Developing a Structural Model of Eating Behaviors Based on Attachment Styles With the Mediating Role of Self-Compassion and Loneliness in Students

A characteristic that affects the quality of life in the general population is eating behaviors. Eating behaviors include restrained eating, eating at night, overeating, strict diets, worrying about body shape and weight, and emotional eating. Factors affecting eating behaviors consist of variables, like attachment styles. Attachment is a biopsychosocial phenomenon that involves social functioning, coping, stress response, psychological wellbeing, health behavior, and morbidity, i.e., why it is so important in psychological research. Research has suggested that studentschr('39') eating behaviors are among the main challenges and could be influenced by some psychological characteristics. The present study aimed to develop a structural model of eating behaviors based on attachment styles concerning self–compassion and loneliness in students.

This was descriptive–correlational research in terms of data collection method and we employed structural equation modeling. The statistical population of this study included all BA and MA students in Tehran City, Iran (age range: 18–40 y) in the fields of humanities, technical engineering, and art in the second semester of the 2019–2020 academic year. The study sample consisted of 522 students; after removing the distorted questionnaires, we analyzed a total of 481 responses. Our sample was randomly selected from the students of Islamic Azad and Government universities. We included Islamic Azad University of Tehran; Science and Research as well as Central Tehran branches. Besides, the included public universities were Allameh Tabatabai, Amirkabir, Science and Technology, and the University of Tehran. A randomized multistage cluster sampling method was employed in this study. Furthermore, the applied research tools included The Dutch Eating Behavior Questionnaire (DEBQ) (van Strien et al., 1986), the Adult Attachment Questionnaire (Simpson, 1990), the Self–Compassion Scale–Short Form (SCS–SF) (Raes et al., 2011), and the Social–Emotion Loneliness Scale for Adults–Short Form (DiTommaso et al., 2003). To analyze the obtained data, SPSS and AMOS as well as structural model methods were used.

The Pearson correlation analysis results suggested a direct and significant correlation between avoidant attachment style and self–compassion, and emotional eating behaviors (p=0.007). There was a direct and significant correlation between self–compassion and external eating (p=0.006). There also existed an inverse relationship between ambivalent and avoidant attachment styles, and loneliness, as well as a significant direct relationship with self–compassion (p=0.001). Besides, there was a direct and significant relationship between self–compassion and emotional eating and external eating (p=0.004). Findings also revealed that secure attachment (p=0.006, β=0.161) and avoidant attachment (p<0.001, β=–0.249) styles, with the mediating role of loneliness, could explain eating behaviors; however, the effect of ambivalent attachment style (p=0.082, β=–0.002) was not significant in this regard. Additionally, the effect of avoidant attachment style on eating behaviors, with the mediating role of self–compassion, was significant (p=0.009, β=–0.177); however, the impacts of secure (p=0.321, β=–0.125) and ambivalent (p=0.224, β=0.143) attachment styles were not significant in this respect. This structural model provided an appropriate fit with the mentioned variables (p≥0.001); Chi–squared=163.325, df=75, X2/df=2.178, RMSEA=0.050, CFI=0.925, NFI=0.911, and GFI=0.957.

Enjoying a safe attachment style and self–compassion, as well as improving insecure attachment styles and loneliness could improve the investigated studentschr('39') eating behaviors. Besides, the obtained data indicated that attachment styles, with the mediating roles of loneliness and compassion, could be a desirable model to explain eating behaviors.