Alcohol & Eating Disorders: Prevalence, Onset, & Personality Profiles Study, Lecture notes of Psychiatry

Download Alcohol & Eating Disorders: Prevalence, Onset, & Personality Profiles Study and more Lecture notes Psychiatry in PDF only on Docsity! Alcohol Use Disorder Comorbidity in Eating A Multicenter Study Disorders: Cynthia M. Bulik, Ph.D.; Kelly L. Klump, Ph.D.; Laura Thornton, Ph.D.; Allan S. Kaplan, M.D.; Bernie Devlin, Ph.D.; Manfred M. Fichter, M.D.; Katherine A. Halmi, M.D.; Michael Strober, Ph.D.; D. Blake Woodside, M.D.; Scott Crow, M.D.; James E. Mitchell, M.D.; Alessandro Rotondo, M.D.; Mauro Mauri, M.D.; Giovanni B. Cassano, M.D., ER.C.Psych.; Pamela K. Keel, Ph.D.; Wade H. Berrettini, M.D.; and Walter H. Kaye, M.D. Background: Eating disorders and alcohol use disorders (AUDs) commonly co-occur, although the patterns of comorbidity differ by eating disorder sub- type. Our aim was to explore the nature of the co- morbid relation between AUDs and eating disorders in a large and phenotypically well-characterized group of individuals. Method: We compared diagnostic and personality profiles of 97 women with lifetime anorexia nervosa only, 282 women with lifetime bulimia nervosa only, and 293 women with a lifetime history of both ano- rexia nervosa and bulimia nervosa or anorexia ner- vosa with binge eating (ANBN) (DSM-IV criteria). All individuals were participants in a multicenter study of the genetics of anorexia nervosa and bulimia nervosa. We explored pattern of onset, Axis I and I1 comorbidity, and personality characteristics of indi- viduals with and without AUDs by eating disorder subtype. Personality characteristics were assessed .with the Multidimensional Perfectionism Scale, the Temperament and Character Inventory, and the Barratt Impulsivity Scale. more prevalent in women with ANBN and bulimia nervosa than in women with anorexia nervosa (p = .0001). The majority of individuals reported primary onset of the eating disorder, with only one third reporting the onset of the AUD first. After eat- ing disorder subtype was controlled for, AUDs were associated with the presence of major depressive disorder, a range of anxiety disorders, and cluster B personality disorder symptoms. In addition, individu- als with AUDs presented with personality profiles marked by impulsivity and perfectionism. Conclusions: Individuals with eating disorders and AUDs exhibit phenotypic profiles characterized by both anxious, perfectionistic traits and impulsive, dramatic dispositions. These traits mirror the pattern of control and dyscontrol seen in individuals with this comorbid profile and suggest that anxiety modu- lation may be related to alcohol use in this group. (J Clin Psychiatry 2OO4;65:10O0-1006) Results: Alcohol use disorders were significantly Received July 16, 2003; accepted Jan. 21, 2004. Fmm the Department qfpsychiatty, Univer.si!v qfNorth Camlina at Chapel Hill, Chapel Hill (De Bulik): the Department qfP.sychologv, Michigan State Univer.sitj East Lansing (De Klump); the Department of Psychiatry, Univer,sit.v qfPittsburgh, Pithburgh. Pa. (Drs. Thornton, Devlin. and Kaye); the Pmgram.for Eating Di.sorder.s and the Department qf Psychiatry, Tomnto General Hospital, Toronto, Ontario, Canada (Dr.9. Kaplan and Wood.side); Klinik Roseneck, Hospita1,for Behavioral Medicine, afiliated with the Universitv ofMunich, Prien, Germany [DE Fichter): New York Presbyterian Hospital, WeiN Medical College of Cornell Univer.sity, White Plains, N. L (De Halmi); the Department qfPsychiatry and Behavioral Science, University qf Cal$ornia at Los Angeles, Los Angeles (DK Stmber): the Department qfpsychiatty, Univer.sity qfMinnesota, Minneapolis (De Cmw); the Neumpsychiatric Research Institute, Fargo, N.D. (De Mitchell); the Department qf P.s.vchiatry, Neumbiologvp Pharmacologv, and Biotechnologies, Univer.sity qfPisa, Pisa, Italy (Drs. Rotondo, Mauri, and Cassano); the Department q f P s y c h o 1 o ~ ~ Univemity qflowa, Iowa City (DK Keel); and the Center,for Netimbiologv and Behavior; Univer.sity ofPennsylvania, Philadelphia (DK Berrettini). Financial disclosure appears at the end qf this article. The authors thank the Price Foundation, Geneva, Switzerland,,for their support qf the clinical collection qfpariicipants and data analysis. The authors are indebted to the participating.familie.s,fir their contribution qf time and effort in support qf this study. Corresponding author and reprints: Cynthia M. Bulik, Ph.D., Department qfP.s.vchiatty, I01 Manning DK CB #7160, University qf North Camlina at Chapel Hill, Chapel Hill, NC 27599-7160 (e-mail: cbulih@med. unc.edii). stimates of the prevalence of comorbid substance E abuse andor dependence in clinical samples of women with bulimia nervosa range between 3% and 50%.' Parameters that have varied across studies and that influence comorbidity estimates include the nature of the clinical service (inpatient vs. outpatient), the definition of and assessment procedures for both eating disorders and substance use disorders, whether current or lifetime diag- noses were assessed, the distorting effects of any exclu- sion criteria for clinical trials, and the age of patients stud- ied.' Overall, the majority of studies have observed an elevated prevalence of alcohol use disorders (AUDs) in clinical samples of women with bulimia nervosa. In a re- view of 25 studies of the prevalence of alcohol abuse or dependence in women with bulimia nervosa in clinical samples, Holderness et al.' calculated a median preva- lence of 22.9%. 1000 Bulik et al. Comorbidity of AUDs appears to differ across sub- types of eating disorders and may reflect differences in patterns of familial liability to psychiatric illness or differ- ences in temperament and personality between individu- als with anorexia nervosa and bulimia nervosa.' In sum, the majority of studies suggest that although AUDs are relatively uncommon in the restricting subtype of ano- rexia nervosa, the prevalence ofAUDs in women with an- orexia nervosa with bulimic features appears to be com- parable to or exceed that in women with normal-weight bulimia n e r v ~ s a . ~ - ~ Careful illumination of onset patterns may shed light on differential mechanisms of comorbid association or unique mechanisms of symptom causation across eating disorder subtypes. For example, if AUD preceding bu- limia nervosa is associated with statistically different pat- terns of Axis I or I1 comorbidity compared with bulimia nervosa preceding AUD, then nested within the larger category of bulimia nervosa may be heterogeneity of etiologic factors. In some cases, for example, AUD may function to moderate binge eating, whereas other onset patterns may imply a common underlying liability. There- fore, variations in risk or liability factors may be implied or imputed through the study of comorbidity, as well as the chronology of onset, of these disorders. In essence, this approach is central to the development of paradigms for refining the study of causative mechanisms under- pinning coaggregated disorders. Several pertinent issues central to this approach have been only partially addressed, due either to the size and nature of the samples investigated or to limitations in the range of variables that have been explored. For example, pattern of onset of AUDs relative to the onset of eating disorders remains unclear, in part because many studies have -explored undifferentiated alcohol and other drug abuse. Wiseman et a1.I' comparcd comorbidity patterns in female patients with an eating disorder and substance dependence with patterns in individuals with an eating disorder and without substance dependence. They did not focus on alcohol only and included a broad range of sub- stances. Of the patients with an eating disorder and sub- stance dependence, 34.8% reported that the onset of the substance abuse preceded the onset of the eating disorder, and the remainder reported that their eating disorder pre- ceded the onset of the substance abuse. The authors found that individuals in this group who developed substance abuse first were more likely to be dependent on a greater number of substances and were also more likely to have cluster B personality disorders compared with those who developed an eating disorder first and those with no co- morbid substance use disorders. Individuals in the group who reported that the onset of eating disorders was first reported the greatest number of comorbid diagnoses and were significantly more likely to have panic disorder and social phobia compared with the other 2 groups. Although these findings are intriguing, the impact of including a broad range of substance use disorders on the results is unclear. Given that AUDs are commonly reported forms of substance abuse reported by women with bulimia, we chose to focus on alcohol abuse and dependence. The goals of the present study were to further charac- terize the nature of the comorbid relationship between eat- ing disorders and AUDs by (1) comparing the prevalence of alcohol abuse and dependence across well-defined eat- ing disorder subtypes, (2) documenting patterns of onset of AUDs relative to eating disorders across eating dis- order subtypes, ( 3 ) determining whether there are differ- ences in comorbidity patterns and personality between in- dividuals who experience eating disorders versus AUDs first, (4) exploring diagnostic comorbidity profiles associ- ated with AUDs in individuals with eating disorders, and ( 5 ) exploring psychological and personality features asso- ciated with AUDs in individuals with eating disorders. METHOD Collaborative Arrangements The current initiative was developed through a co- operative arrangement between the Price Foundation (Geneva, Switzerland), the University of Pittsburgh (Pitts- burgh, Pa.), and other academic sites in North America and Europe. The sites of collaborative arrangement, se- lected on the basis of experience in the assessment of eat- ing disorders and geographical distribution, included Uni- versity of Pittsburgh (W.H.K.); Cornell University, White Plains, N.Y. (K.A.H.); University of California at Los Angeles (M.S.); University of Toronto, Toronto, Ontario, Canada (A.S.K., D.B.W.); University of Munich, Munich, Germany (M.M.F.); University of Pisa, Pisa, Italy (A.R., G.B.C., M.M.); University of North Dakota, Fargo (J.E.M.); University of Minnesota, Minncapolis (S.C.), University of Pennsylvania, Philadelphia (W.H.B.); and Harvard University, Cambridge, Mass. (P.K.). Each site obtained approval separately from its own institutional re- view board. All participants provided written informed consent. Participants Inclusion criteria for probands and affected relatives have been described in detail elsewhere.'' The sample was ascertained on the basis of a proband with bulimia nervosa who also had at least 1 affected relative with an eating dis- order. For probands, acceptance into the study required that they met the following criteria. (1) DSM-IV lifetime diagnosis of bulimia nervosa, purging type. Purging had to include regular vomiting, with other means of purging optional; binge eating and vomiting must have occurred at least twice a week for a duration of at least 6 months. (2) Age between 13 and 65 years. A current or lifetime history of anorexia nervosa was acceptable (e.g., bulimia nervosa 1001 J Clin Psychiatry 65:7, July 2004 ,4lcohol Use Disorder Comorbidity in Eating Disorders Table 3. Logistic Regression Predicting Alcohol Use Disorders on the Basis of Axis I and 11 Disorders and Eating Disorder Groupa borderline personality disorder (24% vs. 11 %), and cluster B symptoms (26% vs. 13%). Axis 1 and I1 Diagnoses ~~~i~~ Disorder Diagnosis x2 pb OR (95% C I ) ~ Diagnosis, x~~ Temperament and Personality Characteristics Fnr thpcp ~ ~ I ~ T J C P C w? compared personality and temperament -.-..'..L'-- I , Major depressive disorder 7 99 ,005 1.84 (1.25 to 2.70) 29.28 Generalized anxiety 0.01 29.69 disorder and without AUD V d I l d U K b across individuals with 1s. Those participants with AUDs I z w = ~3.5 f4.5 vs. 21.4 f 4.2), Obsessive-compulsive 11.41 ,0007 1.73 (1.27 to 2.36) 28.56 scored significantly higher on the motor subscale 4 . . of the BIS (mean ' nn ~ *' Posttraumatic stress 5 3 5 .02 1.75(1.11to276) 3245 higher on the sclt-transcendence scale of the TCI disorder (16.2 k 6.4 vs. 14.6 k 6.7), and higher nn MPS tntnl Social phobia Specific phobia 5.95 ,015 1.77 (1.13 to 2.77) 30.65 over mistakes (31.1 f 9.1 disorder Panic disorder 1.65 28 33 7.61 ,006 1.77 (1.20 to2 60) 32.15 perfectionism (95.9 + 3 3 7 7 Axis I1 ...~_ _ _ _ _._ . _ _ - _- A I 4s. 89.1 f 24.3), concern VS. 28.9 i lO.l), doubts Avoidant 1.87 28.71 about actions (12.8f4.1 vs. 11.8 +4.1), parental 28 51 criticism (12.0 i 4.7 vs. 10.2 rt 4.7), and parental ex- Obsessive-compulsive 3 07 Borderline 14.04 .0002 2.40 (1 55 to 3.70) 25.86 Cluster B 13.16 .0003 2.25 (1 48 to 3.42) 25.42 pectations (14.9 k 5.7 vs. 13.8 & 5.9) than women Cluster C 2.55 28 60 without AUDs. Odds ratios and values are pre- interactions, these were removed from the model. I "As there were no significant alcohol use disorder-by-eating disorder bValue is indicated if significant. .0001. Table 4. Regression Analyses (using GEE) Predicting Alcohol Use Disorders on the Basis of Psychological and Personality Features and Eating Disorder Group Eating Disorder Psychological Variable Diagnosis Scale X2 pa OR (95%CI)a X2 p Barratt Impulsivity Scale DISCUSSION This study serves to replicate previous findings in the literature using well-defined diagnostic sub- groups of individuals with eating disorders and extends the findings by addressing important un- answered questions regarding the nature of the co- morbid relation between eating disorders and AUDs. As previously f ~ u n d , ~ . ~ , ' alcohol abuse or depen- Cognitive 1.33 10.83 0044 dence was significantly less prevalent in individuals 3: Motor 5.25 .02 1.52 (1.12 to 2.08) 8.31 ,0157 with anorexia ne-vnci nnlv thnn in in Non-planning 1.28 9 0 4 ,0109 Temperament and bulimia nervosa ( Character Inventory alcohol abuse or dependei ~ - . _.___. --_ -. . .- . uyu u,l.J bI1yI1 All ... dividuals with mly. Although tlie prevalence of nce did not differ sip- 1 C I I W '1I.Y'. ULlU " U L 1 " A ' U L'Yl *VDLL or the ANBN group were internie- - Harm avoidance 1.52 31 28 ,0001 nificantly betweep th- A N R W anrl hiil;mio n ~ t ? r n c ~ Novelty seeking 1.02 29 19 ,0001 . Reward dependence 0.01 29 75 ,oool groups, the rates fi I Peisistence 0.01 29.17 0001 diate between the rates for the anorexia nervosa Cooperativeness 0.00 30.76 ,0001 Self-directedness 3.46 29.40 ,0001 Self-transcendence 4.53 .03 1.21 (1 02 to 1.43) 27.59 ,0001 tributable to diffel Multidimensional tween individuals with a predominantly anorectic I Perfectionism Scale predominartl.r h..l;-;- o.r-..+-- .-++-. - - A / Concern over mistakes 6 27 .01 1.23 (1 05 to 1.45) 29.74 ,0001 Doubts about actions 4 0.1 .04 1.18 (1.01 to 1.39) 28.01 .0001 Or and ge Personal standards 0.13 30.64 .0001 AUDs tends to be elevated in families of individuals lLly UUll l l l lCl J y l I I p u I I I p u L L c I I I UllW netic factors, as the prevalence of Organization 0.03 30.05 0001 with bulimia nervc-- L.,& --+ -----., :- ---.-"- 19 Parental criticism 16.70 ,0001 1 43 (1.21 to 1.69) 30 49 0001 Parental expectations 4 72 .03 1.20 (1 02 to 1.42) 32 58 ,0001 Patterns of ons Total uerfectionism 9.63 ,002 1.30 (1.10 to 1.54) 31 56 0001 did not differ across eating disorder subtwes. The I J3d VU1 l l U L 0 1 1 U l G A l c l LIGIVUSLL. ;et of eating disorders and AUDs I d * "Value is indicated if significant. Abbreviation: GEE = generalized estimating equation. most common pattern was for the onset of the eating disorder to precede the onset of the AUD. This pat- tern could simply reflect availability, since access to dieting or binge eating is not bound by legal age restrictions as is access to alcohol. Alternatively, this pattern could reflect the typical ages at onset of eating- and alcohol-related disorders. Another hypothesis posits that alcohol abuse or dependence could follow the onset of an eating disorder due to the increased reinforcing effi- cacy of alcohol once food deprivation commences. The onset of both anorexia nervosa and bulimia nervosa is dent effects of eating disorder subgroup. Alcohol use disorders were significantly associated with increased prevalence of major depressive disorder (80% of those with major depressive disorder vs. 67% of those without it had an AUD), obsessive-compulsive disorder (51% vs. 36%), posttraumatic stress disorder (19% vs. 1 l%), social phobia (25% vs. 17%), specific phobia (20% vs. 13%), J Clin Psychiatry 65:7, July 2004 I 1004 Bulik et al. generally marked by periods of food restriction or dieting. In anorexia nervosa, these periods persist indefinitely; in bulimia nervosa, these periods can be prolonged prior to the onset of binge Animal data clearly docu- ment increased self-administration of other reinforcing substances after food deprivation and weight 1oss:4-26 and these principles may be operative in humans who are also vulnerable to the reinforcing properties of ai~01101,~~ al- though this effect has not been observed consistently in the laboratory.?' Given the retrospective nature of this study, we were unable to obtain sufficiently reliable his- torical information to determine whether alcohol use did indeed emerge during periods of intense food restriction. Finally, AUDs were strongly associated with the presence of various anxiety disorders in this population, giving rise to the interpretation that AUDs may emerge in women with eating disorders due to the actual or anticipated anxi- olytic effects of alcohol. The presence of AUDs was associated with various comorbid Axis I and 11 disorders. Specifically, AUDs were associated with an increased risk of depression, obsessive-compulsive disorder, a variety of anxiety disor- ders, and cluster B symptoms and borderline personality disorder. Studies of epidemiologic samples indicate that anxiety and AUDs frequently C O - O C C U ~ * ~ - ~ ~ and that the onset of some anxiety disorders, such as social phobia, which may predispose to drinking,33,34 often predates the onset of AUD, despite the fact that alcohol may not attenuate social anxiety symptom^.^^,^' Substantial data now support the hypothesis that anxi- ety disorders are pervasive in individuals with eating dis- orders and that their onset most commonly predates the onset of the eating di~order .~~-~O Silberg and Bulik have identified a common genetic factor that contributes to vul- nerability to early eating and anxiety disorders (J. Silberg, Ph.D.; C.M.B., manuscript submitted). We hypothesize that food restriction or dieting may be anxiolytic for indi- viduals who are vulnerable to anorexia nervosa and may represent a primary risk and maintaining force for eating disorder^.^' Due to the pervasiveness of anxiety symp- toms, alcohol use may be an additional means whereby individuals with eating disorders attempt to modulate their pervasive anxiety symptoms. These results concur with a previous investigation indicating that the presence of an anxiety disorder in women with bulimia nervosa was associated with both a history of anorexia nervosa and with an earlier age at onset of alcohol or drug use dis- o rde r~ .~ ' It cannot be entirely ruled out, however, that the directionality of the relationship is reversed and that anxi- ety is a consequence of alcohol abuse. In terms of personality profiles, after eating disorder subgroup was controlled for, the presence of AUD was associated with greater motor impulsivity, higher self- transcendence, and higher internal (concern over mis- takes) and external (parental criticism and parental expec- 1005 tations) perfectionism scores. These findings mirror the clinical observation of both overcontrol and dyscontrol in individuals with comorbid eating disorders and AUDs. Elevated self-transcendence could reflect a greater ten- dency toward psychoticism or dissociative experiences or could reflect religious tendencies, possibly related to re- covery attempts. In summary, whereas prior studies have tended to fo- cus on impulsive traits in women with bulimia and AUD, we have shown that both anxious and impulsive traits are characteristic of these individuals. Moreover, the absence of significant interactions between eating disorder sub- type and AUD suggests that similar mechanisms may un- derlie the relation between AUDs and eating disorders across subtypes. Our sample size, however, precluded detection of significant interactions. Finally, although we did not uncover robust predictors of pattern of onset, or an explanation for why it is more typical for the onset of eating disorders to precede or coincide with the onset of AUDs, it is noteworthy that, unlike anxiety disorders, which tend to predate the onset of eating disorder^,^' AUDs tend to follow the onset of eating disorders. Fur- ther studies of predictors of the devclopment of AUDs may assist with identifying those at risk and preventing the emergence of AUDs in individuals with eating dis- orders. The results of this study must be interpreted within the context of some methodological limitations. First, the sample included only probands with bulimia nervosa who also had at least I affected relative with an eating disorder and their affected relatives. Cases that derive from en- riched pedigrees may differ in terms of both severity and comorbidity from sporadic cases. Second, as with all studies that employ lifetime psychiatric history, our data on age at onset of eating disorders and AUDs rely on indi- vidual recall. It is unknown whether accuracy of recall of age at onset differs across disorders or whether other is- sues (e.g., concerns about revealing underage drinking) may influence self-report. Countering these limitations was the fact that we were able to explore the nature of the comorbid relation be- tween eating disorders and AUDs in a large and pheno- typically well-characterized sample of individuals with eating disorders. Our careful documentation of age at on- set of both AUDs and eating disorders enabled a tnore meticulous exploration of patterns and correlates of onset and enabled a more comprehensive comparison across eating disorder subtypes. These data highlight the importance of continued vigi- lance on the part of health care professionals for the emer- gence ofAUDs in women with bulimia nervosa as well as individuals with features of both anorexia and bulimia nervosa. Even if AUDs are not present at initial evalua- tion or commencement of treatment, ongoing attention to the possibility of their later emergence in individuals with J Clin Psychiatry 65:7, July 2004 ~- . Alcohol Use Disorder Comorbidity in Eating Disorders eating disorders is warranted. Particularly vulnerable are those women who manifest comorbid anxiety or depres- sive disorders. Disclosure of ofi-label usage: The authors have determined that, to the best of their knowledge, no investigational information about pharmaceutical agents has been presented in this article that is outside U S . Food and Drug Administration-approved labeling. In the spirit of hll disclosure and in compliance with all ACCME Essential Areas and Policies, the faculty for this CME activity were asked to complete a full disclosure statement. Thc information re- ceived is as follows: Drs. Fichter and Strober received granthesearch support from Price Foundation. Drs. Bulik, Klump, Thomton, Kaplan, Devlin. Halmi, Woodside, Crow, Mitchell, Rotondo, Mauri, Cassano, Keel, Beirettini, and Kaye have no significant commercial relation- ships to disclose relative to their presentations. REFERENCES 1. Bulik CM, Sullivan P, Slof MCT. Comorbidity of eating disorders and substance-related disorders. In: Kranzla HR, Tinsey JA, eds. Dual Diagnosis and Psychiatric Treatment: Substance Abuse and Comorbid Disorders. New York, NY: Marcel Dekker; 2004:317-348 disorders and substance abuse: review of the literature. Int J Eat Disord 2. Holdemess C, Brooks-Gunn J, Warren M. Co-morbidity of eating 1994; 16: 1-35 3. Hcrzog DB, Keller MB, Sacks NR, et al. Psychiatric comorbidity in treatment-seeking anorexics and bulimics. J Am Acad Child Adolesc Psychiatry 1992;3 I :81&818 4. Braun DL, Sunday SR. Halmi KA. Psychiatric comorbidity in patients with eating disorders. Psychol Med 1994;24:859-867 5. Stem S, Dixon K, Sansone R, et al. Psychoactive substance usc disorder in relatives of patients with anorexia nervosa. Compr Psychiatry 1992; 33 :207-2 1 2 6. Henzel H. Diagnosing alcoholism in patients with anorexia nervosa. Am J Drug Alcohol Abuse 391(4;10:461466 7. Hudson J. Pope H, Jonas J, et al. Phenomenologic relationship of eating disorders to major affective disorder. Psychiatry Res 1983;9:345-354 8. Hudson 1, Pope H, Jonas J, et al. Family history study of anorexia and bulimia. Br J Psychiatry 1983;142:133-138 9. Bulik C, Sullivan P, Epstein L, et al. Drug use in women with anorexia and bulimianervosa. Int J Eat Disord 1992;11:213-225 earing disorders: impact of sequence on comorbidity. Compr Psychiatry 10. .Wiseman C, Sunday S, Halligan P, et al. Substance dependence and 1999;40:332-336 11. Kaye WH, Devlin B, Barbarich N, et al. Genetic analysis of bulimia nervosa: methods and sample description. Int J Eat Disord 2004:35: 556570 12. Hebebrand J, Himelmann GW, Heseker H, et al. Use of percentiles for the body mass index in anorexia nervosa: diagnostic, epidemiological, and thwapeutic considerations. Int J Eat Disord 1996;19:359-369 13. Fichter MM, Herperk S, Quadflieg N, et al. Stmctured Interview for Anorexic and Bulimic Disorders for DSM-IV and ICD-IO: updated (third) revision. Int J Eat Disord 1998;24:217-249 14. First M, Spitzer R, Gibbon M, et al. Structured Clinical Interview for DSM-IV Axis I Disorders, Research Version, Patient Edition. New York, NY Biometric Research, New York State Psychiatric Institute; 1997 15. First MB, Gibbon M, Spitzer RL, et al. Structured Clinical Interview for DSM-IV: Personality Disorders. Washington. DC: American Psychiatric Press; 1997 Copnit Ther Res 1990;14:449468 16. Frost R, Marten P, Lahart C. et al. The dimensions ofperfectionism. 17. Cloninga CR, PrzybeckTR, Svrakic DM, et al. The Temperament and Character Inventory (TCI): A Guide to Its Development and Use. St Louis, Mo: Center for Psychobiology of Personality, Washington University; 1994 18. Barratt E. The biological basis of impulsiveness. Pers Individ Dif 19. Lilenfeld L, Kaye W. Greeno C, et al. A controlled family shidy of restricting anorexia and bulimia nervosa: comorbidity in probands and disorders in frst-degree relatives. Arch Gen Psychiatry 1998;55: 603-61 0 20. Bulik C, Sullivan PF, Fear J, et al. Predictors of the development of bulimia nervosa in women with anorexia neivosa. J Nerv Ment Dis 1997;185:704-707 21. Bulik C, Sullivan P, Carter F, et al. Initial manifestation of disordered eating behavior: dieting versus binging. Int J Eat Disord 1997;22: 22. Abraham S, Beumont P. How patients describe bulimia or binge eating. Psychol Med 198212625635 23. Fairbum CG, Cooper JP. Self-induced vomiting and bulimia nervosa: ‘an undetcctedproblcm. BMJ 1982;284:1153-1155 24. Carroll ME, France CP, Meisch RA. Food deprivation increases oral and intravenous drug intake in rats. Science 1979;205:319-321 25. Carroll ME, Meisch RA. Determinants of increased drug self- administration due to food deprivation. Psychopharmacology (Berl) 26. Carroll ME, Mcisch RA Increased drug-reinforced behavior due to food deprivation. In: Thompson T, Dews PB, Barrett E, eds. Advances ui Behavioral Pharmacology. New York, NY Academic Press; 1984:47-88 27. Keyes A, Brozek J. Henschel A. The Biology of Human Starvation. Minneapolis, Minn: University of Minnesota Press; 1959 28. Bulik C, Brinded E. The effect of food deprivation on alcohol consump- tion in bulinlic and control women. Addiction 1993;88:1545-1551 29. Crum RM, Pratt LA. Risk of heavy drinking and alcohol use disorders in social phobia: a prospective analysis. Am J Psychiatry 2001;158: 30. Regier DA, Famier ME, Rae DS, et al. Comorbidity of mental disorders with alcohol and other drug abuse: results from the Epidemiologic Catchment Area (ECA) Study. JAMA 1990;264:2511-2518 31. Kessler RC, Crum RM, Warner LA, et al. Lifetime co-occwence of DSM-111-R alcohol abuse and dependence with other psychiatric disorders in the National Comorbidity Survey. Arch Gen Psychiatry 32. Kushner MG, Sher IU, Beitman BD. The relation between alcohol prob- lems and the anxiety disorders. Am J Psychiatry 1990;147:685-695 33. Magee W, Eaton W, Wittchen H, et al. Agoraphobia, simple phobia, and social phobia in the National Comorbidity Survey. Arch Gen Psychiatry 34. Merikangas KR, Stevens DE, Fenton B, et al. Co-morbidity and familial aggregation of alcoholism and anxiety disorders. Psychol Med 1998;28: 35. h’aftolowitz DF, Vaughn BV, Ranc J, et al. Response to alcohol in social phobia. Anxiety 1994;1:9&99 36. Himle JA, Abelson JL, Haghightgou H, et al. EEect of alcohol on social phobic anxiety. Am J Psychiatry 1999;156:1237-1243 37. Deep A, N a g L, Wcltzin T, et al. Prcinorbid onsct of psychopathology in long-term recovered anorexia nervosa. Int J Eat Disord 1995;17:291-298 38. Bulik C, Sullivan P, Fear J. et al. Eating disorders and antecedent anxiety disorders: a controlled study. Acta Psychiatr Scand 1997;96:101-107 39. Godart NT, Flanient MF, Lecrubier Y, et al. Anxiety disorders in anorexia nervosa and bulimia nervosa: co-morbidity and chronology of appear- ance. Eur Psychiatry 2000;15:3815 40. Kaye W, Bulik C, Thomton L, et al. Anxiety disorders comorbid with bulimia and anorexia nervosa. Am J Psychiatry. In press 41. Kaye W, Barbarich N, Putnam K. et al. Anxiolytic effects of acute trqFto- phan depletion in anorexia nervosa. Int J Eat Disord 2003;33:257-267 42. Bulik C, Sullivan P, Carter F, et al. Lifetime anxiety disorders in women with bulimia nervosa. Compr Psychiatry 1996;37:368-374 1983;4:387-391 195-201 1981;71: 197-200 1693-1700 1997;54:3 13-321 1996;53:159-168 773-788 For the CME Posttest for this article, see pages 1024-1026. J Clin Psychiatry 65:7, July 2004 1006