CentralBringing Excellence in Open Access

Journal of Human Nutrition & Food Science

Cite this article: Mousa T, Beretvas SN (2016) Factor Structure of Scores of an Arabic Version of the Eating Attitude Test. J Hum Nutr Food Sci 4(2): 1084.

*Corresponding authorTamara Yousef Suleiman Mousa, Department of Nutritional Sciences, University of Texas at Austin 103 W 24TH ST A2703, Austin, Tx 78712, USA, Tel: 512-905-4002; Email:

Submitted: 14 March 2016

Accepted: 01 May 2016

Published: 03 May 2016

ISSN: 2333-6706

Copyright© 2016 Mousa et al.

OPEN ACCESS

Research Article

Factor Structure of Scores of an Arabic Version of the Eating Attitude TestTamara Mousa1* and Susan Natasha Beretvas2

1Department of Nutritional Sciences, University of Texas at Austin, USA 2Department of Educational Psychology, University of Texas at Austin, USA

Abstract

Objective: Numerous studies reported different factor dimensions for eating attitude test (EAT-26). Therefore, the goal of this study is to explore the factor structure of scores on an Arabic version of EAT-26 in a non-clinical adolescent population.

Method: Four hundred and thirty two adolescent schoolgirls aged 10-16 years in Amman, Jordan were selected from elementary schools, which had a response rate of 75.46%. Eating attitudes were assessed using a translated version of EAT-26. Exploratory factor analysis conducted using principal component analysis extraction followed by an oblique rotation was performed to assess the dimensionality of EAT-26 item scores. Extraction of factors was based on several criteria including the eigen values greater than one rule, the scree test, and theory.

Results: Three factors consisting of 23 items of EAT-26 were extracted. These factors consisted of dieting and awareness of food content, oral control and perceived social pressures, as well as food preoccupation.

Conclusion: The final factor structure of EAT-26 of the current Arabic population was similar to the structure of the English version of the scale. Further investigation is required to validate the present findings using confirmatory factor analysis.

Keywords•Eating attitude test•Factor structure•Jordan

ABBREVIATIONSU.S: United States; EAT-26 (or 40 or 23): Eating attitude test-

26 (or 40 or 23); Cα: Cronbach’s α; BMI: Body Mass Index; BSQ-34: Body Shape Questionnaire-34

INTRODUCTIONEating disorders are a worldwide alarming illness. It is

the third disease affecting young women and adolescent girls in the United States (U.S.) and other developed countries [1] such as Canada [2] Israel [3], and England [4]. In the U.S, the lifetime prevalence of eating disturbances is approximately 6%. Additionally, abnormal eating behaviors are documented in 25% and 11% of adolescent girls and boys, respectively [5]. Yet, eating problems also are reported in non-Western developing countries. For instance, negative eating attitudes have been observed among adolescent girls in Iran (16.7%) [6], Saudi Arabia (19.6%) [7], Oman (43.1%) [8], and the United Arab Emirates (23.4%) [9]. In Jordan, studies addressing eating disorders are scarce, and a few studies assessed negative eating attitudes using Eating attitude test (EAT-26). Musaiger et al., (2013) reported that 66.7% of obese and 35.9% of non-obese girls aged 15-18 years

had abnormal eating attitudes [10]. Another research indicated that 48.2% of 255 Jordanian female college students desired to lose weight [11]. The author showed that 31%, 1.8% and 0.6% of adolescent girls had eating disorders not otherwise specified (now called; other specified feeding or eating disorder), binge eating disorder and bulimia nervosa, respectively [12].

Disturbances in eating attitudes and behaviors are proposed to be related to negative body image perception in females, particularly dissatisfaction with body weight. This suggestion can be explained by that women generally perceive themselves as overweight because of the belief that female beauty resides in being thin [13]. This thin body ideal imposes pressure on women including adolescent schoolgirls, causing them to be prone to weight and body image concerns [14]. These concerns promote young women to engage in weight control behaviors, predisposing problematic eating attitudes [14,15]. For instance, body image dissatisfaction is found to be significantly associated with negative eating attitudes, and increase the risk of developing eating disorders by 5.2 folds (p < 0.001) [12]. It is also reported that 21.2% of 10-16 year old Jordanian schoolgirls were dissatisfied with their body image disturbances [16] and 35.3%

CentralBringing Excellence in Open Access

Mousa et al. (2016)Email:

J Hum Nutr Food Sci 4(2): 1084 (2016) 2/9

had eating disturbances [12]. Approximately, 38.6% of 1887 Iranian teenage girls were not satisfied with their weight status, where two third of them wanted to lose weight [6]. Furthermore, 25% of 293 adolescent girls had negative body image perception, which is believed to promote eating disorders [15]. Thus, dissatisfaction with body image could cause females to control their weight by dieting and other abnormal eating attitudes.

Eating attitude test-26 is used in non-clinical populations to discern eating disturbances. This scale was developed by Garner and colleagues [17], which is a short version of the original 40-item scale (EAT-40). The EAT-26 is a reliable tool for capturing negative eating attitudes, and identifying psycho-pathological features associated with abnormal eating patterns such as binge eating or self-induced vomiting [4,18]. This scale consists of three factors: dieting, bulimia and food preoccupation, and oral control [17].

This measure has been used in various populations such as Malaysia [19], Iran [6], China [20], Greece [21], Spain [22], the U.S. [23], Canada [2], and Australia [24]. The EAT-26 has been translated into several languages including French, Spanish, Urdu, German, Japanese, and Hebrew [18]. This reflects the essential role of this scale in screening for eating disorders. In addition, using this questionnaire in the mother language of the population under investigation is considered to be critical. In particular, examining eating problems in non-English speakers using a translated version of EAT-26 would provide more accurate estimate of such problems. This accuracy is due to the higher ability of the participants to understand the statements/items of this tool, because of accounting for the effect of culture. This highlights the importance of exploring the factor structure of EAT-26 of an Arabic version.

The Arabic culture values plumpness (the traditional female beauty), family relations and mealtime, and conservative dressing style. Over the past two decades however, Arab countries including Jordan, experienced westernization as a result of globalization through mass media that affected diet, health and fashion. Jordan is a low-middle income developing country that witnessed changes at the social and economical levels. These changes influenced the cultural norms and lifestyle of Jordanians, especially those who reside in the most urbanized city, the capital Amman. For example, attending Western fast-food restaurants, and wearing Western clothes became more popular among Jordanians. Accordingly, Jordanians were predisposed to chronic diseases such as obesity, heart diseases, diabetes mellitus, and cancer [25], as well as eating and weight concerns started to surface among Jordanian women. Therefore, the primary aim of this study is to model the structure of the Arabic version of the EAT-26 by applying factor analysis and comparing it with that of the English version. A secondary aim is to examine associations between the resulting factors of EAT-26 with body image dissatisfaction and weight status.

MATERIALS AND METHODSA cross-sectional study was conducted in spring, 2008 to

assess eating attitudes among adolescent schoolgirls in Amman, Jordan. The protocol is illustrated in brief, where its complete

description can be obtained from the two papers cited by Mousa and others (2010) [12,16].

Participants

Only girls were included in this study because eating problems are more prevalent in women. Thus, 432 adolescent schoolgirls aged 10-16 years were recruited from two public and two private elementary schools to ensure representation from all socioeconomic status. Public schools were allocated by the Ministry of Education, and four private schools were contacted in which two of them consent their participation. From each grade, 18 students, which was the number of girls in each class, were randomly recruited from class lists for participation. With the help of the staff of schools, all parents of participants gave a written informed consent. Information regarding demographic data, eating and body image disturbances was gathered using a set of questionnaires. Data were collected and reported in a confidential manner. Ethics of the protocol of the current study was approved by the Committee of Higher Studies and Scientific Research of the University of Jordan.

Validity and Reliability

The EAT-26 [17] and body shape questionnaire (BSQ-34) [26] were translated by a professor of psychology. Then, a panel of academics of nutrition, psychology and Arabic literature, as well as a panel of Examinee of the Ministry of Education in Jordan assessed and revised the content validity of the translated scales to ensure an accurate translation of the items without altering them from their original form [12,16]. In addition, a pilot study was conducted to standardize and measure the reliability of the Arabic version of the EAT-26 and BSQ-34. Two groups, each consisted of 15 adolescent girls aged 10-16 years were randomly allocated; one from a private school and the other from a public school. The response rate was 80% in which 24 students participated in the pilot study, who were not included in the original research. The scales had a high internal consistency [12,16].

Procedure

A sample of 326 adolescent girls agreed to participate in the study. The researcher explained to participants the purpose of the study, and provided them with an information sheet describing the general background of the research. Consequently, participants completed the questionnaires under the supervision of the researcher. Height and weight of schoolgirls were measured by the researcher using a beam balance scale as described by [27]. Body mass index (BMI) of participants was estimated, and BMI for age percentiles were used to assess weight status [28].

Tools of Assessment

1) Demographic data of participants such as birth date, menarcheal status, and grade level were obtained through a constructed questionnaire [12,16].

2) Negative eating attitudes were evaluated using EAT-26, which consists of 26 items. Each item has a six-point likert scale response, ranging from zero (never) to 3 (always). This score range reflects the frequency of engaging in weight control behaviors that indicate having negative eating attitudes. Total

CentralBringing Excellence in Open Access

Mousa et al. (2016)Email:

J Hum Nutr Food Sci 4(2): 1084 (2016) 3/9

score is the sum of the composite items, ranging from zero to 78. Eating attitudes considered abnormal at a score of 20 or above [17]. The three factors of EAT-26 that Garner et al. reported are: (1) dieting, which reflects restricting intake of high caloric foods and preoccupation with body image/shape. This factor consisted of 13 items including “eat diet foods and avoid foods with sugar in them”; (2) bulimia and food preoccupation describes thoughts regarding food, binging and self-induced vomiting. This dimension included six items such as “have gone on eating binges where I feel I may not be able to stop”, as well as “feel that food controls my life”; and (3) oral control that illustrates the ability to regulate food intake and perceived pressure from others to gain weight. This factor is composed of seven items such as “cut my food into small pieces”, and “feel that others pressure me to eat” [17].

3) Body image dissatisfaction was discerned using BSQ-34. This valid tool consists of 34 six-point likert scale items, such as “Have you been so worried about your shape that you have been feeling that you ought to diet?” The score ranges from one (never) to six (always), which corresponds to the frequency of perceiving body image negatively (i.e.; never to always). A total score of 110 or above reflects dissatisfaction with body image [26].

Statistical Analysis

Statistical analyses were performed utilizing the Graduate Pack SPSS 19.0 for windows 2010. Principal component analysis was used as the extraction method. Exploratory factor analysis followed by oblique rotation was performed to assess the dimensionality of EAT-26 item scores. Selection of the number of factors to extract was based on several criteria including the eigen values greater than one, the scree test, and theory. Based on these criteria, three-, four- and five-factors models were investigated. Factor loadings greater in magnitude than 0.35, were considered substantial. We used a value greater than 0.35, despite that most researchers used a factor loading value equal or greater than 0.4, because it can be rounded to 0.4. Further, Talwar et al., [19] as well as Lee and Lee [20] used factor loadings greater in magnitude than 0.3 and 0.35, respectively. Hence, a factor loading above 0.35 can be considered a sufficiently substantial value. The degree of internal consistency of EAT-26 scale and its factors was tested using Cα. Correlations and linear regression models were performed to examine associations between factors of EAT-26 and with other variables including BMI, and body image dissatisfaction. Multivariate regression models were used to control for confounders. All p-value ≤ 0.05 were considered significant.

RESULTSThe population sample was of low-middle income status,

and adolescent girls had a mean age of 12.9 ± 1.8 years. Mean BMI of the participants was 20.9 ± 3.8 kg/m2 indicating that the participants had a normal body weight. More than half of participants had menarche with a mean menarcheal age of 12.4 ± 1.1 years. Mean EAT-26 score was 16.52 ± 10.4, where 40.5% of the participants had negative eating attitudes (EAT-26 ≥ 20) [12].

The number of factors to extract was evaluated based on several criteria including: eigen values> one criterion, a scree

test, and theory. The scree test resulted in three elbows forming either seven or five or three factors. Eigen values supported extraction of seven factors. Theory suggested extracting three or four or five factors because they are the most common structures of a scale. Extraction of four, five and seven factors was not considered due to having at least one factor with either one or two items loading on this factor. It is also known that the ability of a factor in capturing a trait would be more sensitive if the factor consisted of at least three items. Therefore it is concluded that a three-factor structure is the most appropriate model for the current version EAT-26.

Three-factor analysis resulted in: 14 items loading on the first factor; dieting and awareness of food content that included items describing weight concerns and engagement in weight control attitudes such as “avoid foods with sugar in them” and “think about burning up calories when I exercise”; six items loading on the second factor; oral control and perceived social pressures, which consisted of items that assessed controlling dietary intake and perceiving social pressures regarding body weight and eating, as “feel that others pressure me to eat” and “cut my food into small pieces”; and four items loading on the third factor; food preoccupation that had items such as “find myself preoccupied with food” and “give too much time and though to food” that illustrated thinking about food and eating.

Two items; 15 (i.e.; Take longer than others to eat my meals) and 26 (i.e.; Have the impulse to vomit after meals), were excluded because their factor loadings were less in magnitude than 0.35, and factor analysis was rerun. This exclusion did not affect the pattern of loading of the remaining items on the three factors, which was checked by performing a new exploratory factor analysis. Moreover, item 10 (i.e.; Feel extremely guilty after eating) cross-loaded on factors one and three, thus it was excluded. Accordingly, another factor analysis was undertaken, resulting in 13 items loading on the first factor, six on the second and four on the third. It is noteworthy that some items had negative loadings, yet they were below the acceptable value of 0.35. A negative loading means that the item (e.g.; items 8, 13, and 20) possesses an opposite characteristic of the factor on which they do not load (e.g.; factor 1). Factors, factor loadings, percentage of variance, and reliability values of EAT-26 items are presented in Table 1.

Approximately, 86.2% of the participants frequently dieted and were aware of food content (factor 1), 77.9% more often engaged in oral control and perceived social pressures (factor 2), and 39.6% regularly exhibited food preoccupation (factor 3). The component correlation matrix of EAT-23 is shown in Table 2. The first factor has a low positive correlation with the second (r = 0.09) and third factors (r = 0.10), whereas the last two factors are negatively correlated (r = - 0.05). The weak correlations indicate that each factor measures a specific attitude (dieting and awareness of food content) that is distinctive from the other factor (oral control and perceived social pressures, or food preoccupation).

Correlations between EAT-23 factors, age, BMI and body image dissatisfaction are presented in Table 3. Factors 1 and 3 showed positive correlations with both BMI and body image dissatisfaction (p < 0.01). However, factor 2 (oral control and

CentralBringing Excellence in Open Access

Mousa et al. (2016)Email:

J Hum Nutr Food Sci 4(2): 1084 (2016) 4/9

Table 1: Factor Structure of EAT-23.

Item # Item Reliability (Cα) Variance (%)Factor loadings*

Factor 1 Factor 2 Factor 3Factor 1: Dieting and awareness of food content 0.85 22.31

11 Am preoccupied with a desire to be thinner .716 -.048 .149

12 Think about burning up calories when I exercise .678 -.037 .114

1 Am terrified about being overweight .643 .026 .240

23 Engage in dieting behavior .637 -.197 -.033

17 Eat diet foods .630 -.022 .091

7 Particularly avoid foods with high carbohydrate content (i.e. bread, rice, potatoes, etc.) .610 .008 -.141

24 Like my stomach to be empty .597 .005 .001

14 Am preoccupied with the thought of having fat on my body .574 .123 .305

2 Avoid eating when I am hungry .557 .236 -.105

16 Avoid foods with sugar in them .546 .074 -.067

22 Feel uncomfortable after eating sweets .503 -.091 .104

19 Display self-control around food .464 .055 -.215

6 Aware of the calorie content of foods that I eat .370 -.021 -.025

Factor 2: Oral control and perceived social pressures 0.64 9.85

8 Feel that others would prefer if I ate more -.172 .761 .098

20 Feel that others pressure me to eat -.131 .731 .058

13 Other people think that I am too thin -.276 .574 .017

9 Vomit after I have eaten .240 .547 -.051

25 Enjoy trying new rich foods .236 .517 -.011

5 Cut my food into small pieces .260 .363 -.078

Factor 3: Food preoccupation 0.66 8.61

3 Find myself preoccupied with food .036 .011 .748

21 Give too much time and thought to food .021 .095 .738

4 Have gone on eating binges where I feel that I may not be able to stop -.081 -.025 .737

18 Feel that food controls my life .336 -.026 .430

Table 2: Component Correlation Matrix of EAT-23.

Factor 1 2 3

r

1 1.000

2 .091 1.000

3 .102 -.051 1.000

perceived social pressures) had a negative correlation with BMI (r = -0.36, p < 0.01).

Data obtained by regression models are demonstrated in Table 4. Oral control and perceived social pressures, as well as food preoccupation (after controlling for age, BMI and body image dissatisfaction) had a positive association with dieting and awareness of food content (p < 0.001). Nonetheless, BMI negatively was associated with oral control and perceived social pressures, but positively with dieting and being aware of food content, and preoccupation with food (after controlling for body image dissatisfaction, and factors 1 and 2) (p < 0.05). Furthermore,

body image dissatisfaction had positive associations with dieting and awareness of food content, oral control and perceived social pressures (after controlling for factors 1 and 3), and food preoccupation (p < 0.01).

DISCUSSION Factor analysis of the Arabic version of EAT-26 resulted

in a structure of three dimensions consisting of 23 items (i.e.; EAT-23). This highlights the critical role of culture (including language) in influencing eating attitudes, and their assessment using a translated version of EAT-26. Besides the cultural effect,

CentralBringing Excellence in Open Access

Mousa et al. (2016)Email:

J Hum Nutr Food Sci 4(2): 1084 (2016) 5/9

Table 3: Correlations between EAT-23 Factors and Other Indicators.

Indicator Age BMI Factor 1 Factor 2 Factor 3 BSQ-34

Age (years) 1

BMI (Kg/m2) .384** 1Dieting and awareness of food content

(Factor 1) .125* .420** 1

Oral control and perceived social pressures (Factor 2) -.047 -.362** .058 1

Food preoccupation (Factor 3) .130* .150** .284** -.011 1

Body image dissatisfaction (BSQ-34) .189** .466** .727** -.042 .381** 1* Data are considered statistically significant at p < 0.05. ** Data are considered statistically significant at p < 0.01.Abbreviations: BMI: Body Mass Index; BSQ-34: Body Image Dissatisfaction

Table 4: Association between EAT-23 Factors and other Indicators*.

IndicatorFactor 1 Factor 2 Factor 3

β P β P β PAge (years) 0.55 ** 0.024 0.22 0.027 0.14 ** 0.018

BMI (Kg/m2) 0.42 < 0.001 -0.45 < 0.001 0.06 ¶ 0.049

Body image dissatisfaction 0.15 < 0.001 0.02 ǂ 0.006 0.02 < 0.001Dieting and awareness of food content

(Factor 1) _ _ 0.12 < 0.001 0.07 † < 0.001

Oral control and perceived social pressures (Factor 2) 0.37 < 0.001 _ _ -0.01 0.84

Food preoccupation (Factor 3) 1.18 † < 0.001 -0.02 0.84 _ _* Data are considered statistically significant at p < 0.05. **Controlled for all indicators.†Controlled for age BMI, body image dissatisfaction. ǂ Controlled for factors 1 and 3¶ Controlled for body image dissatisfaction, and factors 1 and 2.Abbreviations: BMI: Body Mass Index

elimination of some items also could be attributed to the low frequency and variability of such behaviors in the current non-clinical population. Factor one is dieting and awareness of food content, which captured attitudes related to weight and eating concerns as well as weight loss. Factor two is oral control and perceived social pressures that described regulating food consumption and perceiving pressure from others regarding dietary intake and weight. Factor three is food preoccupation and assessed the emotional, mental and cognitive engagement in food. Thus, factor structure of EAT-23 is believed to be a valid scale that would assist researchers in investigating a variety of abnormal eating attitudes such as oral control and perceived social pressures in Arabic populations.

The present findings are in agreement with the 3-factor structure of EAT-26 reported by Garner and colleagues [17], except for few differences in the items loaded and composed the resulting factors. For instance, items two (i.e.; avoid eating when I am hungry) and 19 (i.e.; display self-control around food) of the current first factor were included in the oral control factor, as well as items nine (i.e.; vomit after I have eaten) and 25 (i.e.; enjoy trying new rich foods) of the second factor were part of the bulimia and food preoccupation as well as dieting factors of Garner et al., [17], respectively. In spite of these differences, which resulted from the diverse cultural backgrounds, the

present factor structure is found to be a reliable tool to discern different types of eating problems. In contrast to what Garner and others [17] found, bulimia was not part of the present third factor (i.e.; food preoccupation). The current factor only included binge eating but not self-induced vomiting. Hence, bulimia and food preoccupation could be distinct attitudes and independent of each other, which requires further investigation to elaborate the type of relation between these two factors. This suggests that diagnosis of bulimic attitudes should be performed using other questionnaires, which is supported by Nasser [29] and Douka et al., [21].

Factor structure of the current EAT-23 is slightly different from that described by other researchers. For example, EAT-26 was utilized to evaluate eating attitudes in 235 American Caucasian and Hispanic university female students. The resulting factor structure had 16 items that formed four factors; self-perception (e.g.; am terrified about being overweight), dieting (e.g.; think about burning up calories when I exercise), preoccupation (e.g.; find myself preoccupied with food), and food control (e.g.; avoid foods with sugar in them) [30]. Moreover, Koslowsky and others [3] explored eating disturbances and factor structure of EAT-26 in 809 Israeli female soldiers aged 18-19 years. Authors reported four factors of 20 items, namely dieting (e.g.; like my stomach to be empty), oral control (e.g.; other people think I am too thin),

CentralBringing Excellence in Open Access

Mousa et al. (2016)Email:

J Hum Nutr Food Sci 4(2): 1084 (2016) 6/9

awareness of food content (e.g.; particularly avoid foods with high carbohydrate content), and food preoccupation (e.g.; give too much time and thought to food) [3].

In line with this, eating problems in Canadian adolescent and young women aged ≥ 15 years using EAT-26 were studied. Data designated four factors consisting of 19 items. The factors included food preoccupation (e.g.; have gone on eating binges where I feel that I may not be able to stop), body image preoccupation (e.g.; think about burning up calories when I exercise), self-imposed dieting (e.g.; am preoccupied with a desire to be thinner), and perceived external pressure to eat (e.g.; feel that others would prefer if I ate more) [30]. Another study examined eating attitudes in 3000 British adolescent girls using EAT-40. Factor analysis produced five factors comprising of 34 items were obtained upon performing factor analysis including: dieting (e.g.; aware of the calorie content of foods that I eat), preoccupation with food (e.g.; have gone on eating binges where I feel that I may not be able to stop), vomiting (e.g.; have the impulse to vomit after meals), social pressure (e.g.; feel that others pressure me to eat), and social eating (e.g.; enjoy trying new rich foods) [31]. Similar findings have also been documented in Australia [24], Malaysia [19], Spain [22], Greece [21], and China [20].

Overall, the present factor structure is somewhat different in the number of factors produced; nevertheless the items of the current factors are comparable to what others have indicated. The discrepancy in the items composing each factor could be attributed to cultural and ethnic differences. For instance, the Arabic society is an orthodox one that is characterized by specific eating patterns, dressing styles, and cultural norms that value plumpness as a reflection of female beauty. Regardless of these dissimilarities, the current factor structure of the Arabic version of EAT-26 has the ability to diagnose individuals with various forms of negative eating attitudes similar to that of a version in any other language including English.

Interestingly, eating disturbances were evaluated in 351 adolescent school girls in Egypt utilizing an Arabic version of EAT-40. Factor analyses that Nasser [29] completed was based on the three- and five- factor structures of Garner et al., [17] and Eisler and Szmukler [32], respectively. Despite conducting the current study 2 decades later, as well as the difference in the scale version used and type of analyses performed, the present findings support those of Nasser [29] in which that the factor structure of an Arabic version is as reliable as that of an English one.

Displaying oral control and perceiving social pressures as well as preoccupying with food (after controlling for confounders) have been found to increase dieting and awareness of food content among the participants. This can be elucidated by that weight control behaviors, which include dieting, are achieved by reducing food intake, controlling the type/amount of the food consumed, and/or thinking about the food that could facilitate weight loss. These behaviors therefore, would promote dieting and awareness of food content among adolescent girls. In contrast, Koslowsky and colleagues [3] showed that dieting was positively correlated with awareness of food content and food preoccupation, yet negatively with oral control. These discrepancies are attributed to the differences in culture and

ethnicity, as indicated earlier, and probably in the factors and items produced.

Body mass index had direct relationships with dieting and awareness of food content, and food preoccupation as well. One explanation is that being heavy would encourage the individual to diet and try to estimate the value of the food that is going to be eaten in terms of nature and quantity, and how this food would affect weight status. All these attitudes would help adolescent girls to lose and not gain weight. Nonetheless, these associations are opposed by the negative relationships observed in the Canadian population [31]. Overweight women were at lower risk to be preoccupied with food, but not underweight ones who were 3.5 times at higher risk to engage in such behavior (p < 0.05) [31]. The lack of agreement between the two studies is suggested to be due to the differences in the population characteristics and factor structure of EAT-26.

On the other hand, BMI inversely was associated with oral control and perceived social pressures. This is probably due to that participants who are either underweight or of normal weight, would receive remarks regarding body weight such as being thin and the need to increase their food intake. However, those girls might be satisfied with their weight status; hence the received comments would encourage them to control or reduce their food intake in order to remain thin. This is supported by the results designated by Park and Beaudet [31] in which underweight girls were at higher risk to perceive external pressures to eat by 6.5 fold as compared to normal weight ones (p < 0.05).

In addition, positive associations have been found between body image dissatisfaction and the first and third factors of EAT. These relations would be explained by that participants with negative body image perception are criticized about their body weight and/or shape. These critiques would predispose the girls to have negative eating attitudes such as dieting, exerting oral control over food intake, as well as being preoccupied with food and aware of its nutritional value. Consequently, those girls develop such attitudes to lose weight and improve their body shape/image, which is believed to make others accept them. The lack of correlation between factor 2 (oral control and perceived social pressures) and body image dissatisfaction is unknown and worth further investigation. The current findings however are comparable to those described in Israeli [3] and Chinese [20] women.

The present study had few limitations including, not applying a clinical interview for the participants who scored above the cut-off point of the EAT-26 because of the limited resources. This weakened the ability to estimate the exact number of adolescent girls who suffered from an eating disorder such as anorexia nervosa. Another drawback is the inability to generalize the current findings because of its cross-sectional design. Therefore, future longitudinal research studies should be conducted, which are advised to assign a psychiatrist to clinically diagnose participants. Moreover, readjustment of the cut-off point of the current version of EAT-23 is required, because a value of 20 might not be sensitive in capturing disturbances of eating attitudes. Thus, further research in larger populations is needed to confirm our findings and to test for homogeneity of the items of the Arabic factor structure of EAT-26 using confirmatory factor analysis. In

CentralBringing Excellence in Open Access

Mousa et al. (2016)Email:

J Hum Nutr Food Sci 4(2): 1084 (2016) 7/9

addition, factors of the current EAT-23 had an acceptable internal consistency (Cα = 0.64 - 0.85), and it is higher than that reported by Garner et al., (1982) for factors 2 (Cα = 0.61) and 3 (Cα = 0.46) [17]. Nonetheless, future studies are needed to re-assess the internal consistency of the current factor structure, particularly for factors 2 and 3. In spite of these limitations, the present factors of EAT-23 are relatively reliable screening indicators of a variety of eating disturbances in non-clinical Arabic populations who could have weight, shape and body image concerns. This is important when conducting studies to examine the influence of culture or westernization on specific types of eating attitudes and behaviors. In particular, through the comparison of eating problems in Arabic participants residing in their home country, with those living in a western country for a specific period of time (e.g.; few years to pursue their education). This procedure would achieve greater validity and better estimation of such disturbances if researchers used the Arabic version of the factor structure of EAT-23 in both populations.

CONCLUSION In conclusion, the three-factor structure of the Arabic

version of EAT-26 is a useful first step-screening tool for eating disturbances in non-clinical populations. This is attributed to the ability of the factors in distinguishing individuals who exhibit different forms of negative eating attitudes from those who do not. The disagreement of the current factor structure with those reported in other nationalities results from the cultural and ethnic differences. Hence, this supports the essential need to use the present factors of EAT-23 when investigating eating problems in Arabic populations, to obtain more accurate results. Finally, the three dimensions of EAT-23 were associated with each other as well as with BMI and body image dissatisfaction. This would imply that weight status and negative body image perception could function as indicators of dieting and awareness of food content, oral control and perceived social pressures, and food preoccupation.

REFERENCES1. World health Organization. Nutrition in adolescence: Issues and

challenges for the health sector: issues in adolescent health and development. A discussion paper. Geneva; 2005.

2. Jones JM, Bennett S, Olmsted MP, Lawson ML, Rodin G. Disordered eating attitudes and behaviours in teenaged girls: a school-based study. CMAJ. 2001; 165: 547-552.

3. Koslowsky M, Scheinberg Z, Bleich A, Mark M, Apter A, Danon Y, et al. The factor structure and criterion validity of the short form of the Eating Attitudes Test. J Pers Assess. 1992; 58: 27-35.

4. Wood A, Waller G, Miller J, Slade P. The development of eating attitude test scores in adolescence. Int J Eat Disord. 1992; 11: 279-282.

5. Sánchez-Carracedo D, Neumark-Sztainer D, Lopez-Guimera G. Integrated prevention of obesity and eating disorders: barriers, developments and opportunities. Public Health Nutr. 2012; 15: 2295-2309.

6. Pourghassem Gargari B, Kooshavar D, Seyed Sajadi N, Safoura S, Hamed Behzad M, Shahrokhi H. Disordered Eating Attitudes and Their Correlates among Iranian High School Girls. Health Promot Perspect. 2011; 1: 41-49.

7. Al-Subaie A, al-Shammari S, Bamgboye E, al-Sabhan K, al-Shehri S,

Bannah AR. Validity of the Arabic version of the Eating Attitude Test. Int J Eat Disord. 1996; 20: 321-324.

8. Al-Adawi S, Dorvlo AS, Burke DT, Al-Bahlani S, Martin RG, Al-Ismaily S. Presence and severity of anorexia and bulimia among male and female Omani and non-Omani adolescents. J Am Acad Child Adolesc Psychiatry. 2002; 41: 1124-1130.

9. Eapen V, Mabrouk AA, Bin-Othman S. Disordered eating attitudes and symptomatology among adolescent girls in the United Arab Emirates. Eat Behav. 2006; 7: 53-60.

10. Musaiger AO, Al-Mannai M, Tayyem R, Al-Lalla O, Ali EY, Kalam F, et al. Risk of disordered eating attitudes among adolescents in seven Arab countries by gender and obesity: a cross-cultural study. Appetite. 2013; 60: 162-167.

11. Musaiger AO, Al-Mannai M, Tayyem R, Al-Lalla O, Ali EY, Kalam F, et al. Risk of disordered eating attitudes among adolescents in seven Arab countries by gender and obesity: a cross-cultural study. Appetite. 2013; 60: 162-167.

12. Madanat HN, Lindsay R, Campbell T. Young urban women and the nutrition transition in Jordan. . Public Health Nutr. 2011; 14: 599-604.

13. Mousa TY, Al-Domi HA, Mashal RH, Jibril MA. Eating disturbances among adolescent schoolgirls in Jordan. . Appetite. 2010; 54: 196-201.

14. Sweeting H, West P. Gender differences in weight related concerns in early to late adolescence. J Epidemiol Community Health. 2002; 56: 700-701.

15. Cho J, Han S, Kim J, Lee H. Body image distortion in fifth and sixth grade students may lead to stress, depression, and undesirable dieting behavior. Nutr Res Pract. 2012; 6: 175-181.

16. Presnell K, Bearman SK, Stice E. Risk factors for body dissatisfaction in adolescent boys and girls: a prospective study. Int J Eat Disord. 2004; 36: 389-401.

17. Mousa TY, Mashal RH, Al-Domi HA, Jibril MA. Body image dissatisfaction among adolescent schoolgirls in Jordan. Body Image. 2010; 7: 46-50.

18. Garner DM, Olmsted MP, Bohr Y, Garfinkel PE. The eating attitudes test: psychometric features and clinical correlates. Psychol Med. 1982; 12: 871-878.

19. Mintz LB, O’Halloran MS. The Eating Attitudes Test: validation with DSM-IV eating disorder criteria. J Pers Assess. 2000; 74: 489-503.

20. Talwar P. Factorial analysis of the eating attitude uest (EAT-40) among a group of Malaysian university students. MJP. 2011; 20: 1-10.

21. Lee S, Lee AM. Disordered eating in three communities of China: a comparative study of female high school students in hong kong, Shenzhen, and rural hunan. Int J Eat Disord. 2000; 27: 317-327.

22. Douka A, Grammatopoulou E, Skordilis E, Koutsouki D. Factor analysis and cut-off score of the 26-item eating attitudes test in a greek sample. J Bio of Exerc. 2009; 5: 51-68.

23. Rivas T, Bersabé R, Jiménez M, Berrocal C. The Eating Attitudes Test (EAT-26): reliability and validity in Spanish female samples. Span J Psychol. 2010; 13: 1044-1056.

24. Maïano C, Morin A, Lanfranchi M, Therme P. The Eating Attitudes Test-26 revisited using exploratory structural equation modeling. J Abnorm Child Psychol. 2013; 41: 775-788.

25. Gleaves DH, Pearson CA, Ambwani S, Morey LC. Measuring eating disorder attitudes and behaviors: a reliability generalization study. J Eat Disord. 2014; 2: 6.

26. Alwan A, Takrurui H, As’ad A, Belbeisi A. Country profile. In Alwan A KS, editor. Nutrition in Jordan. Amman: Ministry of Health, Ministry of

CentralBringing Excellence in Open Access

Mousa et al. (2016)Email:

J Hum Nutr Food Sci 4(2): 1084 (2016) 8/9

Agriculture, and World Health Organization. 2006. 16-29.

27. O’Brien KM, LeBow MD. Reducing maladaptive weight management practices: developing a psychoeducational intervention program. Eat Behav. 2007; 8: 195-210.

28. Gibson R. Principles of Nutritional Assessment. 1st ed. Oxford: Oxford University Press, NY; 1990.

29. Centers for Disease Control and Prevention (CDC). Body Mass Index-for-age weight status categories for children and teens. 2007.

30. Nasser M. The psychometric properties of the Eating Attitude Test in a non-Western population. Soc Psychiatry Psychiatr Epidemiol. 1994;

29: 88-94.

31. Belon KE, Smith JE, Bryan AD, Lash DN, Winn JL, Gianini LM. Measurement invariance of the Eating Attitudes Test-26 in Caucasian and Hispanic women. Eat Behav. 2011; 12: 317-320.

32. Park J, Beaudet MP. Eating attitudes and their correlates among Canadian women concerned about their weight. Eur Eat Disord Rev. 2007; 15: 311-320.

33. Eisler I, Szmukler GI. Social class as a confounding variable in the Eating Attitudes Test. J Psychiatr Res. 1985; 19: 171-176.

CentralBringing Excellence in Open Access

Mousa et al. (2016)Email:

J Hum Nutr Food Sci 4(2): 1084 (2016) 9/9

Mousa T, Beretvas SN (2016) Factor Structure of Scores of an Arabic Version of the Eating Attitude Test. J Hum Nutr Food Sci 4(2): 1084.

Cite this article