i#or%er#. '3 < P#yc"ia$ry 1487 564H577. IPub0e%J 6. :arEley ('2 0ur."y D(2 *u.aul -<2 :u#" ! 520026 *ri+in in youn a%ul$# &i$" a$$en$ion %efici$ "y.erac$i+i$y %i#or%er7 Dno&le% e2 .erfor3ance2 a%+er#e ou$co3e#2 an% $"e role of eFecu$i+e func$ionin . < 1n$ Neuro.#yc"ol Soc 87 655H672. IPub0e%J 7. (o#enbloo3 !2 Aul$? : 520116 !"ir$y;%ay #elf;re.or$e% ri#Ey %ri+in be"a+ior# of '*=* an% non;'*=* %ri+er#. 'cci% 'nal Pre+ 437 128H133. IPub0e%J 8. !"o3.#on 'L2 0olina :S-2 Pel"a3 A<2 -na y E0 520076 (i#Ey %ri+in in a%ole#cen$# an% youn a%ul$# &i$" c"il%"oo% '*=*. < Pe%ia$r P#yc"ol 327 745H759. IPub0e%J 9. :arEley (' 520016 'cci%en$# an% a$$en$ion;*efici$8=y.erac$i+i$y %i#or%er. Econ Neuro#ci 37 64H68. 10. -ay$on A2 :ailey )2 Aa ner '2 =ar%e#$y @ 519866 (ela$ion#"i. be$&een c"il%"oo% "y.erac$i+i$y an% acci%en$ .ronene##. Perce.$ 0o$ SEill# 637 801H 802. IPub0e%J 11. 0annu??a S2 Dlein (2 'biEoff =2 0oul$on 111 < 520046 Si nificance of c"il%"oo% con%uc$ .roble3# $o la$er %e+elo.3en$ of con%uc$ %i#or%er a3on c"il%ren &i$" '*=*7 ' .ro#.ec$i+e follo&;u. #$u%y. < 'bnor3 )"il% P#yc"ol 327 565H573. IPub0e%J 12. 0annu??a S2 Dlein (-2 0oul$on 111 <L 520086 Life$i3e cri3inali$y a3on boy# &i$" a$$en$ion %efici$ "y.erac$i+i$y %i#or%er7 ' .ro#.ec$i+e follo&;u. #$u%y in$o a%ul$"oo% u#in official arre#$ recor%#. P#yc"ia$ry (e# 1607 237H 246. IP0) free ar$icleJ IPub0e%J 13. Flory D2 0olina :S-2 Pel"a3 AE< -na y E2 S3i$" : 520066 )"il%"oo% '*=* .re%ic$# ri#Ey #eFual be"a+ior in youn a%ul$"oo%. < )lin )"il% P#yc"ol 357 571H577. IPub0e%J 14. A"i$e <A2 :ue"ler ) 520126 '%ole#cen$ #eFual +ic$i3i?a$ion2 '*=* #y3.$o3#2 an% ri#Ey #eFual be"a+ior. < Fa3 @iolence 277 123H132. 15. (ooney 02 )"roni#;!u#cano '2 Goon G 520126 Sub#$ance u#e in colle e #$u%en$# &i$" '*=*. < '$$en *i#or% 167 221H234. IPub0e%J 16. Lee SS2 =u3."rey# DL2 Flory D2 Liu (2 -la## D 520116 Pro#.ec$i+e a##ocia$ion of c"il%"oo% a$$en$ion;%efici$8"y.erac$i+i$y %i#or%er 5'*=*6 an% #ub#$ance u#e an% abu#e8%e.en%ence7 ' 3e$a;analy$ic re+ie&. )lin P#yc"ol (e+ 317 328H341. IP0) free ar$icleJ IPub0e%J 17. Fare " N2 *ere+en#Ey < 520116 -a3blin be"a+ior a3on a%ole#cen$# &i$" a$$en$ion %efici$8"y.erac$i+i$y %i#or%er. < -a3bl S$u% 277 243H256. IPub0e%J 18. *ere+en#Ey <L2 Pra$$ L02 =ar%oon DD2 -u.$a ( 520076 -a3blin .roble3# an% fea$ure# of a$$en$ion %efici$ "y.erac$i+i$y %i#or%er a3on c"il%ren an% a%ole#cen$#. < '%%ic$ 0e% 17 165H172. IPub0e%J 19. :reyer <L2 :o$?e$ '02 Ain$er# D)2 S$inc"fiel% (*2 'u u#$ -2 e$ al. 520096 Goun a%ul$ a3blin be"a+ior# an% $"eir rela$ion#"i. &i$" $"e .er#i#$ence of '*=*. < -a3bl S$u% 257 227H238. IP0) free ar$icleJ IPub0e%J

20. :arEley (' 519976 :e"a+ioral in"ibi$ion2 #u#$aine% a$$en$ion2 an% eFecu$i+e func$ion#7 )on#$ruc$in a unifyin $"eory of '*=*. P#yc"ol :ull 1217 65H94. IPub0e%J 21. Pennin $on :2 O?onoff S 519966 EFecu$i+e func$ion# an% %e+elo.3en$al .#yc"o.a$"olo y. < )"il% P#yc"ol P#yc"ia$ry 377 51H87. IPub0e%J 22. Sonu a;:arEe E<S 520026 P#yc"olo ical "e$ero enei$y in '*8=* ; a %ual .a$"&ay 3o%el of be"a+iour an% co ni$ion. :e"a+ :rain (e# 1307 29H36. IPub0e%J 23. Sa +ol%en !2 <o"an#en E2 'a#e =2 (u##ell @ 520056 ' %yna3ic %e+elo.3en$al $"eory of a$$en$ion;%efici$8"y.erac$i+i$y %i#or%er 5'*=*6 .re%o3inan$ly "y.erac$i+e8i3.ul#i+e an% co3bine% #ub$y.e#. :e"a+ :rain Sci 287 397H468. IPub0e%J 24. !ri.. -2 AicEen# <( 520086 (e#earc" re+ie&7 *o.a3ine $ran#fer %efici$7 ' neurobiolo ical $"eory of al$ere% reinforce3en$ 3ec"ani#3# in '*=*. < )"il% P#yc"ol P#yc"ia$ry 497 691H704. IPub0e%J 25. Lu3an 02 Oo#$erlaan <2 Ser ean$ <' 520056 !"e i3.ac$ of reinforce3en$ con$in encie# on '*8=*7 ' re+ie& an% $"eore$ical a..rai#al. )lin P#yc"ol (e+ 257 183H213. IPub0e%J 26. Sonu a;:arEe E<S 520036 !"e %ual .a$"&ay 3o%el of '*8=*7 'n elabora$ion of neuro;%e+elo.3en$al c"arac$eri#$ic#. Neuro#ci :iobe"a+ (e+ 277 593H604. IPub0e%J 27. Pa$$er#on )02 Ne&3an <P 519936 (eflec$i+i$y an% learnin fro3 a+er#i+e e+en$#7 !o&ar% a .#yc"olo ical 3ec"ani#3 for $"e #yn%ro3e# of %i#in"ibi$ion. P#yc"ol (e+ 1007 716H36. IPub0e%J 28. >uay =) 519976 1n"ibi$ion an% a$$en$ion %efici$ "y.erac$i+i$y %i#or%er. < 'bnor3 )"il% P#yc"ol 257 7H13. IPub0e%J 29. *ou la# @12 Parry P' 519946 Effec$# of re&ar% an% nonre&ar% on fru#$ra$ion an% a$$en$ion in a$$en$ion %efici$ %i#or%er. < 'bnor3 )"il% P#yc"ol 227 281H302. IPub0e%J 30. )rone E'2 <ennin # <(2 @an %er 0olen 0A 520036 Sen#i$i+i$y $o in$erference an% re#.on#e con$in encie# in a$$en$ion;%efici$8"y.erac$i+i$y %i#or%er. < )"il% P#yc"ol P#yc"ia$ry 447 214H226. IPub0e%J 31. -roen G2 0ul%er L<02 Ai,er# ''2 0in%eraa (:2 'l$"au# 0 520096 0e$"yl."eni%a$e i3.ro+e# %i3ini#"e% error an% fee%bacE #en#i$i+i$y in '*=*7 'n e+oEe% "ear$ ra$e analy#i#. :iol P#yc"ol 827 45H53. IPub0e%J 32. 1aboni F2 *ou la# @12 *i$$o : 519976 P#yc"o."y#iolo ical re#.on#e of '*=* c"il%ren $o re&ar% an% eF$inc$ion. P#yc"o."y#iolo y 347 116H123. IPub0e%J 33. Le,ue? )A2 (ea% <P2 Da"ler )A2 (ic"ar%# <:2 (a3#ey SE2 e$ al. 520026 E+alua$ion of a be"a+ioral 3ea#ure of ri#E $aEin 7 !"e balloon analo ue ri#E $a#E 5:'(!6. < EF. P#yc"ol '..l 87 75H84. IPub0e%J 34. Ne&3an <P2 Pa$$er#on )02 Do##on *S 519876 (e#.on#e .er#e+era$ion in .#yc"o.a$"#. < 'bnor3 P#yc"ol 967 145H148. IPub0e%J 35. *au "er$y !D2 >uay =) 519916 (e#.on#e .er#e+era$ion an% %elaye% re#.on%in in c"il%"oo% be"a+ior %i#or%er#. < )"il% P#yc"ol P#yc"ia$ry 327 453H461. IPub0e%J 36. :ec"ara '2 *a3a#io '(2 *a3a#io =2 'n%er#on SA 519946 1n#en#i$i+i$y $o fu$ure con#eNuence# follo&in %a3a e $o "u3an .refron$al cor$eF. )o ni$ion 507 7H15. IPub0e%J

37. :ran% 02 (ecEnor E)2 -raben"or#$ F2 :ec"ara ' 520076 *eci#ion# un%er a3bi ui$y an% %eci#ion# un%er ri#E7 )orrela$ion# &i$" eFecu$i+e func$ion# an% co3.ari#on# of $&o %ifferen$ a3blin $a#E# &i$" i3.lici$ an% eF.lici$ rule#. < )lin EF. Neuro.#yc"ol 297 86H99. IPub0e%J 38. (o er# (*2 E+eri$$ :<2 :al%acc"ino '2 :lacE#"a& '<2 S&ain#on (2 e$ al. 519996 *i##ociable %efici$# in $"e %eci#ion;3aEin co ni$ion of c"ronic a3."e$a3ine abu#er#2 o.ia$e abu#er#2 .a$ien$# &i$" focal %a3a e $o .refron$al cor$eF2 an% $ry.$o."an;%e.le$e% nor3al +olun$eer#7 E+i%ence for 3onoa3iner ic 3ec"ani#3#. Neuro.#yc"o."ar3acolo y 207 322H339. IPub0e%J 39. :ran% 02 Fu,i&ara E2 :or#u$?Ey S2 Dalbe E2 De##ler <2 e$ al. 520056 *eci#ion;3aEin %efici$# of Eor#aEoff .a$ien$# in a ne& a3blin $a#E &i$" eF.lici$ rule#7 '##ocia$ion# &i$" eFecu$i+e func$ion#. Neuro.#yc"olo y 197 267H277. IPub0e%J 40. *rec"#ler (2 (i??o P2 S$ein"au#en =) 520106 *eci#ion 3aEin &i$" uncer$ain reinforce3en$ in c"il%ren &i$" a$$en$ion %efici$8"y.erac$i+i$y %i#or%er 5'*=*6. )"il% Neuro.#yc"ol 167 145H161. IPub0e%J 41. Sc"ere# '2 *i,E#$ra 02 'in#lie E2 :alEan <2 (eynol%# :2 e$ al. 520066 !e3.oral an% .robabili#$ic %i#coun$in of re&ar%# in c"il%ren an% a%ole#cen$#7 Effec$# of a e an% '*=* #y3.$o3#. Neuro.#yc"olo ia 447 2092H2103. IPub0e%J 42. Se uin <(2 'r#eneaul$ L2 !re3blay (E 520076 !"e con$ribu$ion of RcoolS an% R"o$S co3.onen$# of %eci#ion;3aEin in a%ole#cence7 13.lica$ion# for %e+elo.3en$al .#yc"o.a$"olo y. )o n *e+ 227 530H543. 43. *unn :2 *al lei#" !2 La&rence ' 520066 !"e #o3a$ic 3arEer "y.o$"e#i#7 ' cri$ical e+alua$ion. Neuro#ci :iobe"a+ (e+ 307 239H71. IPub0e%J 44. )a#ey :<2 -e$? S2 -al+an ' 520086 !"e a%ole#cen$ brain. *e+ (e+ 287 62H 77. IP0) free ar$icleJ IPub0e%J 45. )o"en < 520056 !"e +ulcani?a$ion of $"e "u3an brain7 ' neural .er#.ec$i+e on in$erac$ion# be$&een co ni$ion an% e3o$ion. < Econ Per#.ec$ 197 3H24. 46. S$einber L 520046 (i#E $aEin in a%ole#cence ; &"a$ c"an e#2 an% &"yP '%ole#cen$ :rain *e+elo.3en$7 @ulnerabili$ie# an% O..or$uni$ie# 10217 51H 58. IPub0e%J 47. -aron N2 0oore )2 Aa#c"bu#c" *' 520066 *eci#ion 3aEin in c"il%ren &i$" '*=* only2 '*=*;anFiou#8%e.re##e%2 an% con$rol c"il%ren u#in a c"il% +er#ion of $"e io&a a3blin $a#E. < '$$en *i#or% 97 607H619. IPub0e%J 48. -eur$# =02 @an %er Oor% S2 )rone E' 520066 =o$ an% cool a#.ec$# of co ni$i+e con$rol in c"il%ren &i$" '*=*7 *eci#ion;3aEin an% in"ibi$ion. < 'bnor3 )"il% P#yc"ol 347 813H824. IPub0e%J 49. =ob#on )A2 Sco$$ S2 (ubia D 520116 1n+e#$i a$ion of cool an% "o$ eFecu$i+e func$ion in O**8)* in%e.en%en$ly of '*=*. < )"il% P#yc"ol P#yc"ia$ry 527 1035H1043. IPub0e%J 50. Lu3an 02 Oo#$erlaan <2 Dnol *L2 Ser ean$ <' 520086 *eci#ion;3aEin in '*=*7 Sen#i$i+e $o freNuency bu$ blin% $o $"e 3a ni$u%e of .enal$yP < )"il% P#yc"ol P#yc"ia$ry 497 712H722. IPub0e%J 51. 0a#una3i !2 OEa?aEi S2 0aeEa&a = 520096 *eci#ion;3aEin .a$$ern# an% #en#i$i+i$y $o re&ar% an% .uni#"3en$ in c"il%ren &i$" a$$en$ion;%efici$ "y.erac$i+i$y %i#or%er. 1n$ < P#yc"o."y#iol 727 283H288. IPub0e%J

52. !o.laE 0E2 <ain /2 !annocE ( 520056 EFecu$i+e an% 3o$i+a$ional .roce##e# in a%ole#cen$# &i$" a$$en$ion;%efici$;"y.erac$i+i$y %i#or%er 5'*=*6. :e"a+ :rain Func$ 17 8. IP0) free ar$icleJ IPub0e%J 53. 0a$$"y# A2 @an -oo?en S=02 *e @rie# =2 )o"en;De$$eni# P2 @an En elan% = 519986 !"e %o3inance of be"a+ioural ac$i+a$ion o+er be"a+ioural in"ibi$ion in con%uc$ %i#or%ere% boy# &i$" or &i$"ou$ a$$en$ion %efici$ "y.erac$i+i$y %i#or%er. < )"il% P#yc"ol P#yc"ia$ry 397 643H651. IPub0e%J 54. Aier# (A2 -unnin A:2 Ser ean$ <' 519986 1# a 3il% %efici$ in eFecu$i+e func$ion# in boy# rela$e% $o c"il%"oo% '*=* or $o .aren$al 3ul$i enera$ional alco"oli#3P < 'bnor3 )"il% P#yc"ol 267 415H430. IPub0e%J 55. =u3."rey# DL2 Lee SS 520116 (i#E $aEin an% #en#i$i+i$y $o .uni#"3en$ in c"il%ren &i$" '*=*2 O**2 '*=*TO**2 an% con$rol#. < P#yc"o.a$"ol :e"a+ '##e## 337 299H307. 56. )rone E'2 @an %er 0olen 0A 520046 *e+elo.3en$al c"an e# in real life %eci#ion 3aEin 7 Perfor3ance on a a3blin $a#E .re+iou#ly #"o&n $o %e.en% on $"e +en$ro3e%ial .refron$al cor$eF. *e+ Neuro.#yc"ol 257 251H 279. IPub0e%J 57. *e@i$o EE2 :lacE&ell '*2 Den$ L2 Er#c"e D*2 )larE L2 e$ al. 520086 !"e effec$# of 3e$"yl."eni%a$e on %eci#ion 3aEin in a$$en$ion; %efici$8"y.erac$i+i$y %i#or%er. :iol P#yc"ia$ry 647 636H639. IP0) free ar$icleJ IPub0e%J 58. *rec"#ler (2 (i??o P2 S$ein"au#en =) 520086 *eci#ion;3aEin on an eF.lici$ ri#E;$aEin $a#E in .rea%ole#cen$# &i$" a$$en$ion;%efici$8"y.erac$i+i$y %i#or%er. < Neural !ran#3 1157 201H209. IPub0e%J 59. ' ay N2 Gec"ia3 E2 )ar3el K2 Le+Eo+i$? G 520106 Non;#.ecific effec$# of 3e$"yl."eni%a$e 5ri$alin6 on co ni$i+e abili$y an% %eci#ion;3aEin of '*=* an% "eal$"y a%ul$#. P#yc"o."ar3acolo y 5:erl 6 2107 511H519. IPub0e%J 60. *uar$e N'2 Aoo%# SP2 (ooney '2 '$Ein#on <=2 -ran$ 1 520126 AorEin 3e3ory %efici$# affec$ ri#Ey %eci#ion;3aEin in 3e$"a3."e$a3ine u#er# &i$" a$$en$ion;%efici$8"y.erac$i+i$y %i#or%er. < P#yc"ia$r (e# 467 492H499. IP0) free ar$icleJ IPub0e%J 61. Ern#$ 02 Di3e# 'S2 Lon%on E*2 0a$oc"iE <'2 El%re$" *2 e$ al. 520036 Neural #ub#$ra$e# of %eci#ion 3aEin in a%ul$# &i$" a$$en$ion %efici$ "y.erac$i+i$y %i#or%er. '3 < P#yc"ia$ry 1607 1061H1070. IPub0e%J 62. 0alloy;*ini? L2 Fuen$e# *2 Lei$e A:2 )orrea =2 :ec"ara ' 520076 13.ul#i+e be"a+ior in a%ul$# &i$" a$$en$ion %efici$8"y.erac$i+i$y %i#or%er7 )"arac$eri?a$ion of a$$en$ional2 3o$or an% co ni$i+e i3.ul#i+ene##. < 1n$ Neuro.#yc"ol Soc 137 693H698. IPub0e%J 63. 0alloy;*ini? L2 Lei$e A:2 %e 0orae# P2 =enriNue Pai+a2 )orrea =2 :ec"ara '2 e$ al. 520086 :ra?ilian .or$u ue#e +er#ion of $"e io&a a3blin $a#E7 !ran#cul$ural a%a.$a$ion an% %i#cri3inan$ +ali%i$y. (e+ :ra# P#iNuia$r 307 144H148. IPub0e%J 64. 0Un$ylU !2 S$ill <2 -ullber S2 *el 0i##ier F 520126 *eci#ion 3aEin in a%ul$# &i$" '*=*. < '$$en *i#or% 167 164H173. IPub0e%J 65. Aeafer <2 0ilic" (2 Fill3ore 0! 520116 :e"a+ioral co3.onen$# of i3.ul#i+i$y .re%ic$ alco"ol con#u3.$ion in a%ul$# &i$" '*=* an% "eal$"y con$rol#. *ru 'lco"ol *e.en% 1137 139H146. IP0) free ar$icleJ IPub0e%J 66. Fi#c"er 02 :arEley ('2 S3alli#" L2 Fle$c"er D 520056 EFecu$i+e func$ionin in "y.erac$i+e c"il%ren a# youn a%ul$#7 '$$en$ion2 in"ibi$ion2 re#.on#e

.er#e+era$ion2 an% $"e i3.ac$ of co3orbi%i$y. *e+ Neuro.#yc"ol 277 107H 133. IPub0e%J 67. 0a$$"ie# S2 P"ili.#en '2 S+al%i < 520126 (i#Ey %eci#ion 3aEin in a%ul$# &i$" '*=*. < :e"a+ !"er EF. P#yc"ia$ry 437 938H946. IPub0e%J 68. Ailber$? -2 +an El#$ L!2 *el a%o 0(2 0aier S2 Fei e :2 e$ al. 520126 Orbi$ofron$al re&ar% #en#i$i+i$y an% i3.ul#i+i$y in a%ul$ a$$en$ion %efici$ "y.erac$i+i$y %i#or%er. Neuroi3a e 607 353H361. IPub0e%J 69. @an *ui,+en+oor%e ')D2 :re%3an <2 =ui?en a = 520126 ' e;rela$e% c"an e# in %eci#ion 3aEin 7 )o3.arin infor3e% an% noninfor3e% #i$ua$ion#. *e+ P#yc"ol 487 192H203. IPub0e%J 70. :uelo& 0!2 Su"r <' 520096 )on#$ruc$ +ali%i$y of $"e io&a a3blin $a#E. Neuro.#yc"ol (e+ 197 102H114. IPub0e%J 71. Pe$ry N0 520126 *i#coun$in of .robabili#$ic re&ar%# i# a##ocia$e% &i$" a3blin ab#$inence in $rea$3en$;#eeEin .a$"olo ical a3bler#. < 'bnor3 P#yc"ol 1217 151H159. IP0) free ar$icleJ IPub0e%J 72. !aEa"a#"i !2 O"3ura G2 Oono =2 (a%for% 0 520096 'lco"ol u#e an% %i#coun$in of %elaye% an% .robabili#$ic ain an% lo##. Neuroen%ocrinol Le$$ 307 749H752. IPub0e%J 73. =al.erin <02 Sc"ul? DP 520066 (e+i#i$in $"e role of $"e .refron$al cor$eF in $"e .a$"o."y#iolo y of a$$en$ion;%efici$8"y.erac$i+i$y %i#or%er. P#yc"ol :ull 1327 560H581. IPub0e%J 74. Faraone S@2 Ailen# ! 520036 *oe# #$i3ulan$ $rea$3en$ lea% $o #ub#$ance u#e %i#or%er#P < )lin P#yc"ia$ry 647 9H13. IPub0e%J 75. )oF *<2 *a+i# 02 0iEa3i 'G2 Sin " =2 0erEel (L2 e$ al. 520126 Lon ;ac$in 3e$"yl."eni%a$e re%uce% colli#ion ra$e# of youn a%ul$ %ri+er# &i$" a$$en$ion; %efici$8"y.erac$i+i$y %i#or%er. < )lin P#yc"o."ar3acol 327 225H230. IPub0e%J

4 Li"a$ .%f 5 Li"a$ .%f 6

Raising Awareness of )leep as a -ealthy Beha$ior

Beral!ine '. Perry, (rP*, .(=, 'usheel P. Patil, "(, Ph(, an! :etitia .. PresleyCantrell, Ph( A!!itional article information 'leep is an essential component of health, an! its timing, !uration, an! %uality are critical !eterminants of health &9). 'leep may play an important role in metabolic regulation, emotion regulation, performance, memory consoli!ation, brain recuperation processes, an! learning &/). 4ecause of the importance of these functions, sleep shoul! be +ie$e! as being as critical to health as !iet an! physical acti+ity. *o$e+er, public health practitioners an! other health care pro+i!ers ha+e not focuse! ma,or attention on the importance of sleep to health. In this essay, $e briefly summariDe the scientific literature about hours of sleep nee!e! an! $hy sleep is an important public health issue. ?e also suggest areas for e-pan!ing sleep research an! strategies for increasing a$areness of the importance of sleep an! impro+ing sleep health. 5inally, $e call for action to bring sleep to the forefront of public health.

-ow Much )leep is eeded and Are 5e ,here6


#he /22> Institute of "e!icine &I6") report .leep "isorders and .leep "eprivation in!icates that the a+erage basal sleep nee!s of a!ults is appro-imately C to 1 hours per night, an! the optimal sleep !uration for a!olescents is 0 hours per night &9). *o$e+er, more than 87A of a!ults report getting fe$er than C hours of sleep !uring a /3-hour perio! &8,3), an! almost C2A of high school stu!ents report getting fe$er than 1 hours of sleep on an a+erage $eeknight &7). 6+erall, about 97 million chil!ren in the Enite! 'tates !o not get sufficient sleep &>). Among a!ults, the reasons for sleep loss appear to be relate! mainly to lifestyle, $ork sche!ules &shift $ork an! long hours), or sleep !isor!ers &9). Appro-imately /2A of $orkers are engage! in shift $ork, $hich often lea!s to longer $ork hours &9). Among a!olescents, insufficient sleep is associate! $ith greater use of social me!ia technology, an! among younger chil!ren it is associate! $ith !epressi+e symptomatology, family !isagreements, an! safety issues aroun! home, school an! neighborhoo! &>).

5hy "s )leep a Public -ealth "ssue6


Insufficient sleep has ma,or health conse%uences in a!ults, a!olescents, an! young chil!ren. 'trong e+i!ence e-ists that among a!ults insufficient sleep has a significant effect on numerous health con!itions, inclu!ing chronic !isease !e+elopment an! inci!ence &9). 5or instance, short sleep !uration &OC hours of sleep per night) an! poor sleep %uality are associate! $ith car!io+ascular morbi!ity an! metabolic !isor!ers such as glucose intolerance, $hich may lea! to obesity, !iabetes, heart !isease, an! hypertension &9). People $ho ha+e short sleep !uration are at 9.31 times greater risk of !e+eloping an! !ying of coronary heart !isease than controls an! 9.97 times more likely

to ha+e a stroke &C). Chil!ren $ho e-perience short sleep !uration are more likely to become obese than those $ho !o not &1). Insufficient sleep also affects immunologic function an! !e+elopment of moo! !isor!ers an! is associate! $ith !epression< !eficits in cognition, memory an! learning< an! re!uce! %uality of life &9). A!ults $ho sleep fe$er than C hours per night ha+e greater !ifficulty concentrating, remembering, an! performing other !aily acti+ities than those $ho sleep C to 0 hours a night &3). Chil!ren an! a!olescents $ho get insufficient sleep ha+e impaire! beha+ior, moo!, an! performance &0). 6ne ma,or conse%uence of insufficient sleep is !aytime sleepiness, $hich re!uces alertness an! causes slo$ reaction time, lea!ing to occupational an! me!ical errors, $orkplace in,uries, impaire! !ri+ing, an! motor +ehicle acci!ents &9). In /220, almost 7A of a!ults in 9/ states reporte! that !uring the pre+ious 82 !ays they ha! no!!e! off or fallen asleep $hile !ri+ing &8). In /227, !ro$sy !ri+ing contribute! to 922,222 motor +ehicle acci!ents an! 97,222 !eaths &92). #he public health bur!en of sleep !epri+ation is enormous. #here are substantial public health in+estments in all areas relate! to sleep, from obesity an! other chronic con!itions to motor +ehicle acci!ents. Insufficient sleep, unlike other health risk factors such as smoking, e-cessi+e alcohol consumption, obesity, an! physical inacti+ity, has historically recei+e! much less attention in the public health an! clinical settings. Insufficient sleep is an important public health risk factor that $oul! benefit from further in+estigation.

/ack of Awareness
(espite strong e+i!ence of the relationship bet$een insufficient sleep an! health problems, most people are una$are of the amount of sleep they nee!, their le+el of sleep !epri+ation, an! the negati+e impact of sleep !epri+ation on health. 4ecause of lack of a$areness, sleep is not commonly incorporate! into public health approaches. In a!!ition, many health care pro+i!ers !o not counsel their patients about healthy sleep habits &99). In a stu!y of health care screening among 9/9 primary care clinics, only 38A inclu!e! sleep-relate! %uestions on their screening batteries compare! $ith 922A for smoking an! alcohol, 08A for healthy eating, an! 1>A for physical acti+ity &99). It is not clear $hy sleep is not inclu!e! in health screenings, but it may be relate! to the clinicianZs lack of kno$le!ge of the importance of sleep. In /22/, only 92A of primary care pro+i!ers !escribe! their kno$le!ge of sleep an! sleep !isor!ers as goo! &9/). Although little e+i!ence e-ists on the effecti+eness of sleep screening an! counseling on sleep beha+ior, screening an! counseling has been sho$n to impro+e the health beha+iors of patients in other areas, such as !ietary habits, smoking cessation, an! physical acti+ity &98). #herefore, gi+ing pro+i!ers information about screening an! counseling for appropriate sleep time an! nee!s coul! better e%uip primary care an! public health professionals $ith the kno$le!ge nee!e! to screen an! counsel patients to promote sleep as a healthy beha+ior &9). *o$e+er, further in+estigation is nee!e! on the effecti+eness of sleep screening an! sleep counseling.

)trategies to "mpro$e Awareness and )leep -ealth


Information about the physiology of sleep an! sleep !isor!ers is $i!ely a+ailable, but much less $ork has been !one on effecti+e strategies to promote sleep as a healthy beha+ior. #his fiel! is prime for public health in+estigations an! inter+entions to re!uce the negati+e effect of insufficient sleep as a common risk factor for many health outcomes. 'ome suggeste! strategies for impro+ing sleep initiation an! sleep maintenance, !uration, an! %uality are consistency in be!time an! rising< maintaining an appropriate sleeping en+ironment &!ark, rela-ing, not too hot or col!)< a+oi!ing tele+ision -$atching before be!, a+oi!ing use of electronics or rea!ing in the be!room< an! a+oi!ing large meals an! physical acti+ity before going to be! &$$$.sleepfoun!ation.org). *o$e+er, more research is nee!e! to e+aluate the effecti+eness of these suggeste! strategies to impro+e sleep beha+ior an! health. #he I6" report calls for se+eral approaches to re!uce the public health bur!en of insufficient sleep through increasing public a$areness of the importance of sleep an! impro+ing !iagnosis an! treatment of sleep !isor!ers &9). .eaching these goals $ill re%uire 9) impro+e! public e!ucation on the nee! for sleep an! the conse%uences of insufficient sleep< /) more training for public health professionals an! health care pro+i!ers on screening an! counseling< an! 8) impro+e! e+i!ence of the bur!en of insufficient sleep ac%uire! through sur+eillance an! monitoring tools. 5e!eral agencies, public health partners, an! pri+ate organiDations are collaborating to employ I6" strategies. 5or e-ample, #he =ational 'leep A$areness .oun!table &='A.#) &$$$.nsart.org), a national coalition of go+ernment, professional, +olunteer, an! other organiDations, is collaborating to raise a$areness about sleep among the public, increase the un!erstan!ing of the importance of sleep, an! re!uce the public health an! safety impact of sleep !epri+ation an! sleep !isor!ers by impro+ing communication an! collaboration among local, state, an! fe!eral agencies. ='A.# member organiDations ha+e contribute! to sleep a$areness by pro+i!ing training $orkshops on healthy sleep for primary care pro+i!ers, by pro!ucing free local initiati+es to e!ucate primary care health pro+i!ers on sleep an! sleep !isor!ers, by promoting (ro$sy (ri+ing Pre+ention ?eek an! =ational 'leep A$areness ?eek to e!ucate the public, an! by publishing research fin!ings. 'e+eral of the Centers for (isease Control an! Pre+entionZs sur+eillance systems ha+e a!!e! %uestions on sleep to pro+i!e state an! national !ata on the bur!en of insufficient sleep &93M9>)< ho$e+er, more national !ata are nee!e! on young chil!ren &age! 2M9/ years). #he /2/2 *ealth 6b,ecti+es &9C) a!!e! sleep as one of its ne$ areas, focusing on increasing the proportion of a!ults an! stu!ents in gra!es 0 through 9/ $ho get sufficient sleep, !ecreasing the number of +ehicular crashes resulting from !ro$sy !ri+ing, an! increasing the proportion of persons $ith sleep apnea symptoms $ho are e+aluate!.

Call to Action

4ecause of the lack of a$areness of the benefits of healthy sleep, multisectoral public health campaigns, similar to those relate! to smoking cessation an! re!ucing e-cessi+e alcohol consumption, are nee!e! to e!ucate the public about the importance of sleep an! the conse%uences of insufficient sleep. 'uggeste! strategies to impro+e sleep health inclu!e the follo$ingI

.esearch on the effecti+eness of screening an! counseling efforts E!ucation of employers on the health effects of long shifts an! insufficient sleep (elaying school start time for high school stu!ents E!ucating the public on the risks of !ro$sy !ri+ing Impro+ing sur+eillance of sleep health, especially among young chil!ren

5inally, the critical public health message isI 'leep is essential for goo! health< it is a necessity, not a lu-ury.

Acknowledgments
?e ackno$le!ge the follo$ing members of ='A.# $ho pro+i!e! insight an! comments !uring the !e+elopment an! re+ie$ of this essayI Phyllis C. See, "(, Ph(< (a+i! B. (a+ila, "(< .obert C. 4asner, "(< (a+i! A. 'chulman, "(, "P*, 5CCP< E!$ar! Bran!i< :ee J. 4rooks, "(< ?ayne Biles, "(, "'< Janet 4. Croft, Ph(< .oger .. .osa, Ph(< an! *arol! ?. Bor!on, Ph(.

*ootnotes
#he opinions e-presse! by authors contributing to this ,ournal !o not necessarily reflect the opinions of the E.'. (epartment of *ealth an! *uman 'er+ices, the Public *ealth 'er+ice, the Centers for (isease Control an! Pre+ention, or the authorsK affiliate! institutions. .uggested citation )or t*is article/ Perry B', Patil 'P, Presley-Cantrell :.. .aising A$areness of 'leep as a *ealthy 4eha+ior. Pre+ Chronic (is /298<92I982219. (6II httpIHH!-.!oi.orgH92.7111Hpc!92.982219.

Article information
Pre+ Chronic (is. /298< 92I E988. Publishe! online /298 August 1. !oiI 92.7111Hpc!92.982219 P"CI(I P"C8C3939/ Beral!ine '. Perry, (rP*, .(=, 'usheel P. Patil, "(, Ph(, an! :etitia .. PresleyCantrell, Ph( Author AffiliationsI 'usheel P. Patil, Johns *opkins 'leep (isor!ers Center, 4altimore, "arylan!< :etitia .. Presley-Cantrell, Centers for (isease Control an! Pre+ention, Atlanta, Beorgia. Correspon!ing author.

Correspon!ing AuthorI Beral!ine ' Perry, (rP*, .(=, :ea! *ealth 'cientist, Epi!emiology an! 'ur+eillance 4ranch, (i+ision of Population *ealth, Centers for (isease Control an! Pre+ention, 3CC2 4ufor! *$y, =E, "ailstop K->C, Atlanta BA 82839. #elephoneI CC2-311-79C3. E-mailI gperryHatHc!c.go+. Copyright notice Articles from Pre+enting Chronic (isease are pro+i!e! here courtesy of Centers for Disease Control and Pre$ention

References
9. Colten *., Alte+ogt 4", e!itors. 'leep !isor!ers an! sleep !epri+ationI an unmet public health problem. Institute of "e!icine &E') Committee on 'leep "e!icine an! .esearch. ?ashington &(C)I #he =ational Aca!emies Press< /22>. [Pub"e! /. 'iegel J". Clues to the function of mammalian sleep. =ature /227<38C&C2>8)I9/>3M C9. !oiI 92.9281Hnature23/17. [Pub"e! [Cross .ef 8. Centers for (isease Control an! Pre+ention Enhealthy sleep-relate! beha+iors b 9/ 'tates, /220. ""?. "orb "ortal ?kly .ep/299<>2&1)I/88M1. [Pub"e! 3. Centers for (isease Control an! Pre+ention Effect of short sleep !uration on !aily acti+ities b Enite! 'tates, /227M/221. ""?. "orb "ortal ?kly .ep/299<>2&1)I/80M 3/. [Pub"e! 7. "cKnight-Eily :., Eaton (K, :o$ry ., Croft J4, Presley-Cantrell :., Perry B'. .elationships bet$een hours of sleep an! health-risk beha+iors in E' a!olescent stu!ents. Pre+ "e! /299<78&3-7)I/C9M8. !oiI 92.929>H,.ypme!./299.2>.2/2. [Pub"e! [Cross .ef >. 'mal!one A, *onig JC, 4ryrne "?. 'leepless in AmericaI ina!e%uate sleep an! relationships to health an! $ell-being of our nationZs chil!ren. Pe!iatrics /22C<990&'uppl 9)I'/0M8C. !oiI 92.973/Hpe!s./22>-/2105. [Pub"e! [Cross .ef C. Cappuccio 5P, Cooper (, (ZElia :, 'traDDullo P, "iller "A. 'leep !uration pre!icts car!io+ascular outcomesI a systematic re+ie$ an! meta-analysis of prospecti+e stu!ies. Eur *eart J /299<8/&9/)I9313M0/. !oiI 92.9208Heurheart,Hehr22C. [Pub"e! [Cross .ef 1. Cappuccio 5P, #aggart 5", Kan!ala =4, Currie A, Peile E, 'tranges ', et al. "etaanalysis of short sleep !uration an! obesity in chil!ren an! a!ults. 'leep /221<89&7)I>90M />. [P"C free article [Pub"e! 0. 5allone B, 6$ens JA, (eane J. 'leepiness in chil!ren an! a!olescentsI clinical implications. 'leep "e! .e+ /22/<>&3)I/1CM82>. !oiI 92.9278Hsmr+./229.290/. [Pub"e! [Cross .ef 92. Paca 5. =ational *igh$ay #raffic 'afety A!ministration notes. (ro$sy !ri+ing. Ann Emerg "e! /227<37&3)I388M3. !oiI 92.929>H,.annemergme!./227.29.297. [Pub"e! [Cross .ef 99. 'orscher AJ. *o$ is your sleepI a neglecte! topic for health care screening. J Am 4oar! 5am "e! /221</9&/)I939M1. !oiI 92.89//H,abfm./221.2/.2C29>C. [Pub"e! [Cross .ef 9/. Papp KK, Penro! CE, 'trohl KP. Kno$le!ge an! attitu!es of primary care physicians to$ar! sleep an! sleep !isor!ers. 'leep 4reath /22/<>&8)I928M0. !oiI 92.9277Hs-/22/8389C. [Pub"e! [Cross .ef

98. Kreuter "?, Chhe!a 'B, 4ull 5C. *o$ !oes physician a!+ice influence patient beha+iorT e+i!ence for a priming effect. Arch 5am "e! /222<0&7)I3/>M88. !oiI 92.9229Harchfami.0.7.3/>. [Pub"e! [Cross .ef 93. 4eha+ioral .isk 5actor 'ur+eillance 'ystem /221 'ur+ey `uestionnaire. Atlanta &BA)I E' (epartment of *ealth an! *uman 'er+ices, Centers for (isease Control an! Pre+ention, /221. httpIHH$$$.c!c.go+HbrfssH%uestionnaires.htm Accesse! June 93, /298. 97. =ational *ealth an! =utrition E-amination 'ur+ey. `uestionnaires, !atasets, an! relate! !ocumentation. /299-/29/ =ational *ealth an! =utrition E-amination 'ur+ey &=*A=E'). 'ur+ey `uestionnaires, E-amination Components an! :aboratory Components /299-/29/. 'ample Person `uestionnaire. Atlanta &BA)I E' (epartment of *ealth an! *uman 'er+ices, Centers for (isease Control an! Pre+ention, =ational Center for *ealth 'tatistics, /29/. httpIHH$$$.c!c.go+HnchsHnhanesHnhanes/299/29/H%ueste-am99X9/.htm Accesse! June 93, /298. 9>. =ational *ealth Inter+ie$ 'ur+ey. `uestionnaires, !atasets, an! relate! !ocumentation. 900C-Present (ata, `uestionnaires an! relate! !ocumentation. /299 =*I' 'ur+ey `uestionnaires. Atlanta &BA)I E' (epartment of *ealth an! *uman 'er+ices, Centers for (isease Control an! Pre+ention, =ational Center for *ealth 'tatistics, /299. httpIHH$$$.c!c.go+HnchsHnhisH%uestX!ataXrelate!X900CXfor$ar!.htm Accesse! June 93, /298. 9C. People *. /2/2. /2/2 #opics ; ob,ecti+es. 'leep *ealth. ?ashington &(C)I E' (epartment of *ealth an! *uman 'er+ices, =o+ember /292. httpIHH$$$.healthypeople.go+ Accesse! June 93, /298.

*ealth an! `uality of :ife 6utcomes 4io"e! Central

-ealth!related 7uality of life of children and adolescents with mental disorders


Katharina ?eitkamp, Ju!ith K (aniels, [... , an! 'ilke ?iegan!-Brefe A!!itional article information

Abstract
Background
#he aim $as to assess the association of internalising an! e-ternalising pathology $ith the chil!Zs health-relate! %uality of life &`o:), an! to !etermine $hich chil! an! en+ironmental characteristics beyon! pathology $ere relate! to poor `o:.

Methods
(ata $as obtaine! for 9/2 chil!ren an! a!olescents &age! > to 91) commencing outpatient psychotherapy treatment. Parents an! chil!ren &age! 99 years an! ol!er) fille! out %uestionnaires. `o: $as measure! $ith the KI('C.EE=-/C.

Results
`o: $as more strongly associate! $ith internalising than e-ternalising pathology accor!ing to both self- an! parent report. "ultiple regression analyses sho$e! that beyon! internalising an! e-ternalising pathology, gen!er, age, family functioning, functional impairment, an! prior mental health treatment $ere associate! $ith in!i+i!ual `o: scales.

Conclusions
#he !ata un!erscore! the relationship bet$een mental pathology an! impaire! `o: e+en if potential item o+erlap $as controlle! for. #his stresses the importance of e-ten!ing therapy goals an! outcome measures from mere pathology to measures of `o: in psychotherapy research particularly for patients $ith internalising pathology. Keywords: `uality of life, Internalising !isor!ers, E-ternalising !isor!ers, Chil!, A!olescent

"ntroduction
.esearch on health-relate! %uality of life &`o:) in chil!ren an! a!olescents $ith psychiatric !isor!ers is still in its early stages. #hough in e+i!ence-base! assessment it has been suggeste! to supplement the measure of pathology $ith the assessment of `o: an! functioning in chil! psychotherapy research [9 . #his seems $arrante!, since suffering from a psychiatric !isor!er !uring chil!hoo! an! a!olescence has a consi!erable impact on the chil!Zs sub,ecti+e satisfaction $ith his or her !ay-to-!ay acti+ities an! social $ell-being [/,8 . `o: in chil!ren $ith mental health problems is not only consi!erably poorer than in healthy chil!ren, but `o: seems to be more se+erely impaire! e+en $hen compare! $ith chil!ren suffering from a chronic somatic illness [3,7 . It seems critical to search for influencing factors beyon! the specific association $ith pathology so that therapists might respon! in a more targete! $ay. In a first stu!y on factors influencing `o: in chil!ren $ith a psychiatric !isor!er, 4astiaansen an! his colleagues [> note! that the impact of psychopathology $as larger in girls than in boys, possibly because chil!ren $ith e-ternalising beha+iour problems &pre!ominantly males) may not e-perience their symptoms as problematic. 'econ!ly, the impact of psychopathology increase! $ith age. *o$e+er, a me!iation by chronicity of the !isor!er

$as not teste!. 5urthermore, poor global report of `o: co-occurre! $ith poor social support, poor family functioning, an! stressful life e+ents. #here is conflicting e+i!ence on $hether !ifferent mental !isor!ers ha+e a specific impact on `o:. 'chubert an! colleagues [C foun! no !ifferences bet$een !iagnostic groups of chil!ren suffering from either emotional, somatoform, or e-ternalising !isor!ers. Contrary to this, t$o further stu!ies reporte! lo$er le+els of emotional functioning, more pain, an! increase! !iscomfort in chil!ren $ith !epressi+e symptoms [3,7 . In contrast, school functioning an! social functioning $ere more likely affecte! in chil!ren $ith attention-!eficit an! !isrupti+e beha+iour !isor!ers [3,7 . A criticism of pre+ious `o: research is the failure to control for item o+erlap bet$een `o: an! mental health assessment [1 . #o our kno$le!ge, only t$o stu!ies controlle! for item o+erlap [7,0 , reporting similar relationships bet$een mental !isor!ers an! `o:. #he aim of the current stu!y $as to assess the !ifferential impact of internalising an! e-ternalising pathology on the chil!Zs `o: an! to follo$ up on 4astiaansenZs [> fin!ing that !iminishe! global `o: $as associate! $ith chil! factors beyon! pathology. #he abo+e-mentione! stu!y ha! se+eral limitations concerning the informant type &relying on a mi-ture of sources for the pre!ictor +ariables like parent, teacher, clinician, an! chil! report) an! the un!ifferentiate! consi!eration of `o: $ith a global `o: score. #he follo$ing hypotheses $ere teste! in the current stu!yI &a) both, internalising an! e-ternalising pathology are significantly relate! to the `o: !imensions< &b) family functioning, chronicity of pathology in!icate! by prior mental health treatment, le+el of impairment, age an!Hor gen!er ha+e an in!epen!ent relationship $ith `o:, beyon! the influence of e-ternalising an! internalising pathology, an! &c) the association bet$een `o: an! pathology remains after controlling for item o+erlap bet$een the constructs.

Methods
Procedure
(ata collection $as carrie! out as part of an effecti+eness trial for chil! an! a!olescent psychotherapy in Bermany. At the commencement of the outpatient therapy, patients &99 years an! ol!er) an! both parents &if a+ailable) $ere aske! to participate by the therapist. 4et$een 'eptember /22C an! June /292, 9/2 of the approache! /C/ families $ith a chil! bet$een > to 91 years agree! to participate an! ga+e their $ritten informe! consent &33.9A), returne! their %uestionnaires, an! hence compose the sample of the current stu!y. A further 97/ families refuse! to participate &77.0A). 5or these patients $e attaine! anonymous basic !ata on age, gen!er, an! !iagnostic status from the therapists. Comparison of the participants an! non-participants yiel!e! no significant !ifferences in terms of age, gen!er, an! comorbi!ity status, an! fre%uency of !isor!ers other than affecti+e !isor!ers. Affecti+e !isor!ers $ere significantly more pre+alent in the group of the participants &89A +s. 90.CA). #he stu!y $as appro+e! by the ethics committee of the "e!ical 4oar! *amburg.

Participants
.eporting on a total of 9/2 patients, C8 chil!renHa!olescents &99 years an! ol!er) an! 928 parents fille! out %uestionnaires. #he sample $as on a+erage 0age @ 9/.77 years ol! &rangeI >M91 years), >8.7A $ere female &n @ CC), 79A came from !i+orce! families. All patients ha! at least one !iagnose! mental !isor!er, >C.8A reporte! comorbi!ities. #he therapists !iagnose! 37.9A $ith an an-iety !isor!er, 89.2A $ith an affecti+e !isor!er, /7.CA $ith a P#'(, 97.0A $ith a !isrupti+e !isor!er, an! 88.>A $ith other !isor!ers &mainly eating !isor!ers, enuresis, encopresis, or sleeping !isor!ers).

Measures
#he assessment instruments presente! in the current stu!y $as part of a broa!er assessment battery compile! for a psychotherapy effecti+eness trial. *ealth-relate! %uality of life $as assesse! $ith the Berman KI('C.EE=-/C [/ . #his instrument consists of /C items an! $as !e+elope! to measure fi+e !imensionsI p*ysical %ell1'eing2 psyc*ological %ell1'eing2 autonomy and parent relation2 social support and peers, as $ell as sc*ool environment. Parallel parent an! chil! self-rating +ersions are a+ailable. Each item is score! on a 7-point :ikert scale ranging from 9 @ ^not at allHne+erZ to 7 ^e-tremelyHal$aysZ. #he items are summe! up for the specific subscales $ith high +alues representing high le+els of `o: an! may be con+erte! into #-+alues &# N 37 are consi!ere! abo+e a+erage). .eliability has been sho$n to be goo! &CronbachZs alphaI c @ .C1 to c @ .18). #o assess the o+erall pathology, sub,ects $ere a!ministere! the Chil! 4eha+ior Checklist for parents &C4C:< [92 ) or Vouth 'elf .eport for chil!ren an! a!olescents &V'.< [99 ), respecti+ely. #he C4C:HV'. consists of 991 items on specific emotional an! beha+ioural problems in chil!hoo! an! a!olescence. An internalising an! an e-ternalising symptom score can be calculate! from the correspon!ing syn!rome scales. Each item stan!s for a specific problem beha+iour an! is rate! on a 8-point scale from 2 @ ^not trueZ to / @ ^+ery true or often trueZ. #he reliability an! +ali!ity of these $i!ely use! instruments ha+e been e-amine! in a number of stu!ies [99,9/ . :e+el of functional impairment $as rate! by the therapists on the Impairment-'core for Chil!ren an! A!olescents &I'-CA< [98 ). #he I'-CA measures functional impairment on four !imensionsI mental, somatic, social-communicati+e, an! performance. Each item is rate! on a fi+e-point scale ranging from 2 @ ^not at allZ to 3 @ ^e-tremelyZ an! may be summe! up to a total score. *igher scale scores &ranging from 2 to /3) stan! for greater functional impairment &cut-off L C< [93 ). .etest-reliability $as high for the total score, rtt @ .13 [98 . 5amily functioning $as assesse! $ith the 5amilienboegen &54< [97 ). #he 54 is a selfreport scale comprising /1 items $hich family members rate on a :ikert-scale ranging from 2 @ ^completely trueZ to 8 @ ^not true at allZ. #he se+en subscales &communication2 role 'e*aviour2 task )ul)ilment2 emotionality2 control2 values and norms2 an! a))ective

esta'lis*ment o) relations) may be summe! into a total score an! transforme! into #+alues. *igher scale scores reflect greater !ysfunction &cut-off # L >2< [97 ). In the current stu!y, reliability for the total score $as satisfying $ith c @ .C0. As an in!icator of chronicity of pathology $e assesse! $hether the youth ha! prior mental health treatment $ith an a! hoc formulate! item asking if an! $hen the patient ha! pre+iously recei+e! treatment for a mental health con!ition. #he item $as !ichotomous ^yesHnoZ.

Analyses
(ata $ere processe! $ith 'P'' 91.2. *ierarchical linear regression analyses $ere calculate! $ith the `o: scales as !epen!ent +ariables an! chil! characteristics as in!epen!ent +ariables for the patient an! the parent report an! for each subscale separately. 5or the analyses of the parent report n @ 928 cases $ere inclu!e!, for the patient report n @ C8 cases coul! be inclu!e!. #he in!epen!ent +ariables $ere centre! on the in!i+i!ual mean. #he potential bias !ue to multicollinearity seeme! negligible &9.2C1 O +ariance inflation factor O 9.3C1). #est po$er $as sufficient &N 02A) &BPo$er< [9> ). #he in!epen!ent +ariables $ere entere! in t$o subse%uent blocks into the regression mo!el. 5irst, internalising an! e-ternalising pathology $ere entere!. In the secon! block, age, gen!er, functional impairment, percei+e! family functioning, an! prior treatment of a mental health con!ition $ere entere!. .esults an! effect siDes $ere e+aluate! base! on establishe! con+entions [9C . #o control for potential item o+erlap, $e e-clu!e! one item on the ability to concentrate in the scale sc*ool environment an! three items on !epresse! moo! in the scale psyc*ological %ell1'eing3 5or these t$o scales regression analyses $ere repeate! an! results compare! $ith results for the ra$ scores of the complete scales. .a$ scores instea! of #-+alues $ere chosen for comparability reasons.

Results
)ample description
'ee #able 9 for !escripti+e statistics of the scales. Compare! $ith norm !ata, patients reporte! re!uce! le+els of `o: regar!ing p*ysical an! psyc*ological %ell1'eing as $ell as satisfaction $ith social support and peers. #he parents reporte! re!uce! le+els of their chil!Zs `o: for p*ysical an! psyc*ological %ell1'eing as $ell as for the satisfaction $ith the sc*ool environment3 5urthermore, patients an! parents reporte! on a+erage clinical le+els of internalising pathology an! ele+ate! but not clinical le+els of e-ternalising pathology. 5unctional impairment as rate! by the therapists $as high an! in the clinical range. A!!itionally, family functioning $as reporte! to be in the normati+e range. About a thir! of the sample in!icate! ha+ing ha! prior mental health treatment.

#able 9 "eans, stan!ar! !e+iations or proportions of `o: measures an! pre!ictor +ariables

Patient report
#he first step of the hierarchical regression analyses &see #able /) tests $hether high le+els of internalising an! e-ternalising pathology $ere relate! to lo$ le+els of `o:. ?hen entere! simultaneously, mainly internalising pathology $as relate! to lo$ le+els of `o:. 5or the chil!renZs self-report, high le+els of internalising pathology $ere significantly relate! to lo$er psyc*ological %ell1'eing &] @ Y.>0), as $ell as lo$er $ellbeing in the realms of social support and peers &] @ Y.3C) an! sc*ool environment &] @ Y3>). *igh le+els of e-ternalising pathology $ere significantly relate! only to lo$er $ell-being $ith parent relations and autonomy &] @ Y.8/). P*ysical %ell1'eing $as neither significantly associate! $ith internalising nor $ith e-ternalising pathology. Psychopathology e-plaine! a significant amount of +ariance of each `o: scale &bet$een R/ @ .98 for autonomy and parent relations an! R/ @ .>9 for psyc*ological %ell1'eing)3

#able / "ultiple linear regression analyses of factors associate! $ith `o: M patient report 5e$ chil! factors e-plaine! some of the +ariance o+er an! abo+e the association bet$een chil! psychopathology an! `o: &see #able /). 6l!er age $as associate! $ith lo$er le+els of P*ysical %ell1'eing &] @ Y.8/). "ale gen!er &] @ Y./7) an! problematic family functioning &] @ Y.7>) $ere associate! $ith lo$ $ell-being $ith parent relations and autonomy3 #he subscales psyc*ological %ell1'eing, social support and peers2 an! sc*ool environment !i! not sho$ a significant increment after the inclusion of chil! factors. 6+erall, bet$een 9CA &for sc*ool environment) an! >2A +ariance &psyc*ological %ell1 'eing) coul! be e-plaine! by pathology an! other chil! characteristics.

Parent report

#he first step of the hierarchical regression analyses tests $hether high le+els of internalising an! e-ternalising pathology $ere relate! to lo$ le+els of `o: for the parent report &see #able 8). Internalising an! e-ternalising pathology $ere significantly relate! to psyc*ological %ell1'eing &internalising pathologyI ] @ Y.88) an! sc*ool environment &e-ternalising pathologyI ] @ Y.8/), $hen entere! simultaneously. #he e-plaine! +ariance in the regression mo!el $as only significant for psyc*ological %ell1'eing &R/ @ .9>) an! `o: $ith social support and peers &R/ @ .21)3

#able 8 "ultiple linear regression analyses of factors associate! $ith `o: M parent report 5e$ chil! factors e-plaine! some of the +ariance of the `o: scales for the parent report, o+er an! abo+e the association $ith chil! psychopathology &see secon! block of the regression mo!el in #able 8). 6l!er age &] @ Y.73) an! pre+ious mental health treatment &] @ Y./2) $ere associate! $ith lo$ le+els of p*ysical %ell1'eing3 Problematic family functioning &] @ Y.88) $as associate! $ith lo$ $ell-being $ith parent relations and autonomy3 A!!itionally, ol!er age &] @ Y.82) an! higher functional impairment &] @ Y./8) $ere associate! $ith lo$er $ell-being $ith the sc*ool environment. 5or `o: of social support and peers an! psyc*ological %ell1'eing2 the inclusion of chil! characteristics yiel!e! no significant increment beyon! the +ariance e-plaine! by pathology. 6+erall, bet$een 1A &for $ell-being $ith autonomy and parent relations4 not significant, p d .27>) an! 37A +ariance &for p*ysical %ell1'eing) coul! be e-plaine! by pathology an! other chil! characteristics as rate! by their parents.

Control for item o$erlap


#he e-clusion of o+erlapping items of the KI('C.EE= scales on psyc*ological %ell1 'eing an! sc*ool environment !i! not change the regression mo!els neither in significance nor magnitu!e compare! $ith the ra$ !ata scales or the #-+alue scales for both self- an! parent report &see #able 3 for regressions $ith ra$ scores an! item o+erlap controlle! scales of the self-report). 5or instance, pre!iction of self-reporte! psyc*ological %ell1'eing $as associate! significantly &all p O .229) $ith internalising pathology in all three teste! scale typesI ]#-+alues @ -.>0< ]ra$-+alues @ Y.C9< ]o+erlap-control @ Y.C/.

#able 3 "ultiple linear regression analyses of factors associate! $ith `o: for psychological $ell-being an! school en+ironment comparing ra$ scores an! item o+erlap control M patient report

Discussion
#he aim of the stu!y $as to assess the association of internalising an! e-ternalising pathology an! the chil!Zs `o: an! to !etermine $hether `o: $as associate! $ith chil! factors such as age, gen!er, functional impairment, poor family functioning or pre+ious mental health treatment as an in!icator of chronic psychiatric pathology. As postulate! in hypothesis &a), both, internalising an! e-ternalising pathology $ere associate! $ith the chil!Zs $ell-being an! functioning. *o$e+er, `o: sho$e! a stronger association $ith internalising than e-ternalising pathology for the self-report. =either internalising nor e-ternalising pathology $as associate! $ith e+ery !imension of `o:. Internalising pathology $as relate! to !iminishe! `o: in terms of psychological $ellbeing, social support an! peers as $ell as $ell-being $ith the school en+ironment $ith mo!erate to large effect siDes. #hese fin!ings are consistent $ith the Australian national sur+ey results [7 . *o$e+er, in contrast to our results, in a (utch sample of chil!ren $ith an-iety an! moo! !isor!ers `o: $as mainly affecte! in terms of emotional functioning. Physical, social, an! school functioning $ere not associate! $ith o+erall pathology [3 . :ike$ise, no association $as foun! bet$een physical $ell-being an! either internalising or e-ternalising pathology in our sample of both patient an! parent report. E-ternalising pathology $as associate! $ith lo$ le+els of percei+e! %uality of parent relations an! autonomy. 5rom the parentsZ perspecti+e, e-ternalising problem beha+iour $as associate! $ith lo$ functioning at school. #he latter fin!ing mirrors the results of the 4astiaansen stu!y [3 . 4ut the fin!ings contrast $ith the Australian stu!y [7 $here e-ternalising pathology $as not relate! to parent-rate! peer contacts or school acti+ities. #he current fin!ings contra!ict the results of 'chubert an! colleagues [C $ho stress that it is not so much the !iagnosis but the o+erall impairment $hich !iminishes `o: of chil!ren. In our stu!y, particularly internalising pathology seeme! to be relate! to impaire! le+els of $ell-being an! functioning. 'ince 'chubert an! colleagues relie! solely on parent reports, this fine-graine! picture of !ifferential impairment of `o: might not ha+e been +isible. Compare! $ith the self-report, the parent reporte! pathology

sho$e! less association $ith the !ifferent aspects of the `o:. Another e-planation for the stronger association of internalising pathology $ith `o: compare! $ith e-ternalising pathology might be an artefact !ue to the more pronounce! internalising symptomatology of the sample. *o$e+er, the !istribution of the e-ternalising scores co+ere! a similar range compare! $ith internalising pathology, sho$e! slightly less +ariance for the selfreport but not for the parent report. #hus, for the self-report the estimation of the association of `o: $ith e-ternalising problems might be conser+ati+ely biase!. *ypothesis &b) $as partly supporte! by the current !ata. :o$ family functioning an! male gen!er $ere relate! to lo$er le+els of the chil!Zs $ell-being an! functioning beyon! the association $ith pathology. Consistent $ith 4astiaansen et al.Zs fin!ings [> , family an! social net$ork factors $ere associate! $ith !iminishe! `o:. In contrast to 4astiaansen et al. [> $ho analyse! `o: $ith a global score, this stu!y in+estigate! !ifferent facets of `o: separately $ith the subscales of the KI('C.EE=-/C. 6ur !ata in!icate a !ifferential association of chil! an! family characteristics $ith these facets. 5or instance, $ell-being $ith the school en+ironment $as particularly associate! $ith internalising pathology. Problematic family functioning, e.g. !ealing $ith conflicts or family rules, $as associate! $ith the chil!Ks e-perience of autonomy an! relations to parents. #he #-scores of the KI('C.EE= $ere alrea!y a!,uste! for age an! gen!er, thus the influence of age an! gen!er on the `o: ratings beyon! the effect of e-ternalising pathology $as e-pectably lo$. =e+ertheless, male gen!er $as associate! $ith more negati+e parent relations an! less e-perience of autonomy as pre+iously reporte! by 4astiaansen et al. [> . #he effect of age on the association of pathology an! `o: reporte! by 4astiaansen et al. $as replicate! as follo$s. 4oth prior mental health treatment as a rough in!icator of chronicity of mental pathology an! age ha! an in!epen!ent association $ith physical $ell-being in the parent report. #his might in!icate that in more chronic con!itions of mental health issues the impro+ement of `o: coul! be $orth consi!ering in the therapeutic process in!epen!ent of targeting pathology. *o$e+er, because of the !ichotomous an! retrospecti+e nature of the item on prior treatment, the influence might be un!erestimate! an! shoul! be interprete! $ith caution. *ypothesis &c) $as supporte! by the current fin!ings. Item o+erlap control !i! not change the regression mo!els neither in significance nor magnitu!e compare! $ith the ra$ !ata scales or the #-+alue scales for both self- an! parent report. Item o+erlap !i! not change the po$er of the pre!ictor +ariables in this sample. As pre+iously sho$n by 'a$yer et al. [91 an! (ey an! colleagues [0 for attention !eficits, controlling for potential confoun!ing mental health problems ma!e little !ifference to the relationships bet$een mental illness an! `o:. 5uture stu!ies might pro+i!e similar control analyses to further !iffuse !oubts about the influence of item o+erlap. #he naturalistic sample of this stu!y has a number of a!+antages an! !isa!+antages attache!I $hile it is %uite representati+e of real life patients, the !istribution of internalising an! e-ternalising pathology turne! out to be %uite imbalance! an! comorbi! occurrence of internalising an! e-ternalising symptoms $as high. Although participation rate $as less than 72A, there !i! not seem to be a large selection bias in terms of age, gen!er, or pathology $ith affecti+e !isor!ers being the e-ception. Another limitation

$hich shoul! be emphasiDe! is the inclusion of relati+ely se+ere cases $hich may ha+e le! to an o+erestimation of the association of `o: $ith internalising an! e-ternalising problems. A!!itionally, the sample siDe $as relati+ely mo!est for the type an! number of analyses con!ucte!, hence caution is $arrante! in !ra$ing conclusions from the analyses. 5urthermore, the cross-sectional !esign pre+ents us from !ra$ing conclusions regar!ing the !irection of influence of factors on `o:. #o sum it up, psychiatric pathology is associate! $ith the impairment of the chil!Zs `o:. #hus, e-ten!ing the outcome measures in psychotherapy research from mere pathology to measures of `o: might enhance sensiti+ity to the patientZs $ell-being an! le+el of functioning. Especially internalising pathology $as associate! $ith impaire! `o: compare! $ith e-ternalising pathology. In +ie$ of the pre!ictors of `o: o+er an! abo+e pathology, one can conclu!e that $ith increasing age an! $ith the persistence of psychiatric problems the consi!eration of `o: for the therapeutic process might gain in importance.

Consent
?ritten informe! consent $as obtaine! from all participating patients &age! 99 years an! ol!er) as $ell as their parents.

Competing interests
#he authors !eclare that they ha+e no competing interests.

Authors# contribution
K? carrie! out the !ata analyses an! !rafting of the manuscript. '?B an! B. concei+e! of the stu!y, participate! in its !esign an! coor!ination, an! super+ise! the !ata analyses an! !rafting of the manuscript. JK( ma!e significant contributions to the !esign of the stu!y an! critically re+ise! the manuscript. All authors rea! an! appro+e! the final manuscript.

Acknowledgements
?e $oul! like to thank the participating therapists, as $ell as the young sub,ects an! their families for their contribution to this stu!y. A!!itionally, $e thank 'an!ra Pollmer for her +aluable contribution to the !ata ac%uisition an! coor!ination of the stu!y. #he Pereinigung analytischer Kin!er- un! Jugen!lichen-Psychotherapeuten &PaKJP) fun!e! the stu!y but ha! no influence on either stu!y !esign or a!ministration.

Article information

*ealth `ual :ife 6utcomes. /298< 99I 9/0. Publishe! online /298 July 89. !oiI 92.991>H93CC-C7/7-99-9/0 P"CI(I P"C8C88>82 Katharina ?eitkamp, 9 Ju!ith K (aniels,/ Beorg .omer,8 an! 'ilke ?iegan!-Brefe9 9 (epartment of Chil! an! A!olescent Psychiatry, Psychotherapy an! Psychosomatics, Eni+ersity "e!ical Centre *amburg-Eppen!orf, *amburg, Bermany / (epartment of Psychiatry, Eni+ersit_tsme!iDin Charit\, 4erlin, Bermany 8 (epartment of Chil! an! A!olescent Psychiatry, Psychosomatics an! Psychotherapy, Eni+ersity *ospital "enster, "enster, Bermany Correspon!ing author. Katharina ?eitkampI k.$eitkampHatHuke.!e< Ju!ith K (anielsI ,u!ith.!anielsHatHcharite.!e< Beorg .omerI g.romerHatHukmuenster.!e< 'ilke ?iegan!BrefeI s.$iegan!-grefeHatHuke.!e .ecei+e! "arch /2, /298< Accepte! July /C, /298. Copyright R/298 ?eitkamp et al.< licensee 4io"e! Central :t!. #his is an 6pen Access article !istribute! un!er the terms of the Creati+e Commons Attribution :icense &httpIHHcreati+ecommons.orgHlicensesHbyH/.2), $hich permits unrestricte! use, !istribution, an! repro!uction in any me!ium, pro+i!e! the original $ork is properly cite!. Articles from *ealth an! `uality of :ife 6utcomes are pro+i!e! here courtesy of BioMed Central

References
1. Voungstrom EA. 5uture !irections in psychological assessmentI combining

/. 8.

4.

7.

>.

e+i!ence-base! me!icine inno+ations $ith psychologyZs historical strengths to enhance utility. J Clin Chil! A!olesc Psychol. /298. pp. 980M970. [Pub"e! [Cross .ef #he KI('C.EE= Broup Europe. #he KI('C.EE= %uestionnaires - han!book. :engerichI Pabst 'cience Publications< /22>. JoDefiak #, :arsson 4, ?ichstrom :, ?allan!er J, "atte,at 5. `uality of life as reporte! by chil!ren an! parentsI a comparison bet$een stu!ents an! chil! psychiatric outpatients. *ealth `ual :ife 6utcomes. /292<99I98>. !oiI 92.991>H93CC-C7/7-1-98>. [P"C free article [Pub"e! [Cross .ef 4astiaansen (, Koot *", 5er!inan! .5, Perhulst 5C. `uality of life in chil!ren $ith psychiatric !isor!ersI self-, parent, an! clinician report. J Am Aca! Chil! A!olesc Psychiatry. /223<99I//9M/82. !oiI 92.920CH29.chi.222292/>13.82/2>.a!. [Pub"e! [Cross .ef 'a$yer "B, ?haites :, .ey J", *aDell P:, BraetD 4?, 4aghurst P. *ealthrelate! %uality of life of chil!ren an! a!olescents $ith mental !isor!ers. J Am Aca! Chil! A!olesc Psychiatry. /22/<99I782M78C. !oiI 92.920CH22223718/22/27222-22292. [Pub"e! [Cross .ef 4astiaansen (, Koot *", 5er!inan! .5. (eterminants of %uality of life in chil!ren $ith psychiatric !isor!ers. `ual :ife .es. /227<99I9700M9>9/. !oiI 92.922CHs9998>-223-CC99-/. [Pub"e! [Cross .ef

7. 'chubert "#, *erle ", ?urst E. IC(-92 (iagnostik un! :ebens%ualit_t. S

Kin!er Jugen!psychiatr Psychother. /228<99I/17M/09. !oiI 92.92/3H93//309C.89.3./17. [Pub"e! [Cross .ef 8. (ey ", :an!olt "A, "ohler-Kuo ". *ealth-relate! %uality of life among chil!ren $ith mental !isor!ersI a systematic re+ie$. `ual :ife .es. /29/<99I9C0CM9193. !oiI 92.922CHs9998>-29/-2920-C. [Pub"e! [Cross .ef 0. (ey ", "ohler-Kuo ", :an!olt ". *ealth-relate! %uality of life among chil!ren $ith mental health problems. A population-base! approach. /29/<99&C8) !oiI 92.991>H93CC-C7/7-92-C8. [P"C free article [Pub"e! [Cross .ef 92. Arbeitsgruppe (eutsche Chil! 4eha+ior Checklist. Elternfragebogen eber !as Perhalten +on Kin!ern un! Jugen!lichen - (eutsche 4earbeitung !er Chil! 4eha+ior Checklist &C4C:H3-91) /. KflnI Arbeitsgruppe Kin!er-, Jugen!- un! 5amilien!iagnostik &KJ5()< 9001. 99. (oepfner ", 4erner ?, :ehmkuhl B. *an!buchI 5ragebogen fer Jugen!liche. 5orschungsergebnisse Dur !eutschen 5assung !es Vouth 'elf .eport &V'.) !er Chil! 4eha+ior Checklist. Arbeitsgruppe Kin!er-, Jugen!- un! 5amilien!iagnostikI Koeln< 9003. 12. Ebesutani C, 4ernstein A, "artineD JI, Chorpita 45, ?eisD J.. #he youth self reportI applicability an! +ali!ity across younger an! ol!er youths. J Clin Chil! A!olesc Psychol. /299<99I881M83>. !oiI 92.9212H978C339>./299.73>239. [Pub"e! [Cross .ef 98. 5ahrig *, Kronmeller K#, *artmann ", .u!olf B. #herapieerfolg analytischer Psychotherapie bei Kin!ern un! Jugen!lichen [#herapy success of analytical psychotherapy of chil!ren an! a!olescence S Psychosom "e! Psychoanal. 900><99I8C7M807. 93. 'tefini A, Beiser-ElDe A, *artmann ", *orn *, ?inkelmann K, Kronmeller K#. 4in!ungsstil un! #herapieerfolg in !er psycho!ynamischen KurDDeittherapie bei Kin!ern un! Jugen!lichen [Attachment style an! therapy success in psycho!ynamic short-term psychotherapy for chil!ren an! a!olescents Psychotherapie un! psychologische "e!iDin. /221. pp. >1MC3. [Pub"e! [Cross .ef 97. Cierpka ", 5re+ert B. (ie 5amilienbfgen - *an!an$eisung. BfttingenI *ogrefe< 9003. 16. 5aul 5, Er!fel!er E, :ang A-B, 4uchner A. BQpo$er 8I a fle-ible statistical po$er analysis program for the social, beha+ioral, an! biome!ical sciences. 4eha+iour .esearch "etho!s. /22C<99I9C7M909. !oiI 92.8C71H452890893>. [Pub"e! [Cross .ef 9C. Cohen J. 'tatistical po$er analysis for the beha+ioral sciences. /. *ills!ale, =JI :a$rence Erlbaum Associates< 9011. 18. 'a$yer "B, Arney 5", 4aghurst PA, Clark JJ, BraetD 4?, Kosky .J, =urcombe 4, Patton BC, Prior "., .aphael 4, #he mental health of young people in AustraliaI chil! an! a!olescent component of the national sur+ey of mental health an! $ell-being. CanberraI Common$ealth of Australia< /222. [Pub"e!

Journal of the Cana!ian Aca!emy of Chil! an! A!olescent Psychiatry Cana!ian Aca!emy of Chil! an! A!olescent Psychiatry

Childhood and Adolescence: Challenges in Mental -ealth


'aurabh .am4iharilal 'hri+asta+a, "(, Prateek 'aurabh 'hri+asta+a, "(, an! Jega!eesh .amasamy, "( A!!itional article information

Abstract
"ental health is an integral an! essential component of health. #he ?orl! *ealth 6rganiDation &?*6) constitution statesI F*ealth is a state of complete physical, mental an! social $ell-being an! not merely the absence of !isease or infirmity.G "ore than 372 million people suffer from mental !isor!ers $orl!$i!e. In In!ia, mental health ser+ices, especially for chil!ren an! a!olescents, are limite! both in terms of number of )acilities as $ell as trained pro)essionals. #he ma,ority of mental health ser+ices are restricte! to urban areas, that is, me!ical colleges or regional mental health institutes. "ere presence of a treatment facility !oes not guarantee that all chil!renHa!olescents suffering from mental illness $ill utiliDe such ser+ices. In fact, most of the time there is a significant !elay from the patient si!e in accessing mental health ser+ices either because of lack of a$areness or associate! stigma. It is high time to promote positi+e mental health in chil!ren, a!olescents an! their parents through health e!ucation. Parental counseling is of utmost importance in or!er to a+oi! the !elay in treatment seeking. Keywords: mental health, challenges, In!ia

R8sum8
:a sant\ mentale est partie int\grante et essentielle !e la sant\. 'elon la Constitution !e lZ6rganisation mon!iale !e la sant\I g :a sant\ est un \tat !e complet bien-htre physi%ue, mental et social, et ne consiste pas seulement en une absence !e mala!ie ou !Zinfirmit\. i Plus !e 372 millions !e gens souffrent !e troubles mentau- !ans le mon!e entier. En In!e, les ser+ices !e sant\ mentale, sp\cialement ceu- pour les enfants et les a!olescents, sont limit\s tant par le nombre dinstallations %ue par celui !e pro)essionnels comp5tents. :a ma,orit\ !es ser+ices !e sant\ mentale se confine au- Dones urbaines, cZest-j-!ire au\coles !e m\!ecine ou au- \tablissements r\gionau- !e sant\ mentale. "hme j cela, la seule pr\sence !Zun \tablissement !e traitement ne garantit pas %ue tous les enfantsHa!olescents souffrant !Zune mala!ie mentale utiliseront ce ser+ice. En fait, la

plupart !u temps, il y a un !\lai significatif !e la part !es patients j acc\!er au- ser+ices !e sant\ mentale, soit par ignorance ou j cause !es stigmates associ\s. Il est plus %ue temps !e promou+oir la bonne sant\ mentale cheD les enfants, les a!olescents et leurs parents au moyen !e lZ\!ucation sur la sant\. :a consultation parentale est !Zune importance primor!iale afin !Z\+iter le retar! !Zaccks au traitement. Mots!cl8s : sant\ mentale, problkmes, In!e "ental health is an integral an! essential component of health. #he ?orl! *ealth 6rganiDation &?*6) constitution statesI F*ealth is a state of complete physical, mental an! social $ell-being an! not merely the absence of !isease or infirmity.G "ore than 372 million people suffer from mental !isor!ers $orl!$i!e &?*6, /292). In In!ia, mental health ser+ices, especially for chil!ren an! a!olescents, are limite! both in terms of number of )acilities as $ell as trained pro)essionals &?*6-AI"', /22>). ?*6 Atlas highlights the lo$ number of mental health professionals in In!ia. #he a+erage national !eficit of psychiatrists is estimate! to be CCA &?*6, /227). #he ma,ority of mental health ser+ices are restricte! to urban areas, that is, me!ical colleges or regional mental health institutes. "ere presence of a treatment facility !oes not guarantee that all chil!renHa!olescents suffering from mental illness $ill utiliDe such ser+ices. In fact, most of the time there is a significant !elay from the patient si!e in accessing mental health ser+ices either because of lack of a$areness or associate! stigma. #his !elay in seeking appropriate treatment may lea! to increase in se+erity of mental illness along $ith prolonge! treatment !uration, making the clinicianZs ,ob e+en more challenging to ha+e a fa+orable outcome. In!ia is home to a pluralistic approach to all types of healthcare. =ot only are there other systems of healthcare other than mo!ern me!icine such as Ayur+e!a, Enani, naturopathy an! homoeopathy, but also people approach religious places for help, especially in case of mental illness. #he main challenge in impro+ement of mental health ser+ices is its accessibility an! a+ailability for those $ho are in nee! of them. It is seen that access to mental health ser+ices for a chil! $ith a mental, emotional or beha+ioral !isor!er is substan!ar!, not pro+i!e! early enough, in sufficient supply. Consi!ering the scarcity of a+ailable facilities an! the longterm follo$-up &psychotherapy or monitoring the !oses an! si!e-effects of !rugs) it becomes !ifficult for chil!ren an! their families to stay a$ay from their home, in the hospital, until the chil!Ha!olescent reco+ers from the illness. E+en if !iagnose! an! treate! in time, the outcome may not be the same as e-pecte! by the clinician mainly because of the lack of compliance pertaining to treatment or follo$-up &Khan!el$al, Jhingan, .amesh, Bupta, ; 'ri+asta+a, /223). It is high time to promote positi+e mental health in chil!ren, a!olescents an! their parents through health e!ucation. Parental counseling is of utmost importance in or!er to a+oi! the !elay in treatment seeking &Kumar, /299). En!er the national mental health program in In!ia, me!ical officers in primary health centres are traine! to !iagnose an! treat chil!renHa!olescents at the earliest stages of their mental illness. #his training of me!ical officers $ill help in e-pansion of the no$ limite! ser+ices &Kumar, /299). 4ut the fact remains that only a minor fraction of them ha+e been traine! until no$ an! e+en though

they are traine!, only a fe$ of the psychiatric me!ications are a+ailable at their e-pense. Inconsistent use of stan!ar!iDe! an! +ali!ate! screening an! assessment tools by the clinicians remains another important challenge, $hich nee!s to be a!!resse!, in the management of mental health !isor!ers &'an! et al., /227). #he importance of psychological $ell-being in chil!ren an! a!olescents, for healthy emotional, social, physical, cogniti+e an! e!ucational !e+elopment, is $ell recogniDe!. *o$e+er, a large gap e-ists in the area of pre+ention an! mental health promotion that mainly re%uires integration of mental health into pe!iatric primary care an! strengthening of the e-isting human resources an! infrastructure.

Acknowledgements 0 Conflicts of "nterest


#he authors ha+e no financial relationships to !isclose.

Article information
J Can Aca! Chil! A!olesc Psychiatry. /298 "ay< //&/)I 13M17. P"CI(I P"C8>3C>// 'aurabh .am4iharilal 'hri+asta+a, "(,9 Prateek 'aurabh 'hri+asta+a, "(,9 an! Jega!eesh .amasamy, "(9 9 (epartment of Community "e!icine, 'hri 'athya 'ai "e!ical College ; .esearch Institute, Kancheepuram, #amil =a!u, In!ia Correspon!ing E-"ailI !rshrishri/221HatHgmail.com .ecei+e! =o+ember />, /29/< Accepte! January 97, /298. Copyright R /298 Cana!ian Aca!emy of Chil! an! A!olescent Psychiatry Articles from Journal of the Cana!ian Aca!emy of Chil! an! A!olescent Psychiatry are pro+i!e! here courtesy of Canadian Academy of Child and Adolescent Psychiatry

References

Khan!el$al 'K, Jhingan *P, .amesh ', Bupta .K, 'ri+asta+a PK. In!ia mental health country profile. International .e+ie$ of Psychiatry. /223<9>&9M/)I9/>M 939. [Pub"e! Kumar A. "ental health ser+ices in rural In!iaI Challenges an! prospects. *ealth. /299<8&9/)IC7CMC>9. 'an! =, 'il+erstein ", Blascoe 5P, Bupta P4, #onniges #P, 6ZConnor KB. Pe!iatriciansZ reporte! practices regar!ing !e+elopmental screeningI (o gui!elines $orkT (o they helpT Pe!iatrics. /227<99>&9)I9C3M9C0. [Pub"e! ?orl! *ealth 6rganiDation "ental healthI 'trengthening our response. /292. A+ailable fromI httpIHH$$$.$ho.intHme!iacentreHfactsheetsHfs//2HenHin!e-.html &Accesse! on /1 'eptember /29/). ?orl! *ealth 6rganiDation "ental health atlas. Bene+aI ?orl! *ealth 6rganiDation< /227.

?*6-AI"' .eport on "ental *ealth 'ystem in Ettarkhan!, In!ia ?*6 an! "inistry of *ealth< (ehra!un, Ettarkhan!, In!iaI /22>.

Journal of the Cana!ian Aca!emy of Chil! an! A!olescent Psychiatry Cana!ian Aca!emy of Chil! an! A!olescent Psychiatry

Multiple De$elopmental Pathways to Conduct Disorder: Current Conceptuali9ations and Clinical "mplications
(ustin Par!ini, Ph( an! Paul J. 5rick, Ph( A!!itional article information

Abstract
'b(ecti$es
.ecent research has unco+ere! se+eral !e+elopmental path$ays through $hich chil!ren an! a!olescents can !e+elop a ten!ency to !isplay the se+ere antisocial beha+ior associate! $ith the !iagnosis of con!uct !isor!er &C().

Methods
#his focuse! re+ie$ is !esigne! to briefly outline three !ifferent etiological path$ays !escribe! in the literature. #hese path$ays are !istinguishe! by the age of onset of the antisocial beha+ior, the presenceHabsence of significant le+els of callous-unemotional traits, an! the presenceHabsence of problems $ith anger regulation.

Results
E+i!ence from !e+elopmental psychopathology research &particularly longitu!inal stu!ies) that support the !ifferent life-course tra,ectories an! putati+e etiological factors associate! $ith antisocial beha+ior across these path$ays is presente!.

Conclusions

:imitations in the a+ailable research on these !e+elopmental path$ays an! implications of this research for the pre+ention an! treatment of chil!ren an! a!olescents $ith C( are !iscusse!. Keywords: con!uct !isor!er, !e+elopmental path$ays, chil!ren, a!olescents

R8sum8
'b(ectifs
:a recherche r\cente a !\cou+ert plusieurs tra,ectoires !\+eloppementales par les%uelles les enfants et les a!olescents peu+ent !\+elopper une ten!ance au comportement antisocial s\+kre associ\ au !iagnostic !u trouble !es con!uites &#C).

M8thodes
Cette \tu!e cibl\e est !estin\e j pr\senter brik+ement trois !iff\rents mo!kles !e tra,ectoires \tiologi%ues !\crits !ans la litt\rature. Ces tra,ectoires se !istinguent par lZlge !Zapparition !u comportement antisocial, la pr\sence ou lZabsence !e !egr\s significatifs !e traits !e !uret\-insensibilit\, et la pr\sence ou lZabsence !e problkmes !e r\gulation !e la colkre.

R8sultats
:es !onn\es probantes !e la recherche sur la psychopathologie !\+eloppementale &en particulier les \tu!es longitu!inales) %ui soutiennent les !iff\rentes tra,ectoires !e parcours !e +ie et les facteurs \tiologi%ues pr\sum\s associ\s au comportement antisocial !ans ces tra,ectoires sont pr\sent\es.

Conclusions
:es limitations !e la recherche !isponible sur ces tra,ectoires !\+eloppementales et les implications !e cette recherche pour la pr\+ention et le traitement !es enfants et a!olescents souffrant !u #C sont !iscut\es. Mots!cl8s : trouble !es con!uites, tra,ectoires !\+eloppementales, enfants, a!olescents Con!uct !isor!er &C() is !efine! as a repetiti+e an! persistent pattern of beha+ior $hich +iolates the rights of others or ma,or age-appropriate societal rules &American Psychiatric Association, /222). 6+er the last se+eral !eca!es, it has become apparent that there are multiple causal factors that un!erlie the beha+ioral manifestations of C( in chil!ren an! a!olescents. ?hile causal heterogeneity is common to all psychiatric !isor!ers, the myria! of !ifferent etiological factors linke! to C( is striking &e.g., genetic, neurocogniti+e, temperamental, peer, family). *o$e+er, recent !e+elopmental psychopathology research has pro+i!e! e+i!ence !ocumenting uni%ue path$ays

associate! $ith the emergence an! continuity of C( o+er time. ?hile a number of !e+elopmental mo!els of C( ha+e been propose! &e.g., (o!ge ; Pettit, /228), the current article is !esigne! to briefly o+er+ie$ three !iffering etiological path$ays to C(. #hese mo!els are base! uponI 9. the !e+elopmental timing of C( symptom emergence< /. the presence of callous-unemotional &CE) traits< an!, 8. the presence of se+ere anger !ysregulation. Intereste! rea!ers can fin! more in-!epth an! comprehensi+e re+ie$s of these mo!els else$here &5rick ; Pi!ing, /220< "offitt, /22>).

,a&onomy of CD based on timing of onset


6ne of the most en!uring subtyping schemes of C( is base! on longitu!inal research in!icating that youth $ith con!uct problems initiate! in chil!hoo! &i.e., chil!hoo!-onset) are at heightene! risk for e-hibiting persistent criminal beha+ior into a!ulthoo!. Vouth $ho !e+elop chil!hoo!-onset C( often ha+e longstan!ing problems relate! to attention!eficit hyperacti+ity !isor!er &A(*() an! oppositional !efiant !isor!er &6(() that emerge prior to their first C( symptom &"offitt, /22>). E+i!ence suggests that the transition to early C( is cause! in part by subtle neurological !eficits &e.g., !eficit inhibitory control, poor +erbal abilities) that lea! to !ifficulties managing peer conflicts, regulating emotions, an! controlling impulses &"offitt, /22>). In a!!ition to these cogniti+e problems, youth $ith chil!hoo!-onset C( often come from families $ith a longstan!ing history of antisocial beha+ior that use harshHinconsistent !iscipline practices &6!gers et al., /221), $hich makes it !ifficult for these chil!ren to ac%uire appropriate social skills an! internaliDe rules for appropriate con!uct. "oreo+er, chil!hoo!-onset C( youth ten! to e-perience escalating aca!emic an! peer !ifficulties o+er time &"offitt, /22>< 6!gers et al., /221). #his casca!e of accumulating risk factors impe!es these youth from making important life transitions &e.g., gra!uating) ser+ing to further entrench them into a criminal lifestyle &"offitt, /22>< 6!gers et al., /221). In contrast to the chil!hoo!-onset path$ay, there is a larger group of youth $ho !o not begin e-hibiting !elin%uent beha+iors until a!olescence &i.e., a!olescent-onset). #hese youth often !o not e-hibit A(*( or 6(( in chil!hoo!, an! are less likely to ha+e early neurological impairments an! se+ere family !ysfunction &"offitt, /22>< 6!gers et al., /221). Instea!, e+i!ence suggests that a!olescent-onset C( emerges in part $hen rebellious a!olescents are poorly monitore! by their parents an! begin affiliating $ith !elin%uent peers &"offitt, /22>). A!olescent-onset C( youth are posite! to be more likely to lea+e their antisocial $ays behin! !uring the transition into a!ulthoo!, as they a!opt prosocial roles &e.g., employment), spen! less time $ith !e+iant peers, an! engage in more mature !ecision-making. *o$e+er, recent e+i!ence suggests that many a!olescent-onset youth continue to engage in criminal beha+ior an! e-perience impairments in se+eral life !omains $ell into early a!ulthoo! &6!gers et al., /221).

4ecause early-onset C( symptoms ha+e been consistently associate! $ith a persistent form of antisocial beha+ior, a chil!hoo!-onset subtyping scheme $as a!!e! to the !iagnosis of C( in ('"-IP, an! is propose! to be retaine! in the upcoming ('"-7 &5rick ; =igg, /29/). *o$e+er, there are limitations associate! $ith relying solely on this !ual !e+elopmental ta-onomy. 'pecifically, it has become clear that a significant portion of chil!hoo!-onset youth !esist from crime by early a!ulthoo! &6!gers et al., /221). #here also remains consi!erable etiological heterogeneity $ithin chil!hoo!-onset C( cases &5rick ; Pi!ing, /220). #herefore, researchers ha+e sought to further refine this subtyping scheme to i!entify more homogenous groups of youth in terms of causal mechanisms an! !e+elopmental outcomes.

CD with callous!unemotional traits


6ne promising metho! for further !istinguishing an etiologically uni%ue group of chil!ren $ith chil!hoo!-onset C( in+ol+es i!entifying callous-unemotional &CE) features. Consistent $ith the affecti+e !imension of a!ult psychopathy, CE traits inclu!e a lack of concern for othersZ feelings, !eficient guilt an! remorse, an! shallo$ affect. #he estimate! pre+alence of high CE traits in youth $ith C( ranges from 92M3>A in community samples to /9M70A in clinic-samples &Kahn, 5rick, Voungstrom, 5in!ling, ; Voungstrom, /29/< Kolko ; Par!ini, /292< .o$e et al., /292). Accumulating e+i!ence in!icates that youth $ith ele+ate! CE traits are at risk for e-hibiting se+ere an! persistent antisocial beha+ior, e+en after controlling for co-occurring !isrupti+e beha+ior !isor!er symptoms &5rick, Cornell, 4arry, 4o!in, ; (ane, /228< "c"ahon, ?itkie$itD, Kotler, ; Con!uct Problems Pre+ention .esearch Broup, /292< Par!ini ; 5ite, /292). As a result, these traits seem to further !elineate chil!hoo!-onset C( cases that are more likely to persist in their antisocial beha+ior into a!ulthoo!. #here also appears to be uni%ue causal factors un!erlying the con!uct problems foun! in chil!ren $ith CE traits, such as lo$ temperamental fear. :ongitu!inal stu!ies ha+e consistently linke! lo$ fearful arousal to the !e+elopment of se+ere antisocial beha+ior, particularly +iolence &:oeber ; Par!ini, /221). "oreo+er, infants an! chil!ren $ith a relati+ely fearless temperament e-hibit impairments in the !e+elopment of empathy an! guilt &5o$les ; Kochanska, /222), in part because they seem to e-perience relati+ely little emotional arousal in response to !istress cues in others or to cues of punishment for misbeha+ior &5rick ; Pi!ing, /220). #ogether, these fin!ings suggest that lo$ temperamental fear may lea! to the !e+elopment of early con!uct problems because it re!uces the effecti+eness of punishment-oriente! socialiDation techni%ues an! fosters the !e+elopment of CE traits &Par!ini, /22>). Consistent $ith this theoretical mo!el, chil!ren $ith CE traits sho$ impairments $hen processing cues of fearful !istress in others &"arsh ; 4lair, /221), possibly !ue to !eficits in atten!ing to emotionally salient facial features &(a!!s et al., /22>). #hey also ten! to make less eye contact $ith caretakers $hen in+ol+e! in emotion !iscussions &(a!!s et al., /29/), $hich may interfere $ith early moral socialiDation. ?hile much of the research aime! at un!erstan!ing the !e+elopment of CE traits has focuse! on chil! characteristics, caregi+er affectionH$armth may protect chil!ren from

!e+eloping CE traits o+er time. "aternal emotional responsi+eness !uring infancy has been associate! $ith higher le+els of empathy &Kiang, "oreno, ; .obinson, /223) an! guilt &Kochanska, 5orman, Aksan, ; (unbar, /227) in chil!hoo!. A $arm an! in+ol+e! parent-chil! relationship has also been sho$n to protect aggressi+e chil!ren $ith lo$ fear from e-periencing increases in CE traits o+er time &Par!ini, :ochman, ; Po$ell, /22C) an! seems to buffer chil!ren $ith high CE traits from !e+eloping more serious con!uct problems &Kroneman, *ip$ell, :oeber, Koot, ; Par!ini, /299< Pasalich, (a!!s, *a$es, ; 4rennan, /299). #aken together, these stu!ies suggest that chil!ren $ho follo$ a CE path$ay to earlyonset C( e-hibit lo$ temperamental fear an! !eficits in atten!ing to salient emotional social cues, $hich can interfere $ith +arious socialiDation processes !esigne! to facilitate the !e+elopment of moral emotions. 'pecifically, chil!ren $ith these characteristics ten! to e-perience little a+ersi+e arousal $hen being punishe!, o+erlook cues of suffering in others, an! are !ifficult to engage in emotional !iscussions. *o$e+er, e-posure to a $arm an! nurturing caregi+er may protect chil!ren from !e+eloping CE traits o+er time, e+en if they ha+e a relati+ely fearless temperament.

CD with se$ere anger dysregulation


(e+elopmental stu!ies ha+e also begun to support another causal path$ay to early-onset C( that in+ol+es problems $ith se+ere anger !ysregulation. 'pecifically, chil!ren $ith con!uct problems in the absence of CE traits ten! to e-hibit high temperamental negati+e emotionality an! ele+ate! le+els of internaliDing problems &*ip$ell et al., /22C< Par!ini, :ochman, ; 5rick, /228). "oreo+er, e+i!ence suggests that !ifficulties regulating anger are particularly important for un!erstan!ing the !e+elopment of early con!uct problems. 5or e-ample, high temperamental anger in infants an! chil!ren has been associate! $ith the !e+elopment of later aggression an! con!uct problems &Arsenio, Cooperman, ; :o+er, /222< :engua ; Ko+acs, /227< .othbart, Aha!i, ; *ershey, 9003), an! !ysregulate! anger represents a core feature of 6((, $hich is a !e+elopmental precursor to early-onset C( &'tringaris, /299). 'ocial-cogniti+e research suggests that chil!ren $ith high le+els of anger ten! to o+er-interpret ambiguous social cues as threatening &'chultD, IDar!, ; 4ear, /223), $hich may lea! them to engage in !efensi+e forms of aggression in response to minor pro+ocation &6robio !e Castro, Peerman, Koops, 4osch, ; "onshou$er, /22/). Importantly, hostile attribution biases an! ele+ate! reacti+e aggression often occur in con!uct problem chil!ren $ithout CE traits &5rick et al., /228). ?hile problems $ith !ysregulate! anger may be partially !ri+en by neurobiological factors, e-posure to harsh an! abusi+e parental !iscipline can also play a role. E-posure to harsh !isciple has been consistently linke! to the !e+elopment of antisocial beha+ior &Bershoff, /22/), particularly among chil!ren $ith lo$ CE traits &Pasalich et al., /299). A!!itionally, chil!ren $ho are e-pose! to high le+els of harsh !iscipline ten! to ha+e !ifficulties !e+eloping appropriate emotion regulation skills &'hiel!s ; Cicchetti, 9001), an! e-hibit an increase! hyper+igilance to cues of potential threat in others &(o!ge, 4ates, Pettit, ; Palente, 9007< Pollak ; 'inha, /22/). ?hile e-posure to harsh !iscipline may facilitate the !e+elopment of early con!uct problems by interfering $ith the

!e+elopment of anger regulation abilities, this association seems to be mo!erate! by genetic factors. 'pecifically, recent stu!ies in!icate that early maltreatment is linke! to the !e+elopment of antisocial beha+ior pre!ominately in chil!ren possessing a gene associate! $ith lo$ le+els of monoamine o-i!ase A &0A#A) enDyme acti+ity &#aylor ; Kim-Cohen, /22C). In sum, chil!ren $ith anger regulation problems often e-hibit early oppositionalH!efiant beha+iors, $hich ten! to prece!e the !e+elopment of C( in chil!hoo!. Vouth $ith high le+els of anger also ten! to ha+e a hostile attribution bias $hen enco!ing cues of potential threat, $hich can perpetuate interpersonal conflicts $ith others. 5inally, $hile early e-posure to abusi+e !iscipline practices may be a particularly important etiological factor in the !e+elopment of anger regulation !ifficulties, this may only occur in youth possessing certain genetic risk factors.

Clinical implications and future directions


Bi+en the research outline! abo+e, one $ay to conceptualiDe !e+elopmental path$ays to C( is to first !ifferentiate bet$een chil!hoo!- an! a!olescent-onset C(, an! then !istinguish those chil!hoo!-onset youth $ith high CE traits an! those $ith se+ere anger !ysregulation. #hese path$ays appear to be %uite promising for gui!ing future research into the etiology of antisocial beha+ior. *o$e+er, there remain many aspects of these mo!els that ha+e not been sufficiently teste!. "oreo+er, there are se+eral unans$ere! %uestions about ho$ to best use the mo!els to gui!e clinical practice. A fe$ key areas that are ripe for future stu!y inclu!eI 9. refining clinical !iagnosis< an!, /. !e+eloping inno+ati+e pre+ention an! treatment inter+entions.

Refining diagnostic criteria


6ne timely issue is $hether to incorporate features of CE traits an! se+ere anger !ysregulation into the upcoming ('"-7 classification system. #here is currently a proposal to a!! a specifier to the !iagnosis of C( &i.e., $ith limite! prosocial emotions) to !esignate those $ith high le+els of CE traits &Par!ini, 5rick, ; "offitt, /292). Initial tests of this specifier ha+e been promising. Chil!ren $ith C( $ho meet criteria for the specifier ten! to ha+e higher le+els of aggressi+e an! cruel beha+iors compare! to youth $ith C( alone &Kahn et al., /29/). In terms of pre!icti+e utility, young girls $ho meet criteria for the specifier ha+e been sho$n to e-hibit more bullying an! C( symptoms at a si--year follo$-up than girls $ith chil!hoo!-onset C( alone &Par!ini, 'tepp, *ip$ell, 'touthamer-:oeber, ; :oeber, /29/). A!olescents $ho meet !iagnostic threshol! for the specifier also appear to be at high risk for e-hibiting antisocial an! criminal beha+ior into a!ulthoo! &"c"ahon et al., /292). *o$e+er, issues regar!ing the optimal metho!s for assessing CE traits at !ifferent ages an! !ealing $ith !iscor!ant information across multiple informants still nee! to be a!!resse!.

#here is currently no propose! metho! for !elineating a subgroup of chil!ren $ith C( $ho ha+e se+ere anger !ysregulation for ('"-7. *o$e+er, allo$ing for co-morbi! !iagnoses of 6(( an! C(, an! labeling three symptoms of 6(( as an angryHirritable !imension may ai! in this respect &5rick ; =igg, /29/). A more contro+ersial proposal is to a!! a ne$ !iagnosis of !isrupti+e moo! !ysregulation !isor!er &("(() to ('"-7 !efine! by persistent irritability an! impulsi+e outbursts of aggression that coul! be !iagnose! in con,unction $ith C( &'tringaris, /299). *o$e+er, the propose! !iagnosis $as not e-plicitly !esigne! to help !elineate a more etiologically homogenous group of youth $ith chil!hoo!-onset C(.

,argeted pre$ention and inter$ention strategies


?hile a number of inter+entions ha+e pro+en effecti+e in treating early emerging con!uct problems, the effecti+eness ten!s to !ecrease in ol!er chil!ren an! a!olescence. #hus, inter+ening early in the !e+elopmental tra,ectory of chil!hoo!-onset C( represents an important a+enue for pre+enting later serious aggression an! antisocial beha+ior. Implementing pre+entati+e-inter+entions $ith chil!ren e-hibiting significant oppositional !efiant beha+iors, !ysregulate! anger or early CE traits !uring the pre-school years &prior to the onset of serious C( symptoms) seems particularly important. Bi+en the large number of risk factors across multiple !omains that ha+e been associate! $ith the !e+elopment of early-onset C(, effecti+e pre+entati+e-inter+entions shoul! be capable of pro+i!ing a comprehensi+e array of ser+ices to families that target multiple risk factors. *o$e+er, it is also important to in!i+i!ualiDe these programs to effecti+ely target the specific !e+elopmental mechanisms un!erlying each chil!Zs antisocial beha+ior. .esearch on the +arious !e+elopmental path$ays to antisocial beha+ior coul! be important for gui!ing in!i+i!ualiDe! treatment approaches for chil!ren $ith C(. 5or e-ample, inter+entions that focus on anger control an! re!ucing harsh an! inconsistent !iscipline may be more effecti+e for C( chil!ren $ith se+ere anger !ysregulation &:ochman ; ?ells, /223). In contrast, treatments focuse! on promoting positi+e parentchil! emotional connecte!ness may be more beneficial for con!uct problem chil!ren $ith high CE traits &#homas ; Simmer-Bembeck, /22C). Bi+en that chil!ren $ith CE traits sho$ relati+ely lo$ le+els of concern about being punishe!, teaching parents ho$ to use positi+e reinforcement to encourage prosocial beha+ior may be particularly beneficial. #o !ate, there is little systematic research testing the utility of matching youth $ith C( to !ifferent types of treatment. An increase! focus on !e+eloping treatments that specifically target the characteristics of chil!ren $ith C( an! CE traits $ill be particularly important mo+ing for$ar!, gi+en the se+ere an! persistent nature of their antisocial beha+ior. 'ome stu!ies ha+e foun! that youth $ith high CE traits e-hibit more !isrupti+e beha+iors both !uring an! after treatment relati+e to youth $ithout these traits &*aas et al., /299< *a$es ; (a!!s, /227). *o$e+er, chil!ren $ith CE traits are by no means Funtreatable,G particularly $hen e-pose! to intensi+e, empirically-base! inter+entions &Kolko ; Par!ini, /292). Importantly, there are no$ se+eral stu!ies in!icating that treatments for young chil!ren $ith con!uct problems can lea! to re!uctions in the CE traits o+er time &*a$es ; (a!!s,

/22C< Kolko et al., /220< "c(onal!, (o!son, .osenfiel!, ; Jouriles, /299< 'omech ; EliDur, /29/). #hese stu!ies promote optimism that it is possible to effecti+ely treat the se+ere con!uct problems foun! in chil!ren $ith CE traits, as $ell as re!uce o+erall le+els of CE features in con!uct problem youth. ?hile there are no$ se+eral manualiDe! inter+entions that ha+e been foun! to pro!uce beha+ioral impro+ements in chil!ren $ith con!uct problems, many chil!ren continue to e-hibit significant beha+ioral impairments at the en! of treatment an! positi+e beha+ioral gains ten! to ero!e o+er time. If inter+entions can be better tailore! to the uni%ue characteristics of chil!ren base! on the !e+elopmental mechanisms un!erlying their con!uct problems, more pronounce! an! sustaine! treatment effects $ill likely be achie+e!. Continue! !e+elopmental research aime! at unco+ering the uni%ue etiological factors un!erlying the beha+ior problems of subgroups of youth $ith C( $ill help to facilitate future inno+ations in these comprehensi+e an! in!i+i!ualiDe! approaches to pre+ention an! treatment.

Acknowledgements 0 Conflicts of "nterest


#he authors ha+e no financial relationships to !isclose.

Article information
J Can Aca! Chil! A!olesc Psychiatry. /298 5ebruary< //&9)I /2M/7. P"CI(I P"C87>7C99 (ustin Par!ini, Ph(9 an! Paul J. 5rick, Ph(/ 9 Eni+ersity of Pittsburgh "e!ical Center, Pittsburgh, Pennsyl+ania, Enite! 'tates / Eni+ersity of =e$ 6rleans, =e$ 6rleans, :ouisiana, Enite! 'tates Correspon!ing e-mailI !ap81HatHpitt.e!u .ecei+e! July /3, /29/< Accepte! 6ctober 8, /29/. Copyright R /298 Cana!ian Aca!emy of Chil! an! A!olescent Psychiatry #his article has been cite! by other articles in P"C. Articles from Journal of the Cana!ian Aca!emy of Chil! an! A!olescent Psychiatry are pro+i!e! here courtesy of Canadian Academy of Child and Adolescent Psychiatry

References

American Psychiatric Association (iagnostic an! statistical manual of mental !isor!ers. 3th e! te-t re+ision. ?ashington, (CI American Psychiatric Association< /222. Arsenio ?5, Cooperman ', :o+er A. Affecti+e pre!ictors of preschoolersZ aggression an! peer acceptanceI (irect an! in!irect effects. (e+elopmental Psychology. /222<8>I381M331. [Pub"e! (a!!s "., Allen J:, 6li+er 4., 5aulkner =, :egge K, "oul C, 'cott '. :o+e, eye contact an! the !e+elopmental origins of empathy +. psychopathy. 4ritish Journal of Psychiatry. /29/</22I909M90>. [Pub"e!

(a!!s "., Perry V, *a$es (J, "erD ', .i!!ell AC, *aines (J, Abeyguna$ar!ane AI. Attention to the eyes an! fear-recognition !eficits in chil! psychopathy. 4ritish Journal of Psychiatry. /22><910I/12M/19. [Pub"e! (o!ge KA, 4ates JE, Pettit B', Palente E. 'ocial information-processing patterns partially me!iate the effect of early physical abuse on later con!uct problems. Journal of Abnormal Psychology. 9007<923I>8/M>38. [Pub"e! (o!ge KA, Pettit B'. A biopsychosocial mo!el of the !e+elopment of chronic con!uct problems in a!olescence. (e+elopmental Psychology. /228<80I830M8C9. [P"C free article [Pub"e! 5o$les (C, Kochanska B. #emperament as a mo!erator of path$ays to conscience in chil!renI #he contribution of electro!ermal acti+ity. Psychophysiology. /222<8CIC11MC07. [Pub"e! 5rick PJ, Cornell A*, 4arry C#, 4o!in '(, (ane *E. Callous-unemotional traits an! con!uct problems in the pre!iction of con!uct problem se+erity, aggression, an! self-report of !elin%uency. Journal of Abnormal Chil! Psychology. /228<89I37CM3C2. [Pub"e! 5rick PJ, =igg J#. Current issues in the !iagnosis of attention !eficit hyperacti+ity !isor!er, oppositional !efiant !isor!er, an! con!uct !isor!er. Annual .e+ie$ of Clinical Psychology. /29/<1ICCM92C. [Pub"e! 5rick PJ, Pi!ing E. Antisocial beha+ior from a !e+elopmental psychopathology perspecti+e. (e+elopment an! Psychopathology. /220</9I9999M9989. [Pub"e! Bershoff E#. Corporal punishment by parents an! associate! chil! beha+iors an! e-periencesI A meta-analytic an! theoretical re+ie$. Psychological 4ulletin. /22/<9/1I780M7C0. [Pub"e! *aas '", ?aschbusch (A, Pelham ?E, King ', An!ra!e 45, Carrey =J. #reatment response in CPHA(*( chil!ren $ith callousHunemotional traits. Journal of Abnormal Chil! Psychology. /299<80I739M77/. [Pub"e! *a$es (J, (a!!s ".. #he treatment of con!uct problems in chil!ren $ith callous-unemotional traits. Journal of Consulting an! Clinical Psychology. /227<C8IC8CMC39. [Pub"e! *a$es (J, (a!!s ".. 'tability an! malleability of callous-unemotional traits !uring treatment for chil!hoo! con!uct problems. Journal of Clinical Chil! an! A!olescent Psychology. /22C<8>I83CM877. [Pub"e! *ip$ell AE, Par!ini (A, :oeber ., 'embo$er ", Keenan K, 'touthamer-:oeber ". Callous-unemotional beha+iors in young girlsI 'hare! an! uni%ue effects relati+e to con!uct problems. Journal of Clinical Chil! an! A!olescent Psychology. /22C<8>I/08M823. [Pub"e! Kahn .E, 5rick PJ, Voungstrom E, 5in!ling .:, Voungstrom JK. #he effects of inclu!ing a callous-unemotional specifier for the !iagnosis of con!uct !isor!er. Journal of Chil! Psychology an! Psychiatry. /29/<78I/C9M/1/. [P"C free article [Pub"e! Kiang :, "oreno AJ, .obinson J:. "aternal preconceptions about parenting pre!ict chil! temperament, maternal sensiti+ity, an! chil!renZs empathy. (e+elopmental Psychology. /223<32I9219M920/. [Pub"e! Kochanska B, 5orman (., Aksan =, (unbar '4. Path$ays to conscienceI Early mother-chil! mutually responsi+e orientation an! chil!renZs moral emotion,

con!uct, an! cognition. Journal of Chil! Psychology an! Psychiatry. /227<3>I90M 83. [Pub"e! Kolko (J, (orn :(, 4ukstein 6B, Par!ini (, *ol!en EA, *art J. Community +s. clinic-base! mo!ular treatment of chil!ren $ith early-onset 6(( or C(I A clinical trial $ith 8-year follo$-up. Journal of Abnormal Chil! Psychology. /220<8CI709M>20. [Pub"e! Kolko (J, Par!ini (A. 6(( !imensions, A(*(, an! callous-unemotional traits as pre!ictors of treatment response in chil!ren $ith !isrupti+e beha+ior !isor!ers. Journal of Abnormal Psychology. /292<990IC98MC/7. [Pub"e! Kroneman :", *ip$ell AE, :oeber ., Koot *", Par!ini (A. Conte-tual risk factors as pre!ictors of !isrupti+e beha+ior !isor!er tra,ectories in girlsI the mo!erating effect of callous-unemotional features. Journal of Chil! Psychology an! Psychiatry. /299<7/I9>CM9C7. [P"C free article [Pub"e! :engua :J, Ko+acs EA. 4i!irectional associations bet$een temperament an! parenting an! the pre!iction of a!,ustment problems in mi!!le chil!hoo!. Journal of Applie! (e+elopmental Psychology. /227</>I/9M81. :ochman JE, ?ells KC. #he Coping Po$er Program for prea!olescent aggressi+e boys an! their parentsI 6utcome effects at the 9-year follo$-up. Journal of Consulting an! Clinical Psychology. /223<C/I7C9M7C1. [Pub"e! :oeber ., Par!ini (. =eurobiology an! the !e+elopment of +iolenceI Common assumptions an! contro+ersies. Philosophical #ransactions of the .oyal 'ociety 4-4iological 'ciences. /221<8>8I/309M/728. [P"C free article [Pub"e! "arsh AA, 4lair .J.. (eficits in facial affect recognition among antisocial populationsI A meta-analysis. =euroscience an! 4iobeha+ioral .e+ie$s. /221<8/I373M3>7. [P"C free article [Pub"e! "c(onal! ., (o!son "C, .osenfiel! (, Jouriles E=. Effects of a parenting inter+ention on features of psychopathy in chil!ren. Journal of Abnormal Chil! Psychology. /299<80I9298M92/8. [Pub"e! "c"ahon .J, ?itkie$itD K, Kotler J', Con!uct Problems Pre+ention .esearch Broup Pre!icti+e +ali!ity of callous-unemotional traits measure! in early a!olescence $ith respect to multiple antisocial outcomes. Journal of Abnormal Psychology. /292<990IC7/MC>8. [P"C free article [Pub"e! "offitt #E. :ife-course persistent +ersus a!olescence-limite! antisocial beha+ior. InI Cicchetti (, Cohen J, e!itors. (e+elopmental psychopathology, /n! e!itionI .isk, !isor!er, an! a!aptation. =e$ VorkI ?iley< /22>. pp. 7C2M701. 6!gers C:, "offitt #E, 4roa!bent J", (ickson =, *anco- .J, *arrington *, Caspi A. 5emale an! male antisocial tra,ectoriesI 5rom chil!hoo! origins to a!ult outcomes. (e+elopment an! Psychopathology. /221</2I>C8MC9>. [Pub"e! 6robio !e Castro 4, Peerman J?, Koops ?, 4osch J(, "onshou$er *J. *ostile attribution of intent an! aggressi+e beha+iorI A meta-analysis. Chil! (e+elopment. /22/<C8I09>M083. [Pub"e! Par!ini (A. #he callousness path$ay to se+ere +iolent !elin%uency. Aggressi+e 4eha+ior. /22><8/I702M701. Par!ini (A, 5ite PJ. 'ymptoms of con!uct !isor!er, oppositional !efiant !isor!er, attention-!eficitHhyperacti+ity !isor!er, an! callous-unemotional traits as uni%ue pre!ictors of psychosocial mala!,ustment in boysI A!+ancing an e+i!ence base

for ('"-P. Journal of the American Aca!emy of Chil! an! A!olescent Psychiatry. /292<30I9983M9933. [P"C free article [Pub"e! Par!ini (A, 5rick PJ, "offitt #E. 4uil!ing an e+i!ence base for ('"-7 conceptualiDations of oppositional !efiant !isor!er an! con!uct !isor!erI Intro!uction to the special section. Journal of Abnormal Psychology. /292<990I>18M>11. [Pub"e! Par!ini (A, :ochman JE, 5rick PJ. CallousHunemotional traits an! socialcogniti+e processes in a!,u!icate! youths. Journal of the American Aca!emy of Chil! an! A!olescent Psychiatry. /228<3/I8>3M8C9. [Pub"e! Par!ini (A, :ochman JE, Po$ell =. #he !e+elopment of callous-unemotional traits an! antisocial beha+ior in chil!renI Are there share! an!Hor uni%ue pre!ictorsT Journal of Clinical Chil! an! A!olescent Psychology. /22C<8>I890M 888. [Pub"e! Par!ini (A, 'tepp ', *ip$ell A, 'touthamer-:oeber ", :oeber .. #he clinical utility of the propose! ('"-7 callous-unemotional subtype of con!uct !isor!er in young girls. Journal of the American Aca!emy of Chil! an! A!olescent Psychiatry. /29/<79I>/MC8. [P"C free article [Pub"e! Pasalich (', (a!!s "., *a$es (J, 4rennan J. (o callous-unemotional traits mo!erate the relati+e importance of parental coercion +ersus $armth in chil! con!uct problemsT An obser+ational stu!y. Journal of Chil! Psychology an! Psychiatry. /299<7/I9821M9897. [Pub"e! Pollak '(, 'inha P. Effects of early e-perience on chil!renZs recognition of facial !isplays of emotion. (e+elopmental Psychology. /22/<81IC13MC09. [Pub"e! .othbart "K, Aha!i 'A, *ershey K:. #emperament an! social beha+ior in chil!hoo!. "errill-Palmer `uarterly. 9003<32I/9M80. .o$e ., "aughan 4, "oran P, 5or! #, 4riskman J, Boo!man .. #he role of callous an! unemotional traits in the !iagnosis of con!uct !isor!er. Journal of Chil! Psychology an! Psychiatry. /292<79I>11M>07. [Pub"e! 'chultD (, IDar! CE, 4ear B. Chil!renZs emotion processingI .elations to emotionality an! aggression. (e+elopment an! Psychopathology. /223<9>I8C9M 81C. [Pub"e! 'hiel!s A, Cicchetti (. .eacti+e aggression among maltreate! chil!renI #he contributions of attention an! emotion !ysregulation. Journal of Clinical Chil! Psychology. 9001</CI819M807. [Pub"e! 'omech :V, EliDur V. Promoting self-regulation an! cooperation in prekin!ergarten chil!ren $ith con!uct problemsI A ran!omiDe! controlle! trial. Journal of the American Aca!emy of Chil! an! A!olescent Psychiatry. /29/<79I39/M3//. [Pub"e! 'tringaris A. Irritability in chil!ren an! a!olescentsI A challenge for ('"-7. European Chil! an! A!olescent Psychiatry. /299</2I>9M>>. [Pub"e! #aylor A, Kim-Cohen J. "eta-analysis of geneMen+ironment interactions in !e+elopmental psychopathology. (e+elopment an! Psychopathology. /22C<90I92/0M928C. [Pub"e! #homas ., Simmer-Bembeck "J. 4eha+ioral outcomes of Parent-Chil! Interaction #herapy an! #riple P-Positi+e Parenting ProgramI A re+ie$ an! metaanalysis. Journal of Abnormal Chil! Psychology. /22C<87I3C7M307. [Pub"e!

10 ' re##i+e be"a+ior 'u$"or 0anu#cri.$ N1= Public 'cce##

)chool 'utcomes of Aggressi$e! Disrupti$e Children: Prediction *rom Kindergarten Risk *actors and "mpact of the *ast ,rack Pre$ention Program
Karen :. 4ierman, John Coie, [... , an! Con!uct Problems Pre+ention .esearch Broup A!!itional article information

Abstract
A multi-gate screening process i!entifie! 109 chil!ren $ith aggressi+e-!isrupti+e beha+ior problems at school entry. 5ast #rack pro+i!e! a multi-component pre+enti+e inter+ention in the conte-t of a ran!omiDe!-controlle! !esign. In a!!ition to psychosocial support an! skill training for parents an! chil!ren, the inter+ention inclu!e! intensi+e rea!ing tutoring in first gra!e, beha+ioral management consultation $ith teachers, an! the pro+ision of home$ork support &as nee!e!) through tenth gra!e. #his stu!y e-amine! the impact of the inter+ention, as $ell as the impact of the chil!Ks initial aggressi+e-!isrupti+e beha+iors an! associate! school rea!iness skills &cogniti+e ability, rea!ing rea!iness, attention problems) on aca!emic progress an! e!ucational placements !uring elementary school &Bra!es 9M3) an! !uring the secon!ary school years &Bra!es CM92), as $ell as high school gra!uation. Chil! beha+ior problems an! skills at school entry pre!icte! school !ifficulties &lo$ gra!es, gra!e retention, placement in a self-containe! classroom, beha+ior !isor!er classification, an! failure to gra!uate). (isappointingly, inter+ention !i! not significantly impro+e these long-term school outcomes.
8eywords* a re##i+e;%i#ru.$i+e c"il%ren2 #c"ool ou$co3e#2 #c"ool rea%ine##2 .re+en$i+e in$er+en$ion

" ,R'D:C,"'
#he aca!emic !ifficulties of chil!ren $ith Fearly startingG aggressi+e-!isrupti+e con!uct problems are $ell-establishe!. Aggressi+e-!isrupti+e beha+iors in the early elementary years pre!ict later aca!emic un!er-achie+ement, nee! for special e!ucation, an!

increase! likelihoo! of school !ropout &Coie ; (o!ge, 9001< 'helley-#remblay, 6K4rien, ; :anghinrichsen-.ohling, /22C). #hese school !ifficulties, in turn, amplify the longterm social han!icaps e-perience! by many aggressi+e chil!ren, inclu!ing un!eremployment, limite! occupational options, financial stress, an! poor %uality of social relationships &Kokko ; Pulkkinen, /222< ?in!le ; ?in!le, 9007). "ultiple factors an! mechanisms may contribute to the school mala!,ustment of aggressi+e chil!ren. Cogniti+e !eficits, inclu!ing lo$er le+els of cogniti+e ability, attention problems, an! !ifficulties learning to rea! ha+e figure! centrally in !e+elopmental mo!els linking early aggression $ith un!erachie+ement an! special e!ucation &*insha$, 900/< "aguin, :oeber, ; :e"ahieu, 9008). In a!!ition, !e+elopmental research has i!entifie! beha+ioral an! social factors that contribute to the stability of aggression o+er time an! un!ermine learning, inclu!ing conflict $ith teachers, peer !ifficulties, an! lo$ le+els of learning engagement &Chen, *uang, Chang, ?ang, ; :i, /292< 'tipek ; "iles, /221). #he 5ast #rack pre+ention program inclu!e! multiple components !esigne! to a!!ress the beha+ioral, social, an! aca!emic nee!s of chil!ren starting school $ith high rates of aggressi+e beha+ior &see Con!uct Problems Pre+ention .esearch Broup [CPP.B , 900/). Boals inclu!e! the promotion of parental support an! effecti+e beha+ior management skills, chil! social competence an! positi+e peer relations, teacher support an! effecti+e classroom management skills, an! chil! rea!ing rea!iness an! school engagement. 'hortterm effects !uring the initial elementary years $ere positi+e, inclu!ing impro+e! scores on measures of social cognitions, peer relations, rea!ing rea!iness, an! re!uce! aggression &see CPP.B, 9000, /22/). In a!!ition, analyses of late a!olescent outcomes !ocument lo$er le+els of criminal arrest among chil!ren in the inter+ention than control group &CPP.B, /292a) an! lo$er rates of con!uct !isor!er !iagnosis among the youth $ith the highest le+els of initial risk $ho recei+e! inter+ention &CPP.B, /299). #he purpose of this paper $as to e-amine broa!er, long-term in!ices of chil!renKs school progress an! e!ucational outcomes, inclu!ing gra!es, gra!e retention, special e!ucation placement in a self-containe! classroom, beha+ior !isor!er classification, an! high school gra!uation. #he t$o ma,or aims $ereI &i) to i!entify the uni%ue contributions that early cogniti+e an! beha+ioral chil! characteristics &cogniti+e ability, rea!ing rea!iness, inattention, an! aggressi+e-!isrupti+e beha+ior) ma!e to school mala!,ustment !uring the elementary an! secon!ary school years in a sample of high-risk aggressi+e chil!ren, an! &ii) to e+aluate the impact of the 5ast #rack pre+ention program on these school outcomes.

D%;%/'PM% ,A/ PR'C%))%) A))'C"A,%D 5",- %AR/< A++R%))"' A D : D%RAC-"%;%M% ,


Chil!ren $ho e-hibit ele+ate! aggression at school entry &e.g., fighting, yelling, harming others) often sho$ concurrent oppositional beha+ior an! rule +iolations &e. g., refusing to

comply $ith a!ult re%uests, !ifficulty follo$ing !irections, temper outbursts) &Coie ; (o!ge, 9001). "any also sho$ ele+ate! impulsi+e or hyperacti+e beha+iors, an! !isplay high rates of off-task beha+iors &*insha$, 900/). (e+elopmental research regar!ing the links bet$een early aggression an! later school mala!,ustment has focuse! alternati+ely on the role playe! by the cogniti+e skill !eficits that often accompany aggressi+e beha+ior, an!Hor on the role of a negati+e socialiDation an! moti+ational casca!e e+oke! by early aggression, inclu!ing intensi+e interpersonal conflict an! subse%uent school !isengagement. Cogniti+e factors often associate! $ith aggressi+e-!isrupti+e beha+ior inclu!e lo$er I`, attention problems, an! lo$ rea!ing rea!iness &"aguin et al., 9008). 4ase! upon a comprehensi+e re+ie$ of stu!ies e-amining the link bet$een e-ternaliDing problems an! aca!emic un!erachie+ement, *insha$ &900/) conclu!e! that inattention-hyperacti+ity is specifically linke! to rea!ing un!erachie+ement in early an! mi!!le chil!hoo!. 'imilarly, after analyDing longitu!inal !ata on the large (une!in sample, "offitt &9008) i!entifie! neuropsychological !eficits &lo$ I`, attention problems) as key pre!icti+e risks associate! $ith stable early-starting aggressi+e con!uct problems an! associate! learning !ifficulties. "aguin et al. &9008) foun! that the combination of early attention problems an! rea!ing !ifficulties characteriDe! many !elin%uent youth. Although these cogniti+e skills &lo$ I`, attention problems, poor rea!ing rea!iness) are implicate! in the school !ifficulties e-perience! by young aggressi+e chil!ren, the relati+e roles that they play in contributing to +arious aspects of school mala!,ustment is not $ell-un!erstoo!. Prior research suggests that these cogniti+e factors may play a greater role in pre!icting school mala!,ustment !uring the elementary school years than in the secon!ary school years &*insha$, 900/< *uesmann, Eron, ; Varmel, 901C) an! may be more important !eterminants of aca!emic progress &e.g., gra!es, gra!e retention) than of school beha+ior problems or e!ucational placement &e.g., beha+ior !isor!er classification, placement in self-containe! classrooms). In a!!ition to cogniti+e skills, beha+ioral, an! social factors appear to play a key role in un!ermining the school a!,ustment an! learning engagement of young chil!ren $ho are aggressi+e. In a longitu!inal stu!y mo!eling !e+elopmental influences !uring the early elementary years, Chen et al. &/292) foun! that aggression ha! uni%ue effects on later social competence an! aca!emic achie+ement after controlling for earlier le+els of social competence an! achie+ement. 'imilarly, in their long-term longitu!inal stu!y, *uesmann et al. &901C) foun! that, although lo$ I` in the early years pre!icte! ele+ate! aggression through age 1, le+els of aggression &rather than I`) pre!icte! intellectual achie+ement in a!ulthoo!. #hey hypothesiDe! that, in early chil!hoo!, lo$ intelligence impairs the learning of aggression control, thereby increasing le+els of aggression !uring the early elementary school years. Aggressi+e beha+ior then impe!es learning an! intellectual !e+elopment in the secon!ary school years an! early a!ulthoo!. #hese fin!ings len! support to a negati+e casca!e mo!el in $hich early aggression un!ermines school a!,ustment o+er time through both !irect an! in!irect processes &see also (o!ge, Breenberg, "alone, ; CPP.B, /221). Aggression at school entry often e+okes negati+e social reactions from peers an! a!ults, increasing interpersonal conflict,

re!ucing social support, an! thereby fueling negati+e attitu!es to$ar!s school. 5or e-ample, follo$ing chil!ren o+er the course of the elementary school years, 'tipek an! "iles &/221) !ocumente! transactional influences bet$een chil! aggression an! teacher conflict, $ith the effects of aggression on achie+ement partially me!iate! by conflictual relationships $ith teachers. Controlling for family an! neighborhoo! !emographics, ?hite an! :oeber &/221) foun! that aggressi+e beha+ior increase! peer !islike an! social marginaliDation an!, along $ith poor aca!emic achie+ement, significantly increase! the likelihoo! of special e!ucation placement. Enfortunately, school-base! inter+entions !esigne! to help struggling stu!ents sometimes increase feelings of alienation an! e-acerbate aggressi+e beha+ior problems. 5or e-ample, Jimerson an! 5erguson &/22C) e-amine! the impact of gra!e retention on aggressi+e beha+ior. #heir statistical analyses, $hich controlle! for prior le+els of aggression, re+eale! that stu!ents $ho $ere retaine! in gra!e, or place! into a Ftransition classroomG to gi+e them an e-tra year of early elementary school became more aggressi+e than non-retaine! stu!ents in mi!!le school an! more likely to !rop out. #he increase! aggression le+els among retaine! stu!ents may reflect feelings of frustration regar!ing their !ecrease! aca!emic status, as $ell as increase! peer relationship challenges face! by stu!ents after gra!e retention. 'ome research suggests that placement in special e!ucation classes is also linke! to the escalation of aggressi+e beha+ior !ue to peer contagion &Pisser, Kunnen, ; Pan Beert, /292), although not all stu!ies fin! this negati+e impact &?hite ; :oeber, /221). #he relati+e impact of cogniti+e +ersus social an! moti+ational factors on the school mala!,ustment of aggressi+e chil!ren may change $ith age, $ith cogniti+e factors playing a greater role in the initial elementary years, an! social, beha+ioral, an! attitu!inal factors playing a greater role in the secon!ary school years &*uesmann et al., 901C). 4y prea!olescence, unpleasant confrontations $ith teachers an! peers create a situation in $hich high-risk chil!ren, $ho ha+e ha! histories of poor school performance an! poor peer relations in elementary school, make the transition to mi!!le school feeling alienate! from a!ults an! !isengage! from school &*a$kins, Catalano, ; "iller, 900/). After the mi!!le-school transition, youth e-perience re!uctions in parent an! teacher support an! monitoring, allo$ing them more autonomy an! more time $ith peers. #his free!om increases the risk that chil!ren $ith lo$ le+els of commitment to school $ill initiate a!olescent problem beha+iors, inclu!ing truancy, substance use, an! antisocial acti+ities that further un!ermine their aca!emic performance &(ishion, Piehler, ; "eyers, /221< *a$kins et al., 900/).

RA,"' A/% *'R M:/,"!C'MP' % , )C-''/! BA)%D " ,%R;% ,"' )


Bi+en the multiple !e+elopmental processes associate! $ith the poor school outcomes e-perience! by early-starting aggressi+e chil!ren, effecti+e pre+enti+e inter+entions likely nee! to a!!ress multiple skill !omains &CPP.B, 900/). Prior research suggests that inter+entions that combine aca!emic tutoring a!!ressing the cogniti+e !elays of early-

starting aggressi+e chil!ren $ith social competence training a!!ressing their self-control an! prosocial skills can impro+e their school a!,ustment in the short term &Coie ; Krehbiel, 9013). Prior research also suggests that training teachers in positi+e classroom management skills can re!uce aggressi+e-!isrupti+e beha+iors an! impro+e the on-task beha+ior of chil!ren $ith cogniti+e !elays &*a$kins, Catalano, Kosterman, Abbott, ; *ill, 9000< "etropolitan Area .esearch Broup [Eron et al. , /22/< ?alker, 900>). 5inally, uni+ersal social-emotional learning programs ha+e pro+en effecti+e in promoting positi+e classroom climate, an! in some stu!ies, enhancing peer relations an! re!ucing classroom rates of aggression, $hich might re!uce the peer influences that make aggressi+e beha+ior functional in the classroom &(urlak et al., /299< Breenberg et al., /228).

,-% *A), ,RACK " ,%R;% ,"'


4ase! on prior !e+elopment an! inter+ention research, the 5ast #rack pre+ention program inclu!e! multiple components, combining se+eral approaches that ha! prior e+i!ence of efficacy in re!ucing chil! e-ternaliDing beha+ior problems an! promoting chil! self-regulation skills. (escribe! further in the metho!s section, the 5ast #rack pre+ention program inclu!e! a Funi+ersalG social-emotional learning program implemente! by classroom teachers in Bra!es 9M7, the PA#*' &Promoting Alternati+e #*inking 'trategies) Curriculum &Breenberg ; Kusche, 9008) along $ith the pro+ision of consultation an! support to classroom teachers in the !e+elopment an! implementation of positi+e beha+ior management programs. #arget chil!ren &e.g., those $ho screene! into the high-risk sample base! upon parent an! teacher ratings of aggressi+e-!isrupti+e beha+iors) $ere in+ite! to atten! e-tracurricular group meetings $hich inclu!e! social skills training for them an! parent management training for their parents. #o enhance the generaliDation of the social skill training program, high-risk chil!ren also recei+e! a Fpeer-pairingG program to support positi+e peer relations at school. #o a!!ress !elays in rea!ing rea!iness, all high-risk chil!ren recei+e! tutoring in rea!ing skills !uring first gra!e, using an e+i!ence-base! phonics approach. 4ase! on annual !iscussions $ith teachers an! in!i+i!ualiDe! case plans, target chil!ren $ere eligible to recei+e a!!itional aca!emic support in subse%uent years through Bra!e 92, primarily focuse! on support for home$ork completion. Prior stu!ies ha+e !ocumente! the positi+e impact of the 5ast #rack Program on youth beha+ioral, social, an! aca!emic a!,ustment intheinitial elementaryschoolyears. At the en! of first gra!e, in comparison to the high-risk control chil!ren, the high-risk inter+ention chil!ren !isplaye! positi+e impro+ements in social-cogniti+e skills &e.g., emotion recognition, emotion coping, an! social problem sol+ing skills), obser+e! positi+e interactions $ith peers at school, an! impro+e! sociometric social preference scores. #eachers rate! the inter+ention chil!ren as ha+ing significantly lo$er rates of aggressi+e, oppositional beha+iors in the school setting &CPP.B, 9000). Aca!emically, chil!ren in the inter+ention con!ition !emonstrate! impro+e! $or! attack skills, attaine! higher languageartsgra!es,an!recei+e!fe$erminutesofspecial e!ucation ser+ices !uring Bra!e 9. At the en! of thir! gra!e, inter+ention chil!ren recei+e! lo$er teacher ratings of con!uct problems compare! $ith chil!ren in the control group, an! continue! to sho$

impro+e! social cogniti+e skills. *o$e+er, incontrast tothe effectsnote! attheen! of the first gra!e, the inter+ention effects $ere no longer e+i!ent on rea!ing achie+ement tests at the en! of thir! gra!e &CPP.B, /22/).

,-% PR%)% , ),:D<


#he purpose of this paper $as to e-amine broa!er, long-term in!ices of chil!renKs school outcomes, inclu!ing gra!es, gra!e retention, special e!ucation placement in a selfcontaine! classroom, beha+ior !isor!er classification, an! high school gra!uation. A key %uestion $as $hether the significant inter+ention effects obser+e! on !imensional measures in the early elementary years affecte! substanti+e school outcomes in later years, in $ays that might ha+e cost sa+ings for schools an! life-course impact for the participants. Bi+en e+i!ence for possible !ifferences in the !eterminants of school a!,ustment in the elementary an! secon!ary school years, $e e-amine! aca!emic outcomes at t$o perio!s of timeb!uring elementary school &Bra!es 9M3) an! !uring the secon!ary school years &Bra!es CM92), as $ell as rates of high school gra!uation. 4ecause $e $ere intereste! in markers of school mala!,ustment that ha! lo$ base rates an! inter!epen!encies across years &e.g., gra!e retention, special e!ucation classification an! placement), $e e-amine! the occurrence of these e+ents $ithin each broa! perio! of time &elementary years< secon!ary years), rather than on an annual basis. #his stu!y e-amine! the impact of the inter+ention, as $ell as the impact of the chil!Ks initial aggressi+e-!isrupti+e beha+iors an! associate! school rea!iness skills &cogniti+e ability, rea!ing rea!iness, attention problems) on aca!emic outcomes in the elementary an! secon!ary school years. #he t$o ma,or aims $ereI &i) to i!entify the uni%ue contributions that early cogniti+e an! beha+ioral chil! characteristics ma!e to school mala!,ustment !uring the elementary an! secon!ary school years in a sample of high-risk aggressi+e chil!ren, an! &ii) to e+aluate the impact of the 5ast #rack pre+ention program on these school outcomes. ?e hypothesiDe! that cogniti+e factors &lo$ I`, inattention, poor rea!ing rea!iness) $oul! be important uni%ue pre!ictors of school mala!,ustment in the elementary school years, particularly gra!es, gra!e retention, an! placement in a selfcontaine! classroom. ?e further hypothesiDe! that the se+erity of aggressi+e- !isrupti+e beha+ior $oul! be an important uni%ue pre!ictor of all forms of school mala!,ustment in the secon!ary school years, an! $oul! also contribute uni%uely to the classification as beha+ior !isor!ere! in the elementary school years. 4ase! upon the positi+e inter+ention fin!ings on re!uce! aggression, impro+e! peer relations, an! impro+e! rea!ing rea!iness at the en! of first gra!e, $e e-pecte! positi+e inter+ention effects on the measures of school a!,ustment an! attainment.

M%,-'D
Participants
5ifty-fi+e schools $ere i!entifie! at four !emographically !i+erse sites &=ash+ille, #=< (urham, =C< 'eattle, ?A< an! rural central PA), selecte! as high risk base! on crime

an! po+erty statistics. 6+er three consecuti+e years &9009M9008), the aggressi+e!isrupti+e con!uct problems of all 0,703 kin!ergarteners atten!ing these schools $ere rate! by teachers, using the #eacher 6bser+ation of Chil! A!,ustment-.e+ise! &#6CA.) Authority Acceptance 'cale &?erthamer-:arsson, Kellam, ; ?heeler, 9009). #he parents of chil!ren scoring in the top 32A $ithin cohort an! site $ere then solicite! for the ne-t stage of screening< 09A agree! to complete an inter+ie$ $hich inclu!e! ratings of chil! aggressi+e-!isrupti+e beha+iors, !ra$n primarily from the Chil! 4eha+ior Checklist &Achenbach, 9009) &for further !etails, see :ochman ; CPP.B, 9007). #he teacher an! parent screening scores $ere stan!ar!iDe! an! combine! to yiel! a total se+erity-of-risk screen score. ?ithin each site, participating schools $ere !i+i!e! into sets matche! for !emographics &siDe, percentage free or re!uce! lunch, ethnic composition), an! the sets $ere ran!omly assigne! to inter+ention or control con!itions. Chil!ren $ere selecte! for inclusion into the stu!y base! on their screen score, mo+ing from the highest score !o$n$ar! until !esire! sample siDes $ere reache! $ithin sites, cohorts, an! con!itions. (e+iations $ere ma!e $hen a chil! faile! to matriculate in the first gra!e at a core school &n @ 70) or refuse! to participate &n @ C7), or to accommo!ate a rule that no chil! $oul! be the only girl in an inter+ention group. #he outcome $as that 109 chil!ren &nKs @ 337 for inter+ention an! 33> for control) participate!. #he mean age of the high-risk chil!ren $as >.7 years &." @ 2.31) at the time of i!entification. Across all sites, the sample $as 79A African American, 3CA European American, an! /A of other ethnicity &e.g., Pacific Islan!er an! *ispanic), reflecting the ethnic !i+ersity of the population at the four sites. 'i-ty-nine percent $ere boys. 5ifty-eight percent of the high-risk chil!ren came from single-parent families, /0A of the parents $ere high school !ropouts, an! 87A of the families $ere in the lo$est socioeconomic class as !etermine! by *ollingshea! scoring &*ollingshea!, 90C7). 6n the kin!ergarten #eacherKs .eport 5orm of the Chil! 4eha+ior Checklist &#.5< Achenbach, 9009), the a+erage E-ternaliDing #-score &a+ailable for 11A of the high risk sample) $as >>.3, an! C>A of these chil!ren score! abo+e a #-score of >2, in!icating sub-clinical to clinical le+els of e-ternaliDing beha+ior problems. In a!!ition to the high-risk sample, a stratifie! normati+e sample of 81C chil!ren $as i!entifie! from the control schools to represent the population at each site. #hese chil!ren $ere not a part of the inter+ention analyses reporte! here, but $ere inclu!e! to pro+i!e a normati+e reference point for chil! outcomes. Informe! parental consent $as obtaine! for all participants, follo$ing APA gui!elines for ethical research.

"nter$ention Procedures

PA,-) Curriculum
In Bra!es 9M7, classroom teachers implemente! the PA#*' &Promoting Alternati+e #*inking 'trategies) Curriculum, a uni+ersal social-emotional learning program !esigne! to promote a positi+e classroom climate an! enhance chil!renKs social skills, emotional a$areness, an! self-control &Breenberg ; Kusche, 9008). #eachers recei+e!

!etaile! manuals an! materials an! participate! in 8-!ay $orkshops le! by certifie! trainers !uring the summer months.

,eacher consultation
5ast #rack E!ucational Coor!inators [ECs &!escribe! belo$) ma!e $eekly +isits to each classroom. #hey also met $ith teachers each month either in!i+i!ually or in small groups. ECs re+ie$e! the implementation of PA#*' $ith teachers, mo!ele! classroom lessons, an! !iscusse! classroom management challenges, offering support in planning an! implementing positi+e beha+ior management programs.

"ndi$idual tutoring
In first gra!e, all chil!ren recei+e! the ?allach rea!ing tutoring program &?allach ; ?allach, 90C>), !eli+ere! by paraprofessionals in three half-hour sessions per $eek. #his highly structure! one-on-one tutoring program $as !esigne! for use by paraprofessionals to ser+e lo$-rea!iness chil!ren from !isa!+antage! backgroun!s. #he program emphasiDe! a phonics-base!, mastery-oriente! approach to the !e+elopment of initial rea!ing skills. In later years, as note! belo$, aca!emic tutoring $as pro+i!e! to a subset of chil!ren $ho sho$e! on-going aca!emic !ifficulties an! fell in the bottom thir! of the class in rea!ing skills. 6nce chil!ren ha! mastere! the ?allach program, tutoring focuse! on home$ork completion an! other learning acti+ities recommen!e! by the classroom teacher.

Peer pairing
In first gra!e, all chil!ren recei+e! peer pairing sessions, !eli+ere! by paraprofessionals in one half-hour session per $eek. 5ollo$ing a scripte! manual, these super+ise! play sessions ga+e target chil!ren an opportunity to practice social skills $ith classmates &partners rotate! o+er the course of the year), to support skill generaliDation an! foster a positi+e peer reputation. In secon! gra!e, peer-pairing $as offere! to the subset of chil!ren $ho continue! to e-perience peer re,ection or e-hibit high rates of aggression to$ar! peers at assessments con!ucte! at the en! of first gra!e. In the later elementary years &Bra!es 8M7) peer support $as offere! to chil!ren $ith ongoing social nee!s, in the conte-t of a $eekly Fsocial clubG or Flunch clubG offere! after school or !uring an in!oor recess perio!.

Middle school transition program


At the point of transition into mi!!le school &$hich happene! follo$ing Bra!e 3, 7, or > for youth in this sample), stu!ents $ere offere! a set of eight small group e-periences. #hese inclu!e! orientation an! tours of the mi!!le school, an! !iscussion acti+ities !esigne! to pro+i!e support in the $eeks imme!iately prece!ing an! follo$ing the transition.

Academic support in the secondary school years


In!i+i!ualiDe! case assessments $ere use! to !etermine the nee! for aca!emic support in the later school years &Bra!es >M92). Primarily, secon!ary ser+ices for chil!ren $ith aca!emic nee!s focuse! on home$ork support, +ocational interest e-ploration, an! in some cases, $orking $ith parents an! teachers to a!!ress beha+ioral or atten!ance issues.

%&tracurricular parent training and child social skill training groups


Parent management skill training groups focuse! on impro+ing parentMchil! relationships an! parenting that pro+i!e! positi+e support, consistent an! non-puniti+e limit-setting, monitoring an! communicating $ith their chil!. Chil! social skill training groups focuse! on promoting prosocial an! self-regulation skills. #hese groups $ere hel! !uring /-hour e-tracurricular sessions hel! outsi!e of the school conte-t. Broups met $eekly in Bra!e 9 &// sessions) an! bi-$eekly in Bra!e / &93 sessions), follo$e! by monthly Fbooster sessionG groups in Bra!es 8M7 &0 sessions each year). In a!!ition, families recei+e! in!i+i!ual home +isits &$ith +arie! fre%uency, base! upon criterion-reference! assessments of nee!) to promote the generaliDation of skills presente! in the group an! to a!!ress in!i+i!ual nee!s &for more !etail, see 4ierman, Breenberg, ; CPP.B, 900>< "c"ahon, 'lough, ; CPP.B, 900>).

%ducation Consultants 1%Cs2


#he ECs at each site $ere e-perience! classroom teachers $ho ha! e-pertise $orking $ith beha+iorally challenging chil!ren. #hey recei+e! 8 !ays of cross-site training each year, an! follo$e! !etaile! manuals !escribing the assessment an! inter+ention protocols for each of the key program components. #hey $ere super+ise! in $eekly meetings at each site< super+isors participate! in $eekly cross-site phone calls $ith the principal in+estigators $ho o+ersa$ the key program components. ECs !eli+ere! the social skill training to target chil!ren, consulte! $ith classroom teachers on PA#*' implementation an! classroom management issues, an! super+ise! paraprofessionals in the implementation of the tutoring an! peer-paring programs.

"nter$ention participation
6nly a fe$ families &>A of the sample) !ecline! to initiate participation in the parent or chil! groups in Bra!e 9. 6f the 03A participating, C0A of the parents an! 02A of the chil!ren atten!e! at least 72A of the sessions offere!. Participation in the inter+ention !ecrease! gra!ually across the years, particularly as families mo+e!. 4y the mi!!le school years, appro-imately 97A of the youth in the sample $ere rate! as lo$ risk an! recommen!e! for minimal ser+ices each year, but a ma,ority of youth continue! to recei+e some ser+ices through tenth gra!e. In terms of the school-base! inter+ention ser+ices, !uring first-gra!e, the program protocol calle! for three tutoring sessions per $eek for all chil!ren assigne! to the inter+ention con!ition. All e-cept for three chil!ren got some tutoring in first gra!e, $ith

an a+erage of 72 sessions !eli+ere! &C8A of the planne! amount). In the subse%uent years, ECs met $ith teachers to !iscuss chil!renKs aca!emic progress an! !etermine their nee! for tutoring. ?hen teachers felt that chil!ren $ere performing Fin the bottom thir!G of the class, chil!ren %ualifie! for ongoing tutoring sessions, sche!ule! 9M/ times per $eek. 6n a+erage, t$o-thir!s of the chil!ren in the sample %ualifie! for tutoring each year &range from >2A to C3A), an! o+er 12A of those $ho %ualifie! recei+e! tutoring each year. 'ome coul! not be ser+e!, primarily !ue to mo+es out of the ser+ice area, an! this number increase! by about 7A each year. 6f those $ho recei+e! tutoring, the a+erage number of sessions $as 32 per year !uring the elementary years an! 92 per year !uring the secon!ary years. All analyses are base! upon an Fintent to treatG mo!el $hich reflects the assignment of chil!ren an! families to inter+ention +ersus control con!itions, regar!less of the !egree to $hich chil!ren an! families actually engage! in the inter+ention, or the amount of inter+ention recei+e!.

Child Characteristics at )chool %ntry: Predictors of )chool 'utcomes


5our measures assesse! chil! cogniti+e an! beha+ioral school rea!iness at school entry. #hese $ere e-amine! as pre!ictors of school !ifficulties.

Cogniti$e ability
#he Pocabulary an! 4lock (esign subtests from the ?echsler Intelligence 'cales for Chil!ren-.e+ise! &?I'C-.< ?echsler, /22/) $ere a!ministere! in!i+i!ually to all chil!ren $hen they $ere in kin!ergarten. #he scale scores $ere stan!ar!iDe! an! a+erage! to form a composite estimate of cogniti+e ability, base! upon prior research suggesting that these t$o subtests are significantly correlate! $ith the full scale I` &rs @ . C3 an! .>1, respecti+ely< 'attler, 900/).

Reading readiness
#he :etter-?or! I!entification subtest from the ?oo!cock-Johnson Psycho-E!ucational 4attery-.e+ise! &?oo!cock ; Johnson, 9010) assesse! rea!ing rea!iness in kin!ergarten. 5i+e items on this subtest assesse! the chil!Ks ability to match a pictorial representation of a $or! $ith an actual picture of an ob,ect< the remaining items assesse! the chil!Ks ability to i!entify letters an! $or!s. 4ase! upon its pre!icti+e +ali!ity, this subtest has been use! by other researchers as a measure of early rea!ing ability &?oo! ; 5elton, 9003). #he number of items ans$ere! correctly on this subtest ser+e! as our measure of early rea!ing achie+ement.

"nattention
#he Attention Problems subscale of the Chil! 4eha+ior Checklistb#.5 &Achenbach, 9009) $as use! to assess inattention. #his scale inclu!es 97 items !escribing concentration !ifficulties an! task-oriente! beha+ior problems, each rate! on a 8-point

scale. In prior research, the Attention Problems scale has !emonstrate! a!e%uate internal consistency, $ith alphas ranging from .1/ to .02. Preliminary analyses re+eale! that inattention ha! a non-linear relationship $ith se+eral of the aca!emic outcomes in this stu!y. 5or this reason, it $as trichotomiDe!, $ith se+ere le+els of inattention in!icate! by a #-score of C2 or higher &se+ere inattention @ /), an! mo!erate le+els of inattention in!icate! by a T-score bet$een >7 an! C2 &mo!erately inattenti+e @ 9), an! normati+e le+els of inattention in!icate! by a #-score belo$ >7 &not inattenti+e @ 2). Applying these criteria, 9>A of the chil!ren in the high-risk sample sho$e! mo!erate inattention in kin!ergarten, an! 9CA sho$e! se+ere inattention in kin!ergarten.

Aggressi$e!disrupti$e beha$ior
#he #6CA-. Authority Acceptance scale &?erthamer-:arsson et al., 9009) $as use! as a measure of aggressi+e-!isrupti+e beha+ior in the school. #he scale inclu!es 92 items !escribing aggressi+e, !isrupti+e, an! oppositional beha+iors, each rate! on a >-point scale &c @ .03). 'cores represent the a+erage item rating that chil!ren recei+e! on this scale.

'utcome Measures of )chool Malad(ustment


5ollo$ing each aca!emic year, !ata regar!ing gra!es an! special e!ucation $ere collecte! from school recor!s. #he schools +arie! $i!ely in the timing of the transition bet$een elementary school an! mi!!le school or ,unior high. "ost transitione! after fifth gra!e, but many transitione! after si-th gra!e, an! a small proportion transitione! after fourth gra!e. Initial inspection of the !ata suggeste! that gra!es an! special e!ucation ser+ices $ere strongly affecte! by the transition. In or!er to create comparability across the large number of schools, $e focuse! on gra!es an! special e!ucation status !uring the t$o perio!s of stability in school conte-t, the elementary years &Bra!es 9M3), an! the secon!ary years, after all chil!ren ha! ma!e the transition into a larger school conte-t &Bra!es CM92).

+rade point a$erage


Each aca!emic year, gra!es $ere recor!e! for four core aca!emic sub,ectsI math, language artsHEnglish, social stu!iesHhistory, an! science. Bra!es $ere score! using a 3point system &A @ 3, 4 @ 8, C @ /, ( @ 9, EH5 @ 2). #hey $ere a+erage! across sub,ects an! years to create a BPA in!e- representing aca!emic gra!es in elementary school &Bra!es 9M3) an! secon!ary school &Bra!es CM92). =ote that, !ue to retentions, some chil!ren ha! more than 3 years of !ata to contribute to these a+erages.

Retention
'chool recor!s $ere co!e! for retention. #hese !ata $ere collapse! into the categories of Fe+er retaine!G !uring Bra!es 9M3 an! Fe+er retaine!G !uring Bra!es CM92.

Beha$ior disorder classifications


'chool recor!s $ere co!e! for IEP an! !iagnostic classification. #hese !ata $ere collapse! into the categories of Fe+erG classifie! as ha+ing a 4eha+ior (isor!er !uring Bra!es 9M3 an! Fe+erG classifie! as ha+ing a 4eha+ior (isor!er !uring Bra!es CM92.

)elf!contained Placement
'chool recor!s $ere co!e! for classroom placement. A youth $as consi!ere! to be in Fself-containe!G placement $hen they spent a ma,ority of their !ay in a special-e!ucation Fself-containe!G classroom, or $hen they $ere place! in a special school setting, such as a !ay treatment or resi!ential program. #hese !ata $ere collapse! into the categories of Fe+erG place! in a self-containe! setting !uring Bra!es 9M3 an! Fe+erG place! in a selfcontaine! setting !uring Bra!es CM92.

Assessment Procedures
At the en! of the kin!ergarten year, teachers complete! the #.5, $hich pro+i!e! an in!e- of inattention at school entry. In the fall of the first gra!e year, teachers complete! the #6CA-., $hich pro+i!e! an in!e- of aggressi+e beha+ior at school entry. At both time points, a research assistant !eli+ere! the measures to teachers, e-plaine! them, an! left them for teachers to complete an! return to the pro,ect. #eachers recei+e! m92 per chil! for completing these measures. (uring the summer follo$ing kin!ergarten, t$o traine! inter+ie$ers +isite! each family home. Inter+ie$ers $ere naa+e regar!ing the inter+ention +ersus control status of the families they $ere assigne! to inter+ie$. ?hile one research assistant inter+ie$e! the primary caregi+er &usually the mother), a secon! assistant inter+ie$e! the chil! in a separate room, an! a!ministere! in!i+i!ual tests an! %uestionnaires, inclu!ing the measures of cogniti+e ability an! rea!ing rea!iness. Parents an! teachers $ere compensate! financially for their participation.

Attrition and missing data


A total of >>2 youth &out of an original 109) ha! complete aca!emic !ata< the others $ere missing !ata for one or more years. Attrition analyses comparing the pre-inter+ention scores of the >>2 youth $ith complete !ata $ith the /89 youth $ho ha! missing !ata re+eale! a fe$ significant !ifferences. Vouth $ith missing !ata ha! higher rea!ing rea!iness scores an! higher cogniti+e ability than youth $ith complete !ata, but the t$o groups !i! not !iffer on initial aggression or inattention. Vouth $ith missing !ata $ere most likely to be from the 'eattle site &8CA $ith missing !ata) an! least likely to be from the central Pennsyl+ania site &9>A $ith missing !ata), reflecting the !ifferential population stability at these sites. Vouth $ith missing !ata !i! not !iffer from those $ith complete !ata on race or gen!er. In the analyses, multiple imputation $as use! to a!!ress missing !ata. "issing !ata $ere multiply impute! &'chafer, 9000) using 'A' Proc "I.

In!epen!ent analyses $ere performe! on /2 impute! !ata sets. Parameter estimates an! +ariances $ere then combine! to obtain an unbiase! estimate of the population +alues.

R%):/,)
Descripti$e Analyses
(escripti+e statistics are presente! in #able I. Initially, t-tests $ere con!ucte! to i!entify +ariables that significantly !ifferentiate! the high-risk youth &the inter+ention an! control groups $ho e-hibite! ele+ate! con!uct problems at school entry) from the normati+e sample of youth atten!ing the same schools as the chil!ren in the control sample. #hese preliminary analyses $ere un!ertaken to !etermine the !egree to $hich high-risk chil!ren $ho $ere selecte! on the basis of ele+ate! aggressi+e-!isrupti+e beha+ior problems also e-hibite! significantly lo$er le+els of school rea!iness at kin!ergarten entry compare! to the normati+e sample !ra$n from the same schools. 'ignificant !ifferences emerge!, $ith aggressi+e-!isrupti+e chil!ren e-hibiting lo$er le+els of cogniti+e ability an! rea!ing rea!iness, an! higher le+els of teacher-rate! inattention an! aggression than the normati+e sample &see #able I). 4ase! upon the t$o scale scores, the I` estimate for youth in the normati+e sample !ra$n from these high-risk schools $as 07.92 &." @ 91.80), $hereas the I` estimate for youth in the aggressi+e high-risk sample $as 17.02 &'( @ 9>.>7). #hese high-risk chil!ren $ent on to e-perience school mala!,ustment at higher rates than the normati+e sample !uring both the elementary an! secon!ary school years, inclu!ing lo$er gra!es, higher rates of beha+ior !isor!er classification an! selfcontaine! placement, an! lo$er rates of high school gra!uation. 6nly gra!e retentions !i! not !ifferentiate the high-risk aggressi+e an! normati+e chil!ren. Interestingly, these fin!ings also illustrate a general increase in the pre+alence of school problems e-perience! by chil!ren &both high risk an! normati+e comparison groups) in the secon!ary school years compare! to the elementary school years. ?hereas mean BPAs for the high-risk youth ho+ere! in the 4HC range in Bra!es 9M3, they plummete! to the CH( range in Bra!es CM92. 4et$een the elementary an! secon!ary school years, high-risk youth e-perience! a t$ofol! increase in rates of 4eha+ior (isor!er classification &from 1.7A to 9>.7A), an! nearly a threefol! increase in rates of self-containe! classroom placements &from 99A to /1A). 4y high school, the high-risk youth $ere fi+e times more likely than normati+e comparison chil!ren to be classifie! as 4eha+ior (isor!ere! &9>.7A +s. 3A) an! three times more likely to be place! in a self-containe! classroom &/1A +s. 92A). #hey also ha! lo$er BPAs &9.C8 +s. /.2/), an! $ere significantly less likely to gra!uate from high school &77.7A +s. >>A). #hese fin!ings replicate prior stu!ies !ocumenting that chil!ren $ho enter school $ith aggressi+e con!uct problems are at increase! risk for school mala!,ustment, in areas of aca!emic learning as $ell as social-beha+ioral a!,ustment.

!':LE 1 *e#cri.$i+e S$a$i#$ic# for Gou$" in $"e =i " (i#E an% Nor3a$i+e Sa3.le#

Pre!icti+e correlations !ocumenting the associations bet$een the school rea!iness measures collecte! at kin!ergarten entry &cogniti+e ability, rea!ing rea!iness, inattention, an! aggression) an! the later school outcomes are sho$n in #able II. Each of the kin!ergarten measures sho$e! robust pre!icti+e po$er, !espite the constraine! sample that inclu!e! only chil!ren at high-risk !ue to ele+ate! beha+ior problems.

!':LE 11 )orrela$ion# :e$&een Din%er ar$en '%,u#$3en$ an% 'ca%e3ic Ou$co3e#

Correlations compute! among the aca!emic outcome measures are presente! in #able III. =egati+e outcomes sho$e! mo!erate continuity across the elementary to secon!ary school years for all outcomes e-cept retention, $ith cross-time correlations ranging from r @ .89 for BPA to r @ .73 for 4eha+ior (isor!er classification, ps O 229. In a!!ition, ha+ing a 4eha+ior (isor!er classification $as significantly associate! $ith a selfcontaine! placement, both concurrently an! prospecti+ely, rs @ .7/ an! .88, but $as not associate! $ith gra!e retention or BPA.

!':LE 111 )orrela$ion# '3on 'ca%e3ic Ou$co3e#

"nter$ention %ffects

"ulti-le+el, hierarchical regression analyses &.au!enbush ; 4ryk, /22/) $ere use! to assess the impact of early-starting con!uct problems an! the pre+enti+e inter+ention on negati+e aca!emic outcomes e-perience! !uring the elementary an! secon!ary school years. #hese mo!els accounte! for the non-in!epen!ence of the !ata at the le+el of initial school placement &$here the inter+ention $as ran!omiDe! an! !eli+ere!), treating it as a :e+el / ran!om effect. Chil! se- an! race $ere inclu!e! as :e+el 9 co+ariates, an! site $as inclu!e! as a :e+el / co+ariate. Continuous responses, specifically the BPA outcomes, $ere assesse! using P.6C "InE( &'A') an! binary outcomes $ere assesse! using P.6C B:I""In. #o assess the impact of the 5ast #rack pre+ention program an! the impact of cogniti+e an! beha+ioral school rea!iness, the mo!els estimate! the uni%ue contributions of the chil!Ks kin!ergarten cogniti+e ability, rea!ing rea!iness, inattention, an! aggression, as $ell as the main effect for inter+ention on each of the school outcomes. .esults for the elementary school outcomes of BPA an! gra!e retention are sho$n in the top part of #able IP. Each of the school rea!iness factors stu!ie! ma!e significant uni%ue contributions to the pre!iction of elementary BPA, $ith ] in the .2>M.91 range, all p O . 229. BPA$as co!e! on a 7-point scale &e.g., A @ 3 to 5 @ 2), so chil!ren $ith the kin!ergarten characteristics ha! BPAs that $ere, on a+erage, .2>M.91 of a gra!e point higher &cogniti+e ability, rea!ing rea!iness) or lo$er &African American ethnicity, inattention, aggression) than chil!ren $ho !i! not share the kin!ergarten characteristic. ?hen elementary gra!e retention $as consi!ere! &right top si!e of #able IP), inattention $as the only kin!ergarten characteristic that ma!e uni%ue contributions. #he magnitu!e of impact $as notableI 32A of the se+erely inattenti+e chil!ren an! 81A of the mo!erately inattenti+e chil!ren e-perience! a gra!e retention bet$een Bra!es 9M3, compare! $ith only 9CA of the chil!ren $ithout attention problems, o!!s ratio @ /.03 for mo!erate inattention an! 3.29 for se+ere inattention, ps O .229. .esults for these t$o outcomes at the secon!ary school le+el are sho$n in the bottom part of #able 3. Chil! se- an! kin!ergarten cogniti+e ability $ere the only significant uni%ue pre!ictors of BPA in the secon!ary school years, $ith males recei+ing lo$er gra!es than females, ] @ .//, p O .229, an! lo$er Kin!ergarten cogniti+e ability pre!icting lo$er gra!es, ] @ .21, p O .29. Kin!ergarten aggression $as the only significant uni%ue pre!ictor of gra!e retention !uring the secon!ary school years ] @ .//, p O .27, o!!s ratio @ 9./7.

!':LE 1@ 1ni$ial )"il% )"arac$eri#$ic# an% 1n$er+en$ion Effec$# on -ra%e# an% -ra%e (e$en$ion

Inter+ention !i! not ha+e a significant effect on any of these school outcomes.

.esults for the elementary school outcomes of 4eha+ior (isor!er classification an! placement in a self-containe! special e!ucation classroom are sho$n in the top part of #able P. 4eha+ior (isor!er classification in elementary school $as more likely for boys, ] @ .17, p O .27, o!!s ratio @ /.83, an! for aggressi+e chil!ren, ] @ .10, p O .229, o!!s ratio @ /.33, but less likely for chil!ren $ith a!e%uate kin!ergarten rea!ing rea!iness, ] @ .39, p O .29, o!!s ratio @ .>>. Placement in a self-containe! classroom in elementary school $as also more likely for boys, ] @ 9.2C, p O .27, o!!s ratio @ /.0/, for aggressi+e chil!ren, ] @ .10, p O .229, o!!s ratio @ /.27, for chil!ren $ith mo!erate or se+ere inattention, ] @ 9.>3, 9./2, respecti+ely, p O .29, o!!s ratio @ 7.9> an! 8.8/, but less likely for chil!ren $ith a!e%uate kin!ergarten rea!ing rea!iness, ] @ .>9, p O .29, o!!s ratio @ .73.

!':LE @ 1ni$ial )"il% )"arac$eri#$ic# an% 1n$er+en$ion Effec$# on :e"a+ior *i#or%er2 Self;)on$aine% Place3en$2 an% =i " Sc"ool -ra%ua$ion

.esults for the secon!ary school outcomes of 4eha+ior (isor!er classification, placement in a self-containe! special e!ucation classroom, an! high school gra!uation are sho$n in the bottom part of #able 7. Kin!ergarten aggression $as the only uni%ue pre!ictor of 4eha+ior (isor!er classification in the secon!ary school years, ] @ .CC, p O .229, o!!s ratio @ /.9>. 'imilar to elementary school, placement in a self-containe! classroom in secon!ary school ha! multiple uni%ue pre!ictors, an! $as more likely for boys, ] @ .C0, p O .29, o!!s ratio @ /./2, for African American chil!ren, ] @ .CC, p O .27, o!!s ratio @ /.9>, for aggressi+e chil!ren, ] @ .>0, p O .229, o!!s ratio @ 9.00, for chil!ren $ith mo!erate or se+ere inattention, ] @ .>0, .C8, respecti+ely, p O .27, o!!s ratio @ 9.00 an! /.21, but less likely for chil!ren $ith a!e%uate kin!ergarten rea!ing rea!iness, ] @ M.30, p O .229, o!!s ratio @ .>9. 5inally, both rea!ing rea!iness &] @ ./3, p O .29, o!!s ratio @ 9./C) an! mo!erate inattention &] @ .37, ] O .27, o!!s ratio @ .>3) ma!e uni%ue contributions to the pre!iction of high school gra!uation, as !i! se-, as boys $ere less likely to gra!uate than girls, ] @ .3/, p O .27, o!!s ratio @ .>>. In neither the elementary nor secon!ary years, !i! any significant main effects emerge! for inter+ention on these outcomes.

D")C:))"'
#hese stu!y fin!ings replicate an! e-ten! !e+elopmental mo!els that i!entify multiple factors contributing to the school mala!,ustment of chil!ren $ho e-hibit high rates of aggressi+e-!isrupti+e beha+ior at school entry. #he aggressi+e-!isrupti+e chil!ren in this sample sho$e! significantly lo$er le+els of school rea!iness than a normati+e

comparison sample !ra$n from the same !isa!+antage! schools, inclu!ing lo$er le+els of cogniti+e ability, inattention, an! rea!ing rea!iness. #hey $ent on to e-perience significantly higher le+els of school mala!,ustment than their normati+e classmates, inclu!ing lo$er gra!es in both the elementary an! secon!ary school years, more fre%uent classification as 4eha+ior (isor!ere!, more fre%uent placement in self-containe! special e!ucation classrooms, an! lo$er rates of high school gra!uation. Cogniti+e factors &cogniti+e ability, inattention, rea!ing rea!iness) as $ell as initial aggression at school entry each ma!e uni%ue contributions to their later school mala!,ustment, !emonstrating the comple- an! multi-!etermine! nature of the school !ifficulties of young, earlystarting aggressi+e-!isrupti+e chil!ren. =otably, the se+erity of a chil!Ks early aggression an! !eficits in the +arious cogniti+e skills stu!ie! ha! !ifferent implications for !ifferent facets of school a!,ustment &e.g., BPA +s. self-containe! placement) at !ifferent !e+elopmental perio!s &the elementary +s. secon!ary school years). As !iscusse! belo$, these fin!ings may ha+e implications for un!erstan!ing the !e+elopmental course of the school !ifficulties e-perience! by young aggressors, as $ell as for the !esign of effecti+e pre+ention programs. Enfortunately, !espite e+i!ence that the 5ast #rack inter+ention impro+e! rea!ing attack skills, re!uce! aggression, an! impro+e! peer relations by the en! of first gra!e &CPP.B, 9000), the inter+ention !i! not ha+e a significant effect on any of these broa! an! important school outcomes. #he !e+elopmental an! pre+ention implications of the fin!ings are !iscusse! further in the follo$ing sections.

De$elopmental Processes /inking %arly Aggression 5ith )chool Malad(ustment


#his stu!y e-amine! the !egree to $hich +arious facets of the school rea!iness of aggressi+e-!isrupti+e chil!ren measure! at school entry pre!icte! their school mala!,ustment in the later elementary an! secon!ary school years. At school entry, correlations bet$een se+erity of aggression an! the three cogniti+e skills &cogniti+e ability, inattention, an! rea!ing rea!iness) $ere all statistically significant, but only small to mo!erate in magnitu!e. As e-pecte!, the kin!ergarten assessment of cogniti+e ability &?I'C-. Pocabulary an! 4lock (esign) $as correlate! $ith all of the later school outcomes. ?ith the other in!ices of school rea!iness controlle!, ho$e+er, kin!ergarten scores on these ?I'C-. subtests ma!e a significant uni%ue contribution only to BPA. #eacher-rate! inattention also ma!e a significant uni%ue contribution to the pre!iction of BPA !uring the elementary &but not the secon!ary) school years, an! inattention $as the only kin!ergarten skill that uni%uely pre!icte! gra!e retention in elementary school. In the secon!ary school years, cogniti+e ability $as the only kin!ergarten skill that uni%uely pre!icte! BPA. At neither gra!e le+el !i! cogniti+e ability nor inattention make uni%ue contributions to 4eha+ior (isor!ere! classification, but inattention !i! make uni%ue contributions to the likelihoo! of self-containe! classroom placement !uring both elementary an! secon!ary years. Apparently, cogniti+e ability uni%uely affects learning an! school performance &BPA). Controlling for cogniti+e ability, inattenti+e beha+ior increases the likelihoo! of special classroom placement, inclu!ing gra!e retention an!Hor self-containe! classroom placement.

In contrast, rea!ing rea!iness an! se+erity of initial aggression emerge! as the t$o uni%ue pre!ictors of classification as 4eha+ior (isor!ere! !uring the elementary years, an! also each increase! risk for placement in a self-containe! classroom. Kin!ergarten rea!ing rea!iness remaine! a significant uni%ue pre!ictor of placement in a self-containe! special e!ucation setting in the secon!ary schools, along $ith aggression an! inattention, although only se+erity of kin!ergarten aggression pre!icte! a 4eha+ior (isor!er classification in Bra!es CM92. E+en $ith the three cogniti+e skills controlle!, le+el of kin!ergarten aggression ma!e a uni%ue contribution to elementary school BPA an! to secon!ary gra!e retention, as $ell as uni%uely pre!icting 4eha+ior (isor!er classification an! self-containe! special e!ucation placement. #hese fin!ings are consistent $ith the mo!el propose! by "offitt &9008) an! the conclusions !ra$n by *insha$ &900/), in $hich neuropsychological !eficits, particularly inattention, play a critical role in the un!erachie+ement an! learning !ifficulties e-perience! by aggressi+e chil!ren. At the same time, the current fin!ings e-ten! this earlier research in t$o $ays. 5irst, they in!icate an important role for rea!ing rea!iness skills in pre!icting the future school !ifficulties of aggressi+e chil!ren, e+en after controlling for their cogniti+e ability an! attention problems. 'econ!, they in!icate that the se+erity of initial kin!ergarten aggression significantly contributes to multiple aspects of school !ifficulty &lo$ BPA in the elementary years, gra!e retention in the secon!ary years, self-containe! classroom placement) e+en $ith concurrent cogniti+e skills inclu!ing attention problems controlle!, an! it is the primary pre!ictor of special e!ucation classification as 4eha+ior (isor!ere!. #hese fin!ings are correlational, an! hence !o not support causal conclusions. *o$e+er, they are consistent $ith !e+elopmental mo!els that suggest that both cogniti+e an! beha+ioral processes contribute to the school mala!,ustment of aggressi+e chil!ren, an! that early e-periences prior to school entry contribute to the !elays in school rea!iness that set the stage for later school problems. "any of the family factors that increase risk for chil! aggressi+e-!isrupti+e beha+ior problems, inclu!ing financial har!ship, lo$ le+els of maternal e!ucation, single parenthoo!, an! maternal !epression also affect the %uality of parentMchil! +erbal interaction an! support. In turn, the fre%uency an! %uality of parentMchil! +erbal interaction !ramatically affects chil!renKs oral language skills an! early literacy !e+elopment, as !oes the %uality of language use in chil!-care an! preschool settings &*art ; .isley, 9007). :o$ le+els of language stimulation an! associate! !elays in language skill !e+elopment may be a particularly important precursor of later school !ifficulties, because of the important role that language skills play in rea!ing comprehension, un!erstan!ing information an! !irections pro+i!e! in class, an! !e+eloping an! maintaining positi+e relations $ith teachers an! peers &Catts, 5ey, Shang, ; #omblin, 9000). 5amily a!+ersity an! lo$-%uality stimulation may also !elay the !e+elopment of the e-ecuti+e function an! self-regulation skills that foster a!apti+e approaches to learning in school, inclu!ing the capacity to participate cooperati+ely in classroom acti+ities, follo$ teacher !irections, focus attention, an! sustain task in+ol+ement &4lair, SelaDo, ; Breenberg, /227< "cClellan!, Acock, ; "orrison, /22>). (elays in e-ecuti+e function

skill !e+elopment are reflecte! in !eficits in inhibitory control, planning, !ecisionmaking, an! social problem-sol+ing skills that fre%uently accompany aggressi+e beha+ior &'helley-#remblay et al., /22C). It is likely that both genetic an! en+ironmental e-periences contribute to !elays in the early !e+elopment of these cogniti+e skills, resulting in poor rea!iness an! +ulnerability to mala!,ustment at school entry. Although the results of the present stu!y +ali!ate the important role of early !elays in these cogniti+e skills for school functioning, they also in!icate an in!epen!ent pre!icti+e role for the impact of the se+erity of a chil!Ks aggressi+e-!isrupti+e beha+ior problems on later school outcomes. =otably, e+en in this sample that $as selecte! for ele+ate! aggressi+e-!isrupti+e beha+iors &an! thus constricte! in !istribution), the se+erity of those beha+iors ma!e uni%ue contributions to chil!renKs school outcomes. #he 5ast #rack inter+ention !esign $as base! upon a near-consensual mo!el of the casca!ing negati+e impact that aggressi+e beha+iors ha+e on interpersonal relations $ith parents, peers, an! teachers, an! the correspon!ing !isruption it causes in the !e+elopment of prosocial an! a!apti+e school beha+iors &CPP.B, 900/). In this !e+elopmental mo!el, temperament an! early e-periences contribute to the !e+elopment of beha+iors an! social-cognitions that pre!ispose some chil!ren to react against an! resist a!ult authority, !efy con+entions, an! engage in a range of rule-breaking acti+ities &E!ry, Ko+enock, "orris, ; +an !en 4erg, 9007< ?illoughby, Chalmers, ; 4usseri, /223). #hese rebellious ten!encies are strengthene! o+er time by responses of peers an! parents, $ho mo!el counter-aggression, an! either positi+ely re$ar! &by laughing or gi+ing in) or negati+ely reinforce &by re!ucing !eman!s) aggressi+e an! rule-breaking beha+iors &see also (o!ge, 4ates, ; Pettit, 9002< (ishion, 5rench, ; Patterson, 900C). An interesting fin!ing of the present stu!y in+ol+e! the magnification of school problems obser+e! in the secon!ary school years &$hen rates of school problems !ouble!), pre!icte! by the se+erity of early aggression. #his fin!ing is consistent $ith the pre!ictions of problem beha+ior theory, $hich suggest that early rebelliousness sub+erts social controls an! un!ermines the socialiDation efforts of parents an! teachers, escalating into serious problem beha+iors at the transition to a!olescence $hen youth gain greater autonomy an! mobility &E!ry et al., 9007).

Decoupling Aggressi$e Beha$ior and )chool )uccess


Although the !e+elopmental course of aggressi+e beha+ior correlates $ith the !e+elopmental course of school success, the fin!ings reporte! here in!icate that the 5ast #rack inter+ention that impro+e! aggressi+e beha+ior !i! not impro+e long-term school success. .an!om assignment to the inter+ention !i! lea! to lo$er self-reporte! serious antisocial beha+ior in high school &CPP.B, /22C) an! less in+ol+ement in ,u+enile arrests &specifically, inter+ention !elaye! the onset of arrest, !ecrease! the probability of arrest for a se+ere &le+el 3H7) crime, an! lo$ere! the ,u+enile arrest se+erity in!e- score, CPP.B, /292a). Apparently, impro+ements in antisocial beha+ior !o not ensure impro+ements in aca!emic outcomes.

,he Challenges of %ffecti$e Pre$ention

4y the en! of first gra!e, the 5ast #rack pre+ention program ha! successfully mo!ifie! the key factors implemente! in the negati+e casca!e mo!el of school mala!,ustment. #hat is, chil!ren in the pre+ention group ha! higher rea!ing achie+ement scores an! higher language arts gra!es, more positi+e social skills an! peer relations, an! re!uce! rates of aggressi+e-!isrupti+e beha+ior at school than chil!ren in the control group. In a!!ition, the inter+ention inclu!e! the on-going !eli+ery of aca!emic support ser+ices through tenth gra!e, albeit at re!uce! le+els an! only for chil!ren assesse! $ith aca!emic nee!s. *ence, it $as !isappointing to fin! that the early aca!emic gains !issipate! o+er time an! faile! to impact the longer-term school a!,ustment of the participating chil!ren. #he !issipation of aca!emic gains $as e+i!ent %uite early, as inter+ention effects on rea!ing achie+ement $ere no longer significant at the en! of thir! gra!e &CPP.B, /22/). *o$e+er, re!uctions in aggressi+e-!isrupti+e beha+ior at school remaine! significant through the early elementary years &CPP.B, /22/), yet faile! to significantly impact the broa! school outcomes assesse! here. #his raises %uestions regar!ing $hy the inter+ention $as not more successful at changing aca!emic outcomes, an! ho$ future pre+ention stu!ies might be mo!ifie! to strengthen their impact on aca!emic outcomes. 6ne possible reason for the lack of impact on the later aca!emic outcomes is that the kin! of aca!emic support that $as pro+i!e! by 5ast #rack $as not effecti+e, or alternati+ely, that it $as pro+i!e! at too lo$ a !ose or intensity to be effecti+e. In the first year of aca!emic tutoring &an! until chil!ren mastere! the skills), 5ast #rack use! the e+i!encebase! ?allach tutoring program &?allach ; ?allach, 90C>). *o$e+er, after chil!ren mastere! that program, 5ast #rack focuse! tutoring on the areas of nee! suggeste! by classroom teachers. 6ften, this inclu!e! an emphasis on helping $ith home$ork completion. Prior stu!ies suggest that chil!ren $ith rea!ing !isabilities benefit most from intensi+e tutoring, utiliDing $ell-informe! an! e+i!ence-base! programs that incorporate a systematic progression of skills an! teaching techni%ues, le! by certifie! teachers &?asik ; 'la+in, 9008). #he $eekly or bi-$eekly home$ork support by paraprofessionals that 5ast #rack pro+i!e! may ha+e been too little an! too unfocuse! to reme!iate the core cogniti+e !eficits un!ermining the school mala!,ustment of many of the high-risk youth. In a!!ition, careful analyses of the 5ast #rack tutoring program in the early school years suggeste! that it $as most effecti+e for chil!ren $ho ha! rea!ing skill !elays $ithout concurrent attention problems< the treatment effect siDe $as much smaller for chil!ren $ith concurrent attention !eficits &.abiner, "alone, ; CPP.B, /223). Impro+ing the achie+ement of chil!ren $ith attention problems is a ubi%uitous challenge, an! the !ifficulty of !oing so suggests that the 5ast #rack program may ha+e been insufficient to meet the significant aca!emic nee!s of the appro-imately 82A of the sample $ith clinically ele+ate! attention !eficits. A!!itional or alternati+e inter+ention approaches are likely nee!e! to a!!ress the attention problems of this sub-group of aggressi+e chil!ren, focuse! on promoting their cogniti+e !e+elopment an! processing, although it is not yet clear $hich inter+ention approaches $ill $ork. Among chil!ren $ith A(*(, longitu!inal stu!ies ha+e suggeste! that early an! sustaine! me!ication may impro+e aca!emic achie+ement &'cheffler et al., /220), but rigorous, ran!omiDe! trials ha+e not !ocumente! positi+e me!ication effects on achie+ement &"olina et al., /221), making it

unclear $hether me!ication coul! foster impro+ements in school success among aggressi+e chil!ren $ith se+ere, comorbi! attention !eficits. In recent years, a gro$ing area of research has focuse! on early chil!hoo! inter+entions !esigne! to foster the !e+elopment of e-ecuti+e function skills &e.g., $orking memory, attention control) &for re+ie$s, see 4ierman ; #orres, in press< (iamon! ; :ee, /299). #he goal of these early inter+entions is to promote gro$th in attention control !uring the !e+elopmental perio! $hen the prefrontal corte- is un!ergoing rapi! gro$th. Clearly, these are areas of pre+ention research that are important to pursue, $ith a particular focus on their utility $ith chil!ren $ho ha+e the combine! problems of serious attention !eficits an! aggressi+e beha+ior. 6ne of the unfortunate conse%uences of many current school-base! inter+entions !esigne! to help stu!ents $ho enter school unprepare! for the aca!emic an! social !eman!s in+ol+es uninten!e! an! negati+e long-term effects. 5or e-ample, Jimerson an! 5erguson &/22C) e-amine! the impact of elementary school gra!e retention or placement in a Ftransition K-9G program to pro+i!e chil!ren $ith more time to mature an! a!,ust to school !eman!s. Controlling for initial le+els of aggression, stu!ents $ho $ere recommen!e! for the transition classroom, but promote!, !isplaye! lo$er aggression in eighth gra!e, compare! to both groups of retaine! stu!ents. #heir research an! other stu!ies suggest that gra!e retention can ha+e short-term gains, but increase risk for aggressi+e beha+ior an! high school !rop-out in later years. 'imilarly, causal mo!eling of the 5ast #rack !ata using propensity score metho!s suggests that placement in a selfcontaine! classroom in the secon!ary school years increase! the likelihoo! that youth $oul! en! school $ith a con!uct !isor!er !iagnosis &Po$ers, Coffman, ; 4ierman, /299). #he reasons for these a!+erse effects are not clear, but they may reflect peer contagion effects associate! $ith placements that increase interactions $ith other highrisk aggressi+e youth or changes in teacher perceptions, e-pectations, an! treatment once youth ha+e been i!entifie! as in nee! of retention or other special e!ucation ser+ices &(ishion et al., /221). It is possible that the positi+e early effects of 5ast #rack $ere un!ermine! in later years, as the inter+ention may ha+e ina!+ertently increase! the +isibility of the participating stu!entsZ aca!emic an! beha+ioral !eficiencies, lea!ing to re!uce! aca!emic e-pectations or biase! attributions on the part of teachers an! school personnel regar!ing their capacity to learn. It is also possible that the timing of pre+enti+e inter+entions re%uires more careful consi!eration. 5ast #rack pro+i!e! the most intensi+e set of ser+ices at school entry, base! upon the hope that a goo! start in school $oul! set chil!ren on a positi+e path that re!uce! the negati+e !e+elopmental casca!e so often associate! $ith aggressi+e beha+ior. (ue to a concern about possible iatrogenic effects of group inter+ention at the transition into a!olescence &(ishion et al., /221), 5ast #rack shifte! to a primary reliance on in!i+i!ualiDe! ser+ices !uring the mi!!le school years. #he current !ata suggest that the transition into mi!!le school is a particularly challenging time for aggressi+e chil!ren< in this high-risk sample, gra!es plummete! an! rates of 4eha+ior (isor!er classification an! self-containe! placements !ouble! post-transition. 'o, although the intensi+e 5ast #rack pre+ention efforts at school entry pro!uce! a number of positi+e an! significant effects, they !i! not buffer high-risk chil!ren sufficiently from the challenges

of the mi!!le school e-perience an! transition into a!olescence &see also CPP.B, /292b). .esearch is nee!e! to i!entify effecti+e pre+enti+e inter+entions that focus on aggressi+e youth !uring the prea!olescent an! early a!olescent years as they prepare for an! make this critical transition. 5inally, although 5ast #rack employe! an a!apti+e inter+ention !esign in or!er to in!i+i!ualiDe the pro+ision of aca!emic inter+ention ser+ices, this aspect of the !esign $as not highly structure!. ECs met $ith classroom teachers to !iscuss stu!entsZ nee!s for inter+ention. A more elaborate! system for e+aluating stu!entsZ aca!emic nee!s an! linking them $ith e+i!ence-base! aca!emic support components might ha+e strengthene! the inter+ention impact &see Collins, "urphy, ; 4ierman, /223 for more !iscussion of this issue).

/imitations and *uture Directions


5ast #rack $as !esigne! primarily to apply psychosocial inter+entions to promote the early competencies of aggressi+e-!isrupti+e chil!ren, an! thereby re!uce their risk for long-term antisocial outcomes. .ecent analyses !emonstrate that 5ast #rack $as effecti+e in re!ucing antisocial outcomes, promoting significant re!uctions in ,u+enile arrests &CPP.B, /292a) an!, among the most aggressi+e youth, re!ucing a!olescent !iagnoses of con!uct !isor!er &CPP.B, /299). *o$e+er, simply re!ucing aggression !i! not result in impro+e! school outcomes, nor !i! the aca!emic tutoring pro+i!e! by 5ast #rack result in long-term gains. #hese are sobering fin!ings, an! $orth consi!ering in terms of the Flessons learne!G for future pre+ention efforts. It shoul! be note! that the mo!els run here focus solely on pre!iction from chil!renKs kin!ergarten rea!iness scores an! 5ast #rack inter+ention, an! !o not take into account the multitu!e of a!!itional factors that influence! chil!renKs progress through school. #he 5ast #rack !ata !o not inclu!e information about the %uality of e!ucational ser+ices that chil!ren recei+e! $ithin their schools. `uite likely, the %uality of classroom teaching, reme!ial support, an! special e!ucation ser+ices +arie! consi!erably across the schools stu!ie! here, an! ha! an impact on youth outcomes in $ays that are not represente! in this stu!y. In a!!ition, this $as not a normati+e sample< all of the chil!ren in the highrisk sample stu!ie! here ha! ele+ate! aggression at school entry. ?hile a focus on this high-risk sample likely attenuate! some relations that $oul! be e+i!ent in a sample that inclu!e! non-aggressi+e chil!ren, the ability to mo!el !i+ersity in the aca!emic outcomes of a large sample of aggressi+e chil!ren allo$e! for a closer e-amination of heterogeneity in cogniti+e skill le+els $ithin the high-risk group, pro+i!ing a more nuance! un!erstan!ing of the uni%ue contributions of particular skill !eficits to the range of school outcomes e-perience! by chil!ren $ho share! the early risk of ele+ate! aggression. In a!!ition, the schools stu!ie! here $ere not typical schools, but rather schools selecte! because they $ere in high-crime neighborhoo!s &or, in the case of rural Pennsyl+ania, because they $ere in small to$ns characteriDe! by economic !isa!+antage.) #he high school gra!uation rates of the normati+e sample is lo$er than the national a+erage,

reflecting the general !isa!+antage an! risk of the stu!ent bo!y for aca!emic !ifficulties. Possibly, the 5ast #rack inter+ention $as not sufficient to o+ercome the en+ironmental !isa!+antages face! by youth in these schools, an! it is possible that the results may ha+e been !ifferent in !ifferent school or neighborhoo! conte-ts. #he fin!ings reporte! here raise as many %uestions as they ans$er, particularly in terms of the implications for pre+enti+e inter+entions. #hey un!erscore the comple-ity of the factors contributing to the poor school outcomes of early-starting aggressi+e chil!ren, an! the significant challenge associate! $ith pre+enting their school mala!,ustment.

ACK '5/%D+M% ,)
#he =ational Institute of (rug Abuse an! the Center for 'ubstance Abuse Pre+ention pro+i!e! support for 5ast #rack through a memoran!um of agreement $ith the =ational Institute of "ental *ealth &=I"*). "ark Breenberg recei+es financial compensation associate! $ith royalties for the PA#*' Curriculum, an! the 5ast #rack inter+ention manuals authore! by the CPP.B are un!er contract for publication. 5or a!!itional information concerning 5ast #rack see $$$.fasttrackpro,ect.org. Contract grant sponsorI #he =ational Institute of "ental *ealth &=I"*)< Contract grant numbersI .91"*31328, .91"*72079, .91"*7207/, .91"*72078, K27"*22C0C, K27(A97//>< Contract grant sponsorI #he =ational Institute of (rug Abuse an! the Center for 'ubstance Abuse Pre+ention &=I(A)< Contract grant numbersI (A9>028, (A29C710, (A297//>< Contract grant sponsorI (epartment of E!ucation< Contract grant numberI '913E8222/.

Article information
' re## :e"a+. 'u$"or 3anu#cri.$4 a+ailable in P0) 2013 'u u#$ 1. Publi#"e% in final e%i$e% for3 a#7 ' re## :e"a+. 2013 0ar;'.r4 395267 114H130. Publi#"e% online 2013 February 5. %oi7 10.10028ab.21467 P0)1*7 P0)3730283 N1=0S1*7 N1=0S498118 Daren L. :ier3an2129 <o"n )oie22 Denne$" *o% e23 0arE -reenber 24 <o"n Loc"3an25 (ober$ 0c0o"an26 Ellen Pin%er"u "e#27 an% )on%uc$ Proble3# Pre+en$ion (e#earc" -rou.a 1 *e.ar$3en$ of P#yc"olo y2 Penn#yl+ania S$a$e /ni+er#i$y2 /ni+er#i$y ParE2 Penn#yl+ania 2 *e.ar$3en$ of P#yc"olo y2 *uEe /ni+er#i$y2 *ur"a32 Nor$" )arolina 3 Public Policy S$u%ie#2 *uEe /ni+er#i$y2 *ur"a32 Nor$" )arolina 4 =u3an *e+elo.3en$ an% Fa3ily S$u%ie#2 Penn#yl+ania S$a$e /ni+er#i$y2 /ni+er#i$y ParE2 Penn#yl+ania 5 *e.ar$3en$ of P#yc"olo y2 /ni+er#i$y of 'laba3a2 !u#caloo#a2 'laba3a 6 *e.ar$3en$ of P#yc"olo y2 Si3on Fra#er /ni+er#i$y2 :urnaby2 :ri$i#" )olu3bia2 )ana%a

Elio$;Pear#on *e.ar$3en$ of )"il% *e+elo.3en$2 !uf$# /ni+er#i$y2 0e%for%2 0a##ac"u#e$$# a 0e3ber# of $"e )PP(- in al."abe$ical or%er inclu%e Daren L. :ier3an 5Penn#yl+ania S$a$e /ni+er#i$y62 <o"n *. )oie 5*uEe /ni+er#i$y62 Denne$" '. *o% e 5*uEe /ni+er#i$y62 0arE !. -reenber 5Penn#yl+ania S$a$e /ni+er#i$y62 <o"n E. Loc"3an 5/ni+er#i$y of 'laba3a62 (ober$ <. 0c0a"on 5Si3on Fra#er /ni+er#i$y62 an% Ellen Pin%er"u "e# 5!uf$# /ni+er#i$y6. 9 )orre#.on%ence $o7 Daren L. :ier3an2 !"e Penn#yl+ania S$a$e /ni+er#i$y2 110 0oore :uil%in 2 /ni+er#i$y ParE2 P' 16802. E3ail7 Eb28a$8.#u.e%u )o.yri "$ no$ice an% *i#clai3er !"e .ubli#"erV# final e%i$e% +er#ion of $"i# ar$icle i# a+ailable a$ ' re## :e"a+

R%*%R% C%)
1. 'c"enbac" !0. 0anual for $"e )"il% :e"a+ior )"ecEli#$7 4;18 an% 1991 .rofile. /ni+er#i$y of @er3on$ *e.ar$3en$ of P#yc"ia$ry4 :urlin $on2 @!7 1991. 2. :ier3an DL2 -reenber 0!2 )PP(- . Social #Eill# $rainin in $"e Fa#$ !racE .ro ra3. 1n7 *e+. Pe$er# (2 0c0a"on (<2 e%i$or#. Pre+en$in c"il%"oo% %i#or%er#2 #ub#$ance abu#e2 an% %elinNuency. Sa e4 !"ou#an% oaE#2 )'7 1996. ... 65H89. 3. :ier3an DL2 !orre# 0. Pro3o$in $"e %e+elo.3en$ of eFecu$i+e func$ion# $"rou " early e%uca$ion an% .re+en$ion .ro ra3#. 1n7 -riffin <'2 Freun% LS2 0c)ar%le P2 e%i$or#. EFecu$i+e func$ion in .re#c"ool a e c"il%ren7 1n$e ra$in 3ea#ure3en$2 neuro%e+elo.3en$ an% $ran#la$ional re#earc". '3erican P#yc"olo ical '##ocia$ion4 Aa#"in $on2 *)7 in .re##. 4. :lair )2 Kela?o P*2 -reenber 0!. !"e 3ea#ure3en$ of eFecu$i+e func$ion in early c"il%"oo%. *e+elo.3en$al Neuro.#yc"olo y. 20054287561H571. IPub0e%J 5. )a$$# =A2 Fey 0E2 K"an O2 !o3blin <:. Lan ua e ba#i# of rea%in an% rea%in %i#abili$ie#7 E+i%ence fro3 a lon i$u%inal in+e#$i a$ion. Scien$ific S$u%ie# of (ea%in . 1999437331H361. 6. )"en O2 =uan O2 )"an L2 Aan L2 Li *. ' re##ion2 #ocial co3.e$ence2 an% aca%e3ic ac"ie+e3en$ in )"ine#e c"il%ren7 ' 5;year lon i$u%inal #$u%y. *e+elo.3en$ an% P#yc"o.a$"olo y. 20104227583H 592. IPub0e%J 7. )oie <*2 *o% e D'. ' re##ion an% an$i#ocial be"a+ior. 1n7 Ei#enber N2 *a3on A2 e%i$or#. =an%booE of c"il% .#yc"olo y2 @ol. 37 Social2 e3o$ional an% .er#onali$y %e+elo.3en$. 5$" e%i$ion Ailey4 Ne& GorE7 1998. ... 779H862. 8. )oie <*2 Dre"biel -. Effec$# of aca%e3ic $u$orin on $"e #ocial #$a$u# of lo&;ac"ie+in 2 #ocially re,ec$e% c"il%ren. )"il% *e+elo.3en$. 198445571465H1478. 9. )ollin# L02 0ur."y S'2 :ier3an DL. ' conce.$ual fra3e&orE for a%a.$i+e .re+en$i+e in$er+en$ion#. Pre+en$ion Science. 2004457185H 196. IP0) free ar$icleJ IPub0e%J

10.)PP(- ' %e+elo.3en$al an% clinical 3o%el for $"e .re+en$ion of con%uc$ %i#or%er#7 !"e F'S! !racE Pro ra3. *e+elo.3en$ an% P#yc"o.a$"olo y. 1992447505H527. 11.)PP(- 1ni$ial i3.ac$ of $"e Fa#$ !racE .re+en$ion $rial for con%uc$ .roble3#7 1. !"e "i ";ri#E #a3.le. <ournal of )on#ul$in an% )linical P#yc"olo y. 19994677631H647. IP0) free ar$icleJ IPub0e%J 12.)PP(- E+alua$ion of $"e fir#$ $"ree year# of $"e Fa#$ !racE Pre+en$ion !rial &i$" c"il%ren a$ "i " ri#E for a%ole#cen$ con%uc$ .roble3#. <ournal of 'bnor3al )"il% P#yc"olo y. 2002430719H35. IPub0e%J 13.)PP(- !"e Fa#$ !racE ran%o3i?e% con$rolle% $rial $o .re+en$ eF$ernali?in .#yc"ia$ric %i#or%er#7 Fin%in # fro3 ra%e# 3 $o 9. <ournal of $"e '3erican 'ca%e3y of )"il% an% '%ole#cen$ P#yc"ia$ry. 2007446510671250H1262. IP0) free ar$icleJ IPub0e%J 14.)PP(- Fa#$ !racE in$er+en$ion effec$# on you$" arre#$# an% %elinNuency. <ournal of EF.eri3en$al )ri3inolo y. 2010a467131H157. IP0) free ar$icleJ IPub0e%J 15.)PP(- !"e %ifficul$y of 3ain$ainin .o#i$i+e in$er+en$ion effec$#7 ' looE a$ %i#ru.$i+e be"a+ior2 %e+ian$ .eer rela$ion#2 an% #ocial #Eill# %urin $"e 3i%%le #c"ool year#. !"e <ournal of Early '%ole#cence. 2010b4307593H624. 16.)PP(- !"e effec$# of $"e Fa#$ !racE .re+en$i+e in$er+en$ion on $"e %e+elo.3en$ of con%uc$ %i#or%er acro## c"il%"oo%. )"il% *e+elo.3en$. 20114827331H345. IP0) free ar$icleJ IPub0e%J 17.*ia3on% '2 Lee D. 1n$er+en$ion# #"o&n $o ai% eFecu$i+e func$ion %e+elo.3en$ in c"il%ren 4 $o 12 year# ol%. Science. 201143337959H964. IP0) free ar$icleJ IPub0e%J 18.*i#"ion !2 Frenc" *2 Pa$$er#on -. !"e %e+elo.3en$ an% ecolo y of an$i#ocial be"a+ior. 1n7 )icc"e$$i *2 )o"en *2 e%i$or#. 0anual of %e+elo.3en$al .#yc"o.a$"olo y. Ailey4 Ne& GorE7 1997. ... 421H471. 19.*i#"ion !<2 Pie"ler !F2 0yer# 0A. *yna3ic# an% ecolo y of a%ole#cen$ .eer con$a ion. 1n7 Prin#$ein 0<2 *o% e D'2 e%i$or#. Peer influence .roce##e# a3on you$". -uilfor%4 Ne& GorE7 2008. ... 72H93. 20.*o% e D'2 :a$e# <E2 Pe$$i$ -S. 0ec"ani#3# in $"e cycle of +iolence. Science. 1990425071678H1683. IPub0e%J 21.*o% e D'2 -reenber 0!2 0alone PS2 )PP(- !e#$in an i%eali?e% %yna3ic ca#ca%e 3o%el of $"e %e+elo.3en$ of #eriou# +iolence in a%ole#cence. )"il% *e+elo.3en$. 200847971907H1927. IP0) free ar$icleJ IPub0e%J 22.*urlaE <'2 Aei##ber (P2 *y3nicEi ':2 !aylor (*2 Sc"ellin er D:. !"e i3.ac$ of en"ancin #$u%en$#B #ocial an% e3o$ional learnin 7 ' 3e$a; analy#i# of #c"ool;ba#e% uni+er#al in$er+en$ion#. )"il% *e+elo.3en$. 20114827405H432. IPub0e%J 23.-reenber 0!2 Du#c"e )'. Pro3o$in #ocial an% e3o$ional %e+elo.3en$ in %eaf c"il%ren7 !"e P'!=S Pro,ec$. /ni+er#i$y of Aa#"in $on Pre##4 Sea$$le2 A'7 1993. 24.-reenber 0!2 Aei##ber (P2 OV:rien 0/2 Kin# <E2 Fre%ericE# L2 (e#niE =2 Elia# 0<. En"ancin #c"ool;ba#e% .re+en$ion an% you$" %e+elo.3en$ $"rou " coor%ina$e% #ocial2 e3o$ional2 an% aca%e3ic learnin . '3erican P#yc"olo i#$. 20034587466H474. IPub0e%J

25.=ar$ :2 (i#ley !(. 0eanin ful %ifference# in $"e e+ery%ay eF.erience of youn '3erican c"il%ren. :rooEe#4 :al$i3ore7 1995. 26.=a&Ein# <*2 )a$alano (F2 Do#$er3an (2 'bbo$$ (2 =ill D-. Pre+en$in a%ole#cen$ "eal$";ri#E be"a+ior# by #$ren $"enin .ro$ec$ion %urin c"il%"oo%. 'rc"i+e# of Pe%ia$ric# C '%ole#cen$ 0e%icine. 199941537226H234. IPub0e%J 27.=a&Ein# <*2 )a$alano (F2 0iller <G. (i#E an% .ro$ec$i+e fac$or# for alco"ol an% o$"er %ru .roble3# in a%ole#cence an% early a%ul$"oo%7 13.lica$ion# for #ub#$ance abu#e .re+en$ion. P#yc"olo ical :ulle$in. 19924112764H105. IPub0e%J 28.=in#"a& SP. EF$ernali?in be"a+ior .roble3# an% aca%e3ic un%erac"ie+e3en$ in c"il%"oo% an% a%ole#cence7 )au#al rela$ion#"i.# an% un%erlyin 3ec"ani#3#. P#yc"olo ical :ulle$in. 199241117127H 155. IPub0e%J 29.=ollin #"ea% ''. ' four fac$or in%eF of #ocial #$a$u#. /n.ubli#"e% 3anu#cri.$. Gale /ni+er#i$y4 Ne& =a+en2 )!7 1975. 30.=ue#3ann L(2 Eron L*2 Gar3el PA. 1n$ellec$ual func$ionin an% a re##ion. <ournal of Per#onali$y an% Social P#yc"olo y. 19874527232H 240. IPub0e%J 31.<i3er#on S(2 Fer u#on P. ' lon i$u%inal #$u%y of ra%e re$en$ion7 'ca%e3ic an% be"a+ioral ou$co3e# of re$aine% #$u%en$# $"rou " a%ole#cence. Sc"ool P#yc"olo y >uar$erly. 20074227314H339. 32.DoEEo D2 PulEEinen L. ' re##ion in c"il%"oo% an% lon ;$er3 une3.loy3en$ in a%ul$"oo%7 ' cycle of 3ala%a.$a$ion an% #o3e .ro$ec$i+e fac$or#. *e+elo.3en$al P#yc"olo y. 20004367463H472. IPub0e%J 33.Loc"3an <E2 )on%uc$ Proble3# Pre+en$ion (e#earc" -rou. Screenin of c"il% be"a+ior .roble3# for .re+en$ion .ro ra3# a$ #c"ool en$ry. <ournal of )on#ul$in an% )linical P#yc"olo y. 19954637549H559. IPub0e%J 34.0a uin E2 Loeber (2 Le0a"ieu P-. *oe# $"e rela$ion#"i. be$&een .oor rea%in an% %elinNuency "ol% for 3ale# of %ifferen$ a e# an% e$"nic rou.#P <ournal of E3o$ional an% :e"a+ioral *i#or%er#. 199341788H100. 35.0c)lellan% 002 'cocE ')2 0orri#on F<. !"e i3.ac$ of Ein%er ar$en learnin ; rela$e% #Eill# on aca%e3ic $ra,ec$orie# a$ $"e en% of ele3en$ary #c"ool. Early )"il%"oo% (e#earc" >uar$erly. 20064217471H 490. 36.0c0a"on (<2 Slou " N2 )on%uc$ Proble3# Pre+en$ion (e#earc" -rou. . Fa3ily;ba#e% in$er+en$ion in $"e Fa#$ !racE .ro ra3. 1n7 Pe$er# (*2 0c0a"on (<2 e%i$or#. Pre+en$in c"il%"oo% %i#or%er#2 #ub#$ance abu#e2 an% %elinNuency. Sa e4 !"ou#an% OaE#2 )'7 1996. ... 90H110. 37.0e$ro.oli$an 'rea )"il% S$u%y (e#earc" -rou.. Eron L*2 =ue#3ann L(2 S.in%ler '2 -uerra N-2 =enry *2 @an'cEer (. ' co ni$i+e8ecolo ical a..roac" $o .re+en$in a re##ion in urban #e$$in #7 1ni$ial ou$co3e# for "i " ri#E c"il%ren. <ournal of )on#ul$in an% )linical P#yc"olo y. 20024705167179H194. IPub0e%J 38.0offi$$ !E. '%ole#cence;li3i$e% an% life;cour#e;.er#i#$en$ an$i#ocial be"a+ior7 ' %e+elo.3en$ $aFono3y. P#yc"olo ical (e+ie&. 199341007674H701. IPub0e%J

39.Po&er# )<2 )off3an *2 :ier3an D. ' re##i+e %i#ru.$i+e #$u%en$# in re#$ric$e% e%uca$ional .lace3en$#7 0o%era$e% influence of .lace3en$ "i#$ory on a%ole#cen$ ou$co3e#.. Po#$er .re#en$e% a$ $"e annual 3ee$in of $"e Socie$y for Pre+en$ion (e#earc"4 Aa#"in $on2 *). 0ay2 2011. 40.(au%enbu#" SA2 :ryE 'S. =ierarc"ical linear 3o%el#7 '..lica$ion# an% %a$a analy#i# 3e$"o%#. 2n% e%. Sa e4 Ne&bury ParE2 )'7 2002. 41.(abiner *L2 0alone PS2 )PP(- !"e i3.ac$ of $u$orin on early rea%in ac"ie+e3en$ for c"il%ren &i$" an% &i$"ou$ a$$en$ion .roble3#. <ournal of 'bnor3al )"il% P#yc"olo y. 20044327273H284. IP0) free ar$icleJ IPub0e%J 42.Sa$$ler <0. '##e##3en$ of )"il%ren. 'u$"or4 San *ie o2 )'7 1992. 43.Sc"afer <L. NO(07 0ul$i.le i3.u$a$ion of inco3.le$e 3ul$i+aria$e %a$a un%er a nor3al 3o%el2 +er#ion 2 #of$&are for Ain%o&# 958988N! Ion; lineJ 1999 '+ailable7 "$$.788&&&.#$a$..#u.e%u8,l#83i#of$&a."$3l. 44.Sc"effler (02 :ro&n !!2 Ful$on :*2 =in#"a& SP2 Le+ine P2 S$one S. Po#i$i+e a##ocia$ion be$&een a$$en$ion;%efici$8 "y.erac$i+i$y %i#or%er 3e%ica$ion u#e an% aca%e3ic ac"ie+e3en$ %urin ele3en$ary #c"ool. Pe%ia$ric#. 2009412371273H1279. IPub0e%J 45.S"elley;!re3blay <2 OV:rien N2 Lan "inric"#en;(o"lin <. (ea%in %i#abili$y in a%,u%ica$e% you$"7 Pre+alence ra$e#2 curren$ 3o%el#2 $ra%i$ional an% inno+a$i+e $rea$3en$#. ' re##ion an% @iolen$ :e"a+ior. 20074127376H392. 46.S$i.eE *2 0ile# S. Effec$# of a re##ion on ac"ie+e3en$7 *oe# conflic$ &i$" $"e $eac"er 3aEe i$ &or#eP )"il% *e+elo.3en$. 200847971721H 1735. IPub0e%J 47./%ry <(2 Do+enocE <2 0orri# N02 +an %en :er :<. )"il%"oo% .recur#or# of a e a$ fir#$ in$ercour#e for fe3ale#. 'rc"i+e# of SeFual :e"a+ior. 19954247329H337. IPub0e%J 48.@i##er 02 Dunnen SE2 @an -eer$ PL). !"e i3.ac$ of con$eF$ on $"e %e+elo.3en$ of a re##i+e be"a+ior in #.ecial ele3en$ary #c"ool c"il%ren. 0in%2 :rain2 an% E%uca$ion. 201044734H43. 49.AalEer =. !"e ac$in ;ou$ c"il%7 )o.in &i$" cla##roo3 %i#ru.$ion. 2r% e%. So.ri#4 Lon 3on$2 )O7 1996. 50.Aallac" 0'2 Aallac" L. !eac"in all c"il%ren $o rea%. /ni+er#i$y of )"ica o Pre##4 )"ica o7 1976. 51.Aa#iE :'2 Sla+in (E. Pre+en$in early rea%in failure &i$" one;$o;one $u$orin 7 ' re+ie& of fi+e .ro ra3#. (ea%in (e#earc" >uar$erly. 19934287179H200. 52.Aec"#ler *. Aec"#ler Pre#c"ool an% Pri3ary Scale of 1n$elli enceW111. P#yc"olo ical )or.ora$ion4 San 'n$onio2 !O7 2002. 53.Aer$"a3er;Lar##on L2 Della3 S-2 A"eeler L. Effec$# of fir#$ ra%e cla##roo3 en+iron3en$ on #"y be"a+ior2 a re##i+e be"a+ior2 an% concen$ra$ion .roble3#. '3erican <ournal of )o33uni$y P#yc"ololo y. 19914197585H602. IPub0e%J 54.A"i$e N'2 Loeber (. :ullyin an% #.ecial e%uca$ion a# .re%ic$or# of #eriou# %elinNuency. <ournal of (e#earc" in )ri3e an% *elinNuency. 20084457380H397. 55.Aillou "by !2 )"al3er# =2 :u##eri 0'. A"ere i# $"e #yn%ro3eP EFa3inin co;occurrence a3on 3ul$i.le .roble3 be"a+ior# in

a%ole#cence. <ournal of )on#ul$in an% )linical P#yc"olo y. 200447271022H1037. IPub0e%J 56.Ain%le ()2 Ain%le 0. Lon i$u%inal .a$$ern# of ."y#ical a re##ion7 '##ocia$ion# &i$" a%ul$ #ocial2 .#yc"ia$ric2 an% .er#onali$y func$ionin an% $e#$o#$erone le+el#. *e+elo.3en$ an% P#yc"o.a$"olo y. 1995477563H585. 57.Aoo% F:2 Fel$on (=. Se.ara$e lin ui#$ic an% a$$en$ional fac$or# in $"e %e+elo.3en$ of rea%in . !o.ic# in Lan ua e *i#or%er. 1994414742H57. 58.Aoo%cocE (A2 <o"n#on 0:. Aoo%cocE;<o"n#on P#yc"oe%uca$ional :a$$ery;(e+i#e%. *L0 !eac"in (e#ource#4 'llen2 !O7 1989.

También podría gustarte