Feeding and Eating Conditions Not Elsewhere Classified (NEC) in DSM-5

Feeding and Eating Conditions Not Elsewhere Classified (NEC) inDSM-5

Tiffany A. Brown, MS; Pamela K. Keel, PhD; and Ruth H. Striegel, PhD

long with changes to the formally recognized eating disorders, the ADiagnostic and Statistical Man- ual, Fifth Edition Eating Disorder Work Group has recommended that the DSM- IV eating disorder not otherwise specified category be renamed: Feeding and eating conditions not elsewhere classified. This change reflects the inclusion of childhood feeding disorders within the eating disorder classification scheme, as well as the focus on reducing the ambiguous and “catch-all” nature of eating disorder not otherwise specified (EDNOS). While there is not currently enough rigorous sci-

Tiffany A. Brown, MS, is a doctoral candi- date in Clinical Psychology, Department of Psychology, Florida State University. Pamela K. Keel, PhD, is Professor of Psychology, De- partment of Psychology, Florida State Uni- versity. Ruth H. Striegel, PhD, is Professor of Psychology, Department of Psychology, Wes- leyan University. Address correspondence to: Ruth H. Strie- gel, PhD, Department of Psychology, Wesley- an University, 207 High Street, Middletown, CT 06459; email: [email protected]. Disclosure: The authors have no relevant financial relationships to disclose. Copyright ©2012, American Psychiatric Association. All rights reserved. doi: 10.3928/00485713-20121105-08 © Shutterstock

PSYCHIATRIC ANNALS 42:11 | NOVEMBER 2012 Healio.com/Psychiatry | 421 entific evidence to categorize the proposed category. The inclusion of BED as a for- specified nature of the residualDSM-IV conditions as disorders, a diagnosis of not mally recognized disorder, along with the EDNOS category itself has been a major elsewhere classified (NEC) is warranted expansion of the AN and BN categories, roadblock to initiating research and pro- if the individual has an eating disturbance has helped reduce what conditions remain viding useful guidelines for clinical prac- that is clinically significant, but does not in the “residual” diagnostic group by re- tice. With this issue in mind, further reor- meet criteria for any other feeding or eat- allocating what were formerly several ganization and specification was needed ing disorder. Thus, the NEC category is atypical or subthreshold presentations within this category to facilitate greater not meant to capture disordered eating, (eg, AN without amenorrhea or BN with understanding regarding some of the al- which can be present among the general once per week binge/purge episodes). ternative syndromes that have been iden- population, but rather, eating conditions tified in research and clinical practice. that may cause significant distress, inter- Consistent with the broader change ference with daily life, and/or increased in the nomenclature from “not otherwise risk of death or disability. specified” to “not elsewhere classified,” From a practitioner’s standpoint, in- the DSM-5 work group has focused dividuals seeking treatment for eating on specifying and defining subgroups problems are likely either experiencing within EDNOS. While DSM-IV EDNOS distress and/or impairment that are clini- included several examples of unnamed cally significant. Proposed conditions in- syndromes, DSM-5 will include several cluded in NEC are: atypical anorexia ner- named syndromes with brief descriptions vosa (atypical AN); subthreshold bulimia (but without formal diagnostic criteria) nervosa (subthreshold BN); subthresh- within three categories: (1) atypical, old binge eating disorder (subthreshold mixed, or subthreshold syndromes; (2) BED); purging disorder (PD); and night other specific syndromes; and (3) condi- eating syndrome (NES). Treatment-seek- Several studies have shown tions with insufficient information. Nota- ing samples of bulimic-related EDNOS bly, for all NEC conditions, no minimum syndromes, which would include many, that with these revisions, the frequency or duration requirements are if not most, of the conditions proposed designated. This decision was made, in for NEC, are at increased risk of death by proportion of NEC diagnoses large part, because there is no available suicide,1 supporting the severity of illness will decrease dramatically. research that would have guided more among patients who will fall within the specific frequency or duration thresh- NEC category. olds. Thus, it is left to clinical judgment Unlike the clear hierarchy proposed for Indeed, several studies have shown to distinguish a clinically significant eat- diagnoses of AN, BN, and BED, there is that with these revisions, the proportion ing disorder from disordered eating. currently no specified diagnostic hierarchy of NEC diagnoses will decrease dramati- for NEC conditions. With this understood, cally.2,3 In a community-based sample, Atypical and Subthreshold practitioners should only diagnose an indi- the use of DSM-IV vs. DSM-5 criteria Presentations vidual with a single NEC condition, based resulted in 63% EDNOS versus 53% Regarding the first category of NEC on the best match between the description NEC, respectively (P < .001).2 Similarly, (atypical AN, subthreshold BN, sub- of NEC conditions and that individual’s among a treatment-seeking sample, esti- threshold BED), descriptive changes specific constellation of symptoms. mates of EDNOS decreased from 53% were warranted based on revisions to to 25% for NEC (P < .001).3 As evident diagnostic criteria for full threshold CLINICAL UTILITY OF NEC from these studies, while proposed revi- disorders. Given that the changes to A large emphasis for revisions of sions will reduce the “catch-all” nature of diagnostic criteria for AN and BN and DSM-5 was to reduce the preponderance the EDNOS category somewhat, DSM-5 the inclusion of BED as a formal diag- of DSM-IV EDNOS diagnoses without NEC is still likely to represent a sizeable nosis reallocated a substantial propor- reducing diagnostic validity of recog- minority of patients in clinical settings tion of patients from DSM-IV EDNOS nized categories, such as AN and BN. and the majority of clinically significant to DSM-5 full-threshold diagnoses, the Perhaps the largest revision is the recom- eating disorders in community settings. definitions of what constitute atypical or mendation that BED be “promoted” from In addition to the need to reduce the subthreshold presentations were in need a form of EDNOS to a formal diagnostic profusion of EDNOS diagnoses, the un- of modification.

422 | Healio.com/Psychiatry PSYCHIATRIC ANNALS 42:11 | NOVEMBER 2012 Purging Disorder contributed to the chronicity of obesity Given that increased freedom has been Regarding other specific syndromes, among a case series of 23 obese individu- designated to the practitioner to define purging disorder (PD), which reflects the als.14 While NES has not been mentioned what qualifies as “low weight” when di- most commonly used specific name for in any iteration of the DSM, advocacy agnosing AN, diagnoses of AN among this symptom configuration across recent from researchers led to a rigorous scien- those who are underweight will largely ex- studies,4 was first introduced as a poten- tific evaluation of NES to determine how clude them from being assigned to NEC. tial description of an atypical eating dis- it may be addressed within DSM-5. Because a central feature of AN is low order in 1986.5 An unnamed version of The largest limitation within the NES weight, the descriptor of “atypical” ap- PD was added as an example descriptor research literature has been lack of agree- pears appropriate for this NEC condition. of EDNOS in DSM-III-R,6 and as Ex- ment on a definitive syndromal definition The other two conditions in this sub- ample 4 in DSM-IV7 (“the regular use of and some debate regarding whether the category of NEC include subthreshold inappropriate compensatory behavior by condition described as NES best repre- conditions. Example 3 from DSM-IV individuals with normal body weight af- sents a syndrome or a set of symptoms.15 EDNOS has been aptly named sub- ter eating small amounts of food”). More Several studies have found support for in- threshold BN, which accurately reflects detailed and rigorous studies of the con- creased pathology on a number of psycho- a syndrome identical to DSM-5 BN, al- dition have produced a clearer descrip- logical correlates for NES as compared to beit with a frequency of less than once tion and name for purging disorder. healthy controls; however, the variables per week or less than 3 months in dura- Several studies have demonstrated examined have differed across research tion. Since the designation of BED as a the clinical significance of PD and its studies (eg, depressive symptoms, anxiety full-threshold eating disorder, subthresh- distinction from non-eating disordered symptoms, substance-use problems, psy- old BED also will be included within the individuals.8-10 A lesser number of stud- chiatric diagnoses) making it difficult to NEC conditions (with identical frequen- ies have demonstrated physiological and draw definitive conclusions overall.15 cy criteria as for subthreshold BN). psychological differences between PD Further, several other methodological Reflecting the definition used in re- and BN;11,12 however, these differences issues exist within the NES research lit- search on PD in recent years, PD will be have been less consistently studied and erature, including lack of data regarding described by recurrent purging behaviors found.13 Finally, studies examining the the prevalence and course of NES, and (eg, self-induced vomiting, laxative, di- distinctiveness of PD from AN, and evi- lack of data supporting the distinctive- uretic, or enema misuse) for the purpose dence of a distinctive course or treatment ness of NES from other eating disorders. of influencing shape or weight, in the outcome for PD have not yet emerged in Thus, to provide a brief definition of this absence of objective binge episodes. The the literature. syndrome to stimulate further research, label of compensatory behaviors was Given the growing body of research the DSM-5 Eating Disorder Work Group narrowed to include only purging behav- on this syndrome, the DSM-5 Eating Dis- decided to introduce NES as a named ex- iors, based on evidence that including order Work Group believed that discus- ample of a NEC condition.15 fasting and excessive exercise in defini- sion of the status of PD in DSM-5 was tions of PD significantly reduced distinc- warranted. Ultimately, it was decided SUGGESTED NEC CHANGES tions from normality.16 that PD should remain as NEC, as op- FOR DSM-5 Further specification regarding the posed to being included as a diagnosis Regarding atypical, mixed, or sub- purpose of purging for weight or shape on par with AN or BN.13 However, the threshold conditions, given changes reasons was included to clarify that this inclusion of a name and clinical descrip- made to full-threshold diagnoses, cer- syndrome should be limited to a patho- tion will make it possible for clinicians to tain syndromes previously described in logical eating condition and should not more easily access literature on this con- DSM-IV EDNOS (Example 1: AN with- include culturally sanctioned purification dition and may facilitate more uniform out amenorrhea) will be diagnosed with rituals or purging due to anxiety.13 DSM- research efforts to examine treatment re- AN in the DSM-5. Example 2 in DSM- 5 criteria also do not make an explicit sponse, course, and outcome in PD. IV has been retained as a descriptor and statement regarding body weight, but named “atypical” AN. The description as per rules regarding NEC conditions NIGHT EATING SYNDROME for atypical AN remains consistent with broadly, criteria for AN cannot be met. The next of these specific syndromes that in DSM-IV, with the exception of the While substantial literature on NES not elsewhere classified, night eating expansion of the weight specifier to in- has evolved, definitions used have large- syndrome (NES), was first introduced in clude individuals both within and above ly varied from study to study.15 As such, 1955 as a pattern of eating that potentially normal weight. the relatively longer description of NES

PSYCHIATRIC ANNALS 42:11 | NOVEMBER 2012 Healio.com/Psychiatry | 423 compared with other NEC syndromes who do not meet criteria for pica, rumi- have included “expanded” diagnos- was designed to provide a uniform nation disorder, or avoidant/restrictive tic groups within their studies,19-22 and working definition for the field. NES food intake disorder. these studies have found efficacy for tentatively encompasses recurrent epi- these treatments among both the strin- sodes of night eating, which are defined APPLICATION TO CLINICAL gently defined and “expanded” diagnos- as either including evening hyperpha- PRACTICE tic groups. To our knowledge, only one gia (consuming a substantial portion of One of the benefits of theDSM-5 study, at a tertiary treatment center, has one’s calories after the evening meal) or NEC for clinical practice is the util- examined treatment outcome in PD. The nocturnal eating (“awakening during the ity of having specifically named syn- authors found no difference in remission night and consuming food before return- dromes, which can help patients with and completion rate for PD patients as ing to sleep”15). these conditions by providing a label compared with AN and BN patients, sug- Several specifications were added to to identify and legitimize their symp- gesting that treatments that have prom- help guide the definition of NES, includ- ise for patients with AN and BN may be ing noting that the individual must be adapted to those with PD.23 aware of his/her behavior, to differentiate While named NEC conditions have While these results are encouraging, this syndrome from sleep disorders that this study was not a RCT and did not include eating without awareness. Fur- several general benefits for clinical focus on a specific form of interven- ther, these behaviors cannot be solely the practice, currently, there is limited tion; thus, no direct recommendations consequence of local norms or as a sec- for treatment can be made for PD from ondary consequence of a disrupted sleep data on treatment outcome and this study. While no studies have specifi- cycle, as the first would preclude a diag- recommendations. cally examined treatments for atypical/ nosis of a mental disorder and the second subthreshold conditions or PD, there would be better accounted for by a sleep, has been research conducted on transdi- rather than an eating, disturbance. toms. This, in turn, can help instill a agnostic treatments, most notably Fair- NES must also be associated with dis- sense of mutual understanding between burn’s transdiagnostic cognitive-behav- tress and/or impairment, which is simi- patient and practitioner and facilitate ioral therapy (CBT) and its “enhanced” lar to the criteria employed for BED, to hope for symptom improvement on the counterpart (CBT-E).24 limit overpathologizing behaviors that part of both individuals.17 Further, the In an RCT of 57 patients with BN, 92 do not disrupt an individual’s function- more specific syndrome names can help patients with EDNOS, and seven patients ing. Several exclusion criteria are also streamline communication between with BED, the authors found that CBT noted to ensure that this syndrome is not practitioners compared with the broad and CBT-E outperformed a waitlist con- best characterized by another existing category of EDNOS, which provided no trol group both at end-of-treatment and disorder (eg, BED, substance-use dis- indication regarding the configuration 60-week follow-up.24 Importantly, treat- orders, and other medical or psychiatric of symptoms. ment outcome did not differ by diagno- disorders). While named NEC conditions have sis, with 52.7% of patients with BN and Also of note, DSM-IV EDNOS Ex- several general benefits for clinical prac- 53.3% of patients with EDNOS having ample 5, which referred to repeatedly tice, currently, there is limited data on a global Eating Disorder Examination chewing and spitting out large amounts treatment outcome and recommenda- score less than one standard deviation of food without swallowing, has been re- tions for NEC conditions. This is likely above community norms at end-of- moved from DSM-5 NEC, due to insuf- due to the lack of randomized controlled treatment. Although the specific forms ficient data on this clinical presentation. trials (RCTs) for conditions that have not of EDNOS included in the sample were The third category of NEC conditions been formally defined.13 For presenta- not specified, this study demonstrates a serves as a residual category for any tions that resemble symptom configu- theoretically sound approach to treating feeding or eating condition that is clini- rations for recognized eating disorders the function of eating disorder behaviors cally significant, but does not meet crite- (eg, AN and atypical AN, BN and sub- and demonstrates the utility of CBT and ria for any other feeding or eating disor- threshold BN, and BED and subthresh- CBT-E for BN and related EDNOS who der or condition. Although discussion of old BED), guidance on treatment may were not underweight. this section has been primarily focused be gained by referring to best treatment Thus, such an approach serves as a on eating conditions, this category also practices for the formal conditions.18 sound starting point for practitioners includes feeding conditions in children Consistent with this, several RCTs treating these conditions. As with treat-

424 | Healio.com/Psychiatry PSYCHIATRIC ANNALS 42:11 | NOVEMBER 2012 ment for any disorder, clinicians should formation available on how to best treat disorder and bulimia nervosa. Arch Gen Psy- conduct careful and regular assess- their patients. chiatry. 2007;64(9):1058-1066. 12. Keel PK, Wolfe BE, Gravener JA, Jimerson ments of symptom levels both prior to Following introduction of BED as DC. Co-morbidity and disorder-related dis- and throughout treatment, to determine a named EDNOS in the DSM-IV, there tress and impairment in purging disorder. Psy- whether the designated intervention is was an explosion of research on its clini- chol Med. 2008;38(10):1435-1442. 13. Keel PK, Striegel-Moore RH. The validity and associated with reductions in the targeted cal correlates, treatment, course, and out- clinical utility of purging disorder. Int J Eat symptom features. come, making it possible for clinicians Disord. 2009;42(8):706-719. Similarly, little treatment research to better assist their own patients with 14. Stunkard AJ, Grace WJ, Wolff HG. The night- eating syndrome; a pattern of food intake has been conducted on NES. Due to the BED. The inclusion of named NEC in among certain obese patients. Am J Med. disruption in circadian eating patterns the DSM-5 may presage a similar expan- 1955;19(1):78-86. observed in NES, the selective serotonin sion of information on these prevalent 15. Striegel-Moore RH, Franko DL, Garcia J. The reuptake inhibitor sertraline has been but understudied conditions. validity and clinical utility of night eating syndrome. Int J Eat Disord. 2009;42(8):720-738. examined. To date, the only randomized 16. Haedt AA, Keel PK. Comparing definitions of placebo-controlled trial for NES was REFERENCES purging disorder on point prevalence and as- conducted using sertraline (50-200 mg) 1. Crow SJ, Peterson CB, Swanson SA, et al. sociations with external validators. Int J Eat Increased mortality in bulimia nervosa and Disord. 2010;43(5):433-439. in a small sample of outpatients over 8 other eating disorders. Am J Psychiatry. 17. Keel PK, Brown TA, Holland LA, Bodell LP. weeks (n=34).25 Sertraline demonstrated 2009;166(12):1342-1346. Empirical classification of eating disorders. significantly greater global improvement 2. Keel PK, Brown TA, Holm-Denoma J, Bodell Annu Rev Clin Psychol. 2012;8:381-404. LP. Comparison of DSM-IV versus proposed 18. National Institute for Clinical Excellence. compared with placebo, but sertraline DSM-5 diagnostic criteria for eating disor- Eating Disorders – Core Interventions in the was not associated with significantly ders: reduction of eating disorder not other- Treatment and Management of Anorexia Ner- greater decrease in night eating symp- wise specified and validity. Int J Eat Disord. vosa, Bulimia Nervosa, and Related Eating toms (nocturnal eating, caloric intake af- 2011;44(6):553-560. Disorders. NICE Clinical Guideline No. 9. 3. Fairburn CG, Cooper Z. Eating disorders, London; 2004. ter the evening meal), compared with the DSM-5 and clinical reality. Br J Psychiatry. 19. Striegel-Moore RH, Wilson GT, DeBar L, et placebo group. Thus, at this time, only 2011;198(1):8-10. al. Cognitive behavioral guided self-help for limited support exists for the use of ser- 4. Keel PK. Purging disorder: subthreshold vari- the treatment of recurrent binge eating. J Con- ant or full-threshold eating disorder? Int J Eat sult Clin Psychol. 2010;78(3):312-321. traline in the treatment of NES. Disord. 2007;40 Suppl:S89-94. 20. McIntosh VV, Jordan J, Carter FA, et al. Three 5. Mitchell JE, Pyle RL, Hatsukami D, Eckert psychotherapies for anorexia nervosa: a ran- ED. What are atypical eating disorders? Psy- domized, controlled trial. Am J Psychiatry. CONCLUSION chosomatics. 1986;27(1):21-28. 2005;162(4):741-747. The DSM-IV EDNOS category has 6. American Psychiatric Association. Diagnostic 21. Schmidt U, Lee S, Beecham J, et al. A ran- signified an overrepresented and uninfor- and Statistical Manual of Mental Disorders. domized controlled trial of family therapy and mative clinical diagnosis, and proposed 3rd ed. Revised ed. Washington, DC: Ameri- cognitive behavior therapy guided self-care for can Psychiatric Publishing; 1987. adolescents with bulimia nervosa and related changes to the DSM-5 NEC category 7. American Psychiatric Association. Diagnostic disorders. Am J Psychiatry. 2007;164(4):591- should reduce some of the clinical ambi- and Statistical Manual of Mental Disorders. 598. guity associated with unspecified eating 4th ed. Text Revision. Arlington, VA: Ameri- 22. Debar LL, Striegel-Moore RH, Wilson GT, can Psychiatric Publishing;2000. et al. Guided self-help treatment for recurrent disorders observed in practice. While it is 8. Pinheiro AP, Bulik CM, Sullivan PF, Machado binge eating: replication and extension. Psy- likely that the NEC group will represent PP. An empirical study of the typology of bu- chiatr Serv. 2011;62(4):367-373. a sizeable minority of patients in clinical limic symptoms in young Portuguese women. 23. Tasca GA, Maxwell H, Bone M, et al. Purg- Int J Eat Disord. 2008;41(3):251-258. ing disorder: psychopathology and treatment practice, proposed changes will hope- 9. Sullivan PF, Bulik CM, Kendler KS. Genetic outcomes. Int J Eat Disord. 2012;45(1):36-42. fully aid practitioners in making more epidemiology of binging and vomiting. Br J 24. Fairburn CG, Cooper Z, Doll HA, et al. Trans- informed and specific diagnoses and Psychiatry. 1998;173:75-79. diagnostic cognitive-behavioral therapy for better guide treatment of their patients’ 10. Striegel-Moore RH, Franko DL, Thompson patients with eating disorders: a two-site trial D, Barton B, Schreiber GB, Daniels SR. An with 60-week follow-up. Am J Psychiatry. feeding and eating problems. To achieve empirical study of the typology of bulimia ner- 2009;166(3):311-319. this goal, researchers will need to ex- vosa and its spectrum variants. Psychol Med. 25. O’Reardon JP, Allison KC, Martino NS, et amine the clinical utility of these newly 2005;35(11):1563-1572. al. A randomized, placebo-controlled trial 11. Keel PK, Wolfe BE, Liddle RA, De Young of sertraline in the treatment of night eating proposed conditions, and clinicians will KP, Jimerson DC. Clinical features and physi- syndrome. Am J Psychiatry. 2006;163(5):893- need to actively seek out the newest in- ological response to a test meal in purging 898.

PSYCHIATRIC ANNALS 42:11 | NOVEMBER 2012 Healio.com/Psychiatry | 425