This study analyzed a sample of Norwegian speaking participants over 18 years of age on their responses to the Norwegian version of the WBIS-M [12], a weight-neutral version of the WBIS [6]. When excluding two proposed versions of the competence item, the remaining 10 items had high factor loadings and good model fit, confirming a one-factor solution, as demonstrated in previous studies [6, 10,11,12]. Therefore, the 10-item version of the Norwegian WBIS-M has satisfactory psychometric properties, and we suggest that it can be used to measure internalized weight bias in Norwegian samples in a weight-neutral fashion. Using a weight-neutral scale such as the WBIS-M over the WBIS has the advantage of being inclusive of all weight categories, improving usability in studies across weight classes, or with individuals who have changed weight status, for instance, due to bariatric surgery, who may present as “normal” or underweight but still experience WBI [12]. Furthermore, using terms such as “overweight” in a sample consisting of only those who self-describe themselves as overweight runs the risk of biasing responses toward greater reported internalized stigma due to priming of negative stereotypes, which the use of this 10-item version of the WBIS-M could minimize. An analysis of the descriptive data showed that most of the sample in the present study identified as normal or overweight. We also found that the degree of internalized weight bias in our Norwegian sample was lower for both genders than in the original validation of the WBIS-M [12], but participants who self-perceived as heavier reported greater levels of WBI and lower overall health.
Our results showing better fit with a 10-item version are in line with previous investigations of the psychometric properties of the WBIS that have indicated that the first item related to competence can be challenging [10, 11]. Initial inspection of answers to the first item halfway through our study indicated that this might also be the case in the Norwegian version of the WBIS-M. Specifically, it became unclear whether agreement with the item indicated presence of WBI or not, due to a distribution of answers indicating possible ambiguity, along with low model fit. To address these concerns, the wording of the item was further discussed amongst the clinician-translators, and two versions of the competence item were tested. To our knowledge, this represents the first attempt to see if different phrasings of this item could improve the fit and clarity of the item that has been published. The results showed that neither version showed satisfactory psychometric validity in the CFA and hence did not contribute to a good model fit. Our work indicates that this item may be particularly hard to convey as weight-neutral – at least in Norwegian. The fact that it is one of only two positively phrased items in the scale could also possibly affect how participants interpret the item and respond to it.
There may also be weight-dependent differences in experience with stigma that may affect the interpretation of this item that our sample did not allow us to explore further. While both underweight and overweight individuals can experience weight stigma, the types of stereotypes directed toward these two weight groups differ [24], which in turn can affect the nature of internalized weight bias in different weight categories. For example, it may be likely that individuals categorized as overweight would be more likely to self-stigmatize relative to their competence than those who are normal or underweight, as those categorized as overweight can frequently experience being labeled as lazy or lacking in willpower in the public discourse [1]. Therefore, an investigation of both versions of the competence item in a larger sample with higher proportions of individuals with under- and overweight might yield different results. There may also be cultural concerns in Norway related to the relationship between weight and competence that have not yet been investigated. As mentioned, our sample showed lower WBI than the sample in the original publication on the WBIS-M [12], which could perhaps also affect the tendency to internalize WBI related to competence. Interestingly, the advent of social media may also mean that sub-cultural differences may arise more quickly than before, particularly in some demographic groups such as the young [25], meaning that cultural influences may be less obvious to pin down than before the social media age. Given some previous issues with this competence-related item when measuring WBI [10, 11], it may also be inherently challenging to operationalize the relationship between weight and competence regardless of the sample, culture, and language. Therefore, further work may be needed to include competence in this scale in a reliable and valid way, at least in a Norwegian context. This could include doing qualitative exploration using a focus group on the underlying concept of WBI related to competence, suggesting new ways of operationalizing, and rephrasing this concept in a weight-neutral way. For instance, it may be worthwhile to use phrasing which does not emphasize a comparative aspect with others, such as “My weight affects how competent I feel”. Such a non-comparative phrasing could perhaps also be beneficial for other language versions of the WBIS-M.
An investigation of gender differences in our material showed a significantly higher degree of internalized weight bias (WBI) among women compared to men. This coincides with other results wherein women experienced more stigma the greater their weight and are more likely to internalize this stigma [9, 26]. Neither self-perceived weight, psychological/emotional state, nor overall health differed among genders in our study, indicating that the higher WBI in the women was likely not a result of differences in these variables. For both genders, higher WBI correlated with higher self-perceived weight, lower overall health, and a more negative psychological state, in the expected directions. In women, higher WBI was also related to both a more negative psychological/emotional state and lower overall health when controlling for self-perceived weight. For men, this was only the case for WBI and psychological state when controlling for self-perceived weight. These results indicate that there is a more robust relationship between WBI and health in women than men, perhaps related to more widespread discrimination against women of higher weight in some settings [13]. However, greater WBI was related to greater psychological distress for both genders, independent of weight, and WBI should therefore be addressed regardless of gender, particularly in clinical settings.
Limitations and future research
This study has some limitations that can be addressed in future research. First, the sample size was too small for further analyses that could have shed more light on WBI across different weight categories, particularly because we did not have many underweight participants. Second, the study did not include objective measurements of BMI. However, this was a deliberate choice founded in feedback from user organizations and clinical work in the research group where we have found that the practice of BMI-measurements can be considered stigmatizing and/or triggering for some individuals, particularly with eating disorders [27]. Moreover, previous investigations on WBI have shown that self-perceived weight is an acceptable and efficient measure when investigating WBI and can be used as an alternative to objective BMI [10]. We do not expect our results would have changed significantly with the use of BMI instead of self-perceived weight, although perhaps an over-estimation of self-perceived weight might be related to more WBI than actual BMI for some respondents.
A further limitation is the low number of men (n = 63) compared to women included in the sample despite the efforts that were undertaken to recruit more men. We had too few men to be able to perform a valid test of measurement invariance among the gender groups, so we cannot rule out that the scale functions differently for men and women. Future studies ought to verify measurement invariance in this Norwegian translation of the WBIS-M to give more confidence in examinations of gender differences. However, inspection of the correlations between WBI and general health, psychological/emotional state, and self-perceived weight in our sample were similar and in expected directions in both genders, strengthening the assumption that the scale measured the same construct in both genders. The relative lack of men could reflect a weakness in the recruitment strategy. It could also indicate that internalized weight bias is perceived as more relevant to women than men. In line with this, we found that the women in our sample showed more WBI than the men. Again, a larger sample would have allowed us to explore possible nuances of this result. For instance, while not observed in our analyses, other studies have shown that men can show more U-shaped tendencies in internalized weight stigma, wherein they experience it most when underweight or overweight [13, 14]. They may also have a higher BMI-threshold before they perceive themselves as overweight [15]. Therefore, as women were overrepresented in our sample, the validity of the Norwegian WBIS-M scale may be more firmly established using samples with a more equal distribution of genders. Studying a more diverse sample in terms of gender identity might also nuance the results further. We had only 1 respondent indicate a gender identity apart from male/female, which did not allow for further exploration, but future studies could aim for a more diverse recruitment strategy in terms of gender.
Furthermore, the study could have incorporated more measures of WBI for a better exploration of convergent validity. However, there is currently a lack of validated measures of WBI in Norwegian, so we chose to approach the question of convergent and external validity in terms of the relationship to measures known to be related to WBI, such as health status, psychological/emotional state, and self-perceived weight. The presence of face and construct validity was also an important part of the discussions during the translation process, which was conducted by a clinical team with long experience working with WBI in patients with higher weight and eating disorders. Future studies should look at adaptation of other measures of WBI into Norwegian, to compare these with the Norwegian translation of the WBIS-M for a more in-depth examination of convergent and construct validity. Finally, further investigations of how to incorporate an item related to competence both in the Norwegian version of the weight-neutral WBIS-M and other scales measuring WBI are still warranted, along with investigations on competence-related WBI across weight categories. This could include qualitative investigations of the concept of competence and WBI in focus groups.
Add Comment