Main International Journal of Eating Disorders Stigmatizing attitudes and beliefs about bulimia nervosa: Gender, age, education and income...

Stigmatizing attitudes and beliefs about bulimia nervosa: Gender, age, education and income variability in a community sample

, , , , ,
How much do you like this book?
What’s the quality of the file?
Download the book for quality assessment
What’s the quality of the downloaded files?
International Journal of Eating Disorders
May, 2014
PDF, 102 KB

To post a review, please sign in or sign up
You can write a book review and share your experiences. Other readers will always be interested in your opinion of the books you've read. Whether you've loved the book or not, if you give your honest and detailed thoughts then people will find new books that are right for them.

Burundi: An ongoing search for durable peace

PDF, 134 KB

Stigmatizing Attitudes and Beliefs About Bulimia
Nervosa: Gender, Age, Education and Income
Variability in a Community Sample
Si^an A. McLean, BSc (Hons)1*
Susan J. Paxton, PhD1
Robin Massey, BPsySc (Hons)1
Phillipa J. Hay, MD, PhD2
Jonathan M. Mond, PhD3
Bryan Rodgers, PhD4

Objective: Stigmatizing attitudes towards
eating disorders negatively impacts
treatment seeking. To determine the
effect of interventions to reduce stigma,
a measure of stigma that is simple to
implement is required. This study
aimed to develop a measure of stigmatizing attitudes and beliefs towards
bulimia nervosa (SAB-BN) and evaluate
the distribution of beliefs across gender,
age, education, and income groups.
Method: Participants were 1828 community adults (890 men; 938 women)
aged 18–65 sampled from the Australian
Electoral Roll responded to a mailed
demographic information and completed
the SAB-BN questionnaire.
Results: Five components of stigmatizing attitudes and beliefs were identified;
advantages of BN, minimization/low seriousness, unreliability, social distance, and
personal responsibility. Stigma was low
except on social distance and personal

Bulimia nervosa (BN) is common in the community with a lifetime prevalence of 1.5% among
women using DSM-IV criteria1 and is associated

Accepted 25 October 2013
Supported by DP1095656 from Australian Research Council.
Author BR supported by Fellowship 471429 from National Health
& Medical Research Council.
*Correspondence to: Ms. Si^
an McLean, School of Psychological
Science, La Trobe University, Melbourne, Victoria, Australia.
School of Psychological Science, La Trobe University, Melbourne, Victoria, Australia
Centre for Health Research, School of Medicine, University of
Western Sydney, Sydney, Australia & School of Medicine, James
Cook University, Townsville, Australia
Research School of Psychology, The Australian National University, Canberra, ACT, Austral; ia
Australian Demographic & Social Research Institute, The Australian National University, Canberra, ACT, Australia
Conflicts of Interest: The authors declare no conflicts of interest, financial or otherwise.
Published online 26 November 2013 in Wiley Online Library
( DOI: 10.1002/eat.22227
C 2013 Wiley Periodicals, Inc.

International Journal of Eating Disorders 47:4 353–361 2014

responsibility sub-scales, which indicated
negative attitudes toward people with
bulimia. Men compared with women
and lower compared with higher education and income groups held significantly
higher stigmatizing attitudes and beliefs.
There were few differences between age
groups in stigma. Differences between
demographic groups provides evidence
for known-groups validity.
Discussion: The SAB-BN questionnaire
provides a potentially useful tool for evaluating stigma in relation to BN. Results
provide insight into components of
stigma and the demographic groups to
whom interventions should be targeted.
C 2013 Wiley Periodicals, Inc.
Keywords: stigma; bulimia nervosa;
questionnaire measure; demographic
(Int J Eat Disord 2014; 47:353–361)

with high levels of functional impairment.2
Although treatments are fairly successful,3 treatment seeking is very low,4 due in part to fear of
stigmatization and shame.5 Stigmatizing attitudes
and beliefs toward eating disorders are likely to
reduce well-being of sufferers as people with BN
fear stigmatization, engage in self-stigma and
experience shame, which impede treatment seeking.5–8 To inform the development of effective
public health interventions to reduce stigma surrounding eating disorders, and BN in particular,
the focus of this study, it would be valuable to
have a comprehensive picture of stigmatizing attitudes and beliefs across a representative range of
demographic groups. Understanding demographic variability enables tailoring of public
health messages to groups in the community who
are likely to be most in need of these. This study
explores the nature and distribution of stigmatizing beliefs about bulimia in a large community
sample following the development of an instrument to assess these beliefs.


Stigma has been defined as a process by which
the labelling of individuals as different from the
norm identifies the individual as possessing undesirable characteristics, which in turn leads to discrimination and social rejection.9 Dimensions of
stigma in relation to mental illness include perceived dangerousness, personal responsibility,
poor prognosis, and avoidance of social interaction.10 Stigmatizing attitudes toward people with
mental illnesses are prevalent and widespread.11
However, research has shown that women are less
likely than men to endorse personal responsibility
for the illness,12 and less likely to desire social distance from individuals with mental illnesses.13
Findings in relation to age differences in stigmatizing beliefs about mental illness are conflicting.
Some studies find that younger age is associated
with lower levels of stigma in relation to desire for
social distance (e.g., as shown in Ref. 14), while
others show greater stigmatization demonstrated
by younger participants on dimensions of social
distance and personal responsibility.15 In addition,
people with higher levels of education are less
likely to hold stigmatizing beliefs related to dangerousness than those with less education.12
A number of studies have explored stigmatizing
beliefs about eating disorders in general and anorexia
nervosa and BN in particular. Stigma in relation to
eating disorders appears to be different from stigma
related to other mental illness, in that perceived dangerousness and the unpredictability dimension of
stigma are rated lower.16 However, beliefs about personal responsibility for the illness have been found
to be higher for eating disorders than other mental
illnesses in college students,17,18 community samples,16,19 and among medical professionals.20
In addition, people with eating disorders are
thought to use the illness to gain attention.18,21 A
number of factors suggest that eating disorders are
trivialized and believed not to need proper treatment.22 These include the perception that it might
not “be too bad” to have an eating disorder given
the possibility of losing weight,18,21,23 and the opinion that people with eating disorders should “pull
themselves together” to recover.16,19
Gender differences in stigmatizing attitudes
towards eating disorders are consistent with those
demonstrated for other mental illnesses. Mond and
Arrighi24 found that undergraduate men considered both anorexia nervosa and BN to be less serious than did women. Similarly, Wingfield et al.25
found that undergraduate men were more likely
than women to think that recovery from eating disorders was easy, thus minimizing the seriousness

of the illness, but in contrast, women rated people
with eating disorders as more self-destructive.
Although these findings provide preliminary
information about stigmatizing attitudes and beliefs
towards eating disorders, they are confined almost
entirely to studies of undergraduate students or
young adult women. Although this age group frequently makes up the peer environment of individuals with an eating disorder, older adults can have
eating disorders as well, and contribute to the social
environment of people with eating disorders in
important ways. However, very little is known about
the attitudes and beliefs of middle-aged and older
men and women. Thus, to guide the development
of public health interventions in community subgroups to contribute to improved treatment seeking
in individuals with BN, it would be valuable to have
a deeper understanding of variability in attitudes
about this disorder across demographic dimensions
including age, education, and income levels.
To facilitate this research, we have developed a
new questionnaire measure of stigmatizing attitudes and beliefs about BN (SAB-BN) to complement and advance current assessment methods. In
previous research, one approach has been to present participants with a vignette describing an individual with an eating disorder followed by
questions about the person described in the
vignette (e.g., as shown in Ref. 22). Another measurement approach has been to use single items to
assess dimensions of stigma (e.g., as shown in Refs.
16, 18, and 19). A multi-item scale would provide a
more comprehensive measure in which related
components could be grouped together.
Thus, to facilitate an investigation of community
attitudes about BN, we aimed first to determine the
factor structure of the SAB-BN measure and to analyze its psychometric properties. It was hypothesized that the factor structure of the measure
would correspond to previously observed elements
of stigmatization including personal responsibility,
attention seeking, trivialization, seriousness of the
illness, and desire for social distance. We next
aimed to examine the distribution of stigmatizing
attitudes and beliefs about BN amongst different
demographic groups within the community. It was
hypothesized that men, younger participants, and
participants with lower education and income levels would exhibit higher levels of stigmatizing attitudes towards BN than women, older participants,
and participants with higher education and income
levels. Differences between groups will be interpreted for known-groups validity.

International Journal of Eating Disorders 47:4 353–361 2014


Participants were recruited from a random sample of
Victorian adults from the Australian Electoral Roll. As it
is compulsory for Australian adults 18 years and older to
vote, the electoral roll provides a convenient database
from which to recruit participants. The sample was
stratified according to gender (men, women), age group
(18–34, 35–45, and 46–65), socioeconomic status (low,
medium, high; derived from the SEIFA index of relative
advantage/disadvantage according to postcode),26 and
location (metropolitan, regional). The stratification was
conducted in order to obtain a sample that represented
the Victorian population on these socio-demographic
Of the 9,357 adults who were invited to participate in
the study, 2,099 responded to the invitation and provided
data (response rate 5 22.4%). Responses from 271 participants were not usable due to missing data. Thus, the
final sample consisted of 1,828 participants.
Reflecting the distribution in the Victorian population,
participants were 51.3% women and 48.7% men.27 In comparison with the Victorian population, younger
(sample 5 25.9% vs. population 5 36.6%) and middle-aged
(sample 5 27.2% vs. population 5 34.9%) groups were
under-represented. The opposite was true for the older
age group (sample 5 46.9% vs. population 5 28.5%). The
distribution of participants across SES groups was approximately comparable to the Victorian population, low SES
group 31.9% (sample) vs 35.9% (population), medium SES
group 41.1 vs. 30.3%, high SES group 27.0 vs. 33.8%. More
participants were from metropolitan (N 5 1,230) than
regional areas (N 5 598). However, the proportion of participants from regional areas was higher than the proportion in the Victorian population (32.7 vs. 27.0%).
The majority of participants were born in Australia
(79.4%) and identified their racial group as being Caucasian/White (89.6%). Almost 70% of participants were
married or living in a de facto relationship (69.8%), the
remainder were single. Just over half the sample were
employed full-time (52.8%), about one quarter were
employed part time (25.5%), a small proportion were
unemployed and looking for work (3.4%), and 17.7%
were not in paid employment. The latter participants
were engaged in home duties, retired, studying, engaged
in a caring role or doing voluntary work. Less than one
percent of the sample (0.5%) did not provide information
about their employment status. In relation to educational
attainment for the sample, just over one fifth (21.3%) did
not complete high school, 32.8% completed either high
school or a trade certificate, 39.5% completed an undergraduate degree or diploma, and 6.0% completed a postgraduate qualification. A small proportion (0.5%) did not
provide education information.
International Journal of Eating Disorders 47:4 353–361 2014

The development of the SAB-BN scale was based on
themes identified from studies examining stigma
towards eating disorders,5,18,19,21,28–31 and other mental
illnesses.12,32 These themes reflected beliefs about personal responsibility for BN, lack of perceived seriousness of
BN, admiration or acceptance of aspects of the disorder,
and desire for social distance from individuals with BN.
Items to assess perceived dangerousness and unpredictability of people with bulimia were not included as these
dimensions are less relevant to eating disorders than to
other mental illnesses.16,19 Initially, 6–8 items were developed by the authors to tap each of these themes, resulting in thirty items that were then pilot tested for clarity
and comprehension (unpublished data). Items that were
deemed to be ambiguous, had a high proportion of nonresponses, or were identified (after prompting) by pilot
study participants as being difficult to understand were
revised or removed. Further items were added to improve
clarity. This resulted in a pool of 32 items for inclusion in
the current study. Items were subject to tests of readability using Microsoft word software to ensure an appropriate level of comprehension. The Flesch-Kincaid Grade
level was 9.0, indicating that the items should have been
able to be understood by an average 9th grade student. A
brief definition of BN was provided (“Bulimia is an eating
disorder (a type of mental illness) that involves binge eating and using extreme ways to control weight. Binge eating means eating more than normal in a short period of
time and losing control of eating. Examples of extreme
ways to control weight are excessive exercise, extreme
dieting, vomiting and abuse of laxatives”) and participants responded to the items on a 6-point scale
(1 5 completely disagree, 2 5 mostly disagree, 3 5 slightly
4 5 slightly
5 5 mostly
6 5 completely agree).
Scores for sub-scales identified from Principal Components Analysis (PCA) were calculated for each participant
as the mean of the items loading onto each component.
Where required, items were reverse scored so that for
sub-scale totals higher scores reflected higher stigmatizing attitudes and beliefs.
Participants provided self-report information for their
age, gender, marital status, country of birth, racial group,
educational attainment, employment status, and income

The La Trobe University Human Ethics Committee
approved the study and participation involved
informed consent. Paper surveys were mailed to
potential participants in nine separate mail-outs
(1000 per mail-out, staggered for logistical

TABLE 1. Factor loadings and cross-loadings from the PCA (rotated factor solution) and CFA for the five component
solution for the SAB-BN items
Pattern coefficients
20. Bulimia is not that bad, because you can eat what you like
without putting on weight.
4. Bulimia goes away without treatment given time.
10. Bulimia is a good way of balancing eating and dieting.
3. There are good things about having bulimia.
1. Bulimia is not a serious problem compared with other
psychological problems.
2. People making themselves sick after binge eating is
nothing to worry about.
17. Bulimia is just a phase that some girls and young women go through.
26. Using laxatives after binge eating is a serious problem.a
32. Fasting (going without food) after binge eating is a serious problem.a
16. Bulimia is a serious eating disorder.a
7. Bulimia with deliberate vomiting is a serious problem.a
21. Bulimia with excessive exercise is a serious problem.a
28. Bulimia is an unacceptable way to control weight.a
30. I would not depend on someone with bulimia to complete
an important task.
14. People with bulimia are poor at their jobs.
13. People with bulimia are unreliable.
22. I wouldn’t employ someone who had bulimia.
27. I’d find it difficult to trust someone with bulimia.
24. I’d feel comfortable dating someone with bulimia.
23. I’d have no problem being friends with someone with bulimia.a
25. Deep down, I would be ashamed of a family member if
they had bulimia.
29. People with bulimia should be less hung up about their looks.
12. People with bulimia should stop obsessing about their
looks and weight.
15. People with bulimia should work on their self-control.
18. People with bulimia should just “pull themselves together”
and “get over it”.
19. Bulimia is a form of attention-seeking.






0.64 (0.77)





0.60 (0.49)
0.59 (0.68)
0.59 (0.50)
0.58 (0.50)





0.54 (0.43)





0.54 (0.60)

0.72 (0.63)
0.72 (0.59)
0.62 (0.61)
0.59 (0.50)
0.57 (0.36)
0.55 (0.46)

20.79 (.81)





20.78 (0.74)
20.77 (0.73)
20.67 (0.63)
20.66 (0.75)

0.71 (20.29)
0.66 (20.41)
20.44 (0.58)






20.75 (0.53)
20.75 (0.53)





20.66 (0.63)
20.56 (0.74)





20.44 (0.63)

Major loadings on components for the pattern coefficients for each item are bolded. Values not in brackets are from the PCA. Values in brackets are loadings from the CFA.
Reverse scored items.
Component 1: Advantages of BN; Component 2: Minimization/Low Seriousness; Component 3: Unreliability; Component 4: Social Distance; Component
5: Personal Responsibility.

reasons). Included in the questionnaire package
were a personalized letter of invitation to participate in a study on community beliefs about eating
problems, and a reply paid envelope for return of
the survey. An A$10 voucher was provided for
return of completed surveys. Nonresponders were
sent an additional questionnaire 1 month following the original questionnaire.

Data Analysis

To examine the factor structure of the SAB-BN
measure, the total sample was randomly split into
two equal sized sub-samples, stratified by age and
gender. Data from the first sub-sample were subject to exploratory PCA, using oblique (Direct Oblimin) rotation to aid interpretation. Data were
examined for suitability for exploratory PCA and all

criteria were met. Sixty-nine cases which had values
above the critical values on Mahalanobis’ distance
were deemed to be outliers and were omitted from
the PCA. The final sample size for sub-sample one
was 840. A Monte Carlo simulation for parallel analysis33 with 100 randomly generated samples of 32
(variables) by 840 (participants) was conducted to
confirm the number of components to be retained.
Criteria for excluding items were as follows; factor
loadings <0.4, loading on more than one component in which none of the loadings were >0.5, and
items with communality values <0.2.34
Data from the second sub-sample were subject
to confirmatory factor analysis (CFA). Prior to conducting the CFA, multivariate outliers were identified using critical values of Mahalanobis’ distance
and removed (N 5 58). The final sample size for the
CFA was 862. The model was specified on the basis
International Journal of Eating Disorders 47:4 353–361 2014


Correlation coefficients between identified components and Cronbach’s alpha values for components

1. Advantages of BN
2. Minimization / Low Seriousness
3. Unreliability
4. Social Distance
5. Personal Responsibility









Cronbach’s alpha



N 5 1828.
p < 0.001.
p < 0.01.

of the outcome of the PCA analysis. Model fit was
assessed with chi-square (small and nonsignificant
values indicate good fit, although in large samples
chi-square will almost always be significant),
Goodness of Fit Index, Confirmatory Fit Index, Parsimony Comparative Fit Index (GFI, CFI,
PCFI;  0.95 indicates good fit) and Root Mean
Square Error of Approximation (RMSEA  0.05 indicates good fit).35
Distributions of sub-scale scores, calculated for
items loading onto each of the identified components, were examined for normality. Most variables
showed strong positive or negative skew. Consequently, examination of differences between demographic groups on the SAB-BN sub-scales was
conducted with nonparametric tests. The MannWhitney U tests were used for dichotomous independent variables (gender) and the Kruskal-Wallis
tests for ordinal independent variables.

Factor Structure of the SAB-BN

The PCA revealed seven components with eigenvalues greater than one. Examination of the scree
plot revealed breaks after the third and fifth components. Support was provided from the results of
the parallel analysis to retain five factors. The five
component solution explained cumulative variance
of 46.94%.
Following oblique rotation the pattern matrix
showed that the majority of items loaded strongly
onto the five components, with the exception of
item 31 which loaded weakly (<0.4) and was therefore excluded (see Table 1). Items 5, 8, 9, and 11
were excluded because they loaded on more than
one component and none of the loadings were
>0.5. Item 6 was also excluded due to a low communality value (0.196).
Data from the second random sample were subject to CFA using the results from the PCA to identify the model to be tested. Twenty-six items were
included with five components (see Table 1). The
International Journal of Eating Disorders 47:4 353–361 2014

initial model had poor fit to the data v2
(289) 5 1302.54, p 5 0.000, GFI 5 0.887, CFI 5 0.855,
PCFI 5 0.760, RMSEA 5 0.064. Inspection of the
modification indices showed large values corresponding to the errors between items 1 and 2, 13
and 14, 12 and 29, 23 and 24, and 27 and 30. These
items were covaried and the analysis repeated. The
second model showed good fit to the data
p 5 0.000,
GFI 5 0.927,
(v2(284) 5 814.24,
CFI 5 0.924, PCFI 5 0.810, and RMSEA 5 0.047).
In the final supported model, seven items were
retained for component one, which was titled
“Advantages of BN.” This component reflects a perception that BN is a disorder for which there may
be some advantages (related to weight loss) and
which is relatively inconsequential. Higher scores
indicated greater perceived advantages of BN. Six
items were retained for component two which was
titled “Minimization/Low Seriousness.” The items
indicated the degree to which BN is considered a
serious illness. Higher scores are indicative of high
levels of minimization and low perceived seriousness of BN. Five items loaded onto the third component, titled “Unreliability.” Higher scores on this
component reflected the perception that individuals with BN are unreliable or untrustworthy. Three
items loaded onto the fourth component, titled
“Social Distance”. Higher scores reflected lack of
comfort with personal interactions with people
who have BN. The final component of five items
that we titled “Personal Responsibility” indicated a
tendency to believe that individuals with BN were
in some way responsible or to blame for their own
illness. Zero-order correlations between components for the total sample are shown in Table 2.

Internal Consistency of the SAB-BN Sub-scales

The items loading on each component were
treated as sub-scales and Cronbach’s alpha values
were calculated for the total sample for each subscale (see Table 2). Most values were above the
acceptable cut-off of 0.7.36 However, the value for
the three-item component, Social Distance, was

TABLE 3. Means, standard deviations, and medians for
sub-scales of the SAB-BN for the total sample and gender
and age groups
Mean (SD)
Advantages of BN
Total sample
1.64 (0.65)
1.76 (0.68)
1.52 (0.60)
Age 18–34
1.61 (0.60)
Age 35–46
1.65 (0.65)
Age 46–65
1.65 (0.68)
Minimization/Low Seriousness
Total sample
1.82 (0.87)
1.96 (0.89)
1.68 (0.82)
Age 18–34
1.83 (0.86)
Age 35–46
1.78 (0.79)
Age 46–65
1.82 (0.91)
Total sample
2.29 (1.00)
2.41 (0.95)
2.18 (1.04)
Age 18–34
2.28 (0.93)
Age 35–45
2.32 (1.03)
Age 46–65
2.28 (1.03)
Social Distance
Total sample
2.68 (0.98)
2.70 (0.97)
2.66 (0.99)
Age 18–34
2.76 (0.96)
Age 35–45
2.75 (0.99)
Age 46–65
2.60 (0.98)
Personal Responsibility
Total sample
2.92 (1.07)
3.09 (1.04)
2.76 (1.08)
Age 18–34
2.99 (1.00)
Age 35–45
2.94 (1.08)
Age 46–65
2.88 (1.11)

95% CI


1.61, 1.67
1.72, 1.81
1.49, 1.56
1.55, 1.66
1.59, 1.71
1.61, 1.70


1.77, 1.86
1.90, 2.02
1.63, 1.73
1.76, 1.81
1.72, 1.85
1.76, 1.88


2.25, 2.34
2.35, 2.48
2.11, 2.25
2.20, 2.37
2.23, 2.41
2.21, 2.35


2.64, 2.73
2.64, 2.77
2.60, 2.73
2.67, 2.85
2.66, 2.84
2.54, 2.67


2.87, 2.97
3.03, 3.16
2.69, 2.83
2.90, 3.08
2.84, 3.03
2.80, 2.95


Total sample (N 5 1,828); Men (N 5 890); Women (N 5 938); Age 18–34
(N 5 471); Age 35–45 (N 5 496); Age 46–65 (N 5 857).
Higher sub-scale scores reflect higher stigma.

low (0.50). This is not unusual for a component
with a small number of items, in which case it is
recommended to use the mean inter-item correlation as an indication of acceptability, a satisfactory
range being between 0.2 and 0.4.37 The mean interitem correlation for the Social Distance component
was r 5 0.25.

Community Sample Scores on SAB-BN Subscales

In general, reports of stigmatizing attitudes were
fairly low, especially in relation to Advantages of
BN, Minimization/Low Seriousness and Unreliability (See Table 3). However, on two indicators of
stigma, Social Distance and Personal Responsibility, stigmatizing attitudes may be considered quite
high. The median responses were 2.7 and 3.0,
respectively, corresponding to the slightly disagree
anchor for the scale, indicating only slight disagreement with items indicative of social distanc358

ing from and personal responsibility and blame for
the disorder. If stigma is defined as a score that corresponds to slightly agree, mostly agree, or completely agree, the rates of stigma were 21.6% for
Social Distance and 30.3% for Personal
Differences Between Demographic Groups on
SAB-BN Sub-scales

Means, standard deviations, 95% confidence
intervals and median values for the total sample
and for gender and age groups for each SAB-BN
sub-scale are shown in Table 3. Examination of gender differences revealed that men compared with
women had significantly higher scores on Advantages of BN (U 5 318,010, z 5 28.880, p < 0.001,
r 5 20.21), Unreliability (U 5 348,856, z 5 26.097,
p < 0.001, r 5 20.14), Minimization/Low Seriousness (U 5 328,826, z 5 27.944, p < 0.001, r 5 20.19),
(U 5 339,803,
z 5 26.891, p < 0.001, r 5 20.16). There were no significant gender differences for Social Distance.
Examination of the three age groups revealed a
H(2) 5 11.03, p 5 0.004. Post-hoc tests showed that
the 46–65 year old age group had lower levels of
Social Distance than both the 18–34 year old age
group, U 5 183,607, z 5 22.859, p 5 0.004, r 5 20.08,
and the 35–45 age group, U 5 195,338, z 5 22.613,
p 5 0.009, r 5 20.07. No differences between age
groups were found on other sub-scales.
To examine differences in attitudes between levels of formal education, highest educational attainment was collapsed into three levels. Participants
who had completed secondary schooling or less,
had completed an apprenticeship or certificate,
and completed some level of university education
were categorized as being of lower, medium or
higher education. The only sub-scale on which significant differences were observed was on Unreliability, H(2) 5 11.14, p 5 0.004. The lower educated
group (Md 5 2.4) had significantly higher scores on
Unreliability than the higher educated group
(Md 5 2.0), U 5 248,583, z 5 22.726, p 5 0.006,
r 5 20.07. The medium educated group (Md 5 2.2)
also had significantly higher levels of Unreliability
than the higher educated group, U 5 124,919,
z 5 22.782, p 5 0.005, r 5 20.08.
Household income was collapsed into three levels, low, medium, and high, representing income
brackets of <A$50,000, A$50,000 to A$100,000, and
>A$100,000, respectively. Significant differences
were observed between income groups on Unreliability H(2) 5 9.52, p 5 0.009 and Personal
International Journal of Eating Disorders 47:4 353–361 2014


Responsibility, H(2) 5 6.79, p 5 0.03. Post-hoc tests
indicated that low income participants (Md 5 2.4)
had higher scores on Unreliability than both
medium (Md 5 2.2), U 5 172, 778, z 5 22.442,
p 5 0.02, r 5 20.07, and high income participants
(Md 5 2.2), U 5 144,439, z 5 22.837, p 5 0.005,
r 5 20.08. In addition, low income participants
(Md 5 3.0) had higher scores on Personal Responsibility than high income participants (Md 5 2.8),
U 5 146519, z 5 22.454, p 5 0.01, r 5 20.07. Significant differences for income level were not found
on the other sub-scales. Descriptive information
for education and income groups for other subscales is available from the authors upon request.

To facilitate our exploration of stigmatizing attitudes and beliefs within a diverse community sample, we developed a new questionnaire and
evaluated its scale properties. The final 26-item
questionnaire demonstrated structural validity
with five components identified. These were used
as sub-scales and named Advantages of BN, Minimization/Low Seriousness, Unreliability, Social
Distance, and Personal Responsibility.
The sub-scales established in our questionnaire
were consistent with previous research using other
methods, for example, identification of attitudes
indicating that people with BN are in some ways
responsible for their disorder,18,19,23 that BN is not
a particularly serious disorder (e.g., as shown in
Ref. 25), and beliefs that BN has some advantages.8,18 Our analysis suggests that additional
components can be identified including perceived
lack of trust mainly related to employment and
lack of comfort with having close personal interactions with people with BN, consistent with studies
of stigma in mental illness more generally (e.g., as
shown in Ref. 10).
It is a positive sign that the majority of people in
our community sample did not believe that BN had
advantages, regarded BN as a fairly serious disorder,
and did not express high levels of stigma related to
trust, especially in the workplace. It is essential for
these beliefs to be present in the community as a
foundation for early identification of the disorder
and support for treatment seeking to be provided to
people with BN. There were, unfortunately more
frequently held attitudes that reflected negatively
on the person with BN. There were higher levels of
stigma related to social distance and personal
responsibility, suggesting that a sizable proportion
of the community believe that people with BN are
International Journal of Eating Disorders 47:4 353–361 2014

in some ways responsible for their disorder. An
appreciation of these beliefs in the community by
individuals with bulimia is likely to increase shame,
which may lead to self-stigma and reduce treatment
seeking in individuals with BN.5
Examination of demographic differences in stigmatizing attitudes in the community sample indicated fairly uniform attitudes. However, the
strongest differences in attitudes between demographic groups were found between men and
women. Consistent with previous research into
mental illness generally and eating disorders in
particular,12,15,24 men had more stigmatizing attitudes about BN than women and this was most
apparent on the Personal Responsibility sub-scale.
The consistency of our findings with previous
research, that males have higher levels of stigma
than females, provides evidence of known-groups
validity for the SAB-BN scale. Although men represent a significant minority of people with disordered eating in the community,38 they may be less
likely than women to be exposed to accurate information about BN or to the extreme appearance
pressures that may contribute to this disorder.
However, overcoming this deficit with welldesigned and targeted public health messages is
crucial as men are frequently in situations in which
they could potentially provide support for treatment seeking.8
There were some differences in stigmatizing attitudes and beliefs between age, education, and
income groups but they were not of great magnitude or consistency. It is notable however, that the
differences between these demographic groups
were for sub-scales that had the highest levels of
stigma. These findings are relevant to the design of
public health interventions as these appear to be
the areas for which intervention is most needed
and they indicate that similar interventions may be
effective across age, education, and income groups,
which could increase economic efficiency.
The main strengths of this research were the use
of a carefully developed measure of stigmatizing
attitudes towards BN in a large community sample
with wide gender, age, education, and income representation. However, there were also limitations.
The major limitation with the self-report questionnaire was the likelihood of a social desirability bias
in which participants are very likely to have underreported stigmatizing attitudes39 and this needs to
be taken into account when interpreting the findings. A further limitation was that although a representative sampling procedure was used the
response rate of 22.4% was low, although typical
for mailed eating disorder surveys with a similar


methodology.40 In addition to young men being
less likely to respond, it is possible that people with
a specific interest in eating disorders, mental
health, or community issues were more likely to
respond. The latter group of participants may be
more sympathetic to people with mental illnesses,
thus levels of stigma may have been underestimated. In addition, people with less flexible time
may have been less likely to respond. These factors
may limit the generalizability of the findings. The
low internal consistency of the social distance subscale was also a limitation. Further research could
examine the properties of this sub-scale following
inclusion of additional items to strengthen the
scale properties.
As treatment seeking for BN is low partly as a
result of fear of stigmatization leading to selfstigmatization, it is essential to understand
community beliefs about BN to inform the development of effective public health interventions.
The development in this research of a measure of
stigma has provided a tool to assess these beliefs.
Future research may also be extended to understand attitudes towards other eating disorders such
as anorexia nervosa and binge eating disorder. In
addition, tracking attitudes across time would provide valuable information about changing levels of
stigma. The findings from this study indicate that
developing a public health intervention to reduce
beliefs about personal responsibility related to BN
that are particularly persuasive in men could
reduce stigma toward BN across the community.

1. Hudson JI, Hiripi E, Pope HG Jr., Kessler RC. The prevalence and correlates
of eating disorders in the National Comorbidity Survey Replication. Biol Psychiatry 2007;61:348–358.
2. Mond JM, Hay P, Rodgers B, Owen C, Crosby R, Mitchell J. Use of extreme
weight control behaviors with and without binge eating in a community
sample: Implications for the classification of bulimic-type eating disorders.
Int J Eat Disord 2006;39:294–302.
3. Hay P. A systematic review of evidence for psychological treatments in eating disorders: 2005–2012. Int J Eat Disord 2013;46:462–469.
4. Hart LM, Granillo MT, Jorm AF, Paxton SJ. Unmet need for treatment in
the eating disorders: A systematic review of eating disorder specific
treatment seeking among community cases. Clin Psychol Rev 2011;31:727–
5. Hepworth NS, Paxton SJ. Pathways to help-seeking in bulimia nervosa and
binge eating problems: A concept mapping approach. Int J Eat Disord 2007;
6. Becker AE, Arrindell AH, Perloe A, Fay K, Striegel-Moore RH. A qualitative
study of perceived social barriers to care for eating disorders: Perspectives
from ethnically diverse health care consumers. Int J Eat Disord 2010;43:
7. Hackler AH, Vogel DL, Wade NG. Attitudes toward seeking professional help
for an eating disorder: The role of stigma and anticipated outcomes. J Couns
Dev 2010;88:424–431.


8. Mond JM, Hay PJ, Paxton SJ, Rodgers B, Darby A, Nillson J, et al. Eating disorders "Mental Health Literacy" in low risk, high risk and symptomatic
women: Implications for health promotion programs. Eat Disord J Treat
Prev 2010;18:267–285.
9. Jones EE, Farina A, Hastorf AH, Markus H, Miller DT, Scott RA. Social Stigma:
The Psychology of Marked Relationships. New York: W. H. Freeman and
Company, 1984.
10. Hayward JA, Bright P. Stigma and mental illness: A review and critique.
J Ment Health 1997;6:345–354.
11. Griffiths KM, Nakane Y, Christensen H, Yoshioka K, Jorm AF, Nakane H.
Stigma in response to mental disorders: a comparison of Australia and
Japan. BMC Psychiatry 2006;6:21.
12. Corrigan PW, Watson A. The stigma of psychiatric disorders and the gender,
ethnicity, and education of the perceiver. Community Ment Health J 2007;
13. H€
ogberg T, Magnusson A, L€
utzen K, Ewalds-Kvist B. Swedish attitudes
towards persons with mental illness. Nord J Psychiatr 2012;66:86–96.
14. Lauber C, Nordt C, Falcato L, Rossler W. Factors influencing social distance
toward people with mental illness. Community Ment Health J 2004;40:265–
15. Reavley NJ, McCann TV, Jorm AF. Mental health literacy in higher education
students. Early Interv Psychiatry 2012;6:45–52.
16. Crisp AH, Gelder MG, Rix S, Meltzer HI, Rowlands OJ. Stigmatisation of people with mental illnesses. Br J Psychiatry 2000;177:4–7.
17. Ebneter DS, Latner JD, O’Brien KS. Just world beliefs, causal beliefs, and
acquaintance: Associations with stigma toward eating disorders and obesity.
Pers Indiv Differ 2011;51:618–622.
18. Roehrig JP, McLean CP. A comparison of stigma toward eating disorders versus depression. Int J Eat Disord 2010;43:671–674.
19. Stewart M-C, Keel PK, Schiavo RS. Stigmatization of anorexia nervosa. Int J
Eat Disord. 2006;39:320–325.
20. Fleming J, Szmukler GI. Attitudes of medical professionals towards patients
with eating disorders. Aust N Z J Psychiatry 1992;26:436–443.
21. Mond JM, Robertson-Smith G, Vetere A. Stigma and eating disorders: Is there
evidence of negative attitudes towards anorexia nervosa among women in
the community? J Ment Health 2006;15:519–532.
22. Mond JM, Hay PJ, Rodgers B, Owen C, Beumont PJV. Beliefs of women concerning the severity and prevalence of bulimia nervosa. Soc Psychiatry Psychiatr Epidemiol 2004;39:299–304.
23. Mond JM, Hay P, Rodgers B, Owen C. Mental health literacy and eating
disorders: What do women with bulimic eating disorders think and
know about bulimia nervosa and its treatment? J Ment Health 2008;17:
24. Mond JM, Arrighi A. Gender differences in perceptions of the severity and
prevalence of eating disorders. Early Interv Psychiatry 2011;5:41–49.
25. Wingfield N, Kelly N, Serdar K, Shivy VA, Mazzeo SE. College students’ perceptions of individuals with anorexia and bulimia nervosa. Int J Eat Disord
26. Australian Bureau of Statistics. Information Paper: An Introduction to SocioEconomic Indexes for Areas (SEIFA) 2006. Canberra, ACT: Australian Bureau
of Statistics, 2008.
27. Australian Bureau of Statistics. 2006 Census of Population and Housing.
2007; Available at:, access on 3 March
28. Mond JM, Hay PJ, Rodgers B, Owen C, Beumont PJV. Beliefs of the public
concerning the helpfulness of interventions for bulimia nervosa. Int J Eat
Disord 2004;36:62–68.
29. Mond JM, Hay PJ, Rodgers B, Owen C, Beumont PJV. Beliefs of women concerning causes and risk factors for bulimia nervosa. Aust N Z J Psychiatry
30. Crisp AH. Stigmatization of and discrimination against people with eating
disorders including a report of two nationwide surveys. Eur Eat Disord Rev
31. Crisafulli MA, Von Holle A, Bulik CM. Attitudes towards anorexia nervosa:
The impact of framing on blame and stigma. Int J Eat Disord 2008;41:333–

International Journal of Eating Disorders 47:4 353–361 2014

32. Angermeyer MC, Holzinger A, Matschinger H. Mental health literacy and attitude towards people with mental illness: A trend analysis based on population surveys in the eastern part of Germany. Eur Psychiat 2009;24:225–232.
33. Watkins MW. Monte Carlo PCA for parallel analysis. State College, PA: Ed &
Psych Associates, 2000.
34. Kline P. The Handbook of Psychological Testing, 2nd ed. London: Routledge, 2000.
35. Hu L-T, Bentler PM. Cutoff criteria for fit indexes in covariance structure
analysis: Conventional criteria versus new alternatives. Struct Equation
Model: A Multidisciplinary J 1999;6:1–55.
36. Nunnally JC. Psychometric theory, 2nd ed. New York: McGraw-Hill, 1978.

International Journal of Eating Disorders 47:4 353–361 2014

37. Briggs SR, Cheek JM. The role of factor analysis in the development and evaluation of personality scales. J Pers 1986;54:106–148.
38. Mitchison D, Mond J, Slewa-Younan S, Hay P. Sex differences in healthrelated quality of life impairment associated with eating disorder features:
A general population study. Int J Eat Disord 2013;46:375–380.
39. Peris TSP, Teachman BAP, Nosek BAP. Implicit and explicit stigma of mental
illness: Links to clinical care. J Nerv Ment Dis 2008;196:752–760.
40. Striegel-Moore RH, Rosselli F, Perrin N, DeBar L, Wilson GT, May A, et al.
Gender difference in the prevalence of eating disorder symptoms. Int J Eat
Disord 2009;42:471–474.