Background
Adolescence is a stage of psychological development during which an individual’s identity and independence from parents becomes established, along with an increased sensitivity to physical developments. Eating disorder is known to commonly occur during late adolescence or early adulthood [
1], and is related to the physical, psychological, and social maturation of adolescents. It is common for patients with eating disorders to have coexisting mental disorders, which include a high likelihood of depression, alcohol dependence, and other anxiety disorders [
2].
Individuals with binge-eating disorder engage in recurrent binge-eating during which they eat an abnormally large amount of food over a short period. Binge-eating disorder is the most common eating disorder. Twelve-month prevalence of binge-eating disorder among U.S. adult (age ≥18 years) females and males is 1.6% and 0.8%, respectively. Life-time prevalence of anorexia nervosa and bulimia nervosa among Koreans is 0.01% and 0.1%, respectively [
3]. However, no reliable studies exist addressing the prevalence of binge-eating disorder in a Korean population. Social and cultural influences are considered one of the major risk factors for binge eating [
4]. Besides daily activity-related stress and social stress, binge eating is considered to be related to psychological issues, especially depression. Previous studies have reported that the majority of people who did binge eat showed a higher level of depression than normal healthy people [
5]. It was also found that binge eating is triggered by events and problems that act as stressors [
6]. Other studies on patients diagnosed with binge-eating disorder reported that they were also diagnosed with various forms of anxiety disorders [
7], and they consumed more food when they experienced anxiety [
8].
The aim of this study was to investigate the relationships between psychological factors (such as depression, anxiety, and stress) and binge eating behaviors in Korean adolescents. Our study participants were a group of female high school students.
Eating behaviors are not only affected by mental and psychological factors such as depression, anxiety, and stress, but can also be affected by external factors, such as social, cultural, and environmental [
9]. Considering this, the participants were selected from among female high school students who attended the same school. Korean high school students from the same school are considered to share a strong emotional bond because they wear identical uniforms, eat the same school-provided lunch, and spend most of their time (from 08:00 to 17:00, 5 days a week) studying at school. By studying these female high school students, we intended to reduce the likelihood of the social, cultural and/or environmental external factors that may affect their eating behavior.
Results
The final 327 participants included in the survey assessment were categorized into the binge-eating and control groups. Overall, the participants’ mean age was 18.0 years, and their mean menarcheal age was 13.4 years. Although the binge-eating group had an older mean age and earlier menarcheal age than the control group, it was not statistically significant (p = 0.051 and p = 0.098, respectively).
The participants’ mean BMI percentile and z-score were 43.34 ± 25.9 and −0.21 ± 0.8, respectively; the binge-eating group had higher BMI than the control group (t = −4.293, df = 325,
p < 0.000). The binge-eating group had a higher PSS than the control group (t = −3.190, df = 325,
p = 0.002). The binge-eating group had higher BDI than the control group (t = −5.300, df = 325,
p < 0.000). The binge-eating group showed statistically significantly higher ASI than control group. (t = −3.821. df = 144.304,
p < 0.000). The binge-eating group had higher TA than the control group (t = −6.303, df = 325,
p < 0.000) (Table
1).
Table 1
Demographic characteristics of study participants
Age (years) | 18.0 | 0.7 | 18.2 | 0.7 | 17.9 | 0.7 | .051 |
Menstrual age (years) | 13.4 | 2.6 | 13.3 | 1.2 | 13.5 | 1.2 | .098 |
BMI (kg/m2) | 20.84 | 2.6 | 21.79 | 2.8 | 20.46 | 2.4 | <.001 |
BMI (percentile) | 43.34 | 25.9 | 52.65 | 25.9 | 39.7 | 24.6 | <.001 |
BMI (z-score) | -0.21 | 0.8 | 0.1 | 0.8 | -0.33 | 0.8 | <.001 |
PSS | 16.13 | 4.5 | 17.38 | 4.9 | 15.63 | 4.2 | .002 |
BDI | 12.89 | 8.2 | 16.58 | 8.4 | 11.44 | 7.7 | <.001 |
ASI | 15.94 | 9.8 | 19.42 | 10.8 | 14.57 | 9.1 | <.001 |
ASI_physical | 6.40 | 5.4 | 7.85 | 5.9 | 5.83 | 5.1 | .004 |
ASI_mental | 4.63 | 4.0 | 6.22 | 4.7 | 4.01 | 3.5 | <.001 |
ASI_social | 5.26 | 2.3 | 5.65 | 2.0 | 5.11 | 2.4 | .041 |
TA | 50.06 | 9.3 | 54.97 | 9.0 | 48.14 | 8.7 | <.001 |
To discover the explanatory factors for binge eating, we studied the correlation between the BITE score (which is a binge eating scale) and stress, depression, trait anxiety, and BMI. The results from a simple correlation analysis showed a quantitative correlation between binge eating and all of the aforementioned factors. Among them, trait anxiety showed the highest correlation with binge eating (r = 0.354,
p < 0.001) (Table
2).
Table 2
Pearson’s correlation coefficient for BITE, symptom scale and BMI
BITE |
PSS | .279a
| | | | | |
BDI | .349a
| .320a
| | | | |
ASI | .303a
| .375a
| .546a
| | | |
TA | .354a
| .371a
| .689a
| .571a
| | |
BMI | .238a
| .063 | .036 | -.020 | -.003 | |
The variables that had a statistically significant correlation from binominal logistic regression analysis were PSS (OR = 1.03, 95% CI = 0.97–1.10), BDI (OR = 1.02, 95% CI = 0.98–1.07), ASI (OR = 1.00, 95% CI = 0.97–1.04), and TA (OR = 1.07, 95% CI = 1.02–1.16). From the results of the binominal logistic regression analysis, using the four variables that were significantly related to binge eating, the variable that was included in the final regression equation was TA. The PSS, BDI, and ASI showed no statistically significant risks (Table
3).
Table 3
Logistic regression analysis demonstrating the contribution of stress, depression, and trait anxiety to binge eating behavior
PSS | .030 | 1.03 | 0.97-1.10 | .362 |
BDI | .022 | 1.02 | 0.98-1.07 | .304 |
ASI | .003 | 1.00 | 0.97-1.04 | .837 |
TA | .066 | 1.07 | 1.02-1.12 | .002 |
Discussion
From this study, 28% (n = 92) of students showed binge-eating behavior with a BITE (Bulimic Investigatory Test Edinburgh) score of >10. The binge-eating group had a significantly higher mean BMI than the control group.
A previous study on the correlation between eating behavior and emotional characteristics reported that depression and self-esteem were the main factors resulting in binge eating, with high levels of depression found in most participants who resorted to binge eating. Another study also reported that binge eating or eating suppression were outward behaviors reflecting participants’ desires to control negative emotions such as depression or anxiety [
28]. Studies have also claimed that eating disorders occur alongside affective disorders, with affective disorders even increasing the risk of developing an eating disorder [
29].
In our study, the higher BITE scores were associated with higher levels of stress, trait anxiety, and depression in the binge-eating group than that in the control. But when the influences were adjusted against each other, the high levels of trait anxiety were found to be associated with the risk of binge eating, but no significant differences were found for stress and depression. These findings are inconsistent with those of previous studies, which showed a high correlation between depression and binge eating [
30]. The mean BDI score of all study participants was 12.89, meaning that most of the participants did not suffer from serious depression [
31]. Although previous studies showed a high correlation between depression and binge eating when evaluated in patients with mood and eating disorders [
32], our study group consisted of a normal population; thus possibly explaining the differing results when compared with previous studies.
A study examining mental disorder coexisting with eating disorders, bulimia nervosa and binge eating showed a higher risk of comorbidity for anxiety disorder, rather than anorexia nervosa [
33]. The comorbidity between anxiety disorder and eating disorder was examined in another study. This investigation revealed that when the participants diagnosed with eating disorders were categorized into an anorexia nervosa group, anorexia nervosa/binge-eating disorder (co-occurring) group, and bulimia nervosa group, the ratio of having one or more co-occurring disorders from anxiety were 55%, 62%, and 68%, respectively [
34]. These results show that if the eating disorder has a higher binge-eating behavior, there is a higher frequency of the onset of anxiety disorder, which is consistent with our finding of a significant difference between onset of binge-eating behavior and increasing anxiety. Contrary to studies indicating that anxiety associates binge eating, many studies have shown evidence that binge eating decreases anxiety. One study suggests that in the presence of an uncontrollable, threatening issue, concentrating on binge eating acts as a process of transferring the negative emotion to the food, which is under one’s control [
35]. A large survey of Korean adolescents reported a high prevalence of interpersonal sensitivity (74.3%), depression (56.9%), and anxiety (48.8%), with girls scoring higher than boys in all three symptoms [
36].
Our study measured levels of binge eating in female high school students using BITE (Bulimic Investigatory Test Edinburgh), and reviewed its relationship with stress, depression and anxiety. We demonstrated that anxiety, among the variables that affect binge eating, plays a more significant role than stress or depression. Trait anxiety is a general and long-standing quality of stress that is characteristic of an individual’s personality. Some people experience a greater appetite than usual when they are anxious, while others experience much less hunger and thirst with anxiety. This difference appears to be associated with the level of trait anxiety. As observed in this study, adolescents with high trait anxiety tend to binge eat. Because food and eating behavior could act as a coping mechanism to stress [
37], high trait anxiety may lead them to craving food when they are suffering from anxiety. We propose public health authorities screen trait anxiety of adolescent students, and provide school or community-based educational programs about healthy eating and proper coping strategies to reduce anxiety. This could diminish the likelihood of adolescents at high risk for developing various kinds of eating disorder from actually doing so. Previous research has already suggested that the primary prevention programs should be offered early enough to prevent the onset of these potentially harmful behaviors [
38]. Therefore, the adolescent binge eating might be prevented by educating schools and communities about the association between trait anxiety and binge eating, and by screening adolescents for mental health including trait anxiety.
Certain study limitations should be noted. First, this study was limited to two girls’ high schools; thus, our results cannot be considered as representative of the eating behavior of the entire female population of Korean high school students. Second, because we used a self-reporting questionnaire, the reliability of the data should be appraised.
Considering the prevalence of eating disorder is much higher in women, a great advantage of this study was that we could investigate eating behavior in a homogenous group of female adolescents, controlling for problematic factors such as sex and age. Culture-specific stressors such as excessive expectations and demands of parents for academic achievement, and peer conflict or rejection might contribute to psychological symptoms in Korean adolescents. Adolescents are more likely to be exposed to a variety of mental disorders due to frustration and increased aggressiveness caused by academic stress. Korean adolescents spent most of their time in school studying. Although sleep is essential for mental and physical health, Korean adolescents do not have enough time even for sleep. Seoul metropolitan government statistics reported that high school students in Seoul only sleep an average of approximately 5.5 h each night, going to bed around 01:00 and waking up at 06:30. This is far shorter than American Academy of Pediatrics’ recommendations (8.5–9.5 h of sleep per night) for adolescents [
39]. Korean adolescents also have little time for leisure or sports activities, which can relieve stress. It has been reported that short sleep durations were associated with increased food intake in European adolescents [
40]. Therefore, future studies should focus on the correlation between study-related stress, sleep deprivation and binge eating, because study-related stress to enter colleges is much higher in Korea than in other countries [
41]. A prospective study for a designated time period might also reveal the cause and effect dynamics between binge eating and emotional factors. An accurate assessment of height, weight, and binge eating using semi-structured interviews by experts would be necessary.
Conclusions
This study examined female adolescents to understand the correlation between the frequency of binge eating, and the emotional characteristics related to binge eating such as stress, depression, and anxiety. From this study, 92 students (28%) from a total 327 female high school students, showed a tendency to binge eat, having a quantitative correlation with BMI, PSS, BDI, ASI, and TA. The binge-eating group showed higher BMI, stress, depression, and anxiety than the control group. However, when the influence of these factors were adjusted against each other, the higher level of trait anxiety was associated with the risk of binge eating, but there was no significant correlation between binge eating, stress, and depression.
Acknowledgements
The authors would like to express the deepest appreciation to 400 students and their parents who participated in this study by providing their valuable information. The authors would like to thank the teachers and staff of two participating high schools. Without their persistent help, this study would not have been possible.