Eating behavior and perception of body image in school children
Body dissatisfaction, which is defined as the discrepancy between perceived and ideal body image, can start during childhood (Smolak 2004) and can lead to eating and weight problems in adolescence (Stice 2002; McKnight 2003). Specifically, Stice 2002 reported that when an adolescent presents with body dissatisfaction the probability of engaging to dieting and eating habits that are precursors of eating disorders was increased. Besides that body dissatisfaction was linked to bulimia (Stice 2002). Moreover, McKnight 2003 suggested that school children who presented with thinner preoccupation were in greater risk of developing an eating disorder (McKnight 2003). In addition, body dissatisfaction has been associated with obesity, since obesity is an eating disorder risk factor (Fairburn, Welch et al. 1997; Mustillo, Worthman et al. 2003) and body image is a moderator of the risk of developing eating disorders (Dounchis, Hayden et al. 2001). Except from body dissatisfaction, eating problems in early childhood can lead to eating disorders in adolescence and early adulthood (Kotler, Cohen et al. 2001).
A number of studies have focused on the relationship between eating styles that precede the development of eating disorders, and the perception of body image in various age groups like school children, adolescents and university students and have reported that individuals with high EAT and ChEAT scores were more likely to choose a thinner ideal body image and consequently had body dissatisfaction (Edlund, Halvarsson et al. 1996; Nishizawa, Kida et al. 2003; Gonçalves, Silva et al. 2012; del Mar Bibiloni, Pich et al. 2013; Kutlu and Civi 2013). In more detail, Kutlu and Civi (2013) investigated the association between symptoms of eating disorders and body perception by using the Eating Attitude Test (EAT-40) in university students and found that individuals who perceived themselves obese reported higher scores in the EAT-40 and thus demonstrated a relation between eating disorders and body image perception. Similar results were presented in a study that investigated the association between self-physique and symptoms of anorexia nevrosa in high school individuals in Japan and showed that students with an eating disorder often chose the thinnest ideal body image when compared to the students who did not have an eating disorder (Nishizawa, Kida et al. 2003). Additionally, Edlund et al. 1996 investigated the relation between the ChEAT, DEBQ questionnaires and body image in school girls and observed that girls with high ChEAT scores were more likely to diet and engage to restrained eating and these girls had a higher discrepancy between ideal and perceived body image.
Other studies have focused on eating behaviors like restrained and emotional eating and the relation with perception of body image (Kapka-Skrzypczak 2012; Wiedemann and Saules 2013; Ohara, Kato et al. 2014). Ohara et al. 2014 investigated the association between eating behavior, by using the DEBQ questionnaire, and discrepancy of body image in Japanese university students and suggested that restrained eating was negatively associated with body dissatisfaction in both males and females. In addition, university female students with emotional eating had higher levels of discrepancy of body image. This pattern was not observed in males (Ohara, Kato et al. 2014). Wiedemann and Saules (2013) suggested that the positive association between emotional eating and weight problem perception, another term for body image perception, could be explained by the fact that individuals that perceive themselves as overweight engage to emotional eating because they experience strong emotions like anger, depression or anxiety more often than individuals who are satisfied with their body image. Kapka-Skrzypczak (2012) studied the relationship between dietary habits in adolescents and young adults and reported that students that weren’t on a diet were more satisfied with their body image as compared to students that tried to lose weight.
Another important fact that arises from the existing literature is that the majority of studies are cross-sectional and examine the association between eating disorders symptoms or eating behavior and body dissatisfaction in a specific point in time. This results in a lack of longitudinal data studying the exact pathway that connects eating behavior and the discrepancy between perceived and ideal body image. Moreover, most of the studies that were found during the literature search included adolescents and university students and only two studies investigated this relationship in school children (Edlund, Halvarsson et al. 1996; Gonçalves, Silva et al. 2012). Therefore, the objective of this study is to explore the association between eating behavior of school children (9 year old), reported by mothers, and the discrepancy of body image. For this purpose data from the Generation R study will be used. Additionally, eating behavior will be assessed by the Children’s Eating Behaviour Questionnaire (CEBQ) and the perception of body image by the Children’s Body Image Scale (CBIS).
In conclusion, future studies should focus on younger ages, between 4-9 years old, because currently it is not known whether the association that is demonstrated by the aforementioned studies for adolescents and university students is present at this age group too. Furthermore, longitudinal studies investigating whether eating behavior of preschool children is a predictor of the perception of body image are needed and besides the individual and parental factors, sociocultural factors should be taken into consideration.
Generation R is a population-based prospective cohort study from fetal life to young adulthood (Jaddoe, van Duijn et al. 2012). This study aims to investigate the causes (environmental and genetic) and the pathway that contribute to a normal or abnormal growth and consequently the state of health throughout fetal life, childhood and adulthood. In Generation R there are certain areas of research interest: maternal health, growth and physical development, behavioral and cognitive development, respiratory health and allergies, diseases in childhood, and health and healthcare for children and their parents. Generation R invited all pregnant women living in Rotterdam and were expected to deliver between April 2002 and January 2006. Furthermore, follow-up studies included children born by mothers that participated in the study. Participants had to sign a written informed consent in order to be able to participate in the study. Physical examinations and questionnaires were included in the assessments, where the majority (86%) of the questionnaires were filled out by parents. Generation R has been approved by the Medical Ethical Committee of the Erasmus Medical Center, Rotterdam. More information and details of Generation R can be found elsewhere (Jaddoe, van Duijn et al. 2012).
Children’s Eating Behaviour Questionnaire (CEBQ)
The CEBQ was developed in order to assess the eating behavior of children and study which eating styles lead to obesity and overweight (Wardle, Guthrie et al. 2001). Children’s eating behavior is reported by parents and consists of 8 eating behavior items: food responsiveness, enjoyment of food, emotional overeating, desire to drink, satiety responsiveness, emotional undereating and fussiness. In order to measure these items a 5-point Likert scale is used and ranges from 1 to 5, where 1 denoted “never” and 5 “always” (Wardle, Guthrie et al. 2001).
Children’s Body Image Scale (CBIS)
The CBIS is an instrument that is used to assess the perception of body size in children (Truby and Paxton 2002). CBIS consists of 7 figures which have different versions for males and females. These 7 figures range from the thinnest body shape to the fattest (3rd to 97th NCHS percentiles) (Truby and Paxton 2002). Usually, children have to choose between these 7 figures the one that represents the perceived and the one that denotes the ideal body image.
Child BMI, gender, ethnicity, maternal BMI and education will be included in the regression analysis as potential confounding variables (Gonçalves, Silva et al. 2012; Baillie and Copeland 2013; Bergmeier, Skouteris et al. 2014; Sukariyah and Sidani 2014). It has been demonstrated that female high school students had higher scores on emotional eating and binge eating than males (Sukariyah and Sidani 2014). Additionally, Baile and Copeland (2013) observed that women had higher score on the Body Shape Questionnaire. An different pattern was reported in Goncalves et al. (2012), where boys who were not satisfied with their body image had higher score in the ChEAT questionnaire, a pattern that was not seen in girls who participated in the same study. Moreover, another study in Korean adolescents found that males had higher percentage of body image distortion and thus were in greater risk of developing an eating disorder than females (Hyun, Jung et al. 2014). BMI has been associated with body image dissatisfaction, since individuals with a higher BMI are more likely to be dissatisfied with their body image and engage to restrained eating (Jones and Crawford 2005; O’Haver, Melnyk et al. 2009; Gonçalves, Silva et al. 2012). Furthermore, differences in the magnitude of body image dissatisfaction concerning ethnicity have been suggested by some studies (Gluck and Geliebter 2002; Baillie and Copeland 2013). Glunk and Geliebter (2002) reported that Caucasians and Asians had a higher discrepancy of body image than African Americans and Caucasians scored higher on the eating disorder questionnaire than Asians and African Americans. Baillie and Copeland (2013) observed that Caucasians had a greater body dissatisfaction with their body image than Chinese but no differences were found concerning the scores on the eating disorder questionnaire (EAT-26). Lastly, maternal BMI and education have been associated with food fussiness in children (Bergmeier, Skouteris et al. 2014). Besides that, Goncalves et al. (2012) found that higher maternal BMI was associated with higher scores on the ChEAT questionnaire.
- Is there an association between eating behavior and the discrepancy between perceived and ideal body image?
- In case an association is present, which items from the CEBQ questionnaire are related with the discrepancy between perceived and ideal body image?
This study will not be based on a specific hypothesis. Instead an exploration of the association between eating behavior and discrepancy of body image will be conducted in 9 year old children from the Generation R study.
In this study data were collected from both mothers and their children. Firstly, mothers provided information on their child’s eating behavior by filling out the CEBQ questionnaire, when their children were at the age of 9. Secondly, children at the age of 9 filled out the CBIS for the assessment of the perception of body image. In total data from approximately 4000 children are available at this point.
CBIS provides ordinal data but usually it is treated as interval data (Collins 1991). From the CBIS questionnaire the variable discrepancy of body image will be created. Discrepancy of body image is defined as the difference of perceived-ideal body image.
- Plots to check whether the variables of interest are normally distributed will be done.
- Descriptive statistics will be performed for eating behavior items and body image (ideal, perceived and discrepancy of perceived-ideal) (mean scores and standard deviations). For this purpose statistical tests, parametric (ANOVA, χ2) and non-parametric tests for not normally distributed data will be done.
- Check correlation between confounding variables, eating behavior and body perception.
- Multiple linear regression will be performed where eating behavior (in the form of CEBQ scores) will be the outcome variable and discrepancy of body perception will be the independent/explanatory variable.
- Sensitivity analysis will be performed for the participants that didn’t have any missing values, in order to check whether individuals with missing data are different from individuals without missing data.
- Multiple imputation techniques will be used in order to check for missing data and avoid bias.
Period of thesis/internship
1 April 2015 – 31 August 2015
1-24/04: Conducting literature search and writing research proposal
27/04-01/05: First acquaintance with dataset
04-08/05: Preparation of data analysis
11/05-05/06: Statistical Analysis and writing part of the report
08-12/06: Interpretation of results for the data analysis
15/06-22/07: Writing report
10-17/08: Correcting and improving manuscript
18-21/08: Outline and submit report
24-31/08: Preparation of presentation and presentation
Dounchis, J. Z., H. A. Hayden, et al. (2001). “Obesity, body image, and eating disorders in ethnically diverse children and adolescents.” Body image, eating disorders, and obesity in youth: Assessment, prevention, and treatment: 67-98.
Kotler, L. A., P. Cohen, et al. (2001). “Longitudinal relationships between childhood, adolescent, and adult eating disorders.” Journal of the American Academy of Child & Adolescent Psychiatry 40(12): 1434-1440.
Sukariyah, M. B. and R. A. Sidani (2014). “Prevalence of and Gender Differences in Weight, Body, and Eating Related Perceptions among Lebanese High School Students: Implications for School Counseling.” Procedia-Social and Behavioral Sciences 159: 184-191.
Wiedemann, A. A. and K. K. Saules (2013). “The relationship between emotional eating and weight problem perception is not a function of body mass index or depression.” Eating behaviors 14(4): 525-528.