Abstract
Background:
The lack of culturally-sensitive tool masks the epidemiology of weight-bias and internalized-stigma. This study aimed to translate, adapt, and validate the Modified Weight-Bias Internalization Scale (WBIS-M) into Hindi.
Methods:
The translation and adaptation process followed the World Health Organization guidelines. Three psychiatrists translated the original WBIS-M into Hindi. The bilingual panel comprised four psychiatrists, two laparoscopic bariatric surgeons, a professor of English, and an individual with severe obesity (with an International English Language Testing System (IELTS) score of Band 8.5) who was familiar with Western culture. The panel identified and resolved inadequate expressions of translations or discrepancies. An independent translator, whose mother tongue was English, back-translated it into the English language. Focused group discussions with individuals with severe obesity were conducted. The final H-WBIS-M was administered to undergraduates (n = 120) after seven days. Test–retest reliability was assessed. The original WBIS-M, Eating Disorder Diagnostic Screen, and Fat Phobia Scale-Short Form were administered to 120 undergraduates and 55 individuals with severe obesity (n = 175)]. Reliability and validity of H-WBIS-M were analyzed.
Results:
Kaiser–Meyer–Olkin measure of sampling adequacy was 0.788, and Bartlett’s test of sphericity was χ2 = 1381.892, p = 0.000. The exploratory factor analysis extracted three components. Spearman-Brown coefficient for H-WBIS-M was 0.875. The internal consistency was α = 0.875 (p = 0.000), intraclass correlation coefficient was 0.857 (95% confidence interval [CI] = 0.812–0.893) (p = 0.000), and Pearson correlation ranged from 0.869 to 0.989 (p = 0.000). Cross-language concordance revealed a significant intraclass correlation coefficient (0.877–0.986) (p = 0.000) and Pearson correlation ranging from 0.781 to 0.972 (p = 0.000).
Conclusions:
The H-WBIS-M demonstrated good psychometric properties based on standard testing procedures.
Keywords: Weight-bias internalization, weight stigma, epidemiology, scale development
Key Messages:
Weight bias internalization has adverse health effects. Hence, it is pertinent to measure and intervene.
There is no culturally sensitive instrument in India to measure it.
The present study culturally adapts and validates the Modified Weight Bias Internalization Scale for use among the Indian population.
The psychometric properties of the Hindi Modified Weight Bias Internalization Scale are acceptable and good and may be used for epidemiological purposes in India.
The terminologies “weight bias,” “weight prejudice,” and “weight stigma” denote negative attitudes, unjustifiable treatment, and discrimination toward individuals on the grounds of body weight. 1 People who are overweight or obese are often considered lazy, unmotivated, noncompliant, sloppy, lack self-discipline, will-power, and less competent. 2 As evident in some multinational studies, these derogatory notions toward persons with higher weight are ubiquitous societal scorn.3,4 Weight bias internalization or weight self-stigma is the devaluation of self due to one’s weight as a result of the assimilation of contemptuous stereotypies. 5
Weight bias internalization is conceptually an amalgam of three factors: awareness of negative stereotypes for social identity, agreement with the marked social identity, and application of those negative stereotypes for self and resultant self-devaluation.3,5,6
Weight bias internalization is known to have adverse mental and physical health effects.7,8 The experience of being teased by peers or family members is related to wide-array of psychosocial problems, including aggressive behavior, poorer self-esteem, depression, suicidal ideation, anxiety, body dissatisfaction, disordered eating, reduced engagement in health-promoting behaviors, worse body image, considerable perceived stress, and weight gain over time.7-10 Chronic stress results in dysregulation of various inflammatory markers, such as cortisol. 10 Higher internalization stigma scores are related to increased weight in the last year. 3
Most of the data related to weight bias internalization and its reverberations come from developed countries. Over the past 50 years, worldwide obesity rates have tripled. 11 By 2013, 62% of the world’s obese population resided in developing countries. 11 Considering the “obesity epidemic” or “globesity” in middle- and lower-income countries, understanding the epidemiology of weight bias internalization, its significant correlates, and their adverse effects is of immense public health significance.12,13
One of the hindrances in identifying the problem statement is the lack of standardized tool for measurement in Hindi or other regional languages. Only three scales are widely used and heavily cited for weight internalization stigma: the Weight Bias Internalization Scale (WBIS) (citation count = 648), the Weight Self-Stigma Questionnaire (WSSQ) (citation count = 295), and the Modified Weight Bias Internalization Scale (WBIS-M) (citation count = 310).5,14-18 The WBIS is an 11-item construct that examines multiple domains of internalized weight bias among persons who are overweight and obese. 5 The WSSQ is a 12-item tool to analyze weight self-stigma and multidimensional characteristics of weight bias internalization. It consists of two sub-scales: self-devaluation and fear of enacted stigma. 14 The WBIS-M is modified from Durso and Latner’s WBIS. The WBIS-M was designed to address the limitations of the original WBIS in terms of generalizability to normal populations across the genders, as it was limited to persons with obesity, and most of the participants in the validation research were females. The term “overweight” was modified to “my weight” in every item of the WBIS-M. The higher WBIS-M levels signify greater weight bias internalization and correspond to a range of psychological problems. 15
The WBIS-M is a popular tool for this purpose and has been validated in many languages, including English, Chinese, German, Spanish, Greek, and Turkish.3,15,19-24 It is an instrument to gauge weight bias internalization and also indicates that weight bias internalization is a distinguishable dimension from antifat attitudes, body image, and self-esteem. 24 The aim of the index study is Hindi translation, validation, and cultural adaptation of the WBIS-M for the Indian population.
Mounting research suggests that the epidemiology of weight bias and internalization is rising in the wake of rapid globalization and Westernization.3,4 However, there needs to be more research from South Asian, low-middle-income, and developing countries. India is the second-most populous country in the world, housing 17.7% of the total world population. 25 Hence, the epidemiological data from the Indian subcontinent can give a notable input to the global picture. The lack of a culturally sensitive tool is one of the major obstacles to such an epidemiological contribution from India. Therefore, the index study was planned with the objective of translating, adapting, and validating the original WBIS-M in Hindi appropriate for epidemiological purposes in India.
Materials and Methods
This study was conducted at the Metabolic Surgery Clinic of a tertiary care teaching hospital in North India. The ethical approval was sought (DRC256/16/3/2019 of the institution). The study followed an analytical, observational, and cross-sectional design. The study participants were enrolled using the purposive sampling technique over three months after obtaining written informed consent.
A rule of an item-to-sample ratio of 1:10 has been suggested for instrument validation. 26 As the WBIS-M has 11 items, a sample size of 110 participants was required.
Using the body mass index (BMI) as a screening tool, an individual with a BMI of >35.5kg/m2 with co-morbidities or BMI >40 kg/m2 with or without co-morbidities was referred to as class-3 obesity or severe obesity. 27
Individuals with severe obesity (n = 55) seeking treatment from the Metabolic Surgery Clinic and aged 18–60 years were recruited. All the participants were bilingual and literate in Hindi and English. An additional 120 medical undergraduate students were recruited for testing test-retest reliability. All the participants (n = 175) were mustered in for cross-cultural validation of the instrument. Special preventive measures for students, including exploitation, providing any credit or rewards, voluntary written informed consent with freedom and without any coercion, and avoidance of retaliation, were strictly practiced during the entire study process. Two authors who belonged from different hospitals other than medical students were involved in the data collection, coding, and analysis. Anonymity and confidentiality were strictly maintained. The rest of the two researchers were unaware of the setting of the student data collection.
Method of Validation and Cultural Adaptation
We followed the World Health Organization (WHO) process of translation, cultural adaptation, and validation of instruments. 28
We began with the translation of the original English WBIS-M into Hindi. A bilingual expert panel was composed for this purpose. It included three psychiatrists whose mother tongue was Hindi. Among these, two were acquainted with Western (English) culture. All were literate in English. The translation focused on the abstract meaning of words or phrases in Indian culture rather than the simple translation. The use of fewer words for the lucidity of the broader population was encouraged. This step helped us frame a simple, crisp, and concise construct.
The second step concentrated on the content validity of the scale using the expert judgment method. Validity is affected by cultural differences in society. Keeping this in mind, a group of experts was collated cautiously. The objective of the second step was to recognize and resolve inadequate interpretations of translation or discrepancies. The expert panel included four psychiatrists (two of them being the original translators, one of them has experience in instrument development and translation), two laparoscopic surgeons with 11 years of experience in metabolic surgeries), one English Professor (Ph.D. in English), one individual with severe obesity (BMI of 35.83 kg/m2), with Band 8.5 score in the International English Language Testing System (IELTS) and well-versed with the western culture. The mother tongue of all the experts was Hindi. Such an expert with severe obesity was deliberately selected to perceive the instrument from the perspective of the target population.
The third step involved a back translation into English by a different translator. The expert translator’s mother tongue was English, and this expert was uninformed about the original WBIS-M. Back translation helped in the comparability of the language and similarity of interpretability. An emphasis was laid on the conceptual equivalence, or the meaning conveyed more than the language. Some challenging words or phrases like “anxious” in item 3, “major way that I judge my value” in item 7, “really fulfilling social life” in item 8, “true-self” in item 10, and “date” in item 11 were notably difficult to adapt. These were resolved with the expert panel and subsequently in cognitive interviewing (details in the Table in the supplementary file).
The Hindi instrument so prepared was discussed with two groups. There were two groups of eight individuals with severe obesity. An in-depth, cognitive interviewing of participants was conducted. The concurrent probing technique was used. Their understanding, interpretation, and expression of every item were discussed. Some targeted probes for challenging words identified in the steps above were specifically discussed. They were inquired about their outlook on the items. Participants were requested to reiterate the items. Their selection basis for a response to an item was evaluated. In the end, respondents were specifically appealed if they found any word impermissible or inappropriate. An experienced interviewer performed the interview using culturally appropriate words or terms. Their suggestions were discussed with the expert panel and incorporated (details in the supplementary file). Thus, the final Hindi WBIS-M was formed.
Figure 1. Algorithm for Hindi Translation, Validation, and Cultural Adaptation of Hindi. Modified Weight Bias Internalization Scale (WBIS-M).
@EDDS: Eating Disorder Diagnostic Screen; #FPS-SF: Fat Phobia Scale-Short Form; WBIS-M: Modified Weight Bias Internalization Scale.
Test–retest reliability was assessed by administering the so-designed Hindi WBIS-M twice a week apart.
For the final step, the individuals with severe obesity (target population) (n = 55) (different from participants of cognitive interview) and undergraduate medical students (n = 120) agreed to participate in the study. Every participant provided written informed consent, and sociodemographic details were recorded. The sample belonged to either gender, 18–60 years of age, and from different socio-economic strata. The final Hindi and English WBIS-M were administered to analyze cross-language concordance. The Eating Disorder Diagnostic Screen (EDDS) and Fat Phobia Scale- Short Form (FPS-SF) were also administered. The relationship of the WBIS-M with the EDDS and FPS-SF (extant eating pathology scales) helped to analyze the criterion validity of the Hindi WBIS-M.
Tools
Sociodemographic profile sheet: A semi-structured questionnaire was developed to record certain variables regarding the patient and students, such as age, sex, total family income, family type, residence, marital status, religion, and locality.
Clinical Profile Sheet: A clinical profile sheet was designed to record the clinical details of the patients, such as symptoms, examination findings, past history, family history, treatment history, current medical condition, medications, and investigations.
Eating Disorder Diagnostic Scale (EDDS): The EDDS is a concise subjective instrument consisting of 22 items for diagnosing anorexia nervosa, bulimia nervosa, and binge-eating disorder. It has good reliability (mean κ = 0.80), criterion validity (mean κ = 0.83), test-retest reliability (r = 0.87), and internal consistency (mean α = 0.89). 29
FPS-SF: The FPS-SF has 14 pairing adjectives in a five-point Likert scale, which records prejudices related to people with obesity. 30 Some items were inverted. The higher the score, the greater the fat phobia. Both EDDS (citation count = 10600 and FPS-SF (citation count = 342) have been used extensively in research.29-34
Analysis of Psychometric Properties of Hindi WBIS-M
For dimension reduction, the principal component analysis using the varimax (orthogonal) rotation technique was performed. The sample size suitability was assessed by the Kaiser–Meyer–Olkin (KMO) test. Bartlett’s test of sphericity examined the level of sample size adequacy. A scree plot was examined. Factor loadings, corresponding Eigenvalue, and their respective variance were analyzed. The Kaiser-Guttman rule was applied to ascertain the optimal number of factors. The temporal stability for reliability was scrutinized. The validity of the so-developed Hindi WBIS-M was explored. The Statistical Package for Social Sciences (SPSS)-20 was employed for analysis.
Results
Qualitative Aspects of Hindi-WBIS-M During Adaptation
Modifications were made to the Hindi version, developed after forward–backward translation and expert panel review to ensure linguistic and cultural nuances were accurately represented when adapting the original WBIS-M into Hindi from an Indian perspective. With their deep understanding of the language and culture, the expert panel recommended changes that were later affirmed through cognitive interviewing.
In item 1, “mere” was initially chosen for “I” during forward translation. However, the panel recommended “apne,” a more culturally resonant term, which was later validated through cognitive interviewing. The panel also advised incorporating “mahsus” to preserve the concept of “feel,” which was absent during the backward–forward translation. During cognitive interviewing, respondents opined that retaining “mahsus” helps to express the emotions related to one’s competency and weight.
In item 2, the expert panel revised “mei apni vazan” with “apne vazan” and “zyadatar anya logo” with “kafi logo,” which were validated during cognitive interviewing as appropriate adaptations.
In item 3, the word “anxious” was represented by “distraught” and “wary” in back-translation. However, the expert panel opined that cultural conceptual equivalence for “anxious” in Hindi is difficult, and hence “vyakul” was preserved. It was approved during cognitive interviewing, too.
During cognitive interviewing, individuals suggested replacing “zabardast” with “ekdum” in item 4. They reasoned that “ekdum,” which is culturally more appropriate in this item as “ekdum” is more commonly used and popular in India to describe a change in one’s weight.
In item 7, “value” was initially translated as “mulya.” However, the panel and cognitive interviewing with individuals with severe obesity recommended “mahatvta” for better cultural resonance. Additionally, “aankta/aankti hun” was altered to “tay karta/karti hun” based on expert feedback, which was approved during cognitive interviewing. During back-translation, “value” came forth as “worth.” Keeping the cultural equivalence of these words from Indian perspectives, “mahatva” was retained in the Hindi scale.
Item 8 was one of the most challenging components to adapt. The portion “a really fulfilling social life” in this item made it particularly difficult to translate and adapt, highlighting the intricacies of the adaptation process. In the initial steps, it was translated to “sampurna samajik jeevan.” However, the expert panel and cognitive interviewing proposed it as “sahi mayane me samajik jeevan.”
For item 9, during back-translation, the word “ok” was represented by “fine.” Seeing the semantic similarity, “thik hai” was preserved.
Item 10 underwent modifications; “asal rup,” being a more popular and simpler word, was preferred over “vastavik swayam” after an expert panel review and cognitive interviewing.
The expert panel shortened the phrase “purush ya mahila” to “vyakti” in item 11 to maintain its meaning and conciseness from an Indian perspective. An additional challenge in this item was posed by the word “date.” During cognitive interviewing, participants suggested that “date” be added with “romance” as it will tap the current usage trends in metropolitan cities and rural populations. It was propounded that this addition would also make the item more understandable for all age groups from Indian cultural perspectives.
Items 3, 5, and 6 were similar to forward–backward translations in the final Hindi WBIS-M. A tabular form has been submitted in the supplementary file.
The bilingual literate students and patients participated in the study. Table 1 shows their sociodemographic, anthropometric, and clinical profiles.
Table 1.
Sociodemographic, Anthropometric, and Clinical Profile of the Participants.
| Sociodemographic variables | Mean (SD) (range) (n = 120) Student participants |
Mean (SD) (range) (n = 55) Patient participants |
| Age (in years) | 18.69(0.75) (range: 17–21) (median: 19) | 43.38 (7.93) (range: 30–60) (median 45) |
| Gender Male Female |
74 (61.7%) 46 (38.3%) |
20 (36.4%) 35 (63.6%) |
| Family income (in rupees per month) | 1,23,100 (232611.34) (5500–2500000) | 31472.73 (21796.07) (8000–100000) |
| Education of the participant Matric Intermediate Graduate |
(0) 0% 120 (100%) 0 (0%) |
13 (23.6%) 9 (16.4%) 33 (60%) |
| Socioeconomic status (as per Kuppuswamy SES scale) Upper-lower: 5–10 Lower-middle: 11–15 Upper-middle: 16–25 Upper >25 |
4 (3.3%) 0 (0%) 90 (75%) 26 (21.7%) |
0 (0%) 22 (40%) 33 (60%) 0 (0%) |
| Religion Hindu Non-Hindu |
117 (97.5%) 3 (2.5%) |
31 (56.4%) 24 (43.6%) |
| Locality Urban Rural |
119 (99.2%) 1 (0.8%) |
55 (100%) 0 (0%) |
| Co-morbid medical illness Yes No |
1 (0.8%) 119 (99.2%) |
34 (61.8%)# 21 (38.2%) |
| Height | 152 (9.00) (152.00–189.00) | 158.34 (11.32) (139.70–180.00) |
| Weight (in kg) | 62.929 (11.88) (43–98) | 108.34 (18.27) (77–170) |
| BMI (in kg/m2) | 21.77 (3.19) (14.02–34.37) | 43.21 (5.53) (35.76-68.53) |
#18 patients (32.7%) had hypertension, 16 (29.1%) had Obstructive sleep apnea, 15 (27.3%) had diabetes mellitus, and 16 (29.1%) had hypothyroidism.
Psychometric Properties of the Hindi WBIS-M Scale
The sample adequacy for exploratory factor analysis was checked. KMO test of sample adequacy across the participants (n = 175) was 0.788; Bartlett’s test of sphericity (approx. χ2 = 1381.892; p = 0.000) was statistically significant. This denotes that the sample size was adequate.
The exploratory factor analysis using the principal component method extracted three components. The initial Eigenvalue for factor-1, factor-2, and factor-3 were 5.663 (variance 51.48%), 1.268 (variance 11.53%), and 1.152 (variance 10.47%), respectively. To simplify the loading of items after principal component analysis, the varimax (orthogonal) rotation technique with Kaiser normalization was used. Three factors emerged. The rotation of sums of squared loadings for factors 1–3 were 4.783, 1.996, and 1.303, with variances of 43.49, 18.15, and 11.85%, respectively (tables for sample adequacy and factor analysis in the supplementary file).
As displayed in Figure 2, the scree-plot showed significant, dramatic tailing after factor 1. Hence, an additional factor would add relatively little to the information already extracted. Therefore, common factor analysis and domain classification were not done for the Hindi version of the scale.
Figure 2. A Scree Plot Showing Principal Component Numbers on the X-axis and Eigenvalues on the Y.
A scree plot showing principal component numbers in X-axis and eigenvalues in Y-axis. The graph has dramatic tailing after factor-1, indicating that other factors will contribute little to the magnitude of available information.
Table 2.
Test–Retest Reliability and Internal Consistency of Hindi WBIS-M Among Student Participants.
| Item | Mean baseline assessment (SD) (n = 120) | Mean assessment time 2 (SD) (n = 120) | Intraclass correlation coefficient | 95% CI | Pearson correlation (p-value) |
| WBIS-M 1 | 4.417(1.886) | 4.558(1.842) | 0.958 | 0.940–0.971 | 0.922*** |
| WBIS-M 2 | 2.876(1.938) | 2.983(1.896) | 0.869 | 0.811–0.908 | 0.768*** |
| WBIS-M 3 | 3.367(2.260) | 3.217(2.151) | 0.944 | 0.920–0.961 | 0.896*** |
| WBIS-M 4 | 3.100(2.124) | 3.108(2.094) | 0.900 | 0.857–0.931 | 0.818*** |
| WBIS-M 5 | 2.758(2.158) | 2.842(2.102) | 0.950 | 0.928–0.965 | 0.905*** |
| WBIS-M 6 | 2.100(1.946) | 1.892(1.748) | 0.892 | 0.844–0.925 | 0.813*** |
| WBIS-M 7 | 2.142(1.807) | 2.133(1.815) | 0.989 | 0.984–0.992 | 0.978*** |
| WBIS-M 8 | 2.192(2.276) | 2.067(2.164) | 0.902 | 0.860–0.932 | 0.823*** |
| WBIS-M 9 | 3.192(1.162) | 3.258(1.220) | 0.976 | 0.965–0.983 | 0.955*** |
| WBIS-M 10 | 2.708(2.239) | 2.508(2.126) | 0.921 | 0.886–0.945 | 0.857*** |
| WBIS-M 11 | 2.758(2.348) | 2.592(2.281) | 0.914 | 0.877–0.940 | 0.843*** |
***Indicates p-value <0.001.
WBIS-M: Modified Weight Bias Internalization Scale.
Reliability Statistics
Split-half reliability: The split-half reliability analysis was done by dividing the Hindi WBIS-M into two halves, consisting of six items and five items, in each half. Cronbach’s alpha for part 1 and part 2 were 0.689 and 0.701, respectively. Spearman-Brown Coefficient for Hindi WBIS-M in both the equal and unequal lengths was 0.793 and 0.794, respectively. Guttman’s Split-Half coefficient was 0.787.
Internal consistency: The internal consistency of 11 items was assessed by Cronbach’s alpha (α), utilizing an absolute agreement type, two-way mixed model, and α = 0.898.
Intraclass correlation coefficient: The two-way random effects model and principal component analysis showed a statistically significant intraclass correlation coefficient of 0.770 (p-value <0.001; 95% confidence interval [CI]: 0.704–0.823).
Test–retest reliability: Weight bias and internalized stigma have adverse consequences that are not restricted to the student’s health. It may compromise the care received by people with obesity and overweight and discrimination in future practice, for example, postponing or termination of consultations, refraining preventive care, and lower continuity of care, poor health outcomes, and quality of life.35-40 Thus, medical students were approached for test–retest reliability. Such sample characteristics encouraged caution regarding the items’ form, wordings, and answers, and the item was constructed to prompt and reflect the lived experiences of weight bias internalization stigma. 41
The scale was given to 120 students a week apart. Table 3 shows the one-week test–retest reliability of the Hindi version of the diagnostic scale. The intraclass correlation coefficient was 0.857 [95% CI (0.812–0.893)] (p-value <0.001) and mean Cronbach’s alpha α = 0.875.
Table 3.
Cross-Cultural Reliability of WBIS-M.
| Item | Hindi Mean (SD) (n = 175) |
English Mean (SD) (n = 175) |
Pearson/Spearman correlation coefficient | Intraclass coefficient$ | 95% CI |
| WBIS-M 1 | 4.474(1.838) | 4.343(1.893) | 0.943*** | 0.969*** | 0.919–0.956 |
| WBIS-M 2 | 2.846(1.871) | 2.743(1.884) | 0.825*** | 0.904*** | 0.871–0.929 |
| WBIS-M 3 | 3.300(2.214) | 3.366(2.262) | 0.924*** | 0.960*** | 0.946–0.970 |
| WBIS-M 4 | 3.291(2.057) | 3.149(2.054) | 0.781*** | 0.877*** | 0.834–0.908 |
| WBIS-M 5 | 2.611(1.991) | 2.520(2.025) | 0.920*** | 0.958*** | 0.943–0.969 |
| WBIS-M 6 | 1.817(1.587) | 1.926(1.722) | 0.805*** | 0.890*** | 0.851–0.918 |
| WBIS-M 7 | 1.914(1.579) | 1.834(1.587) | 0.965*** | 0.981*** | 0.975–0.986 |
| WBIS-M 8 | 1.880(1.904) | 1.914(2.011) | 0.837*** | 0.911*** | 0.880–0.934 |
| WBIS-M 9 | 3.886(1.608) | 3.851(1.630) | 0.972*** | 0.986*** | 0.981–0.989 |
| WBIS-M 10 | 2.240(1.924) | 2.309(2.045) | 0.869*** | 0.929*** | 0.905–0.948 |
| WBIS-M 11 | 2.411(2.118) | 2.474(2.199) | 0.868*** | 0.929*** | 0.905–0.948 |
***Indicates p-value <0.001.
WBIS-M: Modified Weight Bias Internalization Scale.
Cross-language concordance: One-hundred twenty students and 55 individuals with severe obesity attending the Metabolic Surgery Clinic consented to participate for this purpose. As displayed in Table 3, the Pearson (for parametric distribution) and Spearman correlation (nonparametric distribution) were statistically significant for all items. The intraclass correlation coefficient was significant for every item, varying from 0.877 to 0.986 (p-value = 0.000).
Table 4.
Correlation Between Continuous Items of the Hindi Version of WBIS-M and Continuous Variables of EDDS.
| WBIS-M1 | WBIS-M2 | WBIS-M3 | WBIS-M4 | WBIS-M5 | WBIS-M6 | WBIS-M7 | WBIS-M8# | WBIS-M9 | WBIS-M10 | WBIS-M11 | |
| ED1 | –0.238** | 0.158* | –0.244** | 0.413** | 0.240** | 0.264** | 0.176* | 0.231** | |||
| ED2 | –0.208** | 0.187* | 0.445** | 0.239** | 0.265** | 0.198** | 0.183* | ||||
| ED3 | –0.251** | 0.342** | 0.185* | 0.322** | 0.186* | 0.198** | 0.170* | 0.200** | |||
| ED4 | –0.227** | 0.336** | 0.168* | 0.293** | 0.157* | 0.290** | 0.203** | ||||
| ED7# | –0.300** | 0.240** | 0.346* | 0.155* | 0.320* | 0.244** | 0.258** | ||||
| ED8# | –0.331** | 0.216** | 0.361** | 0.272** | 0.188 | 0.165 | 0.236** | ||||
| ED15# | –0.202** | 0.187* | 0.202** | 0.252** | |||||||
| ED16# | –0.181* | –0.224** | –0.223** | 0.176* | –0.158* | ||||||
| ED17# | –0.388** | –0.216** | 0.252** | 0.259** | 0.193* | ||||||
| ED18# | –0.379** | –0.258** | 0.285** | 0.198** | 0.176* | ||||||
| ED19 | –0.151* | 0.515** | –0.240** | ||||||||
| ED20 | 0.171* | –0.191* | |||||||||
| ED21# | –0.159* | 0.273** |
Pearson correlation was calculated to examine the association between various items of EDDS with WBIS-M.
#Spearman correlation was calculated for data with nonparametric distribution.
ED: Eating Disorder Diagnostic Screen; WBIS-M: Modified Weight Bias Internalization Scale.
*p-value <0.05; **p-value <0.01, ***p-value <0.001.
Correlation with Extant Eating Pathology Scales
As shown in Table, item 1 (deals with one’s feeling of competency due to weight as compared to others) of the WBIS-M scale showed a significant negative correlation with items 1–4 (deals with feeling of being fat, fear of gaining weight, judging oneself as a person due to weight or shape) 7–8 (deals with an unusually large amount of food and experienced a loss of control in last six months and three months), 15, 17 and, 18 of EDDS (deals with average frequency in a week in past three months, of inducing vomiting or fasting to avoid weight gain or counter the effects of eating, respectively).
Item-11 (deals with one’s thought arising due to weight with regards to dating desire by someone attractive) of the WBIS-M scale showed a significant positive correlation with items 1–8 (except item-6) of EDDS. Item-2 (deals with cognition of one’s attractiveness due to weight as compared to others) I am less attractive than most other people because of my weight) of WBIS-M showed a significant positive correlation with items 1–6 of EDDS. Similarly, various other items of WBIS-M had a significant correlation with various items of EDDS.
Several dichotomous items of EDDS had a significant association with WBIS-M. Both items 5 and 6 of EDDS (during the past six months, have there been times when you felt you have eaten what other people would regard as an unusually large amount of food given the circumstances? When you ate an unusually large amount of food, did you experience a loss of control) showed a statistically significant correlation with items 1, 2, 3, 4, 7, 10, and 11 (items related to competency, attractiveness, people’s notion about weight, wish to change weight drastically, judging the value as a person and expression of “dating” by someone). Similarly, items 9, 10, 11, and 12 of EDDS (eat much more rapidly than normal?; eat until you felt uncomfortably full?; and eat large amounts of food when you did not feel physically hungry?) had a significant association with multiple items of WBIS-M (Table in supplementary file).
As shown in Table 5, a significant correlation between several items of Hindi WBIS-M and various items of FPS-SF surfaced. Except for items 1, 2, and 4, all the items of WBIS-M revealed a significant positive correlation with item 8 of FPS-SF (weak-strong). Item-3 (I feel anxious about my weight because of what people might think of me) of WBIS-M had a significant positive correlation with items-1 (lazy-industrious), 4 (good self-control-poor self-control), 5 (fast-slow), 7 (active-inactive), 8 (weak-strong), 10 (dislikes food-likes food), and 14 (low self-esteem- high self-esteem) of FPS-SF. However, item-3 negatively correlated with item-11 (shapeless-shapely) of FPS-SF.
Table 5.
Correlation Between Continuous Items of the Hindi Version of WBIS-M and FPS.
| WBIS-M1 | WBIS-M2 | WBIS-M3 | WBIS-M4 | WBIS-M5 | WBIS-M6 | WBIS-M7 | WBIS-M8# | WBIS-M9 | WBIS-M10 | WBIS-M11 | |
| FPS1 | 0.157* | 0.338*** | –0.167* | –0.283*** | 0.180* | ||||||
| FPS2 | –0.365*** | –0.346*** | –0.191* | 0.202** | –0.150* | ||||||
| FPS3 | –0.210** | –0.335** | –0.186* | –0.320* | |||||||
| FPS4 | 0.239** | 0.389*** | |||||||||
| FPS5 | 0.346*** | –0.173* | –0.240** | ||||||||
| FPS6 | –0.198** | 0.350** | 0.243** | ||||||||
| FPS7 | 0.433*** | –0.163* | –0.170* | ||||||||
| FPS8 | 0.180* | –0.364*** | –0.158* | –0.316*** | –0.228* | 0.185* | –0.224** | –0.356*** | |||
| FPS9 | –0.196** | –0.192* | 0.553*** | –0.224** | –0.153* | ||||||
| FPS10 | 0.275*** | 0.201** | 0.380** | –0.254** | |||||||
| FPS11 | –0.318*** | 0.248** | 0.354** | 0.294** | 0.190* | ||||||
| FPS12 | 0.192* | 0.185* | |||||||||
| FPS14 | –0.164* | 0.157* | 0.349*** | 0.218** | 0.198** |
Pearson correlation was calculated to examine the association between various items of FPS-SF with WBIS-M.
#Spearman correlation calculated for variables with nonparametric distribution.
WBIS-M: Modified Weight Bias Internalization Scale.
EDDS: Eating Disorder Diagnostic Screen; FPS-SF: Fat Phobia Scale-Short Form.
*p-value <0.05; **p-value <0.01, ***p-value. <0.001.
Discussion
The study aimed to translate, adapt, and validate WBIS-M for the Indian population. The temporal reliability of Hindi WBIS-M was measured by test–retest reliability, which was acceptable to excellent. The internal consistency was measured using split-half reliability and Cronbach’s alpha methods, which was acceptable. The cross-cultural reliability showed significant Pearson/Spearman correlation, intraclass correlation, and narrow confidence interval.
The expert’s judgment measured the content or curricular validity. For face validity, meticulous and comprehensive pre-testing, including focused-group discussions and cognitive interviews with individuals with severe obesity, was conducted (details in the supplementary file).
For demonstrating criterion validity, extant eating pathology scales EDDS and FPS-SF were used. Most of the items of EDDS and FPS-SF revealed a statistically significant relationship with various items of WBIS-M. Therefore, the criterion validity of the so-designed Hindi WBIS-M is good.
The internal consistency of Pearl’s WBIS was α = 0.94 and had a significant correlation with extant eating pathology scales. 3 The internal consistency and test–retest reliability for Chinese WBIS-M, Cronbach’s alpha, α = 0.79 and r = 0.81, p ≤ 0.001, respectively. 19 The Cronbach’s alpha for German, Turkish, and Spanish versions of WBIS-M are α = 0.87, 0.92, and 0.93, respectively.20-23,42 The intraclass correlation coefficient for Hindi-WBIS-M was 0.857 [95% CI (0.812–0.893)](p-value <0.001), mean Cronbach’s alpha α = 0.875, indicating a high test- retest reliability. 43 The internal consistency for the final Hindi-WBIS-M is comparable to other versions of WBIS-M.
Although confirmatory factor analysis was not done in our study, most of the similar studies involving validation and cultural adaptation of WBIS-M have shown a one-factor structure of WBIS-M after principal component and confirmatory factor analysis.20-23
India is a culturally diverse country. This makes the adaptation and validation of a culturally sensitive standardized tool challenging. For example, in the fourth step of translation and validation, item 11 of WBIS-M (Because of my weight, I do not understand how anyone attractive would want to date me) was found to be the most challenging to translate in Hindi. This may be due to cultural differences in the concept of dating in India as compared to the West.
Nevertheless, the reliability and validity of Hindi WBIS-M are acceptable for use among the Indian population. A plausible explanation may be the sensitive study design and development of the Hindi WBIS-M through a rigorous process. The study setting was the Metabolic Surgery Clinic. The experts for content validity of the construct were chosen cautiously, including an individual with severe obesity, well-versed with Western and Indian culture, proficient in the English language, and had Hindi mother-tongue; two bariatric-laparoscopic surgeons with 11 years of experience in the field. The participants for the correction and interpretation during the pre-testing step of scale development were the target population, i.e., individuals with severe obesity. Medical students were included for test–retest reliability. Those medical undergraduates who perceive negative attitudes toward individuals who are overweight or obese as a norm while pursuing education often have unpleasant patient-related conduct, including a poorer degree of respectfulness, responsiveness, interpersonal communication, and attention. 39 Due to “globesity,” the majority of patients these next-generation healthcare professionals will care for might be overweight or obese.12,44 The culturally sensitive stem words and emotional content of the items of Hindi WBIS-M were perceived and intercepted well.
Future research using H-WBIS-M might help in understanding the weight bias and its implications among Indian medical students. As advocated by previous researchers, the results may guide the concerned authorities for the inclusion of weight bias mitigation strategies in the Indian medical curricula and body diversity in clinical care. 35
The study has certain limitations. The sample size is small, and the patient population could also have been included in the test–retest reliability for better generalization. Due to the lack of validated and culturally adapted similar tools in India, the original EDDS and FPS-SF were used to check the criterion validity of Hindi WBIS-M.
Conclusions
The Hindi version of WBIS-M has good psychometric properties based on standard tests. The scale may be used in the Indian population for epidemiological purposes.
Supplemental Material
Supplementary material for this article is available online.
Supplementary material for this article is available online.
Supplementary material for this article is available online.
Acknowledgments
We are thankful to Rebecca L. Pearl, Department of Psychology at Yale University, for her kind permission to use the original WBIS-M in our study. We thank the experts involved in the translation and cultural adaptation of the instrument. We used the STROBE cross sectional checklist when writing our report. 1
Footnotes
The authors declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.
Ethical Approval: Ethical approval was obtained vide DRC256/16/3/2019 of the institution. The same is mentioned in the first paragraph of the manuscript under the Materials & Methods section.
Funding: The authors received no financial support for the research, authorship and/or publication of this article.
Informed Consent: Written informed consent was obtained from all participants.
Use of Generative Artificial Intelligence: Nil.
References
- 1.LaCroix E, Alberga A, Russell-Mathew S, et al. Weight bias: Systematic review of characteristics and psychometric properties of self-report questionnaires. Obes Facts, 2017; 10: 223–237. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Puhl RM and Heuer CA. The stigma of obesity: A review and update. Obesity, 2009; 20: 941–964. [DOI] [PubMed] [Google Scholar]
- 3.Pearl RL, Puhl RM, Lessard LM, et al. Prevalence and correlates of weight bias internalization in weight management: A multinational study. SSM Popul Health, 2021; 13: 100755. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Romano KA, Heron KE, Sandoval CM, et al. A meta-analysis of associations between weight bias internalization and conceptually-related correlates: A step towards improving construct validity. Clin Psychol Rev, 2022; 92: 102127. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Durso LE and Latner JD. Understanding self-directed stigma: development of the weight bias internalization scale. Obesity, 2008; 16(SUPPL. 2): S80–86. [DOI] [PubMed] [Google Scholar]
- 6.Corrigan PW, Watson AC and Barr L.. The self–stigma of mental illness: Implications for self–esteem and self–efficacy. J Soc Clin Psychol, 2006; 25: 875–884. [Google Scholar]
- 7.Pearl RL and Puhl RM. Weight bias internalization and health: A systematic review. Obes Rev, 2018; 19: 1141–1163. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Puhl RM, Himmelstein MS and Quinn DM. Internalizing weight stigma: prevalence and sociodemographic considerations in US adults. Obesity, 2018; 26: 167–175. [DOI] [PubMed] [Google Scholar]
- 9.Zuba A and Warschburger P.. The role of weight teasing and weight bias internalization in psychological functioning: A prospective study among school-aged children. Eur Child Adolesc Psychiatry, 2017; 26: 1245–155. [DOI] [PubMed] [Google Scholar]
- 10.Tomiyama AJ. Weight stigma is stressful. A review of evidence for the cyclic Obesity/weight-based stigma model. Appetite, 2014; 82: 8–15. [DOI] [PubMed] [Google Scholar]
- 11.Rains M. Science News/ Health. The origins of the obesity epidemic may be further back than we thought [Internet]. 2023, Sep, 13. [cited Sep 16, 2023]: Available from: https://www.science.org/content/article/origins-obesity-epidemic-may-be-further-back-we-thought
- 12.World Health Organization. Controlling the global obesity epidemic. Available from: https://www.who.int/activities/controlling-the-global-obesity-epidemic (2023, accessed Sep 16, 2023).
- 13.Ng M, Fleming T, Robinson M, et al. Global, regional, and national prevalence of overweight and obesity in children and adults during 1980–2013: A systematic analysis for the Global Burden of Disease Study 2013. Lancet, 2014; 384: 766–781. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Lillis J, Luoma JB, Levin ME, et al. Measuring weight self-stigma: the weight self-stigma questionnaire. Obesity (Silver Spring) 2010; 18: 971–976. [DOI] [PubMed] [Google Scholar]
- 15.Pearl RL and Puhl RM. Measuring internalized weight attitudes across body weight categories: Validation of the Modified Weight Bias Internalization Scale. Body Image, 2014; 11: 89–92. [DOI] [PubMed] [Google Scholar]
- 16.Citations for Understanding self-directed stigma: development of the weight bias internalization scale, https://scholar.google.co.in/scholar?hl=en&as_sdt=0%2C5&q=Understanding+self-directed+stigma%3A+development+of+the+weight+bias+internalization+scale&btnG= (2023, accessed Sep 15 2023. ). [DOI] [PubMed]
- 17.Citations for Measuring weight self-stigma: the weight self-stigma questionnaire. https://scholar.google.co.in/scholar?hl=en&as_sdt=0%2C5&q=weight+self-stigma+questionnaire+&btnG= (2023, accessed Sep 15, 2023).
- 18.Citations for Measuring internalized weight attitudes across bodyweight categories: Validation of the Modified Weight Bias Internalization Scale. https://scholar.google.co.in/scholar?hl=en&as_sdt=0%2C5&q=Measuring+internalized+weight+attitudes+across+body+weight+categories%3A+Validation+of+the+Modified+Weight+Bias+Internalization+Scale.+&btnG= (2023, accessed Sep 15 2023. ). [DOI] [PubMed]
- 19.Chen H and Ye Y.. Validation of the weight bias internalization scale for mainland Chinese children and adolescents. Front Psychol, 2021; 11: 3684. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Ciupitu-Plath C, Wiegand S and Babitsch B.. The weight bias internalization scale for youth: Validation of a specific tool for assessing internalized weight bias among treatment-seeking German adolescents with overweight. J Pediatr Psychol, 2018; 43: 40–51. [DOI] [PubMed] [Google Scholar]
- 21.Macho S, Andrés A and Saldaña C. Validation of the modified weight bias internalization scale in a Spanish adult population. Clin Obes, 2021; 11: e12454. [DOI] [PubMed] [Google Scholar]
- 22.Argyrides M, Charalambous Z, Anastasiades E, et al. Translation and validation of the Greek version of the Modified Weight Bias Internalization Scale in an adult population. Clinical Obesity, 2022; 12: e12503. [DOI] [PubMed] [Google Scholar]
- 23.Apay SE, Yılmaz E, Aksoy M, et al. Validity and reliability study of modified weight bias internalization scale in Turkish. Int J Caring Sci, 2017; 10: 1530–1540. [Google Scholar]
- 24.DePierre JA and Puhl RM. Experiences of weight stigmatization: A review of self-report assessment measures. Obes Facts, 2012; 5: 897–918. [DOI] [PubMed] [Google Scholar]
- 25.India Population (2021) - Worldometer. [2023, accessed Sep 16, 2023]. Available from: https://www.worldometers.info/world-population/india-population/
- 26.Cattell RB. The scientific use of factor analysis in behavior and life sciences. Fixing the Number of Factors: The Most Practicable Psychometric Procedures. New York: Plenum, 1978, pp.72–91. [Google Scholar]
- 27.Defining Adult Overweight and Obesity. https://www.cdc.gov/obesity/basics/adult-defining.html (2023, accessed Sep 15, 2023).
- 28.Organization WHO. WHO | Process of translation and adaptation of instruments [Internet]. Who. 2010. [cited 2020. Oct 11]. Available from: https://www.who.int/substance_abuse/research_tools/translation/en/ [Google Scholar]
- 29.Stice E, Telch CF and Rizvi SL. Development and validation of the eating disorder diagnostic scale: a brief self-report measure of anorexia, bulimia, and binge-eating disorder. Psychol Assess, 2000; 12: 123–131. [DOI] [PubMed] [Google Scholar]
- 30.Bacon JG, Scheltema KE and Robinson BE. Fat phobia scale revisited: The short form. Int J Obes Relat Metab Disord, 2001; 25(2): 252–257. [DOI] [PubMed] [Google Scholar]
- 31.Krabbenborg MAM, Danner UN, Larsen JK, et al. The eating disorder diagnostic scale: Psychometric features within a clinical population and a cut-off point to differentiate clinical patients from healthy controls. Eur Eat Disord Rev, 2012; 20: 315–320. [DOI] [PubMed] [Google Scholar]
- 32.Santos IF, Machado PPP and Esteves F.. Development and validation of the Portuguese version of the Eating Disorder Diagnostic Scale. Ansiedad Estres, 2018; 24: 81–89. [Google Scholar]
- 33.Schaefer LM, Anderson LM, Simone M, et al. Gender-based differential item functioning in measures of eating pathology. Int J Eat Disord, 2019; 52: 1047–1051. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Elboim-Gabyzon M, Attar K and Peleg S.. Weight stigmatization among physical therapy students and registered physical therapists. Obes Facts, 2020; 13: 104–116. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Blumberg P and Mellis LP. Medical students’ attitudes toward the obese and the morbidly obese. Int J Eat Disord, 1985; 4(2): 169–175. [Google Scholar]
- 36.Puhl RM, Luedicke J and Grilo CM. Obesity bias in training: attitudes, beliefs, and observations among advanced trainees in professional health disciplines. Obesity (Silver Spring), 2014; 22(4): 1008–1015. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Olson CL, Schumaker HD and Yawn BP. Overweight women delay medical care. Arch Fam Med, 1994; 3(10): 888–892. [DOI] [PubMed] [Google Scholar]
- 38.Drury CAA and Louis M.. Exploring the association between body weight, stigma of obesity, and health care avoidance. J Am Assoc Nurse Pract, 2002; 14(12): 554–561. [DOI] [PubMed] [Google Scholar]
- 39.Phelan SM, Puhl RM, Burgess DJ, et al. The role of weight bias and role-modeling in medical students’ patient-centered communication with higher weight standardized patients. Patient Educ Couns, 2021; 104(8): 1962–1969. [DOI] [PubMed] [Google Scholar]
- 40.Puhl RM, Wharton C and Heuer C.. Weight bias among dietetics students: Implications for treatment practices. J Am Diet Assoc, 2009; 109(3): 438–444. [DOI] [PubMed] [Google Scholar]
- 41.Schinka JA, Velicer WF and Weiner IB. Handbook of psychology: Research methods in psychology. Vol. 2. Hoboken, NJ: John Wiley & Sons, Inc.; 2013. [Google Scholar]
- 42.Hilbert A, Baldofski S, Zenger M, et al. Weight bias internalization scale: Psychometric properties and population norms. PLoS One, 2014; 9: 1–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.Taber KS. The use of Cronbach’s alpha when developing and reporting research instruments in science education. Res Sci Educ, 2018; 48: 1273–1296. [Google Scholar]
- 44.Jayawickrama RS, O’Connor M, Flint SW, et al. Explicit and implicit weight bias among health care students: A cross-sectional study of 39 Australian universities. EClinicalMedicine, 2023; 58: 101894. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Supplementary material for this article is available online.
Supplementary material for this article is available online.
Supplementary material for this article is available online.