Eating and Weight Disorders - Studies on Anorexia, Bulimia and Obesity https://doi.org/10.1007/s40519-022-01453-z ORIGINAL ARTICLE Spanish validation of the Modified Weight Bias Internalization Scale (WBIS‑M) for adolescents Ana Andrés1 · Albert Fornieles‑Deu2,3 · Ana Rosa Sepúlveda4 · Lucía Beltrán‑Garrayo4 · Albert Montcada‑Ribera5 · Anna Bach‑Faig6,7 · David Sánchez‑Carracedo2,8 Received: 11 October 2021 / Accepted: 14 July 2022 © The Author(s) 2022 Abstract Purpose Weight Bias Internalization (WBI) is pervasive and potentially damaging for health. Little is known about WBI in youth. As negative effects of WBI have been observed when controlling for BMI, measures that allow WBI to be assessed across different weight categories are needed. The Modified Weight Bias Internalization Scale (WBIS-M) is one of the most frequently used scales in this field. Our purpose was to obtain a Spanish validated version of the WBIS-M for adolescents across different weight statuses. Methods The data were collected from 298 secondary students (mean age 14.31; 48.32% girls; 18.8% were overweight and 6.4% had obesity). Internal structure was examined by a cross-validation analysis, using both exploratory and confirmatory factor analyses in different subsamples. Results Item 1 showed a psychometric anomalous functioning and was deleted. The one-factor structure of the 10-item version was confirmed with adequate fit ([EFA (KMO = 0.915, χ2(55) = 1075.633, p < 0.0001)]; [CFA (χ2(35) = 200.515; GFI = 0.995; PGFI = 0.992; NFI = 0.991; SRMR = 0.060)]). Internal consistency was high (𝛼 = 0.93; ω = 0.93). Significant correlations with the same set of external variables assessed in the original version (anti-fat bias, self-esteem, mood, body dissatisfaction, drive for thinness, binge eating), all of them correlates of WBI in adolescents, were found. Girls and partici- pants with obesity obtained higher scores. Conclusion The results provide support for the validity and reliability of our WBIS-M version for use with adolescents across weight categories in Spain. Level of evidence Level III, evidence obtained from well-designed cohort studies. Keywords Confirmatory factor analysis · Exploratoy factor analysis · Weight Bias Internalization · Obesity stigma · Adolescents 2 * David Sánchez‑Carracedo Eating and Weight‑Related Problems Unit, Universitat david.sanchez@uab.es Autònoma de Barcelona, 08193 Barcelona, Spain 3 Ana Andrés Department of Psychobiology and Methodology of Health anaav2@blanquerna.url.edu Sciences, Serra Hunter Fellow, Universitat Autònoma de Barcelona, 08193 Barcelona, Spain Albert Fornieles‑Deu 4 albert.fornieles@uab.cat Department of Biological and Health Psychology, School of Psychology, Autonomous University of Madrid, Ana Rosa Sepúlveda 28049 Madrid, Spain sepulveda.ar@gmail.com 5 Head of Community and Health Service, City Council Lucía Beltrán‑Garrayo of Terrassa, Social Rights Section, Barcelona, Spain lucia.beltran@uam.es 6 FoodLab Research Group (2017SGR 83, Faculty of Health Albert Montcada‑Ribera Sciences, Universitat Oberta de Catalunya (Open University Albert.Moncada@terrassa.cat of Catalonia, UOC), 08018 Barcelona, Spain Anna Bach‑Faig 7 Food and Nutrition Area, Barcelona Official College abachf@uoc.edu of Pharmacists, 08009 Barcelona, Spain 1 8 Faculty of Psychology, Education and Sport Sciences, Department of Clinical and Health Psychology, Universitat Blanquerna, Ramon Llull University, Barcelona, Spain Autònoma de Barcelona, 08193 Barcelona, Spain 13 Vol.:(0123456789) Eating and Weight Disorders - Studies on Anorexia, Bulimia and Obesity Introduction Several adaptations of the WBIS-M to other languages have been developed in the recent years. A Turkish validation was Obesity stigma, also known as weight stigma and weight conducted with college students [24]. Two Chinese versions bias, is pervasive in Western societies and has been defined for children and adolescents have been developed based on the as a globalizing health challenge [1]. Weight stigma samples from Hong Kong [25] and mainland China [26]. Two refers to social devaluation and denigration of individuals different German versions are also available. The Weight Bias because of their excess body weight, and can lead to nega- Internalization Scale for Children (WBIS-C), was developed tive attitudes, stereotypes, prejudice, and discrimination for children aged 9–13 years [27] and was based on the previ- [2]. Weight bias, when explicit, refers to overt, consciously ous German version of the WBIS [28]. A recent three-item held negative attitudes, while implicit weight bias refers short version (WBSI-3) has also been developed using a Ger- to automatic, negative attributions and stereotypes which man sample [29]. A Spanish version of the WBIS-M has been exist outside of conscious awareness [2]. developed with a sample of adults from the general population Weight stigma has been associated with psychological [30]. However, there are currently no instruments for assess- distress [3], poorer mental health [4], disordered eating and ing WBI in Spanish children and adolescents across different unhealthy eating behaviours [5], substance use [6], more weight statuses. This study was designed to cover this deficit. physiological stress [7], reduced motivation to engage in The specific aims were: (1) to study the internal structure of physical activity [8], additional barriers to weight loss main- our WBIS-M version using exploratory and confirmatory fac- tenance [9], reduced healthcare seeking behaviour [10], and torial analysis; (2) to study its internal consistency; (3) to study a higher risk for suicidality and mortality [11, 12]. the association between the WBIS-M score and the same set Because weight stigma negatively affects weight-related of external variables analysed in the original version [23] to behaviours and health, it can, paradoxically, contribute provide information about its validity; and (4) to study their to worsened problems associated with obesity and cre- relationships with gender and weight status. ate additional barriers to healthy behaviour change [13]. Therefore, weight stigma is considered a psychosocial con- Methods tributor to obesity [14]. A recent meta-analysis has sup- ported bidirectional relationships between weight stigma Participants and paediatric obesity [15]. Weight bias internalization (WBI) occurs when individ- This study is part of a funded project on weight bias in ado- uals engage in self-blame and self-directed weight stigma lescents carried out using a representative sample of the because of their weight [2]. WBI is also pervasive and city of Terrassa (Barcelona). Schools and classrooms were potentially damaging for health beyond body weight and selected using a multistage cluster sampling. The final sam- experiences of stigma [16, 17], leading to increased research ple of this study consisted of 298 secondary school students attention in recent years, especially in adults. In contrast, (18.75% of the expected sample), up to the point when data not much is known about WBI in youth [18]. Children and collection was stopped due to the COVID-19 lockdown. No adolescents can be particularly vulnerable to weight stigma, specific exclusion criteria were used and all students present with long-lasting consequences that negatively affect their at the time of the assessment with parental consent partici- life course and emotional and physical well being [6]. Thus, pated. Only participants who did not have parental consent, tools for assessing experienced and internalized weight bias refused to participate, did not respond to the parental con- in children and adolescents are needed as a first step to study sent request or were invalid (because of language issues or their negative impact, and as a measure outcome in future because did not pass the surveys controls) were excluded studies focused on how to address weight bias. from the original class lists. Figure 1 shows the flow diagram A systematic review [19] identified two questionnaires of the sample. designed to assess WBI in people who are overweight or The sample came from one public and four grant-aided have obesity: the Weight Self-Stigma Questionnaire (WSSQ) schools and was composed of students from the four years [20] and the Weight Bias Internalization Scale (WBIS) [21]. of Compulsory Secondary Education in the Spanish system. As negative effects of WBI have been observed when con- Table 1 shows the main characteristics of the sample. trolling for BMI, a call for measures that allow for assess- ment of weight stigmatization across different body weight Measures categories was made [22]. The Modified Weight Bias Inter- nalization Scale (WBIS-M; Pearl and Puhl [23]) was conse- Demographics quently developed to be applied across body weight statuses to assess the full impact of this problem. The Four-Factor Index of Social Status [31] was used to determine the socioeconomic status (SES) of the 13 Eating and Weight Disorders - Studies on Anorexia, Bulimia and Obesity household using a weighted average of each parent’s educational and occupational level. Total scores were categorized into five different levels, namely high, medium–high, medium, medium–low and low. Height in cm and weight in kg were measured using a SECA port- able stadiometer, model 214 (20–207 cm; accuracy range of 0.1 cm), and SECA portable scales, model 8777021094 (0–200 kg; accuracy range of 0.1 kg), respectively. Weight status was calculated using the World Health Organization growth reference criteria [32]. Participants also reported information about age, gender, and parental origin. Weight Bias Internalization The WBIS-M [23] was based on the original 11 items from Fig. 1  Participants’ flow diagram the WBIS [21]. Items that included the word “overweight” Table 1  Descriptions and Total (n = 298) Female (n = 144) Male (n = 154) t (p)* tests of significance for age, χ2 (p)** BMI, parental origin, weight status, DFT, Binge Eating, and Age M (SD) 14.31 (1.23) 14.32 (1.28) 14.31 (1.18) − 0.08 (.939) WBIS-M based on comparison BMI M (SD) 20.99 (3.49) 21.04 (3.64) 20.94 (3.35) − 0.26 (0.796) between genders Parental Origin n (%) Spanish 230 (77.2) 110 (76.4) 120 (77.9) 3.80 (0.578) European 3 (1) 2 (1.4) 1 (0.6) Latino 16 (5.4) 10 (6.9) 6 (3.9) North African 17 (5.7) 7 (4.9) 10 (6.5) Mixed origin 5 (1.7) 1 (0.7) 4 (2.6) Others 27 (9.1) 14 (9.7) 13 (8.4) Weight status n (%) Underweight 37 (12.4) 22 (15.3) 15 (9.7) 3.14 (0.370) Normal weight 186 (62.4) 90 (62.5) 96 (62.3) Overweight 56 (18.8) 25 (17.4) 31 (20.1) Obesity 19 (6.4) 7 (4.9) 12 (7.8) EDI-3 M (SD) DFT (0–28) 7.67 (6.83) 9.35 (8.12) 6.10 (4.87) − 4.20 (< 0.001) BINGE EATING n (%) (1) 6 (2) 4 (2.8) 2 (1.3) 1.05 (0.789) (2) 41 (13.8) 21 (14.6) 20 (13.0) (3) 56 (18.8) 27 (18.8) 29 (18.8) (4) 195 (65.4) 92 (63.9) 103 (66.9) WBIS-M M (SD) 2.22 (1.37) 2.57 (1.55) 1.90 (1.09) − 4.29 (< 0.001) (1) Binge eating + loss of control + at least once a week in the last three months + some or a lot of distress; (2) Binge eating + loss of control + at least once a month or a few times a month in the last three months, or a higher frequency but with no or little distress; (3) Binge eating + no loss of control + no distress; (4) No binge eating BMI Body Mass Index, M Mean, SD Standard Deviation, EDI-3 Eating Disorders Inventory-3, DFT Drive for Thinness subscale, BINGE EATING Binge eating severity using DSM-V criteria, WBIS-M Modified Weight Bias Internalization Scale * Degrees of freedom for t test = 296 ** Degrees of freedom for χ2: Parental Origin = 5; Weight status and Binge eating = 3 13 Eating and Weight Disorders - Studies on Anorexia, Bulimia and Obesity were replaced with phrases that instead used the words “my restrictive tendencies in eating and weight behaviours and weight”. Responses are rated on a 7-point Likert scale rang- cognitions. The internal consistency with Cronbach’s alpha ing from “Strongly Disagree” to “Strongly Agree”. Two in nonclinical samples was 0.44–0.95 and test–retest values items (1 and 9) are reverse-scored. The mean of the item were 0.85 to 0.99. In our sample, Cronbach’s alpha was 0.82. responses serves as the participant’s score (range 1–7), with The original development of the WBIS-M included higher scores indicating higher internalized weight bias. The two questions from the Eating Disorder Diagnostic Scale Cronbach’s alpha of the original WBIS-M was 0.94, showed (EDDS) [39] for assessing frequency of binge eating in strong construct validity, and presented a one-factor struc- the past three and 6 months. In our study, binge eating was ture. The translation process of the WBIS-M is described in measured using four previously established and validated the procedures section. questions used in studies focused on weight stigma in ado- We included the same set of measures as were used in lescents [18]. These questions were updated to align with the original versions of both WBIS and WBIS-M, but in Binge Eating Disorder criteria from DSM-V [40], assessing their Spanish versions, to test their construct and convergent the presence of binge eating in the past three months (yes/ validity. no), with or without loss of control (yes/no), frequency of binge eating with loss of control (4-point scale from “every Anti‑fat bias day” to “less than once per month”), and distress over binge eating (4-point scale from “not at all” to “a lot”). Based on The Dislike subscale of the Anti-fat Attitudes Question- the previous research [41], these items were combined to naire (AAQ-D) [33] in its Spanish version [34] was also determine a severity score on a 4-point ordinal scale. administered. It has 7 items with responses ranging from 1 (‘strongly disagree’) to 7 (‘strongly agree’). Cronbach’s Procedure alpha was among 0.70 and 0.86. In our sample, Cronbach’s alpha was 0.74. This study was supported by Terrassa City Council’s Health and Community Service. The Principal Investigator of the Self‑esteem and mood funded research project held meetings with the manage- ment teams of the schools to introduce the project, agree the The Spanish versions of the Rosenberg Self-Esteem Scale assessment scheduling, and obtain the lists of the enrolled (RSE)[35] and the Depression Anxiety Stress Scales (DASS- students. Parental consent was requested. The survey was 21)[36] were included. The Spanish version of the RSE con- designed in an online format on a specific platform of the sists of 10 items based on a 4-point scale from 1 (‘totally company Digital Insights, and included controls as response disagree’) to 4 (‘totally agree’). Cronbach’s alpha (0.85 to ranges and three interspersed control questions to test atten- 0.88) and test–retest correlation (0.84) was found to be satis- tion levels. Participants who did not pass the surveys con- factory. Cronbach’s alpha in our study was 0.88. In the Span- trols were excluded and their data were not registered. The ish version of the DASS-21, respondents evaluate the sever- missing data were avoided with this system. The survey ity/frequency with which they have experienced each of the was administered in classrooms during normal class hours, 21 negative emotional symptoms of depression, anxiety, and under the supervision of a group of previously trained psy- stress during the previous week on a scale from 0 to 3. Both chology master’s students and their regular teacher. Assent a three-factor model and a one-factor model were used. The from each participant was also requested. Groups of 5–7 discriminant validity was satisfactory and Cronbach’s alfa participants were taken to a private room to take height and values were 0.84, 0.70, and 0.82 for the Depression, Anxiety, weight measures, following a standardized protocol [42]. and Stress subscales, respectively. Cronbach’s alpha for the The participants returned to the classroom to continue with whole scale in our sample was 0.92. the survey. Confidentiality was guaranteed throughout the entire process, and data processing was based on anonymous Body image and eating disorder pathology data. The assessments were conducted between March 4–12, 2020. On March 13, the COVID-19 lockdown was decreed A short 10-item Spanish version of the Body Shape Ques- and assessments were stopped. This study was conducted tionnaire (BSQ) [37] and the Spanish version of the Drive in accordance with the Declaration of Helsinki and was for Thinness (DFT) subscale of the Eating Disorders Inven- approved by the Ethics Committee on Animal and Human tory-3 [38] were included. Scores in the BSQ range from 1 Experimentation at the Universitat Autònoma de Barcelona (‘never’) to 6 (‘always’). This version demonstrated metric (CEAAH 3451). invariance and was found to be more consistent than other The adaptation of the WBIS-M from English to Spanish short versions (Warren et al. [37]). In our sample, Cronbach’s was conducted following the International Test Commission alpha was 0.91. The DFT consists of 7 items that measure Guidelines [43]. Two translators whose native tongue was 13 Eating and Weight Disorders - Studies on Anorexia, Bulimia and Obesity the target language (Spanish) translated the original ques- were obtained to assess the relationship between the WBIS- tionnaire from English to Spanish. A panel of experts on M and the other subscales. Moreover, the WBIS-M score was the subject, whose native language was Spanish and who correlated to BMI. Mean comparisons (ANOVA), adjusted had a good level in the source language (English), unified by parental origin, age, and socioeconomic status, were also a preliminary Spanish version. Then, the Spanish version calculated to analyse the relationship between weight status was back-translated into English by two other independent and gender and the scores obtained in the WBIS-M. It must translators. A bilingual expert–reconciliation panel, made be noted that standardised values (z) of BMI were consid- up of the translators and members of the research team, then ered. Simple comparisons were also obtained. concluded a final Spanish version that guaranteed conceptual equivalence. Last, a pilot study was carried out with fifteen participants within the age and group of interest to confirm Results that the final Spanish version could be read and understood. The final document did not need to be modified. Descriptive analysis of items Data analysis Descriptive analyses of items are provided in Table 2. It must be noted that item 8 obtained higher scores in terms The following analytic plan was pre-specified. The psy- of skewness and kurtosis as compared to the rest of items. chometric properties of the WBIS-M for adolescents were analysed through IBM SPSS Statistics, version 24 (IBM Internal structure Corp., Armonk, N.Y., USA), and IBM SPSS Amos Statis- tics, version 20 (IBM Corp., Chicago, I.L., USA). A cross- The data obtained from subsample 1 (n = 149) proved to validation was conducted [44], for which the sample was be adequate for analysis through an EFA (KMO = 0.915, split into two random subsamples. Exploratory factor analy- χ2 = 1075.633, df = 55, p < 0.0001). According to paral- sis (EFA) was conducted with subsample 1 (n = 149) and lel analysis and Kaiser–Guttman rule, one-factor solution a confirmatory factor analysis (CFA) was conducted with was retained, explaining the 54.57% of variance. Item 1 subsample 2 (n = 149). Non-statistically significant differ- showed a low communality (0.092) and the lowest factor ences were observed between samples in terms of school loading (0.302), whereas the rest of the items showed higher (χ 2 = 9.061, df = 5, p = 0.107), year (χ2 = 2.403, df = 3, communalities and adequate factor loadings of above 0.6 p = 0.493) and gender (χ2 = 0,215, df = 1, p = 0.643). The in all cases (Table 3). These results suggested that item 1 adequacy of the data to conduct the EFA was tested using did not fit adequately into the one-factor model. Moreover, the Kaiser–Meyer–Olkin (KMO) test and Bartlett’s spheric- item–total corrected correlation was low (0.299), whereas ity test [45]. In line with the original development of the the rest of the items showed values higher than 0.6. There- WBIS [21], the EFA was conducted through varimax rota- fore, item 1 was excluded from further analysis. tion. In addition, the Kaiser–Guttman [46] rule and parallel CFA was conducted with subsample 2 (n = 149) to test analysis (Hayton, Allen & Scarpello, 2004) were applied to whether the unidimensional structure of the questionnaire retain factors. could be confirmed in another set of data. The ten items CFA was conducted through the unweighted least squares of the questionnaire (excluding item 1) were considered in (ULS) method, because data did not fit multivariate nor- the analysis given that the data showed an adequate fit to mality. The following goodness-of-fit indices were obtained: the one-factor model (χ2 = 200.515, df = 35), GFI = 0.995, goodness-of-fit index (GFI), the parsimony GFI (PGFI), and PGFI = 0.992, NFI = 0.991, and SRMR = 0.060). Standard- the normed fit index (NFI), and the standardized root-mean ised regression weights were adequate in all cases according square residual (SRMR). The cutoff points for these indices to Floyd and Widaman [44], ranging from 0.449 to 0.880. were: equal to or higher than 0.9 for GFI, and NFI, 0.6 for PGFI, and lower than 0.08 for SRMR [47, 48]. The inter- Internal consistency nal consistency of the scale was analysed using Cronbach’s alpha, McDonald’s omega, and item-test corrected correla- Reliability analyses were performed on both subsamples tions. According to Tavakol and Dennick [49], cutoff points and the total sample. As shown in Table 4, internal consist- for acceptable internal consistency indices range from 0.7 to ency was higher than 0.9 in both subsamples and in the total 0.95. The cutoff value of 0.4 was considered for acceptable sample, and according to Cronbach’s alpha and McDonald’s item-test correlations [50]. omega. Moreover, item analyses revealed that all the items A descriptive analysis of the questionnaires was con- contributed to the internal consistency of the scale, since ducted. Spearman’s correlation (for the ordinal variable item–total correlations were higher than 0.6 in both the Binge Eating) and partial correlations controlled by BMI subsamples and in the total sample. The scale showed an 13 Eating and Weight Disorders - Studies on Anorexia, Bulimia and Obesity Table 2  Descriptive statistics of items Item Mean (SD) Skewness Kurtosis 1.Because of my weight, I feel that I am just as competent as anyone/Dado mi peso, siento que soy tan com- 4.71 (2.06) − 0.44 − 1.13 petente como cualquier otra personaa 2.I am less attractive than most other people because of my weight/Soy menos atractivo/a que la mayoría de 2.53 (1.79) 1.02 − 0.10 las personas debido a mi peso 3.I feel anxious about my weight because of what people might think of me/Me siento nervioso/a respecto a 2.79 (2.04) 0.78 − 0.87 mi peso debido a lo que la gente puede pensar de mí 4.I wish I could drastically change my weight/Desearía poder cambiar mi peso drásticamente 2.74 (2.07) 0.93 − 0.52 5.Whenever I think a lot about my weight, I feel depressed/Siempre que pienso mucho sobre mi peso, me 2.30 (1.87) 1.30 0.33 deprimo 6.I hate myself for my weight/Me odio a mí mismo/a por mi peso 1.74 (1.51) 2.32 4.68 7.My weight is a major way that I judge my value as a person/Mi peso es una de las principales maneras por 1.85 (1.54) 2.09 3.71 las que juzgo mi valía personal 8.I don’t feel that I deserve to have a really fulfilling social life, because of my weight/Debido a mi peso, no 1.38 (1.13) 3.59 13.25 siento que merezca tener una vida social que me haga feliz 9.I am OK being the weight that I am/Estoy bien con el peso que tengoa 4.95 (2.09) − 0.60 − 1.09 10.Because of my weight, I don’t feel like my true self/Debido a mi peso, no siento que sea realmente yo 1.72 (1.49) 2.32 4.67 mismo/a 11.Because of my weight, I don’t understand how anyone attractive would want to date me/Debido a mi 2.09 (1.81) 1.60 1.32 peso, no entiendo cómo alguien atractivo querría tener una cita conmigo a Item reverse scored Spanish translation in italics Table 3  Communalities and factor loadings obtained from subsample 1 (n = 149) in the EFA Item Communalities Factor loading 1.Because of my weight, I feel that I am just as competent as anyone/Dado mi peso, siento que soy tan compe- 0.092 0.343 tente como cualquier otra personaa,b 2.I am less attractive than most other people because of my weight/Soy menos atractivo/a que la mayoría de 0.565 0.752 las personas debido a mi peso 3.I feel anxious about my weight because of what people might think of me/Me siento nervioso/a respecto a 0.683 0.827 mi peso debido a lo que la gente puede pensar de mí 4.I wish I could drastically change my weight/Desearía poder cambiar mi peso drásticamente 0.645 0.803 5.Whenever I think a lot about my weight, I feel depressed/Siempre que pienso mucho sobre mi peso, me 0.762 0.873 deprimo 6.I hate myself for my weight/Me odio a mí mismo/a por mi peso 0.667 0.817 7.My weight is a major way that I judge my value as a person/Mi peso es una de las principales maneras por 0.656 0.810 las que juzgo mi valía personal 8.I don’t feel that I deserve to have a really fulfilling social life, because of my weight/Debido a mi peso, no 0.478 0.691 siento que merezca tener una vida social que me haga feliz 9.I am OK being the weight that I am/Estoy bien con el peso que tengo.a 0.427 0.653 10.Because of my weight, I don’t feel like my true self/Debido a mi peso, no siento que sea realmente yo 0.458 0.677 mismo/a 11.Because of my weight, I don’t understand how anyone attractive would want to date me/Debido a mi peso, 0.570 0.755 no entiendo cómo alguien atractivo querría tener una cita conmigo Eigenvalue 6.404 a Item reverse scored b Analysis did not support the inclusion of item 1 Spanish translation in italics 13 Eating and Weight Disorders - Studies on Anorexia, Bulimia and Obesity Table 4  Internal consistency coefficients and item–total correlations in both subsamples and the total sample Subsample 1 (n = 149) Subsample 2 (n = 149) Total sample (n = 298) α (95%CI) ω (95% CI) I–T correla- α (95%CI) ω (95% CI) I–T correla- α (95%CI) ω (95% CI) I–T correla- tion range tion range tion range WBIS-M 0.930 0.936 0.633–0.849 0.928 0.934 (0.918 – 0.681–0.841 0.929 0.935 (0.926 – 0.659–0.845 10-items (0.913– (0.920– 0.950) 0.946) version 0.944) 0.951) I–T Item–total adequate internal consistency in terms of both the reliability to those reported in the original validation studies, except for coefficient and item–total correlations. those corresponding to the Dislike scale, which were slightly lower [21, 23]. WBIS-M showed significant correlations Correlations and differences by gender and weight with all the outcome variables when controlling for BMI. status Figure 2 shows the WBIS-M means based on the weight status (categorized according to zBMI) and gender. Table 5 shows the means and standard deviations of all the In the ANOVA, adjusted by parental origin, age, and scales. The mean of the WBIS-M was 2.22 (SD = 1.37). The socioeconomic status (degrees of freedom are corrected same table shows the results of the WBIS-M correlations for non-homogeneity of variances), the interaction between and the correlations reported by Pearl and Phul [23] for the the two variables was significant (F(1, 287) = 6.14; p = 0.014; original WBIS-M. The original WBIS correlations found by partial η2 = 0.060), as were the effects of weight status Durso and Latner [21] are also shown. (F(1, 287) = 22.65; p < 0.001; partial η2 = 0.193) and gender. A significant positive correlation was found between WBIS-M scores were significantly higher in the girls than in WBIS-M and BMI. The total and partial correlations (con- the boys (F(1, 287) = 33.47; p < 0.001; partial η2 = 0.104), with trolling for BMI) for all the other measures were comparable Table 5  Means, correlations, and partial correlations of WBIS and WBIS-M with study measures Measure Mean (SD) WBIS ­Correlationa WBIS-M WBIS-M Correlation WBIS Partial WBIS-M Partial WBIS- ­Correlationb ­Correlationa,c ­Correlationb,c M Partial ­Correlationc AAQ-Dislike 1.83 (0.84) 0.31** 0.17* 0.13* 0.32** 0.22** 0.12* BSQ 1.60 (0.84) 0.74** 0.77** 0.73** 0.75** 0.72** 0.70** DFT 7.70 (6.80) 0.47** 0.56** 0.64** 0.48** 0.50** 0.59** RSE 2.98 (0.65) − 0.68** − 0.56** − 0.53** − 0.67** − 0.50** − 0.53** DASS21 1.58 (1.18) 0.51** 0.44** 0.40** 0.50** 0.43** 0.40** EDDSe 3Mos 0.25** 0.36** – 0.24** 0.29** – 6Mos 0.32** 0.47** – 0.31** 0.40** – Binge Eating – − 0.16*,d – – – BMI 20.99 (3.49) 0.15 0.47** 0.31** – – – WBIS Weight Bias Internalization Scale, WBIS-M Modified Weight Bias Internalization Scale, AAQ-Dislike Dislike Subscale of Anti-fat Atti- tudes Questionnaire, BSQ Body Shape Questionnaire, DFT Drive for thinness subscale of Eating Disorders Inventory-3, RSE Rosenberg Self- Esteem Scale, DASS-21 21-item Depression Anxiety Stress Scales, EDDS 3Mos and 6Mos Binge eating behaviour in the past 3 and 6 months as measured by the Eating Disorder Diagnostic Scale, Binge Eating Binge eating severity using DSM-V criteria a Statistics also reported in Durso and Latner [21] b Statistics also reported in Pearl and Puhl [23] c Controlling for BMI d Spearman because Binge Eating is an ordinal variable e Reported in Durso and Latner [21] and Pearl and Puhl [23] * p < 0.05, two-tailed ** p < 0.01, two-tailed 13 Eating and Weight Disorders - Studies on Anorexia, Bulimia and Obesity WBIS-M for adults, where item 1 was retained in the final version, despite it showing the lowest factor loading and the lowest standardized regression weight [30]. The removal of item 1 from the original WBIS-M 11-item version has been observed in recent multinational studies on WBI cor- relates [54]. In summary, our exclusion of item 1 based on its anomalous psychometric functioning is supported by evidence coming from the previous studies. After remov- ing this item WBIS-M yielded a one-factorial structure with adequate good fit, replicating the unidimensional structure identified in the original development of the WBIS [21] Fig. 2  Line graph with the means of WBIS-M by gender and weight status and the WBIS-M [23]. This one-factorial structure was also found in all adaptations of the WBIS-M to other languages [24–27, 29, 30]. means of 2.57, SD = 1.55, and 1.90, SD = 1.09, respectively The mean WBIS-M score was slightly below the middle (Cohen’s d = 0,5). range. When compared with other versions of the WBIS-M, A simple contrast shows that participants with obesity which used the original 1–7 range scores, our mean WBIS- status (reference category) scored significantly higher on M score was slighly lower than in versions focused on adults, the WBIS-M than participants with underweight status including the original version [23, 30], and similar to other (p < 0.001), normal weight status (p < 0.001), and over- versions with youth and adolescents [24, 26]. Because WBI weight status (p < 0.001). research has mainly focused on adults, and few studies have been developed with adolescents across different weight sta- tuses, more studies are needed to confirm this trend. Discussion The second objective was to study the internal consist- ency. We obtained excellent reliability results in terms of To the best of our knowledge, this is one of the few studies both the internal consistency coefficient and item–total cor- that provides a validated version of the WBIS-M for use relations in both subsamples and for the whole sample. The with adolescents across weight categories, and the first one internal consistency was very similar to the original version adapted to Spanish. After conducting both an EFA and a [23] and similar or slightly higher than in the other WBIS-M CFA of the WBIS-M with different subsamples, the results versions published to date [24–27, 29, 30]. confirmed adequate psychometric properties of our WBIS-M The third goal was to study the association between the version, with adequate fit of the 10-item version (exclud- WBIS-M score and the same set of external variables ana- ing item 1), high internal consistence, and support for its lysed in the original version. We found that higher scores validity. of WBIS-M were positively associated with an increase in The study had four specific objectives. The first was to dislike of people with obesity, higher body dissatisfaction study the internal structure of this Spanish version with and drive for thinness, lower self-esteem, higher symptoms adolescents using EFA and CFA. Item 1 (‘‘Because of my of stress, depression and anxiety, and a higher frequency and weight, I feel that I am just as competent as anyone”), a severity of binge eating behaviour. Total and partial correla- reverse-keyed item, obtained a low communality when com- tions (adjusted by BMI) were highly comparable to those pared with the rest of the items, as well as a low factor load- reported in the original version, except the total correlation ing and a low item–total correlation. Therefore, it was not for the binge eating measure, which was lower in our study. included in subsequent analyses. These results are consist- Nevertheless, the fact that we used the Spanish versions of ent with previous studies conducted with adults who were the same measures used in the original version, except in the overweight or had obesity, also suggesting that this item case of binge eating, for which we used the stricter DSM-V did not correctly fit the WBIS-M structure [18, 28, 51, 52]. criteria [40], and that this variable was analysed as ordinal A similar result was found with WBIS versions for adoles- (and not as quantitative as in the original), must be taken cents with obesity seeking bariatric surgery [53]. Regarding onto account. The original version of the WBIS-M [23] used WBIS-M versions, the German WBIS-C for children also the same measures as per the original version of the WBIS removed item 1 [27], as did the Mainland Chinese version [21], and DSM-V criteria were not available when the WBIS for children and adolescents [26]. In the Chinese version was originally developed. All the correlated variables have with children and adolescents from Hong Kong, item 1 was been considered as correlates of WBI in adolescents [18, retained, although it showed a very low factor loading [25]. 53, 55–57]. These results provide support for the validity of A similar result was found in the Spanish version of the the questionnaire. 13 Eating and Weight Disorders - Studies on Anorexia, Bulimia and Obesity The fourth and last goal of this study was to assess the detecting real differences in WBI across these groups that relationships between WBI, gender, and weight status. are not attributable to different interpretations of WBIS-M Higher scores of WBI were found in girls and in higher item content between groups. To date, only the German weight statuses. Social pressure to attain the beauty ideal WBIS-C [27] and the WBIS-M version with Chinese ado- is higher among young women than among young men lescents from Hong Kong [25] have provided invariance for [58], which could make women more vulnerable to weight- gender groups, and the latter only partially for weight sta- based stigmatization [59]. Furthermore, young people who tus. We are planning to study measurement invariance for are overweight or are living with obesity are more prone gender, age groups, and weight status once the total sample to experience weight-based discrimination [6]. Our results initially conceived is assessed next year, if the restrictions are in line with research in adults across weight categories derived from the COVID-19 epidemic have been lifted. showing that women and people with higher BMIs show On the other hand, the study has several strengths. This higher WBI levels than men and people with lower BMIs is one of the few studies to provide a validated version of [23, 54]. Regarding children and adolescents, a higher risk the WBIS-M for use with adolescents across weight cat- of reporting weight discrimination has been associated with egories, and the first one adapted to Spanish. The internal the female gender and those with a higher BMI in Chinese structure of our WBIS-M version was analysed by means of adolescents [57]. Higher WBI scores in girls and partici- both EFA and CFA in different subsamples. The translation pants who were overweight than in boys and participants process was done following the International Test Commis- who were not overweight have been reported in previous sion Guidelines [43]. To ensure the quality of data collec- research conducted in Germany with a general population tion, the survey was designed in an online format, included sample [27]. The same results were observed in another Ger- response ranges and three control interspersed questions, man study conducted with adolescents seeking weight loss and was administered in classrooms under the supervision [56]. In contrast, other studies conducted with adolescents of a group of technicians previously trained. And last, the from the USA did not find gender differences regarding WBI adolescents’ weight and height were measured objectively in adolescents [18, 53]. Notably, these last studies were con- using portable and accurate instruments, and following a ducted with samples of adolescents seeking weight loss or standardized protocol, preventing inaccuracy in the self- bariatric surgery. Differences in sample characteristics could reported data in adolescents [60, 61]. explain these results. More studies are needed to clarify the role of gender regarding WBI in adolescents. Together, our results regarding WBI and gender and weight status differ- Conclusion ences support the known-groups validity of our WBIS-M version. Our findings provide support for the validity and reliabil- ity of our WBIS-M version for use with adolescents across Strength and limits weight categories. The availability of a Spanish version of the WBIS-M for adolescents will provide research teams This study has several limitations and strengths. Regarding and practitioners from Spain who are addressing weight limitations, there is a lack of data on the test–retest reliabil- bias in the population with a validated version of one of the ity. Although studies conducted with adults have provided few existing but the most employed instrument for assess- stability data for the WBIS-M (e.g. Macho et al. [30]), up ing WBI. In addition to its value for research and clinical to now only the Mainland Chinese version with a general settings, data coming from the use of this new tool could population sample [26] and the German version with par- contribute as evidence to support the promotion of educa- ticipants seeking weight loss [56] have reported satisfac- tional, regulatory, and legal initiatives designed to prevent tory stability with adolescent samples. Another limitation weight stigma and discrimination, as recommended by a is related to the sample. A bigger sample would provide a recent international consensus statement for ending obesity more accurate analysis of the internal structure of the ques- stigma [2]. tionnaire. Moreover, it has to be noted that the sample is not representative of Spanish adolescents. Last, measurement invariance by relevant groups such as age, gender, or weight What is already known on this subject? status is necessary and this could not be estimated in our study because of the insufficient sample size. Measurement Weight bias internalization (WBI) is pervasive and poten- invariance would provide substantial endorsement for using tially damaging for health, independent of weight status. The the same version of the questionnaire with different groups WBIS-M is one of the few existing but the most employed (e.g., by gender, different ages, weight status) without tar- instrument for assessing WBI, and several validations to dif- geted adaptations to specific group characteristics, and for ferent languages have been developed. There are currently 13 Eating and Weight Disorders - Studies on Anorexia, Bulimia and Obesity no Spanish versions for assessing WBI in children and ado- References lescents across different weight statuses. 1. Brewis AA, SturtzSreetharan C, Wutich A (2018) Obesity stigma What this study adds? as a globalizing health challenge. Global Health 14(20):1– 6. https://​doi.​org/​10.​1186/​s12992-​018-​0337-x 2. Rubino F, Puhl RM, Cummings DE et al (2020) Joint interna- This study provides research teams and practitioners from tional consensus statement for ending stigma of obesity. Nat Med. Spain who are addressing weight bias in the population with https://​doi.​org/​10.​1038/​s41591-​020-​0803-x a validated version of one of the few existing but the most 3. Alimoradi Z, Golboni F, Griffiths MD et al (2020) Weight-related stigma and psychological distress: a systematic review and meta- employed instrument for assessing WBI among adolescents analysis. Clin Nutr 39(7):2001–2013 across different weight statuses. 4. Emmer C, Bosnjak M, Mata J (2020) The association between weight stigma and mental health: a meta-analysis. Obes Rev Acknowledgements The authors would like to thank the participants 21:1–13. https://​doi.​org/​10.​1111/​obr.​12935 and school staffs who participated in this work. 5. Vartanian LR, Porter AM (2016) Weight stigma and eating behav- ior: a review of the literature. Appetite 102:3–14. https://​doi.​org/​ Author contributions Conceptualization: DS-C, AF-D and AA; formal 10.​1016/j.​appet.​2016.​01.​034 analysis: AA and AF-D; sampling and data collection design: AM-R, 6. Puhl RM, Lessard LM (2020) Weight stigma in youth: prevalence, AF-D and DS-C; data curation: DSC and AF-D; translation process: consequences, and considerations for clinical practice. Curr Obes ARS and LB-C; writing— original draft preparation: DS-C, AA and Rep. https://​doi.​org/​10.​1007/​s13679-​020-​00408-8 AF-D; writing—review and editing: all authors; supervision: DS-C. All 7. Tomiyama AJ, Epel ES, McClatchey TM et al (2014) Associations authors have approved the final version of the manuscript. of weight stigma with cortisol and oxidative stress independent of adiposity. Heal Psychol. https://​doi.​org/​10.​1037/​hea00​00107 8. Thedinga HK, Zehl R, Thiel A (2021) Weight stigma experiences Funding Open Access Funding provided by Universitat Autonoma and self- exclusion from sport and exercise settings among people de Barcelona. This research was supported by a Research Grant with obesity. BMC Public Health 21(565):1–18. https://​doi.​org/​ from the Spanish Ministry of Science, Innovation and Universities 10.​1186/​s12889-​021-​10565-7 (RTI2018-099293-B-I00). 9. Puhl RM, Quinn DM, Weisz BM, Suh YJ (2017) The Role of Stigma in Weight Loss Maintenance Among U.S. Adults. Data availability The data presented in this study are available on Ann Behav Med 51(5):754–763. https:// ​ d oi. ​ o rg/ ​ 1 0. ​ 1 007/​ request from the corresponding author. The data are not publicly avail- s12160-​017-​9898-9 able due to privacy or ethical reason. 10. Alberga AS, Edache IY, Forhan M, Russell-Mayhew S (2019) Weight bias and health care utilization: a scoping review. Prim Declarations Health Care Res Dev. https://​doi.​org/​10.​1017/​s1463​42361​90002​ 27 11. Daly M, Robinson E, Sutin AR (2020) Perceived overweight and Conflict of interest None. suicidality among US adolescents from 1999 to 2017. Int J Obes 44:2075–2079. https://​doi.​org/​10.​1038/​s41366-​020-​0620-9 Ethical approval This was approved by the Ethic in Animal and Human 12. Sutin AR, Stephan Y, Terracciano A (2015) Weight discrimination Experimentation Committee from the Universitat Autònoma de Bar- and risk of mortality. Psychol Sci 26:1803–1811. https://​doi.​org/​ celona (CEAAH 3451). Parental consent and adolescent’s approval 10.​1177/​09567​97615​601103 were obtained. 13. Pont SJ, Puhl RM, Cook S et al (2017) Stigma experienced by children and adolescents with obesity. Pediatrics 140:e20173034. Informed consent Informed consent was obtained from all individual https://​doi.​org/​10.​1542/​peds.​2017-​3034 participants included in the study. 14. Puhl RM, Himmelstein MS, Pearl RL (2020) Weight stigma as a psychosocial contributor to obesity. Am Psychol 75:274–289. Open Access This article is licensed under a Creative Commons Attri- https://​doi.​org/​10.​1037/​amp00​00538 bution 4.0 International License, which permits use, sharing, adapta- 15. Ma L, Chu M, Li Y et al (2021) Bidirectional relationships tion, distribution and reproduction in any medium or format, as long between weight stigma and pediatric obesity: a systematic review as you give appropriate credit to the original author(s) and the source, and meta-analysis. Obes Rev. https://​doi.​org/​10.​1111/​obr.​13178 provide a link to the Creative Commons licence, and indicate if changes 16. Pearl RL, Puhl RM (2018) Weight bias internalization and health: were made. The images or other third party material in this article are a systematic review. Obes Rev 19:1141–1163. https://​doi.​org/​10.​ included in the article's Creative Commons licence, unless indicated 1111/​obr.​12701 otherwise in a credit line to the material. If material is not included in 17. Himmelstein MS, Puhl RM, Pearl RL et al (2020) Coping with the article's Creative Commons licence and your intended use is not weight stigma among adults in a commercial weight management permitted by statutory regulation or exceeds the permitted use, you will sample. Int J Behav Med 27:576–590. https://​doi.​org/​10.​1007/​ need to obtain permission directly from the copyright holder. To view a s12529-​020-​09895-4 copy of this licence, visit http://​creat​iveco​mmons.​org/​licen​ses/​by/4.​0/. 13 Eating and Weight Disorders - Studies on Anorexia, Bulimia and Obesity 18. Puhl RM, Himmelstein MS (2018) Weight bias internalization Measurement equivalence across ethnicity and clinical status. Int among adolescents seeking weight loss: implications for eating J Eat Disord 41:265–272. https://​doi.​org/​10.​1002/​eat.​20492 behaviors and parental communication. Front Psychol 9:1–9. 38. Elosua P, López-Jáuregui A, Sánchez-Sánchez F (2010) EDI-3, https://​doi.​org/​10.​3389/​fpsyg.​2018.​02271 Inventario de trastornos de la conducta alimentaria-3 Manual. Tea 19. LaCroix E, Alberga AS, Russell-Mathew S et al (2017) Weight Ediciones, Madrid bias: a systematic review of characteristics and psychometric 39. Stice E, Telch CF, Rizvi SL (2000) Development and validation properties of self-report questionnaires. Obes Facts 10:223–237. of the eating disorder diagnostic scale: a brief self-report measure https://​doi.​org/​10.​1159/​00047​5716 of anorexia, bulimia, and binge-eating disorder. Psychol Assess. 20. Lillis J, Luoma JB, Levin ME, Hayes SC (2010) Measuring weight https://​doi.​org/​10.​1037/​1040-​3590.​12.2.​123 self-stigma: the weight self-stigma questionnaire. Obesity 18:971– 40. American Psychiatric Association (2013) Diagnostic and statisti- 976. https://​doi.​org/​10.​1038/​oby.​2009.​353 cal manual of mental disorders, 5th edn. American Psychiatric 21. Durso LE, Latner JD (2008) Understanding self-directed stigma: Association, Arlington Development of the weight bias internalization scale. Obesity. 41. Neumark-Sztainer D, Wall M, Guo J et al (2006) Obesity, dis- https://​doi.​org/​10.​1038/​oby.​2008.​448 ordered eating, and eating disorders in a longitudinal study of 22. Depierre JA, Puhl RM (2012) Experiences of weight stigmati- adolescents: how do dieters fare 5 years later? J Am Diet Assoc zation: a review of self-report assessment measures. Obes Facts 106:559–568. https://​doi.​org/​10.​1016/j.​jada.​2006.​01.​003 5:897–918. https://​doi.​org/​10.​1159/​00034​6259 42. Sánchez-Carracedo D, López-Guimerà G, Fauquet J et al (2013) 23. Pearl RL, Puhl RM (2014) Measuring internalized weight attitudes A school-based program implemented by community providers across body weight categories: validation of the Modified Weight previously trained for the prevention of eating and weight-related Bias Internalization Scale. Body Image 11:89–92. https://​doi.​org/​ problems in secondary-school adolescents: the MABIC study 10.​1016/j.​bodyim.​2013.​09.​005 protocol. BMC Public Health 13:955. https://​doi.​org/​10.​1186/​ 24. Apay SE, Yılmaz E, Aksoy M, Akalın H (2017) Validity and 1471-​2458-​13-​955 reliability study of Modified Weight Bias Internalization Scale in 43. International Test Comission (2017) The ITC Guidelines for Turkish. Int J Caring Sci 10:1341–1347 Translating and Adapting Tests (Second edition). https://​www.​ 25. Pakpour AH, Tsai MC, Lin YC et al (2019) Psychometric proper- intes​tcom.​org/​files/​guide​line_​test_​adapt​ation_​2ed.​pdf ties and measurement invariance of the Weight Self-Stigma Ques- 44. Floyd FJ, Widaman KF (1995) Factor analysis in the develop- tionnaire and Weight Bias Internalization Scale in children and ment and refinement of clinical assessment instruments. Psychol adolescents. Int J Clin Heal Psychol 19:150–159. https://​doi.​org/​ Assess. https://​doi.​org/​10.​1037/​1040-​3590.7.​3.​286 10.​1016/j.​ijchp.​2019.​03.​001 45. Bartlett MS (1950) Tests of significance in factor analysis. Br J 26. Chen H, Ye YD (2021) Validation of the Weight Bias Internaliza- Stat Psychol. https://d​ oi.o​ rg/1​ 0.1​ 111/j.2​ 044-8​ 317.1​ 950.t​ b0028​ 5.x tion Scale for Mainland Chinese Children and Adolescents. Front 46. Kaiser HF (1960) The application of electronic computers to fac- Psychol 11:1–10. https://​doi.​org/​10.​3389/​fpsyg.​2020.​594949 tor analysis. Educ Psychol Meas. https://​doi.​org/​10.​1177/​00131​ 27. Zuba A, Warschburger P (2018) Weight bias internalization across 64460​02000​116 weight categories among school-aged children. Validation of the 47. Hu LT, Bentler PM (1999) Cutoff criteria for fit indexes in covari- Weight Bias Internalization Scale for Children. Body Image. ance structure analysis: conventional criteria versus new alternatives. https://​doi.​org/​10.​1016/j.​bodyim.​2018.​02.​008 Struct Equ Model. https://​doi.​org/​10.​1080/​10705​51990​95401​18 28. Hilbert A, Baldofski S, Zenger M et al (2014) Weight bias inter- 48. Steiger JH (2007) Understanding the limitations of global fit nalization scale: psychometric properties and population norms. assessment in structural equation modeling. Pers Individ Dif. PLoS ONE 9:1–7. https://​doi.​org/​10.​1371/​journ​al.​pone.​00863​03 https://​doi.​org/​10.​1016/j.​paid.​2006.​09.​017 29. Kliem S, Puls HC, Hinz A et al (2020) Validation of a three- 49. Tavakol M, Dennick R (2011) Making sense of Cronbach’s alpha. item short form of the Modified Weight Bias Internalization Scale Int J Med Educ 2:53–55. https://​doi.​org/​10.​5116/​ijme.​4dfb.​8dfd (WBIS-3) in the German Population. Obes Facts 13:560–571. 50. Ware JE, Gandek B (1998) Methods for testing data quality, scal- https://​doi.​org/​10.​1159/​00051​0923 ing assumptions, and reliability: the IQOLA project approach 30. Macho S, Andrés A, Saldaña C (2021) Validation of the modified international quality of life assessment. J Clin Epidemiol weight bias internalization scale in a Spanish adult population. 51(11):945–952 Clin Obes. https://​doi.​org/​10.​1111/​cob.​12454 51. Durso LE, Latner JD, Ciao AC (2016) Weight bias internalization in 31. Hollingshead A (1975) Four factor index of social status, 1st edn. treatment-seeking overweight adults: psychometric validation and Dept of Sociology Yale University, New Haven associations with self-esteem, body image, and mood symptoms. Eat 32. World Health Organization (2007) BMI-for-age (5–19 years). Behav 21:104–108. https://​doi.​org/​10.​1016/j.​eatbeh.​2016.​01.​011 https://w ​ ww.w ​ ho.i​ nt/t​ ools/g​ rowth-r​ efere​ nce-d​ ata-f​ or-5​ to19-y​ ears/​ 52. Lee MS, Dedrick RF (2016) Weight Bias Internalization Scale: indic​ators/​bmi-​for-​age Psychometric properties using alternative weight status classifica- 33. Crandall CS (1994) Prejudice against fat people: ideology and tion approaches. Body Image 17:25–29. https://​doi.​org/​10.​1016/j.​ self-interest. J Pers Soc Psychol. https://​doi.​org/​10.​1037/​0022-​ bodyim.​2016.​01.​008 3514.​66.5.​882 53. Roberto CA, Sysko R, Bush J et al (2012) Clinical correlates of the 34. Magallares A, Morales JF (2014) Spanish adaptation of the Antifat weight bias internalization scale in a sample of obese adolescents Attitudes Scale/Adaptación al castellano de la Escala de Actitud seeking bariatric surgery. Obesity 20:533–539. https://​doi.​org/​10.​ Antiobesos. Rev Psicol Soc 29:563–588. https://​doi.​org/​10.​1080/​ 1038/​oby.​2011.​123 02134​748.​2014.​972707 54. Pearl RL, Puhl RM, Lessard LM et al (2021) Prevalence and Cor- 35. Martin-Albo J, Nunez JL, Navarro JG, Grijalvo F (2007) The relates of Weight Bias Internalization in Weight Management: a Rosenberg self-esteem scale: translation and validation in uni- multinational study. SSM Popul Heal. https://​doi.​org/​10.​1016/j.​ versity students. Span J Psychol 10:458–467 ssmph.​2021.​100755 36. Bados A, Solanas A, Andrés R (2005) Psycometric properties 55. Marshall RD, Latner JD, Masuda A (2020) Internalized weight of the spanish version of depression, anxiety and stress scales bias and disordered eating: the mediating role of body image (DASS). Psicothema 17:679–683 avoidance and drive for thinness. Front Psychol 10:1–6. https://​ 37. Warren CS, Cepeda-Benito A, Gleaves DH et al (2008) Eng- doi.​org/​10.​3389/​fpsyg.​2019.​02999 lish and Spanish versions of the Body Shape Questionnaire: 13 Eating and Weight Disorders - Studies on Anorexia, Bulimia and Obesity 56. Ciupitu-Plath C, Wiegand S, Babitsch B (2018) The Weight Bias gender discrimination in America. Int J Obes 32:992–1000. Internalization Scale for youth: validation of a specific tool for https://​doi.​org/​10.​1038/​ijo.​2008.​22 Assessing Internalized Weight Bias among Treatment-Seeking 60. Fonseca H, Silva AM, Matos MG et al (2010) Validity of BMI German adolescents with overweight. J Pediatr Psychol 43:40–51. based on self-reported weight and height in adolescents. Acta Pae- https://​doi.​org/​10.​1093/​jpepsy/​jsx079 diatr Int J Paediatr 99:83–88. https://d​ oi.o​ rg/1​ 0.1​ 111/j.1​ 651-2​ 227.​ 57. Sutin AR, Terracciano A, Li G, Wang L (2021) Frequency 2009.​01518.x and correlates of Weight-Based Discrimination among ado- 61. Elgar FJ, Roberts C, Tudor-Smith C, Moore L (2005) Validity of lescents in China. Int J Behav Med. https://​doi.​org/​10.​1007/​ self-reported height and weight and predictors of bias in adoles- s12529-​021-​09982-0 cents. J Adolesc Heal 37:371–375. https://d​ oi.o​ rg/1​ 0.1​ 016/j.j​ adoh​ 58. López-Guimerà G, Levine MP, Sánchez-Carracedo D, Fauquet J ealth.​2004.​07.​014 (2010) Influence of mass media on body image and eating disor- dered attitudes and behaviors in females: a review of effects and Publisher's Note Springer Nature remains neutral with regard to processes. Media Psychol 13:387–416. https://​doi.​org/​10.​1080/​ jurisdictional claims in published maps and institutional affiliations. 15213​269.​2010.​525737 59. Puhl RM, Andreyeva T, Brownell KD (2008) Perceptions of weight discrimination: prevalence and comparison to race and 13