Citalopram Treatment for Impulsive Aggression in JORGE L. ARMENTEROS, M.D.,
Transcription
Citalopram Treatment for Impulsive Aggression in JORGE L. ARMENTEROS, M.D.,
Citalopram Treatment for Impulsive Aggression in Children and Adolescents: An Open Pilot Study JORGE L. ARMENTEROS, M.D., AND JOHN E. LEWIS, PH.D. ABSTRACT Objective: To assess the short-term effect and safety of citalopram in the reduction of impulsive aggression in children and adolescents. Method: Twelve subjects, aged 7 to 15 years, were attending a psychiatric outpatient clinic and had a profile of impulsive aggression. Subjects were treated in an open trial with citalopram for 6 weeks after a 1-week washout period. Dosage was regulated individually over a period of 4 weeks. The starting dose was 10 mg/day followed by 10 mg increments on a weekly basis. The maximum dose was not to exceed 40 mg/day. Outcome measures included the Modified Overt Aggression Scale (MOAS), the Child Behavioral Checklist (CBCL), and the Clinical Global Impressions (CGI). Results: Eleven subjects completed the study. Citalopram produced clinically and statistically significant reductions on target symptoms of impulsive aggression, independent of other behavioral problems, as measured by the MOAS, the CBCL, and the CGI at doses ranging from 20 to 40 mg/day (mean = 27 mg). No major adverse reactions were associated with citalopram use. Conclusion: Citalopram appears to be effective and well tolerated in this sample of children and adolescents with impulsive aggression. J. Am. Acad. Child Adolesc. Psychiatry, 2002, 41(5):522–529. Key Words: impulsive aggression, citalopram, psychopharmacology. Accepted December 18, 2001. Dr. Armenteros is Director of Pediatric Psychopharmacology, Division of Child and Adolescent Psychiatry, University of Miami School of Medicine; Dr. Lewis is Assistant Professor, Department of Psychiatry and Behavioral Sciences, University of Miami School of Medicine. Reprint requests to Dr. Armenteros, Division of Child and Adolescent Psychiatry, University of Miami School of Medicine, 1695 NW 9th Avenue, Suite 1404, FL 33136; e-mail: j.armenteros@miami.edu. 0890-8567/02/4105–05222002 by the American Academy of Child and Adolescent Psychiatry. Several researchers have attempted to define the phenomenology of human aggression and its subtypes (Barratt et al., 1997; Vitiello et al., 1990; Vitiello and Stoff, 1997). Two types of aggressive behaviors have emerged. The first type is considered to have a significant affective component and to be more impulsive. It is characterized by a hair-trigger response to a stimulus that results in an agitated state and culminates in an aggressive act. This impulsive aggression is viewed as an overreaction to a minor provocation and is often accompanied by disinhibition and behavioral dyscontrol. The individual who exhibits impulsive aggression often feels remorse after the aggressive act and vows not to do it again, yet they often do repeat the behavior. The second type of aggression is considered to be nonimpulsive; an instrumental behavior prompted by its anticipated benefits. Brown et al. (1979) were among the first researchers to evaluate the concept of impulsive aggression. The authors found personality disorders generally associated with behavioral impulsivity to have significantly higher aggression indices. A recent study by Seroczynski et al. (1999) provides further support for this concept. The authors found genetic associations between impulsivity and aggression. In children, fight initiation was linked to impulsivity irrespective of whether attention-deficit/hyperactivity disor- 522 J . A M . A C A D . C H I L D A D O L E S C . P S YC H I AT RY, 4 1 : 5 , M AY 2 0 0 2 Aggression is a significant problem in our society, with serious psychological, social, and financial consequences. The problem is not limited to the adult population; aggressive behaviors are extremely common in the daily lives of many juveniles. Aggressive behavior patterns have increased among psychiatrically referred children and adolescents over the past 2 decades (Achenbach and Howell, 1993). A recent study using nationally representative data from the Youth Risk Behavior Survey showed that physical fighting and weapon carrying remains at high levels in minority groups (Brener et al., 1999). In 1991, the U.S. Dept. of Health and Human Services (Public Health Service, 1991) established a reduction in youth violence as a national health objective for the year 2000. Unfortunately, the prevalence of youth violence and aggression is still unacceptably high as this objective has not been met with success. CITALOPRAM FOR IMPULSIVE AGGRESSION der was present or not (Halperin et al., 1995). In addition, impulsivity was related to delinquency, suggesting that children with poor self-control would steal and fight on the spur of the moment (White et al., 1994). In summary, these studies suggest that impulsivity may be a trigger for aggressiveness. In this model, the aggressive act per se is conceived as the observable behavioral manifestation of the underlying impulsivity. The construct of impulsive aggression offers a unitary understanding of a common behavioral problem in children. Although this label has varied among authors (Coccaro et al., 1989; Linnoila et al., 1983), construct and validity studies (Atkins et al., 1993; Barratt et al., 1997) support the soundness of this concept. A basic constituent of the impulsive aggression construct is its dimensional quality. As an independent entity, impulsive aggression does not comprise a separate diagnostic category. Furthermore, no psychiatric diagnosis subsumes impulsive aggression as a pathognomonic sign. Various psychiatric disorders, however, are frequently accompanied by impulsive aggressive behavior. These include conduct disorder, attention-deficit/hyperactivity disorder, intermittent explosive disorder, personality disorders, substance abuse, and psychoses (DSM-IV) (American Psychiatric Association, 1994). The frequency of impulsivity and various forms of aggression was clearly demonstrated by the DSM-IV field trials for disruptive behavior disorders (Frick et al., 1994). The sample consisted of 440 clinic-referred youths who were consecutively admitted to a heterogeneous group of mental health clinics. Although the youths were referred for a variety of reasons, impulsive and aggressive behaviors were very frequent. Consistent with these epidemiological data, a study of 111 children referred to a psychiatric outpatient clinic found that impulsivity was strongly associated with initiation of fighting (Halperin et al., 1995). More importantly, the children who initiate fights were found to be impulsive irrespective of the presence or absence of attention-deficit/hyperactivity disorder or any other disruptive disorder. To reduce diagnostic categorical constraints, impulsive aggression should be considered as a dimensional behavioral disturbance. The association between serotonin (5-HT) and aggression has been established with the consistent observation that abnormalities in central 5-HT function correlate with impulsive aggression, which is supported by an extensive animal literature. Muricidal behavior, shock-induced fighting, and attack are all increased by the chemical or electric lesioning of central 5-HT neurons (Katz, 1980; Sewell et al., 1982; Siegel et al., 1999). Agents that increase central 5-HT activity reversed these behaviors. Moreover, a study of free-ranging rhesus monkeys found a strong negative correlation between cerebrospinal fluid 5-hydroxyindoleacetic acid (CSF 5-HIAA), the main metabolite of serotonin, and aggression (Higley et al., 1992). The hypothesis that central 5-HT dysfunction contributes to impulsive aggression in humans is also well supported by research. In 1976, Asberg and associates published their findings concerning low CSF 5-HIAA concentration in depressed patients. The investigators demonstrated that depressed patients with low CSF 5HIAA concentration were more likely to complete suicide using particularly violent means than depressed patients with higher CSF 5-HIAA concentration. Subsequently, Brown et al. (1979) found a significant negative correlation between impulsive aggression and CSF 5-HIAA within a group of subjects with personality disorder. Moreover, Linnoila et al. (1983) found significantly lower 5-HIAA in the CSF of impulsive violent offenders when compared with offenders who premeditated their acts. These findings suggest that abnormalities in the brain’s serotonergic functioning predispose individuals to impulsive aggression rather than to nonimpulsive, premeditated aggressive behaviors. Using a different strategy, Coccaro et al. (1989) examined the central serotonergic function of patients with affective and personality disorders by means of a fenfluramine challenge. The authors found a blunted prolactin response to fenfluramine, indicating a significant reduction in central serotonergic function in patients with impulsive aggression. Abnormalities in the serotonergic system also abide by a dimensional, rather than a categorical, order. Low serotonergic indices have been associated with major depressive disorder (Meltzer and Lowy, 1987), history of suicide attempt (Asberg et al., 1976; Mann et al., 1995), personality disorders (Brown et al., 1979; Coccaro et al., 1989), and impulsive aggression as reviewed above. These data cumulatively suggest that disturbances in serotonergic function lack nosological specificity and are not linked to a particular psychiatric disorder (van Praag et al., 1987). The serotonergic system in children and adolescents has been studied infrequently because of ethical and practical difficulties in obtaining CSF samples from this pop- J . A M . A C A D . C H I L D A D O L E S C . P S YC H I AT RY, 4 1 : 5 , M AY 2 0 0 2 523 ARMENTEROS AND LEWIS ulation. Kruesi et al. (1990) extended the adult findings to children and adolescents by demonstrating reduced 5HIAA in CSF of aggressive juveniles with disruptive behavior disorders. In contrast, Halperin et al. (1994) found a positive rather than inverse correlation between aggression ratings and the prolactin response to fenfluramine challenge in children with attention-deficit/hyperactivity disorder. Another study also found a higher, rather than lower, prolactin response to the fenfluramine challenge in aggressive children (Pine et al., 1997). It may be possible that a developmental component modulates the relationship between central serotonergic function and impulsive aggression in children. Taken together, these data strongly support the association between abnormalities in the brain’s serotonergic functioning and impulsive aggression, particularly in adults. However, the association in children and adolescents is inconclusive at the present time. On the basis of the existing data, one would expect drugs that selectively affect central serotonergic function to be effective for impulsive aggression. More precisely, the candidate drug should be effective, irrespective of a syndromal or nosological framework. Various psychopharmacological approaches have been used for the treatment of impulsive aggression. An early study of tryptophan, a precursor required for serotonin synthesis, demonstrated a reduced need for injection of antipsychotic medications and sedatives in aggressive psychiatric inpatients (Volavka et al., 1990). Because of the ability of lithium to increase brain tryptophan concentrations and enhance serotonergic activity, it was also studied in the treatment of impulsive aggression (Sheard et al., 1976). Lithium was as effective as haloperidol in reducing aggressiveness and explosiveness in hospitalized children with conduct disorder and a profile of aggressive and explosive behavior (Campbell et al., 1984). Eltoprazide, a mixed serotonin 5-HT1 agonist, was found to be effective in reducing aggressive behavior in a group of patients with schizophrenia (Tiihonen et al., 1993). Theoretically, selective serotonin reuptake inhibitors (SSRIs) should be effective in reducing impulsive aggression, but data on this issue are limited. Clomipramine, a relatively selective serotonin reuptake inhibitor, was superior to placebo in reducing anger in autistic children (Gordon et al., 1993). Among 30 adults with autism, fluvoxamine was also found to be more effective than placebo in reducing aggression (McDougle et al., 1996). A study of depressed adult patients with anger attacks found treat- ment with fluoxetine to be effective in eliminating these attacks in 71% of the patients (Fava et al., 1993). Interestingly, the presence of anger attacks predicted a greater treatment efficacy with fluoxetine irrespective of depressed symptomatology. Using a double-blind crossover design, Vartiainen et al. (1995) examined the effect of citalopram in 14 schizophrenic patients known to be impulsive and aggressive. Citalopram (20–60 mg/day) or placebo was administered as an adjuvant to current neuroleptic medication for a 24-week period. The authors found a reduced frequency of aggressive incidents during citalopram treatment. Moreover, the decrease in aggressive behavior was independent of a decrease in any other mental symptom. Side effects were reported to be mild. For our study, we have chosen citalopram in the treatment of impulsive aggression for several important reasons. First, citalopram is the most specific SSRI for the inhibition of serotonin reuptake (Hyttel, 1982). This will avoid possible unwanted effects on other neurotransmitter systems. The greater the biochemical specificity of a drug, the greater the chance it will be effective against the dimension of impulsive aggression across diagnoses. Second, the plasma concentration of citalopram is proportional to the dose administered (i.e., it has linear pharmacokinetics). This eliminates the problem of metabolic autoinhibition. Third, food intake or times of dosing do not affect the pharmacokinetics of citalopram. Fourth, the mean terminal half-life of citalopram is 33 hours. This makes a single daily dose possible and avoids most discontinuation adverse effects. Lastly, the inhibition of hepatic cytochrome P450 (CYP2D6) isoenzymes caused by citalopram is clinically insignificant. This study presents data concerning the effects of citalopram in the treatment of impulsive aggression in children and adolescents. Our study departs from the generic approach in treating impulsive aggression. Instead of targeting symptoms with no connection to a specific functional hypothesis, our pharmacotherapy emerges from our current understanding of the underlying neurobiology of impulsive aggression. 524 J . A M . A C A D . C H I L D A D O L E S C . P S YC H I AT RY, 4 1 : 5 , M AY 2 0 0 2 METHOD Patient Selection Criteria Participants in the trial were male and female children and adolescents (at least 7 but younger than 16 years of age) who were referred for psychiatric services to the Child and Adolescent Psychiatry outpatient clinic at the University of Miami School of Medicine. To be CITALOPRAM FOR IMPULSIVE AGGRESSION enrolled in the study, patients must have had an established pattern of aggressive behavior of at least 6-months duration with a minimum of three acts of aggression in the past week, two of which had to be acts of physical aggression against other people, objects, or self. Impulsivity needed to be part of the clinical picture, thus inclusion criteria required subjects to score at least one standard deviation beyond a normative sample of peers on either “commission errors,” “hit reaction time,” or “risk taking” on the Conner’s Continuous Performance Test (CPT) (Conners, 1995). In addition, patients were required to have a score greater than or equal to 4 on the Clinical Global Impressions (CGI)-Severity scale. Patients were also required to have laboratory results without significant abnormalities. Children who scored below 75 on Verbal and Performance IQ, measured with the Kaufman Brief Intelligence Test (Kaufman and Kaufman, 1990), were excluded. Previous treatments were assessed by means of the Prior Medication Record (Psychopharmacology Bulletin, 1985). DSM-IV diagnoses were made by means of a semistructured psychiatric interview using the Child and Adolescent Symptom Inventories4 (Gadow and Sprafkin, 2000). These are behavior rating scales that screen for a full range of DSM-IV disorders usually first diagnosed in infancy, childhood, or adolescence. A few exceptions are learning disorders, motor skills disorders, communication disorders, and selective mutism, which are not elicited by the instrument. Using a symptom count scoring procedure, we identified patients with a minimum number of symptoms necessary for a DSM-IV diagnosis. The final diagnoses were assigned through a best-estimate process. The research team, composed of two child and adolescent psychiatrists, reviewed the information obtained from the parent and the youth reports in conjunction with all other clinical information available. The information was summarized, and possible DSM-IV diagnoses were outlined. We arrived at a consensus best-estimate diagnoses for each subject. A basic constituent of the impulsive aggression construct is its dimensional quality; therefore, diagnoses for participants included attention-deficit/hyperactivity disorder, oppositional defiant disorder, major depression, social phobia, generalized anxiety disorder, and conduct disorder as defined by the DSM-IV. Patients were excluded from the study if they had any of the following comorbid disorders: schizophrenia, other psychotic disorders, bipolar I or II disorder, major depression with psychotic features, delirium, dementia, and alcohol or substance abuse/dependence. Patient’s parents or guardians gave their permission for participation in the study and signed an informed consent document approved by the University of Miami School of Medicine Institutional Review Board. Patients also provided assent for their participation. Study Design then summed to yield a total aggression score. We only rated the occurrence of impulsive aggressive events over the previous 1-week period. The MOAS also measures two categories of irritability, subjective and overt, which are summed to yield a total irritability score. Parents/guardians were the respondents for this scale with the child providing information on subjective irritability. We also used the Aggression Questionnaire (AQ) to classify aggression into “predatory” and “affective” subtypes (Malone et al., 1998; Vitiello et al., 1990; Vitiello and Stoff, 1997). The AQ is a 16-item scale consisting of eight items that assess predatory aggression and eight items that assess affective aggression. Items are scored as not true (0), partly true (1), or very true (2). Predatory and affective scores are obtained by adding the corresponding items with a possible range of 0 to 16. The Predatory-Affective index is calculated by subtracting the affective score from the predatory score. Higher index scores indicate a predominantly predatory type of aggression, while lower scores indicate primarily an affective or impulsive subtype. Severity of illness and improvement were assessed weekly on the CGI. The CPT was administered at baseline and endpoint. The CPT consists of a series of letters presented on a computer screen in rapid succession. The subject is required to press a key whenever a target stimulus is presented, and to refrain from pressing the key when the nontarget stimulus is presented. The CPT generates separate measures of inattention and impulsivity. In addition, the parents completed the Child Behavioral Checklist (CBCL) (Achenbach and Howell, 1993) at baseline and endpoint. Adverse events, heart rate, blood pressure, and body weight were determined at each weekly visit. Routine clinical laboratory studies included complete blood cell count and differential, electrolytes, thyroid and kidney function tests, and urine analysis for toxicology and pregnancy (females). These were conducted during the screening phase and at week 6, or upon study withdrawal. Statistical Method Frequency and descriptive statistics were calculated to check all relevant characteristics of the data. The aggression and irritability scores on the MOAS, the AQ, the CGI-Severity of Illness, and the CPT were analyzed using repeated measures analysis of variance in a general linear model design. Mauchly’s (1940) test of sphericity was used to evaluate the form of the variance–covariance matrix of the data and whether the test was significant, and then the Greenhouse-Geisser Epsilon adjustment was used when the sphericity assumption was violated. When a significant overall within-subjects F value was obtained, marginal means were used to assess pairwise comparisons with the least significant difference from the baseline value to the value at each follow-up time point. Data on the CPT and CBCL, reported only at baseline and discharge, were analyzed with paired-sample t tests. The significance level for all analyses was p = .05. After a 1-week washout period, subjects entered a 6-week open trial of citalopram under careful clinical and laboratory monitoring. The dose of citalopram was individually adjusted to obtain maximum benefits with minimal adverse effects. The initial dose of citalopram was 10 mg/day with weekly adjustments of 10 mg in either direction depending on the prevalence of impulsive aggressive behaviors or the emergence of side effects. No other psychotropic medications were allowed during the trial. Subjects were seen at weekly intervals and evaluated with standardized instruments and global assessments for efficacy. The primary efficacy measure was the Modified Overt Aggression Scale (MOAS) (Coccaro et al., 1991). The MOAS retrospectively measures four categories of aggression: (1) verbal aggression, (2) physical aggression against self, (3) physical aggression against objects, and (4) physical aggression against other people. The most severe aggressive event within each category is multiplied by its designated weight factor and Twelve subjects (10 males and 2 females) were enrolled into the protocol. Their ages ranged from 7 to 15 years old (mean = 10.22 years). Table 1 summarizes demographic characteristics and baseline severity of disease on these subjects. Their scores on the AQ predatory-affective index ranged from –11 to 2, indicating that the primary subtype of aggression was impulsive/affective. Therapeutic doses J . A M . A C A D . C H I L D A D O L E S C . P S YC H I AT RY, 4 1 : 5 , M AY 2 0 0 2 525 RESULTS ARMENTEROS AND LEWIS TABLE 1 Baseline Demographic Characteristics and Severity of Disease in 12 Subjects With Impulsive Aggression Subject Age Gender 1 2 3 4 5 6 7 8 9 10 11 12 a 15 10 7 8 8 9 15 8 12 8 12 7 F M M M M M F M M M M M DSM-IV Diagnoses PredatoryAffective Index CGI Severity Maximum Dose (mg/day) –4 –2 –6 –2 2 –11 –6 –6 –8 –1 –3 –4 5 4 5 4 5 6 5 6 5 5 5 5 20 30 30 40 20 40 40 40 40 40 20 20 ODD, CD ADHD-C, CD ODD, CD ODD, CD, SP ADHD-C, ODD, CD ADHD-C, ODD, CD ODD, CD, GA, SP ADHD-C, ODD, CD CD ADHD-C, ODD, CD ADHD-I, ODD, CD ADHD-C Note: ADHD-C = attention-deficit/hyperactivity disorder, combined; ADHD-I = attention-deficit/hyperactivity disorder, inattentive; CD = conduct disorder; ODD = oppositional defiant disorder; SP = social phobia; GA = generalized anxiety disorder. a Subject 12 was discontinued from the study because of his high level of hyperactivity, requiring treatment with a stimulant medication. nificantly different from baseline (F7 = 4.98, p < .001). The weekly scores on the MOAS Aggression Against Objects subscale were also significantly lower from baseline (F6 = 7.9, p < .001). The scores on the MOAS Verbal Aggression, Aggression Against Self, and Aggression Against Other People subscales were nonsignificantly different. Total irritability scores were significantly reduced on weeks 1, 2, 5, and 6 compared with baseline (F7 = 2.95, p = .02). Aggressive behaviors also decreased significantly as measured by the parent’s report on the CBCL. Delinquent behaviors, which are associated with predatory aggres- of citalopram at endpoint ranged from 20 mg/day to 40 mg/day (mean = 27 mg/day). One subject was discontinued from the study because of his high level of hyperactivity, requiring treatment with a stimulant medication. Medications used immediately before the trial and thus required a washout period included methylphenidate, dextroamphetamine, sertraline, and carbamazepine. For the 11 subjects who completed the study, marked improvement was observed on measures of both overt aggression and irritability on the MOAS (Table 2 and Fig. 1). Each weekly mean total score for aggression was sig- TABLE 2 Modified Overt Aggression Scale (MOAS) Total Scores of Subjects Treated With Citalopram Total Aggression Scores Subject 1 2 3 4 5 6 7 8 9 10 11 Mean SD Total Irritability Scores Baseline Wk 1 Wk 2 Wk 3 Wk 4 Wk 5 Wk 6 Baseline Wk 1 Wk 2 Wk 3 Wk 4 Wk 5 Wk 6 22 15 21 14 31 24 18 28 12 16 14 0 18 5 18 10 8 3 22 6 19 9 11 13 2 14 7 NA 10 NA 9 13 3 2 7 15 6 4 14 8 19 17 13 9 2 8 25 8 10 8 2 22 15 18 3 10 6 1 0 7 6 2 25 10 NA 6 3 3 2 0 9 3 1 6 15 25 3 8 6 5 5 6 6 7 5 5 3 6 2 1 3 6 0 6 0 6 2 8 9 1 4 1 7 0 NA 5 NA 6 6 3 2 3 5 2 6 6 1 5 8 6 7 0 4 6 5 0 2 4 6 6 7 3 0 2 0 2 2 6 2 6 7 NA 1 3 3 0 2 1 3 4 1 5 4 3 19.55 6.20 10.73 7.36 9.11 4.34 10.36 5.55 11.00 7.97 7.30 7.10 6.36 7.50 5.64 1.29 3.91 3.21 3.67 2.55 4.64 2.29 3.91 2.43 2.80 2.57 2.64 1.50 Note: Wk = week of study; NA = data are not available. 526 J . A M . A C A D . C H I L D A D O L E S C . P S YC H I AT RY, 4 1 : 5 , M AY 2 0 0 2 CITALOPRAM FOR IMPULSIVE AGGRESSION subjects who completed the protocol, only 4 experienced mild adverse effects that included somnolence, headache, drowsiness, and nightmares. In every case of somnolence, the medication was shifted from morning administration to bedtime administration. This strategy eliminated all complaints of daytime somnolence. Treatment was not discontinued in any of these cases because of adverse side effects. Throughout the study, complete blood cell count and differential, electrolytes, and thyroid and kidney function tests remained within normal limits. DISCUSSION Fig. 1 Temporal assessment of change in Modified Overt Aggression Scale (MOAS) total aggression and irritability scores during a 6-week course of citalopram in children and adolescents with impulsive aggression. In this small, open trial of citalopram, children and adolescents with several DSM-IV diagnoses had clinically and statistically significant reductions in impulsive aggression from baseline to end of treatment across all measures. Citalopram affected both aggression and irritability in 9 out of 11 subjects who completed the study, as measured by the MOAS. Moreover, our results suggest that the effect of citalopram treatment on impulsive aggression is independent of other comorbid behavioral problems. Clinical improvement was confirmed by parental reports on the CBCL. In the present study, the subjects received several DSMIV diagnoses, and all had long standing histories of impulsive aggressive behavior. In all cases, the impulsive aggressive behavior had failed to respond to other treatments. All subjects enrolled in the study completed the 6-week protocol with the exception of subject 12, resulting in a modest attrition rate of 8.3%. The medication was very well tolerated. We did not encounter the agitation and behavioral disinhibition commonly associated with SSRI sion (Vitiello et al., 1990; Vittiello and Stoff, 1997), were nonsignificantly different. Internalizing problems, like anxiety or depression, were also nonsignificantly different. We found a significant reduction in the total problems score on the CBCL, which resulted primarily from a decrease in aggressive behaviors (Table 3). Ratings on the CGI-Severity of illness, which consider the total clinical experience of the rater with this particular population, also declined significantly at the end of citalopram treatment (F7 = 5.78, p < .001). On the Global Improvement item of the CGI at endpoint, which compares the current condition to the subject’s condition before treatment, five subjects were rated as “Much Improved,” four as “Improved,” zero as “Slightly Improved,” and two as “Unchanged.” Citalopram had no effect on CPT scores. Citalopram was generally well tolerated in this sample and most adverse effects were not serious. Out of the 11 TABLE 3 Mean Child Behavior Checklist Scores of Subjects Treated With Citalopram Baseline Endpoint Subscale Mean SD Mean SD p Value Withdrawn Somatic Complaints Anxious/Depressed Social Problems Thought Problems Attention Problems Delinquent Behavior Aggressive Behavior Internalizing Problems Externalizing Problems Total Problems 69.18 60.18 68.73 67.73 71.55 75.18 71.00 77.27 68.27 73.91 74.55 10.61 9.47 11.82 9.38 13.66 14.05 7.04 13.48 11.46 8.60 8.66 64.45 58.09 63.64 63.36 69.73 69.64 66.00 70.64 63.18 68.55 68.82 10.92 7.42 11.69 8.98 10.78 11.73 10.67 13.34 11.56 11.48 9.57 NS NS NS NS NS p = .04 NS p = .01 NS p < .03 p < .02 Note: NS = not significant. J . A M . A C A D . C H I L D A D O L E S C . P S YC H I AT RY, 4 1 : 5 , M AY 2 0 0 2 527 ARMENTEROS AND LEWIS treatment in children and adolescents. Subjects who responded were continued on citalopram at the end of the protocol. Follow-up of subjects after the protocol ended was strictly clinical; therefore, research assessments (e.g., MOAS or CGI) were not repeated. Our best clinical impression is that the benefits of citalopram were maintained and adverse effects did not emerge anew for up to a 4-month period. These data support the hypothesis that a central serotonergic dysfunction leading to impulsive aggression could be ameliorated by treatment with a specific SSRI. Further studies need to correlate abnormalities on biological measures of central serotonergic functioning to demonstrate clearly that the mechanism for reduced impulsive aggression is in fact an improvement in serotonergic functioning. Reports of controlled clinical trials suggest that the aggressive behaviors of many children with conduct disorder are significantly reduced during periods of baseline inpatient hospitalization (Campbell et al., 1984; Cueva et al., 1996; Malone et al., 1997). The aggressive behaviors are clearly affected by the change in environment, because these children left their homes and were hospitalized, and by the hospital milieu and structure. Although the participants in our study continued to live in the same home environments, the possibility exists that weekly visits to the outpatient clinic could exert a therapeutic effect on impulsive aggressive behaviors. This putative effect would be independent of the psychopharmacological effect of citalopram and similar to the “placebo effect” reported elsewhere. A rigorous double-blind and placebo-controlled study would be required to discern these competing effects. The literature review and clinical experience indicate that stimulants may be effective in reducing mild (Werry and Aman, 1975), but not severe, forms of aggression (Conners et al., 1971). More recent research indicates that methylphenidate is effective in a clinical population of children diagnosed with conduct disorder (Klein et al., 1997). In this study, however, aggression was not a target symptom, and the instrument used, the Inattention/ Overactivity With Aggression Conners Teacher Rating Scale (Goyette et al., 1978), does not differentiate between predatory and impulsive types of aggression. Certainly, in milder cases of aggressive behavior, a trial of stimulants would be indicated. and the lack of multiple blind raters. Because the study was not controlled, the observed improvement could have been attributable solely to nonspecific treatment factors (e.g., frequent visits to the clinic, attention to behavioral problems) or the passage of time. However, in spite of these shortcomings, our findings are in agreement with a controlled study of fluoxetine in adults with personality disorder and impulsive aggression (Coccaro et al., 1997). Our results need to be replicated in a larger sample of patients under double-blind and placebo-controlled conditions. Clinical Implications and Conclusions Although these considerations may limit our findings with regard to efficacy, this study presents the first systematic report of citalopram in children and adolescents with impulsive aggression. Clearly, a more extensive evaluation of the reliability and validity of the impulsive aggression model would be necessary. In this study, citalopram 20 to 40 mg daily was associated with clinically and statistically significant reduction in the impulsive aggression dimension irrespective of DSM-IV diagnostic category. These preliminary results are favorable and suggest that citalopram is effective and well tolerated in children and adolescents. A critical assessment of efficacy and safety of citalopram under doubleblind and placebo-controlled conditions seems justified. REFERENCES Some of the limitations of this study include the small number of subjects, the uncontrolled nature of the trial, Achenbach TM, Howell CT (1993), Are American children’s problems getting worse? A 13-year comparison. J Am Acad Child Adolesc Psychiatry 32:1145–1154 American Psychiatric Association (1994), Diagnostic and Statistical Manual of Mental Disorders, 4th edition (DSM-IV). Washington, DC: American Psychiatric Association Asberg M, Traksman L, Thoren P (1976), 5 HIAA in the cerebrospinal fluid: a biochemical suicide predictor? Arch Gen Psychiatry 33:1193–1197 Atkins MS, Stoff DM, Osborne ML, Brown K (1993), Distinguishing instrumental and hostile aggression: does it make a difference? J Abnorm Child Psychol 21:355–365 Barratt ES, Stanford MS, Kent TA, Felthous A (1997), Neuropsychological and cognitive psychophysiological substrates of impulsive aggression. Biol Psychiatry 41:1045–1061 Brener ND, Simon TR, Krug EG, Lowry R (1999), Recent trends in violencerelated behaviors among high school students in the United States. JAMA 282:440–446 Brown GL, Goodwin FK, Ballenger JC, Goyer PF, Major LF (1979), Aggression in humans correlates with cerebrospinal fluid amine metabolites. Psychiatry Res 1:131–139 Campbell M, Small AM, Green WH et al. (1984), Behavioral efficacy of haloperidol and lithium carbonate: a comparison in hospitalized aggressive children with conduct disorder. Arch Gen Psychiatry 41:650–656 Coccaro EF, Harvey PD, Kupsaw-Lawrence E, Herbert JL, Bernstein DP (1991), Development of neuropharmacologically based behavioral assessments of impulsive aggressive behavior. J Neuropsychiatry Clin Neurosci 3:S44–S51 528 J . A M . A C A D . C H I L D A D O L E S C . P S YC H I AT RY, 4 1 : 5 , M AY 2 0 0 2 Limitations CITALOPRAM FOR IMPULSIVE AGGRESSION Coccaro EF, Kavoussi RJ (1997), Fluoxetine and impulsive aggressive behavior in personality-disordered subjects. Arch Gen Psychiatry 54:1081–1088 Coccaro EF, Siever LJ, Klar HM et al. (1989), Serotonergic studies in patients with affective and personality disorders. Arch Gen Psychiatry 46:587–599 Conners CK (1995), Conners’ Continuous Performance Test II: Computer Program for Windows; Technical Guide and Software Manual. Toronto: Multi-Health Systems Conners CK, Kramer R, Rothschild GH, Schwartz L, Stone A (1971), Treatment of young delinquent boys with diphenylhydantoin sodium and methylphenidate. Arch Gen Psychiatry 24:156–160 Cueva JE, Overall JE, Small AM, Armenteros JL, Perry R, Campbell M (1996), Carbamazepine in aggressive children with conduct disorder: a doubleblind and placebo-controled study. J Am Acad Child Adolesc Psychiatry 35:480–490 Fava M, Rosenbaum JF, Pava JA, McCarthy MK, Steingard RJ, Bouffides E (1993), Anger attacks in unipolar depression, I: clinical correlates and response to fluoxetine treatment. Am J Psychiatry 150:1158–1163 Frick PJ, Lahey BB, Applegate B et al. (1994), DSM-IV field trials for the disruptive behavior disorders: symptom utility estimates. J Am Acad Child Adolesc Psychiatry 33:529–539 Gadow KD, Sprafkin J (2000), Child Symptom Inventories Manual. Stony Brook, NY: Checkmate Plus Gordon CT, State RC, Nelson JE, Hamburger SD, Rapoport JL (1993), A double-blind comparison of clomipramine, desipramine, and placebo in the treatment of autistic disorder. Arch Gen Psychiatry 50:441–447 Goyette CH, Connors CK, Ulrich RF (1978), Normative data on revised Conners parent and teacher rating scales. J Abnorm Child Psychol 6:221–236 Halperin JM, Newcorn JH, Matier K, Bedi G, Hall S, Sharma V (1995), Impulsivity and the initiation of fights in children with disruptive behavior disorders. J Child Psychol Psychiatry 36:1199–1211 Halperin JM, Sharma V, Siever LJ et al. (1994), Serotonergic function in aggressive and nonaggressive boys with attention deficit hyperactivity disorder. Am J Psychiatry 151:243–248 Higley JD, Mehlman PT, Taub DM et al. (1992), Cerebrospinal fluid monoamine and adrenal correlates of aggression in free-ranging rhesus monkeys. Arch Gen Psychiatry 49:436–441 Hyttel J (1982), Citalopram: pharmacological profile of a specific serotonin uptake inhibitor with antidepressant activity. Prog Neuropsychopharmacol Biol Psychiatry 6:277–295 Katz RJ (1980), Role of serotonergic mechanisms in animal models of predation. Prog Neuropsychopharmacol Biol Psychiatry 4:219–231 Kaufman AS, Kaufman NL (1990), Kaufman Brief Intelligence Test. Circle Pines, MN: American Guidance Services Klein RG, Abikoff H, Klass E, Ganeles D, Seese LM, Pollck S (1997), Clinical efficacy of methylphenidate in conduct disorder with and without attention deficit hyperactivity disorder. Arch Gen Psychiatry 54:1073–1080 Kruesi MJ, Rapoport JL, Hamburger S et al. (1990), Cerebrospinal fluid monoamine metabolites, aggression, and impulsivity in disruptive behavior disorders of children and adolescents. Arch Gen Psychiatry 47:419–426 Linnoila M, Virkunen M, Scheinin M, Nuutila A, Rimon R, Goodwin FK (1983), Low cerebrospinal fluid 5-hydroxyindolacetic acid concentration differentiates impulsive from nonimpulsive violent behavior. Life Sci 33:2609–2614 Malone RP, Bennett DS, Luebbert JF et al. (1998), Aggression classification and treatment response. Psychopharmacol Bull 34:41–45 Malone RP, Luebbert JF, Delaney MA et al. (1997), Nonpharmacological response in hospitalized children with conduct disorder. J Am Acad Child Adolesc Psychiatry 36:242–247 Mann JJ, McBride PA, Malone KM, DeMeo MD, Keilp J (1995), Blunted serotonergic responsivity in depressed patients. Neuropsychopharmacology 13:53–64 Mauchly JW (1940), Significance test for sphericity of a normal n-variate distribution. Annu Math Stat 11:204–209 McDougle C, Nylor S, Cohen D, Volkmar F, Heninger G, Price L (1996), A double-blind, placebo-controlled study of fluvoxamine in adults with autistic disorder. Arch Gen Psychiatry 53:1001–1008 Meltzer HY, Lowy MT (1987), The serotonin hypothesis of depression. In: Psychopharmacology: Third Generation of Progress, Meltzer HY, ed. New York: Raven, pp 513–526 Pine DS, Coplan JD, Wasserman GA et al. (1997), Neuroendocrine response to fenfluramine challenge in boys: associations with aggressive behavior and adverse rearing. Arch Gen Psychiatry 54:839–846 Psychopharmacology Bulletin (1985), Special feature: rating scales and assessment instruments for use in pediatric psychopharmacology research 21:731–733 Public Health Service (1991), Healthy People 2000: National Health Promotion and Disease Prevention Objectives—Full Report With Commentary. Washington, DC: US Department of Health and Human Services Seroczynski AD, Bergeman CS, Coccaro EF (1999), Etiology of the impulsivity/aggression relationship: genes or environment? Psychiatry Res 86:41–57 Sewell RG, Galus JA, Gault FP, Cleary JP (1982), P-Chlorophenylalanine effects on shock-induced attack and pressing responses in rats. Pharmacol Biochem Behav 17:337–340 Sheard MH, Marini JL, Bridges CI, Wagner E (1976), The effect of lithium on impulsive aggressive behavior in man. Am J Psychiatry 133:1409–1413 Siegel A, Roeling T, Gregg TR, Kruk MR (1999), Neuropharmacology of brain-stimulation-evoked aggression. Neurosci Biobehav Rev 23:359–389 Tiihonen J, Schwartz E, Petrilli R (1993), Eltoprazine for aggression in schizophrenia and mental retardation. Lancet 341:307 van Praag HM, Kahn RS, Asnis GM et al. (1987), Denosologization of biological psychiatry or the specificity of 5-HT disturbances in psychiatric disorders. J Affect Disord 13:1–8 Vartiainen H, Tiihonen J, Putkonen A et al. (1995), Citalopram, a selective serotonin reuptake inhibitor, in the treatment of aggression in schizophrenia. Acta Psychiatr Scand 91:348–351 Vitiello B, Behar D, Hunt J, Stoff D, Ricciuti A (1990), Subtyping aggression in children and adolescents. J Neuropsychiatry Clin Neurosci 2:189–192 Vitiello B, Stoff DM (1997), Subtypes of aggression and their relevance to child psychiatry. J Am Acad Child Adolesc Psychiatry 36:307–315 Volavka J, Crowner M, Brizer D (1990), Tryptophan treatment of aggressive psychiatric inpatients. Biol Psychiatry 28:728–732 Werry JS, Aman MG (1975), Methylphenidate and haloperidol in children: effects on attention, memory and activity. Arch Gen Psychiatry 32:790–795 White JL, Moffitt TE, Caspi A, Bartuch JB, Needles DJ, Stouthamer-Loeber M (1994), Measuring impulsivity and examining its relationship to delinquency. J Abnorm Psychol 103:192–205 J . A M . A C A D . C H I L D A D O L E S C . P S YC H I AT RY, 4 1 : 5 , M AY 2 0 0 2 529