Available online at www.sciencedirect.com

Comprehensive Psychiatry 53 (2012) 238–244www.elsevier.com/locate/comppsych

Childhood trauma, impulsivity, and executive functioningin crack cocaine users

Joana C.M. Narvaeza,b,c, Pedro V.S. Magalhãesa,d, Elise K. Trindadee, Daniel Chaves Vieiraa,d,Márcia Kauer-Sant'Annaa,d, Clarissa S. Gamaa,b, Lisia von Diemend,f,

Natália S. Kapczinskia,d, Flavio Kapczinskia,d,⁎aBipolar Disorders Program and INCT Translational Medicine, Hospital de Clínicas de Porto Alegre,

Universidade Federal do Rio Grande do Sul, Porto Alegre CEP 90035-003, BrazilbPostgraduate Program in Medicine: Medical Sciences, Universidade Federal do Rio Grande do Sul, Porto Alegre CEP 90035-003, Brazil

cCruz Vermelha Brasileira, Filial Rio Grande do Sul, CEP 90035-003, BrazildPostgraduate Program in Medicine: Psychiatry, Universidade Federal do Rio Grande do Sul, Porto Alegre, Porto Alegre CEP 90035-003 Brazil

eInstituto de Diagnóstico e Tratamento em Psicologia, Porto Alegre CEP 90035-003, BrazilfCenter for Drug and Alcohol Research, Hospital de Clínicas de Porto Alegre, Universidade Federal do Rio Grande do Sul, Porto

Alegre CEP 90035-003, Brazil

Abstract

Background: The use of crack cocaine is a major public health concern in Brazil and internationally. Recent data suggest that childhoodtrauma is associated with worse outcomes among cocaine users. This study had the objective of evaluating the relationship of childhoodtrauma with executive functioning and impulsivity in outpatients with crack cocaine use disorders.Methods: This is a cross-sectional study of 84 consecutive outpatients with a primary crack cocaine use disorder who sought treatment inPorto Alegre, Brazil. Childhood trauma was evaluated with the Childhood Trauma Questionnaire; executive functioning, with the WisconsinCard Sorting Test; and impulsivity, with the Barratt Impulsivity Scale.Results: Childhood trauma was strongly associated with executive dysfunction and impulsivity, even when controlled for possible confounders.Conclusions: Childhood trauma may be associated with executive dysfunction and impulsivity in crack cocaine users. The full impact oftrauma needs to be further investigated in longitudinal studies.© 2012 Elsevier Inc. All rights reserved.

1. Introduction

The use of crack cocaine emerged in the United Statesaround 1985 as a major public health problem, and the firstcases in Brazil were detected in 1988 [1,2]. The currentlifetime prevalence of crack cocaine use in Brazil isestimated at 0.7% [3]. Although internationally the preva-lence of use is generally low, use of crack cocaine tends to befrequent in some populations [4,5].

Cognitive changes in crack cocaine users are of majorinterest, as they may be associated with treatment adherence,impulse control, and overall functioning. Studies of

⁎ Corresponding author. Laboratório de Psiquiatria Molecular, Hospitalde Clinicas de Porto Alegre (HCPA), Ramiro Barcelos, 2350, CEP 90035-003, Porto Alegre, RS Brazil. Tel.: +55 51 21018845; fax: +55 5121018846.

E-mail address: kapcz@terra.com.br (F. Kapczinski).

0010-440X/$ – see front matter © 2012 Elsevier Inc. All rights reserved.doi:10.1016/j.comppsych.2011.04.058

cognitive functioning in cocaine users found deficits inmemory [6], attention [7], learning, executive functioning,and verbal fluency [8]. These findings, however, are notuniversal [9,10]. The discrepancies in these results may beexplained by differences in the tests used to evaluatepatients. For example, the results of verbal fluency, whichwas found to be impaired in the study conducted by Cunhaet al [8] but improved in the studies by Hoff et al [9] andO'Malley et al [10], might be affected by the inclusion ofoutpatients or long-term patients under the effect of thesubstance, as cocaine may stimulate verbal productionbecause of dopaminergic agonist modulation and result inan improvement on verbal measures [10,11]. All in all, thestudies indicate that persistent use of crack may causeimprovement in some measures but deterioration in most [9].This leads to more dysfunction and reinforces responses thatare inflexible and poorly adaptive.

e 1ibutions of demographic and clinical characteristics in crack cocaine(n = 84)

acteristic

(y) 29.31 ± 7.62g with partner 25%en 8%e skin 69%ationh school or higher 39%rs of education 7.65 ± 3.51pation

239J.C.M. Narvaez et al. / Comprehensive Psychiatry 53 (2012) 238–244

Neuropsychologic studies may help in the development oftreatment strategies that address neurobehavioral performanceand executive functioning deficits [12]. For example, thedecrease in the capacity to deal with stressors and in emotionalmodulation, functions associated with the prefrontal andtemporal cortex, may reinforce the noxious cycle of chemicaldependence [13]. Crack cocaine affects the electrophysiologicand metabolic activities of the frontal and temporal regions[14], which leads to more neuropsychologic and physiologicdysfunction and reinforces responses that are rigid, inflexible,and poorly adapted to reality. Moreover, the long-termadministration of psychoactive substances changes brainplasticity and reward circuitry [15,16] and is correlated tostructural changes in specific areas of the brain and,consequently, long-lasting addictive behaviors [17].

Childhood maltreatment has been found to be highlyprevalent among crack cocaine users and a strong predictorof drug dependence [18-21]. Recent studies, such as the oneconducted by Savitz et al [22], showed that scores in achildhood sexual abuse scale were negatively associatedwith performances in memory tests. Impulsive behaviorshave been shown to be significant in the initiation andperpetuation of drug use [23] as well as a predictor of poorprognoses. Other studies have suggested that impulsivity isan outcome of traumatic childhood experiences [24,25].

The primary purpose of this study was to evaluate theassociation of childhood trauma with executive functioningand impulsivity among outpatients with a primary diagnosisof crack cocaine misuse. To our knowledge, data on thesespecific hypotheses are lacking at present.

rking/studying 40%working 56%sewife 1%welfare 2%

81.63 ± 11.18at onset (y)substance 13.6 ± 2.9

ck cocaine 24.2 ± 6.5k cocaine uses since last use 3 (1-30)ly use 83%ms per day 10 (5-16)rs of use 3.70 (3.67)s of use of any substance 15.60 (7.92)ently abstinent 10.7%ent useacco 87%ohol 68%nabis 60%aine 37%alants 5%phetamine 2%lucinogens 2%zodiazepines 12%ates 1%ious suicide attempts 26.2%hood trauma 67.1%ime comorbiditiesod disorders 49%er comorbidities 11%

uencies were shown as mean ± SD or median (interquartile range).

2. Methods

This cross-sectional study enrolled 84 consecutive patientsin a program for voluntary treatment of crack cocaine usedisorders in the city of Porto Alegre, a capital in southernBrazil, where crack/cocaine misuse has emerged as a publichealth issue because of its severity and prevalence [3]. Patientswere seen in a Red Cross outpatient facility that specializes inthe treatment of substance use disorders. Red Cross is anongovernmental organization that provides assistance acrossthe globe with the aim of preventing and relieving suffering.The program usually consists of a transdisciplinary approachand includes clinical consults, support groups, and psychoe-ducation. After an initial triage, treatment is prescribedaccording to the patient's needs. Staff in the clinic isvoluntary, and health care is provided at no cost to thepatient. Patients were included in the study if their primarydiagnosis was crack cocaine abuse or dependence. As this wasa naturalistic study, exclusion criteria were kept to aminimum; and multidrug users were not excluded. Patientswere excluded only if they had used any substance on the dayof the interview or refused to provide informed consent. Thisresulted in 8 exclusions, 1 for using crack in the day of theinterview and 7 for not concluding the study protocol. The

diagnoses of substance use disorders were made uponenrollment in the program and were confirmed in a clinicalinterview. Axis I comorbidities were also confirmed by apsychiatrist in a clinical interview based on the StructuredClinical Interview for the Diagnostic and Statistical Manualof Mental Disorders, Fourth Edition. All diagnoses wereassigned by 6 master's level psychiatrists, who had intensivetraining in using the Structured Clinical Interview for theDiagnostic and Statistical Manual of Mental Disorders,Fourth Edition. The study was approved by the local ethicscommittee, and informed consent was obtained from allparticipants before the beginning of procedures.

Patients went through a structured clinical interview toassess the use of crack. They were grouped according to thefrequency and intensity of use of crack cocaine as follows:

1. Abstinence: At least 6 months without any use ofcrack cocaine.

TablDistrusers

Char

AgeLivinWomWhitEducHigYeaOccuWoNotHouOnIQAgeAnyCraCracDayDaiGraYeaYearCurrCurrTobAlcCanCocInhAmHalBenOpiPrevChildLifetMoOth

Freq

240 J.C.M. Narvaez et al. / Comprehensive Psychiatry 53 (2012) 238–244

2. Intensity: Related to use of crack. Patients weredivided into 2 groups: 6 to 12 g/d; more than 12 g/d.

3. Frequency: Patients were divided into daily users andless frequent users.

4. Continuous use: No periods of abstinence in crack usefor the last 2 years.

5. Years of use: The total years of use of crack subtractedfrom total periods of abstinence.

Childhood trauma was assessed with the validatedPortuguese version of the Childhood Trauma Questionnaire(CTQ) [26]. Patients were grouped as “patients who hadtrauma in childhood” and “patients who did not have trauma”according to CTQ criteria. The trauma category includedthose with moderate or severe trauma in any of the CTQsubscales scores (emotional abuse or neglect, physical abuseor neglect, and sexual abuse) as originally proposed by theauthors [27]. Two patients with extreme responses on thedenial scale were excluded.

Impulsivity was assessed using the Barratt ImpulsivityScale (BIS 11) [28], also validates to Portuguese [29]. TheBIS 11 is a 30-item questionnaire with a Likert-like scalefrom 1 (rarely/never) to 4 (always/almost always); its resultsprovide a total impulsivity score that ranges from 30 to 120.

2.1. Neuropsychologic testing

The Wisconsin Card Sorting Test (WCST) [30] is aneuropsychologic test of executive function. It measuresthought flexibility in generating problem-solving strategies.

Table 2Effect sizes for associations of WCST domains and impulsivity with disease char

Characteristic Total correct trials Percentage ofperseverative error

Age † NS NSSex NS NSYears of education † 0.26⁎ NSAge at onsetAny substance † 0.22⁎ NSCrack cocaine † NS NS

Currently abstinent NS NSCrack cocaine use NSHigh intensity 0.56⁎ NSDaily use NS NSYears of use † NS NSNo periods of abstinence 0.65⁎⁎ NS

Current comorbidity NSTobacco NS NSAlcohol NS 0.57⁎

Other illicit substances 0.49⁎ 0.72⁎⁎

Any mental disorder NS NSPrevious suicide attempts NS NSIQ † 0.54⁎⁎ −0.33⁎⁎Digit span † 0.41⁎⁎ −0.33⁎⁎Childhood trauma 0.64⁎⁎ 0.48⁎

Results shown are either Hedges effect sizes for dichotomous predictors (sex, crack coor Spearman coefficients for continuous predictors (age, years of use, IQ, and digitdifferences between groups; and Spearman rank correlation, for association between

⁎ P b .05.⁎⁎ P b .01.

It provides an assessment of the capacity to establish,keep, and change mental categories [31]. It was adminis-tered by experienced psychologists blind to all otherclinical measures. The WCST domains used in theanalyses were those deemed more reliable in assessingexecutive function [32]: number of total correct trials andcorrectly completed categories, the percentage of persev-erative errors and conceptual level responses. TheWechsler Adult Intelligence Scale III [33] was usedaccording to the following subscales: digit spans, evaluatesimmediate memory, capacity of attention, and concentra-tion; vocabulary, measures the capacity to conceptualizeand verbal comprehension; cubes, spatial orientation,analysis, and synthesis [34]. These measures representthe verbal and executive functioning factors and providean approximate measure of functioning IQ. Neuropsycho-logic testing was applied by 4 master's level psychologistswith experience in neuropsychologic testing. In addition,all had specific training on the tests by one of the seniorauthors (NK).

2.2. Statistical analysis

As some variables were not normally distributed, theywere transformed for parametric analysis. Correctly com-pleted categories from the WCST were squared andperseverative errors were log transformed. We used analysisof variance to test for significant associations with executivefunctioning in the bivariate analysis.

acteristics in crack cocaine users (n = 84)

sConcept levelresponses

Correctly completedcategories

BIS 11 score

NS NS NSNS NS NS0.27⁎ 0.28⁎ NS

0.22⁎ 0.22⁎ NSNS NS NSNS NS NS

0.60⁎⁎ 0.68⁎⁎ NSNS NS NSNS NS NS0.61⁎ 0.70⁎⁎ NS

NSNS NS NSNS NS NS0.46⁎ 0.54⁎ NSNS NS NSNS NS NS0.55⁎⁎ 0.55⁎⁎ NS0.42⁎⁎ 0.42⁎⁎ NS0.73⁎⁎ 0.84⁎⁎ 0.81⁎⁎

caine daily use, no periods of abstinence, comorbidities, and childhood trauma)span) (†). Analysis of variance was used to test for statistical significance of2 continuous variables. NS indicates not statistically significant.

Table 3Bivariate associations between childhood trauma and demographic andclinical characteristics of crack cocaine users

Characteristic Childhood trauma

Absent Present

Age (y) 29.04 ± 5.04 29.20 ± 8.46Women 0% 12.7%High school education⁎⁎ 77.8% 20.0%Years of education⁎⁎ 10.42 ± 2.08 6.20 ± 3.25Age at onset ofAny substance 14.19 ± 2.34 13.42 ± 3.18Crack cocaine 24.81 ± 5.18 23.92 ± 7.16Currently abstinent 3.7% 14.5%Crack cocaine useHigh intensity 63.0% 50.0%Daily use 77.8% 84.9%Years of use 3.38 ± 2.78 3.89 ± 4.09No periods of abstinence 22.2% 32.7%Current comorbidityTobacco 63.6% 67.6%Alcohol⁎ 85.2% 58.2%Other illicit substances 72.0% 64.9%Any mental disorder 75% 62%IQ⁎ 85.17 ± 11.40 79.90 ± 10.94Digit span⁎ 6.54 ± 2.06 5.58 ± 1.82Previous suicide attempts 14.8% 32.7%

⁎ χ2 Test or analysis of variance; P b .05.⁎⁎ P b .01.

241J.C.M. Narvaez et al. / Comprehensive Psychiatry 53 (2012) 238–244

For the linear regression model, a principal componentanalysis (PCA) was used to extract the shared variance of the4 WCST domains. We termed this variable executivefunctioning, and it was the outcome of the linear regression.Because the sample was relatively small, the variables usedin the model were those that had a bivariate association withexecutive functioning, in addition to childhood trauma,which was the predictor variable of interest in this report.There was a strong association between estimated IQ anddigit span, and only IQ was included in the regression model.All tests were 2 tailed.

3. Results

The sample included 84 outpatients with crack cocaineuse disorders. Childhood trauma was reported by 67% of thepatients. Of note, every women in this sample had a historyof childhood trauma (χ2 = 3.76; P = .089). Table 1 showssociodemographic and clinical variables of the sample.

Median raw scores for WCST domains were 77 (60-89) fortotal correct trials, 3 (1-6) for correctly completed categories,18.4 (13.5-26.6) for the percentage of perseverative errors,and 47.3 (30.5-60.2) for conceptual level responses. Bivariateassociations of these subcategories are shown in Table 2.Childhood trauma was associated with worse executivefunctioning in the 4 categories. Childhood trauma wasassociated with digit span (F = 4.47; P = .038) and a 5-point difference in estimated IQ (F = 3.99; P = .049), but thisdifference disappeared when controlled for education (P N .05

for both). Table 3 displays overall differences betweenpatients with and without a history of childhood trauma.

The linear regression model of overall executive func-tioning (ie, extracted with the PCA) explained 44% of thevariance. In this model, predictors were childhood trauma(B = −0.32; 95% confidence interval [CI], −0.65 to −0.01;P = .003), IQ (B = 0.51; 95% CI, 0.03-0.060; P b .001),continuous crack use (B = −0.18; 95% CI, −1.11 to −0.23;P = .04), and multidrug use (B = −0.18; 95% CI, −0.78 to0.00; P = .05). Age at onset of substance abuse, intensity ofuse, and education were removed from the model (P N .05).

Impulsivity was associated only with childhood trauma(F = 9.05; P = .004) and education (F = 4.82; P = .032); it wasnot associatedwith any of theWCST domains (P N .05 for all).Trauma remained associated with impulsivity when controlledfor education (B = 6.98; 95% CI, 0.36-13.60; P = .039).

4. Discussion

In this report, childhood trauma was strongly associatedwith poor executive functioning and higher levels ofimpulsivity among crack cocaine users. Previous studieshave indicated that childhood trauma is an indicator of theseverity of prognosis and progression of cocaine use [35]. Inthis study, we suggest that trauma may be an important factorin the psychopathology of crack cocaine use.

We found that 67% of the patients reported childhoodtrauma. Although this prevalence was high, it agrees withprevious studies, where the rate of childhood abuse amongdrug users was about 50% [36]. Studies on the associationbetween trauma and crack cocaine use are usually focused onthe female population [18,19,21,37], although epidemiolog-ical surveys point to a greater incidence of crack cocaine useamong men [3,38-40]. Other studies found a strongcorrelation between childhood abuse and drug dependence[36,37], particularly in the case of crack cocaine [18,19]. Thepresent study report increased rates of trauma in a pre-dominantly male population.

The most marked difference between the groups of crackcocaine users with and without trauma was not in IQ orattention, but a difference was the emotional adaptive andbehavioral organization in terms of functionality as a whole.This finding points to difficulties in more complex skillsinterconnecting the cognitive aspects and the principles oforganization necessary to deal with fluctuating and ambig-uous situations in social relationships [31,41]. Our resultssuggest that childhood maltreatment may affect how crackcocaine users cope with difficulties and apply neuropsycho-logic resources to their everyday lives.

A developmental perspective would suggest that theseresults are related to the effects of the original trauma onself-regulatory systems. These modulate the tolerance toaversive emotional states, heightening subjective stress andpossibly affecting self-esteem, confidence, and sense ofcontrol [42-47]. In the short and long term, they may lead to

242 J.C.M. Narvaez et al. / Comprehensive Psychiatry 53 (2012) 238–244

psychologic [48] and cognitive impairments [22,49]. Thesewould all contribute to create a neuropsychologic vulner-ability that could make the substance attractive as a form ofself-medication [13,50], in an attempt to regulate emotion[47,51]. In addition, studies indicating a greater prevalenceof psychiatric disorders, particularly bipolar disorder[52,53] among patients who experienced childhood trauma,support that notion [54,55].

This understanding may be particularly relevant to peoplewith a diagnosis of crack cocaine dependence, wherecomorbidity with other psychiatric disorders is a rule. It hasbeen postulated that affective instability may be related to asearch for affective regulation via substance use. A relatedargument has been that substance abuse may be a part ofseries of behaviors that intend to alleviate painful traumaticmemories [56], thus creating a cycle of retraumatization [20].

Childhood trauma was also associated with impulsivebehaviors in crack cocaine users. Roy [24] used the samescales and found similar associations between scores, whichsuggest that childhood trauma may be a determinant factorin adult impulsivity [25]. Impulsivity seems to be anunfavorable outcome of trauma for crack cocaine users and arisk factor for the development of a pathologic response totraumatic events.

Continuous drug use may lead to a worse performance inneuropsychologic measures [8]. In our sample, continuoususe of crack cocaine for at least 2 years led to deficits in mostof the WCST neuropsychologic measures, whereas years ofuse did not. This might be due to periods of abstinence. Suchfinding may be of clinical relevance for the treatment fordrug users. In the cases in which use discontinuation is hardto achieve, periods of abstinence may be a damage reductionmeasure, which seems to protect against more markedimpairment or, at least, delay neuropsychologic deficits.

Most of the patients in this study were multidrug users, asin other studies in the literature [57-60]. The association ofcrack cocaine with other illicit drugs reinforces the reducedplasticity to respond to environmental contingencies and theperseverance in using inadequate responses as well as agreater difficulty in forming concepts and in the search formore effective strategies. However common this associationmay be, the fact that trauma remained a significant predictorafter controlling for the use of other drugs strengthens thisresult. As part of the naturalistic design, some patients alsohad psychiatric comorbidities. These, however, were notassociated with executive functioning in this sample.

One of the limitations of this study was the fact thatchildhood maltreatment and impulsivity were based on aself-reported scale, although one with extensive use in thepsychiatric literature [26]. Moreover, our patients were nothospitalized; and, in most cases, we counted only on patientreports to retrieve information about their history of drug useand abuse. Nevertheless, we believe that faking and omittinginformation should not be a major source of bias becauseparticipants sought our service voluntarily and chose toparticipate in this study. Furthermore, the findings here

pertain to men who abuse crack cocaine because theycomprised most of the sample. We did not use controlsubjects to establish comparisons between users and thepopulation in general, and the correlation data were obtainedfrom the study sample. This is an important limitationbecause this way we cannot establish whether this effect oftrauma is specific to this population. Previous studies,however, indicate that the frequency of executive dysfunc-tion in long-term crack cocaine users is higher than inhealthy controls [7].

In conclusion, a history of childhood trauma was stronglycorrelated with worse clinical outcomes for crack cocaineusers—including poorer executive functioning and greaterimpulsivity. One hypothesis for understanding these data isthat this interaction perpetuates a pattern that fails todiscontinue this cycle of self-destructive behavior andreinforces the harmful effects of dependence. Childhoodtrauma was a component of poorer prognoses for drug usersas a possible predictor of severity of neuropsychologic andbehavioral impairment. Our results are potentially relevantfor clinical practice because the early identification andtreatment of children that experience maltreatment may helpto prevent the vulnerability to life stressors or thedevelopment of a less adaptive coping style.

Acknowledgment

Dr. Magalhães is supported by a doctoral scholarshipfrom Coordenação de Aperfeiçoamento de Pessoal de NívelSuperior (CAPES), Brazil.

Clarissa Severino Gama has received research grants fromCNPq-Universal (477974/2009-0), CNPq-INCT-TM,CAPES, and FIPE-HCPA. She has been a paid speaker forLundbeck; Actelion Pharmaceuticals, Ltd; and Astra Zenecaand a consultant for Actelion Pharmaceuticals, Ltd. She hasreceived travel award from Lilly and Janssen.

Dr. Kapczinski has received grant/research support fromAstra-Zeneca, Eli Lilly, the Janssen-Cilag, Servier, CNPq,CAPES, NARSAD, and the Stanley Medical ResearchInstitute; has been a member of the speaker's boards forAstra-Zeneca, Eli Lilly, Janssen, and Servier; and has servedas a consultant for Servier. The other authors declare thatthey have no conflict of interest in the matter. ClarissaSeverino Gama is funded by research grants from CNPq(Universal 477974/2009-0 and PQ 305967/2008-8), Brazil.

References

[1] Siegel RK. Cocaine smoking. J Psychoact Drugs 1982(14):271-359.[2] Leite MC, de Andrade AG. Cocaína e Crack: dos fundamentos ao

tratamento. Porto Alegre: Artes Médicas; 1999.[3] Carlini EA, Galduróz JC, Noto AR, Nappo SA. II Levantamento

Domiciliar Sobre o Uso de Drogas Psicotrópicas no Brasil - 2005. SãoPaulo: CEBRID – Centro Brasileiro de Informações sobre DrogasPsicotrópicas – Universidade Federal de São Paulo; 2006.

[4] Smart RG. Crack cocaine use: a review of prevalence and adverseeffects. Am J Drug Alcohol Abuse 1991;17(1):13-26.

243J.C.M. Narvaez et al. / Comprehensive Psychiatry 53 (2012) 238–244

[5] Miech RA, Chilcoat H, Harder VS. The increase in the association ofeducation and cocaine use over the 1980s and 1990s: evidence for a“historical period” effect. Drug Alcohol Depend 2005;79(3):311-20[Epub 2005 Mar 25].

[6] Rodrigues VS, Caminha RM, Horta RL. Déficits cognitivos empacientes usuários de crack. Revista Brasileira de Terapias Cognitivas2006(2):67-72.

[7] Pace-Schott EF, Morgan PT, Malison RT, Hart CL, Edgar C, WalkerM, et al. Cocaine users differ from normals on cognitive tasks whichshow poorer performance during drug abstinence. Am J Drug AlcoholAbuse 2008(34):109-21.

[8] Cunha PJ, Nicastri S, Gomes LP, Moino RM, Peluso MA. Neuropsy-chological impairments in crack cocaine–dependent inpatients: prelim-inary findings. Revista Brasileira de Psiquiatria 2004(26):103-6.

[9] Hoff AL, Riordan H, Morris L, Cestaro V, Wieneke M, Alpert R, et al.Effects of crack cocaine on neurocognitive function. Psychiatry Res1996(60):167-76.

[10] O'Malley S, Adamse M, Heaton R, Gawin FH. Neuropsychologicalimpairment in chronic cocaine abusers. Am J Drug Alcohol Abuse1992(18):131-44.

[11] Goldberg TE, Bigelow LB, Weinberger DR, Daniel DG, Kleinman JE.Cognitive and behavioral effects of the coadministration of dextroam-phetamine and haloperidol in schizophrenia. Am J Psychiatry1991;148:78-84.

[12] Bolla KI, Cadet JL, London ED. The neuropsychiatry of chroniccocaine abuse. J Neuropsychiatry Clin Neurosci 1998(10):280-9.

[13] Khantzian EJ. Impulse problems and drug addiction: cause and effectrelationships. In: Wishnie H, & Nevis-Olesen J, editors. Working withthe impulsive person. New York: Plenum Press; 1979.

[14] Pascual-Leone A, Dhuna A, Anderson DC. Longterm neurologicalcomplications of chronic, habitual cocaine abuse. Neurotoxicoiogy1991;12:393-400.

[15] Lee KW, Kim Y, Kim AM, Helmin K, Nairn AC, Greengard P.Cocaine-induced dendritic spine formation in D1 and D2 dopaminereceptor-containing medium spiny neurons in nucleus accumbens.Proc Natl Acad Sci U S A 2006;103:3399-404.

[16] Sarti L, Borgland SL, Kharazia VN, Bonci A. Acute cocaine exposurealters spine density and long-term potentiation in the ventral tegmentalarea. Eur J Neurosci 2007;26:749-56.

[17] Russo S, Mazei-Robson M, Ables J, Nestler E. Neurotrophic factorsand structural plasticity in addiction. Neuropharmacology 2009;56(Suppl 1):73-82.

[18] Freeman RC, Collier K, Parillo KM. Early life sexual abuse as a riskfactor for crack cocaine use in a sample of community-recruitedwomen at high risk for illicit drug use. The Am J Drug Alcohol Abuse2002(28):109-31.

[19] Lejuez CW, Bornovalova MA, Reynolds EK, Daughters SB, Curtin JJ.Risk factors in the relationship between gender and crack/cocaine. ExpClin Psychopharmacol 2007(15):165-75.

[20] Fullilove MT, Lown A, Fullilove RB. Crack 'hos and skeezers: traumaticexperiences of women crack users. J Sex Res 1992(29):275-87.

[21] Wallace BC. Crack cocaine smokers as adult children of alcoholics: thedysfunctional family link. J Subst Abuse Treat 1990(7):89-100.

[22] Savitz J, van der Merwe L, Stein DJ, Solms M, Ramesar R. Genotypeand childhood sexual trauma moderate neurocognitive performance: apossible role for brain-derived neurotrophic factor and apolipoproteinE variants. Biol Psychiatry 2007(62):391-9.

[23] Moeller GF, Dougherty DM. Impulsivity and substance abuse: what isthe connection? Addict Dis 2002(1):3-10.

[24] Roy A. Childhood trauma and impulsivity. Possible relevance tosuicidal behavior. Arch Suicide Res 2005(9):147-51.

[25] Brodsky BS, Oquendo M, Ellis SP, Haas GL, Malone KM, Mann JJ. Therelationship of childhood abuse to impulsivity and suicidal behavior inadults with major depression. Am J Psychiatry 2001(158):1871-7.

[26] Grassi-Oliveira R, Stein LM, Pezzi JC. Translation and contentvalidation of the Childhood Trauma Questionnaire into Portugueselanguage. Revista de Saude Publica 2006(40):249-55.

[27] Bernstein D, Fink L. Childhood Trauma Questionnaire. Aretrospective self-report. San Antonio: The Psychological Corpora-tion; 1998.

[28] Patton JH, Stanford MS, Barratt ES. Factor structure of the Barrattimpulsiveness scale. J Clin Psychol 1995(51):768-74.

[29] von Diemen L, Szobot CM, Kessler FH, Pechansky F. Adaptation andconstruct validation of the Barratt Impulsiveness Scale (BIS 11) toBrazilian Portuguese for use in adolescents. Revista Brasileira dePsiquiatria 2007(29):153-6.

[30] Heaton RK, Chleune GJ, Taley JL, Kay GG, Curtiss G. Wisconsin cardsorting test manual (Revised and expanded). Odessa: PsychologicalAssessment Resources; 1993.

[31] Lezak M. Neuropsychological assessment. 3rd ed. New York: OxfordUniversity Press; 1995.

[32] Greve KW, Stickle TR, Love JM, Bianchini KJ, Stanford MS. Latentstructure of the Wisconsin Card Sorting Test: a confirmatory factoranalytic study. Stanford Archives of Clinical Neuropsychology2005(20):355-64.

[33] Wechsler D. WAIS III Test de Inteligencia para adultos. Manual deAdministración y Puntuación. Paidós: Buenos Aires; 2002.

[34] Cunha JA. Psicodiagnóstico – V. Porto Alegre: Artes Médicas; 2000.[35] Hyman SM, Paliwal P, Sinha R. Childhood maltreatment, perceived

stress, and stress-related coping in recently abstinent cocainedependent adults. Psychol Addic Behav 2007(21):233-8.

[36] Dunlap E, Golub A, Johnson BD, Benoit E. Normalization of violence:experiences of childhood abuse by inner-city crack users. J Ethn SubstAbuse 2009(8):15-34.

[37] el-Bassel N, Gilbert L, Schilling RF, Ivanoff A, Borne D, Safyer SF.Correlates of crack abuse among drug-using incarcerated women:psychological trauma, social support, and coping behavior. Am J DrugAlcohol Abuse 1996(22):41-56.

[38] Carlini EA, Galduróz JC, Noto AR, Nappo SA. I LevantamentoDomiciliar Sobre o Uso de Drogas Psicotrópicas no Brasil - 2001. SãoPaulo: CEBRID – Centro Brasileiro de Informações sobre DrogasPsicotrópicas; 2002.

[39] Substance Abuse and Mental Health Services Administration.Results from the 2003 National Survey on Drug Use and Health:national findings. Rockville, MD: Office of Applied Studies,NSDUH Series H-25, DHHS Publication No. SMA 04–3964; 2004.

[40] UNODCCP - United Nations Office for Drug Control and CrimePrevention. Global illicit drug trends 2001 [online]. Vienna:UNODCCP; 2001. Available from: URL: http://www.undcp.org/adhoc/report_2001-06-26_1/report_2001-06-26_1.pdf.

[41] Duffy JD, Campbell III JJ. The regional prefrontal syndromes: atheoretical and clinical overview. J Neuropsychiatry Clin Neurosci1994(6):379-87.

[42] Braquehais MD, OquendoMA, Baca-García E, Sher L. Is impulsivity alink between childhood abuse and suicide? Compr Psychiatry2010;51:121-9.

[43] Cicchetti D, Toth SL. Child maltreatment. Ann Rev Clin Psychol2005;1:409-38.

[44] De Bellis MD. Developmental traumatology: a contributory mecha-nism for alcohol and substance use disorders. Psychoneuroendocrinol-ogy 2002;27:155-70.

[45] Sinha R. Stress and drug abuse. In: Steckler T, Kalin NH, & Reul JM,editors. Handbook on stress and the brain: Part 2: stress: integrative andclinical aspects. Amsterdam: Elsevier Science; 2005.

[46] Mullen PE, Martin JL, Anderson JC, Romans SE, Herbison GP. Theeffect of child sexual abuse on social, interpersonal and sexual functionin adult life. Br J Psychiatry 1994;165:35-47.

[47] Wills TA, Hirky AE. Coping and substance abuse: a theoretical modeland review of the evidence. In: Zeichnec M, & Eudler NS, editors.Handbook of coping: theory, research, and applications. New York:Wiley; 1996. p. 279-302.

[48] Arnow BA. Relationships between childhood maltreatment, adulthealth and psychiatric outcomes, and medical utilization. J ClinPsychiatry 2004;65(Suppl 12):10-5.

244 J.C.M. Narvaez et al. / Comprehensive Psychiatry 53 (2012) 238–244

[49] Breslau N, Lucia VC, Alvarado GF. Intelligence and other predispos-ing factors in exposure to trauma and posttraumatic stress disorder: afollow-up study at age 17 years. Arch Gen Psychiatry 2006;63(11):1238-45.

[50] Weiss RD, Griffin ML, Mirin SM. Drug abuse as self-medication fordepression: an empirical study. Am J Drug Alcohol Abuse 1992;18:121-9.

[51] Boys A, Marsden J, Strang J. Understanding reasons for drug useamongst young people: a functional perspective. Health Educ ResTheory Pract 2001;16(4):457-69.

[52] Garno JL, Goldberg JF, Ramirez PM, Ritzler BA. Impact of childhoodabuse on the clinical course of bipolar disorder. Br J Psychiatry2005;186:121-5 [Erratum in: The British Journal of Psychiatry2005;186:357].

[53] Leverich GS, Post RM. Course of bipolar illness after history ofchildhood trauma. Lancet 2006;367:1040-2.

[54] Brems C, Johnson M, Neal D, Freemon M. Childhood abuse historyand substance use among men and women receiving detoxificationservices. Am J Drug Alcohol Abuse 2004;30:799-821.

[55] Post RM, Leverich GS. The role of psychosocial stress in the onset andprogression of bipolar disorder and its comorbidities: the need forearlier and alternative modes of therapeutic intervention. DevPsychopathol 2006;18:1181-211.

[56] Wasserman DA, Havassy BE, Boles SM. Traumatic events andposttraumatic stress disorder in cocaine users entering privatetreatment. Drug Alcohol Dependence 1997;46:1-8.

[57] Nappo SA. Baqueros e Craquêros - Um estudo etnográfico sobre oconsumo de cocaína na cidade de São Paulo. São Paulo: EscolaPaulista de Medicina, Universidade Federal de São Paulo; 1996.

[58] Ferri CP. Cocaína: padrões de consumo e fatores associados à procurade tratamento. São Paulo: Escola Paulista de Medicina, UniversidadeFederal de São Paulo; 1999.

[59] Ribeiro M. Estudo de seguimento com usuários de crack: mortalidadedurante cinco anos. São Paulo: Escola Paulista de Medicina,Universidade Federal de São Paulo; 2001.

[60] Abi-Saab D, Beauvais J, Mehm J, Brody M, Gottschalk C, Kosten TR.The effect of alcohol on the neuropsychological functioning of recentlyabstinent cocaine-dependent subjects. Am J Addict 2005(14):166-78.