Purging Disorder

May 01, 2008

A syndrome described as purging following the ingestion of normal or small amounts of food in normal-weight persons has gained increasing attention in the field of eating disorders. Various terms have been used in the literature for this newly characterized syndrome, with purging disorder and eating disorder not otherwise specified-purging (only), or EDNOS-P, used most frequently.

A syndrome described as purging following the ingestion of normal or small amounts of food in normal-weight persons has gained increasing attention in the field of eating disorders. Various terms have been used in the literature for this newly characterized syndrome, with purging disorder and eating disorder not otherwise specified-purging (only), or EDNOS-P, used most frequently.1-8

Because purging disorder has not been included in any nosological system, there is no official definition for the disorder. As a consequence, definitions have varied across studies examining the condition, particularly with regard to the minimum required frequency of purging behaviors (ranging from once a month to twice a week for a 3-month period) and absence of objectively large binge-eating episodes (from complete absence to less than twice a week for a 3-month period).9 The Table presents criteria used in the most recent studies published by the University of Iowa Eating Behaviors Research Clinic laboratory.2,3 These criteria were modeled after those used in DSM-IV to define bulimia nervosa, which are being evaluated for possible revision in DSM-V.


Purging versus nonpurging compensatory behaviors
The definition of purging behavior has been provided in DSM-IV and includes self-induced vomiting, laxative abuse, diuretic abuse, and the use of enemas to influence weight or shape. Although some authors have examined a more broadly defined compensatory eating disorder characterized by both purging and nonpurging behaviors in the absence of binge eating, most studies have focused on a syndrome characterized by purging behaviors.10,11 Limited data have been collected to evaluate whether the distinction between purging and nonpurging (eg, fasting and excessive exercise) compensatory behaviors should be maintained either for a diagnosis of bulimia nervosa or for the formation of boundaries for purging disorder.



Frequency criterion
With current suggestions to revise the episode frequency criterion for bulimia nervosa from at least twice a week for 3 months down to at least once a week for 3 months, a similar threshold might be warranted in the definition of purging disorder. Binford and le Grange5 required purging at least once a week for a 6-month period (a duration similar to that recommended for binge-eating disorder in DSM-IV).

Results from their studies were generally consistent with those from studies using higher minimum frequency thresholds, which suggests that there may not be a clinically meaningful difference between a once-versus twice-weekly minimum frequency criterion for purging behavior. Because there is no official threshold, and data for evaluating the optimal frequency threshold are limited, it is important for clinicians and researchers who examine purging disorder to specify the frequency of purging behavior. In a clinical setting, this assessment is crucial for tracking progress in treatment. In a research setting, this information is crucial for communicating possible sources of differences (or the lack thereof) between patients described in one article and patients described in another article.

Absence of objectively large binge-eating episodes
The central difference between a diagnosis of bulimia nervosa and purging disorder is the presence versus the absence of objectively large binge-eating episodes-episodes that involve the consumption of a large amount of food and the loss of control over eating during the episode. However, some investigators have diagnosed purging disorder in those with binge-eating episodes at subthreshold frequency and purging episodes at threshold frequency, while others have excluded those with objectively large binge-eating episodes.1-3,5-7,12 It seems likely that patients with subthreshold binge-eating frequency and threshold purging frequency might fit on a continuum with those patients with threshold binge-eating and purge frequency, potentially reflecting a less severe symptom presentation.7 It does not seem that such individuals would reflect a new or distinct eating disorder different from bulimia nervosa.

In contrast, there is some evidence that defining purging disorder by the absence of objectively large binge-eating episodes identifies a syndrome associated with a distinct subjective and physiological response to food intake compared with bulimia nervosa.2 Specifically, women with purging disorder who have no objectively large binge-eating episodes report significantly greater postprandial fullness and demonstrate a significantly greater postprandial cholecystokinin response to food intake compared with women who have bulimia nervosa.2 Such biological differences may confer differences in response to pharmacological treatments. However, there are no current data to test this hypothesis.

Normal weight range
The central difference between a case of purging disorder and a case of anorexia nervosa with purging is the presence of low weight in anorexia nervosa and its absence in purging disorder. Low weight in combination with binge-purge behaviors in anorexia nervosa has been associated with several distinctive features compared with the presence of binge-purge behaviors among normal-weight persons, including differences in cross-cultural representation,13 longitudinal course and outcome,14 and mortality.15 Thus, it seems likely that low weight also would be associated with clinically meaningful differences between anorexia nervosa with purging and purging disorder. However, there have been no studies that have examined distinctions between anorexia nervosa and purging disorder to my knowledge. Instead, previous studies have used bulimia nervosa as the comparison group in examinations of purging disorder.9 Purging disorder may reside on a continuum with anorexia nervosa-binge eating/purging type with a degree of weight loss/weight below normal that reflects the severity of illness.

Lifetime history of other eating disorders
Some studies have restricted the diagnosis of purging disorder to those with no history of anorexia nervosa or bulimia nervosa, while others have only excluded a history of bulimia nervosa.1,6 Both cases represent attempts to define syndromes that would not be captured by current eating disorder definitions. This approach further separates purging disorder from a residual or partially remitted phase in bulimia nervosa.

Although a diagnosis of bulimia nervosa excludes individuals who concurrently meet criteria for anorexia nervosa, and a diagnosis of eating disorder excludes individuals who concurrently meet criteria for anorexia nervosa or bulimia nervosa, there is no precedent for using lifetime history of eating disorder diagnoses as exclusion criteria for a current eating disorder diagnosis. This has led to criticism of the potential proliferation of distinct eating disorder diagnoses when an individual transitions from meeting criteria for anorexia nervosa to meeting criteria for bulimia nervosa to meeting criteria for binge-eating disorder.16

Indeed, the longitudinal instability of symptom presentation has been one of the greatest criticisms of current eating disorder diagnoses. However, this potential seems to be overstated in many cases because prospective longitudinal studies suggest that individuals are much more likely to retain the diagnosis they received at presentation than they are to cross over to another eating disorder diagnosis.17 The highest reported rates of crossover often come from studies that rely on retrospective reports of past histories of other eating disorder diagnoses.18 Such methods introduce errors and bias because participants may agree with symptom thresholds in the absence of a specific recall of their lowest weight or frequency of binge-eating episodes.19 Thus, it is unclear whether past eating disorder diagnoses should be used as exclusion criteria in the evaluation of current eating disorder symptoms for purging disorder or any other eating disorder.


Most studies that have examined purging disorder have used some form of the Eating Disorders Examination (EDE) to evaluate research participants.9,20 The main advantage of this measure is the careful distinction between objective bulimic episodes and subjective bulimic episodes. Both involve the subjective experience of having eaten too much at one time and feeling a loss of control over eating during the episode. However, objective bulimic episodes involve consuming an amount of food that is objectively larger than most people would eat under similar circumstances. In contrast, subjective bulimic episodes involve consuming an amount of food that is not necessarily larger than most people would eat under similar circumstances.

Several of the women whom we have assessed for purging disorder endorse subjective bulimic episodes (also referred to as subjective binge-eating episodes). Using assessments that do not discriminate the amount of food consumed during self-reported binge-eating episodes would probably misidentify several of these women as having bulimia nervosa-purging type.

Clinicians may question the importance of the amount of food consumed during binge-eating episodes that are followed by purging, and such questions have been raised in research studies as well.10,12,21 The findings from these studies indicate that there is no meaningful difference in the clinical significance of syndromes characterized by purging on the basis of the amount of food consumed before purging. However, there do appear to be significant differences in the levels of associated distress, subjective test-meal responses, and physiological responses to a test meal.2,3,7

Moreover, bulimia nervosa has been defined by the occurrence of large, out-of-control binge-eating episodes since its inception, and all carefully conducted studies of bulimia nervosa have ensured that participants endorse objectively large binge-eating episodes.22 Thus, generalizing results of such studies to individuals who do not have this central feature is problematic because it is not possible to know which findings reflect patterns associated with binge eating and which reflect purging in samples with bulimia nervosa, because both features are present.

Although the EDE has a section devoted to discriminating objective from subjective bulimic episodes, the Structured Clinical Interview for DSM-IV Axis I disorders (SCID-I) can be modified slightly to achieve the same aim.23 Specifically, the SCID-I uses an initial probe into the experience of loss of control over eating, which is potentially relevant for individuals with purging disorder. The SCID-I then probes for consumption of a large amount of food during such episodes. The key modification required to make the SCID-I useful for the assessment of purging disorder is to disregard the skip rules in the assessment. Specifically, if participants deny experiencing a loss of control over eating, standard SCID-I rules advise the interviewer to skip the section and not to probe for the potential use of inappropriate compensatory behaviors (including purging behaviors). Similarly, if a participant endorses a loss of control over eating but denies consuming a large amount of food, the SCID-I rules instruct interviewers to skip the section.

By eliminating these skip rules, information about inappropriate compensatory behavior is probed, whether or not the person reports objectively large binge-eating episodes. A similar modification is required for the frequency/duration criterion because a person with purging disorder will not endorse objectively large binge-eating episodes that recur twice a week over a 3-month period but may endorse this as a minimum frequency for purging. In such instances, study-specific rules can be used to capture information about purging frequency separately from information about binge-eating frequency, and the question concerning the undue influence of weight and shape on self-evaluation can be posed for anyone who reports recurrent purging behavior. Furthermore, there is a designation after the coding for bulimia nervosa for which the interviewer can code purging versus nonpurging subtype.

As may be obvious from the previous paragraph, a notable advantage of the EDE over the SCID-I for the assessment of purging disorder is that the EDE does not use skip rules that have to be revised or ignored to capture cases of purging disorder. Furthermore, the EDE provides individual codings for frequency of specific forms of inappropriate compensatory behavior, such that it is possible to determine the exact frequency of individual purging behaviors, purging behaviors combined (relevant for current research criteria for purging disorder), or all inappropriate compensatory behaviors combined (relevant for coding DSM-IV criteria for bulimia nervosa). However, the SCID-I has the advantage of assessing lifetime histories of anorexia nervosa, bulimia nervosa, and binge-eating disorder, and the eating disorders module (Module H) requires considerably less time to administer than the full EDE even if the SCID-I is modified to eliminate all skip rules.

One question that has emerged is the reliability and validity of assessment of subjective versus objective binge-eating episodes based on recall, with the accuracy of recall for subjective binge-eating episodes being modest.19 This is important for the reliability and validity of distinctions between lifetime diagnoses of bulimia nervosa and purging disorder.


Of interest, accuracy of the recall for self-induced vomiting frequency based on the EDE has been demonstrated to be very high,19 suggesting that the accuracy of lifetime diagnoses of purging disorder that do not require the presence of subjective binge-eating episodes also may be quite high. Several studies report high interrater reliability for interview assessments of bulimia nervosa and purging disorder using the EDE either for current diagnosis or for lifetime diagnosis.1,2,7 Similar data have not been presented for the SCID-I.

Evidence that individuals with purging disorder display distinct physiological responses to a test meal compared with the responses of individuals who have bulimia nervosa further support the validity of a purging disorder diagnosis based on EDE assessment.2 However, when assessments of food consumption are based entirely on self-report, it is likely that some individuals may exaggerate their food intake while others may be unwilling or unable to accurately describe the large amount of food they consume during binge-eating episodes. Finally, the determination of what constitutes a large amount of food can be quite subjective because this often depends on the context in which food is eaten.

In our laboratory, we have adopted a threshold of 1000 kcal within 2 hours, which reflects the upper limit of what healthy controls have been observed to eat within a 2-hour period in feeding laboratory studies24 as well as the upper limit of what college women endorse eating within a 2-hour period on a self-report questionnaire.25 Using this limit reduces problems with interrater reliability, although it does not eliminate the problem of inaccurate self-report. Of note, this is not a unique problem in the diagnosis of purging disorder; it also can be quite challenging to evaluate whether reported episodes are large enough to qualify for a diagnosis of bulimia nervosa.

Treatment considerations
There have been no controlled treatment trials for purging disorder. Thus, there are no evidence-based treatments for this condition. Some have advocated a transdiagnostic cognitive-behavioral approach for the treatment of all eating disorders, and such an approach could be extended to the treatment of purging disorder.26 A central tenet of cognitive-behavioral therapy is the use of repeated assessments throughout treatment to examine whether interventions are favorably impacting symptom levels. Whether a clinician chooses an approach that involves cognitive-behavioral, interpersonal, or psychodynamic interventions, it would be useful for clinicians to assess symptom levels throughout treatment in order to document the potential efficacy of a treatment that uses a single-case series approach (see Barlow and Hersen27 for a detailed description of this approach). In such instances, clinicians can make important contributions to the field by describing possibly effective treatments that could be evaluated in larger randomized controlled trials.

Purging disorder is a form of eating disorder not otherwise specified in DSM-IV. At this time, it is unclear whether this will remain true following the publication of DSM-V. Key criteria for new diagnostic entities in DSM-V include evidence that doing so will improve a clinician's ability to care for patients in terms of understanding that a syndrome has a unique course, set of complications, or treatment needs.28 At this time, we have limited data that suggest that the course of purging disorder is similar to that observed in bulimia nervosa over a 6-month prospective follow-up period and based on retrospective recall of lifetime data.1,6 No data regarding complications of purging disorder have been reported thus far, although it seems reasonable that complications associated with purging behavior, such as hypokalemia, would be evident in this group.29 Finally, as noted above, there are no data on evidence-based treatments for purging disorder. The absence of information on these topics is consistent with the adage that "we study what we define."30

It is notable that information regarding the distinctive course, complications, and treatment response of anorexia nervosa versus bulimia nervosa emerged following the inclusion of these diagnoses in DSM rather than forming the justification for their initial inclusion. However, there are costs associated with the proliferation of diagnoses in subsequent editions of DSM in the absence of supportive data.

One possible solution would be the inclusion of purging disorder in the DSM-V appendix for criteria sets provided for further study. This approach yielded considerable data on binge-eating disorder following the publication of DSM-IV, such that arguments can be made regarding this syndrome's distinctive course and treatment response.31 Although this introduces the risk of reifying diagnostic criteria that have not themselves been adequately vetted,28 this cost seems small in relation to the potential benefits of identifying evidence-based treatments for purging disorder. Published studies indicate that purging disorder affects a substantial minority of late adolescent girls and young women, with lifetime prevalence estimates that are on par with those for anorexia nervosa and bulimia nervosa.6,32 Thus, it is crucial to develop a better understanding of how to help these individuals.




1. Keel PK, Haedt A, Edler C. Purging disorder: an ominous variant of bulimia nervosa? Int J Eat Disord. 2005;38:191-199.
2. Keel PK, Wolfe BE, Liddle RA, et al. Clinical features and physiological response to a test meal in purging disorder and bulimia nervosa. Arch Gen Psychiatry. 2007;64:1058-1066.
3. Keel PK, Wolfe BE, Gravener JA, Jimerson DC. Comorbidity and disorder-related distress and impairment in purging disorder. Psychol Med. In press.
4. Steiger H, Bruce KR. Phenotypes, endophenotypes, and genotypes in bulimia spectrum eating disorders. Can J Psychiatry. 2007;52:220-227.
5. Binford RB, le Grange D. Adolescents with bulimia nervosa and eating disorder not otherwise specified- purging only. Int J Eat Disord. 2005;38:157-161.
6. Wade TD, Bergin JL, Tiggemann M, et al. Prevalence and long-term course of lifetime eating disorders in an adult Australian twin cohort. Aust N Z J Psychiatry. 2006;40:121-128.
7. Wade TD. A retrospective comparison of purging type disorders: eating disorder not otherwise specified and bulimia nervosa. Int J Eat Disord. 2007;40:1-6.
8. Bulik CM, Von Holle A, Hamer R, et al. Patterns of remission, continuation and incidence of broadly defined eating disorders during early pregnancy in the Norwegian Mother and Child Cohort Study (MoBa). Psychol Med. 2007;37:1109-1118.
9. Keel PK. Purging disorder: subthreshold variant or full-threshold eating disorder? Int J Eat Disord. 2007;40(suppl):S89-S94.
10. Tobin DL, Griffing A, Griffing S. An examination of subtype criteria for bulimia nervosa. Int J Eat Disord. 1997;22:179-186.
11. Mond J, Hay P, Rodgers B, et al. Use of extreme weight control behaviors with and without binge eating in a community sample: implications for the classification of bulimic-type eating disorders. Int J Eat Disord. 2006;39:294-302.
12. Keel PK, Mayer SA, Harnden-Fischer JH. Importance of size in defining binge eating episodes in bulimia nervosa. Int J Eat Disord. 2001;29:294-301.
13. Keel PK, Klump KL. Are eating disorders culture-bound syndromes? Implications for conceptualizing their etiology. Psychol Bull. 2003;129:747-769.
14. Herzog DB, Dorer DJ, Keel PK, et al. Recovery and relapse in anorexia and bulimia nervosa: a 7.5-year follow-up study. J Am Acad Child Adolesc Psychiatry. 1999;38:829-837.
15. Keel PK, Dorer DJ, Eddy KT, et al. Predictors of mortality in eating disorders. Arch Gen Psychiatry. 2003;60:179-183.
16. Fairburn CG, Cooper Z. Thinking afresh about the classification of eating disorders. Int J Eat Disord. 2007;40(suppl):S107-S110.
17. Fichter MM, Quadflieg N. Long-term stability of eating disorder diagnoses. Int J Eat Disord. 2007;40(suppl):S61-S66.
18. Tozzi F, Thornton LM, Klump KL, et al. Symptom fluctuation in eating disorders: correlates of diagnostic crossover. Am J Psychiatry. 2005;162:732-740.
19. Peterson CB, Miller KB, Johnson-Lind J, et al. The accuracy of symptom recall in eating disorders. Compr Psychiatry. 2007;48:51-56.
20. Fairburn CG, Cooper Z. The eating disorder examination. In: Fairburn C, Wilson GT, eds. Binge-Eating: Nature, Assessment, and Treatment. 12th ed. New York: Guilford Press; 1993:317-331.
21. Pratt EM, Niego SH, Agras WS. Does the size of a binge matter? Int J Eat Disord. 1998;24:307-312.
22. Russell G. Bulimia nervosa: an ominous variant of anorexia nervosa. Psychol Med. 1979;9:429-448.
23. First MB, Gibbons M, Spitzer RL, et al. Structured Clinical Interview for DSM-IV Axis II Personality Disorders (SCID-II). Washington, DC: American Psychiatric Press; 1997.
24. Kaye WH, Weltzin TE, McKee M, et al. Laboratory assessment of feeding behavior in bulimia nervosa and healthy women: methods for developing a human-feeding laboratory. Am J Clin Nutr. 1992;55:372-380.
25. Keel PK, Cogley CB, Ghosh S, Lester NA. What constitutes an unusually large amount of food for defining binge episodes? Presented at: the Academy for Eating Disorder's 10th International Conference on Eating Disorders; April 25-28, 2002; Boston.
26. Fairburn CG, Cooper Z, Shafran R. Cognitive behaviour therapy for eating disorders: a "transdiagnostic" theory and treatment. Behav Res Ther. 2003;41:509-528.
27. Barlow DH, Hersen M. Single Case Experimental Designs: Strategies for Studying Behavior Change. 2nd ed. Elmsford, NY: Pergamon Press; 1984.
28. Walsh BT. DSM-V from the perspective of the DSM-IV experience. Int J Eat Disord. 2007;40(suppl):S3-S7.
29. Greenfeld D, Mickley D, Quinlan DM, Roloff P. Hypokalemia in outpatients with eating disorders. Am J Psychiatry. 1995;152:60-63.
30. Walsh BT, Kahn CB. Diagnostic criteria for eating disorders: current concerns and future directions. Psychopharmacol Bull. 1997;33:369-372.
31. Wilfley DE, Bishop ME, Wilson GT, Agras WS. Classification of eating disorders: toward DSM-V. Int J Eat Disord. 2007;40(suppl):S123-S129.
32. Favaro A, Ferrara S, Santonastaso P. The spectrum of eating disorders in young women: a prevalence study in a general population sample. Psychosom Med. 2003;65:701-708.