1. Background
Eating disorders are disorders of eating behaviors, associated thoughts, attitudes and emotions, and their resulting physiological impairments. Anorexia nervosa and bulimia nervosa are thought to proximally be derived from one or both of the following factors: (a) an over valuation of the presumed benefits of weight loss or shape change, usually in the context of overvalued beliefs internalized from sociocultural norms promoting the benefits of thinness or shape change, and (b) fear of fat, or somatovisceral discomforts associated with ingesting food that result in functional, medical, psychological, and social impairment (1).
The prevalence of eating disorders was reported to have increased in the recent decades, especially among young girls (2). Disordered eating is more prevalent among female university students than adolescent girls (3). Numerous studies have shown that abnormal eating behaviors and eating disorders frequently emerge in non-Western countries especially among Asian countries (4-6). The lifetime prevalence of eating disorders among female adolescents in Iran is as much as 0.9% for anorexia nervosa (AN), 3.2% for bulimia nervosa (BN), and 6.6 % for the partial syndrome (7). These rates are suggested to be comparable to prevalence rates reported by studies in Western societies, and somewhat higher than what has been reported in other non-Western societies. University students often claim to experience high level of psychological distress, such as depression, stress and anxiety that can have an adverse effect on their academic performance, emotion and health (8). University students may also eat more than usual when they experience these unpleasant emotional experiences (9). The associations between psychological distress and disordered eating are likely to be bi-directional (10, 11). Therefore, the search for specific psychological variables that may contribute to the pathophysiology of these disorders is of great importance. Two such relevant factors are Alexithymia (AL) and dissociation proneness, both considered as strategies of dealing with negative emotions (10)
Alexithymia has been defined as cognitive deficits in identifying and verbalizing emotions and an inability to distinguish between emotional and physical sensations (12, 13). Body weight attitude, age of menarche and smoking have a role in increasing the incidence of eating disorders (14).
Research suggests that poorly-developed ER competencies and the use of strategies that prolong or magnify negative affects, poses a significant risk for the development and maintenance of mental illness. For example, greater use of emotion suppression, self-blame, rumination and catastrophizing, and less use of cognitive reappraisal and refocusing have been associated with higher levels of depression and anxiety and greater peer problems in adolescents (15-18). There is also some existing research linking difficulties in emotional functioning to eating disorders. For example, anorexia nervosa, bulimia nervosa, and binge eating disorder have been variously related to elevated negative affect, alexithymia, suppressed emotion and poor emotional awareness (19-22).
2. Materials and Methods
The present study had a cross sectional design. Students were chosen by random cluster sampling. For this purpose, selected from among eight faculty Shahid Chamran University 4 faculty. Then, three classes were selected from each faculty. The inclusion criteria were being female and having informed consent and a lack mental illness. Exclusion criteria were lack of the listed inclusion criteria. Overall, 325 female undergraduate students were selected by multi-phase random clustering and participants completed the following self-rating questionnaires; eating attitudes Test-26 (EAT-26), alexithymia questionnaire (TAS-20) and dissociative experiences scale (DES-28) individually. Data analysis was done using the Pearson Correlation and stepwise multiple regression analysis with the SPSS16 software.
2.1. Measures and Procedure
Eating Attitudes Test-26 (EAT-26): the EAT-26 is a 26-item measure that assesses the symptoms and characteristics of eating disorders, and yields an overall score. Questions ask the respondent how often they think and feel they have a disorder associated with food, eating, and their body. Items are rated on a six-point Likert scale, ranging from âAlwaysâ to âNever,â although the three least symptomatic responses (never, rarely and sometimes) are given a value of 0. A Cronbachâs alpha of 0.94 for a sample of females has been reported (23). Validity has previously been demonstrated in the form of correlations between the EAT-26 and similar measures, such as those that assess disordered eating, in both females and males. In addition, the EAT-26 distinguishes between females and males with and without eating disorders (24). The EAT-26 is often used to screen for eating disorders and a score of 20 or higher is thought to constitute a high likelihood of having an eating disorder (25). A study at the University of Iran analyzed the validity and reliability of the eating attitude test. In this study 100 male and female students at Shahid Beheshti University were studied, and the factor analysis confirmed three factors, dieting, bulimia and food preoccupation and oral control, which confirmed and validated the test with Cronbach's alpha of 0.86 (26). In this study the reliability was calculated by Cronbachâs alpha coefficient (Îą = 0.83).
Alexithymia questionnaire (TAS-20): the prevalence of alexithymia was assessed using the Finnish version of the 20-item Toronto alexithymia scale (TAS-20) (27-29). Recent studies on Finnish populations have shown the validity of the TAS-20 in adolescents, although no cut-off scores have been established (30, 31). Each TAS-20 item was rated on a five-point Likert scale, with the total score ranging from 20 to 100. The mean value of each item was calculated and missing values were replaced with the mean if less than two responses for the items were unanswered. The scores of three TAS-20 sub-factors (difficulty identifying feelings (DIF), difficulty describing feelings (DDF), and externally oriented thinking (EOT)) were calculated by using the original factor structure of Bagby et al. (28). The DIF scale refers to subjects' difficulty in differentiating their affects from each other or the affective states from their accompanying bodily sensations. The DDF scale measures the subjectsâ capacity to name, describe and verbalize their feelings, and the EOT scale measures the extent to which subjects relate more to objective events than to psychological processes. We used the highest decile of the total scores and sub scores to indicate a high level of these symptoms (32). Cronbachâs alpha was 0.78 for the TAS-20 total scores, 0.83 for DIF, 0.72 for DDF, and 0.47 for EOT. In the current study, the reliability was obtained by Cronbachâs alpha coefficient, that for the overall sample was Îą= 0.74.
Dissociative experiences scale (DES-28): the dissociative experiences scale (DES) is a self-report instrument that measures dissociative symptoms in both normal and clinical populations (32). Items include experiences of amnesia, gaps in the continuity of awareness, depersonalization, de-realization, absorption, and identity alteration. Examples of DES items include having no memory for important past events in oneâs life (autobiographical amnesia), being in a familiar place and finding it strange and unfamiliar (de-realization), feeling as if oneâs body is not oneâs own (depersonalization), becoming so absorbed in watching television or a movie that one is unaware of what is happening (absorption), and feeling as if one is two different people (identity alteration). Instructions on the cover sheet specify that participants should not include experiences that occurred when they were under the influence of alcohol or drugs. To answer DES questions, one must circle the percentage of times that dissociative experiences have occurred for them (given in 10% increments ranging from 0 to 100). Total scores are calculated by averaging the 28-item scores. These findings indicate good concurrent and criterion related validity (32). In the present study, the validity showed one factor by factor analysis on the Varimax rotation and the reliability was calculated by Cronbachâs alpha coefficient (Îą = 0.91).
3. Results
The participants comprised of 325 students in Shahid Chamran University of Ahvaz, aged between 20 and 24, (mean, 22 years). Table 1 shows scores of Pearson correlation between dissociation experiences and eating attitudes, and alexithymia and eating attitudes.
Table 1.Scores of Pearson Correlation Between Dissociation Experiences and Eating Attitudes, and Alexithymia and Eating Attitudes
aCorrelation is significant at the 0.01 level (1-tailed).
bCorrelation is significant at the 0.05 level (1-tailed).
As observed from Table 1, the result of Pearson correlation showed that the dissociative experiences and eating attitudes have significant positive relationships at P = 37.8 (P <0.01). Also, the correlation between alexithymia and eating attitudes was set at P = 11.3 (P < 0.05).
With regards to Table 2, the multi variable correlational coefficient for the linear compound of alexithymia and dissociative experiences with eating attitudes in female students was respectively MR = 37.0 and MR = 11.3, and F ratio was respectively 36.7 and 2.86, which was significant at P = 0.001 and P = 0.092.
Table 2.Regression Analysis for Alexithymia, and Dissociative Experiences and Eating Attitudes, Using the Enter Method in Students
| Dependent Variable | Independent Variable | Multi- Variable Correlational Coefficient (MR) | Coefficient of Determination (RS) | F | P Value | Regression Coefficients (β) | ||
|---|---|---|---|---|---|---|---|---|
| Î | t | P Value | ||||||
| Eating attitudes | Dissociative experiences | 37.0 | 0.14 | 36.7 | 0.001 | 37.8 | 6.06 | 0.001 |
| Alexithymia | 11.3 | 0.01 | 2.86 | 0.092 | 11.3 | 1.69 | 0.092 | |
Table 3 indicates the result of stepwise regression analysis for alexithymia and dissociative experiences with eating attitudes.
Table 3.Stepwise Regression Analysis for Alexithymia and Dissociative Experiences With Eating Attitudes
| Dependent Variable | Independent Variable | Multi-Variable Correlational Coefficient(MR) | Coefficient of Determination (RS) | F | P Value |
|---|---|---|---|---|---|
| Eating attitudes | Dissociative experiences | 0.37 | 0.14 | 36.7 | < 0.001 |
Regression coefficients (β) showed that dissociative experiences are a good predictor of eating attitudes. Furthermore, coefficient of determination (RS = 0.31) also indicates that 14% of eating attitudes are justifiable by dissociative experiences (or independent variable).
4. Discussion
The aim of the current study was to investigate eating attitudes with regards to dissociation and alexithymia in female students of Ahvaz University. Alexithymia and dissociation can be considered distinctive but overlapping phenomena. According to our findings, dissociative experience in general and alexithymia in particular were strong predictors of eating disorders in university students. As expected, correlational analysis revealed a significant positive relationship between dissociative experience and eating attitudes. Correlations also confirmed the relevance of the relationship between alexithymia and eating attitudes. During the past decades several studies had suggested that alexithymia might constitute one of the psychological features of eating disorders. According to previous studies, the prevalence of alexithymia in general psychiatric clinical samples ranges between 11% and 48% (2). However, the result of our regression analysis pointed to a weak but significant relationship between alexithymia and eating disorders. There are several ways in which alexithymia can affect the clinical outcome of eating disorders: via the negative influence it exerts on the clinical expression of the disorders and on the response to therapeutic interventions (33). It must be acknowledged that the predictive power of alexithymia is limited (it explains a small amount of the variance of the clinical outcome); however, it remains significant even after having taken into account the impact of received treatments and the influence of initial clinical severity and depressive symptoms (33). Alexithymia can worsen the outcome of eating disorders by limiting the compliance to care and the efficacy of therapeutic strategies (33).
The relationship between binge eating and dissociation experiences can be considered from different perspectives. According to the cognitive-behavioral theory of eating disorder maintenance, the over valuation of shape and weight (together with âperfectionismâ) is of primary importance in maintaining eating disorders (34). According to the mood modulation theory, dissociation may coexist with bulimic symptoms as separate phenomena with a common function (decreasing self-awareness or negative emotional states); or according to the escape theory, by narrowing awareness, dissociation may enable patients to initiate binging behavior, without having to deal with the long-term consequences of their actions, such as weight gain, guilt and self-dislike (34).
Our study's findings are similar to those of van Strien et al. (35), who reported positive association between emotional overeating and alexithymia. It is important to note that the results of a study done by Fuller-Tyszkiewicz and Mussap suggested that alexithymia in eating disorders is more related to mood disorder, than to eating behavior. This might justify the limited predictive value of alexithymia in our study. The present findings indicate that dissociation is relevant to a broad range of disordered eating symptoms in a nonclinical sample (36).
Some limitations of our study must be acknowledged. First, the participants were non-clinical undergraduate university students; they might not be representative of patients with eating disorder. Furthermore, the externally oriented thinking subscale in TAS-20 did not have enough internal consistency (37). We only used the TAS-20 to measure alexithymia. This can be considered as a limitation, since this scale does not include day dreaming items, which were removed in 1994 because of low corrected item-total correlations and high correlations with social desirability (28). The lack of reliability of this subscale has been reported in some other studies (37).
The generalizability of the results may be limited by the gender distribution of the sample (entirely females, who were university students).
Beside from these limitations, our study tested the hypothesis that eating attitudes is associated with alexithymia and dissociation experiences by examining a group of females; it provided further evidence in support of the relationship between eating attitudes and dissociation experience, and also found that alexithymia played a role in predicting eating attitudes. Further research could investigate the cause of the weak association between alexithymia and eating attitudes. Furthermore, longitudinal studies are needed to determine the effect of dissociation experiences on development and maintenance of eating attitudes. It is hoped that this research may provide further information and insight into the psychological basis and treatment of eating attitudes.
