Topiramate in the Treatment of Binge Eating Disorder Associated With Obesity: A Randomized, Placebo-Controlled Trial

Study participants were outpatients at the University of Cincinnati Medical Center who were between 18 and 60 years of age, met DSM-IV-TR criteria for binge eating disorder, were obese (defined as having a body mass index ≥30 kg/m²) (3, 5), and, to ensure they had marked distress regarding their binge eating, had a score ≥15 on the Yale-Brown Obsessive Compulsive Scale (18), which was modified for binge eating (available from the first author upon request). DSM-IV-TR criteria for binge eating disorder are recurrent episodes of binge eating associated with at least three of five indicators of impaired control and marked distress but without the regular use of inappropriate compensatory behaviors. Also, the binges must occur at least 2 days a week for 6 months. The modified Yale-Brown Obsessive Compulsive Scale measured obsessiveness of binge eating thoughts and compulsiveness of binge eating behaviors. Patients were recruited by a radio advertisement requesting volunteers for a study of a medication for the treatment of binge eating associated with obesity.

We excluded patients from participation if any of the following were present or occurred: 1) substance use disorder (per DSM-IV-TR) within the past 6 months; 2) unstable bipolar disorder (per DSM-IV-TR) within the past 3 months; 3) clinically significant suicidality; 4) any current or past psychiatric disorder that could interfere with diagnostic assessment, treatment, or study adherence; 5) clinically unstable medical illness; 6) a history of nephrolithiasis or seizures (including childhood febrile seizures); 7) clinically significant abnormal laboratory results; 8) need for treatment with any medication that might adversely interact with or obscure the action of topiramate (e.g., stimulants, antidepressants, or carbonic anhydrase inhibitors); 9) treatment with psychoactive medications (other than mood stabilizers or zolpidem) within 2 weeks of random assignment; 10) treatment with an experimental drug or an experimental device within 30 days of random assignment; or 11) previous treatment with topiramate.

The study protocol was approved by the University of Cincinnati Medical Center Institutional Review Board, and the study was conducted in compliance with the Declaration of Helsinki. All patients signed approved informed consent forms after the study procedures had been fully explained. Patients were enrolled from September 2, 1998, through June 4, 2000.

This was a randomized, parallel-group, placebo-controlled, double-blind, flexible-dose (25–600 mg/day) topiramate trial conducted at the University of Cincinnati Medical Center. The trial consisted of a 2- to 5-week screening period, a 14-week treatment period, and a 2-week treatment taper and discontinuation period. Patients were evaluated at least twice (initial screening examination and baseline assessment) during the screening period; after 1, 2, 4, 6, 8, 10, and 14 weeks during the treatment period; and after weeks 15 and 16 during the 2-week treatment taper and discontinuation period.

The following were obtained or conducted at the initial screening visit: written informed consent; assessment with the Structured Clinical Interview for DSM-IV (19) (to determine whether the patient met DSM-IV criteria for binge eating disorder and any comorbid axis I psychiatric disorders); the number of binges and binge days during the previous week; psychiatric and medical history; a physical examination; vital sign monitoring; routine blood chemical and hematological tests; fasting blood levels of glucose, insulin, and lipids; an electrocardiogram; and a urinalysis. Patients were given take-home diaries at this and each of the following visits in which to record, on a daily basis, any binges and, once study medication was initiated, the number of medication tablets taken.

At the last visit of the screening period (the baseline assessment), patients were evaluated to see if they continued to meet entry criteria. Patients continuing to meet these criteria were enrolled in the treatment period and randomly assigned in a 1:1 ratio to therapy with topiramate or placebo. All study medication was in identical 25- and 100-mg tablets supplied in numbered containers and dispensed to patients according to a predetermined randomization schedule (see statistical analysis). The initial dose was 25 mg each evening for a minimum of 3 days. Beginning on day 4, the dose was increased to 50 mg each evening. Beginning on day 7, the dose was increased to 75 or 100 mg each evening. If after 2 weeks there was no response (defined as a 50% or greater reduction in binge frequency), the dose was increased at a maximum of 50 mg/week for 4 weeks to a maximum dose of 300 mg/day at 6 weeks. Subsequently, the dose was increased at a maximum of 75 mg/week for 4 weeks to a maximum dose of 600 mg/day at 10 weeks. Study medication dose was not changed from treatment period weeks 10 through 14 unless a medical reason (e.g., an adverse event) required such a change. If a patient did not tolerate any dose increase, the dose could be decreased to one that was tolerable. Study medication was provided in prepackaged coded plastic bottles containing 60 identical 25- or 100-mg topiramate or placebo tablets.

In addition to the assessment of medication dose, medication compliance (through diary review and tablet count), adverse events, use of any nonstudy medications, weight, and vital signs, efficacy measures were assessed at each subsequent visit. The primary efficacy measure was the number of binge eating episodes (binge frequency) during the 7 days before each visit. A binge was defined according to DSM-IV-TR criteria. Binges were assessed by clinical interview and review of patient take-home diaries, in which patients recorded binges, duration of binges, and food consumed during binges (so that binges could be confirmed by the investigator). A secondary measure of efficacy was the number of binge days (binge day frequency) during the 7 days before each visit. A binge day was defined as a day on which a patient had at least one binge eating episode. For visits that occurred less than 7 days after the previous visit, the number of binges and binge days were normalized to a weekly frequency. Other secondary efficacy measures included scores on the Clinical Global Impression (CGI) improvement and severity scales, the modified Yale-Brown Obsessive Compulsive Scale, and the Hamilton Depression Rating Scale (20) as well as body mass index, weight, waist-to-hip ratio, and percent and total body fat as measured by bioelectrical impedance (Health Management System 1000, Bio/Analogics, Beaverton, Ore.). The CGI improvement scale is a 7-point scale on which 1=very much improved, 2=much improved, 3=minimally improved, 4=no change, 5=minimally worse, 6=much worse, and 7=very much worse. The CGI severity scale is a 7-point scale on which 1=normal, 2=borderline ill, 3=mildly ill, 4=moderately ill, 5=markedly ill, 6=severely ill, and 7=among the most extremely ill. Additional secondary outcome measures included blood pressure and fasting measurements of blood glucose, insulin, and lipids.

At every treatment period visit, we assessed all efficacy measures except for blood tests, waist-to-hip ratio, and percent and total body fat, which were obtained at the week 14 treatment period visit (or the last visit if the patient terminated the study prematurely).

We assessed the following safety measures: adverse events, clinical laboratory data, physical examination findings, and vital signs.

Baseline characteristics of each group were compared by using Fisher’s exact test for categorical variables and the t test for continuous variables.

The primary protocol-defined analysis of efficacy was a repeated-measures random regression analysis comparing the rate of change of binge frequency during the treatment period between groups. The same analysis was applied to binge day frequency, body mass index, weight, and scores on the CGI severity scale, modified Yale-Brown Obsessive Compulsive Scale, and the Hamilton Depression Rating Scale. This analysis was an intent-to-treat analysis and used all observations from all time points from all subjects who completed the baseline visit, even from subjects with only baseline assessments. The difference in rate of change was estimated by random regression methods, as employed in three previous studies of binge eating disorder (8–10) and described in Diggle et al. (21) and Gibbons et al. (22). We used a model for the mean of the outcome variable that included terms for treatment, time, and the treatment-by-time interaction. We modeled time as a continuous variable, expressed as a log (weeks + 1), with weeks ranging from 0 at baseline to 14 at the week 14 visit after random assignment. The logarithmic transformation was used because the response of the efficacy measures was approximately linear on the log scale, as is often found in treatment studies of psychiatric disorders (8–10, 22). For the analyses of binge frequency and binge day frequency, we used the logarithmic transformations log ([binges/week] + 1) and log ([binge days/week] + 1), respectively, to normalize the data and stabilize the variance. The measure of effect was the treatment-by-time interaction, which can be interpreted as the difference in the rate of change (change per unit of time), or the difference in slope with respect to time, of the efficacy measure. To account for the correlation of observations within individuals, we calculated the standard errors of the parameter estimates by using generalized estimating equations, with compound symmetry as the working covariance, as implemented by the PROC GENMOD command in SAS software.

We compared the mean percentage change from baseline in binge frequency and binge day frequency for each treatment group by using the nonparametric Wilcoxon rank sum test for both an intent-to-treat analysis (using the last observation carried forward for each subject with at least one postrandomization assessment) and a completer analysis (patients who completed 14 weeks of treatment). A comparison of the final mean score on the CGI improvement scale in each treatment group was performed by using ANOVA for the same intent-to-treat and completer populations. Response to treatment was also analyzed categorically, with response categories defined by the percentage reduction in binge frequency from baseline: remission=cessation of binges; marked=a 75%–99% reduction; moderate=a 50%–74% reduction; and none=less than a 50% reduction (8–10). Treatment group responder comparisons were made by using the exact trend test for 2-by-k ordered tables in SAS.

For laboratory measures, including weight and body fat, we computed the mean difference between endpoint and baseline measures and then compared the treatment groups by using the t test. We calculated the correlation coefficients by using rank-transformed data (Spearman rank correlation). All statistical tests were two-sided with alpha=0.05.