Commentary: Predictors of Adolescent —More to Learn?

Norman G. Poythress, PhD

J Am Acad Psychiatry Law 29:383-6, 2001

In "Predictors ofAdolescent Psychopathy: The Role PCL-SV scores (I? = .28). Collectively, these find of , Hyperactivity, and Sensation Seek ings led the authors to design a two-stage model ing," Vitacco and Rogers1 continue some previous wherein behavioral dysregulation (primarily impul lines ofresearch on thedevelopment ofpsychopathic sivity) contributes to the development of conduct features in youths and extend them to seriously de problems, and conduct problems in turn contribute linquent adolescents. In their study, they explored to the development ofadolescent psychopathy. two primary subjects: the relationship ofvarious in dicators ofbehavioral dysregulation to psychopathy, Heterogeneity of Psychopathy: More to and Lynam's hypothesis that itis the combination of be Learned from this Study? hyperactive-impulsive-attention (HIA) problems The measures, design, and analyses used by Vi and (CD) problems, rather than tacco and Rogers are reasonable, given the objectives the presence of either problem alone, that is the of their study, and the authors note several limita strongest precursor of psychopathic features in tions and suggest directions for future research. youths.2,3 Standard measures ofbehavioral dysregu Given that the PCL:SV is not recommended for use lation constructs (e.g., sensation seeking, impulsiv inindividuals younger than 16 (Ref. 4, p 17), future ity) and HIA-CD symptoms were obtained in asam investigators may optfor one ofseveral measures de ple of79 adolescent males who had been placed ina signed specifically to assess psychopathic features in maximum-security facility as aresult ofadjudication. youths and adolescents. These include the Child Psy Hierarchical regression analyses were used toexplore chopathy Scale (CPS),5 the Antisocial Process the relationships ofthese predictor variables toscores Screening Device (APSD),6 (previously the Psychop on the Psychopathy Checklist: Screening Version athy Screening Device7), and the Hare Psychopathy (PCL-SV).4 These analyses revealed that among in Checklist: Youth Version (PCL-.YV).8 I suggest that dicators of behavioral dysregulation, only impulsiv there may be more to be learned from this data set ity was independently associated with psychopathy than the authors have presented in their report. (fl2 = .15). Acollateral analysis revealed thatimpul In association with developing theories regarding sivity accounted for the lion's share (R2 = .36) of the etiology of antisocial behavior, in recent years variance explained in a measure ofCD, with symp therehas beena revival of the interestin "subtypes," toms of attention-deficit/hyperactivity disorder which recognizes the potential heterogeneity ofindi (ADHD; I? = .04) and sensation seeking {J? = .03) viduals who appear similar on some omnibus index making additional butminor contributions. Regard of psychopathy.9 Briefly, "primary" psychopathy is ing Lynam's theory, Vitacco and Rogers found that thought tobe associated with aconstitutional deficit only CD, not HIA symptoms, significantly predicted in temperament, such that the individual is less sen sitive to cues of punishment and anticipatory nonre- Dr. Poythress isProfessor and Research Director, Department ofMen tal Health Law andPolicy, Florida Mental Health Institute, University ward (behavioral inhibition system (BIS) deficit), ofSouth Florida, Tampa, FL. Address correspondence to: Norman G. whereas "secondary" psychopathy may result from a Poythress, PhD, Department of Mental Health Law and Policy, different temperament defect marked by increased FMHI-USF, 13301 Bruce B. Downs Boulevard, Tampa, FL 33612- 3899. E-mail: [email protected] sensitivity and reactivity to appetitive cues or antici-

Volume 29, Number 4, 2001 383 Commentary

patory reward (behavioral activation system (BAS) Atfirst blush, these results support thefinding of defect).10 In terms ofmeasurement using psychopa Vitacco and Rogers that impulsivity, but not atten thyinstruments developed in the Haretradition (i.e., tion deficits, is associated with the development of the Hare PCL-revised (PCL-R) and its derivatives), psychopathic features generally. However, Colledge primary psychopathy is thought to be associated andBlair also used partial correlations to control for more with the core affective (e.g., callous/unemo therelationship between the PSD factors themselves. tional, shallow affect, lack ofempathy) andinterper It is worthwhile to review thework byMcHoskey et sonal (e.g., ; conning/manipula al}* and Frick et al.14 on the importance ofpartial tive) features, whereas secondary psychopathy is correlations incontrolling for covariance inpsychop thought to beassociated more with deviant lifestyle athy factor scores. In the context of the foregoing (e.g., impulsivity, parasitic lifestyle) features.10 Con discussion regarding heterogeneity of psychopathy, sequently, investigators have begun to explore rela an interesting pattern of results emerged. Control tionships between constructs of interest not only ling for I/CP reduced to nonsignificance the associ with the total scores on psychopathy measures (the ations between CU and impulsivity (r = .099) and method used by Vitacco and Rogers), but also with inattention (r = .074), whereas I/CP remained sig the separate factor scores for these various measures. nificantly correlated with both impulsivity (r = Factor 1 on the PCL-R and PCL:YV and Part 1 of .569) and inattention (r = .503), when controlling the PCL:SV assess core affective and interpersonal for CU traits. These results suggest that impulsivity features, whereas Factor 2 and Part 2 assess social may contribute strongly to variants or subtypes of deviancy features. Levenson et al.'l developed sepa psychopathy manifest mainly by deviant lifestyle fea rate primary psychopathy (modeled after PCL-R tures, but may berelatively unimportant in ourun Factor 1) andsecondary psychopathy (modeled after derstanding ofthe development ofprimary psychop PCL-R Factor 2) measures. In initial studies using athy, which is associated more strongly with the core Frick's and Hare's APSD (formerly the Psychopathy personality features identified by Cleckley.15 Screening Device (PSD)),6 factor analysis also Similar findings were obtained by Frick et al.,16 yielded two factors labeled Callous/Unemotional who used partial correlations to examine the associ (CU), which corresponds to PCL-R Factor 1, and ationsamong oppositional defiantdisorder(ODD), Impulsivity/ConductProblems (I/CP), whichcorre CD, and ADHD symptoms with psychopathic fea sponds to PCL-RFactor2. tures ina clinical sample (n = 160) ofyouths. These Investigators have found in some instances that DSM-III-R symptom indices were significantly cor potential predictors ofpsychopathy may be differen related with all facets of youth psychopathy when tially associated with (predictive of) these various fac zero-ordercorrelations were obtained. However, the ets of psychopathy. For example, Colledge and significant correlations of CU with ODD (r = .42), Blair examined the relationships ofthe impulsivity CD (r = .38), and ADHD (r = .27) wereall reduced andinattention components ofADHD withtheCU tononsignificance (r = .07, .07, -.09, respectively) and I/CP factor scores of Frick's and Hare's youth when partial correlations were used to control for psychopathy measure6 in 71 male adolescents with other facets (impulsivity, ) of youth psy emotional and behavioral difficulties. Using simple chopathy, whereas partial correlations between these zero-order correlations, both the CU and I/CP factor symptom indices and theimpulsivity and narcissism scores of the PSD were highly correlated with both facets remained significant. Again, these findings the inattention and impulsivity components of permit a different impression concerning theimpor ADHD. Using partial correlations, however, they tance of DSM symptom indices for understanding found that it is the impulsivity component of psychopathy, including CD symptoms, than is ob ADHD thatis more substantially related to psycho tained from theunidimensional analysis reported by pathicfeatures. When inattentionwas controlled for, Vitacco and Rogers—that is, CD symptoms may be impulsivity was still significantly related to both CU important in thedevelopment ofsome facets ofpsy (r = .286) and I/CP (r = .407), whereas controlling chopathy, but they are perhaps less important in our for impulsivity resulted in nonsignificant correla understanding of the development of the core affec tions for inattention with CU (r = .002) and I/CP tive features that are thought to typify primary psy (r = .02). chopathy. (See also O'Brien and Frick, who found

384 The Journal ofthe American Academy of Psychiatry and the Law Poythress that presence or absence of conduct problems was clinical or forensic context, to attach the label of psy unrelated to a youth's performance on a passive chopath to youths or adolescents who are assessed avoidance (reward dominance) task thought to iden with any oftheputative youth psychopathy measures tify individuals with a behavioral inhibition deficit; developed to date. There are some data to suggest youth with high CU scores performed more poorly, that the various youth psychopathy measures corre regardless ofthepresence ofconduct problems.) late poorly with one another,18 and there may be The results reported by Vitacco and Rogers re little basis, other than authority, to select among garding dysregulation behaviors are interesting, and them for those who woulddiagnose psychopathy in they draw reasonable inferences from them for treat youth for either clinical or forensic applications. ment recommendations. However, this brief review As Vitacco and Rogers noted, concerns have been suggests that, at least for purposes ofunderstanding raised about the stability of psychopathy scores in better the development of psychopathic features in adolescents,19 and longitudinal research is needed to youths if not also for treatment-planning purposes, clarify thenumerous areas ofconcern that remain. As some additional insights may beavailable from alter such, these authors' suggestion thatyouths oradoles native analyses oftheir data. Inaddition to theuse of cents may be classified as psychopaths using the thepartial correlations described earlier, some recent PCL-.SV seems ill-advised. The manual for the investigators have also found cluster analyses useful PCL:SV indicates that thosewith high scores on this in examining thefeatures ofyouth with varying pat measure should be further evaluated with the PCL-R terns ofpsychopathic (and other) features.1 (Ref. 4, p22). There is nosuggestion thatthescreen ing version should beused todiagnose psychopathy, HIA-CP and Psychopathic Features even in adults. As noted, there are other consider ations that may militate specifically against the use of in Youths the PCL:SV for diagnosing psychopathy in youths. Vitacco and Rogers found little support for Ly Until a considerably stronger research basis for diag nam's HIA-CD hypothesis, as did Frick et al}6 be nosis is available, it is recommended that both re fore them. However, these investigators used differ search and clinical discourse (if any) use psycho ent research methods (e.g., Lynam used secondary pathic features, callous/unemotional features, or data analysis), different populations (i.e., research similar suchcharacterizations that stop short of cat sample versus clinic sample versus juvenile justice egorical labels. sample), anddifferent measures ofpsychopathic fea tures inyouth (i.e., CPS, Lynam; APSD, Frick etal.; References PCL:SV, Vitacco and Rogers). Given the potential 1. Vitacco M, Rogers R: Predictors of adolescent psychopathy: the array of factors that may have contributed to incon role of impulsivity, hyperactivity, and sensation seeking. J Am Acad Psychiatry Law29:374-82, 2001 sistent outcomes, further research is needed to ex 2. Lynam D: Early identification of chronic offenders: who is the plore Lynam's hypothesis regarding the relationship, fledgling psychopath? Psych Bull 120:209-34, 1996 ifany, ofcurrently accepted diagnostic entities (e.g., 3. Lynam D: Early identification of the fledgling psychopath: locat ADHD, CD) to psychopathic features in youth. ing thepsychopathic child inthecurrent nomenclature. JAbnorm Psychol 107:566-75, 1998 4. Hart SD, Cox DN, Hare RD: The Hare PsychopathyChecklist: A Cautionary Note on Terminology Screening Version. Toronto: Multi-Health Systems, Inc., 1995 5. Lynam D: Pursuing the psychopath: capturing the fledgling psy and Labeling chopath in a normological net.J Abnorm Psychol 106:425-38, Investigating the causes and correlates of psycho 1997 6. Frick PJ, Hare RD: The Antisocial Process Screening Device. pathic features inyouths andadolescents is currently Toronto: Multi-Health Systems, Inc., in press oneof the moreactive and exciting areas of research. 7. Frick PJ, O'Brien BS,Wootton JM, McBurnett K: Psychopathy It is a logical downward extension from the adult andconduct problems inchildren. JAbnormPsychol 103:700-7, 1994 psychopathy literature, in light of current theories 8. Forth AE, Kosson DS, Hare RD: The Psychopathy Checklist: that hypothesize constitutional deficits of one ilkor Youth Version. Toronto: Multi-Health Systems, Inc., in press another to be important etiologic factors. However, 9. Herve HF, Hui LingJY, Hare RD:Criminal psychopathy and its subtypes. Presented at the Biennial Conference of the American the excitement of opening this newarea of research Psychology-Law Society, New Orleans, March 2000 and preliminarily informative findings notwith 10. Lykkcn D: The Antisocial Personalities. Hillsdale, NJ: Lawrence standing, it isstill premature, in my opinion, in any Erlbaum Associates, 1995

Volume 29, Number 4, 2001 385 Commentary

11 Levenson MR, Kiehl KA, Fitzpatrick CM: Assessing psychopathic problems incommunity and clinic-referred samples ofchildren: attributes inanoninstitutionalized population. J Pers Soc Psychol further development ofthepsychopathy screening device. Psychol 68:151-8, 1995 Assess 12:382-93, 2000 12. Colledge E, Blair RJR: The relationship in children between the inattention and impulsivity components of attention deficit and 17. O'Brien BS,FrickPJ:Reward dominance: associations with anx hyperactivity disorder and psychopathic tendencies. Pers Indiv iety, conduct problems, and psychopathy in children. J Abnorm DifF30:1175-87,2001 Child Psychol 24:223-240, 1996 13. McHoskey JW, Worzel W, Szyarto C: Machiavellianism and psy 18 Forth AE, Mailloux DL: Psychopathy in youth: What do we chopathy. J Pers Soc Psychol 74:192-210, 1998 know? inThe Clinical and Forensic Assessment ofPsychopathy: A 14. Frick PJ, Lilienfeld SO, Edens JF, Poythress NG, McBurnett K: Practitioner's Guide. Edited byGacono CB. Hillsdale, NJ: Law The association between anxiety and antisocial behavior. Primary rence Erlbaum Associates, 2000 Psychiatry 7:52-7, 2000 19. EdensJF, Skeem JL, Cruise KR, CaufFman E: Assessments of 15. Cleckley H: The Mask ofSanity (ed 4). St. Louis, MO: Mosby, 1964 juvenile psychopathy and its association with violence: a critical 16. Frick PJ, Bodin SD, Barry CT: Psychopathic traits and conduct review. BehavSci Law 19:53-80, 2001

386 The Journal of the American Academy of Psychiatry and the Law