A Double-Blind, Placebo-Controlled Study of the Opiate Antagonist, Naltrexone, in the Treatment of Kleptomania

A Double-Blind, Placebo-Controlled Study of the Opiate Antagonist, Naltrexone, in the Treatment of Kleptomania Jon E. Grant, Suck Won Kim, and Brian L. Odlaug Background: Kleptomania is a rare psychiatric disorder characterized by recurrent stealing and for which there exists no empirically validated treatments. This study examined the efficacy and tolerability of the opioid antagonist naltrexone in adults with kleptomania who have urges to steal. Methods: An 8-week, double-blind, placebo-controlled trial was conducted to evaluate the safety and efficacy of oral naltrexone for kleptomania. Twenty-five individuals with DSM-IV kleptomania were randomized to naltrexone (dosing ranging from 50 mg/day to 150 mg/day) or placebo. Twenty-three subjects (92%) completed the study. Subjects were assessed every 2 weeks with the Yale Brown Obsessive Compulsive Scale Modified for Kleptomania (K-YBOCS), the urge and behavior subscales of the K-YBOCS, the Kleptomania Symptom Assessment Scale (K-SAS), the Clinical Global Impressions Scale (CGI), and measures of depression, anxiety, and psychosocial functioning. Results: Subjects assigned to naltrexone had significantly greater reductions in K-YBOCS total scores (p ⫽ .001), stealing urges (p ⫽ .032), and stealing behavior (p ⬍ .001) compared with subjects on placebo. Subjects assigned to naltrexone also had greater improvement in overall kleptomania severity (reflected in the CGI scores) (p ⬍ .001). The mean effective dose of naltrexone was 116.7 (⫾44.4) mg/day. Conclusions: Naltrexone demonstrated statistically significant reductions in stealing urges and behavior in kleptomania. Naltrexone was well tolerated. Key Words: Impulse control disorders, kleptomania, naltrexone, pharmacotherapy, shoplifting, theft

K

leptomania is a disorder characterized by persistent and recurrent patterns of stealing. The Diagnostic and Statistical Manual of Mental Disorders (DSM-IV) defines kleptomania by the following criteria: 1) recurrent failure to resist impulses to steal objects that are not needed for personal use or for their monetary value; 2) increasing sense of tension immediately before committing the theft; 3) pleasure, gratification, or relief at the time of committing the theft; 4) the stealing is not committed to express anger or vengeance and is not in response to a delusion or a hallucination; and 5) the stealing is not better accounted for by another disorder. Psychosocial problems are common in kleptomania and include legal consequences, reduced quality of life, and impaired functioning (1– 4). Suicide attempts are also common and have been reported in 25% of persons in treatment for kleptomania (5). More than $13 billion worth of items are stolen each year from retailers (6). Recent data provided through the National Epidemiologic Survey on Alcohol and Related Conditions (NESARC) report rates of lifetime shoplifting of 11.3% in the United States (7). Shoplifting is defined as taking items from a store without paying for them, independent of intent or motivation. Kleptomania, on the other hand, is a psychiatric disorder characterized by an inability to resist recurrent impulses to steal objects that are not needed for their monetary or personal use.

From the Department of Psychiatry, University of Minnesota School of Medicine, Minneapolis, Minnesota. Address reprint requests to Jon E. Grant, J.D., M.D., M.P.H., Department of Psychiatry, University of Minnesota School of Medicine, 2450 Riverside Avenue, Minneapolis, MN 55454; E-mail: [email protected]. Received September 8, 2008; revised November 20, 2008; accepted November 20, 2008.

0006-3223/09/$36.00 doi:10.1016/j.biopsych.2008.11.022

Although shoplifting appears quite common, relatively little is known about the prevalence of kleptomania in the general population. In the only study assessing rates in a general community sample, researchers examined 791 college students and found that although 28.6% reported having stolen an item, only .4% met criteria for kleptomania (8). Other studies examining rates of kleptomania in treatment samples have found lifetime rates of 9.3% among psychiatric inpatients (n ⫽ 204) (9), 3.8% among alcohol-dependent patients (n ⫽ 79) (10), and 3.7% in a sample of depressed patients (n ⫽ 107) (11). Although described for almost two centuries (12), kleptomania remains poorly understood with limited data regarding pathophysiology or treatment. Kleptomania appears to share many phenomenological similarities to substance use disorders: urges or cravings, tolerance, withdrawal, repeated unsuccessful attempts to cut back or stop, and impairment in areas of life functioning (13). Family history data suggest that individuals with kleptomania are significantly more likely than control subjects to have first-degree relatives with alcohol use disorders (14). Given the possible clinical links between kleptomania and substance use disorders, as well as the efficacy of opiate antagonists in treating alcohol and opiate dependence (15–18), the opioid receptor antagonist, naltrexone, was examined in the treatment of kleptomania. The efficacy of opioid antagonists in the treatment of addictive disorders has been proposed to involve opioidergic modulation of mesolimbic dopamine circuitry (19). Behaviorally, opioid antagonist administration leads to diminished urges to engage in the addictive behavior and longer periods of abstinence (15– 18,20,21), consistent with a mechanism of action involving ventral striatal dopamine systems (22–24). If naltrexone could reduce urges to steal as it does to reduce urges for alcohol or drugs, it was hypothesized that it could reduce the behaviors associated with kleptomania. In an earlier open-label study, naltrexone was studied in 10 subjects with kleptomania. After 12 weeks of treatment, 20% of subjects reported full remission of symptoms and 80% had significant improvement in their shopBIOL PSYCHIATRY 2009;65:600 – 606 © 2009 Society of Biological Psychiatry

J.E. Grant et al. lifting urges and behavior (25). Further justification in studying opiate antagonists comes from the only other pharmacological study in kleptomania. Researchers tested escitalopram in 24 kleptomania subjects in an open-label design followed by double-blind discontinuation and found no significant differences between active medication and placebo (26). Because of the hypothesized mechanism of action of naltrexone (i.e., modulation of mesolimbic dopamine) (27,28) and the previous findings of naltrexone’s ability to reduce urges in kleptomania (25), the current study sought to enroll only those individuals with kleptomania who reported stealing secondary to urges or cravings. We hypothesized that naltrexone would reduce the severity of stealing urges and thereby improve behavior and patients’ overall functioning.

Methods and Materials Subjects Men and women aged 17 to 75 with a primary DSM-IV diagnosis of kleptomania were recruited by advertisements and referrals. All subjects met DSM-IV criteria for kleptomania using the clinician-administered Structured Clinical Interview for Kleptomania (SCI-K) (29) and were required to have had at least moderate urges to steal within the week prior to study entry (i.e., score ⱖ2 on the urge component of the Kleptomania Symptom Assessment Scale [K-SAS]) (25) and to have stolen something within 2 weeks prior to enrollment. Women’s participation required a negative beta-human chorionic gonadotropin pregnancy test (performed only at baseline visit) and stable use of a medically accepted form of contraception (subjects were asked at each visit of any change in form of contraception). Exclusion criteria included: 1) infrequent stealing (i.e., less than one time per week) that did not meet DSM-IV criteria for kleptomania; 2) unstable medical illness or clinically significant abnormalities on laboratory tests or physical examination at screening visit; 3) current pregnancy or lactation or inadequate contraception in women of childbearing potential; 4) a need for medication with unfavorable interactions with naltrexone (e.g., narcotics); 5) lifetime history of bipolar disorder type I or II, dementia, schizophrenia, or any psychotic disorder; 6) current or recent (past 3 months) DSM-IV substance use disorder; 7) treatment for kleptomania within the last 6 months; 8) baseline score of ⬎17 on either the 17-item Hamilton Depression Rating Scale (HAM-D) or the Hamilton Anxiety Rating Scale (HAM-A) (30,31); 9) positive urine drug screen at screening; 10) initiation of psychotherapy or behavior therapy within 3 months prior to study; 11) previous treatment with naltrexone; 12) clinically significant suicidality; and 13) treatment with investigational medication within 3 months prior to study. Subjects currently taking psychotropic medications were allowed into the study as long as the dose of medication had been stable for 3 months prior to study and there were no plans to modify the dose during the study. Subjects attending Shoplifters Anonymous were allowed to participate if attendance had been ongoing for at least 6 months prior to study. Subjects who changed doses of medication or started therapy or Shoplifters Anonymous, based on their self-report, were discontinued from the study (changes in treatment were assessed at each study visit and no one was withdrawn due to this reason). Of 47 individuals screened by telephone using the study exclusion and inclusion criteria, 31 individuals were interviewed face-to-face. Of the 31 subjects interviewed in person, 6 failed to meet inclusion/exclusion criteria (4 had positive drug screens; 2

BIOL PSYCHIATRY 2009;65:600 – 606 601 reported lifetime but not current kleptomania symptoms). After the baseline assessment, 25 individuals (18 women [72%] mean age ⫽ 34.3 ⫾ 12.1) with kleptomania were randomized to naltrexone (n ⫽ 12) or placebo (n ⫽ 13). The 25 randomized subjects reported a mean age of DSM-IV kleptomania onset of 23.4 (⫾15.9) years (range 8 – 62). Subjects spent 47.4 (⫾24.3) minutes each week stealing and an additional 114.3 (⫾102.4) minutes each week struggling with urges to steal. Twenty-three subjects (92.0%) had been arrested at least once for stealing. Of the 25 subjects, 6 (24%) had attended Shoplifters Anonymous at least once and 2 (8%) had sought outpatient mental health treatment for stealing. Seven subjects (28%) reported a history of physical or sexual abuse during childhood or adolescence based on clinical interview. Although subjects with bipolar, psychotic, and substance use disorders were excluded, the 25 enrolled subjects reported clinically important comorbid disorders: 19 (76%) subjects had at least one comorbid disorder; 11 (44%) reported histories consistent with major depressive disorder (all in remission currently); 8 (32%) had histories of anxiety disorders (e.g., generalized anxiety disorder, posttraumatic stress disorder, obsessive-compulsive disorder, anxiety disorder not otherwise specified [NOS]) (all in remission); 3 (12%) had another impulse control disorder (most commonly compulsive buying); and 1 (4%) had an eating disorder (i.e., bulimia nervosa). Rates of comorbid disorders did not differ between groups, and no particular comorbid disorder was associated with treatment response. Of the 25 subjects, 19 (76%) were taking a psychotropic medication. All 19 subjects had been on stable doses for at least 6 months: 14 were taking a selective serotonin reuptake inhibitor, 4 were taking a serotonin-norepinephrine reuptake inhibitor, and 1 was taking bupropion. Rates of psychotropic medication use did not differ between groups (76.9% of placebo group and 75% of naltrexone group; ␹2 ⫽ .126; df ⫽ 1; p ⫽ .910). No particular medication was associated with treatment response. The Institutional Review Board for the University of Minnesota approved the study and the informed consent. One investigator discussed potential risks of the study, as well as alternative treatments. This study was carried out in accordance with the Declaration of Helsinki. Data were collected from September 1, 2006, to June 30, 2008. After complete description of the study, written informed consent was obtained. Screening Assessments The screening visit entailed assessments of kleptomania, psychosocial functioning, and psychiatric and physical health. Subjects were evaluated with the Structured Clinical Interview for Kleptomania, a reliable and valid diagnostic instrument using DSM-IV criteria (29). A certificate of confidentiality was obtained from the National Institutes of Health, and subjects were informed about this privacy protection. Psychiatric comorbidity was assessed using the Structured Clinical Interview for DSM-IV (32). Medical history, physical examination, and routine laboratory testing (including liver function tests) were obtained. Investigators assessed kleptomania using the Yale Brown Obsessive Compulsive Scale Modified for Kleptomania (K-YBOCS) (26). Subjects reported kleptomania severity using the self-rated Kleptomania Symptom Assessment Scale (K-SAS) (25). Anxiety and depressive symptoms were rated with the Hamilton Anxiety Rating Scale (HAM-A) (31) and the 17-item Hamilton Depression Rating Scale (HAM-D) (30), respectively. Psychosocial functioning was evaluated using the Sheehan Disability Scale (SDS) (33). www.sobp.org/journal

602 BIOL PSYCHIATRY 2009;65:600 – 606 Study Design Dose range selection was based on naltrexone’s clinical and pharmacokinetic data, studies using naltrexone in other impulse control disorders such as pathological gambling, and the previous open-label study in kleptomania (20,25). Previous studies have suggested that relatively high doses (i.e., 2 to 3 times the recommended therapeutic dose approved for alcohol dependence) may be needed to elicit a therapeutic response (20,25,34). Thus, we selected a dose titration starting at 50 mg/day and ranging to 150 mg/day. After screening, the university’s investigational pharmacy randomized all subjects (in blocks of eight, using computergenerated randomization with no clinical information) to either naltrexone or placebo in a 1:1 fashion. Subjects were seen every 2 weeks for the 8-week period (an 8-week duration was chosen because previous studies of naltrexone in pathological gambling, a related impulse control disorder, reported response within 6 weeks) (35). All eligible subjects were started on naltrexone 50 mg/day or matching placebo for 2 weeks. At week 2, the dose was increased to 100 mg/day, and at week 4, the dose was increased to 150 mg/day unless clinical improvement (i.e., cessation of all stealing behavior and urges for the entire 2-week period as assessed by the investigator) was attained at a lower dose. Eight subjects (61.5%) assigned to placebo and 7 (58.3%) assigned to naltrexone received both dose increases (Fisher’s exact test ⫽ .596). Given the lack of difference between groups in terms of dose increase, change in dose would be unlikely to have affected the blind. Subjects not compliant with medication (i.e., failing to take medication for 3 or more consecutive days per self-report) were discontinued from the study. Subjects were asked at each visit about the number of doses missed per week. Twenty-three subjects reported never missing a dose, and two subjects reported missing at most one or two doses each month. Efficacy and Safety Assessments Subjects were seen every 2 weeks for 8 weeks. Efficacy and safety assessments were performed every 2 weeks. A rater independent of the treating clinician conducted the progress evaluations. The primary outcome measure was the Yale Brown Obsessive Compulsive Scale Modified for Kleptomania (26), a reliable and valid, 10-item, clinician-administered scale that rates symptoms within the last 7 days on a severity scale from 0 to 4 for each item. The K-YBOCS first 5 items comprise the stealing urge/thought subscale (time occupied with urges/thoughts; interference and distress due to urges/thoughts; resistance against and control over urges/thoughts), and items 6 to 10 comprise the behavior subscale (time spent stealing and amount of stealing; interference and distress due to stealing; ability to resist and control behavior). Secondary measures used at each study visit included: 1. Kleptomania Symptom Assessment Scale (25). The K-SAS is an 11-item, reliable and valid, self-rated scale assessing stealing urges, thoughts, and behaviors during the previous 7 days. 2. Clinical Global Impression-Severity (CGI) (36). The CGI is a reliable and valid 7-item scale used to assess severity in clinical symptoms and ranges from 1 ⫽ “not ill at all” to 7 ⫽ “among the most extremely ill.” The CGI was used to refer specifically to kleptomania severity, not overall psychopathology. 3. Sheehan Disability Scale (33). The SDS is a three-item, reliable and valid self-report scale that assesses functioning www.sobp.org/journal

J.E. Grant et al. in three areas of life: work, social or leisure activities, and home and family life. 4. Hamilton Anxiety Rating Scale (31). The HAM-A is a reliable and valid, clinician-administered, 14-item scale providing an overall measure of global anxiety. 5. Hamilton Depression Rating Scale (30). The HAM-D is a valid and reliable, 17-item, clinician-administered rating scale assessing severity of depressive symptoms. Safety assessments at each visit included evaluations of blood pressure, heart rate, and weight. Adverse effects were documented and included time of onset and resolution, severity, action taken, and outcome. The investigator recorded use of concomitant medications in terms of dosage, start and stop dates, and reason for use. Liver function tests were drawn at baseline and then every 2 weeks. Data Analysis Demographic and baseline visit characteristics for both groups were compared using chi-square and t tests to determine if group differences existed at randomization. Primary and secondary measures were examined using repeated measures analysis of variance (ANOVA) (GLM, SPSS for Windows, Version 15.0, SPSS Inc., Chicago, Illinois). The interaction of treatment assignment and visit was the test of primary interest. Analyses were performed on all available data using an intent-to-treat population (last observation carried forward [LOCF]). When Mauchly’s test of sphericity on the repeated measure was significant, the Greenhouse-Geisser correction was used, resulting in fractional degrees of freedom for the withingroup F tests. Because of multiple comparisons, a Bonferroni correction was used; tests were two-tailed and an alpha level of .006 was used to determine statistical significance. To determine when a significant treatment effect was first detected on the primary measure, the interaction of treatment assignment with the contrast of each visit to baseline was examined from the repeated measures ANOVA. When assessing when baseline versus visit by treatment assignment became significant, we set our first measure as primary at .05 and treated the three others as secondary using a Holm’s sequential Bonferroni, resulting in an adjusted significance of p ⫽ .0125. A retrospective power analysis was performed using a bootstrap approach with 10,000 resampling iterations (37). Significance at p ⫽ .05 was achieved in 97.7% of the two-level active/placebo dichotomy for the primary outcome measure, the K-YBOCS (significance at p ⫽ .006 was achieved in 77.7%). These results indicate that significant findings are not likely to be based on chance. Effect sizes were calculated using Cohen’s effect size index d. A d of .2 is considered a small effect size, .5 is medium, and .8 is large.

Results Subject Characteristics Demographics of subjects at baseline are presented (Table 1). There were no statistically significant imbalances regarding age, gender, or marital status between groups. There were significant differences between groups regarding kleptomania severity at baseline with those assigned to naltrexone having more severe symptoms as reflected by the K-SAS and the CGI severity scales. A total of 23 (92%) completed the 8-week study. Twelve (92.3%) of 13 subjects assigned to placebo and 11 (91.7%) of 12 subjects assigned to naltrexone completed the study. The rate of

BIOL PSYCHIATRY 2009;65:600 – 606 603

J.E. Grant et al. Table 1. Baseline Comparison of Kleptomania Subjects Assigned to Naltrexone or Placebo Baseline a

Variable

Age, years Sex, Female, n (%) White, n (%) Married, n (%) Subjects Currently Taking an Antidepressant, n (%) K-YBOCS Total Score K-YBOCS Urge/Thought Subscale K-YBOCS Behavior Subscale Kleptomania Symptom Assessment Scale (K-SAS) Total Score Clinical Global Impression (CGI) – Severity Scale Hamilton Depression Rating Scale Hamilton Anxiety Rating Scale Sheehan Disability Scale

Placebo (n ⫽ 13)

Naltrexone (n ⫽ 12)

Testb

df

p Value

41.4 ⫾ 12.7 [17–64] 10 (76.9) 12 (92.3) 6 (46.2) 10 (76.9) 15.69 (5.44) [7–25] 9.69 (3.17) [4–15] 6.00 (4.67) [0–11] 22.62 (10.21) [10–41]

44.3 ⫾ 12.2 [22–65] 8 (66.7) 12 (100.0) 3 (25.0) 9 (75.0) 20.08 (7.79) [12–36] 10.67 (4.79) [3–20] 9.42 (5.21) [1–19] 29.42 (5.38) [21–41]

.575 f f f F 1.646 .604 1.728 2.105

23 NA NA NA NA 23 23 23 18.489

.571 .673 .999 .267 1.000 .113 .552 .097 .049

4.15 (.69) [3–5] 6.69 (4.96) [0–17] 6.46 (5.24) [0–16] 8.69 (7.60) [0–24]

5.00 (.95) [4–7] 5.58 (3.45) [0–10] 4.92 (3.97) [0–13] 14.33 (8.95) [0–27]

2.559 ⫺.644 ⫺.826 1.704

23 23 23 23

.018 .526 .417 .102

CGI, Clinical Global Impression; K-SAS, Kleptomania Symptom Assessment Scale; K-YBOCS, Yale Brown Obsessive Compulsive Scale Modified for Kleptomania; NA, not applicable. a All variables are mean (⫾SD) [range] unless otherwise indicated. b t test unless otherwise indicated (f ⫽ Fisher’s exact test).

completion did not differ between groups. Of the two subjects who failed to complete the study, one withdrew for an inability to comply with the study schedule and one withdrew due to nausea. Reasons for study withdrawal did not significantly differ between groups. Efficacy Results Efficacy results were tested using an intent-to-treat repeated measure ANOVA analysis with LOCF for missing data. Significantly better results were observed for those assigned to naltrexone on the primary efficacy variable, the K-YBOCS total scores (F ⫽ 6.07; df ⫽ 2.94,67.7; p ⫽ .001) (Table 2; Figure 1). A significant interaction of treatment effect by visit contrast (visit vs.

baseline) was first detected after 6 weeks on active medication (F ⫽ 7.187; df ⫽ 1,23; p ⫽ .013) (Figure 1). Cohen’s effect size after 8 weeks (study end point) was 1.14. The mean effective dose of naltrexone was 116.7 (⫾44.4) mg/day. A placebo response was detected for the first 4 weeks of the study, which then diminished. Yale Brown Obsessive Compulsive Scale Modified for Kleptomania subscales were consistent with the total score. Neither the behavior subscale nor the urge/thought subscale had a significant treatment group main effect. Both subscales had significant visit main effects: F ⫽ 8.52; df ⫽ 4,92; p ⬍ .001 for the behavior subscale and F ⫽ 17.24; df ⫽ 2.27,52.25; p ⬍ .001 for

Table 2. Treatment Responses of Kleptomania Subjects Assigned to Placebo or Naltrexone (Intent-to-Treat Population) Baseline Variablea

Treatment F (df) p Value

Visits F (dfb) p Value

Treatment by Visits F (dfb) p Value

Effect Size Cohen’s dc

17.80 (2.94,67.7) p ⬍ .001 17.24 (2.27,52.25)

6.07 (2.94,67.7) p ⫽ .001 3.29 (2.27,52.25)

1.14

End Point

Placebo

Naltrexone

Placebo

Naltrexone

15.69 (5.44)

20.08 (7.79)

11.46 (7.76)

3.83 (2.86)

K-YBOCS Urge/Thought Subscale

9.69 (3.17)

10.67 (4.79)

6.38 (3.62)

3.17 (2.33)

.16 (1,23) p ⫽ .691 .25 (1,23)

K-YBOCS Behavior Subscale

6.00 (4.67)

9.42 (5.21)

5.08 (5.01)

.67 (2.02)

p ⫽ .619 .06 (1,23)

K-SAS Total Score

22.62 (10.21)

29.42 (5.38)

16.54 (9.99)

7.50 (6.70)

CGI – Severity

4.15 (.69)

5.00 (.95)

3.31 (1.38)

1.75 (.75)

HAM-D

6.69 (4.96)

5.58 (3.45)

4.38 (4.54)

1.50 (2.75)

HAM-A

6.46 (5.24)

4.92 (3.97)

3.38 (3.50)

1.42 (2.71)

Sheehan Disability Scale

8.69 (7.60)

14.33 (8.95)

5.00 (5.51)

1.83 (3.10)

K-YBOCS Total Score

p ⫽ .805 .01 (1.23) p ⫽ .942 .54 (1,23) p ⫽ .470 2.07 (1,23) p ⫽ .164 1.65 (1,23) p ⫽ .212 .65 (1,23) p ⫽ .427

p ⬍ .001 8.52 (4,92) p ⬍ .001 20.64 (3.15,72.37) p ⬍ .001 22.73 (4,92) p ⬍ .001 6.02 (2.67,61.32) p ⫽ .002 6.96 (2.82,64.90) p ⫽ .001 9.35 (2.27,52.19) p ⬍ .001

p ⫽ .039 5.61 (4,92) p ⬍ .001 7.21 (3.15, 72.37) p ⬍ .001 7.64 (4,92) p ⬍ .001 .84 (2.67,61.32) p ⫽ .465 .39 (2.82,64.90) p ⫽ .746 2.79 (2.27,52.19) p ⫽ .064

.80

.89

1.09 1.89 .67 .42 .38

CGI, Clinical Global Impression (Severity) Scale; HAM-A, Hamilton Anxiety Rating Scale; HAM-D, Hamilton Depression Rating Scale; K-SAS, Kleptomania Symptom Assessment Scale; K-YBOCS, Yale Brown Obsessive Compulsive Scale Modified for Kleptomania. a All variables are mean (⫾SD). b When Mauchly’s test of sphericity was significant, the Greenhouse-Geiser correction was used (fractional degrees of freedom). c Cohen’s d at end point calculated using the pooled standard deviation of the two baseline groups.

www.sobp.org/journal

604 BIOL PSYCHIATRY 2009;65:600 – 606 40 36 32 28 24 20 16 12 8 4 0 Naltrexone Placebo

J.E. Grant et al. 7 6 5 4 3

Visit 1

Visit 2

Visit 3

Visit 4

Visit 5

20.08 15.69

12.33 10.08

10.25 9.38

6.08 9.85

3.83 11.46

.485

.293

.013

.001

P-value*

Figure 1. K-YBOCS total scores over time in subjects assigned to naltrexone or placebo. *p value is for the interaction of treatment group by contrast of each visit to baseline (df ⫽ 1,23). With four visits, a Bonferroni adjusted p value would be .0125. K-YBOCS, Yale Brown Obsessive Compulsive Scale Modified for Kleptomania.

the urge/thought subscale (Table 2). The interaction of treatment group by visit was statistically significant on the behavior subscale (F ⫽ 5.61; df ⫽ 4,92; p ⬍ .001) and showed a trend toward significance on the urge/thought subscale (F ⫽ 3.29; df ⫽ 2.27,52.25; p ⫽ .039) (Table 2). A significant interaction of treatment effect by visit contrast was first detected after 6 weeks on active medication for the behavior subscale (F ⫽ 7.263; df ⫽ 1,23; p ⫽ .013). A nearly significant interaction of treatment effect by visit contrast was first seen at 8 weeks on active medication for the urge/thought subscale (F ⫽ 4.248; df ⫽ 1,23; p ⫽ .051). Cohen’s effect size at week 8 was .89 for the behavior subscale and .80 for the urge/thought subscale. Kleptomania Symptom Assessment Scale analysis (Table 2; Figure 2) did not have a significant treatment group main effect but did have a significant visit main effect (F ⫽ 20.64; df ⫽ 3.15,72.37; p ⬍ .001) and demonstrated a statistically significant treatment group by visit interaction in patient-reported stealing symptoms (F ⫽ 7.21; df ⫽ 3.15,72.37; p ⬍ .001). A significant interaction of treatment effect by visit contrast was first detected after 6 weeks on active medication (F ⫽ 13.086; df ⫽ 1,23; p ⫽ 44 40 36 32 28 24 20 16 12 8 4 0 Naltrexone Placebo

Visit 1

Visit 2

Visit 3

Visit 4

Visit 5

29.42 22.62

20.17 16.46

18.08 15.23

9.83 15.23

7.5 16.54

.350

.351

.001

<.001

P-value*

Figure 2. K-SAS total scores over time in subjects assigned to naltrexone or placebo. *p value is for the interaction of treatment group by contrast of each visit to baseline (df ⫽ 1,23). With four visits, a Bonferroni adjusted p value would be .0125. K-SAS, Kleptomania Symptom Assessment Scale.

www.sobp.org/journal

2 1 Naltrexone Placebo P-value*

Visit 1

Visit 2

Visit 3

Visit 4

Visit 5

5

3.58

3.08

2.17

1.75

4.15

3.31

3.08

3.08

3.31

.270

.116

.002

<.001

Figure 3. CGI severity scores over time in subjects assigned to naltrexone or placebo. *p value is for the interaction of treatment group by contrast of each visit to baseline (df ⫽ 1,23). With four visits, a Bonferroni adjusted p value would be .0125. CGI, Clinical Global Impressions Scale.

.001). Cohen’s effect size after 8 weeks on active medication was 1.09. The CGI severity scale demonstrated significant improvement by study end point (Table 2; Figure 3). The CGI severity scale did not have a significant treatment group main effect but did have a significant visit main effect (F ⫽ 22.73; df ⫽ 4,92; p ⬍ .001) and demonstrated a statistically significant treatment group by visit interaction (F ⫽ 7.64; df ⫽ 4,92; p ⬍ .001). A significant interaction of treatment effect by visit contrast was first detected after 6 weeks on active medication (F ⫽ 12.101; df ⫽ 1,23; p ⫽ .002). Cohen’s effect size after 8 weeks on active medication was 1.89 for the CGI severity scale. Subjects assigned to naltrexone demonstrated greater response with respect to psychosocial functioning. The SDS demonstrated a statistical trend toward significance of the treatment group by visit interaction (F ⫽ 2.79; df ⫽ 2.27,52.19; p ⫽ .064) (Table 2). Remission of kleptomania symptoms (i.e., defined as a KYBOCS score of 5 or less) was seen in eight naltrexone subjects (66.7%) and one (7.7%) on placebo (Fisher’s exact test ⫽ .003). Three of 12 naltrexone subjects (25%) failed to respond (i.e., defined as a score of 1 or 2 on the CGI improvement scale— “much” or “very much” improved). Rates of comorbid disorders did not differ between naltrexone responders and nonresponders (Fisher’s exact test ⫽ .470). Completer analyses (n ⫽ 23) demonstrated significant reductions on all measures consistent with the intent-to-treat analyses. An analysis of gender demonstrated that both men and women responded equally well to naltrexone with no significant gender differences. Safety and Tolerability The incidence of adverse experiences in naltrexone-treated subjects was consistent with prior studies (20,35), and no unusual experiences were reported. In the naltrexone group, 4 (33.3%) subjects reported nausea, 1 (8.3%) reported dry mouth, and 1 (8.3%) reported insomnia. In the placebo group, 1 (7.7%) reported headache, 1 (7.7%) reported nausea, and 1 (7.7%) reported diarrhea. Most adverse experiences were of mild to moderate intensity and occurred during the first week of treatment. One naltrexone subject withdrew due to nausea. There

J.E. Grant et al. were no statistically significant differences in the incidence of adverse events between groups. There were no clinically significant changes in laboratory testing, including liver function tests, during treatment. Mean values in HAM-D and HAM-A scores remained at low levels throughout the study in both groups (Table 2).

Discussion This randomized, double-blind, clinical trial found naltrexone to be superior to placebo in the treatment of kleptomania across a spectrum of illness-specific and global outcome measures. The results demonstrate that naltrexone treatment reduces the symptoms associated with kleptomania. In addition, this is the first double-blind pharmacological study in kleptomania and extends our previous open-label findings (25) by demonstrating that naltrexone is beneficial, safe, and well tolerated. The efficacy of naltrexone lends further support to the hypothesis that pharmacological manipulation of the opiate system may target core symptoms of kleptomania (34). Opioid antagonists have been hypothesized to influence dopamine neurotransmission in the nucleus accumbens and linked motivational neurocircuitry, dampening stealing-related excitement and cravings (34). Although modulation of drive and subsequent behavioral output by dopamine, endorphins, and gamma-aminobutyric acid has been hypothesized for impulse control disorders, the specific mechanisms underlying opioid receptor antagonism in kleptomania remain incompletely understood. Opioid antagonists have been effective in treating both impulse control disorders (e.g., pathological gambling) as well as substance addictions (e.g., alcohol dependence) (20,35,38 – 40). Naltrexone’s efficacy in kleptomania lends support to the hypothesis that kleptomania in particular and impulse control disorders in general may be related to substance addictions (41). Adverse events reported in this study were consistent with naltrexone’s previously reported safety profile (20,35). There were no cases of hepatotoxicity during the 8-week trial. Although there has been concern that opioid antagonists may engender depression, depression and anxiety scores (HAM-D and HAM-A) did not worsen during the study. This study represents the only placebo-controlled medication trial in kleptomania performed to date, but there exist several limitations. First, most of the subjects were currently taking antidepressants and this complicates our understanding of the results. Although subjects taking antidepressants had been on the doses for considerable time without any benefit in kleptomania symptoms, it cannot be concluded that naltrexone alone would be effective for kleptomania. Second, the sample size was small and severely limits the power to identify subgroups of subjects who may have responded more robustly to naltrexone. The retrospective power analysis, however, showed estimated power of 97.7% and therefore improvement due to chance is unlikely even with the small sample. Third, the study relied on self-report of stealing behavior. Although granted a certificate of confidentiality, we cannot guarantee that subjects did not underreport stealing or lie for other reasons. No subject was court-ordered to treatment. Assuming that lying or underreporting would be equally distributed across subjects in both groups, however, the findings of improvement on naltrexone still remain strongly indicative of a medication effect. Fourth, kleptomania appears to be a chronic disease that may require long-term therapy. This study did not assess treatment effects beyond the acute 8-week treatment period, and longer-term effects thus require further

BIOL PSYCHIATRY 2009;65:600 – 606 605 evaluation. It is possible that a longer course of therapy could result in continued and even greater reductions in kleptomania symptoms. Although a recent review of the efficacy of naltrexone in patients with alcohol dependence found that naltrexone provided good short-term benefit but that longer-term use provided less compelling evidence of efficacy (42), this is not yet known for kleptomania. In fact, a naturalistic outpatient study demonstrated that subjects treated with naltrexone monotherapy for up to 3 years continued to report overall improvement (43). Fifth, we excluded individuals with severe mental health issues (e.g., bipolar disorder or current substance use disorders) and therefore these results may not generalize completely to the larger population of people with kleptomania. The study did, however, include subjects with depression and anxiety disorders and thereby appears fairly representative of individuals with kleptomania (5,44). This investigation suggests that naltrexone may be effective in the acute treatment of kleptomania when subjects report urges to steal. The long-term effects of this treatment, however, need to be documented in this population. As effective treatments for kleptomania emerge, it becomes increasingly important that physicians and mental health providers screen for kleptomania to provide timely treatment. This research was supported in part by a Career Development Award by the National Institute of Mental Health (K23 MH06975401A1) to Dr. Grant. This study was supported by internal funds. Dr. Grant has received research grants from Forest Pharmaceuticals and GlaxoSmithKline. Dr. Kim and Mr. Odlaug report no biomedical financial interests or potential conflicts of interest. Clinical Trials (A Double-Blind, Placebo-Controlled Study of Naltrexone in Kleptomania; http://www.clinicaltrials.gov/ct2/ show/NCT00332579?term⫽naltrexone⫹kleptomania&rank⫽1; NCT00332579). Supplementary material cited in this article is available online. 1. McElroy SL, Pope HG Jr, Hudson JI, Keck PE Jr, White KL (1991): Kleptomania: A report of 20 cases. Am J Psychiatry 148:652– 657. 2. Aboujaoude E, Gamel N, Koran LM (2004): Overview of kleptomania and phenomenological description of 40 patients. Prim Care Companion J Clin Psychiatry 6:244 –247. 3. Goldman MJ (1991): Kleptomania: Making sense of the nonsensical. Am J Psychiatry 148:986 –996. 4. Grant JE, Kim SW (2002): Clinical characteristics and associated psychopathology of 22 patients with kleptomania. Compr Psychiatry 43:378 – 384. 5. Grant JE, Potenza MN (2008): Gender-related differences in individuals seeking treatment for kleptomania. CNS Spectr 13:235–245. 6. National Association for Shoplifting Prevention (2006): Shoplifting statistics. Available at: http://www.shopliftingprevention.org/WhatNASPOffers/ NRC/PublicEducStats.htm. Accessed July 25, 2008. 7. Blanco C, Grant J, Petry NM, Simpson HB, Alegria A, Liu SM, Hasin D (2008): Prevalence and correlates of shoplifting in the United States: Results from the National Epidemiologic Survey on Alcohol and Related Conditions (NESARC). Am J Psychiatry 165:905–913. 8. Odlaug BL, Grant JE (2008): Prevalence of impulse control disorders in a college sample. Presented at the Annual Meeting of the American Psychiatric Association, May 2008, Washington, DC. NR2-069 [poster]. 9. Grant JE, Levine L, Kim D, Potenza MN (2005): Prevalence of impulse control disorders in adult psychiatric inpatients. Am J Psychiatry 162: 2184 –2188. 10. Lejoyeux M, Feuché N, Loi S, Solomon J, Adès J (1999): Study of impulsecontrol disorders among alcohol-dependent patients. J Clin Psychiatry 60:302–305. 11. Lejoyeux M, Arbaretaz M, McLoughlin M, Adès J (2002): Impulse control disorders and depression. J Nerv Ment Dis 190:310 –314.

www.sobp.org/journal

606 BIOL PSYCHIATRY 2009;65:600 – 606 12. Esquirol E (1838): Des Maladies Mentales. Paris: Bailliere. 13. Grant JE, Odlaug BL (2008): Kleptomania: Clinical characteristics and treatment. Rev Bras Psiquiatr 30(suppl 1):S11–S15. 14. Grant JE (2003): Family history and psychiatric comorbidity in persons with kleptomania. Compr Psychiatry 44(6):437– 441. 15. Volpicelli JR, Alterman AI, Hayashida M, O’Brien CP (1992): Naltrexone in the treatment of alcohol dependence. Arch Gen Psychiatry 49:876 – 880. 16. Anton RF, Moak DH, Waid LR, Latham PK, Malcolm RJ, Dias JK (1999): Naltrexone and cognitive behavioral therapy for the treatment of outpatient alcoholics: Results of a placebo-controlled trial. Am J Psychiatry 156:1758 –1764. 17. Martin WR, Jasinski D, Mansky P (1973): Naltrexone, an antagonist for the treatment of heroin dependence. Arch Gen Psychiatry 23:784 –789. 18. Kosten TR, Kleber HD, Morgan C (1989): Role of opioid antagonists in treating intravenous cocaine abuse. Life Sci 44:887– 892. 19. Ikemoto S, Glazier BS, Murphy JM, McBride WJ (1997): Role of dopamine D1 and D2 receptors in the nucleus accumbens in mediating reward. J Neurosci 17:8580 – 8587. 20. Kim SW, Grant JE, Adson DE, Shin YC (2001): Double-blind naltrexone and placebo comparison study in the treatment of pathological gambling. Biol Psychiatry 49:914 –921. 21. Mason BJ, Salvato FR, Williams LD, Ritvo EC, Cutler RB (1999): A doubleblind, placebo-controlled study of oral nalmefene for alcohol dependence. Arch Gen Psychiatry 56:719 –724. 22. Stewart J (1984): Reinstatement of heroin and cocaine self-administration behavior in the rat by intracerebral application of morphine in the ventral tegmental area. Pharmacol Biochem Behav 20:917–923. 23. Matthews RT, German DC (1984): Electrophysiological evidence for excitation of rat ventral tegmental area dopamine neurons by morphine. Neuroscience 11:617– 625. 24. Broekkamp CL, Phillips AG (1979): Facilitation of self-stimulation behavior following intracerebral microinjections of opioids into the ventral tegmental area. Pharmacol Biochem Behav 11:289 –295. 25. Grant JE, Kim SW (2002): An open label study of naltrexone in the treatment of kleptomania. J Clin Psychiatry 63:349 –356. 26. Koran LM, Aboujaoude EN, Gamel NN (2007): Escitalopram treatment of kleptomania: An open-label trial followed by double-blind discontinuation. J Clin Psychiatry 68:422– 427. 27. Matthews RT, German DC (1984): Electrophysiological evidence for excitation of rat ventral tegmental area and dopamine neurons by morphine. Neuroscience 11:617– 625. 28. Stewart J (1984): Reinstatement of heroin and cocaine self-administration behavior in the rat by intracerebral application of morphine in the ventral tegmental area. Pharmacol Biochem Behav 20:917–923.

www.sobp.org/journal

J.E. Grant et al. 29. Grant JE, Kim SW, McCabe J (2006): Structured Clinical Interview for Kleptomania (SCI-K): Peliminary validity and reliability testing. Int J Methods Psychiatr Res 15:83–94. 30. Hamilton M (1960): A rating scale for depression. J Neurol Neurosurg Psychiatry 23:56 – 62. 31. Hamilton M (1959): The assessment of anxiety states by rating. Br J Med Psychol 32:50 –55. 32. First MB, Spitzer RL, Gibbon M, Williams JBW (1995): Structured Clinical Interview for DSM-IV-Patient Edition (SCID-I/P, Version 2.0). New York: Biometrics Research Department, New York State Psychiatric Institute. 33. Sheehan DV (1983): The Anxiety Disease. New York: Scribner’s. 34. Kim SW (1998): Opioid antagonists in the treatment of impulse-control disorders. J Clin Psychiatry 59:159 –164. 35. Grant JE, Kim SW, Hartman BK (2008): A double-blind, placebo-controlled study of the opiate antagonist naltrexone in the treatment of pathological gambling urges. J Clin Psychiatry 69:783–789. 36. Guy W (1976): ECDEU Assessment Manual for Psychopharmacology. Rockville, Maryland: National Institute of Mental Health. U.S. Dept Health, Education and Welfare publication (ADM) 76 –338, 218 –222. 37. Davison AC, Hinkley DV (1997): Bootstrap Methods and Their Application. Cambridge, United Kingdom: Cambridge University Press. 38. Grant JE, Potenza MN, Hollander E, Cunningham-Williams R, Nurminen T, Smits G, Kallio A (2006): Multicenter investigation of the opioid antagonist nalmefene in the treatment of pathological gambling. Am J Psychiatry 163:303–312. 39. Mason BJ, Salvato FR, Williams LD, Ritvo EC, Cutler RB (1999): A doubleblind, placebo-controlled study of oral nalmefene for alcohol dependence. Arch Gen Psychiatry 56:719 –724. 40. Anton RF, O’Malley SS, Ciraulo CC, Cisler RA, Couper D, Donovan DM, et al. (2006): COMBINE Study Research Group: Combined pharmacotherapies and behavioral interventions for alcohol dependence: The COMBINE study: A randomized controlled trial. JAMA 295:2003–2017. 41. Potenza MN (2006): Should addictive disorders include non-substancerelated conditions? Addiction 101(suppl 1):142–151. 42. Srisurapanont M, Jarusuraisin N (2005): Naltrexone for the treatment of alcoholism: A meta-analysis of randomized controlled trials. Int J Neuropsychopharmacol 8:267–280. 43. Grant JE (2005): Outcome study of kleptomania patients treated with naltrexone: A chart review. Clin Neuropharmacol 28:11–14. 44. Dannon PN, Lowengrub KM, Iancu I, Kotler M (2004): Kleptomania: Comorbid psychiatric diagnosis in patients and their families. Psychopathology 37:76 – 80.