Eating Behaviors 15 (2014) 151–158
Contents lists available at ScienceDirect
Eating Behaviors
Prevalence of body dissatisfaction among a United States adult sample☆ Elizabeth A. Fallon a,b,⁎, Brandonn S. Harris c,d, Paige Johnson b a
Department of Kinesiology & Health, Georgia State University, Atlanta, GA, United States Department of Kinesiology, Kansas State University, Manhattan, KS, United States c Department of Health and Kinesiology, Georgia Southern University, Statesboro, GA, United States d Department of Special Education, Counseling, and Student Affairs, Kansas State University, Manhattan, KS, United States b
a r t i c l e
i n f o
Article history: Received 2 October 2013 Received in revised form 4 November 2013 Accepted 20 November 2013 Available online 5 December 2013 Keywords: Body dissatisfaction Body image Prevalence Adult United States
a b s t r a c t Body dissatisfaction (BD) is a primary determinant of eating disorders and has been linked to chronic disease via decreased likelihood of cancer screening self-exams and smoking cessation. Yet, there are few recent estimates of the prevalence of BD among United States adults. Using an internet-based, opt-in, cross-sectional survey, United States adults (N = 1893) completed assessments of demographic variables, body areas satisfaction, appearance evaluation, fitness evaluation, health evaluation, and overweight preoccupation. Results revealed that the range of BD is 13.4%–31.8% among women and 9.0%–28.4% among men. Compared to previous assessments of prevalence (1973, 1986, 1995, 1997), the prevalence of BD among United States adults may have plateaued or declined over time. © 2013 Elsevier Ltd. All rights reserved.
1. Introduction Body image is defined as a persons body-related self-perceptions and self-attitudes, including thoughts, feelings, and behaviors (Cash, 2003). In some cases, extreme levels of body dissatisfaction (BD) can result in eating disorders (e.g., anorexia nervosa, bulimia nervosa), which have a lifetime prevalence of 2.5% and 0.8% for United States women and men, respectively (Hudson, Hiripi, Pope, & Kessler, 2007). For those who seek treatment, annual treatment costs range from $US 1288 to $US 8042 per person, per year (Stuhldreher et al., 2012). BD is also known to negatively influence behavioral risk factors for chronic disease, which affect an even greater proportion of the US population. For example, breast cancer is the most common cancer among women, with a lifetime prevalence of 12.3% (Howlader et al., 2013). A recent literature review demonstrates that BD is associated with lower likelihood of engaging in breast cancer self-exams (Ridolfi & Crowther, 2013), which could provide early detection of the disease. BD is also associated with lower likelihood of smoking cessation (King et al., 2005), where smoking costs $US 96.8 billion annually in lost productivity and is
☆ The data for this study was collected at Kansas State University, but analyzed and the manuscript composed at Georgia State University. Elizabeth A. Fallon is now at Georgia State University and Brandonn S. Harris is now at Georgia Southern University. Paige Johnson is a recent graduate of Kansas State University, and does not currently maintain a university affiliation.We thank the survey respondents for their time and the students who helped collect the data; specifically, laboratory assistants Joseph Lightner, Abby Banks, Madelaine Ellison, Annie Pipes, Tracy Engstrom, and Laura Moluf. ⁎ Corresponding author at: P.O. Box 3975, Atlanta, GA 30302, United States. E-mail address: [email protected] (E.A. Fallon). 1471-0153/$ – see front matter © 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.eatbeh.2013.11.007
responsible for almost 30% of cancer deaths (Centers for Disease Control, Prevention, 2008). With such robust potential to influence health care costs as well as an individual's health and quality of life across the lifespan, it is imperative for behavioral scientists, clinicians, and public health professionals to be informed of the current overall prevalence of BD among the United States population, changes in BD prevalence over time, as well as understand which sub-populations are at risk. Salient to this study are four preceding national surveys assessing body image/BD among United States adults (Berscheid, Walster, & Bohrnstedt, 1973; Cash & Henry, 1995; Cash, Winstead, & Janda, 1986; Garner, 1997). Since 1972, a series of surveys have documented an increase in dissatisfaction in overall appearance among both men (15% to 43%) and women (23% to 56%). This often-cited research has led many to conclude that there is a ‘normative discontent’ or an ‘epidemic’ of BD in the United States, especially among women (Frederick, Jafary, Gruys, & Daniels, 2012; Rodin, Silberstein, & Striegel-Moore, 1984; Tantleff-Dunn, Barnes, & Larose, 2011). Since Garner (1997), however, very few studies have used national surveys of United States adults, and the few containing national samples, have either not used psychometrically valid measurement tools (Frederick, Peplau, & Lever, 2006), have used single-item measures of body satisfaction (Kruger, Lee, Ainsworth, & Macera, 2008), or have focused on specific United States subgroups (Peplau et al., 2009). Thus, a more recent assessment of overall prevalence of BD among United States adults is timely and the purpose of this study is threefold: a. To provide an overall estimate of the national prevalence of BD among United States men and women,
152
E.A. Fallon et al. / Eating Behaviors 15 (2014) 151–158
b. To examine differences in BD across population subgroups (e.g., sex, race/ethnicity, age, body mass index category), c. To compare the rates of BD of the present sample to previously published national samples.
Table 1 Sample characteristics by participant sex. Women (n = 1246) M
2. Method
Age BMI
Men (n = 647) SD
42.50 25.86
14.00 6.75
M 47.77 26.90
SD 16.27 5.38
2.1. Participants N
Demographic information for adult volunteers (N = 1893; age range 18–90 years) who completed an online survey are presented in Table 1. 2.2. Procedure Prior to data collection, Institutional Review Board approval was obtained for an internet-based, opt-in survey taking participants approximately 20–30 min to complete.1 To recruit United States adults across a wide geographical area, e-mail addresses were collected from publicly available websites in the United States. These websites primarily consisted of, but were not limited to, colleges/universities, public primary schools, state/local government, and faith-based organizations. Invitation emails for the study titled “Attitudes Towards Health Behaviors” contained the IRB informed consent information (e.g., general purpose of the study, IRB and contact information, and that the data would be used for research) and a link to the survey. While the survey respondents could choose to submit the survey anonymously, they could also choose to include their contact information to win a monthly drawing for $15 to a national chain store. This drawing, as well as two followup e-mail reminders sent 1 and 2 weeks after the initial invitation were intended to increase response rate and survey completion rate. After the data collection period (August 2010–April 2011) ended, the data were de-identified. Over the course of the recruitment period, 2665 individuals accessed the online survey. Using a “completers only” analysis for this study, participants failing to complete items assessing demographic variables or the multidimensional body self-relations survey were excluded from the analysis. The final sample for this study is 1893 adults, representing 37 US states. 2.3. Measures 2.3.1. Demographics Consistent with the Behavior Risk Factor Surveillance (U.S. Department of Health and Human Services and Centers for Disease Control, Prevention (2010)) system, participants were asked to self-report their birth date (month, day, year), sex, race, ethnicity, and education level. Age groups were created to be consistent with those from previous research (Cash & Henry, 1995). Self-reported height and weight measurements were used to compute body mass index (BMI; weight [kg]/height [m2]). Subsequently, four BMI categories were created: underweight (BMI b 18.5), healthy weight (BMI 18.5 to b25), overweight (BMI 25.0 to b 30.0), and obese (BMI ≥ 30.0; U.S. Department of Health and Human Services, National Institutes of Health, & National Heart Lung & Blood Institute, 1998). 2.3.2. Body areas satisfaction Satisfaction with specific areas of the body was assessed using the Body Areas Satisfaction subscale of the Multidimensional Body-Self 1
According to publication guidelines (American Psychological Association, 2010), we note that this data set comes from a previously published manuscript (Johnson, Fallon, Harris, & Burton, 2013), which focused on the associations between body satisfaction physical activity behavior change. The purposes of each manuscript necessitated multiple manuscripts for multiple audiences. Furthermore, the data from fitness evaluation, appearance evaluation, health evaluation and health orientation are novel and have not been published elsewhere.
Age 18–24 year 25–34 year 35–44 year 45–54 year 55–64 year 65 and over BMI Underweight Normal weight Overweight Obese Race/ethnicity White/Caucasian Black/African American Asian Hispanic Native American/Native Hawaiian/Pacific Islander Other/don't know/prefer not to answer Education Less than 4-year college Degree 4-year college degree Cohabitation status Married/cohabitating Not married/cohabitating
(%)
N
(%)
107 341 212 242 290 54
8.6 27.4 17.0 19.4 23.3 4.3
51 116 105 99 169 107
7.9 17.9 16.2 15.3 26.1 16.5
32 673 301 240
2.6 54.0 24.2 19.3
6 262 232 147
0.9 40.5 35.9 22.7
1074 30 47 37
86.2 2.4 3.8 3.0
558 9 32 13
86.2 1.4 4.9 2.0
21 37
1.7 3.0
5 30
0.8 4.6
188 1055
15.1 84.7
56 590
8.7 91.2
802 435
64.4 34.9
474 169
73.3 26.1
Relations Questionnaire (MBSRQ; Cash, 2000). This measurement tool uses a 5-point Likert-type scale ranging from very dissatisfied (1) to very satisfied (5) to measure participants' satisfaction with nine specific aspects of their body (e.g., face, height, hair, muscle tone). Using this tool, BD was assessed in two ways. First, for each specific body area, individuals reporting ‘very dissatisfied’ or ‘dissatisfied’ were categorized as having BD. Second, the scores for each of the 9 items were averaged, with higher scores representing higher body satisfaction. According to a protocol from Cash and colleagues (Cash & Henry, 1995; Cash et al., 1986), a mean score of b 3.0 is defined as BD. According to the Frederick and colleagues' protocol (Frederick, Forbes, Grigorian, & Jarcho, 2007; Peplau et al., 2009), a mean score of b2.75 is defined as BD. For this study, we present both protocol cut-off points (Fiske, Fallon, Blissmer, & Redding, under review). Finally, for this study, the internal consistency was good (α = .83). 2.3.3. Appearance evaluation Feelings of physical attractiveness or satisfaction with one's looks are measured using a 7-item Likert-type scale ranging from Definitely Disagree (1) to Definitely Agree (5). Scores on each item are averaged to create a mean score, where higher scores are indicative of greater satisfaction with appearance (Cash, 2000). For this study, the internal consistency was good (α = .89) and we present both the Cash (scale means b 3.0) and Frederick (scale means b2.75) cut-off points. 2.3.4. Overweight preoccupation Overweight preoccupation is a construct reflecting fat anxiety, weight vigilance, dieting, and eating restraint (Cash, 2000). It is assessed using a 4-item Likert-type scale ranging from Definitely Disagree (1) to Definitely Agree (5). Scores on each item are averaged to create a mean score, where higher scores are indicative of greater overweight preoccupation, and therefore greater BD (Cash, 2000). For this study, we present
E.A. Fallon et al. / Eating Behaviors 15 (2014) 151–158
both the Cash (scale means b 3.0) and Frederick (scale means b 2.75) cut-off points. Finally, for this study, the internal consistency was questionable (α = .61). 2.3.5. Fitness evaluation Feelings of being physically fit or unfit were assessed using a 3item Likert-type scale from Definitely Disagree (1) to Definitely Agree (5). Scores on each item are averaged to create a mean score. Individuals scoring highly on this subscale of the MBSRQ regard themselves as physically fit, “in shape”, or athletically active and competent (Cash, 2000). For this study, we present both the Cash (scale means b3.0) and Frederick (scale means b2.75) cut-off points. Finally, for this study, the internal consistency was acceptable (α = .79). 2.3.6. Health evaluation Using a 6-item Likert-type scale from Definitely Disagree (1) to Definitely Agree (5), the Health Evaluation subscale of the MBSRQ assesses feelings of physical health and/or the freedom from physical illness. Higher scores are indicative of feeling the body is in good health (Cash, 2000). Consistent with the other subscales, we present both the Cash (scale means b 3.0) and Frederick (scale means b2.75) cut-off points. The internal consistency for this study was good (α = .80). 2.4. Data analysis Data were examined for outliers, linear dependency, skewness and kurtosis. Frequencies, means, and standard deviations for demographic variables were used to describe the study sample (See Table 1). Means and standard deviations for each of the body image subscales by demographic variable were also calculated (Tables 2 & 3). To provide an estimate of national prevalence of BD among United States adult men and women, we compared the use of three protocols from previous literature. Specifically, the frequency of men and women reporting being ‘very dissatisfied’ or ‘dissatisfied’ with specific body areas is reported (Berscheid et al., 1973; Cash & Henry, 1995; Cash et al., 1986; Garner, 1997; see Table 4). Additionally, we calculated the means and standard deviations for each of the five MBSRQ subscales (body areas satisfaction, appearance
153
evaluation, overweight preoccupation, fitness evaluation, and health evaluation) and calculated BD prevalence based on the Cash (scale mean b3.0; Cash & Henry, 1995; Cash et al., 1986) and Frederick (scale mean b2.75; Frederick et al., 2007; Peplau et al., 2009) protocols. See Tables 5 & 6 for women and men, respectively. To examine differences in each body image measure across population subgroups, we conducted a series of 2-way MANOVAs, with Tukey's post hoc tests, when appropriate. Finally, we visually compared our results to those of previous national United States adult surveys (see Tables 7 & 8).
3. Results More women than men enrolled in the study, and the sample was predominantly white/Caucasian, middle aged, overweight and married/cohabitating. If frequency counts are used to define prevalence of body dissatisfaction, only 9% of men and 13.4% of women were dissatisfied with their overall appearance (see Table 4). If the mean score of the body areas satisfaction measure is used, the estimated prevalence of BD doubles, ranging between 12.2% and 21.2% for men, and 19.2–28.7% for women, depending on choosing more liberal (Cash) or more conservative (Frederick) cut-off protocol. If the appearance evaluation mean score is used, the estimated range of prevalence is 22.9%–28.4% and 26.2%–31.8%, for men and women, respectively (see Tables 5 & 6). When examining differences in MBSRQ subcales across age groups, the Box's Test was significant, violating the assumption of homogeneity of variance [F (165, 659,090) = 1.44, p b .001]. Therefore, Pillai's Trace was used to interpret the MANOVA results. The 2 (Sex) X 6 (Age Category) MANOVA for the five MBSRQ subscales revealed a statistically significant main effect for sex [Pillai's Trace = .06, F (5, 1877) = 25.03, p b .001, η2 = .06], and a significant main effect for age category [Pillai's Trace = .07, F (25, 9405) = 5.58, p b 0.001, η 2 = .02], but no statistically significant interaction emerged between sex and age [Pillai's Trace = .02, F (25, 9405) = 1.19, p = 0.23, η2 = .23]. Follow-up univariate analyses revealed that compared to men, women had lower body areas satisfaction [F (1, 1881) = 19.03, p b .001, η2 = .01, d = .23], lower fitness evaluation [F (1, 1881) = 31.07, p b .001, η2 = .00, d = .28], and greater overweight preoccupation [F (1, 1881) = 73.20, p b .001, η2 = .00,
Table 2 Means and standard deviations for MBSRQ subscales for women across demographic characteristics. Body satisfaction
All women (n = 1246) Age 18–24 years (n = 107) 25–34 years (n = 341) 35–44 years (n = 212) 45–54 year (n = 242) 55–64 years (n = 290) 65 and over (n = 54) BMI Underweight (n = 32) Normal weight (n = 673) Overweight (n = 301) Obese (n = 240) Race White/Caucasian (n = 1074) Black/African American (n = 30) Asian (n = 47) Hispanic (n = 37) Native American/Native Hawaiian/Pacific Islander (n = 21) Other/don't know/prefer not to answer (n = 37)
Appearance evaluation
Overweight preoccupation
Fitness evaluation
Health evaluation
M
SD
M
SD
M
SD
M
SD
M
SD
3.30
0.67
3.27
0.83
2.64
0.8
3.32
0.97
3.83
0.77
3.48 3.46 3.30 3.16 3.15 3.34
0.63 0.66 0.70 0.64 0.65 0.57
3.58 3.52 3.19 3.11 3.06 3.20
0.80 0.83 0.84 0.80 0.79 0.75
2.64 2.57 2.57 2.69 2.77 2.53
0.76 0.84 0.80 0.79 0.75 0.80
3.79 3.39 3.37 3.22 3.13 3.06
0.94 1.01 0.94 0.91 0.94 0.89
3.92 3.86 3.68 3.79 3.86 3.95
0.64 0.73 0.86 0.74 0.79 0.77
3.54 3.55 3.13 2.74
0.66 0.60 0.55 0.59
3.50 3.64 3.04 2.47
0.85 0.67 0.70 0.74
2.15 2.52 2.83 2.81
0.85 0.79 0.74 0.78
3.31 3.45 3.35 2.89
1.06 0.95 0.91 0.97
3.74 4.03 3.79 3.31
0.89 0.68 0.77 0.73
3.29 3.24 3.27 3.31 3.59 3.38
0.66 0.69 0.56 0.70 0.67 0.82
3.25 3.45 3.43 3.35 3.22 3.52
0.83 0.74 0.75 0.92 0.88 1.00
2.63 2.53 2.63 2.78 3.10 2.73
0.79 0.80 0.85 0.78 0.85 1.00
3.29 3.79 3.37 3.56 3.25 3.49
0.96 0.81 0.85 1.04 1.12 1.10
3.83 3.85 3.78 3.93 3.70 3.81
0.77 0.65 0.65 0.87 0.86 0.88
154
E.A. Fallon et al. / Eating Behaviors 15 (2014) 151–158
Table 3 Means and standard deviations of MBSRQ scale scores for men across demographic characteristics.
Overall Men (n = 647) Age 18–24 year (n = 51) 25–34 year (n = 116) 35–44 year (n = 105) 45–54 year (n =99) 55–64 year (n = 169) 65 and over (n = 107) BMI Underweight (n = 6) Normal weight (n = 262) Overweight (n = 232) Obese (n = 147) Race White/Caucasian (n = 558) Black/African American (n = 9) Asian (n = 32) Hispanic (n = 13) Native American/native Hawaiian/Pacific Islander (n = 5) Other/don't know/prefer not to answer (n = 30)
Body satisfaction
Appearance evaluation
Overweight preoccupation
Fitness evaluation
M
SD
M
3.45
0.64
3.58 3.55 3.43 3.35 3.32 3.57
SD
M
SD
M
SD
M
SD
3.3
0.76
2.29
0.76
3.58
0.89
3.86
0.69
0.74 0.61 0.69 0.60 0.61 0.62
3.61 3.50 3.27 3.19 3.13 3.35
0.86 0.72 0.82 0.75 0.75 0.65
2.04 2.14 2.33 2.35 2.36 2.37
0.85 0.85 0.79 0.74 0.69 0.66
3.76 3.70 3.74 3.53 3.43 3.49
1.02 0.87 0.82 0.86 0.88 0.90
3.96 3.98 3.80 3.77 3.76 3.98
0.65 0.62 0.68 0.64 0.76 0.67
3.24 3.64 3.46 3.08
0.42 0.60 0.60 0.64
3.33 3.59 3.35 2.71
0.43 0.63 0.66 0.81
2.04 2.11 2.35 2.52
0.95 0.75 0.73 0.74
3.17 3.60 3.62 3.50
0.86 0.94 0.83 0.88
3.42 3.96 3.94 3.58
0.58 0.65 0.64 0.75
3.44 3.90 3.41 3.52 3.67 3.36
0.63 0.78 0.71 0.63 0.88 0.66
3.29 3.76 3.33 3.45 3.80 3.20
0.77 0.76 0.61 0.66 0.73 0.82
2.28 2.05 2.46 2.08 1.60 2.50
0.74 0.75 0.82 1.17 0.52 0.80
3.57 4.22 3.76 3.46 3.93 3.48
0.87 0.58 0.92 1.02 0.86 1.02
3.88 3.81 3.70 3.76 4.13 3.59
0.68 0.57 0.78 0.80 0.40 0.62
d = .45]. There were no differences between men and women for appearance evaluation [F (1, 1881) = 2.33, p = .13, η2 = .00, d = .03] or health evaluation [F (1, 1881) = 0.61, p = .43, η2 = .00, d = .04]. Follow-up univariate analyses revealed a main effect of age group for body areas satisfaction [F (5, 1881) = 10.66, p b .001, η 2 = .03], appearance evaluation [F (5, 1881) = 16.53, p b .001, η2 = .04], fitness evaluation [F (5, 1881) = 9.06, p b .001, η 2 = .02], health evaluation [F (5, 1881) = 3.61, p = .01, η 2 = .01], and overweight preoccupation [F (5, 1881) = 3.89, p = .01, η2 = .01]. Tukey's post hoc analyses, however, did not reveal any statistically significant differences among the age groups for any of the body image dependent variables (p's N .05). Trends were evident suggesting a curvilinear trajectory for health evaluation (p = .09) and body areas satisfaction (p = .08) that these constructs were generally higher among younger participants (18–24 years) and older adults (N65 years), while lower among the middle-aged. Examining differences in MBSRQ subscales across BMI category, the Box's Test was significant, violating the assumption of homogeneity of
Health evaluation
variance [F (105, 4608) = 1.89, p b .001]. Therefore, Pillai's Trace was used to interpret the MANOVA results. The 2 (Sex) × 4 (BMI category) MANOVA for the five MBSRQ subscales revealed a significant main effect for sex, and a significant main effect for BMI category [Pillai's Trace = .25, F (5, 1881) = 34.11, p b 0.001, η2 = .08], as well as a statistically significant sex by body mass index interaction [Pillai's Trace = .03, F (15, 5649) = 3.38, p b 0.001, η2 = .01]. The main effect for sex has already been noted, and would also be expected for this analysis. Despite a statistically significant main effect in multivariate analysis, Tukey's post hoc analyses did not reveal any differences among the BMI groups for any of the body image dependent variables (p's N .05). Follow-up analyses showed an interaction for body areas satisfaction [F (3, 1885) = 7.29, p b .001, η2 = .01], appearance evaluation [F (3, 1885) = 8.19, p b 001, η2 = .01], fitness evaluation [F (3, 1885) = 5.45, p = .001, η2 = .01], and health evaluation [F (3, 1885) =5.93, p = .001, η2 = .01], such that in the underweight category, women consistently report more positive body image than men. In the overweight and obese categories, however, men consistently report more positive body image than women. No significant interaction was
Table 4 Response frequencies and item means for body areas satisfaction among women and men.
Women (n = 1246) Face Hair Upper torso Middle torso Lower torso Muscle tone Height Weight Overall appearance Men (n = 647) Face Hair Upper torso Middle torso Lower torso Muscle tone Height Weight Overall appearance
Very dissatisfied
Dissatisfied
Very satisfied
Item mean
n
%
n
%
n
Neither %
n
Satisfied %
n
%
M
SD
15 24 68 226 133 91 27 197 25
1.2 1.9 5.5 18.1 10.7 7.3 2.2 15.8 2.0
106 100 215 457 382 339 90 380 142
8.5 8.0 17.3 36.7 30.7 27.2 7.2 30.5 11.4
224 169 309 222 261 345 257 227 324
18.0 13.6 24.8 17.8 20.9 27.7 20.6 19.0 26.0
729 641 553 281 388 409 498 363 684
58.5 51.4 44.4 22.6 31.1 32.8 40.0 29.1 54.9
172 312 101 60 82 62 374 69 71
13.8 25.0 8.1 4.8 6.6 5.0 30.0 5.5 5.7
3.75 3.90 3.32 2.59 2.92 3.01 3.88 2.78 3.51
0.84 0.93 1.03 1.16 1.14 1.05 0.99 1.19 0.84
10 22 23 87 13 21 9 64 9
1.5 3.4 3.6 13.4 2.0 3.2 1.4 9.9 1.4
42 89 105 206 58 127 59 187 49
6.5 13.8 16.2 31.8 9.0 19.6 9.1 28.9 7.6
145 136 175 161 134 180 160 134 180
22.4 21.0 27.0 24.9 20.7 27.8 24.7 20.7 27.8
347 277 276 160 315 265 251 203 350
53.6 42.8 42.7 24.7 48.7 41.0 38.8 31.4 54.1
103 123 68 33 127 54 168 59 59
15.9 19.0 10.5 5.1 19.6 8.3 26.0 9.1 9.1
3.76 3.60 3.40 2.77 3.75 3.32 3.79 3.01 3.62
0.85 1.05 0.96 1.12 0.94 0.99 0.98 1.17 0.81
E.A. Fallon et al. / Eating Behaviors 15 (2014) 151–158
155
Table 5 Comparison of Cash & Frederick recommended cut-off points for determining prevalence of BD among United States women.
Overall women (n = 1246) Age 18–24 years (n = 107) 25–34 years (n = 341) 35–44 years (n = 212) 45–54 years (n = 242) 55–64 years (n = 290) 65 and over (n = 54) BMI Underweight (n = 32) Healthy weight (n = 673) Overweight (n = 301) Obese (n = 240) Race White/Caucasian (n = 1074) Black/AA (n = 30) Asian (n = 47) Hispanic (n = 37) Native American/Native Hawaiian/Pacific Islander (n = 21) Other/don't know/prefer not to answer (n = 37)
Body satisfaction
Appearance evaluation
Overweight preoccupation
Health evaluation
Fitness evaluation
Cash
Frederick
Cash
Frederick
Cash
Frederick
Cash
Frederick
Cash
Frederick
(%)
(%)
(%)
(%)
(%)
(%)
(%)
(%)
(%)
(%)
28.7
19.2
31.8
26.2
38.6
18.2
13.0
10.4
30.2
30.2
19.6 20.2 30.2 36 35.9 24.1
11.2 12.9 19.3 25.6 25.2 13
18.7 19.6 37.3 40.1 39.7 39.7
16.8 17 28.8 33.5 33.4 22.2
37.4 33.1 36.3 43.8 43.8 33.3
14 18.5 15.6 19.8 21.4 11.1
6.5 9.7 16.5 15.7 14.8 11.1
4.7 7.9 14.6 12 11 9.3
15.9 28.7 27.4 31.8 36.6 37
15.9 28.7 274 31.8 36.6 37
18.8 14.1 34.2 64.2
9.4 7.4 22.6 49.2
25 14 42.2 69.6
21.9 10 33.2 63.8
18.8 30.5 50.8 48.8
6.3 15.2 22.9 22.5
12.5 6.4 15 29.2
12.5 4.6 12.6 23.3
31.3 24.5 28.2 48.3
31.3 24.5 28.2 48.3
29.2 30 25.5 29.7 19 21.6
19.4 23.3 17 18.9 14.3 16.2
32.6 23.3 21.3 29.7 38.1 27
27 16.7 14.9 27 23.8 27
37.9 40 36.2 45.9 61.9 40.5
17.6 13.3 19.1 24.3 38.1 21.6
13.4 10 4.3 13.5 19 10.8
10.5 6.7 4.3 10.8 19 10.8
31.4 10 21.3 33.3 32.4 30.2
31.4 10 21.3 18.9 33.3 32.4
found for overweight preoccupation [F (3, 1885) = 1.37, p = .25, η2 = .00]. When examining differences in MBSRQ subscale scores by race/ ethnicity, the Box's Test was significant, violating the assumption of homogeneity of variance [F (150, 18662) = 1.52, p b .001]. Thus, Pillai's Trace was used to interpret the MANOVA results. The 2 (Sex) × 6 (Race/ethnicity) MANOVA for the five MBSRQ subscales revealed a significant main effect of race/ethnicity [Pillai's Trace = .02, F (25, 9405) = 1.61, p = .03, η2 = .00], but no statistically significant sex by race interaction [Pillai's Trace = .02, F (25, 9405) = 1.40, p = .09, η2 = .00]. Follow-up univariate analyses for race/ethnicity did not reveal a main effect for body areas satisfaction [F (5, 1881) = 1.06, p = .38, η2 = .00], appearance evaluation [F (5, 1881) = 1.30, p = .26, η2 = .00], fitness evaluation [F (5, 1881) = 0.85, p = .51, η2 = .00], or health evaluation [F (5, 1881) = 0.92, p = .47, η2 = .00]. A statistically significant difference did emerge for overweight
preoccupation suggesting that compared to non-Hispanic Whites, Black/African–American participants reported lower overweight preoccupation [F (5, 1881) = 2.45, p = .03, η 2 = .01, d = .13]. The Tukey's post hoc analysis, however, was not statistically significant for this interaction (p = .06). Finally, when comparing the results of this study to previously published national samples, it appears that BD prevalence among United States adults has plateaued, or perhaps, slightly decreased over time (see Tables 7 & 8). 4. Discussion The purpose of this research was to provide a recent overall estimate of the national prevalence of BD among United States men and women, examine differences in BD across population subgroups (e.g., sex, race/ ethnicity, age, body mass index category), and compare the rates of
Table 6 Comparison of Cash & Frederick recommended cut-off points for determining prevalence of BD among United States men.
Overall men (n = 647) Age 18–24 years (n = 51) 25–34 years (n = 116) 35–44 years (n = 105) 45–54 years (n = 99) 55–64 years (n = 169) 65 and over (n = 107) BMI Underweight (n = 6) Healthy weight (n = 262) Overweight (n = 232) Obese (n = 147) Race/ethnicity White/Caucasian (n = 558) Black/AA (n = 9) Asian (n = 32) Hispanic (n = 13) Native American/Native Hawaiian/Pacific Islander (n = 5) Other/don't know/prefer not to answer (n = 30)
Body satisfaction
Appearance evaluation
Overweight preoccupation
Health evaluation
Fitness evaluation
Cash
Frederick
Cash
Frederick
Cash
Frederick
Cash
Frederick
Cash
Frederick
(%)
(%)
(%)
(%)
(%)
(%)
(%)
(%)
(%)
(%)
21.2
12.2
28.4
22.9
21.8
8.8
8.8
6.3
19.0
19.0
27.5 14.7 22.9 25.3 21.3 19.6
15.7 10.3 14.3 12.1 13.6 8.4
19.6 21.6 29.5 36.4 31.4 27.1
17.6 18.1 25.7 25.3 26.6 19.6
19.6 20.7 25.7 21.2 21.9 20.6
9.8 8.6 14.3 7.1 6.5 8.4
3.9 5.2 11.4 8.1 11.8 8.4
3.9 4.3 6.7 7.1 8.3 5.6
11.8 18.1 12.4 22.2 22.5 21.5
11.8 18.1 12.4 22.2 22.5 21.5
33.3 10.7 20.3 40.8
16.7 6.9 9.9 25.2
16.7 15.6 23.3 59.9
16.7 11.1 19.0 50.3
16.7 17.2 22.0 29.9
16.7 5.3 8.6 15.0
16.7 5.7 6.0 18.4
0.0 3.4 4.7 14.3
33.3 21.4 15.1 20.4
33.3 21.4 15.1 20.4
21.5 11.1 21.9 15.4 20.0 20.0
12.0 0.0 15.6 15.4 20.0 13.3
29.2 11.1 28.1 15.4 20.0 26.7
23.8 11.1 21.9 7.7 0.0 20.0
21.1 22.2 18.8 23.1 0.0 40.0
8.4 0.0 12.5 0.0 16.7 8.8
8.4 0.0 15.6 15.4 0.0 8.8
6.1 0.0 9.4 15.4 0.0 6.7
19.2 0.0 15.6 23.1 20.0 23.3
19.2 0.0 15.6 23.1 20.0 23.3
156
E.A. Fallon et al. / Eating Behaviors 15 (2014) 151–158
Table 7 Comparisons of BD prevalence among national United States samples over time, as determined by response frequencies. Berscheid et al. (1973)
Body areas satisfaction Face Hair Upper torso (breast/chest) Middle torso (abdomen) Lower torso (hips/upper thigh Muscle tone Height Weight Overall appearance
Cash et al. (1986)
Cash and Henry (1995)
Garner (1997)
Fallon, Harris & Johnson (2013)
Women
Men
Women
Men
Women
Men
Women
Men
Women
Men
%
%
%
%
%
%
%
%
%
%
11.0 19.0 27.0 50.0 49.0 30.0 13.0 48.0 23.0
8.0 20.0 18.0 36.0 12.0 25.0 13.0 35.0 15.0
20.0 n/a 32.0 57.0 50.0 45.0 17.0 55.0 38.0
20.0 n/a 28.0 50.0 21.0 32.0 20.0 41.0 34.0
11.7 16.3 25.1 51.0 47.4 36.9 13.4 46.0 n/a
n/a n/a n/a n/a n/a n/a n/a n/a n/a
n/a n/a 34.0 71.0 61.0 57.0 16.0 66.0 56.0
n/a n/a 38.0 63.0 29.0 45.0 16.0 52.0 43.0
9.7 9.9 22.8 54.8 41.4 34.5 9.4 46.3 13.4
8.0 17.2 19.8 45.2 11.0 22.8 10.5 38.8 9.0
Note: Estimated prevalence of body dissatisfaction (BD) among United States adults is determined by participant responses on each item of the Body Areas Satisfaction subscale of the Multi-dimensional Body Self-Relations Questionnaire. N/A: data is not available from previously published research.
BD of our sample to previously published national samples. Results show that the overall prevalence of BD is varied, depending on the specific measurement tool and cut-off protocol chosen (single-item vs. multiple-item mean scores; conservative vs. liberal cut-off points). When using operational definitions from previous research, the range of BD is 13.4%–31.8% among women and 9.0%–28.4% among men. These ranges are very consistent with the Frederick et al. (2012) estimates of male BD (10%–30%). However, the Frederick et al. (2012) estimate of BD among women (20%–40%) is slightly higher than the range suggested by the present study. Indeed, future studies of BD prevalence using more methodologically sound procedures will be beneficial for more precisely estimating this range (Fiske, Fallon, Blissmer, & Redding, under review). With regard to BD variation across sub-groups, women had greater BD compared to men, with the size of the effect being small for body areas satisfaction (d = .23) and fitness evaluation (d = .28), and approaching medium for overweight preoccupation (d = .45). These effect sizes are consistent with previous research (Feingold & Mazzella, 1998). When examining differences in BD across age groups, a nonstatistically significant trend emerged, such that younger participants (18–24 years) and older adults (N 65 years), reported greater health evaluation (p = .09) and body areas satisfaction (p = .08) than middle-aged groups. This corresponds with previous research suggesting that older men and women report higher body image then their middle-aged counterparts (Feingold & Mazzella, 1998; Grogan, 2011). For race/ethnicity, there were no statistically significant differences for any of the MBSRQ subscales, although there was a strong trend toward a difference between African–American/Black and
White/Caucasian adults for overweight preoccupation. This corresponds with previous research suggesting that African–American adults have more positive body image than White/Caucasian adults, but that the size of the effect is likely small (Grabe & Hyde, 2006). It is important to note that the substantial inequality of sample sizes among racial/ethnic groups in this study likely led to the violation of homogeneity of variances, and a larger sample of racial/ethnic minority groups with homogenous variances may, in fact, find statistical differences across groups. For BMI, our results show that, compared to the prevalence of BD among healthy weight women (7%–14%, as measured by mean scores for body areas satisfaction and appearance evaluation), the prevalence of BD among overweight women is double (33–42%), and the prevalence of BD among obese women is quadruple (49%–69%). This pattern was generally replicated for the prevalence of BD among healthy weight, overweight and obese men. Multivariate analysis revealed that women in the underweight category had lower BD than men in the underweight category, while women in the overweight and obese categories had greater BD compared to their overweight and obese male counterparts. This corresponds with previous research showing that generally, BD increases with weight/BMI (Schwartz & Brownell, 2004), and that men report less BD than women (Feingold & Mazzella, 1998), except when underweight women are closer to the United States cultural ideal (thin ideal for women) than underweight men (muscular ideal for men; McCabe & Ricciardelli, 2004). While not a primary aim of the study, an important finding from the point of view of overweight and obesity as a chronic disease risk factor is that while BD regarding physical appearance is higher in overweight and obese groups, the percent of overweight and obese adults that are
Table 8 Comparisons of BD prevalence among national United States adult samples over time as determined by scale means. Cash et al. (1986)
Cash and Henry (1995)
Fallon, Harris & Johnson (2013) Cash Protocol (mean b 3.0)
Body areas satisfaction Appearance evaluation Overweight preoccupation Fitness evaluation Health evaluation
Frederick Protocol (mean b 2.75)
Women
Men
Women
Men
Women
Men
Women
Men
%
%
%
%
%
%
%
%
n/a 31.0 n/a 29.0 14.0
n/a 24.0 n/a 19.0 9.0
35.6 47.9 48.5 n/a n/a
n/a n/a n/a n/a n/a
28.7 31.8 38.6 30.2 13.0
21.2 28.4 18.2 19.0 8.8
19.2 26.2 21.8 30.2 10.4
12.2 22.9 8.8 19.0 6.3
Note: Estimated prevalence of body dissatisfaction (BD) among United States adults is determined by mean scores on each subscale of the Multi-dimensional Body Self-Relations Questionnaire. N/A: data is not available from previously published research.
E.A. Fallon et al. / Eating Behaviors 15 (2014) 151–158
satisfied with their health and fitness was quite high in these groups. Specifically for fitness evaluation, 69.4%–84.9% of overweight and obese men and 40.8%–71.8% of overweight and obese women were defined as being satisfied with their fitness. Regarding health evaluation, 72.8%–94.0% of overweight and obese men and 56.7%–85.0% of overweight and obese women were defined as satisfied with their health. Such positive perceptions of health and physical fitness among overweight and obese adults may hinder the likelihood of making health behavior changes needed to reduce risk of disease. As previous research has suggested, it could be that some level of BD is beneficial for initiating healthy behavior changes (Heinberg, Thompson, & Matzon, 2001; Schwartz & Brownell, 2004). Finally, the third aim of this study was to compare our findings with that of previously published national samples. Generally, the results of the present study show that the prevalence of BD in the Unites States has not increased above the levels noted from 1973 to 1997. In fact, when focusing on frequency of men and women reporting dissatisfaction with “overall appearance” (Table 7), prevalence of BD appears to have decreased, reverting back to 1973 levels. When using mean scores of the MBSRQ subscales to define BD (Table 8), BD has decreased from the 1990's (body areas satisfaction, appearance evaluation, overweight preoccupation) or remained relatively stable over time, since the 1970's and 1980's (fitness evaluation, health evaluation). While one explanation of this finding is that programs designed to facilitate body appreciation at all ages and sizes have been successful, it is also important to note the numerous methodological limitations of previous prevalence research (Fiske, Fallon, Blissmer, & Redding, under review; Frederick et al., 2012) that hinder a strong conclusion that national prevalence of BD has changed in any direction. Similar to its predecessors, this study of the national prevalence of BD also has several limitations. The sample was a non-random convenience sample, not representative of US adult population in terms of sex, ethnicity, or education. Additionally, we did not assess sexual orientation, a known moderating variable for BD. Finally, the reliability for overweight preoccupation was not optimal, indicating the need for studies examining the psychometric properties of this subscale for a national sample. Although some limitations are evident, this study has several notable strengths. First, we were able to recruit a rather large sample of United States adults, representing at least 37 states and a wide age range. We chose a measure of body image/BD that is recommended for use for both men and women (Thompson, 2004), and has been used in several national US surveys in the past – making our results easily comparable to previous research. Second, we expanded our assessment to include fitness evaluation and health evaluation which are important for gaining a broader perspective on body image, but increasingly important to examining the role of body image with regard to health behaviors that put people at risk for chronic disease. Third, we provided a well-rounded estimate of prevalence by implementing three methods for calculating prevalence (representing conservative and liberal estimates; Fiske, Fallon, Blissmer, & Redding, under review) and calculated prevalence across sex, age, race/ ethnicity, and BMI category within the same study. Finally, as noted by Frederick et al. (2012) the Psychology Today surveys (Berscheid et al., 1973; Cash et al., 1986; Garner, 1997) had different introductions leading up to the survey, providing different rationales for conducting the survey. These forwards may have influenced participant responses and participant self-selection into the study. In an attempt to recruit participants with a wide range of BD, and a wide range of interest in health as a topic, our introductory email requesting participation was brief and did not mention body image or BD, specifically. In conclusion, few studies have sought to examine the national, overall prevalence of BD in the United States using psychometrically sound measurement tools since 1997. Prevalence of BD ranges between 13.4% and 31.8% among women and 9.0%–28.4% among men. Compared to previous national samples using similar measurement tools, we can preliminarily conclude that women remain at higher risk for BD than men and that the prevalence of BD may have plateaued or even declined
157
over time. Future research in this field should employ randomly selected, nationally representative samples and explore the mental, physical, and behavioral health outcomes for individuals who are at both ends of the BD spectrum. Role of Funding Sources There were no funding sources. Contributors Fallon and Harris developed the manuscript concept and measurement protocols. Under the supervision of Fallon and Harris, Johnson submitted the IRB, translated the survey into an online data collection program, recruited participants and managed the data. Fallon analyzed the data and wrote the first draft of the manuscript. All authors consulted throughout the process and contributed to the manuscript. All authors approved the final manuscript. Conflict of Interest All authors declare that they have no conflict of interest.
References American Psychological Association (2010). Writing for the behavioral and social sciences. Publication Manual of the American Psychological Association (pp. 13–15) (6th edition ). New YorkWashington, DC.: American Psychological Association. Berscheid, E., Walster, E., & Bohrnstedt, G. (1973). Body image. The happy American body: A survey report. Psychology Today, 7, 119–131. Cash, T. F. (2000). The multidimensional body-self relations questionnaire (3rd ed.) (Available from the author at www.body-images.com) Cash, T. F. (2003). Body image: Past, present, and future. Body Image, 1, 1–5. http://dx.doi.org/ 10.1016/S1740-1445(03)00011-1. Cash, T. F., & Henry, P. E. (1995). Women's body images: The results of a national survey in the USA. Sex Roles, 33, 19–28. http://dx.doi.org/10.1007/BF01547933. Cash, T. F., Winstead, B. W., & Janda, L. H. (1986). The great American shape-up: Body image survey report. Psychology Today, 20, 30–37. Centers for Disease Control, Prevention (2008). Smoking-attributable mortality, years of potential life lost, and productivity losses —United States, 2000–2004. Morbidity and Mortality Weekly Report, 57, 1226–1228 (Retrieved from http://www.cdc.gov/ mmwr/preview/mmwrhtml/mm5745a3.htm) Feingold, A., & Mazzella, R. (1998). Gender differences in body image are increasing. Psychological Science, 9, 190–195. Fiske, L., Fallon, E. A., Blissmer, B., & Redding, C (under review). Prevalence of body dissatisfaction among United States adults: A review and recommendations for future research. Eating Behaviors. Frederick, D. A., Forbes, G. B., Grigorian, K. E., & Jarcho, J. M. (2007). The UCLA body project I: Gender and ethnic differences in self-objectification and body satisfaction among 2206 undergraduates. Sex Roles, 57, 317–327. http://dx.doi.org/ 10.1007/s11199-007-9251-z. Frederick, D. A., Jafary, A.M., Gruys, K., & Daniels, E. A. (2012). Surveys and the epidemiology of body image dissatisfaction. In T. F. Cash (Ed.), Encyclopedia of body image and human appearance, vol. 2. (pp. 766–774)San Deigo, CA: Elsevier. http://dx.doi.org/ 10.1016/B978-0-12-384926-7.00121-3. Frederick, D. A., Peplau, L. A., & Lever, J. (2006). The swimsuit issue: Correlates of body image in a sample of 52,677 heterosexual adults. Body Image, 3, 413–419. http://dx.doi.org/10.1016/j.bodyim.2006.08.002. Garner, D.M. (1997). The 1997 body image survey results. Psychology Today, 1, 30–44. Grabe, S., & Hyde, J. S. (2006). Ethnicity and body satisfaction among women in the United States: A meta-analysis. Psychological Bulletin, 132, 622–640. Grogan, S. (2011). Body image development in adulthood. In T. F. Cash, & L. Smolak (Eds.), Body image: A handbook of science, practice and prevention (pp. 93–100) (2nd ed.). New York: Guilford Press. Heinberg, L. J., Thompson, K., & Matzon, J. L. (2001). Body image dissatisfaction as a motivator for healthy lifestyle change: Is some distress beneficial? In R. H. Streigel-Moore, & L. Smolak (Eds.), Eating Disorders: Innovative directions in research and practice, vol. 2. (pp. 215–232)Washington, DC: American Psychological Association. Howlader N, Noone AM, Krapcho M, Garshell J, Neyman N, Altekruse SF, Kosary CL, Yu M, Ruhl J, Tatalovich Z, Cho H, Mariotto A, Lewis DR, Chen HS, Feuer EJ, Cronin KA (eds). SEER Cancer Statistics Review, 1975-2010, National Cancer Institute. Bethesda, MD, http://seer.cancer.gov/csr/1975_2010/, based on November 2012 SEER data submission, posted to the SEER web site, April 2013. Hudson, J. I., Hiripi, E., Pope, H. G., & Kessler, R. C. (2007). The prevalence and correlates of eating disorders in the National Comorbidity Survey Replication. Biological Psychiatry, 61, 348–358. http://dx.doi.org/10.1016/j.biopsych.2006.03.040. Johnson, P., Fallon, E. A., Harris, B. S., & Burton, B. (2013). Body Satisfaction is associated with Transtheoretical Model constructs for Physical Activity Behavior Change. Body Image: An International Journal of Research, 10, 163–174. http://dx.doi.org/10.1016/j.bodyim. 2012.12.002 PMID: 23339856. King, T. M., Matacin, M., White, K. S., & Marcus, B. H. (2005). A prospective examination of body image and smoking cessation in women. Body Image, 2, 19–28. http://dx.doi.org/10.1016/j.bodyim.2005.01.003. Kruger, J., Lee, C., Ainsworth, B. E., & Macera, C. A. (2008). Body size satisfaction and physical activity levels among men and women. Obesity, 16, 1976–1979. http://dx.doi.org/ 10.1038/oby.2008.311.
158
E.A. Fallon et al. / Eating Behaviors 15 (2014) 151–158
McCabe, M., & Ricciardelli, L. (2004). Weight and shape concerns of boys and men. In J. K. Thompson (Ed.), Handbook of eating disorders and obesity (pp. 606–634). New York: Wiley. Peplau, L. A., Frederick, D. A., Yee, C., Maisel, N., Lever, J., & Ghavami, N. (2009). Body image satisfaction in heterosexual, gay and lesbian adults. Archives of Sexual Behavior, 38, 713–725. http://dx.doi.org/10.1007/s10508-008-9378-1. Ridolfi, D. R., & Crowther, J. H. (2013). The link between women's body image disturbances and body-focused cancer screening behaviors: A critical review of the literature and a new integrated model for women. Body Image, 10, 149–162. http://dx.doi.org/10.1016/j.bodyim.2012.11.003. Rodin, J., Silberstein, L. R., & Striegel-Moore, R. H. (1984). Women and weight: A normative discontent. In T. B. Sonderegger (Ed.), Nebraska symposium on motivation: Psychology and gender (pp. 267–307). Lincoln, NE: University of Nebraska Press. Schwartz, M. B., & Brownell, K. D. (2004). Obesity and body image. Body Image: An International Journal of Research, 1, 43–56. http://dx.doi.org/10.1016/S1740-1445(03) 00007-X.
Stuhldreher, N., Konnopka, A., Wild, B., Herzog, W., Zipfel, S., Löwe, et al. (2012). Cost-of-illness studies and cost-effectiveness analyses in eating disorders: A systematic review. International Journal of Eating Disorders, 45, 476–491. Tantleff-Dunn, S., Barnes, R. D., & Larose, J. G. (2011). It's not just a “woman thing:” The current state of normative discontent. Eating Disorders: The Journal of Treatment & Prevention, 19(5), 392–402. http://dx.doi.org/10.1080/10640266.2011.609088. Thompson, J. K. (2004). The (mis)measurement of body image: Ten strategies to improve assessment for applied and research purposes. Body Image, 2, 7–14. http://dx.doi.org/10.1016/S1740-1445(03)00004-4. U.S. Department of Health and Human Services, & Centers for Disease Control, Prevention (2010). Behavioral risk factor surveillance system survey questionnaire. Retrieved from. http://www.cdc.gov/brfss/questionnaires.htm U.S. Department of Health and Human Services, National Institutes of Health, & National Heart Lung and Blood Institute (1998). Clinical guidelines on the identification, evaluation, and treatment of overweight and obesity in adults (NIH publication no. 98-4083). Retrieved from. http://www.nhlbi.nih.gov/guidelines/ obesity/ob_gdlns.pdf
Contents lists available at ScienceDirect
Eating Behaviors
Prevalence of body dissatisfaction among a United States adult sample☆ Elizabeth A. Fallon a,b,⁎, Brandonn S. Harris c,d, Paige Johnson b a
Department of Kinesiology & Health, Georgia State University, Atlanta, GA, United States Department of Kinesiology, Kansas State University, Manhattan, KS, United States c Department of Health and Kinesiology, Georgia Southern University, Statesboro, GA, United States d Department of Special Education, Counseling, and Student Affairs, Kansas State University, Manhattan, KS, United States b
a r t i c l e
i n f o
Article history: Received 2 October 2013 Received in revised form 4 November 2013 Accepted 20 November 2013 Available online 5 December 2013 Keywords: Body dissatisfaction Body image Prevalence Adult United States
a b s t r a c t Body dissatisfaction (BD) is a primary determinant of eating disorders and has been linked to chronic disease via decreased likelihood of cancer screening self-exams and smoking cessation. Yet, there are few recent estimates of the prevalence of BD among United States adults. Using an internet-based, opt-in, cross-sectional survey, United States adults (N = 1893) completed assessments of demographic variables, body areas satisfaction, appearance evaluation, fitness evaluation, health evaluation, and overweight preoccupation. Results revealed that the range of BD is 13.4%–31.8% among women and 9.0%–28.4% among men. Compared to previous assessments of prevalence (1973, 1986, 1995, 1997), the prevalence of BD among United States adults may have plateaued or declined over time. © 2013 Elsevier Ltd. All rights reserved.
1. Introduction Body image is defined as a persons body-related self-perceptions and self-attitudes, including thoughts, feelings, and behaviors (Cash, 2003). In some cases, extreme levels of body dissatisfaction (BD) can result in eating disorders (e.g., anorexia nervosa, bulimia nervosa), which have a lifetime prevalence of 2.5% and 0.8% for United States women and men, respectively (Hudson, Hiripi, Pope, & Kessler, 2007). For those who seek treatment, annual treatment costs range from $US 1288 to $US 8042 per person, per year (Stuhldreher et al., 2012). BD is also known to negatively influence behavioral risk factors for chronic disease, which affect an even greater proportion of the US population. For example, breast cancer is the most common cancer among women, with a lifetime prevalence of 12.3% (Howlader et al., 2013). A recent literature review demonstrates that BD is associated with lower likelihood of engaging in breast cancer self-exams (Ridolfi & Crowther, 2013), which could provide early detection of the disease. BD is also associated with lower likelihood of smoking cessation (King et al., 2005), where smoking costs $US 96.8 billion annually in lost productivity and is
☆ The data for this study was collected at Kansas State University, but analyzed and the manuscript composed at Georgia State University. Elizabeth A. Fallon is now at Georgia State University and Brandonn S. Harris is now at Georgia Southern University. Paige Johnson is a recent graduate of Kansas State University, and does not currently maintain a university affiliation.We thank the survey respondents for their time and the students who helped collect the data; specifically, laboratory assistants Joseph Lightner, Abby Banks, Madelaine Ellison, Annie Pipes, Tracy Engstrom, and Laura Moluf. ⁎ Corresponding author at: P.O. Box 3975, Atlanta, GA 30302, United States. E-mail address: [email protected] (E.A. Fallon). 1471-0153/$ – see front matter © 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.eatbeh.2013.11.007
responsible for almost 30% of cancer deaths (Centers for Disease Control, Prevention, 2008). With such robust potential to influence health care costs as well as an individual's health and quality of life across the lifespan, it is imperative for behavioral scientists, clinicians, and public health professionals to be informed of the current overall prevalence of BD among the United States population, changes in BD prevalence over time, as well as understand which sub-populations are at risk. Salient to this study are four preceding national surveys assessing body image/BD among United States adults (Berscheid, Walster, & Bohrnstedt, 1973; Cash & Henry, 1995; Cash, Winstead, & Janda, 1986; Garner, 1997). Since 1972, a series of surveys have documented an increase in dissatisfaction in overall appearance among both men (15% to 43%) and women (23% to 56%). This often-cited research has led many to conclude that there is a ‘normative discontent’ or an ‘epidemic’ of BD in the United States, especially among women (Frederick, Jafary, Gruys, & Daniels, 2012; Rodin, Silberstein, & Striegel-Moore, 1984; Tantleff-Dunn, Barnes, & Larose, 2011). Since Garner (1997), however, very few studies have used national surveys of United States adults, and the few containing national samples, have either not used psychometrically valid measurement tools (Frederick, Peplau, & Lever, 2006), have used single-item measures of body satisfaction (Kruger, Lee, Ainsworth, & Macera, 2008), or have focused on specific United States subgroups (Peplau et al., 2009). Thus, a more recent assessment of overall prevalence of BD among United States adults is timely and the purpose of this study is threefold: a. To provide an overall estimate of the national prevalence of BD among United States men and women,
152
E.A. Fallon et al. / Eating Behaviors 15 (2014) 151–158
b. To examine differences in BD across population subgroups (e.g., sex, race/ethnicity, age, body mass index category), c. To compare the rates of BD of the present sample to previously published national samples.
Table 1 Sample characteristics by participant sex. Women (n = 1246) M
2. Method
Age BMI
Men (n = 647) SD
42.50 25.86
14.00 6.75
M 47.77 26.90
SD 16.27 5.38
2.1. Participants N
Demographic information for adult volunteers (N = 1893; age range 18–90 years) who completed an online survey are presented in Table 1. 2.2. Procedure Prior to data collection, Institutional Review Board approval was obtained for an internet-based, opt-in survey taking participants approximately 20–30 min to complete.1 To recruit United States adults across a wide geographical area, e-mail addresses were collected from publicly available websites in the United States. These websites primarily consisted of, but were not limited to, colleges/universities, public primary schools, state/local government, and faith-based organizations. Invitation emails for the study titled “Attitudes Towards Health Behaviors” contained the IRB informed consent information (e.g., general purpose of the study, IRB and contact information, and that the data would be used for research) and a link to the survey. While the survey respondents could choose to submit the survey anonymously, they could also choose to include their contact information to win a monthly drawing for $15 to a national chain store. This drawing, as well as two followup e-mail reminders sent 1 and 2 weeks after the initial invitation were intended to increase response rate and survey completion rate. After the data collection period (August 2010–April 2011) ended, the data were de-identified. Over the course of the recruitment period, 2665 individuals accessed the online survey. Using a “completers only” analysis for this study, participants failing to complete items assessing demographic variables or the multidimensional body self-relations survey were excluded from the analysis. The final sample for this study is 1893 adults, representing 37 US states. 2.3. Measures 2.3.1. Demographics Consistent with the Behavior Risk Factor Surveillance (U.S. Department of Health and Human Services and Centers for Disease Control, Prevention (2010)) system, participants were asked to self-report their birth date (month, day, year), sex, race, ethnicity, and education level. Age groups were created to be consistent with those from previous research (Cash & Henry, 1995). Self-reported height and weight measurements were used to compute body mass index (BMI; weight [kg]/height [m2]). Subsequently, four BMI categories were created: underweight (BMI b 18.5), healthy weight (BMI 18.5 to b25), overweight (BMI 25.0 to b 30.0), and obese (BMI ≥ 30.0; U.S. Department of Health and Human Services, National Institutes of Health, & National Heart Lung & Blood Institute, 1998). 2.3.2. Body areas satisfaction Satisfaction with specific areas of the body was assessed using the Body Areas Satisfaction subscale of the Multidimensional Body-Self 1
According to publication guidelines (American Psychological Association, 2010), we note that this data set comes from a previously published manuscript (Johnson, Fallon, Harris, & Burton, 2013), which focused on the associations between body satisfaction physical activity behavior change. The purposes of each manuscript necessitated multiple manuscripts for multiple audiences. Furthermore, the data from fitness evaluation, appearance evaluation, health evaluation and health orientation are novel and have not been published elsewhere.
Age 18–24 year 25–34 year 35–44 year 45–54 year 55–64 year 65 and over BMI Underweight Normal weight Overweight Obese Race/ethnicity White/Caucasian Black/African American Asian Hispanic Native American/Native Hawaiian/Pacific Islander Other/don't know/prefer not to answer Education Less than 4-year college Degree 4-year college degree Cohabitation status Married/cohabitating Not married/cohabitating
(%)
N
(%)
107 341 212 242 290 54
8.6 27.4 17.0 19.4 23.3 4.3
51 116 105 99 169 107
7.9 17.9 16.2 15.3 26.1 16.5
32 673 301 240
2.6 54.0 24.2 19.3
6 262 232 147
0.9 40.5 35.9 22.7
1074 30 47 37
86.2 2.4 3.8 3.0
558 9 32 13
86.2 1.4 4.9 2.0
21 37
1.7 3.0
5 30
0.8 4.6
188 1055
15.1 84.7
56 590
8.7 91.2
802 435
64.4 34.9
474 169
73.3 26.1
Relations Questionnaire (MBSRQ; Cash, 2000). This measurement tool uses a 5-point Likert-type scale ranging from very dissatisfied (1) to very satisfied (5) to measure participants' satisfaction with nine specific aspects of their body (e.g., face, height, hair, muscle tone). Using this tool, BD was assessed in two ways. First, for each specific body area, individuals reporting ‘very dissatisfied’ or ‘dissatisfied’ were categorized as having BD. Second, the scores for each of the 9 items were averaged, with higher scores representing higher body satisfaction. According to a protocol from Cash and colleagues (Cash & Henry, 1995; Cash et al., 1986), a mean score of b 3.0 is defined as BD. According to the Frederick and colleagues' protocol (Frederick, Forbes, Grigorian, & Jarcho, 2007; Peplau et al., 2009), a mean score of b2.75 is defined as BD. For this study, we present both protocol cut-off points (Fiske, Fallon, Blissmer, & Redding, under review). Finally, for this study, the internal consistency was good (α = .83). 2.3.3. Appearance evaluation Feelings of physical attractiveness or satisfaction with one's looks are measured using a 7-item Likert-type scale ranging from Definitely Disagree (1) to Definitely Agree (5). Scores on each item are averaged to create a mean score, where higher scores are indicative of greater satisfaction with appearance (Cash, 2000). For this study, the internal consistency was good (α = .89) and we present both the Cash (scale means b 3.0) and Frederick (scale means b2.75) cut-off points. 2.3.4. Overweight preoccupation Overweight preoccupation is a construct reflecting fat anxiety, weight vigilance, dieting, and eating restraint (Cash, 2000). It is assessed using a 4-item Likert-type scale ranging from Definitely Disagree (1) to Definitely Agree (5). Scores on each item are averaged to create a mean score, where higher scores are indicative of greater overweight preoccupation, and therefore greater BD (Cash, 2000). For this study, we present
E.A. Fallon et al. / Eating Behaviors 15 (2014) 151–158
both the Cash (scale means b 3.0) and Frederick (scale means b 2.75) cut-off points. Finally, for this study, the internal consistency was questionable (α = .61). 2.3.5. Fitness evaluation Feelings of being physically fit or unfit were assessed using a 3item Likert-type scale from Definitely Disagree (1) to Definitely Agree (5). Scores on each item are averaged to create a mean score. Individuals scoring highly on this subscale of the MBSRQ regard themselves as physically fit, “in shape”, or athletically active and competent (Cash, 2000). For this study, we present both the Cash (scale means b3.0) and Frederick (scale means b2.75) cut-off points. Finally, for this study, the internal consistency was acceptable (α = .79). 2.3.6. Health evaluation Using a 6-item Likert-type scale from Definitely Disagree (1) to Definitely Agree (5), the Health Evaluation subscale of the MBSRQ assesses feelings of physical health and/or the freedom from physical illness. Higher scores are indicative of feeling the body is in good health (Cash, 2000). Consistent with the other subscales, we present both the Cash (scale means b 3.0) and Frederick (scale means b2.75) cut-off points. The internal consistency for this study was good (α = .80). 2.4. Data analysis Data were examined for outliers, linear dependency, skewness and kurtosis. Frequencies, means, and standard deviations for demographic variables were used to describe the study sample (See Table 1). Means and standard deviations for each of the body image subscales by demographic variable were also calculated (Tables 2 & 3). To provide an estimate of national prevalence of BD among United States adult men and women, we compared the use of three protocols from previous literature. Specifically, the frequency of men and women reporting being ‘very dissatisfied’ or ‘dissatisfied’ with specific body areas is reported (Berscheid et al., 1973; Cash & Henry, 1995; Cash et al., 1986; Garner, 1997; see Table 4). Additionally, we calculated the means and standard deviations for each of the five MBSRQ subscales (body areas satisfaction, appearance
153
evaluation, overweight preoccupation, fitness evaluation, and health evaluation) and calculated BD prevalence based on the Cash (scale mean b3.0; Cash & Henry, 1995; Cash et al., 1986) and Frederick (scale mean b2.75; Frederick et al., 2007; Peplau et al., 2009) protocols. See Tables 5 & 6 for women and men, respectively. To examine differences in each body image measure across population subgroups, we conducted a series of 2-way MANOVAs, with Tukey's post hoc tests, when appropriate. Finally, we visually compared our results to those of previous national United States adult surveys (see Tables 7 & 8).
3. Results More women than men enrolled in the study, and the sample was predominantly white/Caucasian, middle aged, overweight and married/cohabitating. If frequency counts are used to define prevalence of body dissatisfaction, only 9% of men and 13.4% of women were dissatisfied with their overall appearance (see Table 4). If the mean score of the body areas satisfaction measure is used, the estimated prevalence of BD doubles, ranging between 12.2% and 21.2% for men, and 19.2–28.7% for women, depending on choosing more liberal (Cash) or more conservative (Frederick) cut-off protocol. If the appearance evaluation mean score is used, the estimated range of prevalence is 22.9%–28.4% and 26.2%–31.8%, for men and women, respectively (see Tables 5 & 6). When examining differences in MBSRQ subcales across age groups, the Box's Test was significant, violating the assumption of homogeneity of variance [F (165, 659,090) = 1.44, p b .001]. Therefore, Pillai's Trace was used to interpret the MANOVA results. The 2 (Sex) X 6 (Age Category) MANOVA for the five MBSRQ subscales revealed a statistically significant main effect for sex [Pillai's Trace = .06, F (5, 1877) = 25.03, p b .001, η2 = .06], and a significant main effect for age category [Pillai's Trace = .07, F (25, 9405) = 5.58, p b 0.001, η 2 = .02], but no statistically significant interaction emerged between sex and age [Pillai's Trace = .02, F (25, 9405) = 1.19, p = 0.23, η2 = .23]. Follow-up univariate analyses revealed that compared to men, women had lower body areas satisfaction [F (1, 1881) = 19.03, p b .001, η2 = .01, d = .23], lower fitness evaluation [F (1, 1881) = 31.07, p b .001, η2 = .00, d = .28], and greater overweight preoccupation [F (1, 1881) = 73.20, p b .001, η2 = .00,
Table 2 Means and standard deviations for MBSRQ subscales for women across demographic characteristics. Body satisfaction
All women (n = 1246) Age 18–24 years (n = 107) 25–34 years (n = 341) 35–44 years (n = 212) 45–54 year (n = 242) 55–64 years (n = 290) 65 and over (n = 54) BMI Underweight (n = 32) Normal weight (n = 673) Overweight (n = 301) Obese (n = 240) Race White/Caucasian (n = 1074) Black/African American (n = 30) Asian (n = 47) Hispanic (n = 37) Native American/Native Hawaiian/Pacific Islander (n = 21) Other/don't know/prefer not to answer (n = 37)
Appearance evaluation
Overweight preoccupation
Fitness evaluation
Health evaluation
M
SD
M
SD
M
SD
M
SD
M
SD
3.30
0.67
3.27
0.83
2.64
0.8
3.32
0.97
3.83
0.77
3.48 3.46 3.30 3.16 3.15 3.34
0.63 0.66 0.70 0.64 0.65 0.57
3.58 3.52 3.19 3.11 3.06 3.20
0.80 0.83 0.84 0.80 0.79 0.75
2.64 2.57 2.57 2.69 2.77 2.53
0.76 0.84 0.80 0.79 0.75 0.80
3.79 3.39 3.37 3.22 3.13 3.06
0.94 1.01 0.94 0.91 0.94 0.89
3.92 3.86 3.68 3.79 3.86 3.95
0.64 0.73 0.86 0.74 0.79 0.77
3.54 3.55 3.13 2.74
0.66 0.60 0.55 0.59
3.50 3.64 3.04 2.47
0.85 0.67 0.70 0.74
2.15 2.52 2.83 2.81
0.85 0.79 0.74 0.78
3.31 3.45 3.35 2.89
1.06 0.95 0.91 0.97
3.74 4.03 3.79 3.31
0.89 0.68 0.77 0.73
3.29 3.24 3.27 3.31 3.59 3.38
0.66 0.69 0.56 0.70 0.67 0.82
3.25 3.45 3.43 3.35 3.22 3.52
0.83 0.74 0.75 0.92 0.88 1.00
2.63 2.53 2.63 2.78 3.10 2.73
0.79 0.80 0.85 0.78 0.85 1.00
3.29 3.79 3.37 3.56 3.25 3.49
0.96 0.81 0.85 1.04 1.12 1.10
3.83 3.85 3.78 3.93 3.70 3.81
0.77 0.65 0.65 0.87 0.86 0.88
154
E.A. Fallon et al. / Eating Behaviors 15 (2014) 151–158
Table 3 Means and standard deviations of MBSRQ scale scores for men across demographic characteristics.
Overall Men (n = 647) Age 18–24 year (n = 51) 25–34 year (n = 116) 35–44 year (n = 105) 45–54 year (n =99) 55–64 year (n = 169) 65 and over (n = 107) BMI Underweight (n = 6) Normal weight (n = 262) Overweight (n = 232) Obese (n = 147) Race White/Caucasian (n = 558) Black/African American (n = 9) Asian (n = 32) Hispanic (n = 13) Native American/native Hawaiian/Pacific Islander (n = 5) Other/don't know/prefer not to answer (n = 30)
Body satisfaction
Appearance evaluation
Overweight preoccupation
Fitness evaluation
M
SD
M
3.45
0.64
3.58 3.55 3.43 3.35 3.32 3.57
SD
M
SD
M
SD
M
SD
3.3
0.76
2.29
0.76
3.58
0.89
3.86
0.69
0.74 0.61 0.69 0.60 0.61 0.62
3.61 3.50 3.27 3.19 3.13 3.35
0.86 0.72 0.82 0.75 0.75 0.65
2.04 2.14 2.33 2.35 2.36 2.37
0.85 0.85 0.79 0.74 0.69 0.66
3.76 3.70 3.74 3.53 3.43 3.49
1.02 0.87 0.82 0.86 0.88 0.90
3.96 3.98 3.80 3.77 3.76 3.98
0.65 0.62 0.68 0.64 0.76 0.67
3.24 3.64 3.46 3.08
0.42 0.60 0.60 0.64
3.33 3.59 3.35 2.71
0.43 0.63 0.66 0.81
2.04 2.11 2.35 2.52
0.95 0.75 0.73 0.74
3.17 3.60 3.62 3.50
0.86 0.94 0.83 0.88
3.42 3.96 3.94 3.58
0.58 0.65 0.64 0.75
3.44 3.90 3.41 3.52 3.67 3.36
0.63 0.78 0.71 0.63 0.88 0.66
3.29 3.76 3.33 3.45 3.80 3.20
0.77 0.76 0.61 0.66 0.73 0.82
2.28 2.05 2.46 2.08 1.60 2.50
0.74 0.75 0.82 1.17 0.52 0.80
3.57 4.22 3.76 3.46 3.93 3.48
0.87 0.58 0.92 1.02 0.86 1.02
3.88 3.81 3.70 3.76 4.13 3.59
0.68 0.57 0.78 0.80 0.40 0.62
d = .45]. There were no differences between men and women for appearance evaluation [F (1, 1881) = 2.33, p = .13, η2 = .00, d = .03] or health evaluation [F (1, 1881) = 0.61, p = .43, η2 = .00, d = .04]. Follow-up univariate analyses revealed a main effect of age group for body areas satisfaction [F (5, 1881) = 10.66, p b .001, η 2 = .03], appearance evaluation [F (5, 1881) = 16.53, p b .001, η2 = .04], fitness evaluation [F (5, 1881) = 9.06, p b .001, η 2 = .02], health evaluation [F (5, 1881) = 3.61, p = .01, η 2 = .01], and overweight preoccupation [F (5, 1881) = 3.89, p = .01, η2 = .01]. Tukey's post hoc analyses, however, did not reveal any statistically significant differences among the age groups for any of the body image dependent variables (p's N .05). Trends were evident suggesting a curvilinear trajectory for health evaluation (p = .09) and body areas satisfaction (p = .08) that these constructs were generally higher among younger participants (18–24 years) and older adults (N65 years), while lower among the middle-aged. Examining differences in MBSRQ subscales across BMI category, the Box's Test was significant, violating the assumption of homogeneity of
Health evaluation
variance [F (105, 4608) = 1.89, p b .001]. Therefore, Pillai's Trace was used to interpret the MANOVA results. The 2 (Sex) × 4 (BMI category) MANOVA for the five MBSRQ subscales revealed a significant main effect for sex, and a significant main effect for BMI category [Pillai's Trace = .25, F (5, 1881) = 34.11, p b 0.001, η2 = .08], as well as a statistically significant sex by body mass index interaction [Pillai's Trace = .03, F (15, 5649) = 3.38, p b 0.001, η2 = .01]. The main effect for sex has already been noted, and would also be expected for this analysis. Despite a statistically significant main effect in multivariate analysis, Tukey's post hoc analyses did not reveal any differences among the BMI groups for any of the body image dependent variables (p's N .05). Follow-up analyses showed an interaction for body areas satisfaction [F (3, 1885) = 7.29, p b .001, η2 = .01], appearance evaluation [F (3, 1885) = 8.19, p b 001, η2 = .01], fitness evaluation [F (3, 1885) = 5.45, p = .001, η2 = .01], and health evaluation [F (3, 1885) =5.93, p = .001, η2 = .01], such that in the underweight category, women consistently report more positive body image than men. In the overweight and obese categories, however, men consistently report more positive body image than women. No significant interaction was
Table 4 Response frequencies and item means for body areas satisfaction among women and men.
Women (n = 1246) Face Hair Upper torso Middle torso Lower torso Muscle tone Height Weight Overall appearance Men (n = 647) Face Hair Upper torso Middle torso Lower torso Muscle tone Height Weight Overall appearance
Very dissatisfied
Dissatisfied
Very satisfied
Item mean
n
%
n
%
n
Neither %
n
Satisfied %
n
%
M
SD
15 24 68 226 133 91 27 197 25
1.2 1.9 5.5 18.1 10.7 7.3 2.2 15.8 2.0
106 100 215 457 382 339 90 380 142
8.5 8.0 17.3 36.7 30.7 27.2 7.2 30.5 11.4
224 169 309 222 261 345 257 227 324
18.0 13.6 24.8 17.8 20.9 27.7 20.6 19.0 26.0
729 641 553 281 388 409 498 363 684
58.5 51.4 44.4 22.6 31.1 32.8 40.0 29.1 54.9
172 312 101 60 82 62 374 69 71
13.8 25.0 8.1 4.8 6.6 5.0 30.0 5.5 5.7
3.75 3.90 3.32 2.59 2.92 3.01 3.88 2.78 3.51
0.84 0.93 1.03 1.16 1.14 1.05 0.99 1.19 0.84
10 22 23 87 13 21 9 64 9
1.5 3.4 3.6 13.4 2.0 3.2 1.4 9.9 1.4
42 89 105 206 58 127 59 187 49
6.5 13.8 16.2 31.8 9.0 19.6 9.1 28.9 7.6
145 136 175 161 134 180 160 134 180
22.4 21.0 27.0 24.9 20.7 27.8 24.7 20.7 27.8
347 277 276 160 315 265 251 203 350
53.6 42.8 42.7 24.7 48.7 41.0 38.8 31.4 54.1
103 123 68 33 127 54 168 59 59
15.9 19.0 10.5 5.1 19.6 8.3 26.0 9.1 9.1
3.76 3.60 3.40 2.77 3.75 3.32 3.79 3.01 3.62
0.85 1.05 0.96 1.12 0.94 0.99 0.98 1.17 0.81
E.A. Fallon et al. / Eating Behaviors 15 (2014) 151–158
155
Table 5 Comparison of Cash & Frederick recommended cut-off points for determining prevalence of BD among United States women.
Overall women (n = 1246) Age 18–24 years (n = 107) 25–34 years (n = 341) 35–44 years (n = 212) 45–54 years (n = 242) 55–64 years (n = 290) 65 and over (n = 54) BMI Underweight (n = 32) Healthy weight (n = 673) Overweight (n = 301) Obese (n = 240) Race White/Caucasian (n = 1074) Black/AA (n = 30) Asian (n = 47) Hispanic (n = 37) Native American/Native Hawaiian/Pacific Islander (n = 21) Other/don't know/prefer not to answer (n = 37)
Body satisfaction
Appearance evaluation
Overweight preoccupation
Health evaluation
Fitness evaluation
Cash
Frederick
Cash
Frederick
Cash
Frederick
Cash
Frederick
Cash
Frederick
(%)
(%)
(%)
(%)
(%)
(%)
(%)
(%)
(%)
(%)
28.7
19.2
31.8
26.2
38.6
18.2
13.0
10.4
30.2
30.2
19.6 20.2 30.2 36 35.9 24.1
11.2 12.9 19.3 25.6 25.2 13
18.7 19.6 37.3 40.1 39.7 39.7
16.8 17 28.8 33.5 33.4 22.2
37.4 33.1 36.3 43.8 43.8 33.3
14 18.5 15.6 19.8 21.4 11.1
6.5 9.7 16.5 15.7 14.8 11.1
4.7 7.9 14.6 12 11 9.3
15.9 28.7 27.4 31.8 36.6 37
15.9 28.7 274 31.8 36.6 37
18.8 14.1 34.2 64.2
9.4 7.4 22.6 49.2
25 14 42.2 69.6
21.9 10 33.2 63.8
18.8 30.5 50.8 48.8
6.3 15.2 22.9 22.5
12.5 6.4 15 29.2
12.5 4.6 12.6 23.3
31.3 24.5 28.2 48.3
31.3 24.5 28.2 48.3
29.2 30 25.5 29.7 19 21.6
19.4 23.3 17 18.9 14.3 16.2
32.6 23.3 21.3 29.7 38.1 27
27 16.7 14.9 27 23.8 27
37.9 40 36.2 45.9 61.9 40.5
17.6 13.3 19.1 24.3 38.1 21.6
13.4 10 4.3 13.5 19 10.8
10.5 6.7 4.3 10.8 19 10.8
31.4 10 21.3 33.3 32.4 30.2
31.4 10 21.3 18.9 33.3 32.4
found for overweight preoccupation [F (3, 1885) = 1.37, p = .25, η2 = .00]. When examining differences in MBSRQ subscale scores by race/ ethnicity, the Box's Test was significant, violating the assumption of homogeneity of variance [F (150, 18662) = 1.52, p b .001]. Thus, Pillai's Trace was used to interpret the MANOVA results. The 2 (Sex) × 6 (Race/ethnicity) MANOVA for the five MBSRQ subscales revealed a significant main effect of race/ethnicity [Pillai's Trace = .02, F (25, 9405) = 1.61, p = .03, η2 = .00], but no statistically significant sex by race interaction [Pillai's Trace = .02, F (25, 9405) = 1.40, p = .09, η2 = .00]. Follow-up univariate analyses for race/ethnicity did not reveal a main effect for body areas satisfaction [F (5, 1881) = 1.06, p = .38, η2 = .00], appearance evaluation [F (5, 1881) = 1.30, p = .26, η2 = .00], fitness evaluation [F (5, 1881) = 0.85, p = .51, η2 = .00], or health evaluation [F (5, 1881) = 0.92, p = .47, η2 = .00]. A statistically significant difference did emerge for overweight
preoccupation suggesting that compared to non-Hispanic Whites, Black/African–American participants reported lower overweight preoccupation [F (5, 1881) = 2.45, p = .03, η 2 = .01, d = .13]. The Tukey's post hoc analysis, however, was not statistically significant for this interaction (p = .06). Finally, when comparing the results of this study to previously published national samples, it appears that BD prevalence among United States adults has plateaued, or perhaps, slightly decreased over time (see Tables 7 & 8). 4. Discussion The purpose of this research was to provide a recent overall estimate of the national prevalence of BD among United States men and women, examine differences in BD across population subgroups (e.g., sex, race/ ethnicity, age, body mass index category), and compare the rates of
Table 6 Comparison of Cash & Frederick recommended cut-off points for determining prevalence of BD among United States men.
Overall men (n = 647) Age 18–24 years (n = 51) 25–34 years (n = 116) 35–44 years (n = 105) 45–54 years (n = 99) 55–64 years (n = 169) 65 and over (n = 107) BMI Underweight (n = 6) Healthy weight (n = 262) Overweight (n = 232) Obese (n = 147) Race/ethnicity White/Caucasian (n = 558) Black/AA (n = 9) Asian (n = 32) Hispanic (n = 13) Native American/Native Hawaiian/Pacific Islander (n = 5) Other/don't know/prefer not to answer (n = 30)
Body satisfaction
Appearance evaluation
Overweight preoccupation
Health evaluation
Fitness evaluation
Cash
Frederick
Cash
Frederick
Cash
Frederick
Cash
Frederick
Cash
Frederick
(%)
(%)
(%)
(%)
(%)
(%)
(%)
(%)
(%)
(%)
21.2
12.2
28.4
22.9
21.8
8.8
8.8
6.3
19.0
19.0
27.5 14.7 22.9 25.3 21.3 19.6
15.7 10.3 14.3 12.1 13.6 8.4
19.6 21.6 29.5 36.4 31.4 27.1
17.6 18.1 25.7 25.3 26.6 19.6
19.6 20.7 25.7 21.2 21.9 20.6
9.8 8.6 14.3 7.1 6.5 8.4
3.9 5.2 11.4 8.1 11.8 8.4
3.9 4.3 6.7 7.1 8.3 5.6
11.8 18.1 12.4 22.2 22.5 21.5
11.8 18.1 12.4 22.2 22.5 21.5
33.3 10.7 20.3 40.8
16.7 6.9 9.9 25.2
16.7 15.6 23.3 59.9
16.7 11.1 19.0 50.3
16.7 17.2 22.0 29.9
16.7 5.3 8.6 15.0
16.7 5.7 6.0 18.4
0.0 3.4 4.7 14.3
33.3 21.4 15.1 20.4
33.3 21.4 15.1 20.4
21.5 11.1 21.9 15.4 20.0 20.0
12.0 0.0 15.6 15.4 20.0 13.3
29.2 11.1 28.1 15.4 20.0 26.7
23.8 11.1 21.9 7.7 0.0 20.0
21.1 22.2 18.8 23.1 0.0 40.0
8.4 0.0 12.5 0.0 16.7 8.8
8.4 0.0 15.6 15.4 0.0 8.8
6.1 0.0 9.4 15.4 0.0 6.7
19.2 0.0 15.6 23.1 20.0 23.3
19.2 0.0 15.6 23.1 20.0 23.3
156
E.A. Fallon et al. / Eating Behaviors 15 (2014) 151–158
Table 7 Comparisons of BD prevalence among national United States samples over time, as determined by response frequencies. Berscheid et al. (1973)
Body areas satisfaction Face Hair Upper torso (breast/chest) Middle torso (abdomen) Lower torso (hips/upper thigh Muscle tone Height Weight Overall appearance
Cash et al. (1986)
Cash and Henry (1995)
Garner (1997)
Fallon, Harris & Johnson (2013)
Women
Men
Women
Men
Women
Men
Women
Men
Women
Men
%
%
%
%
%
%
%
%
%
%
11.0 19.0 27.0 50.0 49.0 30.0 13.0 48.0 23.0
8.0 20.0 18.0 36.0 12.0 25.0 13.0 35.0 15.0
20.0 n/a 32.0 57.0 50.0 45.0 17.0 55.0 38.0
20.0 n/a 28.0 50.0 21.0 32.0 20.0 41.0 34.0
11.7 16.3 25.1 51.0 47.4 36.9 13.4 46.0 n/a
n/a n/a n/a n/a n/a n/a n/a n/a n/a
n/a n/a 34.0 71.0 61.0 57.0 16.0 66.0 56.0
n/a n/a 38.0 63.0 29.0 45.0 16.0 52.0 43.0
9.7 9.9 22.8 54.8 41.4 34.5 9.4 46.3 13.4
8.0 17.2 19.8 45.2 11.0 22.8 10.5 38.8 9.0
Note: Estimated prevalence of body dissatisfaction (BD) among United States adults is determined by participant responses on each item of the Body Areas Satisfaction subscale of the Multi-dimensional Body Self-Relations Questionnaire. N/A: data is not available from previously published research.
BD of our sample to previously published national samples. Results show that the overall prevalence of BD is varied, depending on the specific measurement tool and cut-off protocol chosen (single-item vs. multiple-item mean scores; conservative vs. liberal cut-off points). When using operational definitions from previous research, the range of BD is 13.4%–31.8% among women and 9.0%–28.4% among men. These ranges are very consistent with the Frederick et al. (2012) estimates of male BD (10%–30%). However, the Frederick et al. (2012) estimate of BD among women (20%–40%) is slightly higher than the range suggested by the present study. Indeed, future studies of BD prevalence using more methodologically sound procedures will be beneficial for more precisely estimating this range (Fiske, Fallon, Blissmer, & Redding, under review). With regard to BD variation across sub-groups, women had greater BD compared to men, with the size of the effect being small for body areas satisfaction (d = .23) and fitness evaluation (d = .28), and approaching medium for overweight preoccupation (d = .45). These effect sizes are consistent with previous research (Feingold & Mazzella, 1998). When examining differences in BD across age groups, a nonstatistically significant trend emerged, such that younger participants (18–24 years) and older adults (N 65 years), reported greater health evaluation (p = .09) and body areas satisfaction (p = .08) than middle-aged groups. This corresponds with previous research suggesting that older men and women report higher body image then their middle-aged counterparts (Feingold & Mazzella, 1998; Grogan, 2011). For race/ethnicity, there were no statistically significant differences for any of the MBSRQ subscales, although there was a strong trend toward a difference between African–American/Black and
White/Caucasian adults for overweight preoccupation. This corresponds with previous research suggesting that African–American adults have more positive body image than White/Caucasian adults, but that the size of the effect is likely small (Grabe & Hyde, 2006). It is important to note that the substantial inequality of sample sizes among racial/ethnic groups in this study likely led to the violation of homogeneity of variances, and a larger sample of racial/ethnic minority groups with homogenous variances may, in fact, find statistical differences across groups. For BMI, our results show that, compared to the prevalence of BD among healthy weight women (7%–14%, as measured by mean scores for body areas satisfaction and appearance evaluation), the prevalence of BD among overweight women is double (33–42%), and the prevalence of BD among obese women is quadruple (49%–69%). This pattern was generally replicated for the prevalence of BD among healthy weight, overweight and obese men. Multivariate analysis revealed that women in the underweight category had lower BD than men in the underweight category, while women in the overweight and obese categories had greater BD compared to their overweight and obese male counterparts. This corresponds with previous research showing that generally, BD increases with weight/BMI (Schwartz & Brownell, 2004), and that men report less BD than women (Feingold & Mazzella, 1998), except when underweight women are closer to the United States cultural ideal (thin ideal for women) than underweight men (muscular ideal for men; McCabe & Ricciardelli, 2004). While not a primary aim of the study, an important finding from the point of view of overweight and obesity as a chronic disease risk factor is that while BD regarding physical appearance is higher in overweight and obese groups, the percent of overweight and obese adults that are
Table 8 Comparisons of BD prevalence among national United States adult samples over time as determined by scale means. Cash et al. (1986)
Cash and Henry (1995)
Fallon, Harris & Johnson (2013) Cash Protocol (mean b 3.0)
Body areas satisfaction Appearance evaluation Overweight preoccupation Fitness evaluation Health evaluation
Frederick Protocol (mean b 2.75)
Women
Men
Women
Men
Women
Men
Women
Men
%
%
%
%
%
%
%
%
n/a 31.0 n/a 29.0 14.0
n/a 24.0 n/a 19.0 9.0
35.6 47.9 48.5 n/a n/a
n/a n/a n/a n/a n/a
28.7 31.8 38.6 30.2 13.0
21.2 28.4 18.2 19.0 8.8
19.2 26.2 21.8 30.2 10.4
12.2 22.9 8.8 19.0 6.3
Note: Estimated prevalence of body dissatisfaction (BD) among United States adults is determined by mean scores on each subscale of the Multi-dimensional Body Self-Relations Questionnaire. N/A: data is not available from previously published research.
E.A. Fallon et al. / Eating Behaviors 15 (2014) 151–158
satisfied with their health and fitness was quite high in these groups. Specifically for fitness evaluation, 69.4%–84.9% of overweight and obese men and 40.8%–71.8% of overweight and obese women were defined as being satisfied with their fitness. Regarding health evaluation, 72.8%–94.0% of overweight and obese men and 56.7%–85.0% of overweight and obese women were defined as satisfied with their health. Such positive perceptions of health and physical fitness among overweight and obese adults may hinder the likelihood of making health behavior changes needed to reduce risk of disease. As previous research has suggested, it could be that some level of BD is beneficial for initiating healthy behavior changes (Heinberg, Thompson, & Matzon, 2001; Schwartz & Brownell, 2004). Finally, the third aim of this study was to compare our findings with that of previously published national samples. Generally, the results of the present study show that the prevalence of BD in the Unites States has not increased above the levels noted from 1973 to 1997. In fact, when focusing on frequency of men and women reporting dissatisfaction with “overall appearance” (Table 7), prevalence of BD appears to have decreased, reverting back to 1973 levels. When using mean scores of the MBSRQ subscales to define BD (Table 8), BD has decreased from the 1990's (body areas satisfaction, appearance evaluation, overweight preoccupation) or remained relatively stable over time, since the 1970's and 1980's (fitness evaluation, health evaluation). While one explanation of this finding is that programs designed to facilitate body appreciation at all ages and sizes have been successful, it is also important to note the numerous methodological limitations of previous prevalence research (Fiske, Fallon, Blissmer, & Redding, under review; Frederick et al., 2012) that hinder a strong conclusion that national prevalence of BD has changed in any direction. Similar to its predecessors, this study of the national prevalence of BD also has several limitations. The sample was a non-random convenience sample, not representative of US adult population in terms of sex, ethnicity, or education. Additionally, we did not assess sexual orientation, a known moderating variable for BD. Finally, the reliability for overweight preoccupation was not optimal, indicating the need for studies examining the psychometric properties of this subscale for a national sample. Although some limitations are evident, this study has several notable strengths. First, we were able to recruit a rather large sample of United States adults, representing at least 37 states and a wide age range. We chose a measure of body image/BD that is recommended for use for both men and women (Thompson, 2004), and has been used in several national US surveys in the past – making our results easily comparable to previous research. Second, we expanded our assessment to include fitness evaluation and health evaluation which are important for gaining a broader perspective on body image, but increasingly important to examining the role of body image with regard to health behaviors that put people at risk for chronic disease. Third, we provided a well-rounded estimate of prevalence by implementing three methods for calculating prevalence (representing conservative and liberal estimates; Fiske, Fallon, Blissmer, & Redding, under review) and calculated prevalence across sex, age, race/ ethnicity, and BMI category within the same study. Finally, as noted by Frederick et al. (2012) the Psychology Today surveys (Berscheid et al., 1973; Cash et al., 1986; Garner, 1997) had different introductions leading up to the survey, providing different rationales for conducting the survey. These forwards may have influenced participant responses and participant self-selection into the study. In an attempt to recruit participants with a wide range of BD, and a wide range of interest in health as a topic, our introductory email requesting participation was brief and did not mention body image or BD, specifically. In conclusion, few studies have sought to examine the national, overall prevalence of BD in the United States using psychometrically sound measurement tools since 1997. Prevalence of BD ranges between 13.4% and 31.8% among women and 9.0%–28.4% among men. Compared to previous national samples using similar measurement tools, we can preliminarily conclude that women remain at higher risk for BD than men and that the prevalence of BD may have plateaued or even declined
157
over time. Future research in this field should employ randomly selected, nationally representative samples and explore the mental, physical, and behavioral health outcomes for individuals who are at both ends of the BD spectrum. Role of Funding Sources There were no funding sources. Contributors Fallon and Harris developed the manuscript concept and measurement protocols. Under the supervision of Fallon and Harris, Johnson submitted the IRB, translated the survey into an online data collection program, recruited participants and managed the data. Fallon analyzed the data and wrote the first draft of the manuscript. All authors consulted throughout the process and contributed to the manuscript. All authors approved the final manuscript. Conflict of Interest All authors declare that they have no conflict of interest.
References American Psychological Association (2010). Writing for the behavioral and social sciences. Publication Manual of the American Psychological Association (pp. 13–15) (6th edition ). New YorkWashington, DC.: American Psychological Association. Berscheid, E., Walster, E., & Bohrnstedt, G. (1973). Body image. The happy American body: A survey report. Psychology Today, 7, 119–131. Cash, T. F. (2000). The multidimensional body-self relations questionnaire (3rd ed.) (Available from the author at www.body-images.com) Cash, T. F. (2003). Body image: Past, present, and future. Body Image, 1, 1–5. http://dx.doi.org/ 10.1016/S1740-1445(03)00011-1. Cash, T. F., & Henry, P. E. (1995). Women's body images: The results of a national survey in the USA. Sex Roles, 33, 19–28. http://dx.doi.org/10.1007/BF01547933. Cash, T. F., Winstead, B. W., & Janda, L. H. (1986). The great American shape-up: Body image survey report. Psychology Today, 20, 30–37. Centers for Disease Control, Prevention (2008). Smoking-attributable mortality, years of potential life lost, and productivity losses —United States, 2000–2004. Morbidity and Mortality Weekly Report, 57, 1226–1228 (Retrieved from http://www.cdc.gov/ mmwr/preview/mmwrhtml/mm5745a3.htm) Feingold, A., & Mazzella, R. (1998). Gender differences in body image are increasing. Psychological Science, 9, 190–195. Fiske, L., Fallon, E. A., Blissmer, B., & Redding, C (under review). Prevalence of body dissatisfaction among United States adults: A review and recommendations for future research. Eating Behaviors. Frederick, D. A., Forbes, G. B., Grigorian, K. E., & Jarcho, J. M. (2007). The UCLA body project I: Gender and ethnic differences in self-objectification and body satisfaction among 2206 undergraduates. Sex Roles, 57, 317–327. http://dx.doi.org/ 10.1007/s11199-007-9251-z. Frederick, D. A., Jafary, A.M., Gruys, K., & Daniels, E. A. (2012). Surveys and the epidemiology of body image dissatisfaction. In T. F. Cash (Ed.), Encyclopedia of body image and human appearance, vol. 2. (pp. 766–774)San Deigo, CA: Elsevier. http://dx.doi.org/ 10.1016/B978-0-12-384926-7.00121-3. Frederick, D. A., Peplau, L. A., & Lever, J. (2006). The swimsuit issue: Correlates of body image in a sample of 52,677 heterosexual adults. Body Image, 3, 413–419. http://dx.doi.org/10.1016/j.bodyim.2006.08.002. Garner, D.M. (1997). The 1997 body image survey results. Psychology Today, 1, 30–44. Grabe, S., & Hyde, J. S. (2006). Ethnicity and body satisfaction among women in the United States: A meta-analysis. Psychological Bulletin, 132, 622–640. Grogan, S. (2011). Body image development in adulthood. In T. F. Cash, & L. Smolak (Eds.), Body image: A handbook of science, practice and prevention (pp. 93–100) (2nd ed.). New York: Guilford Press. Heinberg, L. J., Thompson, K., & Matzon, J. L. (2001). Body image dissatisfaction as a motivator for healthy lifestyle change: Is some distress beneficial? In R. H. Streigel-Moore, & L. Smolak (Eds.), Eating Disorders: Innovative directions in research and practice, vol. 2. (pp. 215–232)Washington, DC: American Psychological Association. Howlader N, Noone AM, Krapcho M, Garshell J, Neyman N, Altekruse SF, Kosary CL, Yu M, Ruhl J, Tatalovich Z, Cho H, Mariotto A, Lewis DR, Chen HS, Feuer EJ, Cronin KA (eds). SEER Cancer Statistics Review, 1975-2010, National Cancer Institute. Bethesda, MD, http://seer.cancer.gov/csr/1975_2010/, based on November 2012 SEER data submission, posted to the SEER web site, April 2013. Hudson, J. I., Hiripi, E., Pope, H. G., & Kessler, R. C. (2007). The prevalence and correlates of eating disorders in the National Comorbidity Survey Replication. Biological Psychiatry, 61, 348–358. http://dx.doi.org/10.1016/j.biopsych.2006.03.040. Johnson, P., Fallon, E. A., Harris, B. S., & Burton, B. (2013). Body Satisfaction is associated with Transtheoretical Model constructs for Physical Activity Behavior Change. Body Image: An International Journal of Research, 10, 163–174. http://dx.doi.org/10.1016/j.bodyim. 2012.12.002 PMID: 23339856. King, T. M., Matacin, M., White, K. S., & Marcus, B. H. (2005). A prospective examination of body image and smoking cessation in women. Body Image, 2, 19–28. http://dx.doi.org/10.1016/j.bodyim.2005.01.003. Kruger, J., Lee, C., Ainsworth, B. E., & Macera, C. A. (2008). Body size satisfaction and physical activity levels among men and women. Obesity, 16, 1976–1979. http://dx.doi.org/ 10.1038/oby.2008.311.
158
E.A. Fallon et al. / Eating Behaviors 15 (2014) 151–158
McCabe, M., & Ricciardelli, L. (2004). Weight and shape concerns of boys and men. In J. K. Thompson (Ed.), Handbook of eating disorders and obesity (pp. 606–634). New York: Wiley. Peplau, L. A., Frederick, D. A., Yee, C., Maisel, N., Lever, J., & Ghavami, N. (2009). Body image satisfaction in heterosexual, gay and lesbian adults. Archives of Sexual Behavior, 38, 713–725. http://dx.doi.org/10.1007/s10508-008-9378-1. Ridolfi, D. R., & Crowther, J. H. (2013). The link between women's body image disturbances and body-focused cancer screening behaviors: A critical review of the literature and a new integrated model for women. Body Image, 10, 149–162. http://dx.doi.org/10.1016/j.bodyim.2012.11.003. Rodin, J., Silberstein, L. R., & Striegel-Moore, R. H. (1984). Women and weight: A normative discontent. In T. B. Sonderegger (Ed.), Nebraska symposium on motivation: Psychology and gender (pp. 267–307). Lincoln, NE: University of Nebraska Press. Schwartz, M. B., & Brownell, K. D. (2004). Obesity and body image. Body Image: An International Journal of Research, 1, 43–56. http://dx.doi.org/10.1016/S1740-1445(03) 00007-X.
Stuhldreher, N., Konnopka, A., Wild, B., Herzog, W., Zipfel, S., Löwe, et al. (2012). Cost-of-illness studies and cost-effectiveness analyses in eating disorders: A systematic review. International Journal of Eating Disorders, 45, 476–491. Tantleff-Dunn, S., Barnes, R. D., & Larose, J. G. (2011). It's not just a “woman thing:” The current state of normative discontent. Eating Disorders: The Journal of Treatment & Prevention, 19(5), 392–402. http://dx.doi.org/10.1080/10640266.2011.609088. Thompson, J. K. (2004). The (mis)measurement of body image: Ten strategies to improve assessment for applied and research purposes. Body Image, 2, 7–14. http://dx.doi.org/10.1016/S1740-1445(03)00004-4. U.S. Department of Health and Human Services, & Centers for Disease Control, Prevention (2010). Behavioral risk factor surveillance system survey questionnaire. Retrieved from. http://www.cdc.gov/brfss/questionnaires.htm U.S. Department of Health and Human Services, National Institutes of Health, & National Heart Lung and Blood Institute (1998). Clinical guidelines on the identification, evaluation, and treatment of overweight and obesity in adults (NIH publication no. 98-4083). Retrieved from. http://www.nhlbi.nih.gov/guidelines/ obesity/ob_gdlns.pdf
