A prevalence study and description of alli use by patients with eating disorders

A Prevalence Study and Description of alli1 Use by of six (16.7%) with purging disorder, and three of 80 (3.8%) with an eating disorder (6.2%) reported a history of alli1 use. Ofthose, 15 (57.7%) met criteria for an eat- ing disorder, including one of 29 patients nervosa; binge eating disorder; diet pill; subtype, six of 66 patients (9.1%) with full or subthreshold bulimia nervosa, four of49 (8.2%) with binge eating disorder, one Currently, alli1 is the only FDA-approved weightloss medication available over-the-counter. The Orlistat (Xenical1) has been available by prescrip- majority of U.S. adults are currently classified as tion as a weight loss aid in the United States since overweight or obese.1 For these individuals, the 1999. In 2007, orlistat was also approved by the ability to readily obtain orlistat without a prescrip- U.S. Food and Drug Administration (FDA) for non- tion may be advantageous. A substantial percent- prescription sales under the brand name of alli1, at age of patients with an eating disorder engage in one-half the daily dose of the prescription product.
over-the-counter medication and herbal productuse to promote weight loss2 and often continue these agents despite experiencing side effects.3 Portions of this manuscript have been presented in poster format at the Eating Disorders Research Society Meeting, Brooklyn Therefore, the nonprescription availability of alli1 warrants investigation of the frequency of use of this product in patients with eating disorders.
*Correspondence to: Kristine Steffen, Neuropsychiatric Research Institute, 120 8th Street South, Fargo, North Dakota 58103.
alli1 is FDA-approved for nonprescription use in the United States by overweight patients ages 18 1 Neuropsychiatric Research Institute, Fargo, North Dakota2 and older who are also on a reduced calorie, low-fat Department of Clinical Neuroscience, University of North Dakota School of Medicine and Health Sciences, Grand Forks, diet.4 alli1 is formulated in 60 mg capsules, which are to be taken within 1 hour of each fat-containing 3 Department of Psychiatry, University of Chicago, Chicago, meal, up to three capsules per day.4 Orlistat’s phar- macological effect occurs through the inhibition of Department of Psychiatry, University of Minnesota, gastric and pancreatic lipases in the gastrointestinal 5 Columbia University College of Physicians and Surgeons, tract, which prevents triglyceride hydrolysis and results in the decreased absorption of dietary fats, Department of Psychiatry at Chapel Hill, University of North which are excreted through the feces.5 alli1 reduces 7 Laureate Psychiatric Clinic and Hospital, Tulsa, Oklahoma dietary fat absorption by approximately 25% at the 8 GlaxoSmithKline Consumer Healthcare, Parsippany, New Jersey recommended dosage.4 Efficacy increases in a dose- Published online 7 May 2010 in Wiley InterScience dependent manner up to approximately 300–400 (www.interscience.wiley.com). DOI: 10.1002/eat.20829 mg per day, at which point a plateau is observed.5 International Journal of Eating Disorders 43:5 472–479 2010 PREVALENCE STUDY AND DESCRIPTION OF ALLI1 Orlistat’s pharmacological actions occur locally supported through a research grant provided by GlaxoS- in the gut6 and less than 2% of the drug is absorbed systemically.7 Therefore, the adverse effect profile Participants filled out the Survey of Eating and Related associated with orlistat predominantly consists of a Behaviors, which is a 38-item self-report questionnaire variety of gastrointestinal side effects such as soft designed for this study to capture demographic and diag- stools, abdominal pain, steatorrhea, fecal urgency, nostic information, binge eating frequency, and compen- flatulence, and other less common side effects, satory behavior methods and frequency. Participant’s such as fecal incontinence.5 These adverse effects self-reported height and weight were used to determine increase in response to the amount of fat con- body mass index (BMI). Participants were not informed sumed,5 although they typically diminish over time that the purpose of the study was to collect data on alli1 as patients gain experience using the medication.8 use, and the questions concerning this were embedded The alli1 package label instructs patients to take a among multiple other compensatory behavior questions multiple vitamin on a daily basis at bedtime while (e.g. diuretics, Syrup of Ipecac). Probable current eating using orlistat since absorption of fat-soluble vita- disorder diagnoses were determined based upon partici- pants’ responses to survey items according to the criteria Not uncommonly, patients with eating disorders sets described in Table 2, which were used to assign each misuse medications such as laxatives, diuretics, and diet pills to compensate for binge eating and/ The study was approved by the Institutional Review or to promote weight loss. Rates of laxative abuse Boards for all five study sites and all participants pro- among outpatients who have an eating disorder vided written informed consent. In addition, participants have been reported to be 26.4% in the month prior ages 12 through 17 were also required to provide written to assessment in one study.9 Similarly, a study of informed consent from a parent or legal guardian prior patients with bulimia nervosa found that 64% of to taking the survey. All data were de-identified and the sample had used diet pills, with 18% of the maintained in a central database at the Neuropsychiatric sample having used them in the month prior to the Research Institute in Fargo, ND. Data were examined study assessment.10 This study also found the fre- quency of diuretic use to be 31%, with 21% of the Data were examined descriptively. The small sample sample having used them in the month prior to the size of alli1 users, and the discrepancy between the sam- study assessment. Therefore, it is possible that the ple size of those who used alli1 and those who had not nonprescription availability of alli1 could lead to prohibited performing valid statistical comparisons.
inappropriate use by patients with eating disorders.
Cases of nonresponse to items were treated as a negative Indeed, a few case reports of use of orlistat by indi- response to the question to enhance manuscript read- viduals with eating disorders have appeared in the ability and this is also indicated as appropriate in Tables.
literature. The majority of these case reports werepublished before orlistat became available over-the-counter. Details of these cases are summarizedin Table 1. This study was developed to quantifythe frequency of alli1 use among patients with eat- ing disorder symptoms. Therefore, a treatment-seeking sample was examined and results are sub- A total of 428 participants completed the survey.
Of those, 417 completed the question regarding ahistory of alli1 use and were explored in greaterdetail. Participants who had prior bariatric surgery(N 5 22) or who had an undeterminable BMI orbariatric surgery history (N 5 44) were excluded from eating disorder diagnostic categorization and This study consisted of a survey that was administered are presented separately. Survey numbers obtained at five eating disorder treatment facilities across the from each of the study sites are as follows: Neuro- United States, including inpatient and outpatient facili- psychiatric Research Institute, 99; University of ties, between June of 2008 and March of 2009. Partici- Minnesota, 98; University of Chicago, 100; Univer- pants ages 12 and older who presented for evaluation or sity of North Carolina, 48; Columbia University, 81.
who were in ongoing treatment programs were eligible to Data were split on the basis of whether patients participate. Those who completed the survey were com- reported a history of alli1 use or not, and are pre- pensated with a 10-dollar gift card. This study was sented along with descriptive information for the International Journal of Eating Disorders 43:5 472–479 2010 International Journal of Eating Disorders 43:5 472–479 2010 PREVALENCE STUDY AND DESCRIPTION OF ALLI1 Criteria for assigning cases to probable eating disorder diagnoses 1. Current BMI  17.5 kg/m22. No binge eating, vomiting, laxative or diuretic use in the past month3. No history of bariatric surgery.
1. Current BMI  17.5 kg/m22. Any binge eating, vomiting, laxative, or diuretic use within the past month3. No history of bariatric surgery 2. A minimum average frequency of binge eating of once per week 3. A minimum average frequency of vomiting, laxative, or diuretic use of once 2. A minimum average frequency of binge eating of once per week over the past month3. No vomiting, laxative, or diuretic use in the past month4. No history of bariatric surgery 1. BMI [ 17.5 kg/m22. No binge eating in the past month3. A minimum average frequency of vomiting, laxative, or diuretic use of once per week4. No history of bariatric surgery 1. BMI  25 kg/m22. Does not meet any other diagnostic criteria3. No history of bariatric surgery 1. BMI [ 25 but \ 30 kg/m22. Does not meet any other diagnostic criteria3. No history of bariatric surgery 1. BMI  30 kg/m22. Does not meet any other diagnostic criteria3. No history of bariatric surgery BMI or bariatric surgery history unreported 1. Participant did not report current height, weight, or both2. No history of bariatric surgery or unknown history of bariatric surgery complete sample in Table 3. The majority of the reported that they had taken alli1. Table 4 summa- sample was female and Caucasian, with a mean rizes the distribution of these patients according to age of 33.2 (613.1) years and a mean BMI of 28.3 diagnostic category. None of the 26 participants who (611.8) kg/m2. As expected in a treatment-seeking reported symptoms consistent with anorexia nerv- sample of patients with eating disorder symptoms, osa restricting subtype (AN-R), and one of 29 (3.4%) binge eating and compensatory methods such as of those with anorexia nervosa binge-purge subtype laxatives, diuretics, and vomiting in the past month (AN-BP) indicated a history of alli1 use. The fre- were relatively common, both in those who had quency of alli1 use was higher in those who were used alli1 and in those who had not. The group categorized as having full or subthreshold bulimia with a history of alli1 use was found to have a nervosa (BN), where six of 66 (9.1%) reported that higher percentage of patients who engaged in laxa- they had used the drug. Reported frequencies of tive, diuretic, diet pill, Syrup of Ipecac, and herbal alli1 use for the other eating disorder diagnoses fat burner use in the past month compared to the were four of 49 (8.2%) for full or subthreshold binge group who had not used alli1. Given the small eating disorder (BED), one of six (16.7%) for purging sample size associated with the group who had disorder (PD), three of 80 (3.8%) for eating disorder used alli1, this observation cannot be confirmed not otherwise specified (EDNOS), one of 10 (10%) for overweight, four of 85 (4.7%) for obese, six of 44(13.6%) of those with an unknown BMI or bariatric Frequency of alli1 Use Among Patients with surgery history, and none of the 22 who had previ- Of the 417 participants who responded to the Six of the 26 (23.1%) participants who had used question regarding a history of alli1 use, 26 (6.2%) alli1 reported that they had exceeded the maxi- International Journal of Eating Disorders 43:5 472–479 2010 Characteristics of sample according to history mean (6SD) BMI and age. alli1 is indicated for adult patients who are overweight. The sample consisted of a small number of participants below N 5 20). One of these patients reported a history of alli1 use and was found to be in theEDNOS category. Of the 26 participants who had used alli1, 10 (38.5%) had a BMI below 25 kg/m2 at the time the survey was conducted. Given the ret- rospective nature of this survey, however, it is pos- sible that the current BMI does not accurately rep- resent the BMI at the time when alli1 was used.
As shown in Table 4, of the 26 participants who reported that they had used alli1, 12 (46.2%) had done so in the past month, consisting of those with BN (N 5 4), BED (N 5 1), EDNOS (N 5 1), overweight (N 5 1), obese (N 5 3), and unknown BMI or bariatric Eating disorder behaviors in the past monthc surgery history (N 5 2). Characteristics of patients who met criteria for an eating disorder (or EDNOS), and reported use of alli1 in the past month (N 5 6) are described in a brief case series according to their survey responses in Table 6. Ages ranged from 24 to 58 years old, BMI ranged from 18.9 to 42.9 kg/m2, most of the patients had a history of using other med- ications for weight loss or to compensate for binge eating, and all patients shared an extreme fear of a Percentages in each column were computed as the number of positive responses divided by the total sample in the respective alli1 use group orthe complete sample (N 5 26, N 5 391, or N 5 417). Only participantswho responded to the question asking if they had ever used alli1 areincluded.
b Values based upon available data; not all participants completed all c Missing data were treated as a negative response to the question.
The population described in this study represents atreatment-seeking sample collected from five eating mum recommended dose of alli1. Data on the disorder treatment facilities across the United States.
extent to which they exceeded the recommended The study was comprised of patients who reported a dosage were not collected. The majority of these variety of eating disorder symptoms and the sample patients (N 5 4) were in the BN diagnostic category.
represented a broad range of ages and BMIs. The The mean duration of alli1 use varied according to results of this study suggest that a small subset of diagnostic category. These data are also presented patients who are presenting for evaluation or are in Table 4. Ten of the 26 patients who had used engaged in treatment in eating disorder care facili- alli1 (38.5%) reported that they had experienced ties have used alli1. The sample size associated with side effects while using the drug. Summarized alli1 use was too small in several of the diagnostic according to the descriptions provided by the par- groups to draw definite conclusions. The rates of ticipants, these included: diarrhea (N 5 3), extreme alli1 use by patients with BN (9.1%) and BED (8.2%) diarrhea (N 5 1), loose bowels (N 5 1), stomach suggest that clinicians should inquire about alli1 cramps and pain (N 5 1), fat/oily diarrhea/stools use along with other compensatory behaviors when (N 5 2), gas (N 5 2), racing/increased heart rate/ interviewing patients with eating disorders.
palpitations (N 5 2), panic (N 5 1), dizziness/faint- Patients with BN frequently use medications to ness (N 5 1), and depression (N 5 1).
compensate for binge eating through purging and/orto promote weight loss. Notably, several of thepatients who indicated alli1 use in this survey also Characteristics of Patients Who Used alli1 engaged in the use of other weight loss methods. This Patients who reported a history of alli1 use (N 5 is consistent with prior literature which suggests that 26) are further described in Table 5 according to a subset of patients with BN, as well as AN, use multi- International Journal of Eating Disorders 43:5 472–479 2010 PREVALENCE STUDY AND DESCRIPTION OF ALLI1 Description of alli1 use among patients with eating disorders a Percentages in each column were computed as the number of positive responses divided by the total sample in each diagnostic group who responded to the question asking if they had ever used alli1.
b Missing data were treated as a negative response to the question.
c Values based upon available data; not all participants completed all questions.
Age and BMI characteristics of patients with eating disorders who reported a history of alli1 use a Percentages in each column were computed as the number of positive cases divided by the total sample in each diagnostic group who responded to the question asking if they had ever used alli1.
b Values based upon available data; not all participants completed all questions.
ple purging methods. This practice has been associ- the prescription dosage of orlistat (120 mg three ated with a higher lifetime prevalence of significant times daily) to be efficacious for reducing body psychopathology, including mood, substance abuse, weight in patients who are obese with BED in com- and cluster B personality disorders.15 The use of mul- bination with either cognitive behavioral therapy18 tiple purging methods has also been associated with a or a reduced calorie diet.19 However, prior case higher level of eating disorder severity,16 and a longi- reports suggest that unmonitored use of this drug tudinal investigation of a college sample showed that by patients who binge eat can be problematic.
multiple purging methods at baseline predicted Given orlistat’s mechanism of action, gastrointesti- higher eating disorder severity at 10 year follow-up.17 nal side effects are more pronounced following a Although data addressing this issue are not available, high-fat meal. Therefore, using alli1 as a strategy to as suggested by Cumella and colleagues (2), the risk compensate for a binge eating episode with high of fat-soluble vitamin deficiency with alli1 in patients fat content could increase the adverse effect bur- with an eating disorder should be considered.
den associated with the drug. To provide over- Orlistat has been examined as a potential treat- weight or obese patients who have BED with the ment for BED.18,19 Two controlled trials have shown highest likelihood of effectiveness from orlistat, International Journal of Eating Disorders 43:5 472–479 2010 International Journal of Eating Disorders 43:5 472–479 2010 PREVALENCE STUDY AND DESCRIPTION OF ALLI1 clinicians should consider prescribing it in the dos- age used in the two extant controlled trials (120 mg Visit: http://www.ce-credit.com for additional informa- three times daily) and providing careful monitoring tion. There may be a delay in the posting of the article, so in the context of a structured treatment program continue to check back and look for the section on Eating which should also include a diet and exercise com- Disorders. Additional information about the program isavailable at www.aedweb.org Ten of the 26 participants (38.5%) in this study who had used alli1 reported that they had experi-enced side effects with the drug. The adverseeffects listed by participants were generally consist- ent with what is expected with alli1, including a va- 1. National Center for Health Statistics. Chartbook on Trends in the riety of gastrointestinal complaints. Cardiovascular Health of Americans. Hyattsville, MD: Public Health Service, 2006.
complaints including palpitations and increased 2. Cumella EJ, Hahn J, Woods BK. Weighing alli’s impact. Eating dis- heart rate and psychiatric symptoms were each order patients might be tempted to abuse the first FDA-approvednonprescription diet pill. Behav Healthc 2007;27:32–34.
reported by two patients, which are not commonly 3. Steffen KJ, Roerig JL, Mitchell JE, Crosby RD. A survey of herbal attributed to orlistat use.5 From this survey, it is not and alternative medication use among participants with eating possible to determine whether these symptoms disorder symptoms. Int J Eat Disord 2006;39:741–746.
were related to alli1, to an eating disorder, to con- 4. Alli Homepage for HealthCare Professionals, GlaxoSmithKline.
comitant medications, or to another etiology.
http://www.allihcp.com. Accessed May 2009.
5. Micromedex1 Healthcare Series (n.d.). Retrieved January, 2009, Along with the self-report nature of these data, from http://www.thomsonhc.com. Greenwood Village, CO: other limitations of this study include the inability to determine the precise temporal sequence of alli1 use 6. Filippatos TD, Derdemezis CS, Gazi IF, Nakou ES, Mikhailidis in relationship to the use of other medications for DP, Elisaf MS. Orlistat-associated adverse effects and drug inter-actions. Drug Safety 2008;31:53–65.
weight loss and binge eating. Other than asking spe- 7. Anderson J. Orlistat for the management of overweight individ- cifically for information on alli1 use, this survey did uals and obesity: A review of potential for the 60mg, over-the- not include questions designed to collect data regard- counter dosage. Exp Opin 2007;8:1733–1742.
ing which specific types of laxatives, diuretics, and 8. Bray GA. Lifestyle and pharmacological approaches to weight loss: diet pills participants were using. Also, BMI at the Efficacy and safety. J Clin Endocrinol Metab 2008;93:S81–S88.
9. Bryant-Waugh R, Turner H, East P, Gamble C, Mehta R. Misuse time the survey was completed may not have repre- of laxatives among adult outpatients with eating disorders: sented the BMI at the time alli1 was used since Prevalence and profiles. Int J Eat Disord 2006;39:404–409.
patients were asked if they had ever taken alli1.
10. Roerig J, Mitchell J, de Zwaan M, Wonderlich S, Kamran S, Eng- Therefore, alli1 use in the past month may be of bloom S, et al. The eating disorders medicine cabinet revisited: greatest relevance for this comparison. Nonpurging A clinician’s guide to appetite suppressants and diuretics. Int JEat Disord 2003;33:443–457.
weight loss methods were not assessed, such as food restriction and excessive exercise. Data on diuretic Martinez L, Turon-Gil V, Vallejo-Ruiloba J. Bulimia nervosa and mis- use in the last month that were collected in this sur- use of orlistat: Two case reports. Int J Eat Disord 2001;30:458–461.
vey were assumed to be for weight loss purposes, 12. Cochrane C, Malcolm R. Case report of abuse of orlistat. Eat although it is possible that patients were using them 13. Malhotra S, McElroy SL. Orlistat misuse in bulimia nervosa. Am for hypertension or other medical purposes.
The results of this survey suggest that patients with 14. Hagler-Robinson A. Orlistat misuse as purging in a patient with eating disorders do use alli1, although in comparison binge-eating disorder. Psychosomatics 2009;50:177–178.
to published prevalence rates of other inappropriate 15. Keel PK, Fichter M, Quadflieg N, Bulik CM, Baxter MG, Thornton et al. Application of a latent class analysis to empirically compensatory weight loss methods such as laxative misuse, the use of alli1 appears relatively uncommon at this time. The cost to purchase alli1, in compari- 16. Elder C, Haedt AA, Keel PK. The use of multiple purging meth- son to some of the other nonprescription medica- ods as an indicator of eating disorder severity. Int J Eat Disord tions, may be one factor that has led to the lower 17. Haedt AA, Edler C, Heatherton TF, Keel PK. Importance of mul- reported rates of misuse of this product relative to tiple purging methods in the classification of eating disorder other medication classes. No serious adverse effects subtypes. Int J Eat Disord 2006;39:648–654.
that could be clearly attributed to the drug were 18. Grilo CM, Masheb RM, Salant SL. Cognitive behavioral therapy reported. Given the potential for the inappropriate guided self- help and orlistat for the treatment of binge eating use of this medication by patients with eating disor- trial. Biol Psychiatry 2005;57:1193–1201.
ders, clinicians are encouraged to monitor for alli1 19. Golay A, Laurent-Jaccard A, Habicht F, Gachoud JP, Chabloz M, use along with all other medications for weight loss Kammer A, et al. Effect of orlistat in obese patients with binge eating disorder. Obes Res 2005;13:1701–1708.
International Journal of Eating Disorders 43:5 472–479 2010

Source: http://www.ce-credit.com/articles/101723/alli-101723.pdf