|Year : 2017 | Volume
| Issue : 3 | Page : 219-224
Eating attitudes and body shape concerns among medical students in Chandigarh
Nitin Gupta1, Rachna Bhargava2, Bir Singh Chavan1, Pratap Sharan3
1 Department of Psychiatry, Government Medical College and Hospital, Chandigarh, India
2 Department of Clinical Psychology, National De-addiction and Treatment Centre, New Delhi, India
3 Department of Psychiatry, All India Institute of Medical Sciences, New Delhi, India
|Date of Web Publication||14-Sep-2017|
Department of Psychiatry, All India Institute of Medical Sciences, New Delhi
Source of Support: None, Conflict of Interest: None
Background/Objectives: Eating disorders are rarely encountered in the Indian subcontinent. Surprisingly, there is sparse literature related to eating attitude distortions and body dissatisfaction in the Indian population. The objective of this study was to explore the eating attitudes and body shape concerns in medical students, with the objectives of determining any gender differences on eating attitudes and body shape concerns, and any predictors of psychological morbidity using demographic and psychosocial parameters. Materials and Methods: Sample comprised medical undergraduate students from all years (I-V) pursuing MBBS course at Government Medical College and Hospital, Sector 32, Chandigarh. It was a prospective, cross-sectional study. Instruments used were socioclinical sheet, acculturation index (AI), eating attitudes test-26 (Hindi version), and body shape questionnaire (BSQ) (Hindi version). Following written informed consent, all the questionnaires were administered in a group setting to each MBBS batch. Statistical analysis was carried out with Statistical Package for the Social Sciences (SPSS) version 17.0 using descriptive analysis (frequency, percentage, mean), univariate analysis (Chi-square, t-test), Pearson's correlational analysis, and binary logistic regression (backward). Results: A total of 250 medical students from 5 batches with mean age of 20.15 +/- 1.32 (18-28) years. And 55% (n = 137) were males and 45% (n = 113) were females. On gender comparison, males were more likely to be Hindu and have higher body weight (actual and ideal); females had significantly higher scores on dieting subscale of eating attitudes test-26 and BSQ (total score). On AI, there were no gender differences on family domain variables, with significantly less females reporting pure Western preference for social/peer domain variables. Correlation and logistic regression analysis showed high score on BSQ (median divide) as the only statistically significant predictor of eating (disorder) morbidity, whereas high score on AI (median divide) approached significance (P = 0.062). Conclusion: High score on BSQ is the mediating (proximate) risk factor for eating (disorder) morbidity. Influence of other/distal risk factors (especially AI) may be mediated through it.
Keywords: Acculturation index, body shape concerns, eating attitudes, eating disorders, India
|How to cite this article:|
Gupta N, Bhargava R, Chavan BS, Sharan P. Eating attitudes and body shape concerns among medical students in Chandigarh. Indian J Soc Psychiatry 2017;33:219-24
|How to cite this URL:|
Gupta N, Bhargava R, Chavan BS, Sharan P. Eating attitudes and body shape concerns among medical students in Chandigarh. Indian J Soc Psychiatry [serial online] 2017 [cited 2019 Aug 19];33:219-24. Available from: http://www.indjsp.org/text.asp?2017/33/3/219/214605
| Introduction|| |
Eating disorders (EDs) historically have been reported to be uncommon in South Asian countries.,, Recent literature also suggests that EDs remain rare in the Indian subcontinent. However, research into body shape and eating attitudes from nonWestern settings is increasing. In the last decade, research from many Asian countries (e.g., Japan, Korea, Taiwan, China, and Pakistan) has shown high rates of body dissatisfaction and eating attitude distortions, with gender differences being reported in some studies.,,,,,, The research is perhaps fuelled by the expected impact of globalization and Westernization on eating and body image of individuals in transitional Asian societies.
As a society and country, India has also been exposed and is undergoing the process of “Westernization” at various levels. Surprisingly, there is sparse literature related to eating attitude distortions and body dissatisfaction in the Indian population.,,,
This study explores the eating attitudes and body shape concerns in medical students, as this segment of youth is drawn to the West both for extracurricular and curricular pursuits/requirements (almost all books are Western, desire for further degrees from or career in developed countries where English is the main language).
Aim and Objectives
To determine the eating attitudes and body shape concerns among medical students.
- To determine any gender differences on eating attitudes and body shape concerns.
- To determine any predictors of psychological morbidity using demographic and psychosocial parameters.
| Materials and Methods|| |
Medical undergraduate students from all years (I-V) pursuing MBBS course at Government Medical College and Hospital, Sector 32, Chandigarh. Each batch comprised of maximum 50 students.
Prospective; Cross-sectional study.
- Socioclinical sheet: This was specially developed by the investigators and contained relevant sociodemographic (age, gender, etc.) and personal (weight, height, etc.) variables
- Acculturation index (AI): This was specially developed by the corresponding author (PS) and colleagues to assess the influence of Westernization. It comprises of six items related to preferences for movies, books, food, dress, language, and relationship with the opposite sex. Each item is scored on a 3-point scale. Scores for English-Western; for both Hindi-Traditional and English-Western; and Hindi-Traditional were 3, 2, 1 respectively. The AI has been shown to have face validity, and adequate internal consistency and item-total correlations. Its score range from 6 to 18
- Eating Attitudes Test (EAT-26): The EAT is most widely used screening questionnaire for abnormal eating attitudes. EAT has been widely translated and adapted in different languages, cultures, and countries. The EAT-26 was originally developed by Garner et al. and evaluate attitudes, feelings, and preoccupations in relation to food, weight, and exercise. It has three subscales viz., dieting, bulimia and food preoccupation, and oral control. A Hindi version of the EAT-26, translated by the authors and with adequate internal consistency and convergent validity [with body shape questionnaire (BSQ)], was used for this study. It comprises of 26 items; however, its rating format was changed from likert type (0-3 rating score) to a ‘forced choice’ (no/yes = 1/2) type to accommodate the less sophisticated population and also to counteract the tendency to mark intermediate responses. Hence, the total score ranges from 26-52
- BSQ: Originally developed by Cooper et al.; it is a measure of dissatisfaction with body shape, in association with EDs. It has been shown to measure a single dimension of concern for body shape. A Hindi version of the BSQ, translated by the authors and with adequate internal consistency and convergent validity (with EAT-26), was used for this study. It comprises of 34 items with “forced choice” (yes/no) responses and the total score ranges from 34 to 68.
Each MBBS batch of the identified sampling frame was approached by one of the investigators (RB) at the end of the theory class being attended by them. The students were explained about the study and questionnaires were distributed to them after obtaining written informed consent (as detailed below). The questionnaires were group administered in class rooms that assured privacy regarding responses. Each student was encouraged to respond to all questions. However, a small number of respondents (ranging from 1 to 5 in each batch) did leave some items blank. This was never in excess of 5% of items.
Written informed consent was taken from the respondent. The respondent was made aware about the purpose of the study and nature of the questionnaires. He/she was also informed that they had the freedom not to participate, and had the liberty of withdrawing their consent at any time during the study. Anonymity and confidentiality was ensured for each respondent.
Statistical analysis was carried out using SPSS version 17.0.
Descriptive analysis in the form of calculation of frequency, percentage, and mean (standard deviation) was carried out. Univariate analysis in the form of Chi-square and t-test were carried out for objective 1.
For objective 2, correlational analysis of the key demographic and psychosocial variables was carried out using Pearson's correlational analysis. Following that, Binary logistic regression (backward) was carried out; with 95th percentile divide on EAT as the dependent/response variable. The cut off at 95th percentile was taken as a signifier of psychological morbidity as this corresponds to two standards deviation (for directional hypotheses) on a normally distributed measure (here EAT score). The predictor variables entered in the regression were all categorical: Either they had been measured in categories (e.g., gender) or they were categorized using median divide (e.g., BSQ score). A median divide (with averages for the two consequent subgroups centered on 25th percentile and 75th percentile, respectively) serves to delineate clinically (easily) identifiable low and high scorers on such variables.
The level of statistical significance was kept as P < 0.05 for all tests.
| Results|| |
A total of 250 medical students from 5 batches comprised the sample for this study. The mean age was 20.15 + 1.32 (18-28) years. In terms of gender distribution, 55% were males (n = 137) and 45% were females (n = 113). A total of 75% (n = 187) belonged to Hindu religion; 91% (n = 229) belonged to nuclear family set-up; and 70% (n = 176) were residents of Chandigarh.
In order to evaluate objective 1, the whole sample was divided into two groups on the basis of gender, i.e., Group A (males; n = 137) and Group B (females; n = 113).
On comparing both groups across the demographic and personal variables, males were more likely to be Hindu by religion and have higher body weight (actual and ideal) than females, who were more likely to be Sikh [Table 1].
On comparing both (gender) groups across the psychological variables, females had significantly higher scores on dieting subscale of EAT-26 and BSQ (Total score). Both groups were comparable on other subscales and parameters [Table 2].
|Table 2: Gender differences on Eating attitude, body shape, and acculturation index|
Click here to view
An attempt was made to evaluate gender-related differences in relation to the AI. Both groups had comparable scores on ‘kind of food’ and ‘kind of language’ used at home. However, there were significant differences on the items related to movies, books, dress and (gender) mingling. Significantly higher percentage of males reported preference for Western movies, Hindi (regional) books, Western dresses, and uninhibited mingling between boys and girls. Significantly higher percentage of females reported preference for Hindi (regional) movies, English (Western) books, both kind (Western and traditional) of dresses outside home, and partial mingling between boys and girls [Table 3].
In order to evaluate Objective 2, correlation analysis was carried out using all the key demographic and psychological variables [Table 4]. EAT-26 correlated with BSQ, Acculturation Index, Age and Batch. BSQ correlated with Age and Batch. Acculturation Index did not correlate with any demographic variable.
Using EAT-26 score as a dependent variable, the significant variables were entered into a Logistic Regression (backward) model. The cut-off was taken as the 95th percentile (indicative of psychological morbidity defined in statistical terms). High score on BSQ (median divide) was the only statistically significant predictor of eating (disorder) morbidity, whereas high score on AI (median divide) approached significance (P = 0.062) as a predictor of eating (disorder) morbidity [Table 5].
|Table 5: Logistic regression (backward) with eating attitude as dependent variable (n=243)|
Click here to view
| Discussion|| |
EDs are uncommon in the Indian sub-continent as compared to rates reported in the West., Various factors and hypotheses have been put forward for the same; differences in attitude towards body weight and shape probably being the most critical factor., In the Indian sub-continent, clinical presentations of EDs are not only uncommon but also differ in nature and severity.,,, Due to the same, researchers have argued that a simplistic focus on trying to identify the well-defined major EDs alone will not fulfill the need for identifying determinants of these disorders. Considering the cultural specificity of EDs, many experts have emphasized the need to study the attitudes related to these disorders in the non-Western settings.,,
The sample had a mean age of around 20 years, which was in keeping with the expected average age for five consecutive MBBS batches. Just over half of the sample consisted of males. The majority of students were Hindus and from nuclear families; in keeping with the profile of families from this part of the North India. Majority belonged to the city of Chandigarh, which was as per the admission system of the medical college.
Both genders were comparable on all demographic variables, except that males were more likely to be Hindu (and females more likely to be Sikh) by religion. This difference is likely to be a function of the demographic profile of the region (e.g., migrants are more likely to be male and Hindu). Regarding personal parameters, as expected, males had a significantly higher body weight; this being maintained for their “expected ideal weight” too. Hence, the demographic and personal profile did not appear to reveal any critical or unexpected gender related differences [Table 1].
On comparing the psychosocial variables [Table 2], it was seen that, apart from the dieting subscale, both genders were comparable on other subscales and total score of EAT-26. Higher scores on “dieting” for females could be reflective of the expected tendency of females to be more focused on dieting and avoidance of fatty food, even in the Indian culture., Comparable scores on other subscales is probably reflective of the lack of internal focus of females on these parameters (as reflected by overall low scores), but this is conjectural. Additionally, it is interesting to note that both groups were overall comparable on EAT-26 indicating similar presence/absence of eating-related psychological distress in both females and males.
The females scored significantly higher on the BSQ than the males. It is of interest to note that though there has been very little research on dissatisfaction with body shape, studies from India have reported conflicting results., This may be attributable to the differences in methodological framework, respondent sample, etc., Although Srinivasan et al. had studied gender differences in medical students, but we are unable to make any direct comparison with their findings as they used the Bulimia Investigatory Test rather than BSQ. We were unable to find any other study from India that assessed gender differences and body shape concerns in medical students.
In the current era where rapid sociocultural changes are impacting values, attitudes, habits, and lifestyle, their effect of cultural tradition and acculturation on attitudes, behaviors, and psychopathology relating to eating and body shape needs to be studied., However, understanding the role of culture is not easy as the concept of cultural orientation is a complex phenomenon. Attempts have been made by researchers to measure this cultural orientation in the form of “Acculturation questionnaire,” “Acculturation scale,” and “AI” from the USA, UK, and India (by the authors of the current study), respectively. The underlying concept centers around the potential conflicts generated due to the interplay of traditional and Western cultural orientation and attitudes that occur between the individual and the environment. Keeping this in perspective, we shall proceed now onto the discussion related to acculturation.
Males and females were comparable regarding the total scores obtained on the AI, indicating that both groups were experiencing similar levels of influence of Westernization [Table 2]. Of the six items of the AI, two are in the family domain (“kind of food” and “kind of language” used at home), two in social/peer domain (“kind of dress worn” and “mingling between girls and boys”) and two in personal domain (preferred books and movies). On comparing the two groups for each of the six items of AI, it was seen that they were comparable on items in the family domain (“kind of food” and “kind of language” used at home). This is not unexpected as family influences are likely to be equally applicable across genders. For social/peer domain variables, significantly less number of females reported preference for pure Western dresses and uninhibited (gender) mingling; preferring both kind of dresses outside home and partial (gender) mingling, i.e., they appeared less Westernized than males in this domain. On the personal domain variables, significantly lesser number of females reported preference for Western movies, while significantly more females reported preference for Western books. It can be hypothesized that females are more inhibited in the social domain, possibly due to a complex interplay of individual, group, family, social, and cultural factors. The greater salience of sociocultural factors for females can potentially make them intrapsychically ambivalent regarding their value orientation (traditional Indian and Western values), thereby making them more prone to developing pathologies related to eating and body shape; similar to the finding by Mumford et al. amongst Asian girls in the UK. This issue shall be discussed in further detail under Objective 2. Nevertheless, it merits further study and is an area for future research.
A two-stage analysis was carried out to understand the complex interplay between demographic, psychosocial variables and eating attitudes, and body shape issues. Initially, a correlational matrix was generated using all the relevant variables [Table 4]. Following this, logistic regression was applied using EAT-26 as the dependent variable. It was seen that EAT-26 and BSQ correlated with each other and variables of batch and age. Hence, results indicated that cohort (batch) maybe more important variable than age, but this should be viewed as tentative than definite.
The Hindi version of EAT-26 does not have a specified cut-off score for identifying EDs. On the face of it, this may appear to be a limitation of the study. However, the issue of cut-off needs to be examined against the intricacies inherent in the assessment and diagnosis of EDs in non-Western contexts. Although EAT-26 has been translated for use in other languages/countries, researchers have often needed to modify it (EAT-13 in Greece) or have faced difficulties in using it for identification of EDs., King and Bhugra have demonstrated the conceptual limitations of translating and using EAT-26 in Hindi, e.g., the misunderstanding between dieting and religious fasting. This was largely overcome by us in this study by using Hindi phrases (instead of single English words) to convey the meaning of the original and occasionally including alternative English terms when the translations were inexact (e.g., urja/calories). On balance, available evidence mandates that Western cut-offs of EAT-26 should not be applied in a “non-Western” culture like India, without adequate standardization in multiple samples.
The cut-off for eating (disorder) morbidity as defined by EAT-26 scores was taken as the 95th percentile (statistically defined morbidity). Application of binary regression statistical model for determining potential predictors of morbidity revealed that BSQ (median divide) was the only significant predictor [Table 5]; suggesting that high score on BSQ is the mediating (proximate) risk factor for eating (disorder) morbidity. Influence of other/distal risk factors may be mediated through it. This conclusion, however, needs to be tempered by the fact that we have studied a limited number of predictor variables in a special sub-population and in a single study. “Inexorable drive for thinness” has been seen as the meta-narrative for the emergence and spread of eating and body shape pathology all over the world.
However, it was also observed that high scores on AI approached significance as a predictor for eating (disorder) morbidity (P = 0.062) [Table 5]. Keeping in view its theoretical relevance (and the limited range of scores offered by this very brief measure), we feel it is relevant to discuss it in brief here. Acculturation may be particularly relevant in settings where cultural conflict and consequent intrapsychic ambivalence are salient for concerned individuals, for example, in societies and areas undergoing rapid cultural transition rather than in those showing less shifts in value orientation (either traditional or Western). Additionally, acculturation issues may be more relevant in the female than male gender; an issue also highlighted by Mallick et al. However, these conjectures should be viewed as hypotheses that need to be tested in further studies.
Our study was limited by the following factors viz., cross-sectional design, single sample, a special subpopulation that is not representative of the general population, absence of a priori cut-off scores for Hindi EAT-26 and BSQ, and nonapplication of structured/semistructured clinical diagnostic assessment. However, while these limitations suggest caution in regarding the study's conclusions as definitive; some were unavoidable as there is no standardized cut-off for EAT-26 and BSQ and there are no standardized interviews for diagnostic assessment of eating pathology for use in India. The study also has some strengths: It was adequately powered for the analyses that were conducted, had a very high response rate, addressed both genders rather than focusing only on females, used appropriate and psychometrically sound instruments adapted for use in the local language, and addressed the complex issue of ‘culture’ by using the AI.
| Conclusions|| |
In a special subpopulation of medical undergraduate students from North India based on gender, similarities outweighed differences (on BSQ) when compared for eating attitudes and body shape concerns. Nevertheless, the differences that emerged on BSQ were robust and consistent even when subjected to multivariate analysis. Additionally, cultural orientation and intrapsychic-intergenerational conflict appeared to be another relevant variable highlighting the relevance of cultural factors in relation to EDs. Future studies need to focus on trying to understand the cultural variations in eating attitudes, body shape concerns, and EDs in the Indian subcontinent using a longitudinal design, culturally appropriate instruments, thereby additionally being able to address the limitations mentioned above.
Financial support and sponsorship
Conflict of interest
There are no conflicts of interest.
| References|| |
Srinivasan TN, Suresh TR, Jayaram V. Emergence of eating disorders in India: Study of eating distress syndrome and development of a screening questionnaire. Indian J Soc Psychiatry 2000;16:81-90.
Soh NL, Touyz SW, Surgenor LJ. Eating and body image disturbances across cultures: A review. Eur Eat Disord Rev 2006;14:54-65.
Balhara YPS, Mathur S, Kataria DK. Body shape and eating attitudes among female nursing students in India. East Asian Arch Psychiatr 2012;22:70-4.
Yang SJ, Choi Y, Lee HY. Eating attitude, ego-identity, and self-esteem among the urban high school girls. J Korean Neuropsychiat Assoc 2001;40:477-86.
Sakamaki R, Toyama K, Amamoto R, Liu CJ, Shinfuku N. Nutritional knowledge, food habits and health attitude of Chinese university students- a cross sectional study. Nutrition J 2005;4:4.
Yamamoto C, Uemoto M, Shinfuku N, Maeda K. The usefulness of body image tests in the prevention of eating disorders. Kobe J Med Sci 2007;53:79-91.
Liao Y, Knoesen NP, Castle DJ, Tang J, Deng Y, Bookun R, et al
. Symptoms of disordered eating, body shape, and mood concerns in male and female Chinese medical students. Compr Psychiatr 2010;51:516-23.
Chang YJ, Lin W, Wong Y. Survey on eating disorder-related thoughts, behaviours, and their relationship with food intake and nutritional status in female high school students in Taiwan. J Am College Nutr 2011;30:39-48.
Liao Y, Liu T, Cheng Y, Wang J, Deng Y, Hao W, et al
. Changes in eating attitudes, eating disorders and body weight in Chinese medical university students. Int J Soc Psychiatr 2013;59:578-85.
Memon AA, Adil SE, Siddiqui EU, Naeem SS, Ali SA, Mehmood K. Eating disorders in medical students of Karachi, Pakistan: A cross-sectional study. BMC Res Notes 2012;5:84-90.
Mallick N, Ray S, Mukhopadhyay S. Eating behaviours and body weight concerns among adolescent girls. Advances in Public Health 2014;2014:8 Doi:10.1155/2014/257396
Shroff H, Thompson JK. Body image and eating disturbance in India: Media and interpersonal influences. Int J Eat Disord 2004;35:198-203.
Balhara YPS, Yadav T, Arya K, Mathur S, Kataria DK. A cross-sectional study of body shape and eating attitude among Indian female healthcare students. Int J Psychiatr Med 2012;43:309-23.
Upadhyah A, Misra R, Parchwani D, Maheria P. Prevalence and riak factors for eating disorders in Indian adolescent females. Nat J Physiol Pharmacy Pharmacol 2014; 4:153-7.
King MB, Bhugra D: Eating Disorders: Lessons from a cross-cultural study. Psychol Med 1989;19: 955-8.
Garner DM, Olmsted MP, Bohr Y, Garfinkel PE. The Eating Attitudes Test: Psychometric features and clinical correlates. Psychol Med 1982;12:871-8.
Cooper PJ, Taylor MJ, Cooper Z, Fairburn CG. The development and validation of the Body Shape Questionnaire. Int J Eat Disord 1987;4:485-94.
Mumford DB, Whitehouse AM, Choudhary IY. Survey of eating disorders in English medium schools in Lahore, Pakistan. Int J Eat Disord 1992;11:173-84.
Garner PE, Moldofsky H, Garner DM. The heterogeneity of anorexia nervosa: Bulimia as a distinct subgroup. Arch Gen Psychiatry 1980;37:1036-40.
Khandelwal SK, Saxena S. Anorexia nervosa in people of Asian extraction. Brit J Psychiatry 1990;157:784.
Mendhekar DN, Arora K, Lohia D, Aggarwal A, Jiloha RC. Anorexia nervosa: An Indian perspective. Nat Med J India 2009;22:181-2.
Chandra PS Abbas S, Palmer R. Are eating disorders a significant clinical issue in urban India? A survey among psychiatrists in Bangalore. Indian J Eat Disord 2012; 45: 443-6.
Herzog DG, Keller MB, Lovori PW, Sacks NR. The course and outcome of bulimia nervosa. J Clin Psychiatry 1991;52: 4-8.
Patton GC, King MBN. Epidemiological study of eating disorders: Time for a change of emphasis. Psychol Med 1991;21:287-91.
Srinivasan TN, Suresh TR, Jayaram V, Fernandez MP. Eating disorders in India. Indian J Psychiatr 1995;37:26-30.
Khandelwal SK, Sharan P, Saxena S. Eating disorders: An Indian perspective. Int J Soc Psychiatr 1995;41:132-46.
Mumford DB, Whitehouse AM, Platts M. Sociocultural correlates of eating disorders among Asian schoolgirls in Bradford. Brit J Psychiatr 1991;158:222-8.
Pumariega AJ. Acculturation and eating attitudes in adolescent girls: A comparative and correlation study. J Am Acad Child Psychiatr 1986;25:276-9.
Douka A, Grammatopolou E, Skordilis E, Koutsouki D. Factor analysis and cut-off score of the 26-item eating attitudes test in a Greek sample. Biol Exercise 2009; 5.1: 51-67.
[Table 1], [Table 2], [Table 3], [Table 4], [Table 5]
|This article has been cited by|
||Eating disorders risk among medical students: a global systematic review and meta-analysis
| ||Haitham Jahrami,Mai Sater,Ahmed Abdulla,Mo’ez Al-Islam Faris,Ahmed AlAnsari |
| ||Eating and Weight Disorders - Studies on Anorexia, Bulimia and Obesity. 2018; |
|[Pubmed] | [DOI]|
||The relationship between risk of eating disorders, age, gender and body mass index in medical students: a meta-regression
| ||Haitham Jahrami,Zahraa Saif,Mo’ez Al-Islam Faris,Michael P. Levine |
| ||Eating and Weight Disorders - Studies on Anorexia, Bulimia and Obesity. 2018; |
|[Pubmed] | [DOI]|