nueva página del texto (beta)
Inglés (pdf)
Artículo en XML
Referencias del artículo
Traducción automática
Enviar artículo por email
Citado por SciELO 
Similares en
SciELO 
Permalink
Introduction
Disordered eating behaviors (DEB) are conditions that may lead to eating disorders (ED), which are most common during adolescence and early adulthood,1 and due to the morbidity, mortality, and disability associated with them, they qualify as a public health issue.2,3,4,5Although there are as yet no unified criteria in the literature by which to define DEB, they include all the manifestations of ED in similar form but with lesser frequency and intensity.6 DEB can be grouped into three categories: restrictive behaviors, purging behaviors and binge eating.7
Prospective evidence indicates, for example, that voluntary dieting is associated with a higher incidence of ED.8 DEB have also been associated with functional impairments and mental problems.9,10People that have a genuine desire to maintain a healthy weight do not engage in DEB, such as restrictive dieting, excessive exercising or purging behaviors.
Women and adolescents are considered most at risk for ED, although some researchers now include university students,5 adolescents and young men as risk groups.11,12 A study of the young adult population, and particularly of university students, represents a much-needed new direction for the public health agenda. Entering the university brings a series of problems that can trigger a variety of mental health pathologies, including ED; therefore, their detection and timely treatment call for the attention of health care professionals.
Professions and activities that place great value on thin bodies influence the appearance and development of ED.5,13,14Until a few decades ago, ED were considered illnesses of the higher socioeconomic levels.15 A review of the literature16 suggests that between 1979 and 1990, socioeconomic status (SES) did not affect the risk of ED. However, in Mexico, Palma and colleagues17 found that the sample of young people aged 10 to 19 years who participated in the National Survey of Health and Nutrition (Ensanut, by its Spanish acronym) in 2006 showed a greater prevalence of DEB than those with a higher SES.
Overweight and obesity are risk factors for ED, owing to an increased desire for thinness, dissatisfaction with one’s body, and attempts to lose weight.18 Overweight and obese men and women have a greater frequency of DEB than those whose weight is low to normal, because the former tend to adopt a greater number of unhealthy eating strategies in order to lose weight.19
Thompson and Stice20 have found correlations between the internalization of the aesthetic thin ideal (IATI), body dissatisfaction (BD), and ED. Longitudinal studies have shown the presence of IATI and body image disorders as two of the most consistent risk factors for ED.21While most people evaluate themselves based on such aspects of their lives as personal relationships or performance at work or at sports, those with ED focus their self-esteem on their weight and their ability to control it.22 There is also evidence from longitudinal studies that low self-esteem is present prior to the appearance of an ED.1
Because the ideal of masculine beauty is a mesomorphic muscular body, it is necessary to evaluate the drive for that type of body as a precursor to ED in men and the health implications of substance use, dieting, and exercise to attain such a body.23,24
Research on ED and DEB in Mexico has focused mostly on adolescents. According to Ensanut, the prevalence of DEB in subjects aged 10 to 19 years doubled between 200625 and 2012,26 going from 0.9 to 1.9% in women and from 0.4 to 0.8% in men. Among college students, the frequency of DEB has been reported to be between 5.0 and 18.0% in women;27 among men, the figures range from 4.0% in students, in Hidalgo,28 to 13.0%, in Yucatán.29
The purpose of this study was to estimate the prevalence of DEB by sex, and their association with demographic variables (age, area of study, and socioeconomic status), psychological variables (internalization of the aesthetic thin ideal, self-esteem, body dissatisfaction, and drive for muscularity), and body mass index in freshman students at a public university in Mexico City.
Materials and methods
Sample
A representative sample consisting of 892 subjects (502 women and 390 men) was drawn from freshman students of both sexes in all the academic majors offered at the Metropolitan Autonomous University, campus Xochimilco (UAM-X, by its Spanish acronym), in Mexico City. The sample constituted 51.8% of the freshman students enrolled in the last term of 2012. All students in their first trimester were invited to volunteer. Those aged 18 years or more were given an informed consent form, while those aged 17 or less were given an informational letter and consent form for their parents.
The research protocol on which this article is based was reviewed and approved by the Health and Biological Sciences Academic Board of UAM-X.
Instruments
The study had an analytic cross-sectional design. A survey including demographic data (age, sex, and academic major), body mass index (BMI) and physical activity was performed. Assessments of SES, DEB, ITI, self-esteem, drive for muscularity and body dissatisfaction were performed using the AMAI Index30 (Spanish acronym), a Brief Questionnaire for the Measurement of Risky Eating Behaviors31 (CBCAR, Spanish acronym), an Attitudes Towards Appearance Questionnaire,32 a Self-esteem Questionnaire,33a Drive for Muscularity Questionnaire34,35 and an analog scale of body figures,36,37 respectively. See Table I for a detail description of the questionnaires included in the survey and the BMI assessment.38,39
Statistical analysis
Data were analyzed by sex. Two age groups were defined: the first included adolescents, according to the criteria of the 2006 and 2012 Ensanut (≤19 years),26 while the second consisted of those aged 20 years and over. A descriptive analysis of the population yielded simple frequencies and chi-squares, as all the variables were categorical. Given the existence of three categories in the outcome variable DEB -no risk (DEB-NR), moderate risk (DEB-MR), and high risk (DEB-HR)-, a bivariate analysis was carried out using a 2 x 3 contingency table.
The effect of the BMI on the association of DEB with the other variables was measured. Finally, a multinomial logistic regression using all the variables, in which the elimination criteria were based on a plausibility test, was utilized to estimate the association between DEB and its covariates. A value of p<0.05 was considered statistically significant, and confidence intervals (CI) were estimated at 95.0%. Data were processed using the statistical package STATA 11.*
Results
The sample included 892 students (56.2% women and 43.7% men). The average age of the women was 19.3 years (SD=2.4), and that of the men was 20.4 years (SD=3.3) (data not shown in the tables). Table II shows a descriptive analysis by sex. A significant difference was found between the sexes according to age range (p<0.000). The analysis of SES by sex showed a statistically significant difference between women and men. A larger number of women fell into the categories “C” and “D+/D,” which correspond to the lowest socioeconomic levels in the sample. No individuals were found in category “E,” which represents the lowest level of the AMAI Index. Analysis of the distribution between women and men according to area of study found a significant difference (p=0.008), with a greater proportion of women in the biological and health sciences, and of men in design, arts and sciences, and humanities and social sciences.
The analysis of DEB by age, SES and area of study showed no statistical significance for either men or women (table III). The differences in the prevalence of DEB-MR and DEB-HG were statistically significant for both female and male students. In order to determine whether the BMI changed the association of the DEB with the other variables, a logistic regression analysis including the interaction terms was carried out. The BMI showed no effect on these associations; the hypothesis tests showed no significant differences (data not shown in the tables).
Table III Prevalence of disordered eating behaviors by age, SES, and BMI. Mexico City, 2012
Source: Direct survey. SES: Socio-economic status; BMI: Body mass index
A multinomial regression analysis generated two models (table IV). In Model 1, the results show a significant association (p<0.05) among women between DEB-MR and BD (OR=2.8), IATI (OR=6.6), self-esteem (OR=2.3), BMI (OR=1.5), and physical activity >2 h/day (OR=1.4). In men with DEB-MR, there are significant associations with BD (OR=2.0), IATI (OR=4.8), drive for muscularity (OR=4.0), and self-esteem (OR=2.6). Model 2 shows significant associations for women (p<0.05) between DEB-HR and IATI (OR=22.0), self-esteem (OR=3.8), BMI (OR=2.0), and physical activity >2 h/day (OR=1.6). Men with DEB-HR showed significant associations (p<=0.05) only with IATI (OR=17.3), drive for muscularity (OR=5.3), and self-esteem (OR=3.3).
Discussion
The prevalence of DEB-HR found in this sample of Mexico City university students (6.8% in women and 4.1% in men) is similar to that reported among higher education students in Hidalgo (7.9% in women and 4.2% in men).28 Both sets of frequencies are higher than the estimated prevalence of DEB in the national sample of adolescents who participated in the Ensanut in 200625 and 2012.26 The international literature has reported a prevalence of 20.0% in women and 15.0% in men in Spain,40 44% in women and 10.0% in men in Colombia,5 and 9.0% in women and 2.0% in men in Venezuela.12
In contrast to the associations found in adolescents who participated in the 2006 Ensanut,26 the students at the UAM-X showed no significant associations between SES and the prevalence of moderate or high levels of DEB. In a study of female students at a university in southern Brazil, no differences were found in the risk of ED by income level.41 These findings suggest that the variability between socioeconomic levels in samples of university students does not allow for the detection of differences between groups.
The results also showed no differences by sex, age, or area of study for moderate or high levels of DEB. In contrast with reports in the literature, this study shows no evidence that women have a greater risk of exhibiting eating disorders.1,17,42,43The results here confirm that adolescents and young adults share a risk of DEB, in accordance with the findings of other studies.44 The lack of association with the area of study among the men and women with DEB in this sample also contrasts with the results obtained for female university students in the State of Mexico, where those studying administrative and social sciences showed the highest prevalence of DEB.45
The findings in the bivariate analysis with respect to BMI coincide with literature reports, where a greater BMI increases the prevalence of DEB.28 However, the significance of this association in men disappeared in the multivariate regression. Analysis of the frequency of DEB among different categories of BMI found results similar to those in the international literature1 as well as in other studies conducted in Mexico,19 where subjects with greater BMI consistently showed higher frequencies of DEB. The association between DEB and the BMI in men was not, however, constant when controlled for the other variables. The interaction of BMI was tested with the other variables, and the hypothesis tests showed no significant differences.
The multivariate analysis also examined the association between DEB and SES, area of study, and age group. Other associations with increased risk of DEB in women and men were found here: in women, this increase was related to IATI, self-esteem, BMI, and physical activity >2 h/day, while in men, it was associated with IATI, self-esteem and the drive for muscularity.
Although the literature describes the existence of distinct risk factors -such as age and sex- that are associated with the development of DEB and ED, this finding is not borne out by the present study. A differentiation can be observed, however, between the associated variables for men (drive for muscularity) and those for women (BMI and physical activity). It is noteworthy that no significant differences in the prevalence of DEB between women and men were found in this study. The increase in media images of lean and muscular male bodies46 may be generating pressure to achieve this ideal among male university students, motivating them to change their eating behaviors and limit their intake of certain food groups and macronutrients.24 Given this and other findings regarding increased risk of ED in men13 and a lack of interest in diagnosis and treatment,47 the precursor conditions to these disorders require further study, and instruments must be developed for the evaluation of risk factors specific to men.
We recommend carrying out cohort studies in Mexican children and adolescent populations in order to identify ED risk factors and implement prevention strategies.
This study illustrates the limitations of cross-sectional studies, in the sense that it was not possible to differentiate DEB from risk factors. Another limitation is that, because the sample includes only freshman college students, and therefore represents neither the rest of the population of young people nor university students in general, the non-probabilistic sampling results cannot be generalized and must be taken with discretion.
