Eating and Weight Disorders - Studies on Anorexia, Bulimia and Obesity (2023) 28:92

https://doi.org/10.1007/s40519-023-01620-w

RESEARCH

The Nepean Belief Scale (NBS) as a tool to investigate the intensity

of beliefs in anorexia nervosa: psychometric properties of the Italian
version
Arianna Sciarrillo1 · Francesco Bevione1 · Marta Lepora1 · Federica Toppino1 · Maria Carla Lacidogna1 ·
Nadia Delsedime1 · Matteo Panero1 · Matteo Martini1 · Giovanni Abbate Daga1 · Antonio Preti1

Received: 31 August 2023 / Accepted: 18 October 2023

© The Author(s) 2023

Abstract
Background People with anorexia nervosa (AN) show a peculiar impairment of insight regarding their condition, often
manifesting a denial of extreme emaciation and sometimes hiding or underreporting socially undesirable abnormal eating
patterns. Sometimes the intensity of the beliefs held by patients with AN reach a delusional intensity.
Objectives In this study, the Italian version of the Nepean Belief Scale was applied to a sample of patients diagnosed with
AN to investigate the intensity of their beliefs and convictions and its clinical correlates.
Methods The Nepean Belief Scale (NBS) was translated and adapted to Italian and applied to a sample of patients diagnosed
with AN based on the Structured Clinical Interview for DSM-5 (SCID-5).
Results The Italian version of the 5-item NBS showed excellent reliability. Convergent validity was proved by negative
association with levels of insight measured with the Schedule for the Assessment of Insight in Eating Disorders. Beliefs
of delusional intensity were reported by 10% of participants. Those with a greater intensity of beliefs, either overvalued or
delusional ideas, were more likely to report poorer general cognitive performances on the Montreal Cognitive Assessment.
No association was observed between NBS score and age, body mass index, symptoms of eating disorders, body dissatis-
faction, or levels of depression. Fear of weight gain and control seeking were the most often reported themes at the NBS.
Conclusions The Italian version of the NBS is a reasonably reliable, valid, and usable tool for the multidimensional assess-
ment of insight in AN.
Level of evidence Level III, Evidence obtained from well-designed cohort or case–control analytic studies.

Keywords Insight · Awareness of illness · Anorexia nervosa/diagnosis · Beliefs · Conviction · Psychometrics/methods ·

Body mass index

Introduction of the world, their behavior, and their decision-making pro-

cesses [3]. Beliefs can play an important role in the develop-
The concept of belief is not easy to define. At first, a belief ment and maintenance of mental disorders [4–7]. In some
can be conceived as a mental attitude or conviction that cases, certain beliefs or cognitive biases can contribute to
something is true, real, or exists, without necessarily hav- the onset of mental disorders or make them worse (e.g., [8]).
ing proof to support it [1]. Beliefs can derive from personal These beliefs can cause significant distress and interfere with
experiences, or be formed through cultural and social influ- the ability to function in daily life [9]. Moreover, people with
ences [2]. Beliefs are thought to shape a person’s perception mental disorders may hold negative beliefs about themselves
and their condition due to societal stigma and shame [10,
11]. These beliefs can contribute to feelings of isolation,
* Arianna Sciarrillo shame, and reluctance to seek treatment [12, 13].
[email protected] People diagnosed with anorexia nervosa (AN) may
1 hold irrational beliefs about their body size or weight,
Eating Disorders Unit, Department of Neuroscience “Rita
Levi Montalcini”, University of Turin, Via Cherasco 15, sometimes heavily contradictory to the evidence [14].
10126 Turin, Italy Often they may deny or minimize the severity of their

13
Vol.:(0123456789)
 92 Page 2 of 9 Eating and Weight Disorders - Studies on Anorexia, Bulimia and Obesity (2023) 28:92

symptoms, resist treatment, or be unaware of the conse- Aims

quences of their behaviors [15]. In a subgroup of patients
diagnosed with AN, the intensity of beliefs about shape This study illustrates the first Italian application of the NBS
and weight may overextend from mild overvalued ideas to to patients diagnosed with AN. The aims of this study were:
overtly delusional intensity [16]. It is unclear how much (a) to assess the reliability of the Italian version of the NBS
the deficit of insight, which is described in patients with in a sample of patients diagnosed with AN; (b) to ascertain
AN, contributes to the intensity of these irrational beliefs the links between the intensity of beliefs in patients with AN
or if it is the ego-syntonic nature of the disorder, leading and their degree of insight as measured by the SAI-ED; (c)
the patients to adhere to their convictions even when in to evaluate the relationship of the intensity of beliefs with
contrast to other people's concerns [17, 18]. Fear of stigma symptoms of eating disorder and of anxiety and depression.
and the firm willingness to maintain control over their eat-
ing behavior may support, too, the intensity of irrational
beliefs in AN [19].
Methods
Insight is a concept commonly used in psychiatry and
refers to a patient’s understanding and acceptance of their
This study was approved by the Ethical Committee of the
mental illness or disorder [20, 21]. It includes recognizing
local academic hospital (Prot. n. 0014269; code: 446/2021;
that one has a mental illness, understanding the symptoms
signed on February 8th, 2022), and all participants signed
and their impact on one’s life, and accepting the need for
written informed consent.
treatment. Insight and intensity of conviction are related but
distinct concepts. In the context of mental illness, insight
may involve recognizing that one’s thoughts or behaviors are Participants
a symptom of a mental health condition, while the intensity
of conviction may refer to how strongly a person believes All consecutive admissions among those seeking voluntary
in the content of those thoughts or behaviors. Overall, both admission for AN to a major academic Eating Disorders
insight and intensity of conviction are important factors to Center between March 2022 and June 2023 were invited to
consider in the diagnosis and treatment of a mental disorder, participate in the study.
but they should be assessed separately and in the appropri- Participants were included when they were: (a) aged
ate context. In AN, insight is an important factor and can between 18 and 50 years old; (b) had a confirmed diagno-
moderate the severity of certain cognitive and behavioral sis of AN according to the Structured Clinical Interview
components of the disorder [22]. Poor insight is associated for Diagnostic and Statistical Manual of Mental Disorders,
with later access to treatment and prolonged symptomatol- 5th edition (SCID-DSM-5); (c) had the capacity of reading
ogy [23]. It is unclear how the deficit of insight in AN is the Italian language; (d) had no current or past psychotic or
related to the intensity of their irrational beliefs about weight bipolar disorder or current substance use disorder.
and shape.
With the spreading of interest in illness awareness in psy- Measures
chiatry, several tools have been developed to measure insight
in people with mental disorders. Most of these tools have The study applied the semi-structured NBS interview to the
been developed to measure the deficit of insight in psycho- participants (see supplementary material Table S1). Stand-
sis and might be not suited to measure insight in AN [20]. ard translation and back-translation procedures were used to
Recently a semi-structured interview has been devised to translate the Italian version of the NBS from the English ver-
specifically rate insight in patients diagnosed with an eating sion [28]. Additional tools used in this study were the SAI-
disorder, the Schedule for the Assessment of Insight in Eat- ED interview, the Montreal Cognitive Assessment (MoCA),
ing Disorders-SAI-ED [22]. Evidence of its reliability and the Eating Disorder Examination Questionnaire (EDE-Q),
validity has been reported in an Italian sample of patients the Body Shape Questionnaire (BSQ), the State-Trait Anxi-
diagnosed with AN (Sciarrillo, submitted manuscript). In ety Inventory (STAI); and the Beck Depression Inventory
the past, the intensity of beliefs in AN has been measured second edition (BDI-II) [29].
with the Brown Assessment of Beliefs Scale (BABS), which The NBS is a five-item semi-structured interview that
was developed to measure anomalous conviction in people after the identification of a main belief of the patient, rates
with psychosis [24, 25]. More recently, the Nepean Beliefs five belief characteristics: conviction, fixity, fluctuation,
Scale (NBS), devised to measure the intensity of beliefs resistance, and awareness that the belief is unreasonable.
in patients diagnosed with obsessive–compulsive disorder Each item is rated on a 5-point Likert scale (from 0 to 4).
(OCD) [26, 27], has been used to rate the intensity of beliefs A total score of the intensity of the belief is calculated by
in patients with AN [16].

13
Eating and Weight Disorders - Studies on Anorexia, Bulimia and Obesity (2023) 28:92 Page 3 of 9 92

summing up all item scores, ranging from 0 to 20 (The Ital- test's reliability, a consequence which is more often observed
ian version is reproduced in Appendix 1). Internal consist- for cognitive and neuropsychological measures [35].
ency and inter-rater reliability of the NBS were found good The EDE-Q is a self-report questionnaire consisting
(> 0.70) in past studies on clinical samples [16, 27, 30]. of 28 items that assess four main symptom areas of eat-
Based on the replies on the BABS in a past study [16], the ing disorders, namely dietary restraint, eating concerns,
following thresholds were applied to the scores on the NBS: weight concerns, and shape concerns [36]. Participants are
0–10, “good/fair insight”; 11–15, “over-valued idea”; 16 or required to rate the intensity of their behavior or thoughts
higher, “delusional belief”. during the last 28 days on a scale ranging from 0 to 6,
The SAI-ED is a semi-structured interview that measures with a global score also being produced. Clinical samples
three major components of insight, such as the ability to indicate good reliability of the tool, typically > 0.80, which
recognize that one has a mental illness, the ability to rela- was confirmed in the Italian validation, too [37].
bel unusual mental events as pathological, and compliance The BSQ is a 34-item self-report questionnaire spe-
with treatment [22]. The SAI-ED includes two questions on cifically aimed at assessing body dissatisfaction prompted
awareness of the illness and the need for treatment, rated on by the feelings of being fat [38]. Participants rate their
a 0–2 scale; four questions on the recognition of symptoms responses on a six‐point scale, with a time interval focused
of eating disorder and their relabeling as attributable to the on the past 4 weeks, where higher scores indicate greater
illness, plus a question on the hypothetical contradiction dissatisfaction with body shape. The validity and reliabil-
between own’s view of the condition and concerns of other ity of the BSQ-34 have been proved [39], with reliability
people, all rated on a 0–4 scale. There is also a final ques- judged excellent (around 0.90), as confirmed in the Italian
tion on treatment adherence, aimed at collecting therapist validation [40].
impressions, rated on a 0–5 scale. The SAI-ED produces two The STAI is a self-report questionnaire comprising 40
global scores: a subtotal on awareness of illness and symp- items that measure anxiety levels [41]. For the purposes
toms, the sum of the first seven items (ranging from 0 to 24); of this study, the version Y of the STAI has been used
and a total score, measuring the global level of insight, the [42]. Participants rate their levels of anxiety on a scale
sum of all items (ranging from 0 to 29). The reliability of the from 1 (never) to 4 (always) for each item, based on their
SAI-ED was reported to be high [22]. In the initial valida- current state anxiety (20 items) or habitual trait anxiety
tion study, the lack of insight was likely for a global score (20 items). The reliability of the STAI varies from 0.65 to
on the SAI-ED ≤ 18 [22]. In the Italian validation study, the 0.90, depending on the sample [43]. The Italian validation
SAI-ED had moderate reliability (intraclass correlation coef- confirmed the psychometric properties of the STAI.
ficient [ICC] = 0.71; 95%CI 0.49–0. 86) (Sciarrillo, submit- The BDI-II is a self-report questionnaire consisting of
ted manuscript). 21 items that assess the presence and severity of depres-
The MoCA is a widely used cognitive screening tool that sive symptoms [29]. Participants rate their behavior or
is designed to assess various cognitive domains, including thoughts on a scale of 0 to 3, with higher scores indicat-
attention and concentration, executive functions, memory, ing greater severity of depression symptoms. Scores above
language, visuoconstructional skills, conceptual thinking, 16 are indicative of severe depressive symptomatology.
calculations, and orientation. The MoCA consists of a brief The tool has a high internal consistency, with a reliability
questionnaire and a series of tasks that assess different cog- of around 0.90 across samples [44]. The reliability and
nitive ability [31]. It takes approximately 10–15 min to com- validity of the BDI-II have been confirmed in the Italian
plete. The MoCA is commonly used to screen for cognitive version [45].
impairment and it is also used in research studies to assess
cognitive function and to track changes in cognitive ability
over time. The MoCA has acceptable internal consistency Procedure
(Cronbach’s α > 0.70) [31, 32]. Acceptable inter-rater and
intra-rater reliability were reported for the Italian validation, Participants were individually interviewed with the NBS,
too, when used to detect cognitive impairment [33]. How- SAI-ED, and the MOCA in a quiet room, with the inter-
ever, the MoCA was developed to be sensitive to mild cogni- views typically lasting an average of 15 min for each task.
tive impairment in geriatric populations: in healthy people, Furthermore, participants were also requested to complete
inter-rater reliability, when measured with the kappa coef- additional questionnaires through the REDCap platform
ficient, varied from 0.46 to 0.94 [34]. This is likely to depend (Research Electronic Data Capture, Vanderbilt University),
on the ceiling effect, which occurs when a large percentage using a tablet or smartphone. The time taken to complete
of participants achieve the highest score on a test. The ceil- these questionnaires was not measured. Prior to the assess-
ing effect limits the ability of a test of discriminating sub- ment, participants were required to provide informed written
jects at the top level of the performance, thus decreasing the consent, and anonymity was ensured.

13
 92 Page 4 of 9 Eating and Weight Disorders - Studies on Anorexia, Bulimia and Obesity (2023) 28:92

Data analysis Italian language, and 6 were outside the age cutoff (n. = 2
were < 18 years old, and n. = 4 were > 50 years old).
Analysis was done with the Statistical Package for the Social The final sample included 38 patients diagnosed with AN,
Sciences (SPSS) version 28 (IBM Corp. Released 2021) of whom, 28 with AN restricting type (AN-R), and 10 with
and dedicated packages running in R [46]. The significance AN binge-eating/purging type (AN-BP).
threshold was set at p < 0.05. All analyses were two-tailed. Participants included one male (3%) and 37 females
Continuous variables were reported as means with stand- (97%), of whom six did not adhere to a binary classifica-
ard deviations and range. Skewness and kurtosis were also tion of the genre. Age in the sample ranged from 18 to
reported, with values ≥ 2 indicating deviation from the nor- 49 years, with mean = 24 ± 9. BMI ranged from 11 to 17,
mal distribution according to a conservative threshold for with mean = 14 ± 2. Duration of illness ranged from 0.5 year
sample size ≤ 50 [47]. to 35 years, with mean = 7 ± 9.
Categorical variables were reported as counts and per-
centages. Analysis of continuous variables was performed
with Student’s, or Welch’s t-test in case of violation of the Reliability
homogeneity of the variance, or with analysis of variance
(ANOVA). Categorical variables were analyzed with Chi- Inter-rater reliability was fair for MoCA (ICC = 0.53; 95%CI
square, with Yates correction when necessary. Pearson cor- 0.32 to 0.73), and moderate to excellent for SAI-ED (0.74;
relation coefficient was used to measure the association 0.56–0.86), and NBS (0.80; 0.67–0.89). Cronbach’s alpha
between continuous variables. Effect size in between-group for questionnaire varied from good to excellent for EDE-Q
comparisons and correlations was assessed with reference (Cronbach’s alpha = 0.93), BSQ (0.97), STAI (0.95), and
to the thresholds suggested by Cohen [48]. BDI-II (0.95).
The reliability of the self-report questionnaires was meas-
ured with Cronbach’s alpha to facilitate comparisons with
previous studies. According to a shared rule, Cronbach’s Scores on the measures used in the study
alpha is assumed to be fair when it is equal to or greater
than 0.70 in a questionnaire measuring one latent dimen- Table 1 lists the scores for each measure used in the study.
sion, good when is greater than 0.80, excellent when exceeds No deviation from the normal distribution was observed
0.90 [49]. based on skewness and kurtosis. As for the MoCA, a ceil-
Reliability of the SAI-ED, NBS, and MOCA was assessed ing effect was found: 9 participants (24%) scored 30, the
as inter-raters agreement with ICC with 95% confidence maximum value on the test, and an additional 8 (21%) scored
interval (CI). The calculation was with a two-way random- 29. Overall, 45% of the sample scored at maximum or near
effects model. According to a shared rule of thumb, ICC maximum value on the MOCA. The high ceiling effect justi-
values between 0.40 and 0.59 indicate fair reliability; 0.60 fies the observation of a relatively low reliability of the test
to 0.74 indicate moderate reliability; values 0.75 and above in the sample.
indicate excellent reliability [50].
This is a pilot validation study, and sample size require-
ments were kept to a minimum. For reliability purposes, a Table 1  Mean scores of the measures used in the study in the sample
sample size of 25/30 participants can be enough to maintain (n. = 38)
a balance between the power and the precision of reliability Mean (SD) Range Skew Kurtosis
estimates [51]. Validity was estimated by correlational anal-
EDE-Q
ysis. For a two-tailed alpha at 0.05, power at 80%, and for a
Restraint 3.7 (1.8) 0–6 − 0.5 − 0.9
minimum clinically meaningful effect size of 0.45 (= 20% of
Eating concern 3.1 (1.5) 0.2–5.4 − 0.4 − 0.9
shared variance), the required sample size was 33. The cal-
Shape concern 4.3 (1.6) 0.9–6 − 0.6 − 1.00
culation was done with G*Power 3.1 [52]. The final sample
Weight concern 3.5 (1.9) 0.2–6.0 − 0.3 − 1.3
size in the study was above these minimum requirements.
BSQ 122 (43) 46–187 − 0.2 − 1.2
STAI
State anxiety 57 (13) 31–77 − 0.6 − 0.2
Trait anxiety 60 (11) 34–74 − 0.8 − 0.4
Results
BDI-II 25.3 (15.9) 3–56 0.5 − 0.9
MoCA 27.2 (2.8) 19–30 − 1.1 0.6
Overall, 60 patients were invited to take part in the study.
SAI-ED 21.5 (4.5) 12–29 − 0.1 − 0.9
Among them, 11 refused, 4 were too ill to be involved in the
NBS 9.9 (4.9) 2–18 − 0.2 − 1.2
study, 1 was discarded because had difficulties in reading the

13
Eating and Weight Disorders - Studies on Anorexia, Bulimia and Obesity (2023) 28:92 Page 5 of 9 92

Intensity of belief at the NBS Table 3  Correlations between socio-demographics and clinical fea-
tures and the Nepean Beliefs Scale total scores in patients diagnosed
with anorexia nervosa (n. = 38)
The mean beliefs score in the sample for the NBS was
9.9 ± 4.9; range: 2 to 18. Item-total correlation was good Nepean Beliefs Scale
(> 0.70), except for item 4 (Resistance). Maximum intensity Age (years) r = − 0.09; p = 0.59
of the beliefs (score = 4) was rated by less than one-fifth of Body mass index (kg/m2) r = − 0.09; p = 0.59
the sample for each item (Table 2). EDE-Q
According to the predefined thresholds, 19 (50%) par- Restraint r = 0.05; p = 0.76
ticipants showed good/fair insight, 15 (39.5%) overvalued Eating concern r = 0.18; p = 0.28
ideas and 4 (10.5%) had a delusional intensity of beliefs. Shape concern r = 0.10; p = 0.55
Participants with poor insight at the SAI-ED were 11 (29%). Weight concern r = 0.19; p = 0.25
There was no statistically significant association between BSQ r = 0.09; p = 0.59
these three degrees of belief and categorically rated poor STAI
insight: χ2 = 3.35; df = 2; p = 0.19. However, several cells State anxiety r = 0.42; p < 0.01
have less than 5 counts, thus the result must be taken with Trait anxiety r = 0.27; p = 0.10
caution. BDI-II r = 0.18; p = 0.28
MoCA r = − 0.57; p < 0.01
Association of clinical variables with the intensity SAI-ED
of beliefs on the NBS Awareness of illness and symptoms r = − 0.46; p < 0.01
Adherence to treatment r = − 0.39; p = 0.02
The intensity of belief, as rated continuously on the NBS, Level of insight r = − 0.54; p < 0.01
was unrelated to age, BMI, symptoms of eating disorders as
rated on the EDE-Q, body dissatisfaction rated on the BSQ,
levels of depression on the BDI-II (Table 3).
A positive association was observed between state but and other related (13%) or multiple/unclassified categories
not trait anxiety and total scores on the NBS. The inten- (11%).
sity of beliefs, rated on the NBS, was negatively related to
general cognition, as measured with MoCA, and treatment
adherence and insight, as measured by the SAI-ED. Put in Discussion
another way, those with a greater intensity of beliefs, either
overvalued or delusional ideas, were more likely to report The results of this pilot study indicate that the Italian version
poorer performances on the MoCA, adhere less to treatment, of the NBS shows moderate-to-excellent intra-rater reliabil-
and be rated with a lower insight. ity and possesses good concurrent validity, as revealed by
the inverse correlation between the intensity of the belief
Emerging themes during the NBS interview and levels of insight as measured on the SAI-ED. Moreover,
NBS scores were negatively related to treatment adherence,
The self-reported beliefs of the participants that emerged confirming that beliefs of particular intensity, assimilable to
during the interview with the NBS were analyzed to identify overvalued or delusional ideas, impair treatment adherence
any recurring thematic connections. Various themes emerged and might negatively affect therapy outcomes.
from the sample, listed in Table 4, including concerns about As already noted, a key advantage of the NBS is the offer-
weight gain (39%), efforts to exert control (21%), percep- ing of an operational definition of the belief and its charac-
tions of self-worth (5%), critical self-evaluation (11%), teristics, and this facilitates the accurate recording of beliefs

Table 2  Mean scores of the Mean SD Median Range Rated 4 Correlation

Nepean Beliefs Scale scores with total
score

Conviction 2.6 1.1 3 0–4 8 (21%) 0.88

Fixity 2.1 1.3 2 0–4 6 (16%) 0.69
Fluctuation 2.2 1.2 2 0–4 6 (16%) 0.79
Resistance 1.7 1.2 1 0–4 2 (5%) 0.51
Reasonableness of 1.5 1.5 1 0–4 2 (5%) 0.83
belief

13
 92 Page 6 of 9 Eating and Weight Disorders - Studies on Anorexia, Bulimia and Obesity (2023) 28:92

Table 4  Emerging themes observed in participants' beliefs

Theme Example n. (%)

Fear of weight gain “Adding the legumes to my diet will surely make my weight increase by 1 kg.” 15 (39%)
Control seeking “If I keep under control my diet, then I will consequently take control of all aspects of my 8 (21%)
life.”
Self-worth “I’m convinced I’m not thin enough despite my weight.” 2 (5%)
Negative evaluation “My legs are fat and ugly despite what everybody tells me.” 4 (11%)
Other “As more I move, as more my emotions come out and I feel them.” 5 (13%)
Multiple/unclassified “Some foods are damaging to my body, even if nobody told me they are [ex. bread, pasta, 4 (11%)
floury, fruit].”

in patient populations [16, 30]. In this study, the mean score might influence the finding of delusional intensities on the
for conviction was relatively high—2.6 (see Table 2), and NBS. However, the negative association of NBS scores with
one-fifth circa of patients rated their belief as probably a measure of general cognition might depend on the negative
true. Fixity of the belief was on average high, too (mean impact of malnutrition on cognition, albeit we did not find
score = 2.1; median = 2), and one-fifth of patients endorsed an association of BMI with NBS scores.
their belief as true even against contrary evidence. This was As in past studies [16, 53], we did not find a relationship
confirmed by relatively high scores on fluctuation, with 16% between body dissatisfaction and intensity of beliefs in our
of the patients holding their belief as true over time. A lower sample. Instead, we found a link between trait anxiety and
fraction of patients expressed resistance to the belief (5%) the intensity of beliefs on the NBS. This might suggest that
or communicate the awareness that their belief was unrea- those with overvalued or delusional ideas experience greater
sonable (5%). Overall, results were closer to those found in anxiety because of their beliefs, thus making them less ready
patients with OCD [27] than in patients with psychosis [30]. to have their highly regarded beliefs challenged during treat-
The percentage of patients with beliefs of delusional ment. On a therapeutic ground, this is important since the
intensity (10%) was similar to the one found in past studies primary treatment of delusional ideas is with antipsychot-
with different tools, e.g., 16% in Hartmann et al. [53], 10% ics, and the efficacy of antipsychotics such as olanzapine or
in Mountjoy et al. [54], but lower to the percentage observed aripiprazole in the reduction of delusional beliefs in AN has
in a past study with the NBS (23.7%) [16] or when measured yet to be established [66, 67].
with BABS (e.g., 20% in [25]; 22.7% in [22]). Discrepan-
cies might depend on some unmeasured characteristics of Strengths and limitations of the study
the samples, such as psychiatric or somatic comorbidity,
which was rarely or never reported in past studies. Indeed, The use of two interviews to investigate beliefs and insight in
in research, psychiatric comorbidity should be better inves- the sample is the most significant strength of the study since
tigated as clinicians often tend to underdiagnose the present self-report tools are more exposed than interviews to misun-
clinical pictures that are different from ED [55]. derstanding, hiding, and social desirability replies. It must
In this study, MoCA scores were negatively related to be considered that while the SAI-ED has been developed
scores on the NBS, thus indicating that as poorer the cogni- specifically for patients with eating disorders, the NBS has
tive performances on the MoCA, as greater the chances of been devised for use in patients with obsessive–compulsive
finding delusional intensities on the NBS. In patients with disorder. Nevertheless, its application in this and a past study
psychosis, the NBS showed good convergent validity with [16] indicates that the tool can be applied to patients with
the BABS [30]. Patients with schizophrenia-spectrum psy- AN. It is worth noting that the Italian validation of the SAI-
chosis are known to display cognitive impairment on the ED has not yet been published and that the Italian NBS has
MoCA [56]. Comorbidity of AN with psychosis has long not been validated so far in people with obsessive–compul-
been debated [57–59], but some features such as rigidity sive disorder (OCD), who were the original clinical target of
of thought [60, 61], abnormal bodily perception [62], and the tool. Thus, evidence on the reliability and validity of the
hallucinations-like experiences such as the “anorexic voice” Italian NBS in patients with AN should not be extrapolated
[63], might be assimilated to psychotic-like traits. Moreover, to Italian patients with OCD before appropriate testing.
the phenotypic relationships between eating disorders and Some limitations have to be taken into account, too. The
psychosis seem supported by shared genetic vulnerability sample size was not large enough to allow more complex
[64], as confirmed by their familial co-aggregation [65]. The statistical analyses, e.g., a comparison between AN-R and
fraction of patients with AN also displaying psychotic traits AN-BP participants. The small sample size might also have

13
Eating and Weight Disorders - Studies on Anorexia, Bulimia and Obesity (2023) 28:92 Page 7 of 9 92

prevented the detection of a statistically significant asso- Author contributions All authors contributed to the study’s conception
ciation between the three degrees of belief drawn from the and design. Material preparation and data analysis were performed by
AS, FB, MM, and AP. GAD oversaw and had responsibility for the
NBS and categorically rated poor insight on the SAI-ED. We research activity planning and execution. AS, FB, MCL, MM, FT, ML,
could not address the role of psychiatric or somatic comor- ND contributed to the organization of the study and the acquisition of
bidity in the sample since information was limited. Partici- the data. AS, FB, MM and MP had responsibility for data process-
pants included help-seeking individuals requiring inpatient ing and management. AP wrote the first draft of the manuscript and
all authors commented on the subsequent versions of the manuscript.
treatment, which implies some insight among them. Moreo- GAD contributed to the interpretation and discussion of the findings
ver, the most severe patients were not recruited for ethical and revised the initial draft into a more general framework. All authors
reasons. The sample cannot be considered completely repre- read and approved the final manuscript.
sentative of the population from which it was drawn. All but
Funding Open access funding provided by Università degli Studi
one patient were women. Therefore it cannot be assured that di Torino within the CRUI-CARE Agreement. Open access funding
the findings will hold in male patients with AN. provided by Università degli Studi di Torino within the CRUI-CARE
Agreement. This research was performed as a part of the institutional
activity of the units, with no specific funding. All expenses, including
What is already known on this subject? salaries of the investigators, were covered by public research funds
assigned to the units.
Insight and intensity of beliefs about weight and body con- Data availability The corresponding author had full access to all the
cerns influence adherence to treatment and outcome in data in the study and takes responsibility for the integrity of the data
patients with AN [15, 19, 23]. The NBS has been devised to and the accuracy of data analysis. Data sharing is not applicable since
measure intensity of beliefs in patients with obsessive–com- consent to publish was for aggregated data only.
pulsive disorder as a proxy of insight [27]. NBS was proved Code availability Standard codes were used in the analysis. Sample
to be applied to patients with psychosis, too, a popula- codes will be shared upon demand.
tion with difficulties in achieving awareness of illness and
insight about it [30]. Overall, the NBS is an agile tool that Declarations
can be applied in the short term with adequate reliability,
Ethics approval and consent to participate This study was performed
a clear advantage in patients that may be challenging to be in line with the principles of the Declaration of Helsinki. Local ethics
evaluated. committee approved the study.

Consent to publish Participants provided consent to publish the

What does this study add? acquired data only in anonymized and aggregated format.

Informed consent Written informed consent was provided by partici-

The Italian version of the NBS showed good psychometric
pants.
properties in terms of reliability and concurrent and diver-
gent validity. The findings of this study confirm and extend Competing interests The authors have no relevant financial or non-
the results of a past application of the NBS to a sample of financial interests to disclose.
patients with AN [16]. In this population, the NBS allows
Open Access This article is licensed under a Creative Commons Attri-
the identification of a subgroup of patients with beliefs of bution 4.0 International License, which permits use, sharing, adapta-
delusional intensity. tion, distribution and reproduction in any medium or format, as long
as you give appropriate credit to the original author(s) and the source,
provide a link to the Creative Commons licence, and indicate if changes
were made. The images or other third party material in this article are
Conclusions included in the article’s Creative Commons licence, unless indicated
otherwise in a credit line to the material. If material is not included in
the article’s Creative Commons licence and your intended use is not
While it is still unknown the prognosis of those patients permitted by statutory regulation or exceeds the permitted use, you will
with AN that nurture delusional beliefs, their identification need to obtain permission directly from the copyright holder. To view a
is helpful for therapeutic purposes. Future studies should copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
test whether there are differences across subtypes of AN in
terms of the intensity of beliefs, and evaluate which relevant
factors influence the occurrence of beliefs with delusional
intensity. The sensitivity of delusional beliefs in patients
with AN to treatment with antipsychotics should be explored
with randomized-controlled trials.
Supplementary Information The online version contains supplemen-
tary material available at https://d​ oi.o​ rg/1​ 0.1​ 007/s​ 40519-0​ 23-0​ 1620-w.

13
 92 Page 8 of 9 Eating and Weight Disorders - Studies on Anorexia, Bulimia and Obesity (2023) 28:92

References 19. Konstantakopoulos G, Tchanturia K, Surguladze SA, David AS

(2011) Insight in eating disorders: clinical and cognitive corre-
lates. Psychol Med 41:1951–1961. https://​doi.​org/​10.​1017/​S0033​
1. Connors MH, Halligan PW (2015) A cognitive account of belief:
29171​00025​39
a tentative road map. Front Psychol. https://d​ oi.o​ rg/1​ 0.3​ 389/f​ psyg.​
20. Konstantakopoulos G (2019) Insight across mental disorders: A
2014.​01588
multifaceted metacognitive phenomenon. Psychiatriki 30:13–16.
2. Usó-Doménech JL, Nescolarde-Selva J (2016) What are belief
https://​doi.​org/​10.​22365/​jpsych.​2019.​301.​13
systems? Found Sci 21:147–152. https:// ​ d oi. ​ o rg/ ​ 1 0. ​ 1 007/​
21. David AS (2020) Insight and psychosis: the next 30 years. Br J
s10699-​015-​9409-z
Psychiatry 217(3):521–523. https://​doi.​org/​10.​1192/​bjp.​2019.​217
3. Seitz RJ, Angel H-F (2020) Belief formation—a driving force
22. Konstantakopoulos G, Georgantopoulos G, Gonidakis F, et al The
for brain evolution. Brain Cogn 140:105548. https://​doi.​org/​10.​
development of the Schedule for the Assessment of Insight in Eat-
1016/j.​bandc.​2020.​105548
ing Disorders (SAI-ED): Dimensions and associations of insight
4. Ehlers A, Clark DM (2000) A cognitive model of posttraumatic
in anorexia nervosa and bulimia nervosa
stress disorder. Behav Res Ther 38:319–345. https://​doi.​org/​10.​
23. Arbel R, Latzer Y, Koren D (2014) Revisiting poor insight into
1016/​S0005-​7967(99)​00123-0
illness in anorexia nervosa: true unawareness or conscious disa-
5. Kube T, Schwarting R, Rozenkrantz L et al (2020) Distorted
greement? J Psychiatr Pract 20:85–93. https://d​ oi.o​ rg/1​ 0.1​ 097/0​ 1.​
cognitive processes in major depression: a predictive process-
pra.​00004​45243.​00683.​30
ing perspective. Biol Psychiatry 87:388–398. https://​doi.​org/​10.​
24. Eisen AL, Eisen JL, Phillips KA, et al (1998) The Brown Assess-
1016/j.​biops​ych.​2019.​07.​017
ment of Beliefs Scale: Reliability and Validity
6. Pillny M, Schlier B, Lincoln TM (2020) “I just don’t look for-
25. Steinglass JE, Eisen JL, Attia E et al (2007) Is anorexia nervosa a
ward to anything”. How anticipatory pleasure and negative
delusional disorder? An assessment of eating beliefs in anorexia
beliefs contribute to goal-directed activity in patients with
nervosa. J Psychiatr Pract 13:65–71. https://​doi.​org/​10.​1097/​01.​
negative symptoms of psychosis. Schizophr Res 222:429–436.
pra.​00002​65762.​79753.​88
https://​doi.​org/​10.​1016/j.​schres.​2020.​03.​059
26. Brakoulias V, Starcevic V, Berle D et al (2013) Further support
7. Kube T, Rozenkrantz L (2021) When beliefs face reality: an
for five dimensions of obsessive-compulsive symptoms. J Nervous
integrative review of belief updating in mental health and ill-
Mental Dis 201:452–459. https://​doi.​org/​10.​1097/​NMD.​0b013​
ness. Perspect Psychol Sci 16:247–274. https://​doi.​org/​10.​1177/​
e3182​94804e
17456​91620​931496
27. Brakoulias V, Starcevic V, Milicevic D et al (2018) The Nepean
8. Ingram R, Miranda J, Segal Z (1998) Cognitive vulnerability to
Belief Scale: preliminary reliability and validity in obsessive–
depression. Guilford Press, New York
compulsive disorder. Int J Psychiatry Clin Pract 22:84–88. https://​
9. Vîslǎ A, Flückiger C, Grosse Holtforth M, David D (2016)
doi.​org/​10.​1080/​13651​501.​2017.​13744​13
Irrational beliefs and psychological distress: a meta-analysis.
28. Beaton DE, Bombardier C, Guillemin F, Ferraz MB (2000) Guide-
Psychother Psychosom 85:8–15. https://​doi.​org/​10.​1159/​00044​
lines for the process of cross-cultural adaptation of self-report
1231
measures. Spine 25(24):3186–3191. https://​doi.​org/​10.​1097/​
10. Corrigan P (2004) How stigma interferes with mental health care.
00007​632-​20001​2150-​00014
Am Psychol 59:614–625. https://​doi.​org/​10.​1037/​0003-​066X.​
29. Beck A, Steer R, Brow G (1996) BDI-II: beck depression inven-
59.7.​614
tory manual, 2nd edn. Psychological Corporation, San Antonio
11. Eliasson ET, McNamee L, Swanson L et al (2021) Unpacking
30. Rajendran P, Van Winssen C, Viswasam K et al (2022) The
stigma: meta-analyses of correlates and moderators of personal
psychometric properties of the Nepean belief scale as a tool for
stigma in psychosis. Clin Psychol Rev 89:102077. https://​doi.​org/​
assessing delusions in schizophrenia and related psychotic dis-
10.​1016/j.​cpr.​2021.​102077
orders. Compr Psychiatry. https://​doi.​org/​10.​1016/j.​compp​sych.​
12. Jorm AF (2000) Mental health literacy: public knowledge and
2022.​152337
beliefs about mental disorders. Br J Psychiatry 177:396–401
31. Nasreddine ZS, Phillips NA, Bédirian V et al (2005) The Mon-
13. Angermeyer MC, van der Auwera S, Carta MG, Schomerus G
treal cognitive assessment, MoCA: a brief screening tool for mild
(2017) Public attitudes towards psychiatry and psychiatric treat-
cognitive impairment. J Am Geriatr Soc 53(4):695–699. https://​
ment at the beginning of the 21st century: a systematic review and
doi.​org/​10.​1111/j.​1532-​5415.​2005.​53221.x
meta-analysis of population surveys. World Psychiatry 16:50–61.
32. Pereiro AX, Ramos-Lema S, Lojo-Seoane C et al (2017) Nor-
https://​doi.​org/​10.​1002/​wps.​20383
mative data for the Montreal cognitive assessment (MOCA) in a
14. Gadsby S (2017) Explaining body size beliefs in anorexia. Cogn
Spanish sample of community-dweller adults. Eur Geriatr Med
Neuropsychiatry 22:495–507. https://​doi.​org/​10.​1080/​13546​805.​
8:240–244. https://​doi.​org/​10.​1016/j.​eurger.​2017.​04.​003
2017.​14015​31
33. Pirrotta F, Timpano F, Bonanno L et al (2015) Italian validation of
15. Kostecka B, Kordynska KK, Murawiec S, Kucharska K (2019)
Montreal cognitive assessment. Eur J Psychol Assess 31:131–137.
Distorted body image in women and men suffering from Ano-
https://​doi.​org/​10.​1027/​1015-​5759/​a0002​17
rexia Nervosa—a literature review. Archiv Psychiatry Psychother
34. Bruijnen CJWH, Dijkstra BAG, Walvoort SJW et al (2020)
21:13–21
Psychometric properties of the montreal cognitive assessment
16. Barton R, Aouad P, Hay P et al (2022) Distinguishing delu-
(MoCA) in healthy participants aged 18–70. Int J Psychiatry Clin
sional beliefs from overvalued ideas in Anorexia Nervosa: an
Pract 24:293–300. https://​doi.​org/​10.​1080/​13651​501.​2020.​17463​
exploratory pilot study. J Eat Disord. https://​doi.​org/​10.​1186/​
48
s40337-​022-​00600-2
35. Gaudet CE, Konin J, Faust D (2021) Immediate post-concussion
17. Vitousek K, Watson S, Wilson GT (1998) Enhancing motivation
and cognitive testing: ceiling effects, reliability, and implications
for change in treatment-resistant eating disorders. Clin Psychol
for interpretation. Arch Clin Neuropsychol 36:561–569. https://​
Rev 18(4):391–420. https://​doi.​org/​10.​1016/​s0272-​7358(98)​
doi.​org/​10.​1093/​arclin/​acaa0​74
00012-9
36. Fairburn CG, Beglin SJ (1994) Assessment of eating disor-
18. Konstantakopoulos G, Varsou E, Dikeos D et al (2012) Delusion-
ders: interview or self-report questionnaire? Int J Eat Disord
ality of body image beliefs in eating disorders. Psychiatry Res
16(4):363–370
200:482–488. https://​doi.​org/​10.​1016/j.​psych​res.​2012.​03.​023

13
Eating and Weight Disorders - Studies on Anorexia, Bulimia and Obesity (2023) 28:92 Page 9 of 9 92

37. Calugi S, Milanese C, Sartirana M et al (2017) The Eating Dis- ideas in anorexia nervosa versus body dysmorphic disorder. Psy-
order Examination Questionnaire: reliability and validity of the chiatry Res 210:1129–1135. https://​doi.​org/​10.​1016/j.​psych​res.​
Italian version. Eat Weight Disord 22:509–514. https://d​ oi.o​ rg/1​ 0.​ 2013.​08.​010
1007/​s40519-​016-​0276-6 54. Mountjoy RL, Farhall JF, Rossell SL (2014) A phenomenological
38. Cooper PJ, Taylor MJ, Cooper Z, Fairburn CG (1987) The devel- investigation of overvalued ideas and delusions in clinical and
opment and validation of the body shape questionnaire. Int J Eat subclinical anorexia nervosa. Psychiatry Res 220:507–512. https://​
Disord 6:485–494. https://​doi.​org/​10.​1002/​1098-​108X(198707)​ doi.​org/​10.​1016/j.​psych​res.​2014.​07.​073
6:4%​3C485::​AID-​EAT22​60060​405%​3E3.0.​CO;2-O 55. Longo P, Toppino F, Martini M et al (2022) Diagnostic concord-
39. Rosen JC, Jones A, Ramirez E, Waxman S (1996) Body shape ance between research and clinical-based assessments of psychiat-
questionnaire: Studies of validity and reliability. Int J Eat Disord ric comorbidity in Anorexia Nervosa. J Clin Med 11:7419. https://​
20:315–319. https://​doi.​org/​10.​1002/​(SICI)​1098-​108X(199611)​ doi.​org/​10.​3390/​jcm11​247419
20:3%​3c315::​AID-​EAT11%​3e3.0.​CO;2-Z 56. Rosca EC, Cornea A, Simu M (2020) Montreal cognitive
40. Marzola E, Martini M, Longo P et al (2022) Psychometric assessment for evaluating the cognitive impairment in patients
properties of the Italian body shape questionnaire: an investiga- with schizophrenia: a systematic review. Gen Hosp Psychiatry
tion of its reliability, factorial, concurrent, and criterion valid- 65:64–73
ity. Eat Weight Disord 27:3637–3648. https://​doi.​org/​10.​1007/​ 57. Miotto P, Pollini B, Restaneo A et al (2010) Symptoms of psycho-
s40519-​022-​01503-6 sis in anorexia and bulimia nervosa. Psychiatry Res 175:237–243.
41. Spielberger C (1983) Manual for the state-trait anxiety inventory. https://​doi.​org/​10.​1016/j.​psych​res.​2009.​03.​011
Consulting Psychologists Press, Palo Alto 58. Bou Khalil R, Hachem D, Richa S (2011) Eating disorders and
42. Spielberger CD (1983) Manual for the State-Trait Inventory STAI schizophrenia in male patients: a review. Eat Weight Disord
(Form Y). Mind Garden, Palo Alto, CA 16(3):e150–e156. https://​doi.​org/​10.​1007/​BF033​25126
43. Barnes LLB, Harp D, Jung WS (2002) Reliability generalization 59. Seeman MV (2014) Eating disorders and psychosis: Seven
of scores on the spielberger state-trait anxiety inventory. Educ hypotheses. World J Psychiatry 4:112. https://​doi.​org/​10.​5498/​
Psychol Meas 62(4):603–618. https://​doi.​org/​10.​1177/​00131​ wjp.​v4.​i4.​112
64402​06200​4005 60. Abbate-Daga G, Amianto F, Delsedime N et al (2013) Resist-
44. Wang YP, Gorenstein C (2013) Psychometric properties of the ance to treatment in eating disorders: a critical challenge. BMC
beck depression inventory-II: a comprehensive review. Rev Bras Psychiatry. https://​doi.​org/​10.​1186/​1471-​244X-​13-​294
Psiquiatr 35:416–431 61. King JA, Korb FM, Vettermann R et al (2019) Cognitive over-
45. Sica C, Ghisi M (2007) The Italian versions of the beck anxi- control as a trait marker in anorexia nervosa? Aberrant task- and
ety inventory and the beck depression inventory-II: psychometric response-set switching in remitted patients. J Abnorm Psychol
properties and discriminant power. Leading-edge psychological 128:806–812. https://​doi.​org/​10.​1037/​abn00​00476
tests and testing research. Nova Science Publishers, Hauppauge, 62. Gaudio S, Brooks SJ, Riva G (2014) Nonvisual multisensory
pp 27–50 impairment of body perception in anorexia nervosa: a systematic
46. R Core Team (2022) R: A language and environment for statistical review of neuropsychological studies. PLoS ONE 9(10):e110087.
computing. In: URL https://​www.R-​proje​ct.​org/. https://​doi.​org/​10.​1371/​journ​al.​pone.​01100​87
47. Kim H-Y (2013) Statistical notes for clinical researchers: assessing 63. Aya V, Ulusoy K, Cardi V (2019) A systematic review of the ‘eat-
normal distribution (2) using skewness and kurtosis. Restor Dent ing disorder voice’ experience. Int Rev Psychiatry 31:347–366
Endod 38:52. https://​doi.​org/​10.​5395/​rde.​2013.​38.1.​52 64. Watson HJ, Palmos AB, Hunjan A et al (2021) Genetics of eating
48. Cohen J (1992) A power primer. Psychol Bull 112:155–159. disorders in the genome-wide era. Psychol Med 51:2287–2297
https://​doi.​org/​10.​1037/​0033-​2909.​112.1.​155 65. Zhang R, Larsen JT, Kuja-Halkola R et al (2021) Familial co-
49. Nunnally J, Bernstein I (1994) Psychometric theory, 3rd edn. aggregation of schizophrenia and eating disorders in Sweden and
McGraw-Hill, New York Denmark. Mol Psychiatry 26:5389–5397. https://d​ oi.o​ rg/1​ 0.1​ 038/​
50. Cicchetti DV (1994) Guidelines, criteria, and rules of thumb for s41380-​020-​0749-x
evaluating normed and standardized assessment instruments in 66. Marzola E, Desedime N, Giovannone C et al (2015) Atypical
psychology. Psychol Assess 6(4):284–290. https://​doi.​org/​10.​ antipsychotics as augmentation therapy in anorexia nervosa. PLoS
1037/​1040-​3590.6.​4.​284 ONE 10:e0125569. https://d​ oi.o​ rg/1​ 0.1​ 371/j​ ourna​ l.p​ one.0​ 12556​ 9
51. Cocchetti DV (1999) Sample size requirements for increasing the 67. Cassioli E, Sensi C, Mannucci E et al (2020) Pharmacological
precision of reliability estimates: problems and proposed solu- treatment of acute-phase anorexia nervosa: evidence from rand-
tions. J Clin Exp Neuropsychol 21:567–570. https://​doi.​org/​10.​ omized controlled trials. J Psychopharmacol 34:864–873. https://​
1076/​jcen.​21.4.​567.​886 doi.​org/​10.​1177/​02698​81120​920453
52. Faul F, Erdfelder E, Buchner A, Lang A-G (2009) Statistical
power analyses using G*Power 3.1: tests for correlation and Publisher's Note Springer Nature remains neutral with regard to
regression analyses. Behav Res Methods 41:1149–1160. https://​ jurisdictional claims in published maps and institutional affiliations.
doi.​org/​10.​3758/​BRM.​41.4.​1149
53. Hartmann AS, Thomas JJ, Wilson AC, Wilhelm S (2013) Insight
impairment in body image disorders: delusionality and overvalued

13