Abstract Rationale: The role of serotonin in human ag-gression and impulsivity was evaluated by administeringparoxetine or placebo for 3 weeks and comparing the ef-fects on laboratory measures of aggression and impulsi-vity among male subjects with a history of conduct dis-order. Methods: Twelve male subjects with a history ofcriminal behavior participated in experimental sessions,which measured aggressive and impulsive responses. Sixsubjects were assigned to placebo treatment and six sub-jects to placebo and paroxetine treatment. Aggressionwas measured using the point subtraction aggression par-adigm (PSAP), which provides subjects with an aggres-sive and monetary reinforced response options. Impul-sive responses were measured using a paradigm thatgives subjects choices between small rewards after shortdelays versus larger rewards after longer delays. Results:Chronic administration of paroxetine (20 mg/day) for 21days produced significant decreases in impulsive re-sponses. Decreases in aggressive responses were evidentonly at the end of paroxetine treatment. Decreases in im-pulsive and aggressive responses could not be attributedto a non-specific sedative action because monetary rein-forced responses were not decreased as has been ob-served following CNS sedation. Conclusions: Inhibitionof serotonin reuptake by paroxetine is the possible mech-anism for reductions in aggressive and impulsive re-sponses. These results support other data linking seroto-nin function and aggression and impulsivity.Keywords Aggression · Impulsivity · Serotonin · ParoxetineIntroductionA large number of scientific articles support a relation-ship between serotonin (5-HT) and human impulsive andaggressive behavior (Linnoila et al. 1983; Virkkunen etal. 1989). Studies in non-human species have also linkedreduced 5-HT levels to aggressive and impulsive behav-ior (e.g. Higley et al. 1996). Pharmacological manipula-tion of the 5-HT system through tryptophan depletion orsupplementation resulted in changes in predicted direc-tions in self-reported mood/hostility (Cleare and Bond1995) and laboratory measures of aggression (Bjork etal. 1999). Assessments of 5-HT response through chal-lenge agents have revealed blunted 5-HT response inchildren with conduct disorder (Stoff et al. 1992) andpersonality-disordered individuals reporting high levelsof aggression and impulsivity (Moss et al. 1990).A number of studies with non-human subjects haveestablished a relationship between 5-HT and laboratorymeasures of impulsivity. Decreases in impulsive behav-ior have been reported following administration of 5-HTreuptake inhibitors, 5-HT agonists (Soubrie 1986), and5-HT releasing agents (Poulos et al. 1996). Lesioning of5-HT pathways produced decreased impulse control (Hoet al. 1998). Collectively, these investigations suggestthat reduced 5-HT plays a role in impulsive and aggres-sive behavior.Serotonin reuptake inhibitors (SSRIs) such as fluoxe-tine have been found to decrease aggression in severalspecies of animals (see review by Fuller 1996). Some in-vestigators have suggested that patients with impulse and/or aggression disorders may respond favorably to SSRIs(e.g. Boyer 1992). Many clinicians have reported a re-duction in anger outbursts among depressed, affect labileand post-traumatic stress disorder patients treated withfluoxetine (Rosenbaum et al. 1993). Fluoxetine com-pared to placebo treatment reduced self-report measuresof irritability and aggression among personality disor-dered participants (Coccaro and Kavoussi 1997).Critical to understanding the biology of impulsive/ag-gressive behavior is accurate measurement of these be-D.R. Cherek (✉) · S.D. Lane · C.J. Pietras · J.L. SteinbergHuman Psychopharmacology Laboratory, Department of Psychiatry and Behavioral Science, University of Texas-Houston Health Science Center, 1300 Moursund Street, Houston, TX 77030-3497, USAe-mail: [email protected].: +1-713-5002797, Fax: +1-713-5002618Psychopharmacology (2002) 159:266–274DOI 10.1007/s002130100915ORIGINAL INVESTIGATIONDon R. Cherek · Scott D. Lane · Cynthia J. Pietras Joel L. SteinbergEffects of chronic paroxetine administration on measures of aggressive and impulsive responses of adult males with a history of conduct disorderReceived: 21 March 2001 / Accepted: 10 August 2001 / Published online: 12 October 2001© Springer-Verlag 2001267haviors. In the present study, we describe a laboratory-based procedure in which subjects were exposed to testsof both impulsive and aggressive responding in the sameexperimental day. The impulsivity component involved awell-documented delay of gratification task (Mazur1987; Logue 1995) in which subjects chose between asmall reward available after a short delay and a larger re-ward available only after a longer delay. This procedurehas demonstrated sensitivity in detecting impulsivenessin populations with impulse control difficulties (Logue1995), and in measuring drug effects on impulsive be-havior (Cherek and Lane 2000). The aggressive respond-ing component employed the point subtraction aggres-sion paradigm (PSAP) (Cherek 1992). The external va-lidity of this procedure has been established in studiesdemonstrating differences between violent and non-vio-lent individuals (Cherek et al. 1997).The present investigation assessed the effects of par-oxetine (a 5-HT reuptake inhibitor) on aggressive andimpulsive behavior of individuals with a history of anti-social behavior. We postulated that, in accord with previ-ous human studies, increasing 5-HT activity in the CNSwould produce decreases in aggressive and impulsive re-sponding. An earlier study with d,l-fenfluramine, whichreleases 5-HT and dopamine, reported significant de-creases in aggressive and impulsive responses among agroup of CD male subjects (Cherek and Lane 1999). Thepresent experiment would allow us to contrast the resultsof chronic 5-HT reuptake inhibition with our previousstudies using 5-HT releasing agents.Materials and methodsSubjectsTwelve male subjects on parole were recruited via newspaper ad-vertisements into a laboratory study involving measures of behav-ior and paroxetine administration. Informed consent was obtainedduring intake interviews. All subjects had been convicted of atleast one felony and were incarcerated for some period of time.The specific consent forms and all procedures were reviewed andapproved by the IRB for the Health Science Center.Seven subjects reported no current alcohol use,