Vol.

11: 1-13, March 2006

ORIGINAL
RESEARCH The Body Uneasiness Test (BUT):
PAPER Development and validation of a new
body image assessment scale
M. Cuzzolaro*, G. Vetrone**, G. Marano***, and P.E. Garfinkel****
*Department of Child and Adolescent Psychiatry, Eating Disorders Unit, University of Rome La Sapienza,
** Department of Philosophical Research, University of Rome Tor Vergata, *** Department of Psychology,
University of Bologna, Italy, and **** Centre for Addiction and Mental Health, University of Toronto, ON,
Canada

ABSTRACT. Objective: To investigate the psychometric properties of the Body Uneasiness

Test (BUT), a 71-item self-report questionnaire that consists of two parts: BUT•A which mea-
sures weight phobia, body image concerns, avoidance, compulsive self-monitoring, detach-
ment and estrangement feelings towards one’s own body (depersonalization); and BUT•B
which looks at specific worries about particular body parts or functions. Methods: We
recruited a clinical sample of 531 subjects (491 females) suffering from eating disorders and a
general population sample of 3273 subjects (2016 females) with BMI <25 and Eating Attitudes
Test-26 scores under the cut-off 20. Results: The exploratory and confirmatory analyses
confirmed a structural five-factor model for BUT•A and an eight-factor model for BUT•B.
Internal consistency was satisfactory. The test-retest correlation coefficients were highly sig-
nificant. Concurrent validity with other tests (Eating Disorder Inventory, EDI-2; Eating
Attitudes Test, EAT-26; Symptom Check List, SCL-90R and Coopersmith Self-Esteem
Inventory, SEI) was evaluated. Normative values for BUT scores in non-clinical samples of
normal-weight non eating disordered subjects, from adolescence to old age, males and
females, were calculated. The differences between males and females were highly significant,
above all in the 18-39-age range. As for the comparison between women with eating disor-
ders and controls, the results demonstrated a good predictive validity for anorexia nervosa
and bulimia nervosa. Conclusions: The BUT is psychometrically sound. It can be a valuable
tool for the screening and the clinical assessment of abnormal body image attitudes and eat-
ing disorders.
(Eating Weight Disord. 11: 1-13, 2006). ©2006, Editrice Kurtis

INTRODUCTION preoccupation with a real or imagined

defect in appearance concerning a particular
Body image is a multidimensional concept area of the individual’s body. Moreover, an
that has generated a great deal of attention increasing amount of interesting research is
as people in the Western world have analyzing body image disturbances and
become preoccupied with youth, body shape their relations with eating habits in obesity
and control over personal health (1). Body (2-6). Rosen has coined the term negative
image has assumed a significant role clini- body image (7) to include all cases of clinical-
Key words: cally with regard to the eating disorders ly significant uneasiness due to body image
Body Uneasiness Test (BUT), (ED) in which affective and, at times, percep- in subjects with objective anomalies con-
body image, eating disorders, tual distortions have been documented. cerning appearance (e.g. people with obesi-
psychometrics. ty) and in subjects with basically imaginary
There are three other clinical areas in which
Correspondence to: body image is of relevance: in some neuro- or exaggerated, or at times both, aesthetic
Massimo Cuzzolaro
University of Rome logical conditions there may be a faulty per- problems [e.g. people with, anorexia ner-
La Sapienza ception of one’s own body, for example fail- vosa (AN), bulimia nervosa (BN)].
Department of Child and ure to recognize a hemiplegia after a stroke; Because of this clinical interest the psycho-
Adolescent Psychiatry some people may suffer from a physical dis- metric assessment of body image has been
Via dei Sabelli 108 of importance and a large number of psy-
00185 Roma, Italy.
figurement through illness or surgery,
E-mail: m.cuzzolaro@flashnet.it whereby one’s bodily self–perception is chometric instruments have been created to
Received: September 18, 2005 altered; another condition, body dysmorphic assess the phenomena related to it (8, 9).
Accepted: January 9, 2006 disorder (BDD), involves an exaggerated Schilder originally had characterized body

1
 Body Uneasiness Test

image as “the picture of our own body which avoidance and compulsive self-monitoring
we form in our mind, that is to say the way in behavior, detachment and/or estrangement
which the body appears to ourselves” (10). feelings towards the subject’s own body, spe-
Modifications have since been made to his cific worries towards body parts, features (e.g.
operational definition. Most noteworthy, Slade smell) or functions (e.g. sweating). We named
(11) expanded the concept of body image by the scale Body Uneasiness Test (BUT), deliber-
emphasizing the affective domain – the strong ately staying away from the term body dissatis-
reactions that individuals may entertain in faction. This seemed to fit particularly well the
response to their own bodies. He stated that it not well-defined and more general phenomena
is “the picture we have in our own minds of the we wanted to investigate. Indeed, in our opin-
size, shape and form of our bodies; and our ion, to describe these phenomena, we must use
feelings concerning these characteristics and not only dissatisfaction but also words such as
our constituent body parts” which is “influ- anxiety, alarm, trepidation, worry, mistrust,
enced by a variety of historical, cultural and misgiving, doubt, suspicion and embarrass-
social, individual and biological factors, which ment, which in English Dictionaries are usually
operate over varying time spans”. Following linked to the word uneasiness.
this many measures have focused on the affec- In 1998 we presented a preliminary validation
tive dimension dissatisfaction. study of the questionnaire at the 5 th
In developing an approach to assessing body International Conference on Eating Disorders
image we felt it was important to take a more (New York, 24-26 April 1998) and in 1999 we
broad view of the construct because: published it in Italian (14). In this preliminary
1. People with body image disturbances display study a 76-item version (39 statements and 37
not only dissatisfaction with particular body body parts) of the scale was administered to a
parts, shapes or functions, but also an clinical sample of 175 women suffering from ED
uneasiness, which is more general, less spe- (AN, BN, BED, EDNOS) according to DSM-IV
cific and often very difficult to describe (age: 32.26±11.84), who were attending an
2. Body-uneasiness can be found also among Anorexia, Bulimia and Obesity Outpatient Unit
subjects, especially teenagers, who don’t of the University of Rome, La Sapienza. The
meet full criteria needed to diagnose specific same preliminary version was administered to a
psychiatric disorders (e.g. following ICD-10 non-clinical sample of 725 high-school students
or DSM-IV-TR) (12, 13) (F 379, M 346; age 16.41±0.69). On the basis of
3. Body dissatisfaction and body uneasiness exploratory and confirmatory factor analyses, a
can arouse avoidance behavior (to the point final 71-item version (BUT•A 34 statements and
of social phobia) and compulsive checking BUT•B 37 body parts - see Appendix 1) was
behavior (e.g. long and painful self-monitor- administered to a new non-clinical sample of
ing at the mirror) 380 high-school students (F 194, M 186; age:
4. In body image disturbances, detachment and 16.4±0.73) and to a new clinical sample of 125
estrangement feelings, to the point of soma- eating disordered young women (age:
to-psychic depersonalization, are very 26.83±7.14). The BUT showed good psychomet-
important issues, but have been little investi- ric properties: the scale loading, the Cronbach’s
gated in psychometric studies. alpha values and the split-test values were very
From this view-point we can say that there high (14). The first study, however, involved
are many aspects to body uneasiness that need only young subjects: women with ED and a
overall consideration in psychometric assess- control group made up of adolescents.
ment. Accordingly, we developed a self admin- The present article describes a new validation
istered rating scale designed to explore several study of the BUT in different age groups - from
areas in clinical and non-clinical populations: adolescence to old age - and in samples of
- body shape and/or weight dissatisfaction patients suffering from ED. We chose this diag-
- avoidance nostic group because body image has a long
- compulsive control behaviors and storied association with eating and weight-
- detachment and estrangement feelings related problems (15).
towards one’s own body Three separate sections present internal con-
- specific worries about particular body parts, sistency and internal structure of the BUT, test-
shapes or functions. retest reliability, validity and normative com-
The initial instrument was developed through parisons. The English version of the BUT can
clinical interviews with some subjects affected be found in the Appendix 1 (the Italian and the
by BDD and many subjects with ED. We French versions of the test can be requested to
recorded typical and recurrent statements the Authors). Scoring instructions are reported
about: uneasiness relating to body or weight, in Appendix 2.

Eating Weight Disord., Vol. 11: N. 1 - 2006 2

 M. Cuzzolaro, G. Vetrone, G. Marano, et al.

TABLE 1 indicating greater body uneasiness). The test

Samples. was administered in a self-reported fashion
(i.e., the investigators did not assist the subjects
Age range Number of subjects Females % Males %
in the compilation of the questionnaire).
Non clinical sample
13-15 840 57.7 42.3 Data analyses
16-17 864 59.2 40.8 The 34 items of BUT•A and the 37 items of
18-39 854 73.4 26.6 BUT•B were subjected to an exploratory factor
40-65 484 75.8 24.2 analysis and a series of confirmatory factor
> 65 231 62.2 37.8
Total (13-80) 3273 61.6 38.4 analyses to find the structural equation model
best fitting with the data. The statistical analy-
Clinical sample
ses were carried out on the whole clinical sam-
13-15 7 100.0 0.0 ple and a subsample of 1886 subjects (M 638, F
16-17 20 100.0 0.0 1248), taken from the non-clinical population,
18-39 502 92.0 8.0
40-65 2 100.0 0.0
in which different age groups, socio-economic
> 65 0 0.0 0.0 and schooling levels were equally represented.
Total (13-78) 531 92.5 7.5 At first, to evaluate the structural properties
and the reliability of the dimensions of the test,
an analysis of the main components was car-
ried out for each of the subscales. The homo-
STUDY 1: INTERNAL geneity of the different subscales was further
CONSISTENCY AND INTERNAL studied by using the mean of the corrected cor-
STRUCTURE OF THE BUT relations between each item belonging to a
given scale and the total of the same (CITC) and
Participants also using Cronbach’s alpha coefficient values.
We studied a total clinical sample of 531 (F The analyses were performed using SPSS 10.0
491, M 40) subjects suffering from ED accord- (SPSS, San Diego, CA) and JMP 6 (SAS
ing to DSM-IV-TR diagnostic criteria (12) and a Institute, Cary, NC).
total non-clinical sample of 3273 (F 2016, M
1257) subjects with a BMI value <25 and an Results
EAT-26 score under the cut-off point 20 (scores The exploratory and confirmatory analyses
higher than 20 indicate possible cases of ED) confirmed the structural five-factor model for
(16). Both the clinical sample and the control BUT•A which had already emerged in the 1999
group were made up of subjects resident in dif- preliminary study. These five factors are
ferent regions in the north, centre and south of defined as follows:
Italy. A review board of the University of Rome - WP, Weight Phobia (fear of being or becom-
“La Sapienza” approved this study. Informed ing fat)
consent was obtained, prior to completion of - BIC, Body Image Concerns (worries related to
the questionnaire, from the participants, and in physical appearance)
the case of minors, from their parents. - A, Avoidance (body image related avoidance
Table 1 shows the frequency and the distribu- behavior)
tion by gender and age for the whole sample - CSM, Compulsive Self-Monitoring (compul-
according to the two sub-groups. From these sive checking of physical appearance)
populations, subsamples were derived on - D, Depersonalization (detachment and
which the statistical analyses described in this estrangement feelings toward the body).
paper were carried out. In the same way, as far as BUT•B is con-
The final version of the BUT (see Appendix 1) cerned, the exploratory and confirmatory
was administered to all the subjects, who rated analyses confirmed the structural model with
each item on a six-point Likert-type scale eight factors that had already emerged in the
(range 0-5, from “never” to “always”, high rates preliminary study.

TABLE 2
Structural models.
Model c2 DF p Dc2 p GAFI
BUT•A: Five-factor uncorrelated 1388.18 517 0.000 216.86 0.04 0.87
BUT•B: Eight-factor uncorrelated 1004.6 406 0.000 3988.6 0.0000 0.93

3 Eating Weight Disord., Vol. 11: N. 1 - 2006

 Body Uneasiness Test

TABLE 3
BUT•A Factor loading
Weight Phobia Body Image Avoidance Compulsive Depersonalization
Concern Self-Monitoring
BUT 09 0.79 BUT 03 0.70 BUT 05 0.72 BUT 01 0.66 BUT 02 0.62
BUT 10 0.81 BUT 04 0.78 BUT 08 0.83 BUT 11 0.75 BUT 07 0.71
BUT 18 0.79 BUT 06 0.72 BUT 13 0.80 BUT 17 0.75 BUT 14 0.67
BUT 21 0.63 BUT 12 0.78 BUT 16 0.80 BUT 20 0.70 BUT 26 0.80
BUT 24 0.62 BUT 15 0.76 BUT 19 0.63 BUT 27 0.72 BUT 28 0.80
BUT 31 0.71 BUT 22 0.72 BUT 30 0.74 BUT 29 0.79
BUT 32 0.73 BUT 23 0.85
BUT 33 0.72 BUT 25 0.80
BUT 34 0.86
% var. expl. 56.3 62.1 57.7 54.8 49.4

TABLE 4
BUT•B Factor loading.
I II III IV V VI VII VIII
Mouth 0.83 Face Shape 0.75 Thighs 0.84 Legs 0.80 Arms 0.78 Moustache 0.84 Skin 0.82 Blushing 0.73
Lips 0.82 Head Shape 0.73 Hips 0.81 Ankles 0.80 Chest 0.76 Beard 0.74 Hair 0.82 Sweating 0.66
Eyes 0.68 Forehead 0.71 Knees 0.79 Feet 0.73 Shoulders 0.74 Hairs 0.69 Noises 0.65
Eyebrow 0.59 Chin 0.71 Abdomen 0.79 Hands 0.67 Breasts 0.66 Buttocks 0.55
Nose 0.57 Neck 0.68 Stomach 0.78 Stature 0.66 Genitals 0.59 Odours 0.49
Teeth 0.56 Ears 0.58
% var. expl. 57.0 48.2 64.5 62.4 50.1 54.8 67.2 50.1

The BUT•B eight factors are represented by TABLE 5

Roman numerals: Homogeneity and internal consistency.
-I eyebrow, eyes, nose, mouth, lips, teeth
Subscale Mean CITC Cronbach
- II shape of the head, shape of the face,
forehead, ears, chin, neck BUT•A Weight phobia 0.71 0.84
- III stomach, abdomen, hips, thighs, knees BUT•A Body image concern 0.70 0.90
- IV stature, legs, ankles, feet, hands
-V arms, shoulders, chest, breasts, genitals BUT•A Avoidance 0.63 0.79
- VI moustache, beard, hairs BUT•A Compulsive self-monitoring 0.56 0.82
- VII hair, skin
BUT•A Depersonalization 0.68 0.85
- VIII sweating, blushing, noises, odours, but-
tocks. BUT•B I 0.70 0.88
The results are exposed in Table 2: for each BUT•B II 0.58 0.81
model the values of the chi-square and the
delta chi-square, the degrees of freedom (DF), BUT•B III 0.65 0.84
the likelihood level (p) and the Goodness of BUT•B IV 0.55 0.75
Adjusted Fit-Index (GAFI) are shown.
BUT•B V 0.77 0.90
Tables 3 and 4 show the factor loadings relat-
ing to BUT•A and BUT•B. In both Tables the BUT•B VI 0.76 0.87
last line states the percentage of variance BUT•B VII 0.48 0.69
explained by the only extracted component.
Table 5 displays other psychometric features BUT•B VIII 0.58 0.77

Eating Weight Disord., Vol. 11: N. 1 - 2006 4

 M. Cuzzolaro, G. Vetrone, G. Marano, et al.

of each subscale: mean of the Corrected Item- TABLE 6

Total Correlations (CITC) and Cronbach’s Test-retest reliability.
alpha coefficient.
Control subjects ED subjects
The suggested solution converged normally
(the maximum residual cosine was close to BUT A
zero). The Chi-square was 2784.5 per 464 Global Severity Index 0.91 0.90
degrees of freedom (p<0.21). The Steiger-Lind Weight phobia 0.90 0.92
RMSEA index was equal to 0.08, indicating a Body image concerns 0.89 0.88
very good adaptation of the model to the data,
even compensating for the parsimoniousness Avoidance 0.71 0.83
of the model itself. Compulsive self-monitoring 0.73 0.80
Depersonalization 0.85 0.94

STUDY 2: TEST-RETEST BUT B

Positive symptom total 0.85 0.75
RELIABILITY
Positive symptom distress index 0.81 0.84
Participants BUT•B I 0.87 0.86
For this section we randomly selected from
BUT•B II 0.92 0.92
the total samples 118 subjects, 38 eating disor-
dered patients (F 32, M 6; age 24.48±6.12) and BUT•B III 0.87 0.77
80 non-eating disordered normal-weight sub- BUT•B IV 0.82 0.74
jects (F 56, M 24; age 31.74±8.32). Each subject
BUT•B V 0.94 0.78
completed the BUT at a one-week interval.
Calculations were made of the correlation coef- BUT•B VI 0.81 0.68
ficients between the results of the first and sec- BUT•B VII 0.87 0.80
ond administration for the global measures
BUT•B VIII 0.78 0.71
(Appendix 2), and for each subscale of BUT•A
and BUT•B.

Results
The test-retest correlation coefficients were the analyses of the multi-method/multi-trait
highly significant for both samples. They are matrices was not easily applicable: on the one
reported in Table 6. hand it was difficult to interpret large sized
matrices and on the other, high correlations
were present between different subscales of
STUDY 3: CONCURRENT AND each single test. We then carried out a second
DISCRIMINANT VALIDITY, order factorial analysis. The details will be the
NORMATIVE COMPARISONS object of a separate publication: we will present
AND PREDICTIVE VALIDITY here only the conclusions.

A. Concurrent and discriminant validity Results

Participants The first analysis compared the BUT•A sub-
For this section of the study we randomly scales with the BUT•B subscales. Two canoni-
selected 188 subjects from the non-clinical cal correlations were significant (both with
sample among the 16-year old boys and girls p<0 .001). The variables most associated with
(104 F, 84 M). To evaluate concurrent validity, the first canonical correlation were Body
the participants were administered, besides the Image Concern and Weight Phobia of BUT•A
BUT, the EDI-2 (17), the EAT-26 (16), the with the subscales III and IV of BUT•B. The
Derogatis SCL90-R (18) and the Coopersmith variables most associated with the second
Self-Esteem Inventory, SEI (19). canonical correlation were Avoidance and
Depersonalization with the subscales II and V
Data analyses of BUT•B. The Bartlett test showed that there
We calculated the canonical correlations weren’t any other significant correlations.
between the BUT•A and the BUT•B subscales The second analysis compared the BUT•A
and those between the BUT subscales and the subscales and the EDI-2 subscales.
EDI-2 subscales. As far as concurrent validity Two canonical correlations were significant
between the BUT subscales and those of the (both with p<0.05): Weight Phobia and Body
other tests are concerned, given the large num- Image Concerns with EDI Drive for Thinness
ber of subscales, the analysis systems based on and Body Dissatisfaction; and, Compulsive Self-

5 Eating Weight Disord., Vol. 11: N. 1 - 2006

 Body Uneasiness Test

Monitoring, Depersonalization and Avoidance TABLE 7

with EDI Ineffectiveness, Interpersonal Distrust Comparison between the BUT•A subscales and those
and Social Insecurity. of EAT-26, SCL-90R and SEI.
The third analysis compared the BUT•B sub-
Scales First Rank Factors
scales and the EDI-2 subscales. Two canonical
correlations were significant (both with p<0.05). SCL-Anxiety 1°
In the first we found the subscale III correlated SCL-Somatization 1°
with EDI Drive for Thinness and Interoceptive
SCL-Obsession-compulsion 1°
Awareness. In the second, the subscales I, II, V
and VII were correlated with Social Insecurity, SCL-Depression 1°
Interoceptive Awareness and Asceticism. SCL-Psychoticism 1°
As regards the comparison between the
SCL-Paranoid ideation 1° - 4°
BUT•A subscales and those of the other tests
(EAT-26, SCL-90R and SEI), in view of our BUT - Depersonalization 1° - 2° - 3°
objectives, we will only comment on the way in SCL-Phobic anxiety 1° - 3°
which the subscales were related to the latent
BUT-Weight phobia 2°
factors of first order (Table 7).
In Table 7 we see how the SCL subscales con- BUT-Body image concern 2°
tributed to constituting a first latent factor. This BUT-Compulsive self-monitoring 2°
did not apply to the Hostility subscale that con-
BUT-Avoidance 2° - 3°
nected with another latent factor (the fourth).
All the BUT•A subscales contributed to con- EAT-Dieting 2° - 3°
stituting a second factor. The EAT subscales EAT-Bulimia 3°
formed the third factor and two of the SEI sub-
EAT-Oral control 3°
scales the fifth. The BUT•A subscales exclusive-
ly linked to the second factor were Weight SEI-Home 4°
Phobia, Body Image Concern and Compulsive SCL-Hostility 4°
Self-Monitoring. The Avoidance subscale also
SEI-School 5°
connected with the factor linked to the EAT
and the Depersonalization subscale of the BUT SEI-Social 5°
shared both the first and the third latent factor.
In view of these results we tried rereading
the correlations matrix between the single sub-
scales. We found that the link between the BUT Phobic Anxiety (r=0.53) subscale and with the
Depersonalization subscale and the factor con- SCL Psychoticism (r=0.51) subscale.
nected with the SCL subscales took place On the other hand, the link of the same
because of the high correlations with the SCL Depersonalization subscale with the third latent

TABLE 8
BUT•A scores in a non-clinical sample: means (standard deviations).
Age 13-15 16-17 18-39 40-65 >65
Global severity index M 0.75 (0.66) 0.74 (0.62) 0.45 (0.52) 0.58 (0.67) 0.54 (0.48)
F 1.51 (1.02) 1.31 (0.78) 1.32 (0.91) 0.90 (0.81) 0.98 (0.67)
Weight phobia M 0.95 (0.88) 0.98 (0.95) 0.70 (0.65) 0.95 (0.55) 0.75 (0.33)
F 1.91 (1.16) 1.97 (1.17) 1.92 (1.21) 1.35 (0.91) 1.36 (1.05)
Body image concern M 0.94 (0.88) 1.06 (0.72) 0.48 (0.59) 0.67 (0.45) 0.56 (0.40)
F 1.86 (1.12) 1.65 (0.94) 1.62 (1.19) 1.05 (0.93) 1.20 (0.86)
Avoidance M 0.42 (0.64) 0.29 (0.70) 0.13 (0.38) 0.22 (0.30) 0.50 (0.47)
F 0.86 (1.02) 0.61 (0.86) 0.56 (0.85) 0.45 (0.69) 0.50 (0.51)
Compulsive self-monitoring M 0.89 (0.81) 0.83 (0.81) 0.63 (0.55) 0.72 (0.39) 0.63 (0.45)
F 1.64 (1.02) 1.37 (0.85) 1.61 (1.01) 0.96 (0.64) 0.99 ( 0.58)
Depersonalization M 0.33 (0.69) 0.26 (0.59) 0.14 (0.45) 0.27 (0.33) 0.08 (0.11)
F 0.80 (1.03) 0.64 (0.98) 0.84 (1.12) 0.40 (0.67) 0.48 (0.60)

Eating Weight Disord., Vol. 11: N. 1 - 2006 6

 M. Cuzzolaro, G. Vetrone, G. Marano, et al.

TABLE 9 As regards the Global Severity Index the

Predictive validity for eating disorders: samples. interaction between sex and age groups was
significant (p=0.00034). In males there were no
Diagnosis (number of subjects) Age BMI
significant differences between the different
ANbp (65) 23.17±5.77 15.80±1.55 age groups, whereas the females under 40 had
ANr (108) 20.71±5.34 15.65±1.80 significantly higher scores compared to the
others. In the Weight Phobia subscale the
BNp (142) 22.96±5.08 21.31±5.08
interaction was significant (p=0.01): the
Non-ED (477) 20.37±9.38 20.46±2.74 females appeared more worried than the males
Total (792) 21.11±8.09 19.58±3.75 in the first three age groups; however, in the
two age groups 40-65 and >65 there was no
ANbp: anorexia nervosa binge-purging type; ANr: anorexia nervosa restrictive significant difference between males and
type; BNp: bulimia nervosa purging type; Non-ED: non eating disorder
females. On the Body Image Concern subscale,
the interaction was significant (p=0.00002) in
all the age groups except the fourth (40-65). On
the Avoidance subscale, the interaction was
factor, connected with the EAT, depends large- significant (p=0.0004) only in the second (16-
ly on the high correlation with the EAT Bulimia 17) and the third (18-39) age groups. On the
(r=0.46) subscale. Finally, the Avoidance BUT Compulsive Self-Monitoring subscale the
subscale connected with the EAT subscales fac- interaction was significant (p=0.0003) for the
tor by virtue of the particularly elevated link first three age groups. At last, on the
with the EAT Dieting subscale. Depersonalization subscale the females’ scores
were significantly greater (p<0.00001) than
B. Predictive validity for sex and age those of males only in the first (13-15) and the
Participants third (18-39) age groups.
For this segment of the study we examined
the whole non-clinical sample of 3273 (F 2016, C. Predictive validity for eating
M 1257) normal-weight and EAT-26 low-scor- disorders
ers (under the cut-off point 20). The aim was to Participants
calculate the capacity of BUT to distinguish The purpose of this study was to evaluate the
between subjects on the basis of sex and age. capacity of the BUT to single out the
We used the Global Severity Index score (GSI, presence/absence of an ED, using the scores of
mean BUT• A score) and those of the five sub- the five BUT•A subscales. In addition, we won-
scales of BUT•A. dered if the differences of BUT scores were to
be attributed to the presence of an ED or to dif-
Data analyses ferences in age and/or BMI. Finally, we calcu-
A multivariate analysis of variance was car- lated for all subscales sensitivity and specificity
ried out. Contrast analysis was used for all the taking into account the different diagnostic cat-
subscales, considering the interactions with egories, age levels and BMI values.
p<0.05 significant. For this part of our study we extracted from
the clinical sample 315 females suffering from
Results AN full syndrome or BN full syndrome. The
Table 8 shows the scores obtained by males diagnosis was formulated on the basis of the
and females on the different BUT•A subscales semi-structured clinical interview, EDE 12.0D
for the different age groups. (20). The EDNOS were excluded because the

TABLE 10
Predictive validity for eating disorders: analyses of covariance.
Observed values (expected values for age 21.11 and BMI 19.58).
WP BIC A CSM D
ANbp 3.14* (2.82)* 2.56* (1.63) 1.88* (1.42)* 1.99* (1.61) 1.89* (1.42)*
ANr 2.73* (2.86)* 2.28* (1.62) 1.59* (1.40)* 2.03* (2.42)* 1.59* (1.39)*
BNp 3.31* (3.24)* 2.67* (2.49)* 1.79* (1.66)* 2.35* (2.37)* 1.80* (1.66)
Non-ED 1.73 (1.65) 1.46 (1.35) 0.48 (0.43) 1.16 (1.10) 0.48 0.43

*p<0.05; ANbp: anorexia nervosa binge-purging type; ANr: anorexia nervosa restrictive type; BNp: bulimia nervosa purging type; Non-ED: non eating disorder

7 Eating Weight Disord., Vol. 11: N. 1 - 2006

 Body Uneasiness Test

diagnostic criteria for partial syndromes are Results

much less clear. Then we selected from the Table 10 summarizes the results of the
total non clinical sample a control group of 477 analyses of covariance. It displays, for each
normal-weight EAT-26-low-scoring females diagnostic group, the observed scores and, in
with comparable age. Males were excluded between brackets, the expected scores for the
because the clinical sample lacked the neces- mean age (21.11) and BMI (19.57) values of the
sary numbers. Table 9 shows age and BMI val- total sample. The observed scores of the clini-
ues of the clinical and control groups. cal groups (ANr, ANbp and BNp) were always
significantly different from those of the con-
Data analyses trol group. Taking into account age and BMI,
A set of analyses of covariance (ANCOVA) the expected scores of the clinical groups
was carried out in order to evaluate the dis- (ANr, ANbp and BNp) were also significantly
criminatory ability of the BUT•A subscales. In different from those of the control group with
each analysis we used a BUT•A subscale as few exceptions: BIC and CSM for ANr, BIC
dependent variable, age and BMI as covariates for ANbp.
and the diagnosis (ANr, ANbp, BNp and non- Referring to the discriminant analyses, we
ED) as factor. Then, according to the results of found for each diagnostic group a significant
the ANCOVA, we performed a set of discrimi- discriminant function. Table 11 summarizes the
nant analyses to compare each diagnostic results. It displays, for each diagnostic group,
group with a control group matched for age. the variables (BUT•A scales) that were inserted
We used the stepwise method: at each step, the into the discriminant function, Wilks’ lambdas,
variable that minimizes the overall Wilks’ lamb- significance and percentage of correctly classi-
da is entered. At last, we used the Receiver fied subjects.
Operating Characteristics (ROC) curves to cal- In Table 12 we suggest for each diagnostic
culate sensitivity and specificity of the ques- category, age group and BMI level one BUT•A
tionnaire against the clinical diagnosis. scale with the cutoff point that according to the

TABLE 11
Predictive validity for eating disorders: discriminant analyses.
AGE
<19 19-26.9 > 26.9
BMI V Wl % cc V Wl % cc V Wl % cc
ANr <18.5 D 0.63 ** 82 D, WP 0.64 ** 87.3 A 0.80 * 68.2
ANbp <18.5 A, D 0.41 ** 92.5 WP, D 0.50 ** 87.5 D 0.61 * 81
BNp <18.5 D, A, CSM 0.68 ** 86.3 WP, BIC 0.36 ** 94.1 CSM 0.35 * 90.9
BNp 18.5-25.0 D, CSM 0.85 ** 86.8 D, A, BIC 0.59 ** 80.6 WP, A 0.50 ** 89.7

V: variable (BUT•A scale); Wl: Wilk’s lambda; %cc: percentage of correctly classified subjects. *p<0.05; **p<0.001
ANbp: anorexia nervosa binge-purging type; ANr: anorexia nervosa restrictive type; BNp: bulimia nervosa purging type

TABLE 12
Predictive validity for eating disorders: cutoffs.
AGE
<19 19-26.9 > 26.9
BMI BUT•A scale sens. spec. BUT•A scale sens. spec. BUT•A scale sens. spec.
(cutoff) (cutoff) (cutoff)
ANr <18.5 D (0.5) 90% 72% D (0.5) 90% 92% A (0.6) 66% 84%
ANbp <18.5 A (1.0) 100% 88% D (0.5) 97% 92% D (0.5) 93% 67%
BNp <18.5 D (1.0) 80% 85% D (0.5) 95% 92% D (0.5) 99% 100%
BNp 18.5-25.0 D (1.0) 100% 84% D (0.5) 96% 70% WP (1.5) 93% 78%

ANbp: anorexia nervosa binge-purging type; ANr: anorexia nervosa restrictive type; BNp: bulimia nervosa purging type

Eating Weight Disord., Vol. 11: N. 1 - 2006 8

 M. Cuzzolaro, G. Vetrone, G. Marano, et al.

ROC curves ensured, in our opinion, the most found some significant and understandable
balanced sensitivity/specificity ratio. correlations: Weight Phobia and Body Image
Concerns are correlated with Drive for
DISCUSSION Thinness and Body Dissatisfaction; Compulsive
Self-Monitoring, Depersonalization and
Internal Consistency and factor structure Avoidance are correlated with Ineffectiveness,
The BUT’s internal consistency is satisfactory Interpersonal Distrust and Social Insecurity.
in terms of both the significant homogeneity As for the comparison between the subscales
(indicated by the Item-Total Correlations) and of BUT•A and those of EAT-26, SEI and SCL-
the clearly one-dimensional structure shown by 90R, the results of the second order factorial
each of the subscales according to the analyses analysis seem to reflect certain interesting clini-
of the main components. The levels of cal data: all the five BUT•A subscales make up
Cronbach’s alpha coefficients range between the same factor to which three of these (Weight
0.69 (only BUT•B VII) and 0.90. Therefore all the Phobia, Body Image Concern and Compulsive
subscales but one showed acceptable consisten- Self-Monitoring) are connected exclusively,
cy as they were greater than 0.7 (21, 22). The fac- without correlating with any subscales of the
torial structure studied via confirmatory analy- other tests. This seems to indicate a particular
ses is characterised by a good adaptation of the autonomy and specificity on their part.
structural model hypothesised to the experimen- On the other hand, the BUT•A Avoidance
tal data: the quotient chi2 /DF falls within the subscale shows correlations with subscales of
level considered as highly acceptable (from 2 to the other tests, indicative of a high concurrent
5, in accordance with Marsh et al. (23). validity. The Avoidance subscale has a particu-
larly significant link with the Dieting subscale
Test-retest reliability of EAT-26 made up of items many of which
The test-retest reliability too is satisfactory: all refer to explicit avoidance behavior (e.g. Avoid
the subscales but one (BUT•B VI in the clinical foods with sugar in them) or to feelings which
sample) showed good test-retest reliability as the tend to cause them (e.g. Like my stomach to be
correlation coefficients were greater than 0.7 (22). empty or Feel extremely guilty after eating).
The Depersonalization subscale shows high
Concurrent and discriminant validity correlations with the Phobic Anxiety and
The comparison between BUT•A and BUT•B Psychoticism subscales of the SCL-90R and with
showed that certain BUT•A subscales are signif- the Bulimia subscale of the EAT-26. In spite of
icantly correlated with certain BUT•B subscales: criticisms about the validity of the SCL-90R sub-
Body Image Concern and Weight Phobia with scales, these data appear interesting because
the subscales III (stomach, abdomen, hips, they are consistent with certain well-known clin-
thighs, knees) and IV (stature, legs, ankles, feet, ical phenomena. The association between the
hands); Avoidance and Depersonalization with depersonalization symptoms, panic attacks and
the subscales II (shape of the head, shape of the phobic behavior and the fact that Body
face, forehead, ears, chin, neck) and V (arms, Dysmorphic Disorder and Social Phobia seem to
shoulders, chest, breasts, genitals). Some of belong to the same spectrum is not surprising.
these correlations appear comprehensible: the The correlation with the Psychoticism subscale
first is that between Body Image Concern, might correspond with the psychotic character
Weight Phobia and the BUT•B III and IV sub- of certain phenomena explored by items which
scales. BUT•B III includes thighs, hips, abdomen make up the BUT Depersonalization subscale
and stomach, body parts which in both sexes - (e.g. I don’t trust my appearance: I’m afraid it
but surely much more in women - are consid- will change suddenly). Finally, as regards the
ered particularly unpleasant and socially stigma- high correlation with the EAT-26 Bulimia sub-
tised indicators of overweight. BUT•B IV scale, we might mention the links indicated
includes legs and stature: it would seem plausi- many times between binge eating, body shame
ble that worries about shortness are associated and dissociative phenomena (24-26).
with greater worries about overweight. Another The Coopersmith Self-Esteem Inventory and
significant correlation is that between the the BUT do not share any latent factor.
Depersonalization and Avoidance subscales and These data suggest that the BUT is a specific
the BUT•B II and BUT•B V subscales. These two questionnaire with respect to the other psycho-
subscales include body parts and physical fea- metric instruments with which it was compared.
tures (e.g. face, head, breasts, genitals) which
seem especially linked to the personal and gen- Predictive validity
der identity of the subjects. We performed an extensive normative study
In comparing the BUT with the EDI-2 we for the Italian population: this article presents

9 Eating Weight Disord., Vol. 11: N. 1 - 2006

 Body Uneasiness Test

APPENDIX 1
BUT•A
Mark with an X the answer which best expresses your experience at the moment
never seldom sometimes often very often always
1 I spend a lot of time in front of the mirror 0 1 2 3 4 5
2 I don’t trust my appearance: I’m afraid it will change suddenly 0 1 2 3 4 5
3 I like those clothes which hide my body 0 1 2 3 4 5
4 I spend a lot of time thinking about some defects of 0 1 2 3 4 5
my physical appearance
5 When I undress, I avoid looking at myself 0 1 2 3 4 5
6 I think my life would change significantly if I could 0 1 2 3 4 5
correct some of my aesthetic defects
7 Eating with others causes me anxiety 0 1 2 3 4 5
8 The thought of some defects of my body torments me 0 1 2 3 4 5
so much that it prevents me being with others
9 I’m terrified of putting on weight 0 1 2 3 4 5
10 I make detailed comparisons between my appearance 0 1 2 3 4 5
and that of others
11 If I begin to look at myself, I find it difficult to stop 0 1 2 3 4 5
12 I would do anything to change some parts of my body 0 1 2 3 4 5
13 I stay at home and avoid others seeing me 0 1 2 3 4 5
14 I am ashamed of the physical needs of my body 0 1 2 3 4 5
15 I feel I am laughed at because of my appearance 0 1 2 3 4 5
16 The thought of some defects of my body torments 0 1 2 3 4 5
me so much that it prevents me studying or working
17 I look in the mirror for an image of myself which satisfies
me and I continue to search until I am sure I have found it 0 1 2 3 4 5
18 I feel I am fatter than others tell me 0 1 2 3 4 5
19 I avoid mirrors 0 1 2 3 4 5
20 I have the impression that my image is always different 0 1 2 3 4 5
21 I would like to have a thin and bony body 0 1 2 3 4 5
22 I am dissatisfied with my appearance 0 1 2 3 4 5
23 My physical appearance is dissappointing compared to 0 1 2 3 4 5
my ideal image
24 I would like to undergo plastic surgery 0 1 2 3 4 5
25 I can’t stand the idea of living with the appearance I have 0 1 2 3 4 5
26 I look at myself in the mirror and have a sensation of 0 1 2 3 4 5
uneasiness and strangeness
27 I am afraid that my body will change against my will, 0 1 2 3 4 5
in a way I don’t like
28 I feel detached from my body 0 1 2 3 4 5
29 I have the sensation that my body does not belong to me 0 1 2 3 4 5
30 The thought of some defects of my body torments me 0 1 2 3 4 5
so much that it prevents me having a sexual life
31 I observe myself in what I do and ask myself how 0 1 2 3 4 5
I seem to others
32 I would like to decide what appearance to have 0 1 2 3 4 5
33 I feel different to how others see me 0 1 2 3 4 5
34 I am ashamed of my body 0 1 2 3 4 5
Continued

Eating Weight Disord., Vol. 11: N. 1 - 2006 10

 M. Cuzzolaro, G. Vetrone, G. Marano, et al.

the norms for normal-weight non eating disor- As for the comparison between female
dered adolescents and adults, males and patients affected by ED and controls, the
females. The discriminant functions concerning results of the analyses of covariance and those
differences between males and females are high- of the discriminant analyses demonstrated a
ly significant, above all in the 18-39-age range. good predictive validity for AN (restrictive and

APPENDIX 1 - (Continued)
BUT•B
Mark with an X the answer which best expresses your experience at the moment
Of my body, in particular, I hate
never seldom sometimes often very often always
1 height 1 2 3 4 5
2 the shape of my head 1 2 3 4 5
3 the shape of my face 1 2 3 4 5
4 skin 1 2 3 4 5
5 hair 1 2 3 4 5
6 forehead 1 2 3 4 5
7 eyebrows 1 2 3 4 5
8 eyes 1 2 3 4 5
9 nose 1 2 3 4 5
10 lips 1 2 3 4 5
11 mouth 1 2 3 4 5
12 teeth 1 2 3 4 5
13 ears 1 2 3 4 5
14 neck 1 2 3 4 5
15 chin 1 2 3 4 5
16 moustache 1 2 3 4 5
17 beard 1 2 3 4 5
18 hairs 1 2 3 4 5
19 shoulders 1 2 3 4 5
20 arms 1 2 3 4 5
21 hands 1 2 3 4 5
22 chest 1 2 3 4 5
23 breasts 1 2 3 4 5
24 stomach 1 2 3 4 5
25 abdomen 1 2 3 4 5
26 genitals 1 2 3 4 5
27 buttocks 1 2 3 4 5
28 hips 1 2 3 4 5
29 thighs 1 2 3 4 5
30 knees 1 2 3 4 5
31 legs 1 2 3 4 5
32 ankles 1 2 3 4 5
33 feet 1 2 3 4 5
34 odour 1 2 3 4 5
35 noises 1 2 3 4 5
36 sweat 1 2 3 4 5
37 blushing 1 2 3 4 5

11 Eating Weight Disord., Vol. 11: N. 1 - 2006

 Body Uneasiness Test

binge-purging types) and for BN purging type. interesting to note that this important psy-
We think that our data are particularly signifi- chopathological dimension is often neglected
cant because in our study we took into account in the psychometric assessment of body image
and verified the influence of age and BMI on disturbances.
body uneasiness of females normalweight and We will present many other details about the
underweight. clinical use of BUT•A and BUT•B and the sug-
Furthermore, the results of the discriminant gested cutoff points for each scale in the forth-
analyses and those of the ROC curves showed coming manual.
that the BUT subscale D (depersonalization)
has the best predictive validity for ED. It is
LIMITATIONS

APPENDIX 2
There are some limitations of this study. A
SCORING clear limitation is the absence of samples of
patients suffering from BDD, obesity and con-
BUT•A (page 1: items 1-34) genital or acquired disfiguring illnesses, out-
Global measure
comes of mutilating or deforming surgical
operations. A second limitation is the absence
GSI Global Severity Index of a sample of males with eating disorders.
The average rating of all 34 items constituting the
BUT•A (1-34)
Finally, it could be useful to compare the BUT
with a diagnostic semi-structured interview
Subscales for body image disturbances, e.g. BDD
WP Weight Phobia
Examination (27-29).
Average (9+10+18+21+24+31+32+33)
BIC Body Image Concerns Clinical implications and future directions
Average (3+4+6+12+15+22+23+25+34) We think that the BUT takes its place in the
A Avoidance range of instruments employed to study body
Average (5+8+13+16+19+30) image with good psychometric characteristics
CSM Compulsive Self-Monitoring and interesting possibilities of use as a diag-
Average (1+11+17+20+27) nostic and prognostic test (30). As a screening
D Depersonalization tool in prevention programmes it might allow
Average (2+7+14+26+28+29) us to single out subjects at risk in whom body
uneasiness and disordered eating are associat-
BUT•B (page 2: items 1-37) ed. In addition, it might be of use in the study
Global measures of the clinical course and outcome of illnesses
PST, Positive Symptom Total like AN and BN in which the improvement of
The number of symptoms rated higher than zero body image disturbance not always goes with
PSDI, Positive Symptom Distress Index the improvement of body weight. The study of
The average rating of those items constituting the negative body image associated with obesity
Positive Symptom Total and the inquiry of the effects on body image of
Subscales medical or surgical therapies aimed at reduc-
BUT•B I Mouth ing weight might be another remarkable field
Average (7+8+9+10+11+12) of application.
BUT•B II Face Shape
Average (2+3+6+13+14+15)
REFERENCES
BUT•B III Thighs
Average (24+25+28+29+30) 1. Dorian L., Garfinkel P.E.: Culture and body image in
BUT•B IV Legs Western society. Eat. Weight Disord., 7, 1-19, 2002.
Average (1+21+31+32+33) 2. Sarwer D.B., Thompson J.K., Cash T.F.: Body image
and obesity in adulthood. Psychiatr. Clin. North Am.,
BUT•B V Harms 28, 69-87, 2005.
Average (19+20+22+23+26)
3. Thompson J.K., Smolak L. (Eds.): Body image, eating
BUT•B VI Moustache disorders and obesity in youth. Assessment, preven-
Average (16+17+18) tion and treatment. Washington, DC, APA, 2001.
BUT•B VII Skin 4. Thompson J.K.: Introduction: body image, eating dis-
orders and obesity. An emerging synthesis. In:
Average (4+5)
Thompson J.K. (Ed.), Body image, eating disorders
BUT•B VIII Blushing and obesity. An integrative guide for assessment and
Average (27+34+35+36+37) treatment. Washington, DC, APA, 1996, pp. 1-20.
5. Stunkard A., Mendelson M.: Obesity and the body

Eating Weight Disord., Vol. 11: N. 1 - 2006 12

 M. Cuzzolaro, G. Vetrone, G. Marano, et al.

image. 1. Characteristics of disturbances in the body 19. Coopersmith S.: Self-Esteem Inventories. Palo Alto,
image of some obese persons. Am. J. Psychiatry, 123, CA, Consulting Psychology Press, 1981.
1296-1300, 1967. 20. Fairburn C.G., Cooper Z.: The Eating Disorder
6. Stunkard A., Burt V.: Obesity and the body image: II. Examination, 12th ed. In: Fairburn C.G., Wilson G.T.
Age of onset of disturbances in the body image. Am. J. (Eds.), Binge eating. Nature assessment and treatment.
Psychiatry, 123, 1443-1447, 1967. New York, The Guilford Press, 1993, pp. 317-360.
7. Rosen J.C.: Negative body image. In: Blechman E.A., 21. Nunnally J.: Psychometric theory. New York,
Brownell K.D. (Eds.). Behavioral medicine and women. McGraw-Hill, 1978.
A comprehensive handbook. New York, The Guilford 22. Thompson J.K., Penner L., Altabe M.: Procedures,
Press, 1998, pp. 380-385. problems and progress in the assessment of the body
8. Thompson J.K., Van den Berg P.: Measuring body images. In: Cash T., Pruzinsky T. (Eds.), Body Images -
image among adolescents and adults. In: Cash T., Assessment Development and Change. New York, The
Pruzinsky T. (Eds.), Body image: A handbook of theo- Guilford Press, 1990.
ry, research and clinical practice. New York, Guilford, 23. Marsh H., Hocevar D.: Application of confirmatory
2002, pp. 142-154. factor analysis to the study of self-concept: First- and
9. Radika L., Hayslip B.: Projective techniques to assess higher order factor models and their invariance
body image. In: Cash T., Pruzinsky T. (Eds.), Body across groups. Psychol. Bull., 97, 562-582, 1985.
image: A handbook of theory, research and clinical
24. Hallings-Pott C., Waller G., Watson D., Scragg P.:
practice. New York, Guilford, 2002, pp. 155-162.
State dissociation in bulimic eating disorders: an
10. Schilder P.: The image and appearance of the human experimental study. Int. J. Eat. Disord., 38, 37-41,
body. New York, International Universities Press, 1935. 2005.
11. Slade P.D.: Body image in anorexia nervosa. Br. J. 25. Nagata T., Kiriike N., Iketani T., Kawarada Y., Tanaka
Psychiatry, (Suppl.), 2, 20-22, 1988. H.: History of childhood sexual or physical abuse in
12. American Psychiatric Association: Diagnostic and Japanese patients with eating disorders: relationship
Statistical Manual of Mental Disorders, DSM-IV-TR. with dissociation and impulsive behaviours. Psychol.
Washington, DC, APA, 2000. Med., 29, 935-942, 1999.
13. World Health Organization: The ICD-10 Classification 26. Vandereycken W., van Houdenhove V.D.: Stealing
of Mental and Behavioural Disorders. Diagnostic behavior in eating disorders: characteristics and asso-
Criteria for Research (ICD-10: DCR-10). Geneva, ciated psychopathology. Compr. Psychiatry, 37, 316-
World Health Organization, 1993. 321, 1996.
14. Cuzzolaro M., Vetrone G., Marano G., Battacchi M.W.: 27. Cuzzolaro M., Aveni F.: Versione italiana e adattamen-
BUT: una nuova scala per la valutazione del disagio to dell'intervista: Body Dysmorphic Disorder
relativo all'immagine del corpo. Psichiatria Examination (BDDE) di J.C. Rosen e J.L. Reiter
dell'Infanzia e dell'Adolescenza, 66, 417-428, 1999. [Codice di identificazione: #767]. In: Conti L. (Ed),
15. Thompson J.K.: Assessment of body image. In: Allison Repertorio delle scale di valutazione in psichiatria, vol
D.B. (Ed.), Handbook of assessment methods for eat- 3. Firenze, SEE, 2000, pp. 1723-1753.
ing behaviors and weight-related problems. Thousand 28. Rosen J.C., Reiter J.: Development of the body dys-
Oaks, CA, Sage, 1995, pp. 119-148. morphic disorder examination. Behav. Res. Ther., 34,
16. Garner D., Garfinkel P.: The Eating Attitudes Test: an 755-766, 1996.
index of the symptoms of anorexia nervosa. Psychol. 29. Rosen J.C., Reiter J., Orosan P.: Assessment of body
Med., 9, 273-279, 1979. image in eating disorders with the body dysmorphic
17. Garner D.: EDI 2. Eating Disorder Inventory-2. disorder examination. Behav. Res. Ther., 33, 77-84,
Professional Manual. Odessa, FL, Psychological 1995.
Assessment Resources Inc., 1991. 30. Jacobi C., Abascal L., Taylor C.B.: Screening for eat-
18. Derogatis L.: Symptom Checklist-90 Manual. Baltimore, ing disorders and high-risk behavior: caution. Int. J.
Johns Hopkins University Press, 1977. Eat. Disord., 36, 280-295, 2004.

13 Eating Weight Disord., Vol. 11: N. 1 - 2006