Printer Friendly

Risk of eating behavior disorders in adolescents from Cartagena, Colombia/Riesgo de trastorno del comportamiento alimentario en adolescentes de Cartagena, Colombia/Risco de transtorno do comportamento alimentario em adolescentes de Cartagena, Colombia.


In current society, the greatest concern is for the physical appearance; the idea of thinness has been generalized and the risk of eating behavior disorders (REDB) increased. From the perspective of collective and public healthcare REDB is a priority matter of high frequency in adolescent population. (1) In the clinical context, eating behavior disorders gather three main diagnostic categories: anorexia, bulimia, and unspecified eating behavior disorders, according to the American Psychiatric Association and the World Health Organization. (2) However, epidemiological studies frequently identify only individuals in REDB.

Within the global context, in high school students, according to the characteristics of the population and the measurement instrument, the frequency of REDB can be reached by 22%. (3,4) In Colombia, prevalence is observed between 3 and 25% REDB in school-age adolescents (5). To date, studies available in Colombia have been limited to estimating the prevalence of REDB and a reduced number of associated variables and consistently reported that REDB was higher in women and in those manifesting depressive symptoms with clinical importance. (5)

Given that the REDB is a complex phenomenon related to different variables, which interact according to the population group studied, (6) and that currently, it is one of the most relevant mental health problems in adolescents, given the difficulty of early diagnosis and its consequences; thereby, nurses play a fundamental role in timely detection and offering nursing care within the context of an interdisciplinary therapeutic strategy, as well as in evaluating the care needs, diagnoses, and nursing activities that make up the care plan in this type of mental disorder. (7) The approach from the socio-cultural, biological, and psychological perspective, without losing sight of the family will enable this action to be efficient and affective. (7)

This research analyzed a broad number of possible variables associated to REDB, not included in preceding Colombian studies, and controlled eventual bias. (8) This study sought to know the prevalence and variables related to REDB in school-age adolescents from Cartagena, Colombia.


This work is a sub-product of a previous cross-sectional observational analytic study revised and approved by the Ethics Committee at Universidad de Cartagena. The parents authorized their children's participation and the students agreed to participate in the research. A sample was calculated of 1067 high school students, attending public schools in the city of Cartagena, Colombia, during 2012; the frequency expected was 50% for any of the events that could be taken as dependent variable, with 3% margin and 5% alpha error. Additionally, the initial sample was multiplied by 2.5; a design effect to have close confidence intervals in all the estimations of associations and in the fit through logistic regression. The sampling was random through conglomerates; each classroom was taken as one of them.

The students completed the questionnaire, without indicating identification. The form inquired on demographic data (age in years, sex, educational level, and economic level of residence) and the scales presented ahead. Sexual orientation was evaluated with a point that included the definition of the categories: bisexual, heterosexual, homosexual, and unsure of the sexual orientation. For analysis, two groups were re-categorized: heterosexual and non-heterosexual (bisexual, homosexual, and unsure about sexual orientation).

Religiosity was evaluated through the Francis Scale of Attitude toward Christianity short-form, with five items exploring the affective response to God, Jesus, and prayer; each point offers five response options. The scores range from 0 to 20; scores equal to or below 18 correspond to low religiosity. (9) The level of physical activity was quantified through questions adapted from the YOUTH 2007 questionnaire from the United States. Physical inactivity was considered when engaging in less than three days of activities during the last week. (10) Problematic consumption of alcohol was estimated through the CAGE questionnaire; problematic consumption of alcohol was considered with an affirmative response to two or more of the four points of the questionnaire. (11)

To evaluate consumption of any illegal substance at any moment in life, the participants were asked about their use of ecstasy, cannabis, crack, cocaine, and other substances (inhalants, taking sleeping pills or pills for the nerves). Questions were asked about cigarette smoking at any moment in life and during the last month. The presence of depressive symptoms with clinical importance (DSWCI) was quantified through the psychological general wellbeing index (WHO-5), which has five points that quantify DSWCI during the last month. Scores above six were considered DSWCI. (12) Family functioning was quantified with the family APGAR scale, which has five items, exploring the perception of family functioning during the last six months. Scores below 16 were taken with family dysfunction. (13)

The REDB was quantified with the SCOFF questionnaire. This instrument comprises five questions related to eating behavior disorder (EBD) symptoms during the last three months. (14) This questionnaire has been validated in Colombian adolescent population and habitually shows low internal consistency; nevertheless, with good performance against a reference criterion, good sensitivity and specificity to identify possible cases. (15)

The STATA statistical package was used. All the variables studied were dichotomized. According to their ages, the students were divided into age groups over and under 15 years of age. The association between low religiosity and the variables studied was quantified through odds ratio (OR) with 95% confidence intervals (CI95%). Finally, the associations were adjusted for age and sex via binomial logistic regression. The Hosmer-Lemeshow goodness fit was estimated for the final model. The internal consistency of the scales used in this study had the following Cronbach's alpha coefficients: 0.91 for Francis; CAGE: 0.58; WHO-5: 0.75; family APGAR: 0.77; and SCOFF: 0.39.


A total of 2625 students satisfactorily completed the whole questionnaire. Mean age of the participants was 13.8 [+ or -] 2.0 years; 60.4% of the students were between 10 and 15 years of age and 39.6% were between 16 and 20 years of age.

Regarding sex, 54.3% were women. The students were between 6th and 11th grades: 67.7% in basic secondary (between 6th and 9th grades) and 32.3% middle vocational (10th and 11th grades).

In terms of socioeconomic level, 72.8% of the students reported coming from low socioeconomic levels and 27.2% from middle and high socioeconomic levels. According to sexual orientation, 91.9% were classified as heterosexual and 8.1% as non-heterosexual (gay, lesbian, and unsure). In relationship to REDB, a group of 852 students (32.5%; CI95%: 30.7%-34.3%) obtained risk scores for these disorders. Distribution of other variables studied indicated: 32.7% has low religiosity, 27.6% has a high level of physical activity, 15% have smoked at any moment in life and 7.9% has consumed some illegal substance, 8.8% has problematic consumption of alcohol, 7.2% has suffered from depressive symptoms with clinical importance, and almost one of every two (48.0%) has family dysfunction.

The internal consistency of the scales quantifying construct was found between 0.39 and 0.91. Cronbach's alpha coefficients were: 0.91 for Francis; CAGE: 0.58; WHO-5: 0.75; family APGAR: 0.77, and SCOFF: 0.39. With bivariate analysis, we obtained the values for OR of all the variables of interest that can be noted in Table 1. Except for socioeconomic level, religiosity, and level of physical activity, the remaining variables showed association among these risk factors and REDB.

Table 2 shows that in the multivariate analysis, the variables: problematic consumption of alcohol, female sex, consumption of any illegal substance at any moment in life, non-heterosexual sexual orientation, depressive symptoms with clinical importance, cigarette smoking at any moment in life, and family dysfunction were associated to REDB, with OR values between 1.84 and 1.24. The model fit adequately.


This study showed that approximately one third of adolescent students are under REDB. The risk is significantly higher in students with problematic consumption of alcohol and in women. With a similar instrument, REDB prevalence of 32.5% in this research exceeds that documented in other research, which found prevalence that did not surpass 25%. (3,4)

Likewise, this study showed that age was independent of REDB. The few studies available in this population are contradictory: Herpertz-Dahlmann et al., (4) and Unikel et al., (16) observed that REDB increased with age, while Gonzalez-Juarez et al., (17) and Vega et al., (18) found that this risk diminished with age.

In the current research it was noted that women showed a higher proportion of REDB compared to men; said data agrees with most. (4,5,18) We also observed that cigarette smoking at any moment in life was associated to REDB, although few investigations have explored this relationship; Martinez-Mantilla et al., (19) found that daily cigarette smoking in high school students was independent of REDB. Also, Babio et al., (20) demonstrated that this consumption was only associated to REDB in women.

Our results revealed that socio-economic level was independent of REDB, a similar observation to that reported by Angel et al. (5) In terms of the relationship between depressive symptoms with clinical importance and REDB our study showed relationship between these two variables, a finding that coincides with that reported by other authors like Herpertz-Dahlmann et al., (4) and Pineros et al., (21); however, Babio et al., (20) specified that depressive symptoms are related to REDB exclusively in women. Family dysfunction was another of the variables associated significantly to REDB in our study group, a finding that is similar to the research by Gonzalez-Quinonez et al., (22) who reported that family dysfunction increases the of REDB.

Consumption of legal (alcohol and cigarettes) and illegal substances is related to REDB, although no studies were found with adolescent population exploring these associations, research with university students suggests that consumption of substances can be a strategy to lower anxiety and food intake. (23) Regarding sexual orientation, non-heterosexual students showed higher risk of REDB compared to heterosexual students; the aforementioned has been explained a greater concern for the body shape and weight control, particularly in non-heterosexual males. (24)

In our research, religiosity and physical activity did not show statistically significant relationship with REDB. The religious beliefs or convictions can be important factors in REDB in some contexts, above all in that related to self-punishments and sacrifices in some religions. (25) Physical activity can be an indicator of REDB in some populations, where exercise is used as a weight-control strategy by adolescents. (26)

The findings from the current study corroborate the complexity of the problem of eating behavior disorders in adolescent populations, which is why it is necessary for the strategies of promoting healthy eating behavior and prevention and reduction of REDB to be comprehensive, with an interdisciplinary approach and guaranteeing participation from different sectors to encompass all the aspects involving this problem. (3,5)

One of the main intervention strategies is education for individuals, family, and community in which nursing professionals have the necessary training to carry out prevention-type programs, at all levels of care, especially in the first level of care, as an entryway into the healthcare system. (7) From this study, it may be concluded that the prevalence of REDB is high in adolescent students from Cartagena, Colombia. The REDB is significantly higher in those reporting problematic consumption of alcohol, in women, adolescents with history of consuming any illegal substance, non-heterosexuals, those presenting depressive symptoms with clinical importance, in cigarette smoking at any moment of their lives, and those with family dysfunction. Interventions are needed aimed at carrying out comprehensive prevention of REDB in high school students from this Colombian city.


(1.) Van der Wal JS, Gibbons JL, Grazioso MP. The sociocultural model of eating disorder development: Application to a Guatemalan sample. Eat Behav. 2008; 9: 277-84.

(2.) Organizacion Mundial de la Salud. Clasificacion Internacional de las Enfermedades (CIE). Trastornos mentales y del comportamiento. Criterios diagnosticos de investigacion. 10 Edicion. Madrid: Meditor; 1993.

(3.) Rivas T, Bersabe R, Castro S. Prevalencia de los trastornos de la conducta alimentaria en adolescentes de Malaga (Espana). Salud Mental. 2001; 24: 25-31.

(4.) Herpertz-Dahlmann B, Wille N, Holling H, Vloet TD, Ravens-Sieberer U, BELLA study group. Disordered eating behaviour and attitudes, associates psychopathology and health-related quality of life: results of the BELLA study. Eur Child Adolesc Psychiatry. 2008: 17: 82-91.

(5.) Angel LA, Martinez LM, Gomez MT. Prevalencia de trastornos del comportamiento alimentario (T.C.A.) en estudiantes de bachillerato. Rev Fac Med Univ Nac. 2008; 56: 193-210.

(6.) Soh N, Surgenor LJ, Touyz S, Walter G. Eating disorders across two cultures: does the expression of psychological control vary? Aust N Z J Psychiatry. 41: 351-8.

(7.) Perea JM, Espina A, Ortego A. La valoracion enfermera en los Trastornos de la Conducta Alimentaria. Rev Presencia [Internet]. 2007 [cited 2013 Feb 22]; 3(5). Available from: http://www.

(8.) Hernandez-Avila M, Garrido F, Salazar-Martinez E. Sesgos en estudios epidemiologicos. Salud Publica Mex. 2000; 42: 438-46.

(9.) Campo-Arias A, Oviedo HC, Cogollo Z. Internal consistency of a five-item form of the Francis scale of attitude toward Christianity among adolescent students. J Soc Psychol. 2009; 149: 258-62.

(10.) Youth Risk Behavior Survey, 2007. Center for Disease Control. Disponible en (fecha de acceso septiembre 17 de 2009).

(11.) Ewing JA. Detecting alcoholism--The CAGE questionnaire. JAMA 1984; 252: 1905-7.

(12.) World Health Organization. Regional Office for Europe. Well-being measures in primary health care: The DepCare Project. Consensus meeting, Stockholm; 1998.

(13.) Smilkstein G. The family APGAR: a proposal or a family function test and its use by physician. J Fam Pract. 1978; 6: 1231-9.

(14.) Morgan JF, Reid F, Lacey JH. The SCOFF questionnaire: assessment of a new screening tool for eating disorders. Br Med J. 1999; 319:1467-8.

(15.) Rueda GE, Diaz LA, Ortiz DP Pinzon C, Rodriguez J, Cadena LP Validacion del cuestionario SCOFF para el cribado de los trastornos del comportamiento alimentario en adolescentes escolarizados. Aten Primaria. 2005; 35:89-94.

(16.) Unikel C, Saucedo-Molina T, Villatoro J, Fleiz C. Conductas alimentarias de riesgo y distribucion del Indice de Masa Corporal en estudiantes de 13 a 18 anos. Salud Mental. 2002; 25: 49-57.

(17.) Gonzalez-Juarez C, Perez-Perez E, Martin B, Mitja I, Roy I, Vasquez P Deteccion de adolescentes en riesgo de presentar trastorno de la alimentacion. Aten Primaria. 2007; 39: 189-94.

(18.) Vega AT, Rasillo MA, Lozano JE, Rodriguez G, Franco M. Eating disorders. Prevalence and risk profile among secondary school students. Soc Psychiatry Psychiatr Epidemiol. 2005; 40:980-7.

(19.) Martinez-Mantilla JA, Amaya-Naranjo W, Campillo HA, Diaz-Martinez LA, Campo-Arias A. Daily cigarette smoking among Colombian high school students: gender related psychosocial factors. Rev Lat Am Enfermagem. 2008; 16:903-7.

(20.) Babio N, Canals J, Pietrobelli A, Perez S, Arija V. A two-phase population study: relationship between overweight, body composition and risk of eating disorders. Nutr Hosp. 2009; 24: 485-91.

(21.) Pineros S, Molano J, Lopez C. Factores de riesgo de los trastornos de la conducta alimentaria en jovenes escolarizados en Cundinamarca (Colombia). Rev. Colomb. Psiquiatr. 2010; 39:313-28.

(22.) Gonzalez-Quinones JC, de la Hoz-Restrepo F. Relaciones entre comportamientos de riesgo psicosociales y la familia en adolescentes de Suba, Bogota. Rev Salud Publica. 2011; 13:67-78.

(23.) Dunn EC, Neighbors C, Fossos N, Larimer ME. A cross-lagged evaluation of eating disorder symptomatology and substance-use problems. J Stud Alcohol Drugs. 2009; 70:106-16.

(24.) Rosen DS. Eating disorders in adolescent males. Adolesc Med. 2003; 14:677-89.

(25.) Huline-Dickens Sarah. Anorexia nervosa: Some connections with the religious attitude. Br J Med Psychol.2000; 73(Pt 1):67-76.

(26.) Goni A, Rodriguez A. Eating disorders, sport practice and physical self-concept in adolescents. Actas Esp Psychiatr. 2004; 32:29-36.

Zuleima Cogollo [1]

Edna Margarita Gomez-Bustamante [2]

[1] RN, Master. Professor Universidad de Cartagena, Colombia.


[2] RN, Master. Professor Universidad de Cartagena, Colombia.


Article linked to research: Riesgo de trastorno del comportamiento alimentario en adolescentes de Cartagena, Colombia.

Subventions: Vicerrectoria de investigaciones Universidad de Cartagena.

Conflicts of interests: none.

Receipt date: Mar 5, 2013.

Approval date: Aug 20, 2013.
Table 1. Associations for REDB and its risk factors in high school
students from Cartagena, Colombia

Variable                                         OR    OR [CI.sub.95%]

Age in years: 10-15 / 16-20                     1.25      1.06-1.48
Sex: female/male                                1.52      1.29-1.80
Education: basic secondary/medial vocational    1.52      1.27-1.82
Economic level: low/medium-high                 1.20      1.00-1.45
Sexual orientation: non-heterosexual/           1.60      1.21-2.13
Religiosity: low/high                           1.17      0.98-1.39
Level of physical activity: high/low            0.98      0.82-1.18
Cigarette smoking at any moment in life/        1.65      1.32-2.05
  Never experimented
Consumption of any illegal substance at any     1.74      1.31-2.33
  moment in life: yes/no
Problematic consumption of alcohol: yes/no      2.01      1.53-2.63
Depressive symptoms with clinical importance:   1.82      1.35-2.45
Family dysfunction: yes/no                      1.39      1.18-1.64

Table 2. Multivariate model for variables related to REDB in
adolescents from Cartagena, Colombia

Variable                                         OR    OR [CI.sub.95%]

Problematic consumption of alcohol              1.84      1.39-2.45
Female sex                                      1.68      1.42-2.00
Consumption of any illegal substance at any     1.56      1.16-2.11
Non-heterosexual sexual orientation             1.55      1.16-2.08
Depressive symptoms with clinical importance    1.52      1.11-2.07
Cigarette smoking at any moment                 1.49      1.18-1.89
Family dysfunction                              1.28      1.08-1.52

Hosmer-Lemeshow test chi square = 7.754; degrees of freedom = 6;
probability = 0.257
COPYRIGHT 2013 Universidad de Antioquia, Facultad de Enfermeria
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2013 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:Original article/Articulo original/Artigo original
Author:Cogollo, Zuleima; Gomez-Bustamante, Edna Margarita
Publication:Investigacion y Educacion en Enfermeria
Date:Dec 1, 2013
Previous Article:Risk factors of foot ulceration in patients with Diabetes Mellitus type 2/Factores de riesgo de ulceracion de pies en pacientes con Diabetes Mellitus...
Next Article:Using the internet to exchange information and experience on cystic fibrosis/Uso del Internet para el intercambio de informacion y experiencia sobre...

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters