Objective
The main aim of the present study is to establish the prevalence of eating disorders (ED) in adolescents of both genders. To our knowledge, such data have not previously been published using both DSM-IV and DSM-III-R criteria.
Method
The study sample consisted of 1960 adolescents (1026 girls and 934 boys), 14–15 years of age. The participants completed the Survey for Eating Disorders (SEDs), including DSM-III-R and DSM-IV diagnoses for all subcategories of ED.
Results
Lifetime prevalence of any ED among girls was 17.9% anorexia nervosa (AN) 0.7%, bulimia nervosa (BN) 1.2%, binge eating disorder (BED) 1.5%, and EDs not otherwise specified (EDNOS) 14.6%. Corresponding numbers for boys for any ED is 6.5%, AN 0.2%, BN 0.4%, BED 0.9%, and EDNOS 5.0%.
Discussion
Our prevalence rates on AN, BN, and BED largely support previous school/community-based studies, while our figures on EDNOS were rather high. Generally, we found high numbers for boys with ED.
Eating disorders (EDs) seem to be an increasing problem in the western world Andersen, 1990, O'Dea & Abraham, 2002, Pyle et al., 1991 and Willi & Grossmann, 1983. In a review, McCallum (1993) refers to a prevalence in white adolescent female community samples in Western countries of approximately 1% for anorexia nervosa (AN) and 2–4% for bulimia nervosa (BN). The prevalence of BN in young males were approximately 0.2%. Patton, Selzer, Coffey, Carlin, and Wolfe (1999) have found a prevalence rate of BN 2.7% and AN 0.5% among females aged 14–15 years, and BN 0.3% among boys that age. The prevalence of binge eating disorder (BED) for any age group is not clear, although some nonpatient community-based samples suggest prevalence rates from 0.7 to 10% American Psychiatric Association, 1994, Basdevant et al., 1995, Bruce & Agras, 1992, Drewnowski et al., 1988, Götestam & Agras, 1995, Spitzer et al., 1992 and Spitzer et al., 1993.
Early detection of ED is important and recovery appears to be best for patients treated early in their course (Lock, LeGrange, Agras, & Dare, 2001).
Studies have suggested that approximately 10% of individuals who present with AN and BN and 25% of those presenting BED are men American Psychiatric Association, 1994 and Fairburn & Beglin, 1990. Furthermore, EDs may be increasing among young men in Western society Andersen, 1990, Carlat et al., 1997 and O'Dea & Abraham, 2002. Studies of adolescents have found that although boys report less body dissatisfaction than girls do, significant numbers of boys (5–20%) report restrained eating, vomiting, laxative abuse, or smoking cigarettes for weight control O'Dea & Abraham, 1996, Wertheim et al., 1992 and Worsley et al., 1990.
Studies have found that women and men with EDs suffer similar psychosocial morbidity, as well as course and outcome of the illnesses Eliot & Baker, 2001, Margo, 1987 and Woodside et al., 2001. ED behavior seems to have similar pattern for both sexes, but men and women may develop ED in different ways (Andersen & DiDomenico, 1992).
There are several advantages and disadvantages for self-administered questionnaires versus interviews when screening to identify persons at risk to develop ED. The questionnaire is economical and relatively rapid compared to a “gold standard” clinical interview or interviews by trained staff Black & Wilson, 1996 and Fichter et al., 1998. They may yield more accurate data on sensitive or embarrassing topics because they are more anonymous.
On the other hand, many concepts are difficult to assess accurately with self-administered questionnaires on eating behavior. Questions about “a large amount of food” and “overconcern about weight and shape” may not be clear, with similar concerns about the meaning of “loss of control” and “binge eating.” Items that are even more complex conceptually may be hard to interpret and answer. Items left blank may create problems for analysis and interpretation. These issues are prominent when attempting to classify ED adolescents relying exclusively upon the participant's responses to a self-report questionnaire. Data from interview-based studies generally confirm the findings of questionnaire studies, although with somewhat higher frequencies of ED in the questionnaire studies (Fairburn & Beglin, 1990).
The aim of the present study was to collect questionnaire information on an adolescent sample of both genders, to give DSM-IV- and DSM-III-R-based diagnoses of AN, BN, BED, and EDNOS. To our knowledge, the relative prevalence of all the principal ED categories using DSM-IV and DSM-III-R has not been studied previously in an adolescent population of both genders. We were also interested in the patterns of ED, as well as the relationship between body perception and self-reported weight. Since this study comprised an adolescent sample, methodological and diagnostic issues were given concern.
The life and point prevalence of AN, BN, and BED were consistent with earlier studies in adolescent populations and figures from the American Psychiatric Association (1994), with somewhat elevated figures for boys. Our figures on EDNOS were alarming, and although based on self-reporting, they may indicate an elevated risk potential for ED in 14- to 15 year-olds.
The male–female ratios in lifetime ED (DSM-IV) was for any ED 1:2.8, AN 1:3.5, BN 1:2, BED 1:1.7, and EDNOS 1:2.9. DSM-III-R and DSM-IV gave corresponding numbers on total ED. DSM-III-R gave higher BN numbers than DSM-IV for both genders, and lower figures for EDNOS. An alarming number of young girls and boys expressed feelings of overweight when classed as underweight or normal weight based on BMI classifications.
