European Child & Adolescent Psychiatry [Suppl 1] 12:4753 (2003) DOI 10.

1007/s00787-003-1107-7

Maria Grigoroiu-Serbanescu Sanda Magureanu Stefan Milea Iuliana Dobrescu Elvira Marinescu

Modest familial aggregation of eating disorders in restrictive anorexia nervosa with adolescent onset in a Romanian sample

Prof. Dr. Maria Grigoroiu-Serbanescu () Psychiatric Genetics and Developmental Psychopathology Research Unit The Alexandru Obregia Psychiatric Hospital Sos. Berceni, 10, O. P. 8 R-75622 Bucharest, Romania Fax: +40-1/3 34-71 64 E-Mail: mserban@dnt.ro Prof. Dr. Sanda Magureanu, M. D. Prof. Dr. Stefan Milea, M. D. Dr. Iuliana Dobrescu, M. D. Carol Davila Medical University The Alexandru Obregia Psychiatric Hospital Sos. Berceni, 10, O. P. 8 R-75622 Bucharest, Romania Elvira Marinescu, M. S. W. Psychiatric Genetics and Developmental Psychopathology Research Unit The Alexandru Obregia Psychiatric Hospital Sos. Berceni, 10, O. P. 8 R-75622, Bucharest, Romania

I Abstract The study of the familial psychopathology in relatives of restrictive anorexia nervosa (AN) probands whose diagnosis was verified during a long-term follow-up was aimed at determining behavioural phenotypes with which AN could share the genetic liability. A total of 185 first degree relatives of 68 restrictive AN patients with adolescent onset followed up for 5 to 18 years and 198 first degree relatives of 68 normal women were investigated. DSM-III-R criteria were used. The lifetime rate of clinical AN was 1 % and the rate of any eating disorders was 2 % in female proband relatives versus 0 % in control relatives. No case of bulimia nervosa (BN) was found in proband relatives. The heritability of AN was low (0.18) when only the

full-blown AN was considered in relatives and modest (0.36) when also a case of subthreshold AN was added. There were significantly higher rates of anxiety disorders (14.6 %) and unipolar major depression (8.3 %) in female proband relatives and schizo-spectrum disorders (8.3 %) and alcoholism (13.1 %) in male proband relatives compared to relatives of controls. Restrictive AN might share partial liability with phenotypes expressing emotional restraint and anxiety. A sex effect of the heterotypically affected relative on the vulnerability for AN was suggested. I Key words anorexia nervosa family study genetics comorbidity

Introduction
Twin studies evidenced the involvement of a genetic component in anorexia nervosa (AN) [11, 21]. Holland et al. [11] found a concordance rate of 56 % for AN in monozygotic twins but the concordance rate decreased dramatically in dizygotic twins (5 %). Reports about monozygotic twins discordant for AN were also published [7, 35]. In a large twin female sample (N = 2163) drawn from the general population [36] no genetic component was detected in AN. Bulik et al. [5] reviewed the twin studies of AN and came to the conclusion that no precise contribution of additive genetic and environmental factors to AN may be inferred. But twins are a

special population that is not representative for the general population of anorexic patients. Family studies based on the ascertainment of anorexic patients recruited from clinical settings or from general population do not always converge in their findings concerning the degree of familial aggregation of the AN and other eating disorders; some studies found moderate to high familiality [2931], while other studies reported no significant difference in familial aggregation of AN and of other eating disorders between relatives of AN subjects and relatives of healthy controls [4, 15, 16]. Except the study by Strober et al. [31] all the other studies mentioned above did not consider the longitudinal course of AN when enrolling the probands into the

I/48

European Child & Adolescent Psychiatry, Vol. 12, Supplement 1 (2003) Steinkopff Verlag 2003

family-genetic study. Outcome research of AN documented that an important percent of the initially restrictive anorexic patients develop bulimic episodes after some years [26, 27, 34]; thus the proband phenotype changes. Lilenfeld et al. [14] analysed separately the families of purely restrictive anorexic patients excluding those patients that could have developed bulimic episodes during the natural course of the disorder in a longitudinal study design (3-year follow-up of 31 AN probands). In contrast with previous studies, this study reported significantly higher rates only of eating disorders (ED) not otherwise specified and of any ED in relatives of anorexic patients compared to relatives of control women screened only for ED, but rates of full-blown AN (or BN) were not different. A case-control study by the same group [31] with a sample of 152 purely restrictive anorexics also followed up for minimum 5 years and a control sample screened for any psychiatric disorder led to somewhat different conclusions in the sense that the familiality of AN was significant and evidence for a common genetic liability for AN and bulimia nervosa was found. But families with multiply affected relatives beyond the proband were rare also in this study. Because of its not striking familiality and of its clinical picture different family and comorbidity studies tried to link AN to other phenotypes such as affective disorders [6, 9, 35], addictive disorders [19, 33], obsessive-compulsive disorder [4, 31]. Recent Scandinavian studies pointed out the relationship between AN and autistic traits in proband relatives [16, 17]. The aims of our work were 1) the study of the familial psychopathology in first degree relatives of purely restrictive AN probands with adolescent onset whose phenotype was defined during a long-term follow-up (5 to 18 years) compared to normal controls screened for any psychiatric disorder and 2) the study of the comorbidity of AN in probands in connection with the familial psychopathology. Both aims were subordinated to the objective of determining those behavioural phenotypes that have the closest relationship to the restrictive AN and with which AN could share the genetic liability.

The patients took part in a long-term follow-up study on outcome of AN started in 1984 in our clinic and were retained from a sample of 90 AN patients collected over several years.All patients had a restrictive AN at the first psychiatric investigation. Twenty-two patients (24.4 %) were excluded either because they developed bulimic episodes after entering the study or because they could not be followed-up for at least 5 years. At the last investigation the probands were aged 18 to 26 (mean = 22.3; SD = 2.7).A subsample of 41 probands was also included in a transcultural follow-up study [26] being diagnosed according to two criteria sets (DSM-III-R and ICD10) with complete overlap.

I Probands families
The AN probands had 185 first degree relatives; 147 (79.4 %) of them were directly interviewed. Psychopathological information about the remainder 38 relatives was indirectly collected through the family history method. The informants were all the other family members directly interviewed as well as the patients themselves. Sex distribution of the proband first degree relatives is shown in Table 1. There is no significant difference between male (N = 87) and female (N = 98) relatives.

I Normal controls
The controls were 68 normal women aged 1827 (mean = 23.5; SD = 2.9). The controls were ascertained among high school and university students as well as among patients of a general practice outpatient service. The controls had to have no lifetime history of ED or of any other Axis I psychiatric disorder. The controls were matched with the probands for age and socio-cultural level. The age matching occurred within a difference of at most two years since we had to simultaneously consider the socio-cultural level of the family. Five sociocultural levels were defined according to the education and income level of the parents with level one being the lowest and level five the highest.

Method
I Families of controls I Probands
The proband sample reported in the present paper consisted of 68 AN girls with age at onset between 12 and 16 (mean = 14.3; SD = 1.9) and with age 12 to 18 (mean = 14.7; SD = 2.2) at the first psychiatric investigation. The mean BMI at admission in the study was 14.4 (SD = 2.1); this indicates that our sample consisted of severe AN cases. The control women had 198 first degree relatives (105 female relatives and 93 male relatives). The breakdown by relative sex is shown in Table 1. One-hundred twentyone (61.1 %) of the control relatives were directly interviewed. The remaining 38.9 % [77] were investigated through the family history method with information provided by the controls and a family member when available. Twenty-eight relatives out of 77 refused the di-

M. Grigoroiu-Serbanescu et al. Familial aggregation of eating disorders Table 1 Number and age at investigation of first degree relatives of restrictive anorexic probands and controls Probands Age at investigation (Mean, s. d.) 43.8 (7.4) 17.8 (5.1) 46.2 (9.6) 17.5 (6.3) 37.7 (12.5) Controls Age at investigation (Mean, s. d.) 45.7 (8.2) 21.4 (6.3) 48.5 (7.2) 19.8 (8.4) 40.6 (7.3)

I/49

Relatives mothers sisters fathers brothers Total

N 68 30 68 19 185

N 68 37 68 25 198

psychiatric examination intended to exclude subjects with clinical psychiatric disorders. They had to have no lifetime diagnosis of ED and of any other psychiatric disorder on the SCID-I interview in order to be considered normal.We paid special attention to selecting normal controls and not controls with other psychiatric disorders than ED because this may alter the familial psychopathology picture. As mentioned by Schechter et al.[22],many controlled studies find surprisingly high rates of psychopathology in controls.

I Assessment of proband and control relatives

rect interview (11.9 % of the total control relative sample). As the psychopathological investigation of the proband relatives started in 1987, all relatives were diagnosed according to DSM-III-R criteria using SCID-I and SCID-II for the subjects interviewed in person and a checklist containing symptoms and DSM-III-R criteria for those investigated through family history. Control relatives were investigated with the same instruments.

I Clinical assessment of probands and normal controls Diagnostic criteria

Between 1984 and 1987 patients were diagnosed according to the criteria defined by DSM-III [1]. Since 1987 the criteria outlined in DSM-III-R [2] have been applied. Old cases could be pooled together with new cases because the percent of body weight loss of 25 % in DSM-III is more severe than the 15 % weight loss required by DSM-III-R. Additionally, the criterion of amenorrhoea not specifically mentioned by the DSM-III criteria was covered by the diagnostic instrument used by us the Schedule for Affective Disorders and Schizophrenia for School-age Children Epidemiologic Version (K-SADSE) [20]. The AN section of this interview contains one item (item 7) addressing the issue of amenorrhoea. But the severe weight loss required by DSM-III was always associated with amenorrhoea.

I Final diagnosis of probands and relatives

The best estimate procedure based on the review of all available information was used. The team reviewed all the case documents in diagnostic conferences without knowing the identity of the subjects. Names were replaced by figures and the subjects were drawn at random during the review process. The diagnostic procedure was consensual both for probands and family members.

I Statistical analysis
Unadjusted lifetime rates of psychiatric disorders in probands, their relatives and control relatives were computed. The significance of the rate differences between proband and control relatives was estimated with the ztest for proportions. Age-corrected rates were not computed with the survival procedure because the frequency of most disorders was low and age correction would not have brought important differences. The heritability (h2) of the AN was estimated under the multifactorial transmission model according to the procedure described by Falconer [8]; this model assumes that additive genetic and environmental factors determine the disease liability. The lifetime prevalence of AN in the general population was set at 0.5 %. This is an intermediate value in accordance with prevalences reported by epidemiologic studies that found a prevalence of AN ranging between 0.22 % and 0.84 % in the general population [27]. The mean population preva-

Diagnostic instruments
The K-SADS-E interview for DSM-III criteria and since 1987 the version for DSM-III-R criteria of the same interview [18] were administered both to the patient and at least to one of the parents about the patient for investigating Axis I psychopathology by age 17. After age 17, the Structured Clinical Interview for DSM-III-R Patient Version (SCID-I/P) [24] was administered to the patient and often to a parent about the patient. Information provided by the medical records was also available. All patients were hospitalised in the Obregia Psychiatric Hospital of Bucharest at least once. Axis II disorders (personality disorders) were diagnosed with the Structured Clinical Interview for DSMIII-R Personality Disorders (SCID-II) [25]. The control sample was uniformly diagnosed according to DSM-III-R criteria. The controls had to pass the

I/50

European Child & Adolescent Psychiatry, Vol. 12, Supplement 1 (2003) Steinkopff Verlag 2003

lence of 0.5 % was frequently used to compute the heritability of AN [29, 30]. Logistic regression and odds ratio estimated the effect of familial psychopathology loading on comorbidity of AN.

personality, generalised anxiety disorder) in female relatives. One case of bipolar disorder was found in the sample (the sister of a proband). Drug abuse or drug dependence was not found in our sample as drugs were not available in our country by 1994 and the few patients recruited thereafter did not use drugs.

Results
I Prevalence of familial psychopathology
Table 2 shows the unadjusted lifetime rates of all psychiatric disorders found in first degree relatives of AN probands and controls by sex. The lifetime rate of fullblown AN of 1 % in female first degree relatives of probands was not significantly higher than the lifetime rate of AN of 0 % in the control relatives. The lack of significance persisted when considering any ED that also included a case of partial restrictive AN among relatives of probands (z = 1.43, p > 0.05). There was no case of BN among proband and control relatives. Four disorder groups significantly differentiated the proband relatives from control relatives: alcoholism (13.1 %) and any schizo-spectrum disorders (8.3 %) (schizophrenia, schizoaffective disorder depressive type, and schizoid personality) in males and any affective disorders (10.4 %), especially unipolar major depression (8.3 %) and any anxiety disorders (14.6 %) (obsessive-compulsive disorder, obsessive-compulsive
Table 2 Psychiatric disorders in first degree relatives of AN probands and in first degree relatives of controls by sex of the relatives

I Heritability of AN
Since only females were affected with AN the h2 was computed for female relatives. The prevalence of the pure AN was 1.02 % (1/98) in the female first degree relatives of the probands. This gave a correlation in liability between female first degree relatives of 0.09 and an h2 value of 0.18 (S. E. = 0.08). When considering the prevalence of 2.04 % for any ED the heritability value doubled (h2 = 0.36; S. E. = 0.13) in the proband relative sample. This figure supports a moderate familiality of AN in our sample.

I Familial psychopathology and comorbidity in AN probands

The fact that all patients were followed-up for 5 to 18 years from onset gave us the opportunity to follow-up also the development of comorbid disorders during the

Diagnosis

Proband relatives Females (N = 98) Males (N = 87)

Control relatives Females (N = 105) 0.0% (0) 0.0% (0) 0.0% (0) 3.0% (3)a 0.9% (1) 0.0% (0) 4.0% (4)a 0.0% (0) 0.0% (0) 0.0% (0) 0.0% (0) 0.0% (0) 3.8% (4) 0.0% (0) 0.9% (1) 0.9% (1) 5.9% (6)b 0.0% (0) 0.0% (0) 0.0% (0) 0.0% (0) 3.3% (3) 12.4% (13/105) Males (N = 93) 0.0% (0) 0.0% (0) 0.0% (0) 0.0% (0) 0.0% (0) 0.0% (0) 0.0% (0) 0.0% (0) 0.0% (0) 1.1% (1) 1.1% (1)a 0.0% (0) 1.1% (1) 0.0% (0) 0.0% (0) 0.0% (0) 1.1% (1) 0.0% (0) 1.1% (1) 4.4% (4)a 2.2% (2) 2.2% (2) 11.8% (11/93)

AN Partial AN Any eating disorder Unipolar major depression Dysthymia Bipolar disorder Any affective disorder Schizophrenia Schizoaffective disorder depressive type Schizoid personality Any schizo-spectrum disorder Obsessive-compulsive disorder Simple phobia Obsessive-compulsive personality Generalised anxiety disorder Avoidant personality disorder Any anxiety disorder Chronic tic disorder Somatisation disorder Alcoholism Conduct disorders Other personality disorders Total psychopathology rate
ap

1.0% (1) 0.0% (0) 1.0% (1) 0.0% (0) 2.0% (2) 0.0% (0) 8.3% (8)a 2.4% (2) 1.0% (1) 1.0% (1) 1.0% (1) 0.0% (0) 10.4% (10)a 4.6% (4) 0.0% (0) 1.2% (1) 2.0% (2) 1.2% (1) 0.0% (0) 4.8% (4) 2.0% (2) 8.3% (7)a 2.0% (2) 0.0% (0) 4.2% (4) 0.0% (0) 4.2% (4) 2.4% (3) 3.1% (3) 0.0% (0) 1.0% (1) 0.0% (0) 14.6% (14)b 3.4% (3) 0.0% (0) 1.2% (1) 0.0% (0) 0.0% (0) 1.0% (1) 13.1% (11)a 0.0% (0) 3.6% (3) 0.0% (0) 3.9% (4) 39.6% 36.8% (29/98) (32/87)

< 0.05; b p < 0.01

M. Grigoroiu-Serbanescu et al. Familial aggregation of eating disorders

I/51

natural course of the AN in probands. Table 3 exhibits the unadjusted lifetime rates of comorbid disorders and Table 4 displays the most frequent comorbid disorder groups in probands. The comorbidity structure in probands was similar in its variability to the familial psychopathology; unipolar major depression and schizoid, obsessive-compulsive, schizotypal and avoidant personality were the most common comorbid disorders in probands (Table 3). Only one personality diagnosis in each AN patient was made. The personality disorders were subordinated to clusters A and C. Schizoid plus schizotypal personality disorders (28 %) had the highest frequency. Personality disorders pertaining to cluster B were not observed. Substance dependence was completely absent in probands, although in male first degree relatives alcoholism had a high rate compared with the control relatives. When grouping Axis I and Axis II disorders in probands according to common traits (Table 4), three

Table 3 Prevalence of DSM-III-R Axis I and Axis II comorbid disorders in AN patients Axis I disorders Unipolar major depression Schizophrenia Schizoaffective disorderdepressive type Obsessive-compulsive disorder Generalised anxiety disorder Social phobia Somatisation disorders Total Axis I disorders Axis II (personality) disorders Schizoid Schizotypal Avoidant Obsessive-compulsive Passive-aggressive Personality disorder NOS Total Axis II (personality) disorders Rate 11.8% (8) 1.5% (1) 1.5% (1) 4.5% (3) 5.9% (4) 2.9% (2) 3.0% (2) 30.9% (21) Rate 20.6% (14) 7.4% (5) 10.3% (7) 13.2% (9) 2.9% (2) 7.4% (5) 61.8% (42)

classes of disorders appeared to be strongly related to the restrictive AN, namely anxiety disorders (36.7 %; 25/68) (obsessive-compulsive disorder and obsessivecompulsive personality, generalised anxiety disorder, social phobia, avoidant personality), disorders belonging to the schizo-spectrum (30.9 %; 21/68) (schizoid and schizotypal personality, schizophrenia, schizoaffective disorder-depressive type) and unipolar major depression (11.8 %; 8/68). Disorders characterised by emotional instability were not found in probands. Under the hypothesis that AN with Axis I comorbidity could express a stronger genetic diathesis than the AN without comorbidity, we compared the 22 AN patients with Axis I comorbidity with the 46 AN patients without Axis I comorbidity in terms of familial loading. The proportion of affected relatives was 34.5 % (19/55) in the families of probands with Axis I comorbidity versus 21.5 % (28/130) in the families of the probands without Axis I comorbidity (Z = 5.13, p < 0.001). The logistic regression also supported a significant effect of the presence of familial psychopathology on the likelihood to develop a comorbid disorder during the course or after the recovery of AN (Wald statistic = 3.72, df = 1, P < 0.02). The odds ratio of 1.85 (95 % confidence interval) showed a nearly 2-fold increase in the risk for any psychopathology in relatives of AN probands with Axis I comorbidity compared to relatives of probands without Axis I comorbidity. All severe diagnoses (schizophrenia, schizoaffective disorder, bipolar disorder, obsessive-compulsive disorder, and eight out of the ten unipolar major depression cases) were found among the relatives of the anorexic probands with Axis I comorbidity.

Discussion
The data of the present study suggest that restrictive AN might share partial transmission liability with at least three disorder groups: anxiety disorders,unipolar major depression and disorders of schizo-spectrum since these disorders were the most frequent both in first degree relatives and in probands themselves as comorbid disorders. The relationship between restrictive AN and disorders of schizo-spectrum is a new link highlighted by our study as a difference to other recently published family studies of AN. Nilsson Wentz et al. [17] found also an overrepresentation of autism spectrum disorders and cluster C personality disorders among AN patients. Moreover, 31 % of the first degree relatives of the Swedish anorexic probands showed autistic traits [16]. This concept overlaps with the schizo-spectrum disorders the term used by us. Similar to our study, Westen and Harnden-Fischer [37] found that schizoid, schizotypal and obsessive-compulsive personality disorders are the most common in restrictive anorexic patients. In

Table 4 Most frequent groups of comorbid disorders in AN patients Disorder group Any Axis I anxiety disorder (generalised anxiety disorder, social phobia, obsessive-compulsive disorder) Any anxiety disorder (Axis I disorders plus avoidant and obsessive-compulsive personality disorders) Any schizo-spectrum disorder (schizophrenia, schizoaffective disorder, schizotypal personality, schizoid personality) Unipolar major depression Rate 13.2 % (9)

36.7 % (25)

30.9 % (21)

11.8 % (8)

I/52

European Child & Adolescent Psychiatry, Vol. 12, Supplement 1 (2003) Steinkopff Verlag 2003

line with our results is also the report by Hebebrand et al. [12] who found low body weight in male children and adolescents with schizoid personality and Aspergers disorder. Looking at the common features of the co-occurring clinical and personality disorders in probands, it emerges that the liability for restrictive AN could be more generally connected to the diathesis of the emotional restraint and anxiety. Our study also revealed for the first time that the familial psychopathology that could create the propensity to AN is differentiated by the sex of the relative. Female relatives were significantly afflicted with anxiety and depressive disorders and male relatives with schizo-spectrum disorders and alcoholism. The selection of a proband sample consisting of severe cases of restrictive anorexics (according to the mean BMI and hospitalisation), with unitary age of onset [1216] and with a known illness course led to collecting familial data that showed modest familial aggregation of AN in first degree relatives. AN cases were not missed by the investigators because they remained in contact with the probands families for years and other psychopathology was acknowledged by the interviewed subjects. Other recent studies [4, 14], in which the probands phenotype was limited to the purely restrictive form of AN, failed to show significant differences between relatives of AN probands and relatives of normal controls as to AN or ED prevalence. Our data do not lend support to those studies linking AN to addictive mechanisms [19, 33]. Addictive disorders were completely absent among comorbid disorders in probands, though alcoholism was significantly increased in male first degree relatives. Logue et al. [15] and Halmi et al. [10] also failed to find an association between AN and alcoholism. High rates of alcohol and substance abuse/dependence in anorexics reported by some studies [32] seem to be culturally influenced. The high rate of personality disorders (61.8 %) in probands and the relatively various structure of the psychopathology in proband relatives and of the comorbidity in probands point at the connection of AN with a more general vulnerability for psychiatric disorders. Those probands having familial psychopathology also developed comorbid disorders and all severe diagnoses were found in relatives of the anorexic probands with Axis I comorbidity. This finding also raises the question whether AN in adolescence is only the age expression of a genetic predisposition to other psychiatric disorders in the adult life in those anorexic probands with heterotypic familial psychopathology. This hypothesis is strengthened by the conclusion of a survey of 119 outcome studies of AN published in the twentieth century [28]; in a significant proportion of patients the outcome

of AN led to (1) affective disorders; (2) neurotic disorders including mainly unspecified anxiety disorders and phobias; (3) obsessive-compulsive disorder; (4) schizophrenia; (5) histrionic personality disorder; (6) unspecified personality disorders, including borderline states; (7) obsessive-compulsive personality disorder; (8) substance use disorders. (The numbers in parentheses rank the frequency of the disorders in the outcome of AN.) The restrictive and bulimic forms of AN were not separately treated in the survey. Both the presence of comorbid disorders in probands and the presence of a high rate of heterotypic familial psychopathology is in line with the conclusions of recent overviews of molecular genetic research of AN [13] stating that broad and narrow pathophysiological pathways might lead to anorexic behaviour, to which none of the candidate genes proved to be unequivocally related yet.

I Research implications
Beyond the often mentioned connections of AN to depressive and anxiety disorders, our family and comorbidity data highlight links between restrictive AN and the phenotype of schizo-spectrum disorders on the one hand and the sex influence of the heterotypically affected first degree relatives on the vulnerability for AN on the other hand. This means that the subdivision of anorexic probands by comorbidity and familial psychopathology could ensure more homogeneous phenotypes for genetic studies. The relationship of the restrictive AN to the male sex of the heterotypically affected first degree relative (the father) should be explored in the presence of schizospectrum disorders either in the proband or in the family. A paternal effect on autistic traits was recently described for the chromosome 7q [3]. The clinical implications of our study are related to therapy and long-term outcome prediction. The association of AN with different groups of disorders implies differentiated measures. The association with schizoid and schizotypal personality disorders induces high resistance to therapeutic interventions and a slower recovery. In a previous paper [23], we showed that anorexics with schizoid and schizotypal personalities had the poorest outcome measured on the Morgan-Russell Index after seven years from onset. The relatively small dimensions of our proband sample confer limitations to our results but the longitudinal assessment of the probands phenotype over many years and the long-term contact with their families ensure data authenticity. Nevertheless, replication with other samples drawn from other populations would validate our findings.

M. Grigoroiu-Serbanescu et al. Familial aggregation of eating disorders

I/53

References
1. American Psychiatric Association (1980) Diagnostic and Statistical Manual of Mental Disorders (3rd ed). Washington DC: APA 2. American Psychiatric Association (1987) Diagnostic and Statistical Manual of Mental Disorders (3rd Rev ed). Washington DC: APA 3. Ashley-Koch A, Wolpert CM, Zaem L, Menold MM, Basu S, Donnelly SL, Ravan SA (1999) Autistic disorder and chromosome 7: evidence for a paternal effect. Molecular Psychiatry 4(suppl 1):15 4. Bellodi L, Cavallini MC, Bertelli S, Chiapparino D, Riboldi C, Smeraldi E (2001) Morbid risk for obsessive-compulsive spectrum disorders in first-degree relatives of patients with eating disorders. Am J Psychiatry 158:563569 5. Bulik CM, Sullivan PF, Wade TD, Kendler KS (2000) Twin studies of eating disorders: a review. International Journal of Eating Disorders 27:120 6. Cantwell DP, Sturzenberger S, Burrough J, Salkin B, Green JK (1977) Anorexia nervosa: an affective disorder? Archives of General Psychiatry 34:10871093 7. Elbadawy MH, Cliffe MJ, James PT (1985) A monozygotic twin pair discordant for anorexia nervosa. Canadian Journal of Psychiatry 30:544545 8. Falconer DS (1965) The inheritance of liability to certain diseases estimated from the incidence among relatives. Annals of Human Genetics 29:5176 9. Gershon ES, Schreiber JL, Hamovit JR, Dibble ED, Kaye W, Nurnberger JI, Andersen AE, Ebert M (1984) Clinical findings in patients with anorexia nervosa and affective illness in their relatives. American Journal of Psychiatry 141:14191422 10. Halmi KA, Eckert E, Marchi P, Sampurgnaro V,Apple R,Cohen J (1991) Comorbidity of psychiatric diagnoses in anorexia nervosa. Archives of General Psychiatry 48:712718 11. Holland A, Sicotte N, Treasure J (1988) Anorexia nervosa: evidence for genetic basis. Journal of Psychosomatic Research 32:561571 12. Hebebrand J, Henninghausen K, Nau S, Schfer H, Schulz E, Remschmidt H (1997) Low body weight in male children and adolescents with schizoid personality disorder or Aspergers disorder. Acta Psychiatrica Scandinavica 96: 6467 13. Hinney A, Remschmidt H, Hebebrand J (2000) Candidate gene polymorphisms in eating disorders.European Journal of Pharmacology 410:147159 14. Lilenfeld LR, Kaye WH, Greeno KG, Merikangas KR, Plotnicov K, Pollice C (1998) A controlled family study of anorexia nervosa and bulimia nervosa. Archives of General Psychiatry 55: 603610 15. Logue CM, Crowe RR, Bean JA (1989) A family study of anorexia nervosa and bulimia. Comprehensive Psychiatry 30:179188 16. Nilsson Wentz E, Gillberg C, Rastam M (1998) Familial factors in anorexia nervosa: a community-based study. Comprehensive Psychiatry 39:392399 17. Nilsson Wentz E, Gillberg C, Gillberg IC, Rastam M (1999) Ten-year follow-up of adolescent-onset anorexia nervosa: personality disorders. Journal of the American Academy of Child and Adolescent Psychiatry 38:13891395 18. Orvaschel H, Puig-Antich J (1987) Schedule for Affective Disorders and Schizophrenia for School-age Children Epidemiologic Version (K-SADS-E) 4th version. Pittsburg: Western Psychiatric Institute and Clinic 19. Ploog DW, Pirke KM (1987) Psychobiology of anorexia nervosa. Psychological Medicine17:843859 20. Puig-Antich J, Orvaschel H, Tabrizi MA, Chambers WJ (1981) Schedule for Affective Disorders and Schizophrenia for School-age Children Epidemiologic Version (K-SADS-E). 3rd Working Draft. Dept. of Child and Adolescent Psychiatry, New York: New York State Psychiatric Institute 21. Rutherford J, McGuffin P, Katz RJ, Murray RM (1993) Genetic influences on eating attitudes in a normal female twin population. Psychological Medicine 23:425436 22. Schechter D, Singer TM, Kuperman J, Endicott J (1998) Selection of normal control subjects for psychiatric research: update on a model for centralised recruitment. Psychiatric Research 79:175185 23. Serbanescu M, Magureanu S (1993) Anorexie und Persnlichkeitsstrung im Verlauf der Erkrankung (Personality disorder in the course of anorexia). In: Poustka F, Lehmkuhl U (eds) Gefhrdung der kindlichen Entwicklung. Mnchen, Quintessenz Verlag, pp 151158 24. Spitzer RL, Williams JWB, Gibbon M (1986) Structured Clinical Interview for DSM-III-R Patient version (SCIDI/P). Biometrics Research Department, New York, New York State Psychiatric Institute 25. Spitzer RL, Williams JBW, Gibbon M (1987) Structured Clinical Interview for DSM-III-R Personality Disorders (SCID-II). Biometrics Research Department, New York, New York State Psychiatric Institute 26. Steinhausen HCH, Boyadjieva S, Grigoroiu-Serbanescu M, Seidel R, WinklerMetzke C (2000) A transcultural outcome study of adolescent eating disorders. Acta Psychiatrica Scandinavica 101:6066 27. Steinhausen HCh (2002) Anorexia and bulimia nervosa. In: Rutter M, Taylor E (eds) Child and Adolescent Psychiatry. Modern Approaches. 4th ed. Oxford: Blackwell Scientific Publications, pp 555570 28. Steinhausen H-Ch (2002) The outcome of anorexia nervosa in the twentieth century. American Journal of Psychiatry (in press) 29. Strober M, Morrell W, Burroughs J, Salkin B, Jacobs C (1985) A controlled family study of anorexia nervosa. Journal of Psychiatric Research 19:239246 30. Strober M, Lampert C, Morrell W, Burroughs J, Jacobs C (1990) A controlled family study of anorexia nervosa: evidence of familial aggregation and lack of shared transmission with affective disorders. International Journal of Eating Disorders 9:239253 31. Strober M, Freeman R, Lampert C, Diamond J, Kaye W (2000) Controlled family study of anorexia nervosa and bulimia nervosa: evidence of shared liability and transmission of partial syndromes. American Journal of Psychiatry 157:393401 32. Sullivan P, Bulik CM, Fear JL, Pickering A (1998) Outcome of anorexia nervosa: a case-control study. American Journal of Psychiatry 155:939946 33. Szmuckler GI, Tantam D (1984) Anorexia nervosa: starvation dependence. British Journal of Medical Psychology 57:303310 34. Theander S (1985) Outcome and prognosis in anorexia nervosa and bulimia: some results of previous investigations, compared with those of a Swedish longterm study. Journal of Psychiatric Research 19:493508 35. Wade TD, Bulik CM, Neale M, Kendler KS (2000) Anorexia nervosa and major depression: shared genetic and environmental risk factors. American Journal of Psychiatry 157:469471 36. Walters EE, Kendler KS (1995) Anorexia nervosa and anorexic-like syndromes in a population-based female twin sample. American Journal of Psychiatry 152:6471 37. Westen D, Harnden-Fischer J (2001) Personality profiles in eating disorders: rethinking the distinction between Axis I and Axis II. American Journal of Psychiatry 158:547562

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.