Introductıon
Childhood obesity and overweight is considered an important public health problem due to its increasing prevalences in many low and middle income countries (1). Many psychiatric factors such as internalizing and externalizing symptoms are found to be associated with childhood obesity (2).
Alexithymia is defined as inability to find words to describe one’s own feelings (3). As a way of coping with emotions, several studies suggest that the alexithymic patients might develop negative eating behaviors (4)(5)(6). Although there are many studies showing that alexithymic features are more common in obese individuals (7), studies that show no significant relationship between body mass index (BMI) and alexithymic characteristics are also reported (8)(9). Therefore, the role of alexithymia in the development of obesity is not fully understood (7).
Emotion recognition or regulation problems may increase emotional eating which can be defined as increased amount of eating in negative affective situations such as stress. Therefore this mechanism may also cause obesity (10)(11). People with high alexithymic properties are reported to be more prone to relieve their emotional tension by excessive eating (12).
Insulin resistance, known as impaired function of insulin in balancing plasma and hepatic glucose levels (13), is considered as a risk factor for many health problems such as essential hypertension (14) and Type 2 Diabetes Mellitus (15), and is frequently associated with obesity (16). It is known that insulin resistance can cause increased food consumption and disinhibition of eating (17).
Studies examining the relationship between alexithymia and eating behaviour in obese people have been reported mostly in the adult age group. In children and adolescents, the relationship between alexithymia levels and effect of emotional and external factors on the amount of food consumption is less investigated (18). Also, to our knowledge, there are no studies conducted in Turkish population. Understanding the relationship between emotional eating and alexithymia can be a guide for psychosocial interventions in the treatment process of obese youth. We predict that obese adolescents or overweight adolescents (O+OW) have more alexithymic features than normal weight adolescents (N). Therefore, O+OW are more prone to have problematic eating behaviours. Main objective of this study is to investigate the relationship between alexithymia levels and problematic eating behaviours such as emotional, restrained and external eatings in obese and overweight adolescents. Possible predictors of obesity/overweight among these factors are determined. To reveal effect of insuline resistance on eating behaviour, the association between insulin resistance and emotional, restrained and external eating levels is investigated.
Method
The data used in this study is obtained from Zehra Koyuncu’s doctoral thesis (referanslara ekleyebilirsiniz). Cases with subtreshold depression and anxiety symptoms are used in the analysis.
Participants:
A hundred and four adolescents (11-16 years old) with a BMI percentile above 85 were recruited from Department of Pediatrics, Adolescent Outpatient Clinic Unit, Cerrahpaşa Medical School. The reference values determined by Neyzi et al. (19) for Turkish children were used to determine the percentile of the BMI. Patients with BMI above 95 were considered Obese (O) and those between 85 and 95 were assumed Overweight (OW) (20). Adolescents who had antipsychotic or corticosteroid medicine in the last 6 months, who were diagnosed with mental retardation, autism spectrum disorder and schizophrenia spectrum disorder and who were illiterate were not included. Participants with depression/anxiety scores higher than clinical treshold were excluded from the statistical analysis due to the reported relationship between alexithymia and depression/anxiety disorder (21)(22). Seventeen adolescents were excluded because of using antipsychotic drugs, missing data and having RCADS-CV score above clinical treshold (70). Therefore, statistical analysis was conducted with 87 patients.
Comparison group were recruited from middle and high schools in Istanbul. Adolescents with a BMI percentile between 5 and 85 with similar sociodemographic characteristics were enrolled in as the comparison group. The objective of the study and the tests were explained to the adolescents and their parents. Those who accepted were included in the study. One hundred and fifty two adolescents were tested for the purpose of creating the comparison group. 51 members of the tested adolescents were removed because of missing forms (n=25), inappropriate BMI (n=16), being diagnosed with a psychiatric disorder during the study (n=1), having RCADS-CV t score above the clinical treshold (n=9). As a result, the comparison group was formed with 101 adolescents. Permissions were obtained from relevant educational institutions and written consent was obtained from each participants and their parents. The participants were asked to fill in surveys which lasted about 30 minutes. In addition, HOMA-IR value, which is a reliable method to evaluate insulin resistance, was calculated in obese adolescents. The formula for the HOMA-IR value is\(\frac{\text{fasting\ insulin\ concentration\ x\ fasting\ glucose\ concentration}}{22.5}\)(23).
The Declaration of Helsinki was used as the standard of medical ethics in the study design. The Istanbul University-Cerrahpasa, Cerrahpasa Medical School Ethics Committee reviewed and approved all study materials (83045809-604.01.02-).
Instruments:
The Sociodemographic Data Form was created by researchers to record information on age, gender, number of siblings, number of children, presence of chronic diseases, psychiatric disease history, weight and height, family income, family integrity, parental loss, age, educational level and occupation of parents.
Alexithymia Questionnaire for Children (AQC) was developed by Rieffe et al.(24) to measure children’s alexithymia levels. The scale consists of 20 items (24). Koçak et al. (25) adapted this scale to Turkish and conducted a reliability study. Reliability coefficient values was very close to the original scale. The Cronbach alpha internal consistency coefficient of the scale was .78. Higher total score from the scale indicates more alexithymic properties (24).
The Dutch Eating Behavior Questionnaire (DEBQ) was developed in 1986 by Van Strien et al. (26). The DEBQ is used to identify external stimuli and internal stimuli other than fasting, that affect eating behavior. In addition to the ’Restrained Eating’ scale which assesses food intake restriction, a new questionnaire was developed by adding the ’Emotional Eating’ scale and the ’External Eating’ scale which asseses eating in response to food-related stimuli (26). The questionnaire was adapted to Turkish and the validity and reliability study was performed by Bozan (27). The internal consistency coefficient of the whole scale was found to be .94 (27).
Revised Child Anxiety and Depression Scale- Child Version (RCADS-CV) was developed in 2000 by Chorpita et al. (29) based on the Spence Children’s Anxiety Scale (28). The Turkish validity study was carried out by Gormez et al.(30). Cronbach alfa value was calculated as 95 (30). A T score of 70 was reported as clinical treshold (31).
Statistical Analysis
All data were analyzed using SPSS software package 22.0. Mean, standard deviation, median, lowest, highest, frequency and ratio values were used in descriptive statistics of the data. The distribution of variables was measured by the Shapiro Wilk test. In the analysis of quantitative independent data, Student t test and Mann-Whitney test were used. Chi-square test was used for the analysis of qualitative independent data and fischer test was used when the chi-square test conditions were not satisfied. Spearman correlation analysis was used in the correlation analysis. Lastly, potential predictors of obesity and overweight were determined by Multiple Linear Regression Analysis. Statistically, values of p <0.05 were considered significant.
Results
4.1 Comparison of Obese+Overweight with Normal
The mean age of the O+OW was 13.5 ± 1.2 years and the mean age of the control group was 13.6 ± 1.3 year (p> 0.05). When O+OW were compared with N, weight and BMI were significantly higher in O+OW (p< 0,001). There was no significant difference (p> 0.05) between the height, parental education, family income and integrity between N and O+OW (Table 1). AQC scores were significantly higher in O+OW than N (p:0.009). The DEBQrestrained eating and emotional eating subscale scores of O+OW were significantly higher than the scores of N (p<0,001 for both). The DEBQ external eating subscale score did not differ significantly between O+OW and N (p> 0.05). (Table 2).