Introduction
The DSM-IV recognizes the existence of mental disorders related with food consumption--disorders characterized by intense preoccupation with food, weight gain and body image changes.1 These disorders constitute a diagnostic category termed «eating disorders» (ED). This category is subdivided into anorexia nerviosa, bulimia nerviosa and nonspecific ED.
Eating disorders are currently a health problem that is characteristic of developed countries, and their incidence and prevalence are increasing2 to the extent that they have been considered by some to be reaching epidemic proportions. Among adolescents, ED are the third most common chronic illness.3 The ages at which they appear most often (12-25 years) are a further motive for concern, and it has become clear that the age of onset is falling. These disorders appear mainly in girls and women at a ratio 10:1 in comparison to boys and men.
The epidemiological data compiled by Saldaña for several countries indicate that anorexia occurs in 1 out of every 250 young persons between the ages of 12 and 14 years.4 Cervera noted that 1 out of every 100 young women now has anorexia.5 In 10 years the incidence of anorexia and bulimia in Spain have equalled the figures for other European countries, making this the third most common disease among Spanish adolescents.6-8 Toro and colleagues have noted that in Western countries, from 0.2% to 0.8% of the patients have anorexia.9
Children and adolescents are one of the main risk groups for nutritional deficiency, a problem that has worsened in recent years because of cultural, social and demographic changes that have influenced eating patterns in industrialized countries.
This situation, as Moraleda et al10 have noted, requires preventive strategies and an active search for patients with ED in view of the high prevalence and large percentage of adolescents with risk behaviors.
The overall aim of this study was to identify risk factors related with ED in a community of adolescents. Our specific aims were to evaluate the nutritional status and eating habits in a population of students aged 12 to 15 years, to explore family relations and eating habits, and to identify the degree of influence of the mass media on students' habits.
Methods
Design
This descriptive, cross-sectional study was carried out during 2001 in the province of Seville in Southern Spain.
Study population
The sample population consisted of 789 first and second year students in compulsory secondary school (aged 12 to 15 years). The students were recruited at five secondary schools within the Virgen del Rocío and Virgen de Macarena health service areas in Seville. The schools were selected to ensure variety in rural or urban location, socioeconomic level, and health service district:
- Instituto Pablo Neruda (135 students) in Castilleja de la Cuesta. Camas health service district.
- Instituto Beatriz de Suabia (213 students) in Seville (city center). Este-Oriente health service district.
- Instituto Fernando de Herrera (179 students) in Seville. Este-Oriente health service district.
- Instituto Pablo Picasso (119 students), in Barriada de Alcosa, Sevilla. Sur-Guadalquivir health service district.
- Instituto Torre del Rey (143 students) in Pilas. Aljarafe health service district.
Information obtained
The directors and parents' associations of each school were contacted to request authorization for the students to participate in the study.
The questionnaires were handed out to students and collected during class hours.
The validated Spanish version11 of the Eating Attitudes Test (EAT-40) designed by Garner and Garfinkel12 was used to evaluate behaviors and attitudes toward food, weight and exercise. The participant chose one of six possible answers for each of the 40 items in this inventory, and the answers were scored from 0 to 3. The cut-off score was set at 30, for a sensitivity of 67.9% and a specificity of 85.9%.
The CIMEC-26, designed by Toro, Salamero and Martínez,13 is intended to evaluate the influence of agents and situations that transmit prevalent models of body image. A cut-off score of >=23-24 was used; higher scores indicated that the participant was easily influenced by and highly receptive to external agents.
El Family Environment Scale (FES)14,15 evaluates and characterizes the relationships between family members, the most important personal growth factors and basic family structure. This scale consists of 90 items grouped into 10 subscales that measure three dimensions: a) relationships: communication and free expression in the family; b) personal growth (importance of certain processes for the family), and c) system maintenance: family organization and degree of control of some members over others. The highest possible score in each of the 10 subscales is 9.
After the questionnaires were collected, each student was weighed and measured to calculate body mass index (BMI).
The EAT-40 and CIMEC-26 questionnaires were completed by the students, and the FES was given to the students in a sealed envelope for them to take home to their parents. The students later returned the completed questionnaires to their school advisor.
Statistical analysis
All data were entered into a database for treatment with version 10.0 of the SPSS program for Windows. Analyses were done with techniques for descriptive statistics (frequency distribution and mean values with 95% confidence intervals) and the *2 test.
Results
A total of 764 valid questionnaires were received (97%) from 334 adolescent girls and 430 adolescent boys. Mean age was 13.2 years, with a standard deviation of 0.9 years.
The distribution of returned questionnaires by school was Pablo Picaso 13.5%, Pablo Neruda 17.5%, Torre del Rey 15.2%, Beatriz de Suabia 27.6%, and Fernando de Herrera 23.2%.
More than one-third (37.5%) of the families returned useful FES questionnaires. Of the questionnaires that the students turned in, 13% were invalid because of errors in responding to some items. The distribution of family members who completed the FES was: father 6.9%, mother 55.9%, both 25.3%, others 1.9%.
The distribution of the number of siblings in the family was: only child, 3.2%; 1 sibling, 52.5%; 2 siblings, 23.7%; 3 siblings, 7.1%; 4 or more siblings, 3.5%.
The results for nutritional status are summarized in Table 1. Body mass index was <15 (severely underweight, malnutrition) in 0.3% of the students and 15-17 (moderately underweight, slight malnutrition) in 7%.
Parents' observations on students' eating behavior, exercise and leisure time activity were as follows: 42.6% reported that their child studied or worked a lot; 35.1% noted that their child was almost always on his/her feet; 22.5% noted poor appetite; 19.8% indicated that their child was fussy or choosy about food; 19.1% reported that their child complained that servings were too large; 7.7% reported binge eating; 6.9% noted that their child cut food into small pieces; 5.5% indicated that their child was tense during meals; 5.2% noted complaints about body image; 5.3% noted that their child ate alone or in secret; 4.2% reported that their child showed interest in diet or low-fat foods.
With regard to family relations, 32.4% of the parents noted that being successful in life was important to their child; 31.3% indicated that their child compared his or her own achievements against those of other members of the family, and 26% indicated that family members provided little help and support for each other.
Family eating habits were investigated in some detail. To shop for food, a list of needed items was made in 83.5% of the families; food chosen depending on availability in the store in 12%; and each family member chose his or her own foods in 2%.
The type of diet (if any) followed in the family was vegetarian in 2%, slimming diet in 7.3%, and other in 6.6%.
Among the nutritional factors considered by the family, 38.4% did not eliminate fat-rich foods, 52% did not limit sugar-rich foods, 16.5% did not eliminate snack foods, 18% of the families snacked between meals, and 23.5% of the families did not try to balance foods consumed when they ate out, or to balance food intake between lunch and supper.
All members ate together three times a day in 17.2% of the families, twice a day in 56.5%, once a day in 16%, and only on weekends in 5%.
During meals, relations between members led to arguments in 4.6% of the families, news or information being exchanged in 14%, communication between members in 66.4%, giving orders in 1.1%, and watching TV or reading newspapers in 12%.
Only 12% of all students ate a large breakfast, and 43% had only a light breakfast (pastry or toast and milk) (figure 2).
Scores on the questionnaires
Mean score on the EAT-40 was 15.27, with a standard deviation of 9.99. Sixty-eight participants (8.8%) had a score of more than 30 (probable ED or at risk for ED) and 25 (3.3%, 20 girls and 5 boys) satisfied DSM-IV criteria for ED.
Mean score on the CIMEC-26 was 16.55, with a standard deviation of 11.17. Of all participants, 104 (13.5%) were significantly vulnerable to pressure by the mass media, and 85 (11.1%) were considered highly vulnerable.
The scores on the FES did not bring to light any findings of interest (Table 2).
We noted that parents' evaluations of their children's eating behavior were related to and scores on the EAT-40 and CIMEC questionnaires (P=.01).
Students in schools in rural areas (Pilas and Castilleja) or in suburbs (Alcosa) scored higher on the EAT-40 (P=.04) and on the CIMEC (P=.01) than students in schools located in urban areas.
Discussion
The EAT-40 is considered an index of socioeconomic status rather than personality trait,13 and therefore the score cannot be considered diagnostic of ED. However, this instrument is useful to detect certain risk behaviors related with eating disorders.
We found positive scores on the EAT-40 in 3.3% of the participants, a lower figure than in similar studies.16-19
The FES, an indicator of social climate, did not provide data of interest regarding family relationships. The low rate of participation by parents (37.5%) may have been due to the large number of items--some covering similar topics--in the questionnaire, and to the manner in which the questionnaire was delivered and collected.
Eating behavior disorders can be detected early and prevented.20 Evaluation of eating patterns with input from both the subjects and their families is a more effective approach to early detection of the population at risk, and to preventing the health problems that underlie eating disorders.
Thus responsibility for early detection falls to primary care professionals who have diversified their community-level activities to include health education, but who use a variety of nonstandardized methods. Moreover, primary care professionals are unfamiliar with eating problems and their milieu, and their work is more often aimed at disease prevention, vaccination, AIDS, diabetes, etc. Nevertheless, quality criteria should implicitly involve health education activities. Because eating habits are established at an early age, it is important to begin nutritional education before behaviors and attitudes become fixed. Interventions should take advantage of the ties between persons, i.e., their social context and the community-level dimension (family, neighborhood and school). An understanding of the young person's environment can help make health education activities more efficacious and effective.
