Body Image 11 (2014) 438–445
Contents lists available at ScienceDirect
Body Image journal homepage: www.elsevier.com/locate/bodyimage
Ethnic identity, thin-ideal internalization, and eating pathology in ethnically diverse college women Liya M. Rakhkovskaya, Cortney S. Warren ∗ Department of Psychology, University of Nevada, Las Vegas, NV, United States
a r t i c l e
i n f o
a b s t r a c t
Article history: Received 13 December 2013 Received in revised form 5 July 2014 Accepted 6 July 2014 Keywords: Ethnic identity Ethnicity Thin-ideal internalization Eating concerns Weight concerns
Although much research suggests that ethnic identity is positively correlated with psychological health for ethnic minority women, research examining ethnic identity’s relationships to thin-ideal internalization, weight concerns, and eating concerns is sparse. Consequently, this study examined these relationships in European American, African American, Latina, and Asian American college women (N = 816). As expected, univariate analyses of variance indicated that European American women scored lowest on ethnic identity and highest on eating and weight concerns, whereas African American women scored lowest on thin-ideal internalization. Hierarchical regression analyses indicated that ethnic identity was negatively associated with eating and weight concerns, while body mass index and thin-ideal internalization were positively associated. Ethnic identity moderated the relationship between thin-ideal internalization and eating concerns such that the relationship was stronger for participants with lower ethnic identity. These results suggest ethnic identity may be a direct or interactive protective factor against eating concerns in ethnically diverse college women. © 2014 Elsevier Ltd. All rights reserved.
Introduction A growing body of research suggests that ethnic identity is associated with psychological health, particularly for ethnic minorities ˜ in the United States (Martinez & Dukes, 1997; Nagel, 1994; UmanaTaylor, Wong, Gonzales, & Dumka, 2012). According to prominent ethnic and racial formation theories (Helms, 1990; Phinney, 1996), ethnic identity encompasses a sense of belonging to or acceptance of the norms and practices of one’s cultural or subcultural group. As such, ethnic identity contributes to a sense of community, culture formation, and self-enhancement (Nagel, 1994). For example, Martinez and Dukes (1997) found positive associations between ethnic identity and life purpose, self-esteem, and self-confidence in a sample of 12,386 high school students. Similarly, in a study examining gender, ethnic identity, and academic adjustment, ethnic identity buffered the negative link between discrimination and externalizing behaviors in Mexican American middle school stu˜ et al., 2012). dents (Umana-Taylor Given the positive associations between ethnic identity and mental health, ethnic identity may serve as a protective factor against eating pathology. Research conducted almost exclusively
∗ Corresponding author at: Department of Psychology, University of Nevada, Las Vegas, 4505 Maryland Parkway, Las Vegas, NV 89154, United States. E-mail address:
[email protected] (C.S. Warren). http://dx.doi.org/10.1016/j.bodyim.2014.07.003 1740-1445/© 2014 Elsevier Ltd. All rights reserved.
with African American samples lends some support for this theory (Henrickson, Crowther, & Harrington, 2010; Stein, Corte, & Ronis, 2010; Stojek, Fischer, & Collins, 2010; Turnage, 2004). For example, Turnage (2004) found positive associations between ethnic identity, global self-esteem, and positive appearance evaluation in a community sample of 105 African American female high school students. More recently, Henrickson et al. (2010) found a negative relationship between ethnic identity and disordered eating symptoms (e.g., purging, binge-eating, preoccupation with food, body shape or weight) in a sample of 93 undergraduate African American women. Although the inverse relationship between ethnic identity and disordered eating has empirical support in African American samples, research investigating this relationship among Asian American, Latina, and other ethnic minority groups is very sparse with less consistent results. On one hand, Stein et al. (2010) found that ethnic identity negatively predicted binge eating behaviors in community Mexican American women. On the other hand, Barry and Garner (2001) found that ethnic identity was not significantly correlated with the desire to become thin in a community sample of female East-Asian immigrants. Likewise, Rhea and Thatcher (2013) found that ethnic identity, combined with high self-esteem, correlated with lower eating pathology in African American, but not European American or Mexican American, female high school students. Consequently, to further explore these relationships, this study examined the relationships between ethnic identity,
L.M. Rakhkovskaya, C.S. Warren / Body Image 11 (2014) 438–445
thin-ideal internalization, weight concerns, and eating concerns in a sample of European American, African American, Hispanic/Latina, and Asian American women. Ethnic Identity, Thin-Ideal Media Internalization, and Eating Pathology One way that ethnic identity could be protective against eating pathology in ethnic minority women is by helping women reject mainstream Western media values and the thin ideal as self-relevant. The media refers to widespread Western mass communication outlets such as the Internet, television, magazines, advertisements, and video games. In Western cultures, these outlets promote the thin ideal by providing unrealistic depictions of women as predominantly thin, young, hypersexual, and Caucasian (Scharrer, 2013). Furthermore, Western culture’s emphasis on individualism and personal responsibility exacerbates the pressure to become thin by describing overweight individuals as weak, undisciplined, and responsible for their undesirable size (Brownell et al., 2010). Overall, endorsement of a thin ideal as personally relevant, referred to as thin-ideal internalization, is a key riskfactor for weight and eating concerns (Homan, 2010; Levine & Murnen, 2009; Scharrer, 2013). For example, a meta-analysis of media endorsement, body image, and eating pathology research found that exposure to thin-ideal media is a risk factor of negative body image and disordered eating symptoms (Levine & Murnen, 2009). Ethnic identity may be protective against thin-ideal internalization, weight concerns, and eating concerns because values and ideals of appearance differ greatly across ethnic groups. Specifically, the thin ideal promoted in Western media is characteristic of White mainstream values and ideals of beauty that may not be salient to individuals who identify with a group that values a larger or different physical ideal or places less emphasis on physical appearance as a determinant of female value (Warren, Gleaves, Cepeda-Benito, Fernandez, & Rodriguez-Ruiz, 2005). For example, African American culture tends to evaluate attractiveness based on a self-confident attitude and good style (Parker, Nichter, Nichter, Vuckovic, Sims, & Ritenbaugh, 1995; Poran, 2006; Rubin, Fitts, & Becker, 2003), as well as toward a curvy, rather than thin, figure (Overstreet, Quinn, & Agocha, 2010; Poran, 2002). Similarly, traditional Latino culture idealizes a larger physique and close, mutually dependent relationships (Chamorro & FloresOrtiz, 2000; Santiago-Rivera, Arredondo, & Gallardo-Cooper, 2002; Warren et al., 2005). While upholding similar collectivist values, Asian Americans tend to idealize a thin figure, as well as a paler complexion and European facial features (Hall, 1995; Mintz & Kashubeck, 1999). Consequently, strong ethnic identity may help some ethnic minority women (i.e., African Americans and Latinas) reject extreme thin beauty ideals. Meanwhile, women who strongly identify with their collectivist culture (i.e., Latinas and Asian Americans) may be protected from body dissatisfaction by placing less emphasis on personal appearance as a determinant of worth, although this link may be further shaped by the extent the individual’s collectivistic culture idealizes the thin ideal. Theoretically, differences in levels of ethnic identity may result in variable levels of eating pathology across ethnic groups. Indeed, a large body of research testing the relationship between disordered eating and ethnicity (rather than ethnic identity) lends some support for this hypothesis. Although results are mixed (Ferguson, 2013; Grabe & Hyde, 2006; Nouri, Hill, & Orrell-Valente, 2011), a meta-analysis by Grabe and Hyde (2006) found that European American women had significantly higher body dissatisfaction than African American women. In contrast, in a recent meta-analysis of 204 studies, Ferguson (2013) found little evidence for ethnic differences in the association between media exposure and
439
body dissatisfaction. Similarly, results of a study comparing Asian American and European American college students indicated that thin-ideal internalization explained the association between media exposure and body dissatisfaction equally for both groups (Nouri et al., 2011). Studies examining the direct association between ethnic identity and eating pathology would offer more straightforward evidence for its potential protective effects. Unfortunately, this type of research is very sparse. In one of the only studies to examine this possibility, Rogers, Wood, and Petrie (2010) found that ethnic identity was negatively associated with internalization of mainstream Western beauty ideals in a sample of 322 African American female undergraduates. Further investigations of ethnicity, ethnic identity, and other cultural attributes as factors associated with body image and eating-related concerns are warranted. Ethnic Identity as a Moderator In addition to being negatively correlated with thin-ideal internalization, weight concerns, and disordered eating for certain ethnic groups, it is possible that ethnic identity may act as protective factor against eating pathology by weakening the relationships between thin-ideal internalization and weight or eating concerns for women who belong to ethnic groups that value a larger or curvier ideal (i.e., African American; Latina). Indeed, limited research on ethnic identity in African American women supports this hypothesis (Rogers et al., 2010; Williams, 2009). For example, Williams (2009) found that African American female undergraduates with a strong ethnic identity exhibited a diminished drive for thinness and reported feeling less pressured to become thin. It is important to note that for individuals who belong to cultures that traditionally value the thin ideal (i.e., European Americans and Asian Americans), the relationship between ethnic identity and eating pathology is less clear. On one hand, it is possible that ethnic identity is protective because having a sense of belonging to a cultural group is, in and of itself, a resiliency factor. In support of this hypothesis, Stojek et al. (2010) found that stronger ethnic identity predicted lower thinness expectancies and symptoms of bulimia nervosa (e.g., binge eating, self-induced vomiting, excessive exercise) in a study of 493 ethnically diverse college women. Notably, the Stojek et al. (2010) sample included both European American and minority (i.e., African American, Asian American, Latina, etc.) participants. Conversely, if having a stronger sense of ethnic identity means stronger adherence to attaining the thin-ideal (as would be true of European American and Asian American cultures), ethnic identity may have less of a protective effect on weight and eating concerns. However, research supporting this alternative hypothesis is lacking. Current Study Despite extensive research suggesting ethnic differences in attitudes toward appearance and beauty standards (Chamorro & Flores-Ortiz, 2000; Hall, 1995; Mintz & Kashubeck, 1999; Poran, 2002, 2006; Rubin et al., 2003; Warren et al., 2005) and positive psychological variables associated with stronger ethnic identity (Henrickson et al., 2010; Martinez & Dukes, 1997; Nagel, 1994; ˜ Stein et al., 2010; Stojek et al., 2010; Umana-Taylor et al., 2012), research examining the relationship between ethnic identity and eating pathology is sparse (Henrickson et al., 2010; Stein et al., 2010; Turnage, 2004). Furthermore, this existing research focuses predominantly on comparing African American and European American women (Soh & Walter, 2013) and rarely includes Asian American and Hispanic/Latina samples. Consequently, to build on extant research, this study examined the relationships among ethnic identity, thin-ideal internalization,
440
L.M. Rakhkovskaya, C.S. Warren / Body Image 11 (2014) 438–445
weight concerns, and disordered eating in a sample of European American, African American, Asian American, and Latina female undergraduates. Specifically, we examined (a) mean levels of ethnic identity, sociocultural attitudes toward appearance (i.e., thin-ideal internalization) and eating pathology symptoms by ethnic group; (b) strengths of relationships between these constructs; and (c) whether ethnic identity or ethnicity moderated the relationship between thin-ideal internalization and eating pathology. We predicted that endorsement of the thin ideal would be lower among participants belonging to ethnic groups preferring larger, curvier figures (i.e., African American and Latina women) compared to ethnic groups preferring thin figures (i.e., European American and Asian American women). Additionally, in light of extensive research on the association between ethnic identity and mental health (Henrickson et al., 2010; Martinez & Dukes, 1997; Nagel, ˜ 1994; Stein et al., 2010; Stojek et al., 2010; Umana-Taylor et al., 2012), we predicted that ethnic identity would be negatively correlated with thin-ideal internalization, weight concerns, and eating concerns for women of all ethnic groups. In accordance with extant research (e.g., Stojek et al., 2010), we believe that the protective effects of ethnic identity would still extend to cultures valuing the thin ideal (i.e., European American and Asian American). Furthermore, we theorized that ethnic identity would serve as a moderator between thin-ideal internalization and weight and/or eating concerns such that the relationships would be weaker for individuals with stronger ethnic identity than for those with weaker ethnic identity.
Method Procedure Eligible students were recruited from introductory psychology courses via the University of Nevada, Las Vegas psychology department’s online research management system. Students meeting inclusionary criteria (i.e., female, age 18 and older) who registered for this study were automatically assigned a unique numeric code, devoid of any personal identifiers. An individualized link containing that code was sent to the participant through e-mail. Clicking on the e-mailed link directed participants to the informed consent sheet and the questionnaires, presented on online survey software (i.e., Qualtrics). This study was approved by the University of Nevada, Las Vegas’ Institutional Review Board. The questionnaires took participants approximately 45 min to 1 hour to complete. After completing the questionnaires, participants were awarded course credit. Notably, course requirements allowed participants the option of completing four hours of any study included on the research management system or summarize four articles on an assigned topic in psychology. As such, participants had no incentive to complete our study in particular.
Participants Participants were selected from the data of 849 female undergraduates at University of Nevada, Las Vegas, from a larger, ongoing study examining anti-fat attitudes, eating pathology, and cultural attributes among college women (Claudat, Warren, & Durette, 2012; Warren, Gleaves, & Rakhkovskaya, 2013; White & Warren, 2013). Although the larger study lacked exclusionary criteria, only women who self-identified as belonging to one of the four ethnic groups of interest were included in this study. A screen of the data revealed that no participants were male. No participants were excluded due to incomplete responses. This yielded a final sample of 816 participants.
On average, participants were about 21 years old (M = 21.05, SD = 5.47) and of average body mass index (BMI M = 23.59, SD = 5.14; De Onis & Habicht, 1996). A total of 47.0% (n = 384) self-identified as European American, 11.0% (n = 90) as African American, 22.5% (n = 184) as Asian American, and 19.3% (n = 158) as Latina/Hispanic. Participants also varied in generational status: 12.0% were 1st generation (i.e., born outside of the U.S.); 26.8% were 2nd generation (i.e., born in the U.S., to parents not born in the U.S.); 6.5% were 3rd generation; 11.4% were 4th generation; and 39.3% were 5th or 6th generation Americans. Measures Demographics. Participants were asked to report their age, height, weight, ethnicity, and generational status. Ethnic identity. The Ethnic Identity subscale of the Multigroup Ethnic Identity Measure (MEIM-EI; Phinney, 1992) is a 14-item scale measuring ethnic attitudes and behaviors. Responses are recorded on a 4-point Likert-type scale, ranging from 1 (Strongly Disagree) to 4 (Strongly Agree) and summed, with higher scores indicating stronger ethnic identity. A meta-analysis of the psychometric properties of the MEIM found satisfactory to good internal consistency (Ponterotto, Gretchen, Utsey, Stracuzzi, & Saya, 2003). In this study, Cronbach’s alphas were acceptable for the entire sample (˛ = .88), as well as within each ethnic group (European Americans = .86; African Americans = .90; Asian Americans = .85; Latinas = .89). Thin-ideal internalization. The Internalization-General subscale of the Sociocultural Attitudes Towards Appearance Questionnaire-3 (SATAQ-INT; Thompson, van den Berg, Roehrig, Guarda, & Heinberg, 2004) is a 9-item measure of thin-ideal internalization, or endorsement of sociocultural beauty standards in mainstream Western media. Items are rated on a 5-point Likert scale ranging from 1 (Definitely Disagree) to 5 (Definitely Agree) and summed, with higher scores indicating higher thin-ideal internalization. Previous research suggests that the overall SATAQ-3 has high internal consistency among college women (Thompson et al., 2004) and eating disordered patients (Calogero, Davis, & Thompson, 2004). The current sample yielded internally consistent scores in the overall sample (˛ = .96), as well as in each ethnic group (European Americans = .96; African Americans = .96; Asian Americans = .96; Latinas = .97). Weight and eating concerns. The Eating Disorder Examination Questionnaire (EDEQ) is a 41-item measure designed to measure self-reported symptoms of eating pathology over the past 28 days, with higher summed scores indicating higher levels of disordered eating symptoms (Fairburn & Beglin, 1994). We used two EDEQ subscales which examine participants’ concerns about their weight (EDEQ-W, five items) and eating (EDEQ-E, five items). Items are measured on a 7-point rating scale, ranging from 0 (No days) to 6 (Every day) and averaged, with higher subscale scores indicating more pathology. Luce and Crowther (1999) showed that the EDEQ has acceptable internal consistency in undergraduate women (i.e., Cronbach’s alphas ranging from .78 to .93). Internal consistency was adequate to moderately strong for the current overall sample (for ˛ = .79 for EDEQ-E; ˛ = .87 for EDEQ-W), as well as in each ethnic group (for EDEQ-E: European Americans = .79; African Americans = .81; Asian Americans = .74; Latinas = .73; for EDEQ-W: European Americans = .87; African Americans = .84; Asian Americans = .84; Latinas = .85). Statistical analyses. We used IBM SPSS Statistics version 20 for Windows to conduct analyses of variance (ANOVAs), Pearson’s r correlations, and regression analyses. Missing data were handled using pairwise deletion. We examined descriptive statistics and Pearson’s r correlations for all measures. Using VassarStats (Lowry, 2001), we conducted Fisher’s z comparisons to compare the
L.M. Rakhkovskaya, C.S. Warren / Body Image 11 (2014) 438–445
441
Table 1 Means (SDs) and ANOVAs examining demographic and outcome variables by ethnic group. Variable
European American
African Americans
Asian Americans
Latinas
F
p
2
Age BMI Generation status Ethnic identity Thin-ideal internalization Weight concerns Eating concerns
21.71 (6.39)a 23.72 (5.11) 4.22 (1.17)a 2.60 (0.53)a 30.82 (9.98)a 3.46 (1.73)a 2.07 (1.30)a
21.69 (6.31) 24.91 (6.20)a 4.54 (1.19)a 3.03 (0.56)b 22.79 (10.60)b 2.75 (1.62)b 1.78 (1.24)
19.87 (3.08)b 22.18 (3.69)b 2.17 (1.39)b 2.88 (0.46)b 29.45 (10.04)a 3.16 (1.60) 1.95 (1.68)
20.48 (4.40) 23.14 (5.63)a 2.49 (1.26)b 2.96 (0.53)b 28.77 (10.99)a 3.18 (1.63) 1.73 (1.00)b
5.75 7.29 165.35 31.05 14.20 4.23 3.27
<.01 <.01 <.01 <.01 <.01 .01 .02
.02 .03 .38 .10 .05 .02 .01
Note. Means in the same row that do not share subscripts differ at p < .05 on post hoc tests with Bonferroni correction. BMI = body mass index. N = 816 (n = 384 for European Americans; n = 90 for African Americans; n = 184 for Asian American; n = 158 for Latinas.
strengths of the correlations by ethnic group. One-way ANOVAs, with p values adjusted using the post hoc Bonferroni correction, were used to examine whether mean values differed by ethnic group. Additionally, we conducted two separate 3-step hierarchical regression analyses to test the unique contribution of ethnic identity on eating pathology: one testing weight concerns and one testing eating concerns as the criterion variable (see Aiken & West, 1991). We did not mean-center the variables, but z-transformed the scores. In Step 1, we entered demographic variables (i.e., age, ethnicity, BMI) to test and control for group differences. Ethnicity was coded as follows: European Americans as 1, African Americans as 2, Asian Americans as 3, and Latinas as 4. In Step 2, we entered the thin-ideal internalization and ethnic identity simultaneously. In Step 3, we added the thin-ideal internalization × ethnic-identity interaction term. When a statistically significant interaction emerged, we examined moderating effects via a categorical dummy variable for the level of ethnic identity, as suggested by Aiken and West (1991). Specifically, we separated participants into three groups: a low ethnic identity subgroup (i.e., participants with ethnic identity scores over one standard deviation below the mean), a medium ethnic identity subgroup (i.e., participants with ethnic identity scores within one standard deviation of the mean), and a high ethnic identity group (participants with ethnic identity scores above one standard deviation of the mean). We then examined the slopes between the low ethnic identity and the high ethnic identity subgroups. We used Interaction software (Soper, 2010) to plot any significant interaction terms. Results Descriptive Analyses and Means by Ethnic Group As shown in Table 1, ethnic groups differed significantly by age, F(3, 809) = 5.75, p < .01, and BMI, F(3, 805) = 7.29, p < .01, although the effect sizes were small (2 = .02 and .03, respectively). Specifically, Asian Americans were younger than European Americans and reported a lower BMI than all other groups. European Americans and African Americans reported significantly higher generational status (i.e., had lived in the U.S. for significantly more generations) than Asian Americans and Latinas, F(3, 803) = 165.35, p < .01. With regard to eating pathology, European Americans scored significantly higher than African Americans on weight concerns, F(3, 696) = 4.23, p = .01, 2 = .02, and higher than Latinas on eating concerns, F(3, 696) = 3.27, p = .02, 2 = .01. For thin-ideal internalization, African Americans scored significantly lower than all other groups, F(3, 767) = 14.20, p < .01, 2 = .05. Finally, European Americans scored significantly lower than all other groups on ethnic identity, F(3, 808) = 31.05, p < .01, 2 = .10. Correlations As shown in Table 2, ethnic identity and weight concerns were significantly negatively correlated in the overall sample and for
Latinas, but not for European Americans, African Americans, or Asian Americans. Additionally, ethnic identity and eating concerns were significantly negatively correlated in the overall sample, but not significantly correlated for any ethnic group. Furthermore, ethnic identity did not correlate significantly with thin-ideal internalization in the overall sample or for any ethnic group. Additionally, as expected, thin-ideal internalization, weight concerns, and eating concerns were all significantly positively correlated for the overall sample and for each ethnic group. Fisher’s z comparisons did not indicate significant ethnic differences in any of the aforementioned relationships. Hierarchical Regression Analyses Hierarchical regression analyses predicting eating concerns in the overall sample are presented in Table 3. All variables were z-transformed prior to analyses. Ad-hoc descriptive statistics showed that BMI and age were highly kurtotic (5.95 and 15.50, respectively). Inverse transformation (i.e., 1/age; 1/BMI resulted in minimized kurtosis for both variables (.07 for BMI and 2.66 for age). Thus, we used inverted age and BMI for our final regression analyses. Eating concerns. The overall model accounted for 27% of the variance in eating concerns. In Step 1, ethnicity and BMI were significantly associated with eating concerns, as was the model as a whole, F(4, 666) = 16.02, p < .001. In Step 2, BMI, ethnic identity and thin-ideal internalization each explained unique variance in eating concerns, as indicated by a significant change in R2 from Step 1 to Step 2, R2 = .10, F(2, 660) = 75.13, p < .001. In Step 3, ethnic identity was no longer a significant predictor, but BMI and the ethnic identity by thin-ideal internalization interaction (MEIM-EI × SATAQ-INT) significantly predicted eating concerns, as indicated by a significant change in R2 from Step 2 to Step 3, R2 = .01, F(1, 659) = 7.47, p = .01. To understand the nature of the interaction, we re-ran the regression analyses by ethnic identity level and examined significance of the simple slopes in Interaction software (Fig. 1). Results indicated that the relationship between thin-ideal internalization and eating concerns was stronger (p < .05) for the low ethnic identity group, B = .51, SE B = .05, t = 10.60, p < .001, than for the high ethnic identity group, B = .33, SE B = .05, t = 6.72, p < .001. Groups also differed on mean levels of eating concerns, such that the low ethnic identity group scored significantly higher than the high ethnic identity group, F(1, 218) = 5.58, p = .02. The groups did not differ on mean thin-ideal internalization scores, F(1, 246) = 0.13, p = .72. Weight concerns. Hierarchical regression analyses predicting weight concerns are presented in Table 4. The overall model accounted for 48% of the variance in weight concerns. In Step 1, ethnicity and BMI were significantly associated with weight concerns, as was the model as a whole, F(4, 666) = 61.10, p < .001. In Step 2, ethnic identity and thin-ideal internalization predicted additional unique variance in weight concerns, as indicated by a significant change in R2 from Step 1 to Step 2, R2 = .21, F(2, 660) = 134.40,
442
L.M. Rakhkovskaya, C.S. Warren / Body Image 11 (2014) 438–445
Table 2 Bivariate correlations for the sample and by ethnicity. Group Total sample Age BMI Generation status Ethnic identity Thin-ideal internalization Weight concerns European Americans Age BMI Generation status Ethnic identity Thin-ideal internalization Weight concerns African Americans Age BMI Generation status Ethnic identity Thin-ideal internalization Weight concerns Asian Americans Age BMI Generation status Ethnic identity Thin-ideal internalization Weight concerns Latinas Age BMI Generation status Ethnic identity Thin-ideal internalization Weight concerns
BMI
Generation status
Ethnic identity
Thin-ideal int.
Weight concerns
Eating concerns
.22** –
.04 .10** –
.01 −.02 −.24** –
−.02 .04 −.02 −.03 –
.08* .47** .03 −.10** .48** –
.05 .28** .06 −.12** .42** .75**
.23** –
−.04 .06 –
.07 −.05 −.20** –
−.09 .05 .01 .07 –
.04 .47** .04 −.06 .45** –
.02 .30** .06 −.10 .40** .76**
.34** –
−.21 −.02 –
.04 .11 −.06 –
.03 .03 .02 .03 –
.23* .54** −.02 −.03 .55** –
.13 .39** .10 −.04 .50** .81**
.15 –
−.13 .12 –
−.03 .03 .00 –
−.03 .12 −.04 .10 –
.00 .59** .04 −.03 .44** –
.01 .31** .01 −.11 .36** .75**
.08 –
.03 .05 –
.00 −.13 −.31** –
.15 .06 −.06 −.12 –
.11 .47** −.05 −.17* .51** –
.11 .21* −.05 −.06 .46** .70**
Note. N = 816 (n = 384 for European Americans; n = 90 for African Americans; n = 184 for Asian American; n = 158 for Latinas. BMI = body mass index. * p < .05. ** p < .01.
Table 3 Regression analyses predicting eating concerns. Predictor
B
SE B
ˇ
t
p
R2
Step 1 BMIa Agea Ethnicity Generation status
−0.30 0.03 −0.11 −0.02
0.04 0.04 0.04 0.04
−0.30 0.03 −0.11 −0.02
−7.86 0.80 −2.44 −0.37
<.01 .43 .02 .71
.10**
Step 2 BMIa Agea Ethnicity Generation status Thin-ideal internalization Ethnic identity
−0.27 0.02 −0.04 0.01 0.40 −0.10
0.03 0.03 0.04 0.04 0.03 0.03
−0.27 0.02 −0.04 0.01 0.40 −0.10
−7.95 0.49 −0.96 0.19 11.88 −2.79
<.01 .63 .34 .85 <.01 .01
.26**
Step 3 BMIa Agea Ethnicity Generation status Thin-ideal internalization Ethnic identity Interaction
−0.26 0.01 −0.04 0.01 0.90 0.17 −0.57
0.03 0.03 0.04 0.04 0.19 0.11 0.21
−0.26 0.01 −0.04 0.01 0.90 0.17 −0.57
−7.59 0.41 −1.02 0.15 4.84 1.65 −2.73
<.01 .68 .31 .88 <.01 .10 .01
.27*
Note. Interaction = thin-ideal internalization by ethnic identity interaction term. a Inverted variables. * in R2 is significant at p < .05. ** in R2 is significant at p < .01.
L.M. Rakhkovskaya, C.S. Warren / Body Image 11 (2014) 438–445
443
Table 4 Regression analyses predicting weight concerns. Predictor
B
SE B
ˇ
t
p
R2
Step 1 BMIa Agea Ethnicity Generation status
−0.52 0.04 −0.10 −0.06
0.03 0.03 0.04 0.04
−0.52 0.04 −0.10 −0.06
−15.31 1.18 −2.57 −1.59
<.01 .24 .01 .11
.27**
Step 2 BMIa Agea Ethnicity Generation status Thin-ideal internalization Ethnic identity
−0.49 0.03 −0.03 −0.03 0.45 −0.09
0.03 0.03 0.03 0.03 0.03 0.03
−0.49 0.03 −0.03 −0.03 0.45 −0.09
−17.13 0.90 −0.85 −0.97 16.06 −2.95
<.01 .37 .39 .33 <.01 <.01
.48**
Step 3 BMIa Agea Ethnicity Generation status Thin-ideal internalization Ethnic identity Interaction
−0.49 0.02 −0.03 −0.03 0.69 0.04 −0.27
0.03 0.03 0.03 0.03 0.16 0.09 0.17
−0.49 0.02 −0.03 −0.03 0.69 0.04 −0.27
−16.82 0.85 −0.89 −1.00 4.40 0.46 −1.52
<.01 .39 .38 .32 <.01 .65 .13
.48
Note. Interaction = thin-ideal internalization by ethnic identity interaction term. a Inverted variables. * in R2 is significant at p < .05. ** in R2 is significant at p < .01.
Fig. 1. Thin-ideal internalization and eating concerns in participants with low vs. high ethnic identity.
p < .001. In Step 3, the ethnic identity by thin-ideal internalization interaction term was not a significant individual contributor to weight concerns, as indicated by a nonsignificant change in R2 from Step 2 to Step 3, R2 = .00, F(1, 659) = 2.31, p = .13. Discussion The present study examined the relationships between ethnic identity, thin-ideal internalization, and eating and weight concerns in a diverse sample of college women and yielded some important findings. First, significant differences emerged on the major study variables among ethnic groups. Specifically, European Americans scored lowest on ethnic identity and highest on eating and weight concerns, whereas African Americans scored lowest on thin-ideal internalization. These findings are consistent with existing literature on ethnic differences in levels of eating pathology and endorsement of mainstream media (Warren et al., 2005; Warren, Holland, Billings, & Parker, 2012) and ethnic identity (Schwartz et al., 2012; Turnage, 2004). Although mean ethnic group differences emerged
among the major study variables, it is important to note that the strength of the correlations between the study variables did not differ across ethnic groups. Therefore, the variables were related to the same extent for all women, regardless of their ethnic group. Second, thin-ideal internalization positively correlated with both eating and weight concerns in the overall sample as well as in each ethnic group. This finding is consistent with expectations, as well as with a relatively large body of previous findings (Homan, 2010; Levine & Murnen, 2009; Scharrer, 2013). Additionally, ethnic identity negatively correlated with eating and weight concerns in the overall sample. This finding supports a small body of extant findings on the negative association between ethnic identity and eating pathology (e.g., Rogers et al., 2010; Williams, 2009). However, these relationships were not statistically significant when examined by ethnic group, with the exception of the significant negative correlation between ethnic identity and weight concerns in Latinas. Furthermore, while significant, these correlations were small in effect size (i.e., accounting for approximately 1% of variance in the overall sample and 3% in Latinas). Third, in the regression analyses ethnic identity was a statistically significant predictor of both eating and weight concerns and the relationship between thin-ideal internalization and eating concerns was qualified by a significant interaction with ethnic identity. Although small in effect size, as hypothesized, that relationship was stronger for individuals with lower ethnic identity. The thin-ideal internalization by ethnic identity interaction was not a significant predictor of weight concerns. These data suggest that ethnic identity can serve as a protective factor against eating pathology, which warrants attention in clinical practice and research (Rogers et al., 2010; Williams, 2009). Limitations Despite the importance of these findings, they must be considered in light of several limitations. First, as the study involved correlational measures, we cannot make causal inferences or ensure the direction of the relationships in question. In addition, our measure lacked validity questions (e.g., “Please mark Strongly Agree if you are paying attention”). As a result, we cannot say with absolute confidence that participants responded truthfully
444
L.M. Rakhkovskaya, C.S. Warren / Body Image 11 (2014) 438–445
and attentively. We also did not counterbalance the measures. As such, the order of the questionnaires as well as response fatigue may have influenced responding. Second, the thin-ideal internalization by ethnic identity interaction accounted for only 1% of the variance in eating concerns. That said, McClelland and Judd (1993) outlined the difficulties of detecting significant interactions in correlational research. Specifically, interaction terms are unlikely to explain significant and/or large proportions of the variables in criterion variables, unless the study design oversamples. As such, small, significant interaction terms can nevertheless be practically meaningful in cross-sectional community studies, such as ours (McClelland & Judd, 1993). Third, we examined these constructs in a sample of college women from four major U.S. ethnic groups. As such, the results may not be generalizable to other ethnicities (e.g., Native American), other age groups (e.g., older adults) or men. The study also neglected to account for some within-group differences (e.g., Japanese American vs. Korean American), resulting in potentially overgeneralized findings. Nevertheless, we lacked the statistical power to examine ethnic differences in the moderating effects of ethnic identity on the relationship between thin-ideal internalization and eating concerns. As such, a replication of this study in a larger, ethnically diverse sample is warranted. Finally, our study was conducted at an urban university in the Southwestern U.S., in a city with a highly hyper-sexualized microculture. As such, our participants may have endorsed higher levels of thin-ideal internalization than most U.S. college students. In addition, participants attend one of the most ethnically diverse universities in the U.S. As a result, we cannot generalize our findings to other geographic regions and other (predominantly European American) universities in the United States. Implications and Future Directions Despite these limitations, the present study’s findings shed light on ethnic identity as potentially protective against eating and weight concerns. Furthermore, results indicate ethnic differences in levels of ethnic identity, thin-ideal internalization, eating concerns, and weight concerns. Accordingly, future research should examine ethnic differences in ethnic identity as a predictor for eating and/or weight concerns, as well as a moderator of the relationship between thin-ideal internalization and eating and/or weight concerns. Additionally, future studies should explore these relationships in less populous ethnic groups (e.g., Native Americans), within major ethnic/cultural groups (e.g., Jewish Americans), in different age groups (e.g., older adults), and in men. Furthermore, an examination of the relationships between ethnic identity and other risk-factors and symptoms of eating pathology, such as dissatisfaction with specific, racially salient areas of the body (i.e., hair texture, complexion) is of interest. Notably, Warren (2012) showed that Latina women report higher dissatisfaction with their eyes and nose, compared to European American and African American women. In addition, European American and Latina women reported higher dissatisfaction with their overall facial features, lips, lower body, and overall body than African American women. Warren (2012) found ethnic identity negatively predicted body dissatisfaction in most appearance areas. Given these findings, it is possible that, for some ethnic groups, ethnic identity may be protective against dissatisfaction with racially salient body areas in addition to or in lieu of overall body dissatisfaction. Future research examining ethnic identity, racially salient body area dissatisfaction and their potential relationships to thinideal culture is warranted. Additionally, the meaning of the construct of ethnic identity remains somewhat unclear for European American women. In this study, European American participants reported very low levels
of ethnic identity. This is unsurprising given the fact that European Americans are the ethnic majority in the U.S. and often do not endorse race or ethnicity as a part of their core identity (Rodriguez, Schwartz, & Krauss Whitbourne, 2010). As such, an examination of cultural factors that are more salient to identity in European Americans may be more beneficial. One such factor is American identity, or a sense of identifying with and attachment to the U.S. (Schildkraut, 2007). As individuals of European descent were historically in the majority in the U.S., European American values and cultural attributes largely comprise American identity today (i.e., American = White association; Devos & Banaji, 2005; Devos & Heng, 2009). Although research on the relationships between American identity and eating pathology is lacking, a similarly protective effect for European Americans with strong American identity is theoretically possible. Finally, we hope that our findings will inform culturally sensitive interventions for women with disordered eating symptoms. Future clinical directions include assessments of ethnic identity in determination of potential risk or protective factors. In addition, future interventions may focus on cultural beauty norms (e.g., a curvaceous body shape, good style, etc. for African American women), rather than unpacking majority culture messages (i.e., the thin ideal).
References Aiken, L. S., & West, S. G. (1991). Multiple regression: Testing and interpreting interactions. Thousand Oaks, CA: Sage Publications. Barry, D. T., & Garner, D. M. (2001). Eating concerns in East Asian immigrants: Relationship between acculturation, self-construal, ethnic identity, gender, psychological functioning, and eating concerns. Eating and Weight Disorders, 6, 90–98. http://dx.doi.org/10.1007/BF03339757 Brownell, K. D., Kersh, R., Ludwig, D. S., Post, R. C., Puhl, R. M., Schwartz, M. B., & Willett, W. C. (2010). Personal responsibility and obesity: A constructive approach to a controversial issue. Health Affairs (Project Hope), 29, 379–387. http://dx.doi.org/10.1377/hlthaff.2009.0739 Calogero, R. M., Davis, W. N., & Thompson, J. K. (2004). The Sociocultural Attitudes Toward Appearance Questionnaire (SATAQ-3): Reliability and normative comparisons of eating disordered patients. Body Image, 1, 193–198. http://dx.doi.org/10.1016/j.bodyim.2004.01.004 Chamorro, R., & Flores-Ortiz, Y. (2000). Acculturation and disordered eating patterns among Mexican American women. International Journal of Eating Disorders, 28, 125–129. http://dx.doi.org/10.1002/(SICI)1098108X(200007)28:1<125::AID-EAT16>3.0.CO;2-9 Claudat, K., Warren, C. S., & Durette, R. T. (2012). The relationships between body surveillance, body shame, and contextual body concern during sexual activities in ethnically diverse female college students. Body Image, 9, 448–454. http://dx.doi.org/10.1016/j.bodyim.2012.05.007 De Onis, M., & Habicht, J. P. (1996). Anthropometric reference data for international use: Recommendations from a World Health Organization Expert Committee. American Journal of Clinical Nutrition, 64, 650–658. Devos, T., & Banaji, M. R. (2005). American = White? Journal of Personality and Social Psychology, 88, 447–466. http://dx.doi.org/10.1037/0022-3514.88.3.447 Devos, T., & Heng, L. (2009). Whites are granted the American identity more swiftly than Asians: Disentangling the role of automatic and controlled processes. Social Psychology, 40, 192–201. http://dx.doi.org/10.1027/1864-9335.40.4.192 Fairburn, C., & Beglin, S. (1994). Assessment of eating disorders: Interview or self-report questionnaire? International Journal of Eating Disorders, 16, http://dx.doi.org/10.1002/1098-108X(199412)16:4<363::AID363–370. EAT2260160405>3.0.CO;2-# Ferguson, C. J. (2013). In the eye of the beholder: Thin-ideal media affects some, but not most, viewers in a meta-analytic review of body dissatisfaction in women and men. Journal of Popular Media Culture, 2, 20–37. http://dx.doi.org/10.1037/a0030766 Grabe, S., & Hyde, J. S. (2006). Ethnicity and body dissatisfaction among women in the United States: A meta-analysis. Psychological Bulletin, 132, 622–640. http://dx.doi.org/10.1037/00332909.132.4.622 Hall, C. (1995). Asian eyes: Body image and eating disorders of Asian and Asian American women. Eating Disorders: The Journal of Treatment & Prevention, 3, 8–19. http://dx.doi.org/10.1080/10640269508249141 Helms, J. E. (1990). Black and White racial identity: Theory, research, and practice. New York, NY, England: Greenwood Press. Henrickson, H. C., Crowther, J. H., & Harrington, E. F. (2010). Ethnic identity and maladaptive eating: Expectancies about eating and thinness in African American women. Cultural Diversity Ethnic Minority Psychology, 16, 87–93. http://dx.doi.org/10.1037/a0013455
L.M. Rakhkovskaya, C.S. Warren / Body Image 11 (2014) 438–445 Homan, K. (2010). Athletic-ideal and thin-ideal internalization as prospective predictors of body dissatisfaction, dieting, and compulsive exercise. Body Image, 7, 240–245. http://dx.doi.org/10.1016/j.bodyim.2010.02.004 Levine, M. P., & Murnen, S. K. (2009). Everybody knows that mass media are/are not [pick one] a cause of eating disorders: A critical review of evidence for a causal link between media, negative body image, and disordered eating in females. Journal of Social and Clinical Psychology, 28, 9–42. http://dx.doi.org/10.1521/jscp.2009.28.1.9 Lowry, R. (2001). VassarStats: Web Site for Statistical Computation. (Website). Retrieved from: http://faculty.vassar.edu/lowry/vsleft.html Luce, K. H., & Crowther, J. H. (1999). The reliability of the Eating Disorder Examination— Self-Report Questionnaire Version (EDE-Q). International Journal of Eating Disorders, 25, 349–351. http://dx.doi.org/10. 1002/(SICI)1098-108X(199904)25:3<349::AID-EAT15>3.0.CO;2-M Martinez, R. O., & Dukes, R. L. (1997). The effects of ethnic identity, ethnicity, and gender on adolescent well-being. Journal of Youth and Adolescence, 26, 503–516. http://dx.doi.org/10.1023/A:1024525821078 McClelland, G. H., & Judd, C. M. (1993). Statistical difficulties of detecting interactions and moderator effects. Psychological Bulletin, 114, 376–390. http://dx.doi.org/10.1037/0033-2909.114.2.376 Mintz, L. B., & Kashubeck, S. (1999). Body image and disordered eating among Asian American and Caucasian college students. Psychology of Women Quarterly, 23, 781–796. http://dx.doi.org/10.1111/j.1471-6402.1999.tb00397.x Nagel, J. (1994). Constructing ethnicity: Creating and recreating ethnic identity and culture. Social Problems, 41, 152–176. http://dx.doi.org/10.1525/sp.1994.41.1.03x0430n Nouri, M., Hill, L. G., & Orrell-Valente, J. K. (2011). Media exposure, internalization of the thin ideal, and body dissatisfaction: Comparing Asian American and European American college females. Body Image, 8, 366–372. http://dx.doi.org/10.1016/j.bodyim.2011.05.008 Overstreet, N. M., Quinn, D. M., & Agocha, V. B. (2010). Beyond thinness: The influence of a curvaceous body ideal on body dissatisfaction in black and white women. Sex Roles, 63, 91–103. http://dx.doi.org/10.1007/s11199-010-9792-4 Parker, S., Nichter, M., Nichter, M., Vuckovic, N., Sims, C., & Ritenbaugh, C. (1995). Body image and weight concerns in African American and White adolescent females: Differences that make a difference. Human Organization, 54, 103–112. Phinney, J. S. (1992). The Multigroup Ethnic Identity Measure: A new scale for use with diverse groups. Journal of Adolescent Research, 7, 156–176. http://dx.doi.org/10.1177/074355489272003 Phinney, J. S. (1996). When we talk about American ethnic groups, what do we mean? American Psychologist, 51, 918–927. http://dx.doi.org/10.1037/0003-066X.51.9.918 Ponterotto, J. G., Gretchen, D., Utsey, S. O., Stracuzzi, T., & Saya, R., Jr. (2003). The Multigroup Ethnic Identity Measure (MEIM): Psychometric review and further validity testing. Educational and Psychological Measurement, 63, 502–515. http://dx.doi.org/10.1177/0013164403063003010 Poran, M. A. (2002). Denying diversity: Perceptions of beauty and social comparison processes among Latina, Black, and White women. Sex Roles, 47, 65–81. http://dx.doi.org/10.1023/A:1020683720636 Poran, M. A. (2006). The politics of protection: Body image, social pressures, and the misrepresentation of young Black women. Sex Roles, 55, 739–755. http://dx.doi.org/10.1007/s11199-006-9129-5 Rhea, D. J., & Thatcher, W. (2013). Ethnicity, ethnic identity, self-esteem, and at-risk eating disordered behavior differences of urban adolescent females. Eating Disorders: The Journal of Treatment & Prevention, 21, 223–237. http://dx.doi.org/10.1080/10640266.2013.779177 Rodriguez, L., Schwartz, S. J., & Whitbourne, S. (2010). American identity revisited: The relation between national, ethnic, and personal identity in a multiethnic sample of emerging adults. Journal of Adolescent Research, 25, 324–349. http://dx.doi.org/10.1177/0743558409359055 Rogers, Wood, N. A., & Petrie, T. A. (2010). Body dissatisfaction, ethnic identity, and disordered eating among African American women. Journal of Counseling Psychology, 57, 141–153. http://dx.doi.org/10.1037/a0018922
445
Rubin, L. R., Fitts, M. L., & Becker, A. E. (2003). ‘Whatever feels good in my soul’: Body ethics and aesthetics among African American and Latina women. Culture, Medicine and Psychiatry, 27, 49–75. http://dx.doi.org/10. 1023/A:1023679821086 Santiago-Rivera, A. L., Arredondo, P., & Gallardo-Cooper, M. (2002). Counseling Latinos and la familia: A practical guide. Thousand Oaks, CA: Sage Publications, Inc. Scharrer, E. L. (2013). 15 Representations of gender in the media. In K. E. Dill (Ed.), The Oxford handbook of media psychology (pp. 267–273). New York, NY: Oxford University Press. Schildkraut, D. J. (2007). Defining American identity in the twenty-first century: How much ‘there’ is there? The Journal of Politics, 69, 597–615. http://dx.doi.org/10.1111/j.1468-2508.2007.00562.x ˜ Schwartz, S. J., Park, I. K., Huynh, Q., Zamboanga, B. L., Umana-Taylor, A. J., Lee, R., & Agocha, V. (2012). The American Identity Measure: Development and validation across ethnic group and immigrant generation. Identity: An International Journal of Theory and Research, 12, 93–128. http://dx.doi.org/10.1080/15283488.2012.668730 Soh, N. L. W., & Walter, G. (2013). Publications on cross-cultural aspects of eating disorders. Journal of Eating Disorders, 1, 4. http://dx.doi.org/10.1186/2050-2974-1-4 Soper, D. S. (2010). Interaction (Version 1.7.2211). (Software). Retrieved from http://www.danielsoper.com/interaction/ Stein, K. F., Corte, C., & Ronis, D. L. (2010). Personal identities and disordered eating behaviors in Mexican American women. Eating Behaviors, 11, 197–200. http://dx.doi.org/10.1016/j.eatbeh.2010.02.001 Stojek, M., Fischer, S., & Collins, B. (2010). Thinness and restricting expectancies mediate the influence of ethnic identity on bulimic Personality and Individual Differences, 49, 102–106. symptoms. http://dx.doi.org/10.1016/j.paid.2010.03.012 Thompson, J. K., van den Berg, P., Roehrig, M., Guarda, A. S., & Heinberg, L. J. (2004). The Sociocultural Attitudes Towards Appearance Scale-3 (SATAQ-3): Development and validation. International Journal of Eating Disorders, 35, 293–304. http://dx.doi.org/10.1002/eat.10257 Turnage, B. F. (2004). Influences on adolescent African American females’ global self-esteem: Body image and ethnic identity. Journal of Ethnic & Cultural Diversity in Social Work: Innovation in Theory, Research & Practice, 13, 37–45. http://dx.doi.org/10.1300/J051v13n04 02 ˜ Umana-Taylor, A. J., Wong, J. J., Gonzales, N. A., & Dumka, L. E. (2012). Ethnic identity and gender as moderators of the association between discrimination and academic adjustment among Mexican-origin adolescents. Journal of Adolescence, 35, 773–786. http://dx.doi.org/10.1016/j.adolescence.2011.11.003 Warren, C. S. (2012). Body area dissatisfaction in White, Black and Latina female college students in the USA: An examination of racially salient appearance areas and ethnic identity. Ethnic and Racial Studies, 37, 537–556. http://dx.doi.org/10.1080/01419870.2012.716520 Warren, C. S., Gleaves, D. H., Cepeda-Benito, A., del Carmen Fernandez, M., & Rodriguez-Ruiz, S. (2005). Ethnicity as a protective factor against internalization of a thin ideal and body dissatisfaction. International Journal of Eating Disorders, 37, 241–249. http://dx.doi.org/10.1002/eat.20102 Warren, C. S., Gleaves, D. H., & Rakhkovskaya, L. M. (2013). Score reliability and factor similarity of the Sociocultural Attitudes Towards Appearance Questionnaire3 (SATAQ-3) among four ethnic groups. Journal of Eating Disorders, 1, 14. http://dx.doi.org/10.1186/2050-2974-1-14 Warren, C. S., Holland, S., Billings, H., & Parker, A. (2012). The relationships between fat talk, body dissatisfaction, and drive for thinness: Perceived stress as a moderator. Body Image, 9, 358–364. http://dx.doi.org/10.1016/j.bodyim.2012.03.008 White, E. K., & Warren, C. S. (2013). Body checking and avoidance in ethnically diverse female college students. Body Image, 10, 583–590. http://dx.doi.org/10.1016/j.bodyim.2013.04.003 Williams, N. M. (2009). The moderating effects of perfectionism and ethnic identity on the relationship between sociocultural pressure and body dissatisfaction (Unpublished master’s thesis). Kent, OH: Kent State University.