data was collected once a time. The population of
this study was nutrition college students in Indonesia
metropolitan cities. The inclusion criteria were 1)
live in Jabodetabek agglomeration area (Jakarta,
Bogor, Depok, Tangerang, and Bekasi); 2) Not in any
low energy, high energy, vegetarian diet, or
ketogenic diet. This research used purposive
sampling technique. 83 subjects were included.
2.2 Measures
2.2.1 Eating Behavior
Eating behavior was assessed using The Dutch
Eating Behavior Questionnaire (DEBQ) (Strien et
al., 1986). It consisted of three subscales: restrained,
emotional, and external eating behavior. There are
10, 13, and 10 questions for restrained, emotional,
and external eating behavior respectively. The
answer was scored 1-5 as follows: never = 1, seldom
= 2, sometimes = 3, often = 4, very often = 5. Each
subject was cateogorized experiencing restrained,
emotional, and external eating if had score more than
50% of total score for each subscale. The dominant
eating behavior was obtained from the highest score
among three subscales.
2.2.2 Nutritional Status and Body
Composition
Body Mass Index (BMI) was used as indicator of
nutritional status. BMI was calculated by formula:
body weight (kg)/square of body height (m). BMI
was categorized into: underweight (BMI < 18.5),
normal (BMI = 18.5-24.9), overweight (BMI = 25-
26.9), and obese (BMI ≥ 27). Body composition
indicators used in this study were percentage of total
body fat, visceral fat, total subcutaneous fat, trunk
subcutaneous fat, total skeletal muscle, and trunk
muscle. Body composition was measured with
bioelectrical impedance analysis (BIA) Omron
Karada Scan.
2.2.3 Data Analyses
Data was analysed using Statistical Package for
Social Sciences (SPSS) version 22. In descriptive
analysis the data was performed as mean (SD) for
normally distributed data and median (interquartile
range/IQR) for skewed data. Prevalence or
proportion of categorized variable was performed as
n (%). Bivariate analysis between eating behavior
with nutritional status and body composition used
Pearson correlation test for normally distributed data
and Spearman correlation test for skewed data.
Analysis of the differences DEBQ question item
between low-normal BMI and high BMI subjects
used independent t-test. Low-normal subjects were
referred to underweight and normal BMI subjects.
High BMI normal subjects were referred to
overweight dan obese subjects.
3 RESULT
The median age of subjects was 19 years old. Most of
subjects were female (96.4%) and stay with their
family in family house. The majority of subjects lived
in South Jakarta. The overweight and obese
prevalence were 13.2% and 22.9% respectively. Cut
off point for normal total body fat is < 30% for female
and < 25% for male. The mean value of total body fat
showed that there were many female subjects who
had normal body fat. The same result was showed by
visceral fat variable. The mean value of visceral fat
was lower than cut off for high visceral fat. So, many
subjects had normal visceral fat. The mean value of
restrained, emotional, and external eating was not
much different although there were more questions to
assess emotional eating (13) than restrained and
external eating (10) (Table 1).
Table 1: Descriptive characteristics of subjects.
Characteristics
Mean (SD), median
(IQR), or n (%)
A