Online ordering is currently unavailable due
to technical issues. We apologise for any
delays responding to customers while we
resolve this. For further updates please visit
our website:
https://www.cambridge.org/news-and-
insights/technical-incident
We use cookies to distinguish you from
other users and to provide you with a
better experience on our websites. Close
this message to accept cookies or find
out how to manage your cookie settings.
Products and Services
Discover Content
Home > Journals > Public Health Nutrition
> Volume 24 Issue 13
> Validation of the Arabic version of the Eating Attitude...
Access
English Français
32
Cited by
Validation of the Arabic
version of the Eating Attitude
Test in Lebanon: a population
study
Published online by Cambridge University Press: 08
September 2020
Chadia Haddad, Chloe Khoury,
Pascale Salameh, Hala Sacre, Rabih Hallit,
Nelly Kheir, Sahar Obeid and
Souheil Hallit
Article
Article contents
Rights &
Save PDF Share Cite
Permissions
Abstract
Objectives:
To validate an Arabic version of the Eating
Attitude Test (EAT-26) and identify factors (such
as depression, stress, anxiety and body
dissatisfaction) that might be associated with
disordered eating among a sample of the
Lebanese population.
Design:
Cross-sectional study.
Setting:
All Lebanese governorates.
Participants:
A total of 811 participants randomly selected
participated in this 5-month study (January–
May 2018).
Results:
The EAT-26 scale items converged over a
solution of six factors that had an eigenvalue
over 1, explaining a total of 60·07 % of the
variance (Cronbach’s α = 0·895). The prevalence
of disordered eating attitudes was 23·8 %.
Higher EAT-26 scores (disordered eating
attitudes) were significantly associated with
higher depression (β = 0·325), higher emotional
eating (β = 0·083), daily weighing (β = 3·430),
higher physical activity (β = 0·05), starving to
reduce weight (β = 4·94) and feeling pressure
from TV/magazine to lose weight (β = 3·95).
Conclusions:
The Arabic version of EAT-26 can be a useful
instrument for screening and assessing
disordered eating attitudes in clinical practice
and research. Some factors seem to be
associated with more disordered eating
attitudes among participants for whom
psychological counseling may be needed. Yet,
our findings are considered preliminary, and
further studies are warranted to confirm them.
Keywords
Eating Attitudes Depression
Emotional eating Arabic Lebanon
Type
Research paper
Information
Public Health Nutrition , Volume 24 , Issue 13 ,
September 2021 , pp. 4132 - 4143
DOI: https://doi.org/10.1017/S1368980020002955
Copyright
© The Author(s), 2020. Published by Cambridge
University Press on behalf of The Nutrition Society
‘Eating Attitude’ is a term used globally to
describe the beliefs, thoughts, feelings and
(1)
behaviours related to food . People develop
positive or negative attitudes leading to actions
consistent with cognitions and emotions
towards food. Diverse approaches to food
might have consequences on general well-
being, and psychological components of eating
attitudes might be important determinants of
health.
Rapid changes in food choices and behaviours
have occurred lately, with people desiring to
(2–5)
stay slim . Unhealthy eating attitudes,
related to abnormal preoccupations with food
and weight, such as continuous concerns about
controlling and losing weight, food restriction
and amplified anger and bad temper when
hungry, had quickly emerged and resulted in
(6,7)
severe eating disorders . Eating disorders,
defined as abnormal or disturbed eating
behaviours, are highly encountered and include
harmful eating attitudes with severe distress or
(8)
concerns about body weight or shape .
According to many studies, the prevalence of
unfavourable eating attitudes varies between
(9,10)
10 and 16 % in America , 3–13·7 % in
(11) (2,5,12,13)
Europe and 4–30 % in Asia . Arab
countries, especially Lebanon, are a!ected by
westernisation and tend to adopt western
traditions in everything, including eating
attitudes. Thus, Lebanese people (mostly
women) favour fasting or restrained eating to
(6)
lose weight .
Psychological factors and emotional status play
very important roles in food choices and
(14)
behaviours . Negative emotions could have
an impact on the behaviour of individuals by
reducing their actions, which usually involves
(14)
avoidance of the stimulus or the activity .
Emotional regulation can a!ect the eating
attitude as it plays an influential role in daily
food intake and may cause what is known as
(15)
‘emotional overeating’ . Furthermore, stress
(often coupled with depression and anxiety
when chronic) and anxiety are also associated
(16,17)
with disordered eating . Indeed, studies
have shown that anxiety appears much earlier
than eating disorders, making it a cause of
(18)
binge eating .
Moreover, anxiety showed some association
with fasting, especially ‘social appearance
anxiety’, where people get anxious about what
society may think of their appearance. Studies
have shown that a higher frequency of use of
social media resulted in high rates of eating
disorders: pressure may lead to body
dissatisfaction, lowered self-esteem and urge to
become extremely skinny, especially in
(6,7,19)
women . Consequently, the frequency of
restrained eating is increasing worldwide;
people stop eating voluntarily to lose weight
(17)
and become skinny , without knowing that
this will increase the food intake after some
time, leading to binge eating again.
Furthermore, studies have demonstrated that
intense physical activity is linked to eating
(20,21)
disorders . An excessive workout is often
used as an approach to cope with emotional
conditions, either by reinforcing mood or by
avoiding destructive thoughts related to not
(22)
exercising . On the other hand, people with
food addiction are less active and prefer
spending their leisure time sitting rather than
(23)
walking or doing any physical activity . Also,
education level has an impact on food
(24–26)
behaviours and attitudes : uneducated
persons tend to consume large amounts of
(26)
junk, high energetic content food , while
more educated people read and analyse health
(24)
facts and choose healthier lifestyles .
In Lebanon, there is a need to assess the risk of
eating-related diseases and eating attitudes
using appropriate tools; one of the most used is
(27)
the Eating Attitude Test (EAT-26) . The EAT-26
was designed to diagnose anorexia nervosa but
has been found useful in non-clinical settings to
screen for disordered eating attitudes,
particularly atypical preoccupation with food
(27)
and weight . This test was validated in Mexico
as a secondary prevention tool to modify public
(28) (29,30)
health programmes , and in Spain ,
where it is known as Disorder Eating Attitude
Scale. It was also validated as a very sensitive
(31)
method in Italy , as a sensitive and specific
(32)
useful tool in non-western countries and as
the most reliable and commonly used self-
report tool for eating disorders worldwide.
Additionally, the EAT-26 was validated in Arabic
among a representative sample of female
(33)
students in Riyadh, Saudi Arabia . However, it
was not validated in the Lebanese dialect and
among the general population; hence, the
importance of its validation to use it as a
screening tool for disordered eating attitudes.
Therefore, the primary objective of the study is
to validate the Arabic version of the EAT-26 and
identify factors (such as depression, stress,
anxiety, emotion regulation and body
dissatisfaction) that might be associated with
disordered eating among a sample of the
Lebanese population.
Methods
Participants
This cross-sectional study conducted between
January and May 2018, enrolled 811 community
dwelling participants using a proportionate
random sample from all Lebanese
governorates (Beirut, Mount Lebanon, North,
South and Bekaa). Each governorate is divided
into Caza (stratum). Two villages were randomly
selected from the list of villages provided by the
Central Agency of Statistics in Lebanon.
Participants were randomly selected from each
village. Prior to participation, individuals were
briefed on the study objectives and
methodology and were assured of the
anonymity of their participation. No financial
reward was given to the participants, and they
had the right to accept or refuse to participate
in the study. Those who agreed to participate
were asked to read and sign a written consent
form.
All participants above 18 years of age were
eligible to participate. Excluded were those who
refused to fill out the questionnaire and those
su!ering from cognitive impairment as
reported by a family member. Data collection
was performed through personal interviews
with participants by a trained study-
independent clinical psychologist, who clinically
evaluated the level of psychiatric illness to
exclude participants with psychiatric problems.
The same methodology was used in previous
(34–36)
papers from this project .
Minimal sample size
calculation
(37)
According to Comrey and Lee , ten
observations are needed for each scale item for
the validation process. Since the EAT includes
twenty-six items, a minimal sample of 260
participants was deemed necessary.
Procedure
Study-independent trained clinical
psychologists collected data by performing
personal interviews with the participants that
required approximately 60 min. Completed
questionnaires were collected back by the
interviewer and sent for data entry. During the
data collection process, the anonymity of the
participants was guaranteed.
The full sample was randomly divided into two
separate samples (406 for the first sample and
405 for the second sample). To ensure the
validity of the results from sample 1, the score
of the EAT scale used in sample 1 was tested on
another sample (sample 2).
Questionnaire
The questionnaire used during the interview
was in Arabic, the native language of Lebanon.
The first part assessed the socio-demographic
details of the participants. The BMI was
calculated as follows: weight (kg) divided by the
square of the height (metres), as self-reported
by each participant. Consumption of alcohol,
ca!einated beverages and tobacco was
categorised into dichotomous variables
(yes/no). The physical activity index is a
frequently used indicator of physical activity at
the population level. This index is based on
responses to a series of questions about the
intensity, frequency and duration of
participation in physical activity during leisure
time. The Total Physical Activity Index was
calculated by multiplying the intensity, duration
(38)
and frequency of daily activity . The
education level was categorised into four
groups: primary, complementary, secondary
and university; however, when stratifying over
education, the four groups were reduced to
two, low level of education (illiterate, primary
and complementary) and high level of
education (secondary, university).
The second part of the questionnaire consisted
of the perception of eating habits among
participants. The questions were selected from
(39–41)
previous articles about eating disorders .
The final part included the following scales.
Eating Attitude Test
The EAT is used to assess disordered food
(42)
attitude . The questionnaire comprises
twenty-six questions each with six response
options, varying from infrequently/almost
never/never (0) to always (3). The total score is
calculated by summing all questions answers
and can vary from 0 to 78. A score of 20 or
above indicates possible disordered food
(43)
attitudes . In the current study, the
Cronbach’s α was 0·908.
Body dissatisfaction subscale of
the Eating Disorder Inventory
second version
In the present study, body dissatisfaction
subscale of Eating Disorder Inventory second
version was used to measure body disturbance.
The body dissatisfaction subscale contains nine
items assessing levels of dissatisfaction with
their overall body shape as well as specific body
parts. A four-point Likert scale was used
ranging from 0 (sometimes, rarely and never) to
3 (always). Higher scores indicate more body
(44)
dissatisfaction . In the current study, the
Cronbach’s α was 0·779.
Perceived Stress Scale
The questions in the Perceived Stress Scale ask
about feelings and thoughts during the last
(45)
month . The Perceived Stress Scale is a ten-
item scale, with answers ranging from never (0)
to almost always (4). Items 4, 5, 7 and 8 are
reversed items. The total score is calculated by
summing the ten items with higher scores
(45)
indicating more perceived stress . In the
current study, the Cronbach’s α was 0·709.
Hamilton Anxiety Rating Scale
The Hamilton Anxiety Rating Scale, validated in
(46)
Lebanon , consists of fourteen symptom-
defined elements and targets both
psychological and somatic symptoms. Each
item is scored on a basic numeric scoring of 0
(not present) to 4 (severe). The total score
ranging from 0 to 56, with higher scores
(47)
indicating higher anxiety . In the current
study, the Cronbach’s α was 0·912.
Hamilton Depression Rating Scale
The Hamilton Scale for Depression, validated in
(48)
Lebanon , is used to evaluate the severity of
depression in patients who are already
diagnosed as depressed. The total score is
based on the sum of the first seventeen items
only. Higher scores indicated higher
(49)
depression . In the current study, the
Cronbach’s α was 0·879.
Emotion Regulation
Questionnaire (ERQ)
The Emotion Regulation Questionnaire is used
to measure respondents’ tendency to regulate
their emotions in two ways: (1) cognitive
reappraisal and (2) expressive suppression. The
cognitive reappraisal facet is a way of
managing and controlling attention and
cognitively changing the meaning of
(50)
emotionally stimulating stimuli . It is
considered a healthy emotion regulation
(50)
strategy . The expressive suppression
involves inhibition of emotional expressive
behavior, thereby changing the emotional
(51)
impact of a situation . It is considered a less
(51)
healthy emotion regulation strategy . A ten-
item scale ranging from 1 (strongly disagree) to
7 (strongly agree). Items 1, 3, 5, 7, 8 and 10
make up the cognitive reappraisal facet and
items 2, 4, 6 and 9 make up the expressive
suppression facet. Each facet’s scoring is kept
separate. The higher the scores, the greater the
(52)
use of the emotion regulation strategy . In
the current study, the Cronbach’s α values for
the cognitive reappraisal facet and the
expressive suppression facet were 0·744 and
0·732, respectively.
The emotional eating, restrained eating,
orthorexia nervosa and binge eating scales
were used to test the convergent validity of the
EAT scale.
Emotional Eating Scale
The Emotional Eating Scale, validated in
(53)
Lebanon is a twenty-five-item scale with
three-factor analytically derived subscales:
anger, anxiety and depression. Participants rate
the extent to which certain feelings lead to the
urge to eat using a five-point Likert scale
ranging from 0 (no desire to eat) to 4 (an
overwhelming urge to eat). Higher scores
indicate a reliance on using food to help
(54)
managing emotions . In the current study,
the Cronbach’s α was 0·957.
Dutch Restrained Eating Scale
Recently, Dutch Restrained Eating Scale,
(55)
validated in Lebanon , is composed of ten
items scored using a five-point Likert scale,
varying from 1 (never) to 5 (always). By dividing
the total items score by the total number of
items, the score for this scale was acquired. A
greater score would show higher degree of
(56)
restrained eating (Cronbach’s α = 0·928).
Orthorexia Nervosa Scale – ORTO-
15
The Orthorexia Nervosa Scale, validated in
(57)
Lebanon , is a measure instrument
comprising fifteen multiple-choice items.
Orthorexia Nervosa Scale is a self-reported
(58)
questionnaire with a four-point Likert scale
(never, sometimes, often and always). Lower
scores would indicate higher levels of
orthorexia tendencies and behaviours. In the
current study, the Cronbach’s α was 0·822.
Binge Eating Scale
The Binge Eating Scale, validated in
(59)
Lebanon , was originally developed to
identify binge eaters within an obese
(60)
population . It does not specify a time frame
and presents a series of di!erently weighted
statements for each item, from which
respondents select the statement that best
describes their attitudes and behaviours. This
yields a continuous measure of binge eating
pathology of 0–46. The severity of binge eating
was divided into three categories as follows: <
17 as non-binging, between 18 and 26 as
moderate binging and ≥ 27 as severe
(61)
binging . The Binge Eating Scale has good
test–retest reliability (r = 0·87, P < 0·001). In the
current study, the Cronbach’s α was 0·862.
Forward and back translation
procedure
Forward translation was first conducted by a
single bilingual translator, whose native
language is Arabic and fluent in English. An
expert committee formed by healthcare
professionals and a language professional
verified the Arabic-translated version. A
backward translation was then performed by a
native English speaker translator, fluent in
Arabic and unfamiliar with the concepts of the
scales. The back-translated English
questionnaire was subsequently compared with
the original English one, by the expert
committee, aiming to discern discrepancies and
to solve any inconsistencies between the two
versions. The process of forward-back
translation was repeated until all ambiguities
disappeared.
Statistical analysis
SPSS software version 25 was used to conduct
data analysis. A descriptive analysis was done
using the counts and percentages for
categorical variables and mean and SD for
continuous measures. We checked the
distribution normality for the EAT scale using
the Shapiro Wilk test. Since the assumption of
normality was not normally distributed, the
comparison of means was performed using the
non-parametric tests (Kruskal–Wallis and
Mann–Whitney tests). Spearman correlation
was used for the linear correlation between
continuous variables. For categorical variables,
2
the χ and Fisher exact tests were used.
A stepwise linear regression was conducted,
taking the EAT score as the dependent variable.
All variables that showed a P < 0·1 in the
bivariate analysis were considered important
variables to be entered in the model in order to
eliminate potentially confounding factors as
much as possible. Afterwards, other stepwise
linear regressions taking the EAT as the
dependent variable but stratified over
education level were conducted (stratification
analysis).
Two di!erent methods were used to confirm
the EAT questionnaire construct validity. First, a
factor analysis was run using the principal
component analysis technique, run on sample
1. Since the extracted factors were found to be
significantly correlated, the promax rotation
technique was used. To ensure the model’s
adequacy, the Kaiser–Meyer–Olkin measure of
sampling adequacy and Bartlett’s test of
sphericity were calculated. Factors with an
eigenvalue higher than one were retained.
Moreover, Cronbach’s α was recorded for
reliability analysis for each scale.
Second, a confirmatory factor analysis was
carried out in sample 2. To assess the structure
of the instrument, the maximum likelihood
method for discrepancy function was used.
Several goodness-of-fit indicators were
2
reported: relative chi-square (χ /df), root mean
square error of approximation, goodness of fit
index and the adjusted goodness of fit index.
The goodness of fit index was calculated by the
2
value of χ /df (cut-o! values <2–5). The root
mean square error of approximation tests the
fit of the model to the covariance matrix. As a
guideline, values of <0·05 indicate a close fit
and values below 0·11 an acceptable fit. The
goodness of fit index and adjusted goodness of
2
fit index are χ -based calculations independent
of df. The recommended thresholds for
(62)
acceptable values are ≥0·90 . A P < 0·05 was
considered significant.
Results
Out of 1000 questionnaires distributed, 811
(81·1 %) were completed and collected. The
