Contrary to minority stress theory, racial and ethnic identity has been described as a protective factor for eating disorders, primarily due to social support within their family [7] . Such social support among Black and Hispanic women may be a buffer from acculturative stress and decrease the personal relevance of thin-ideal pressures typically associated with White appearance ideals [8] . However, results are inconclusive when examining specific types of eating disorders, their prevalence, and symptomatology among different racial and ethnic groups [9] .

Regarding people of Asian backgrounds, past research reveals that Asian Americans do not differ from White, Black, or Hispanic Americans on lifetime or past-year diagnoses of either AN, BN, or BED [5] . Yet, compared to White Americans, Asian Americans were shown to have a greater lifetime and past-year presence of any binge-eating behaviors [5] . These findings were the same for both men and women. Thus, while Asian Americans do partake in a certain amount of disordered eating behaviors, they do not seem to be at greater risk than other racial or ethnic groups for the more severe eating disorders.

When exploring people of Hispanic backgrounds, the findings of previous research differ based on the eating disorder. Compared to White Americans, Hispanic Americans had lower lifetime diagnoses of AN and equivalent lifetime diagnoses of BED [5] [10] . However, the findings pertaining to BN are mixed, with some studies showing no difference between White and Hispanic Americans on lifetime diagnoses of BN [5] [6] , and other studies showing that Hispanic Americans had higher lifetime diagnoses of BN than White Americans [10] . A systematic review of studies investigating racial and ethnic differences in binge eating revealed that most studies found no differences among Hispanic, White, and Black Americans on sub-clinical binge eating behaviors [11] . Taken together, these past research findings suggest that although Hispanic ethnicity may be a protective factor against AN, Hispanic Americans may potentially be at heightened risk for BN.

Regarding people of Black backgrounds, the research findings vary as a function of gender and type of eating disorder. For American women, Black women had lower lifetime diagnoses of AN and equivalent lifetime diagnoses of BN and BED in comparison to White women, along with no differences in comparison to Asian and Hispanic women [5] . In contrast, for American men, Black men had higher lifetime diagnoses of BN and equivalent lifetime diagnoses of AN and BED in comparison to White men, along with no differences in comparison to Asian and Hispanic men [5] . Regarding subclinical levels of disordered eating, whereas research shows no differences between Black and White Americans when men and women are combined in the same sample [12] , other research found that Black women exhibited lower levels of disordered eating behaviors than White women [13] . Thus, a Black woman intersectional identity seems to be a protective factor against disordered eating behaviors, particularly those associated with AN.

Orthorexia nervosa (ON) is a term that was initially derived from the Greek “right” and “appetite” to describe an obsession with “correct” eating [14] . Not currently classified as a disorder in the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5) [15] , ON is a proposed condition where an individual pursues an extreme diet, increasing time and effort to plan, prepare, and consume foods deemed purely healthy [16] . Moreover, four common symptoms have been overwhelmingly agreed upon among researchers and clinicians [16] [17] . The first is an obsessive preoccupation with the purity or healthfulness of food accompanied by rigid dietary rules and restrictions to abstain from foods that the individual deems to be impure, unhealthy, or toxic. The second is an experience of emotional distress such as shame, guilt, and anxiety from violations of their dietary rules. The third is an experience of psychosocial impairments such as impaired interpersonal relationships and impaired concentration due to preoccupation with the diet. The fourth is an experience of physical impairments due to a lack of nutrition from extreme dietary restrictions. Unfortunately, by the time that health problems develop due to an eating disorder, the disordered thinking (regarding the need for a rigid and restrictive diet, the perceived dangers of most foods, and the drive to obtain or maintain unrealistic body ideals) is so ingrained in the individual, that they find it extremely difficult to change their diet and overcome the eating disorder [16] [17] .

To our knowledge, only one study has examined ethnic differences in ON [18] . The results show that Hispanic Americans scored higher than non-Hispanic Americans on the Bratman orthorexia self-test [19] . Unfortunately, the study did not include a sufficient sample of some racial groups (e.g., only 10 Black participants) to conduct an analysis of racial differences. In addition, the Bratman self-test is lacking in reliability and validity in comparison to some of the other measures of ON [20] .

Past research reveals many factors associated with the development of eating disorders, including a distorted body image related to internalization of the thin ideal (predominantly among women) or the muscular ideal (predominantly among men), as well as the presence of a mood or anxiety disorder (often comorbid with BN or BED) or obsessive-compulsive disorder (often comorbid with AN or ON) [6] [8] [21] . However, past research on disordered eating has often ignored racial and ethnic influences. The current study aims to fill this gap in the research, exploring potential racial and ethnic differences in ON symptomatology, as well as general disordered eating behaviors including excessive unhealthy dieting behaviors associated with AN, as well as binging and purging behaviors associated with BN. Based on past research, it is expected (a) that Asian Americans will not differ from the other racial and ethnic groups on either unhealthy dieting behaviors associated with AN or binging and purging behaviors associated with BN, (b) that Black and Hispanic Americans will have lower levels of unhealthy dieting behaviors associated with AN than White Americans, and (c) that Hispanic Americans may have slightly higher levels of binging and purging behaviors associated with BN than White Americans. Given the lack of past research investigating ON differences, we have no a priori hypotheses regarding potential racial or ethnic differences in ON symptomatology.

The principal focus of this study was to investigate whether ON symptomatology differed across racial and ethnic groups (White, Black, Hispanic, Asian or Pacific Islander, and biracial or multiracial). A power analysis using G*Power revealed that with a standard α criterion of 0.05, a minimum sample size of 305 participants (61 per group) would be optimal for detecting group differences with a medium effect size. Two methods of recruitment were used to help achieve a sufficient sample size based on the a priori power analysis. First, recruitment emails were sent to undergraduate students enrolled in multiple nutrition courses and psychology courses at Texas State University. Second, participants were recruited through social media advertisements on Twitter, Reddit, Facebook, and Instagram. The audience interests for these advertisements were set to target both healthy eaters (e.g., “clean eating”, “healthy eating”, “paleo”) and normal or unhealthy eaters (e.g., “Cooking Panda”, “desserts”, “comfort food”). The only exclusionary criterion was an age of less than 18 years. The recruitment emails and social media advertisements indicated that participants completing the eating-habits survey would have the option of providing their email address for entry into a raffle for an opportunity to win one of ten $25 Amazon gift cards.

For this study, which was approved by the Institutional Review Board of Texas State University, participants provided informed consent and then completed an online survey through Qualtrics. In attempt to prevent bots, the “Prevent multiple submissions” and “Bot detection” features of the Qualtrics survey were enabled. The survey began with a demographic questionnaire, followed by measures of ON and disordered eating. The demographic questionnaire included questions asking about gender identity (three response options: man, woman, and nonbinary or genderqueer), race and ethnicity (six response options: White, Black, Hispanic or Latino/a/x, Asian or Pacific Islander, biracial or multiracial, and other), age, weight and height for calculation of body mass index (BMI), and past or current eating disorder diagnosis (five response options: no, yes—anorexia nervosa, yes—bulimia nervosa, yes—binge eating disorder, and yes—other).

ON was assessed with the Orthorexia Nervosa Inventory (ONI) [21] , which is comprised of 24 statements about eating, to which participants use a 4-point scale (ranging from 1 = “not at all true” to 4 = “very true”) to indicate how true each statement is for them based on their current eating habits. The ONI includes three subscales: 1) Behaviors with nine statements relevant to following a restrictive diet believed to be healthy and pure (e.g., “I follow a health-food diet rigidly, only eating what my diet allows and not allowing myself any deviations from this diet”), 2) Impairments with 10 statements relevant to either physical impairments (e.g., “The stricter I become with my diet, the more I seem to experience one or more physical symptoms such as fatigue, faintness, heart racing, nausea, diarrhea, pain, etc.”) or psychosocial impairments (e.g., “My healthy eating is a significant source of stress in my relationships”), and 3) Emotions with five statements relevant to extreme emotional distress with feelings of shame and anxiety in response to violations of the restrictive diet (e.g., “When I stray from my healthy diet, I can only think about what a failure I am”). Regarding the reliability of the ONI, Cronbach’s alpha was 0.94 for the overall ONI and ranged from 0.88 to 0.90 for its three scales in the original validation study [21] .

Disordered eating was assessed with the Eating Attitudes Test-26 (EAT-26) [22] that is comprised of 26 statements about eating, to which participants use a 6-point scale (ranging from 1 = “never” to 6 = “always”) to indicate how often they do or feel what is indicated in the statement. The EAT-26 includes three subscales: 1) Dieting with 13 statements relevant to dieting for weight loss (e.g., “I engage in dieting behaviors,” “I am terrified about being overweight”), 2) Bulimia and Food Preoccupation with six statements relevant to binging/purging behaviors or preoccupation with food (e.g., “I vomit after I have eaten,” “I feel that food controls my life”), and 3) Oral Control with seven statements relevant to self-control with eating (e.g., “I display self-control when I am around food,” “I avoid eating when I am hungry”). Regarding reliability of the EAT-26, Cronbach’s alpha was 0.90 for the overall EAT-26 and ranged from 0.83 to 0.90 for its three scales in the original validation study [22] .

First, a MANCOVA was conducted to assess whether the racial and ethnic groups differed in their scores on the three ONI subscales (Behaviors, Impairments, and Emotions) after controlling for three covariates that have been related to ON symptomatology in past research: gender, BMI, and EAT-26 Total scores. Second, a MANCOVA was conducted to assess whether the racial and ethnic groups differed in their scores on the three EAT-26 subscales (Dieting, Bulimia and Food Preoccupation, and Oral Control) after controlling for three covariates: gender, BMI, and ONI Total scores. For both MANCOVAs, Pillai’s Trace F test was used for the omnibus test due to its robustness to heterogeneous variances, Fisher’s LSD tests were used for the post-hoc tests in the case of a significant omnibus test, and η p 2 was used as the measure of effect size with a value of 0.01 considered a small effect, 0.06 a medium effect, and 0.14 a large effect. A standard alpha criterion of 0.05 was used to determine statistical significance.

This study evaluated the impact of racial and ethnic differences in ON symptomatology and disordered eating behaviors (dieting, bulimia and food preoccupation, and oral control). Regarding the former, to the best of our knowledge, no study has examined racial differences in ON symptomatology, and only one study has examined ethnic differences in ON symptomatology [18] , finding that Hispanic Americans scored higher than non-Hispanic Americans on the Bratman orthorexia self-test [19] . However, Bratman and Knight indicated that the 10-item measure was only meant to be used as an informal self-test, and research has indicated that the Bratman self-test is lacking in reliability and validity in comparison to the ONI and to some of the other measures of ON [20] . Using the ONI, the current study found that the racial and ethnic groups did not significantly differ in ON symptomatology, after controlling for gender, BMI, and other disordered eating attitudes.

One possible explanation of this finding is that BMI and disordered eating attitudes are correlated with ON, regardless of racial and ethnic differences. Regarding body weight and size, higher BMI tends to positively correlate with higher ON symptomatology [18] [23] [24] . This relationship may be due to body image, dieting, and weight loss culture. In societies with more prevalent dieting cultures, individuals are more likely to experience distorted body image and an extensive desire to engage in behaviors that reduce weight or affirm healthy eating choices, thus leading to a higher obsession with clean eating and greater ON symptomatology [24] [25] . Similarly, greater distorted eating attitudes are associated with greater ON symptomatology. These disordered eating attitudes are influenced by culture, social media, and psychopathological factors, such as obsessive-compulsive disorder [26] [27] [28] , which may be more relevant to ON symptomatology than race or ethnicity.

The second objective of our present study was to examine the racial and ethnic differences in disordered eating behaviors. Regarding our first hypothesis, the results supported that Asian Americans did not significantly differ from other racial and ethnic groups on unhealthy dieting behaviors (e.g., restrictive eating, food avoidance, excessive exercise), although they were higher on binge eating and purging behaviors compared to Black Americans. Perhaps the general lack of differences between Asian Americans and the other racial and ethnic groups is due to Asian Americans also being influenced by the Western body image standards of thinness and muscularity [29] [30] .

Consistent with our second hypothesis, Black Americans had significantly lower levels of unhealthy dieting behaviors compared to White Americans, as well as significantly lower binging and purging behaviors than White, Hispanic, and Asian Americans. Past research has shown that Black Americans engage in lower dietary restraint behaviors and have lower weight concern compared to White Americans [31] . Moreover, while Black Americans tend to have higher average BMIs compared to White and Asian Americans, their overall body satisfaction tends to be higher [30] [32] . One possible explanation for this is resiliency levels and social support among the Black community stemming from historical and social events (e.g., discrimination, racism, inequality). Some studies have suggested that higher resiliency moderates the relationship between stress and eating disorders [33] . Thus, higher resiliency among Black Americans may serve as a buffer and/or protective factor against unhealthy dieting behaviors. Future research would benefit from examining the relationship between resiliency and eating behaviors in racial and ethnic groups.

Contrary to our third hypothesis, White Americans reported significantly higher levels of binging and purging behaviors compared to Hispanic Americans. Many researchers have indicated that Hispanic Americans may be more likely to engage in binging and purging behaviors compared to other ethnic groups [5] [34] , especially Black Americans [35] . However, a more recent study has shown that when compared to White women, Hispanic women reported significantly less binging and purging behaviors [36] . One possible explanation for this recent study and for our finding is that stigma may have prevented our Hispanic participants from reporting these behaviors. Typically, racial and ethnic minority groups tend to report unfavorable behaviors less. Therefore, future research should try to explore this phenomenon.

This present study has some limitations. First, more than half of our sample was White participants. Although we reached our minimum sample size to detect group differences, the overrepresentation of one group can introduce biases that limit generalizability. This may be one possible explanation for why, contrary to past research, White participants reported higher binging and purging behaviors compared to Hispanic participants in our study. Moreover, it is important to consider that due to stigma, ethnic and racial minority groups tend to report less adverse behaviors. Second, 74% of our sample overwhelmingly identified as women. Past studies have indicated gender differences among racial and ethnic groups in disordered eating behaviors and attitudes [34] [37] . Some studies have reported that compared to White men, Black men engage in higher binge eating and purging behaviors, while others have reported the inverse [38] [39] . There have also been studies suggesting that both groups engage in these behaviors at similar rates [5] . Future studies should further investigate the effects of intersectional identities on disordered eating behaviors and ON. Third, participants completed self-report measures that can introduce social desirability bias. In light of the sensitive nature of eating disorders, it is probable that some participants were reluctant to admit these behaviors. Lastly, while this study examined various racial and ethnic groups, it overlooked important intracultural differences within these communities (e.g., African Americans vs Ghanian Americans or Korean Americans vs Malaysian Americans).