Introduction

Young people with CDC-defined overweight/obesity frequently underestimate their weight status (i.e., indicate their weight is ‘about right’), suggesting they may misinterpret or choose not to identify with CDC-defined operationalized definitions of weight categories [1, 2]. Though the occurrence of weight misperception among young adults has been established, the impact of weight misperception on individuals’ mental and physical health is less clear. Historically, weight misperception, and more specifically weight category underestimation, has been viewed as having a negative impact on engagement in healthy lifestyle behaviors. Historically, it has been posited that underestimating one’s weight may lead to decreased physical activity, lower consumption of healthier food, and greater sedentary behaviors [3,4,5]. More recent research has identified weight misperception as a potential protective factor against future weight gain [6], depression [7], and disordered eating behaviors [8].

RacialFootnote 1 and ethnic health disparities in weight status have been well established, with prevalence rates indicating that Black and Hispanic/Latinx young people have higher rates of obesity than their white counterparts [11]. In addition, studies show that individuals who are of a minoritized racial or ethnic background are more likely to indicate that their weight is ‘about right' [12,13,14]. The association between weight misperception and physical and mental health, as well as weight loss behaviors, are influenced by race and ethnicity status. Specifically, Black individuals classified with overweight or obesity and who misperceived their weight, were less likely to report weight loss attempts than their white and Hispanic/Latinx counterparts [5]. Weight misperception has also been found to be a stronger protective factor against depressive symptomatology for white young adults relative to their Black counterparts [7].

These findings suggest that factors contributing to racial and ethnic disparities in weight status are complex. Despite the established existence of weight status disparities among minoritized individuals, identifying mechanisms facilitating these differences continues to be an understudied research area. Accuracy of weight perception has been found to influence weight loss behaviors, and accuracy of weight perception differs across various racial and ethnic groups. Thus, there is a clear need to disentangle the complex relations among race, weight perception, body mass index (BMI), and physical and mental health outcomes. In order to effectively disentangle these associations, culturally responsive tools for measuring weight perception are needed.

Figure rating scales, which utilize visual imagery of body silhouettes to guide participants’ responses to a variety of questions about their bodies, have been used to assess accuracy of weight perception for males and females [15]. Despite established racial differences in weight perception [12,13,14], the most ubiquitous figure rating scale, developed by Stunkard and colleagues [15], does not reflect cultural variations in body shape. Since the development of Stunkard’s figure rating scale, additional figure rating scales have been created and culturally adapted [16,17,18]. However, these adapted figure rating scales tend to be used less frequently in research than the well-known Stunkard scale. Furthermore, researchers have not yet established which scales are most appropriate for differing racial, ethnic, and gender groups.

The objective of the present study was to explore intersectional differences in accuracy of weight perception in a diverse sample of young adults by examining which of four separate figure rating scales [15,16,17,18] were related to greatest accuracy of CDC-defined weight status for varying groups based on race (Black vs. white) and gender (male vs. female).

Methods

Subjects and procedures

With Institutional Review Board approval, undergraduate students were recruited into the WRAP (Weight Reporting from Adolescents Perspectives) study cohort at a public university in the Mid-Southern region of the United States. Participants were excluded from the study if they were pregnant or endorsed taking medication or having a medical condition affecting their weight status, dietary choices, or physical activity. After participants completed electronic questionnaires, trained research assistants measured their height and weight using a Tanita BWB 800S Doctors Scale (Tanita Corporation, Arlington Heights, IL) and a standard PE-WM-BASE stadiometer (Perspective Enterprises, Portage, MI).

The WRAP study cohort included 471 participants, who were age 18–65 years (Mage = 21.2, SD = 6.0) and weight diverse (15.7–57.2 kg/m2, MBMI = 25.9, SD = 6.53). For the present analyses, only participants between the ages of 18 and 25 categorized as having CDC-defined “normal or healthy weight” henceforth referred to as lean weight (BMI = 18.5–24.9), CDC-defined “overweight” (BMI = 25–29.9), and weight in the CDC-defined “obese” range (BMI ≥ 30) were included. Participants over age 25 and participants with a BMI < 18.5 were excluded from analyses given the study’s focus on young adults and lean vs. high weights. The final sample for the present study consisted of 322 young adults (Mage = 19.7, SD = 1.8) who were majority women (74%), with BMI ranging from 18.5 to 57.2 (MBMI = 26.01, SD = 6.46). For the present study, analyses were restricted to participants identifying as Black or African American and white or European American based upon research supporting unique cultural perceptions of weight among Black and white individuals [19, 20].

Measures

Demographics, BMI, and weight category

Participants completed a demographic questionnaire that included items assessing self-reported age, gender, race, and ethnicity (Table 1). Participant BMI was calculated using measured height (m) and weight (kg) by dividing each participant’s weight (kg) by height in meters squared (kg/m2) [21]. Congruent with recent literature [6], participants indicated how they perceived themselves in terms of their weight by choosing from the following CDC-defined weight status categories: “very underweight,” “underweight,” “healthy weight,” “overweight,” and “very overweight/obese.” For the present analyses, the final two perception categories were collapsed, which is congruent with previous findings that perception of overweight or obesity is associated with maladaptive outcomes [6].

Table 1 Demographic variables by overall sample and by weight status (M ± SD)
Full size table

Figure rating scales

Using four different visual figure rating scales (FRS; see Fig. 1), body size perception was assessed by asking participants, “Which figure is closest to your current body size?” [17]. The first FRS scale, hereafter referred to as the Silhouette Scale, by Anderson and colleagues [16] displayed 18 shadow figure silhouettes (9 men and 9 women) with sizes ranging from 1 (obese) to 9 (underweight), which were reverse coded for the present study. The second FRS scale, hereafter referred to as the Stunkard Scale, consisted of 18 white figure silhouettes (9 men and 9 women) with sizes ranging from 1 (underweight) to 9 (obese) [15]. A third FRS scale, hereafter referred to as the Culturally-Adapted Scale, by Mitola and colleagues [17] was initially adapted from Stunkard and colleagues [15] and consisted of 18 culturally adapted figure silhouettes (9 men and 9 women) which were brown-skinned, with sizes ranging from 1 (underweight) to 9 (obese). On the fourth and final FRS scale, hereafter referred to as the Photo-Based Scale by Harris and colleagues [18], participants responded to questions based on 20 photo-based silhouettes of white individuals (10 men and 10 women) with sizes ranging from 1 (underweight; BMI < 18.5) to 10 (class III obesity; BMI ≥ 40). Overall, high scores on the figure rating scales indicated that participants chose figures with a higher BMI.

Fig. 1
figure 1

Figure rating scales

Full size image

Analytic plan

Using IBM’s Statistical Package for the Social Sciences (SPSS) version 25, Receiver Operating Characteristics (ROC) curve analyses were conducted to examine and compare the accuracy of weight and body perception predictors (i.e., weight perception, four figure rating scales) in classifying CDC-defined “overweight or obesity” (OWOB) status among young adults. The present analyses were stratified to examine the intersection of race and gender (i.e., white female, Black male, Black female, white male). The dependent variable (i.e., CDC-defined weight status) was dichotomized into OWOB (BMI ≥ 25; coded as 1) or lean weight (LW; BMI = 18.5–24.9; coded as 0).

ROC curve analyses estimated the probability that a participant will be classified as OWOB vs. LW based upon their responses on the weight perception item and the four figure rating scales. The area under the curve (AUC) statistics for each predictor were examined. This statistic indicated the probability that a randomly chosen participant with OWOB would have higher weight and body perception than a randomly chosen participant without OWOB [22]. AUC values are evaluated using the following criteria [23]: ≥ 0.90 “excellent”, ≥ 0.80 “good”, ≥ 0.70 “fair”, < 0.70 “poor”. To compare the performance of the five predictors in discriminating weight status, Hanley and McNeil’s [24] methodology for comparing within sample AUCs was implemented using Youngstrom [22] excel spreadsheets containing the necessary calculations.

Results

Area under the curve (AUC) statistics indicated all scales were significantly better at classifying OWOB than chance (Table 2, Fig. 2). With respect to intersectional analyses, among Black females, the Culturally Adapted Scale was “excellent” at classifying OWOB and had the strongest discriminatory ability (AUC = 0.93, SE = 0.02, p < 0.001, 95% CI = 0.89–0.97). Weight perception labels had the weakest discriminatory ability (AUC = 0.84, SE = 0.03, p < 0.001, 95% CI = 0.77–0.91). When comparing AUC estimates for predictors among Black females, the Silhouette Scale, the Stunkard Scale, the Culturally Adapted Scale, and the Photo-Based Scale statistically differed from and were superior to weight perception labels (all p < 0.05).

Table 2 ROC curve analyses: AUC statistics by race and gender
Full size table
Fig. 2
figure 2

Receiver operating characteristics (ROC) curves by race and gender

Full size image

When examining AUC estimates among Black males, the Culturally Adapted Scale was “excellent” at classifying OWOB and had the strongest discriminatory ability (AUC = 0.93, SE = 0.04, p < 0.001, 95% CI = 0.86–1.00). Weight perception labels had the weakest discriminatory ability (AUC = 0.77, SE = 0.07, p < 0.01, 95% CI = 0.63–0.92). Comparison of AUC estimates indicated that the Silhouette Scale and the Culturally Adapted Scale were statistically superior to weight perception labels (p < 0.05) at discriminating OWOB from LW.

Among white females, the Silhouette Scale was “excellent” at classifying OWOB and had the strongest discriminatory ability (AUC = 0.93, SE = 0.03, p < 0.001, 95% CI = 0.88–0.99). Weight perception labels had the weakest discriminatory ability (AUC = 0.82, SE = 0.05, p < 0.001, 95% CI = 0.73–0.91). When comparing AUC estimates for predictors among white females, the Silhouette Scale was significantly superior to the Stunkard Scale (p < 0.05), the Culturally Adapted Scale (p < 0.05), and weight perception labels (p < 0.05) at discriminating OWOB from LW. The Photo-Based Scale was also significantly better than weight perception labels (p < 0.05).

Among white males, the Photo-Based Scale was “good” at classifying OWOB and had the strongest discriminatory ability (AUC = 0.84, SE = 0.06, p = 0.001, 95% CI = 0.71–0.96). Weight perception labels had the weakest discriminatory ability (AUC = 0.76, SE = 0.08, p < 0.01, 95% CI = 0.60–0.92). However, when statistically comparing AUC estimates for predictors among white males, no predictors significantly differed.

Discussion

Given racial and ethnic disparities in weight status and varying levels of weight misperception across race and ethnicity, it is pivotal to examine culturally responsive tools for assessing an individual’s weight perception. Accordingly, this study examined differences in weight perception accuracy across four figure rating scales and weight perception labels among a diverse sample of young adults by race and gender. The results of this study add to a growing body of literature examining weight perception among young adults. In addition, this study addresses a gap in the weight perception literature by examining this construct using widely used, but not culturally tailored, figure rating scales as well as culturally tailored figure rating scales. While previous research has examined differences across race and ethnicity in weight perception, our study moves the field forward by examining weight perception using multiple figure rating scales to determine which, if any, are associated with greatest weight accuracy perception. In addition, our study also moves the field forward by acknowledging research indicating that weight misperception may be a protective factor against negative outcomes among young people [7]. Acknowledgement of such research serves to balance the research literature that primarily focuses on risk and negative outcomes in the context of overweight and obesity.

The objective of the present study was to explore intersectional differences in accuracy of weight perception by comparing weight status labels and four separate figure rating scales [15,16,17,18]. Overall, findings underscore the utility of figure rating scales rather than weight status labels (i.e., underweight, very underweight, healthy weight, overweight, or very overweight) when examining weight perception among young adults given that all participants more accurately reported their weight using figure rating scales relative to weight status labels. Thus, clinicians and researchers should consider utilizing figure rating scales rather than weight labels when assessing weight perception among young adults. This approach is supported by research showing that weight labels such as “obese” are not preferred by patients [25, 26] and may perpetuate weight stigma [27]. Furthermore, weight terminology preferences vary in perceived desirability and acceptability across populations and settings [28] and as such, patients’ understanding of such stigma and personal preferences may influence their choice of weight status labels.

Our findings also support the notion that racial identification, as well as gender, impact the accuracy of figure rating scales when assessing weight perception. This finding is well situated in the literature given that prior research has found differences in weight perception accuracy by race and ethnicity [2, 4, 14, 29]. Furthermore, our finding that a culturally adapted figure rating scale [17] was the strongest classifier of overweight/obesity perception among Black young adults underscores the importance of using culturally tailored measures of weight perception. Findings that the shadow silhouette and culturally adapted ratings scales were the strongest classifiers of overweight and obesity among young adult white and Black women, respectively, whereas the culturally adapted and photo-based figure rating scale were the strongest classifier of overweight and obesity among young adult Black and white men, respectively, also highlight established differences in weight perception across gender.

Given prior research indicating that weight misperception may be associated with engagement in health-promoting behaviors among Black individuals [1, 5], it is imperative to identify and utilize accurate and culturally appropriate assessment measures of weight perception. The results of the current study suggest there is clinical utility in administering figure rating scales rather than using weight status labels when assessing weight perception among young adults. The current study further underscores the importance of utilizing culturally adapted figure rating scales when assessing weight perception among young adults across both assessment and intervention settings.

Limitations

While this study adds novel information to the growing literature addressing weight perception in a culturally responsive manner and highlights the importance of utilizing culturally adapted figure rating scales, there are some limitations. The cross-sectional design of our study limits our ability to make temporal or directional statements about the association between the variables of interest. Additionally, restricting the sample to an age range of 18–25-year-old college students may limit the generalizability to young adults in non-college settings or individuals of varying ages. Our study consisted of Black and white male and female participants with a higher proportion of female participants. This sample distribution does not represent the full variety and proportions of racial, ethnic, and gender backgrounds.

Future directions

Given our sample restrictions, future studies should aim to replicate our findings in more diverse populations across gender, race, and ethnicity to allow for a fuller examination of the impact of intersectional identities on weight perception. Additionally, future research is needed to explore the health effects of weight perception accuracy when utilizing culturally-adapted scales. Specifically, given the protective effects of weight misperception, it is not clear whether improved accuracy should be the ultimate goal in efforts to improve the health behaviors and habits of young adults. It may be that understanding the underlying mechanisms by which weight misperception is protective could lead to improved strategies for improving the health of young adults.

Conclusions

This study is novel in that we examined weight perception accuracy by race and gender using weight status labels and four figure rating scales in a sample of diverse young adults. Our study moves the field forward by highlighting the importance of utilizing culturally-adapted figure rating scales across both assessment and intervention settings. While findings will need to be replicated across other demographic groups, it is important to highlight that the use of weight perception labels is likely an ineffective method of assessing weight status among young adults. Future research should explore why certain figure rating scales are more or less salient for young adults. Such knowledge could provide key insights for future work focused on disentangling the complex relations among weight perception, health outcomes, gender, race, and ethnicity among young adults.

What is already known on this subject?

Silhouette scales are a good method to assess body size perception. It is unknown whether accuracy of body size reporting varies as a function of scale type and demographics. We examined figure rating scales accuracy by gender and race.

What this study adds?

Race and gender influence body size reporting accuracy, suggesting the need for tailored figure rating scales in the assessment of weight status.