Cognitive flexibility and attention to detail in adolescents and adults with severe forms of anorexia nervosa

Abstract

Objective: To determine if adolescents and adults diagnosed with anorexia nervosa differ in their levels of cognitive flexibility and attention to detail independently of potential confounds.

Method: Sixty-two adolescents and 54 adults were assessed while receiving inpatient treatment and completed the following self-reports: Eating Disorders

Results: Comparisons of the adolescents and adults with anorexia nervosa revealed no significant differences for any of the neuropsychological test scores even after adjusting for potential confounding factors. Neither cognitive flexibility nor attention to detail were associated with level of eating disorder symptomatology, depression, anxiety or obsessive-compulsive symptom-atology. Unlike age, illness duration was found weakly associated with perseverative errors Wisconsin Card Sorting Test and with the central coherence index of the Rey Complex Figure recall condition.

Conclusions: Set-shifting and central coherence performance were independent of age, clinical symptoms severity and emotional status. Additional studies on the relationship between the duration of anorexia nervosa and neuropsychological difficulties are needed.

Full text

Received: 21 June 2021
-
Revised: 14 December 2021
-
Accepted: 24 December 2021
DOI: 10.1002/erv.2883
BRIEF REPORT
Cognitive exibility and attention to detail in adolescents
and adults with severe forms of anorexia nervosa
Sylvie Berthoz
1,2
|
Anne‐Solène Maria
3
|Damien Ringuenet
4
|
Léna Bourdier
5
|Isabelle Nicolas
2
|Corinne Blanchet
6,7
|
Christine Foulon
8
|
Guillaume Lavoisy
8
|Nathalie Godart
6,9,10
|Caroline Barry
6
1
Université Bordeaux, CNRS, EPHE,
INCIA, UMR 5287, Bordeaux, France
2
Department of Psychiatry for Adolescents
and Young Adults, Institut Mutualiste
Montsouris, Paris, France
3
Private Practice, Stimulus Consulting Ltd,
Cascavelle, Mauritius
4
APHP, Hôpital Paul Brousse,
Département de Psychiatrie, Unité
Troubles du Comportement Alimentaires,
Villejuif, France
5
Addictology Department, University
Hospital of Bordeaux, Bordeaux, France
6
Université Paris‐Saclay, Université Paris‐
Sud, UVSQ, CESP, INSERM, Villejuif,
France
7
APHP, Hôpital Cochin, Maison des
Adolescents‐Maison de Solenn, Paris,
France
8
Clinique Villa Montsouris, Paris, France
9
UFR Simone Veil, UVSQ, Univ. Paris‐
Saclay, Montigny‐le‐Bretonneux, France
10
Fondation Santé des Etudiants de
France, Paris, France
Correspondence
Sylvie Berthoz, INCIA CNRS UMR 5287,
Bat 2a, Zone Nord, 146 rue Léo Saignat,
33076 Bordeaux Cdx, France.
Email: sylvie.berthoz-landron@inserm.fr
Funding information
Protocole Hospitalier de Recherche
Abstract
Objective: To determine if adolescents and adults diagnosed with anorexia
nervosa differ in their levels of cognitive exibility and attention to detail
independently of potential confounds.
Method: Sixty‐two adolescents and 54 adults were assessed while receiving
inpatient treatment and completed the following self‐reports: Eating Disorders
Examination Questionnaire, Maudsley Obsessive Compulsive Inventory and
Hospital Anxiety and Depression scale. Performance‐based evaluations
included the Wisconsin Card Sorting Test Computerised Version, the
Comprehensive Trail Making Test, the Brixton Spatial Anticipation Test, the
Rey Complex Figure and the Group Embedded Figures Test.
Results: Comparisons of the adolescents and adults with anorexia nervosa
revealed no signicant differences for any of the neuropsychological test
scores even after adjusting for potential confounding factors. Neither cognitive
exibility nor attention to detail were associated with level of eating disorder
symptomatology, depression, anxiety or obsessive‐compulsive symptom-
atology. Unlike age, illness duration was found weakly associated with
perseverative errors Wisconsin Card Sorting Test and with the central
coherence index of the Rey Complex Figure recall condition.
Conclusions: Set‐shifting and central coherence performance were indepen-
dent of age, clinical symptoms severity and emotional status. Additional
studies on the relationship between the duration of anorexia nervosa and
neuropsychological difculties are needed.
KEYWORDS
anorexia nervosa, central coherence, development, neurocognition, set‐shifting
Abbreviations: AN‐R, Anorexia Nervosa Restrictive type; AN‐BP, Anorexia Nervosa Binge/Purging type; BMI, Body Mass Index; BSAT, Brixton
Spatial Anticipation Test; CI, condence interval; CTMT F5‐F1, Comprehensive Trait Making Test time of completion of form 5 minus form 1;
EDE‐Q, Eating Disorder Examination‐Questionnaire; GEFT, Group Embedded Figures Test; HADS, Hospital Anxiety and Depression Scale; IOTF,
International Obesity Task Force; MOCI, Maudsley Obsessive Compulsive Inventory; RCF CCI, Rey Complex Figure Central Coherence Index; SD,
standard deviation; WAIS, Wechsler Adult Intelligence Scale; WCST, Wisconsin Card Sorting Test.
© 2022 Eating Disorders Association and John Wiley & Sons Ltd.
Eur Eat Disorders Rev. 2022;1–9. wileyonlinelibrary.com/journal/erv
-
1

Clinique, Grant/Award Number: PHRC‐
2010‐AOR10099; Research Award from
the Institut Benjamin Delessert, Grant/
Award Number: IBD AO‐2012; Regional
Council of the Parisian Area Ile de France,
Grant/Award Number: PICRI‐2011
Highlights
�Past studies among people suffering from anorexia nervosa that investigated
if difculties in cognitive exibility and poor central coherence are inde-
pendent of age have provided inconsistent results
�Adolescents and adults with severe anorexia nervosa obtained similar per-
formances on a variety of task indexing set‐shifting and attention to details
�While age and neurocognitive performances were not associated, the longer
the illness duration the lower the performances
1
|
INTRODUCTION AND AIMS
Cognitive difculties may play a critical role in the devel-
opment and maintenance of anorexia nervosa (AN; Trea-
sure & Schmidt, 2013). People with AN would lack
cognitive exibility (also called set‐shifting), a high‐level
cognitive process that refers to the ability to change (i.e.
shift) selectively between mental processes or thoughts as a
function of environmental stimuli or demands and to make
appropriate behavioural adjustments (Treasure &
Schmidt, 2013; Zhou et al., 2018). The balance between a
detail‐oriented and a global processing of information, is
another form of cognitive dysfunction that has been
incriminated in AN.
While diminished cognitive exibility and central
coherence have been posited as viable endophenotypes of
the disorder (Treasure & Schmidt, 2013; Zhou et al., 2018),
which would imply that they are traits independent of age
effects, the studies examining the presence of these dif-
culties in youth with AN and their potential differences
relative to adults diagnosed with this disorder have pro-
vided inconsistent results (Lang et al., 2015; Smith
et al., 2018; Weinbach et al., 2019; Wu et al., 2014). In the
rst meta‐analysis of case‐control studies using tests of set‐
shifting in young people diagnosed with AN, Lang, Stahl,
et al. (2014) found a null pooled effect size (d= −0.005) for
the Trail‐Making Test (TMT; based on ve studies) and a
small but non‐signicant pooled effect size (d= 0.196) for
the Wisconsin Card Sorting Test (WCST; three studies
included), these two tests being the most commonly used.
In a meta‐analysis including adolescent and adult case‐
control studies using the WCST (Westwood et al., 2016),
a small but non‐signicant pooled effect size was observed
for the adolescents (d= 0.25; four studies included) while a
signicant medium pooled effect size was observed for the
adults (d= 0.48; 18 studies included). In a meta‐analysis
that included a larger number of studies and a variety of
set‐shifting tasks (Wu et al., 2014), found medium pooled
effect sizes, with no signicant difference between ado-
lescents and adults (−0.38 vs. −0.45, p= 0.627; 11 studies
with adolescents and 37 with adults). Nevertheless, pos-
terior case‐control studies in samples of young AN patients
reported no differences (e.g. see Weinbach et al., 2019). In
regards to central coherence, no specic meta‐analysis of
case‐control studies in youth with AN has been conducted
but Lang et al. (2016) collated data from 8 datasets and
found that patients with AN (mean age 26.1(8.1)) had
weaker central coherence than healthy controls (d=0.26).
In their systematic review on this issue, Lang and Tchan-
turia suggested that young and adult patients with AN have
similar difculties on the Rey Complex Figure test (RCF),
one of the most commonly used tests for measuring central
coherence along with the Group Embedded Figures Test
(GEFT; Lang & Tchanturia, 2014). However, this conclu-
sion was based on the visual inspection of graphical rep-
resentations of results obtained from children and
adolescents compared to studies of adults with AN.
Moreover, meta‐analyses of central coherence in AN
revealed overall medium pooled effects sizes with the RCF
and the GEFT, but with no information on potential dif-
ferences between adolescent and adult populations (Lang,
Lopez, et al., 2014; Keegan et al., 2021).
Among the methodological concerns about past
research on this topic are the fact many studies (a) relied
on a single task; (b) included both adolescents and adults,
while drawing conclusions about the cognitive charac-
teristics of adult patients only, and (c) overlooked the
potential inuence of confounds, notably heterogeneity
in nutritional status, emotional status and psychotropic
medications (Fuglset, 2021; Smith et al., 2018). Moreover,
there is limited research on the relationship between
neurocognitive functioning and the severity of the eating
disorder symptomatology.
The present study aimed to address these shortcomings
in a large sample of adolescents and adults inpatients with
AN. In line with the suggestion that weaknesses in cogni-
tive exibility and central coherence are endophenotypic
biomarkers of AN, and not illness‐acquired difculties,
we expected similar performances among the adolescents
and the adults. Moreover, based on the Cognitive‐
Interpersonal model of AN, we expected that greater dif-
culties in cognitive exibility and central coherence
would be associated with more severe psychological and
behavioural symptoms of AN in the overall sample.
2
-
BERTHOZ
ET AL.

2
|
METHOD
2.1
|
Participants
The sample was composed of patients hospitalised for a
severe form of AN. Participants were required to be
women, between the ages of 15 and 40 years old, hos-
pitalised for AN, uent in French. Exclusion criteria
were: current substance use disorder, schizophrenia,
history of neurological disorder, or serious, progressive or
life‐threatening physical disease pathology, as well as
neuroleptic or psychotropic treatment at doses causing
drowsiness or concentration problems or colour
blindness.
The study protocol was approved by an independent
ethics committee and registered (Comité de Protection
des Personnes Ile de France VII; clinicaltrial.gov id:
NCT01772394). Participants (as well as their legal repre-
sentatives in case of minor participants) provided their
written informed consent.
Participants were enroled once they were admitted to
one of the four ED inpatient care units that were part of
the study. Baseline assessment occurred after the partic-
ipants had gained at least one quarter of their discharge
target weight and were considered medically stabilised by
the inpatient care unit team.
2.2
|
Measures
The Eating Disorders Examination Questionnaire (EDE‐Q
v5, Fairburn & Beglin, 1994) is a self‐report composed of
33 items evaluating the frequency of eating disorders
symptoms during the last 28 days with four subscores:
Restraint, Eating Concern, Shape Concern and Weight
Concern. High scores indicate a higher frequency of
symptoms and greater clinical severity.
The Hospital Anxiety and Depression Scale (HADS,
Lépine et al., 1985; Zigmond & Snaith, 1983) is a self‐
report comprising seven items evaluating depression
and seven evaluating anxiety. Higher scores indicate a
higher level of symptoms for each dimension.
The Maudsley Obsessive Compulsive Inventory (MOCI,
Hantouche & Guel, 1993; Hodgson & Rachman, 1977) is a
self‐report including 30 items that refer to the most com-
mon complaints of subjects suffering from obsessional
symptoms. In addition to a total score, four partial scores
can be calculated: Checking, Washing, Doubting and
Slowness.
The Wechsler Adult Intelligence Scale short form D
(WAIS, Grégoire & Wierzbicki, 2009) comprises four
subtests (Vocabulary, Similarities, Block Design and
Matrix Reasoning) and gives an approximation of the
Intellectual Quotient (IQ). Norms are the same as the full
version of the WAIS‐III (Wechsler, 1997) with a mean
score of 100 and a standard deviation of 15.
The Wisconsin Card Sorting Test Computerised
Version 4 (WCST, Research Edition, Heaton & Psycho-
logical Assessment Resources, 2003) is a neuropsycho-
logical evaluation assessing cognitive rigidity. The
objective is to match stimuli cards with one of four
category cards. Category cards include several stimuli
(different colours, shapes and numbers of shapes), each
of them dening a sorting rule. The participant is not
aware of the sorting rule and has to discover it, by trial
and error, with the help of a feedback after each sort
attempt. After 10 correct sorts, the sorting rule changes
and the participant has to discover and adapt to the new
rule. The test comprises ve rule shifts (Colour–Shape–
Number–Colour–Shape–Number). It ends when the
participant has completed 10 correct sorts in each cate-
gory, or when the 128 cards are sorted. The following
parameters were used: card animation move time of
1.5 s; visual feedback time of 1 s; frame time of 10 ms. In
the present study, the variable of interest was the per-
centage of perseverative errors which is considered the
most representative index of set‐shifting difculties in
AN and is associated with the largest effect size in studies
of AN inpatients (Tchanturia et al., 2012; Westwood
et al., 2016).
The Comprehensive Trail Making Test (CTMT, Rey-
nolds, 2002) is an evaluation of exibility based on the
TMT (Reitan, 1956) that comprises ve different parts.
Here we have used the Forms 1 and 5 of the CTMT,
which are very similar to the original version of the TMT.
The main difference between the two is that the CTMT
comprises blank stimuli in addition to the relevant ones
that serve as distractors. It is considered to be a more
sensitive test for participants with subclinical cognitive
difculties, as it is the case for patients with AN. In Form
1, participants are asked to connect circles containing
numbers from 1 to 25 in a chronological manner and as
quickly as possible. In Form 5, the task sheet comprises
circles containing numbers (from 1 to 13) and letters
(from A to L) that the participant has to connect in a
chronological way, switching between letters and
numbers, as quickly as possible. While Form 1 evaluates
the speed of execution in a neutral condition, Form 5
assesses the level of set‐shifting. The variable of interest
was the difference between the time to complete Form 5
and that needed to complete Form 1, with higher values
indicating lesser exibility.
The Brixton Spatial Anticipation Test (BSAT,
Burgess & Shallice, 1997) is an evaluation of cognitive
exibility in which participants are asked to predict the
movement of a blue circle among 10 positions over 56
BERTHOZ ET AL.
-
3

trials. The movement rule changes during the test and the
participant has to detect the new rule as quickly as
possible. Here, we used a semi‐computerised version of
the task. The variable of interest was the number of
perseverative errors.
The Rey Complex Figure (RCF, Osterrieth, 1944;
Rey, 1941) is a visuospatial evaluation in which the
participant is asked to reproduce the gure while it is
in view (copy) and again from memory, 20 min later,
without prior warning (recall). A particular scoring has
been developed (Savage et al. (1999), Booth &
Denise (2006)) to assess the strategies used to repro-
duce the gure. For each condition (copy and recall), a
central coherence index (CCI) can be computed, based
on the order of the construction scores (order in which
the different elements of the gure are drawn) and
Style scores (continuity of drawing; see https://www.
katetchanturia.com/clinical‐work‐packages‐‐protocols).
Scores range from 0, which results from a detailed
approach of the task, to 2, which results from a ‘bigger
picture’ approach.
The Group Embedded Figures Test (GEFT, Witkin
et al., 1971) is a visuospatial evaluation in which the
participant is asked to identify and trace a hidden
target simple shape within an embedding context (i.e.
complex gure). The variable of interest is the num-
ber of shapes correctly identied among the 18 tar-
gets. Higher values result from a more detailed
approach.
2.3
|
Statistical analyses
Internal consistency for each of the self‐report measure
was assessed using Cronbach Alphas (see Table S1).
Adolescent‐adult comparisons (below 18 years vs.
18 years or more) were rst conducted using independent
student t‐tests or chi square tests. Effect sizes were
expressed using Cramer's Vor Cohen's dwith their cor-
responding 95% condence intervals. Associations be-
tween cognitive performance or clinical severity with age
group were also tested by means of multiple regression
analyses with, respectively, the variables of interest as
outcomes and age group, AN subtype, IOTF BMI and
neuroleptic medication as regressors. Correlations be-
tween the neuropsychological and clinical feature vari-
ables of interest were assessed in the overall sample using
Pearson coefcients.
Analyses were conducted using R 3.6.1 (R Core
Team, 2019). Missing data were not imputed; and
p< 0.05 was considered statistically signicant.
3
|
RESULTS
3.1
|
Sample's characteristics
The descriptive statistics of the study population are
presented in the Table S2. The sample was comprised of
116 women aged from 15 to 40 years (mean
age = 20.58 years; SD = 5.94) with a mean illness duration
of 4.32 years (SD = 5.39). The sample had a mean number
of previous hospitalisation for AN of 2.24 (SD = 2.17),
and a Mean Body Mass Index (BMI) of 14.11 (SD = 1.35)
at hospital admission and of 15.64 (SD = 1.55) at the
time of the evaluation, and 56.9% of the sample had
a IOTF BMI <16. Regarding the subtype of AN, 75%
of the sample presented a restrictive subtype, 25% a
binge‐purging subtype, and 12.1% were premenarcheal
forms. Antidepressants were consumed by 39.1%
(N= 45), Anxiolytics by 53.9% (N= 62), Neurolep-
tics by 33.9% (N= 39), and Hypnotics by 14.8%
(N= 17).
The descriptive statistics for the clinical question-
naires and cognitive performance test scores by age group
are displayed in Table 1.
Concerning the clinical prole, as expected, the
duration of illness was signicantly longer for the adults
than for the adolescents, with a large effect size (p< 0.001;
d= 0.93). There were no statistically signicant differ-
ences between the adolescents and adults for the AN
subtypes (restrictive: 80.6% vs. 68.5%; binge‐purging:
19.4% vs. 31.5%, p= 0.132, V= 0.12; premenarcheal: 14.5%
vs. 9.3%, p= 0.386, V= 0.05) or for medication (antide-
pressants: 41% vs. 37%, p= 0.665, V= 0.02; anxiolytics:
52.5% vs. 55.6%, p= 0.740, V= 0.11; neuroleptics: 41% vs.
25.9%, p= 0.089, V= 0.14; hypnotics: 14.8% vs. 14.8%,
p= 0.993, V= 0.0). Grade 3 IOTF thinness concerned
fewer adolescents than adults (48.4% vs. 66.7%; p= 0.047,
V= 0.17). In addition, effect sizes were negligible with no
statistically signicant difference for any of the clinical
questionnaire scores, with the exception of EDE‐Q Re-
straint scores, which were signicantly and moderately
higher among the adolescents than the adults (p= 0.042;
d= 0.38, 95% CI (−0.74; −0.01)).
3.2
|
Power calculation
Power simulations were performed a priori, based on the
number of adolescents and adults available in the dataset.
There were 62 adolescents (mean age = 16.67 years;
SD = 0.91) and 54 adults (mean age = 25.06 years;
SD = 6.12). A two sample t‐test with a type I error of 0.05
4
-
BERTHOZ
ET AL.

therefore provides a power of 99% for detecting a large
effect size (Cohen's d>= 0.8), a power of 76% for
detecting a medium effect size (d= 0.5), and of 19% for
detecting a small effect size (d= 0.2).
3.3
|
Comparisons of the
neuropsychological performances
between the adolescents and the adults
Concerning the neuropsychological tests, adolescents and
adults did not differ signicantly on any of the variables.
In consideration of IOTF thinness grade 3 differences,
and the slightly different proportion of participants by
AN subtype and under neuroleptic medications between
the two age groups (with 0.10 < Cramer's V< 0.20),
multiple regression analyses were conducted adjusting
for these potential confounding factors. The results
remained unchanged (see Table S3).
The correlation matrix between the variables of interest
in the overall sample is presented in Table 2. We found no
signicant association (all r< 0.20) between cognitive
exibility or attention to detail and any of the measures of
psychological and behavioural symptoms of AN. Illness
duration, but not age, was signicantly but weakly
associated with WCST% of perseverative errors (r= 0.227;
p= 0.015) and the Recall Central Coherence Index
(r= −0.185; p= 0.048). These associations would no longer
remain signicant if corrected for multiple comparisons.
4
|
DISCUSSION
In this study, we used different performance‐based eval-
uations of cognitive exibility and attention to detail to
compare the neurocognitive prole of two relatively large
samples of older adolescents and adults hospitalised for
AN. To the best of our knowledge, this is the rst study
comparing the performances of adolescents and adults
with AN using the most commonly used tests of set‐
shifting and central coherence in this disorder.
In addition to an absence of signicant differences,
the observed effect sizes were negligible (<0.20). Further,
age and neurocognitive performances were not signi-
cantly correlated in the overall sample. While this study
is cross‐sectional and did not include a control group of
unaffected participants, we believe the fact adolescents
and adults with severe AN have similar levels of perfor-
mance argues against the suggestion that poor cognitive
rigidity and weak central coherence are merely illness‐
TABLE 1Clinical questionnaires and cognitive performance tests' scores by age group
Adolescents Adults t‐tests 95% CI
Mean SD Mean SD pvalue Cohen's d Lower Upper
EDE‐Q total 2.96 1.57 2.55 1.24 0.1242 −0.29 −0.66 0.08
EDE‐Q restraint 2.38 1.98 1.71 1.44 0.0422 −0.38 −0.74 −0.01
EDE‐Q eating 2.42 1.53 2.03 1.14 0.1227 −0.29 −0.66 0.08
EDE‐Q shape 3.93 1.83 3.57 1.55 0.2554 −0.21 −0.58 0.16
EDE‐Q weight 3.12 1.69 2.88 1.49 0.4315 −0.15 −0.52 0.22
MOCI 10.41 5.42 9.37 6.26 0.3422 −0.18 −0.55 0.19
HADS depression 7.60 3.70 7.30 3.57 0.6584 −0.08 −0.45 0.29
HADS anxiety 10.58 4.55 10.43 3.98 0.8470 −0.04 −0.41 0.33
WAIS 101.25 11.06 100.96 12.51 0.8972 −0.02 −0.4 0.35
CTMT F5‐F1 14.52 22.71 15.91 16.33 0.7093 0.07 −0.3 0.44
BSAT 10.98 7.35 11.15 6.29 0.8983 0.02 −0.35 0.4
WCST 11.49 6.10 12.63 7.96 0.3867 0.16 −0.21 0.53
RCF CCI copy 1.56 0.20 1.60 0.19 0.3094 0.19 −0.18 0.56
RCF CCI recall 1.40 0.37 1.36 0.31 0.6306 −0.09 −0.46 0.28
GEFT 15.00 2.80 15.72 2.73 0.1651 0.26 −0.11 0.63
Abbreviations: BSAT, Brixton Spatial Anticipation Test; CTMT F5‐F1, Comprehensive Trait Making Test time of completion of form 5 minus form 1; EDE‐Q,
Eating Disorder Examination‐Questionnaire; GEFT, Group Embedded Figures Test; HADS, Hospital Anxiety and Depression Scale; MOCI, Maudsley
Obsessive Compulsive Inventory; RCF CCI, Rey Complex Figure Central Coherence Index; SD, Standard Deviation; WAIS, Wechsler Adult Intelligence Scale.
BERTHOZ ET AL.
-
5

TABLE 2Correlation matrix (Pearson) between the variables of interest in the overall sample
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
1. Age ***** 0.757 −0.077 −0.090 −0.109 −0.036 0.032 −0.053 −0.079 −0.027 −0.072 −0.012 0.042 0.153 −0.049 −0.158 −0.063
2. Illness duration <0.001 ***** −0.053 0.054 0.016 0.115 0.147 0.098 −0.026 0.008 0.061 −0.032 0.115 0.227 −0.092 −0.185 −0.140
3. BMI 0.409 0.573 ***** 0.100 0.073 0.127 0.133 0.123 0.240 −0.010 −0.004 0.111 −0.072 −0.078 −0.108 −0.045 −0.249
4. EDE‐Q restraint 0.339 0.562 0.284 ***** 0.650 0.609 0.656 0.827 0.284 0.447 0.499 −0.038 0.131 0.013 −0.139 −0.147 −0.081
5. EDE‐Q eating 0.247 0.864 0.439 <0.001 ***** 0.801 0.789 0.896 0.277 0.485 0.558 −0.025 0.122 0.156 −0.002 −0.154 −0.103
6. EDE‐Q shape 0.702 0.221 0.177 <0.001 <0.001 ***** 0.872 0.919 0.397 0.420 0.524 −0.037 0.089 0.126 −0.022 −0.114 −0.113
7. EDE‐Q weight 0.733 0.117 0.156 <0.001 <0.001 <0.001 ***** 0.928 0.340 0.483 0.536 −0.021 0.094 0.124 0.018 −0.081 −0.123
8. EDE‐Q total 0.571 0.297 0.191 <0.001 <0.001 <0.001 <0.001 ***** 0.366 0.512 0.592 −0.034 0.122 0.120 −0.045 −0.143 −0.119
9. MOCI 0.401 0.781 0.010 0.002 0.003 <0.001 <0.001 <0.001 ***** 0.194 0.387 0.157 −0.012 0.055 0.025 0.043 −0.121
10. HADS Dep. 0.777 0.935 0.916 <0.001 <0.001 <0.001 <0.001 <0.001 0.037 ***** 0.438 0.021 0.153 0.158 −0.088 −0.120 −0.019
11. HADS anxiety 0.444 0.516 0.965 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 ***** 0.039 0.156 0.117 −0.001 −0.032 −0.350
12. CTMT F5‐F1 0.899 0.736 0.237 0.684 0.793 0.696 0.824 0.718 0.095 0.825 0.677 ***** 0.043 0.118 −0.033 −0.086 −0.219
13. BSAT 0.659 0.221 0.443 0.162 0.198 0.345 0.322 0.197 0.901 0.103 0.096 0.650 ***** 0.212 −0.180 −0.323 −0.093
14. WSCT 0.103 0.015 0.409 0.889 0.097 0.180 0.190 0.205 0.564 0.091 0.212 0.208 0.023 ***** 0.041 −0.223 −0.212
15. RCF CCI copy 0.607 0.332 0.252 0.141 0.981 0.818 0.847 0.636 0.797 0.349 0.995 0.730 0.056 0.668 ***** 0.527 0.063
16. RCF CCI recall 0.094 0.048 0.637 0.118 0.104 0.231 0.394 0.130 0.651 0.203 0.736 0.362 <0.001 0.018 <0.001 ***** 0.095
17. GEFT 0.501 0.136 0.007 0.392 0.273 0.230 0.193 0.208 0.198 0.842 0.708 0.019 0.327 0.023 0.507 0.313 *****
Note: Upper part of the diagonal: coefcients of correlation; lower part of the diagonal: pvalues.
Abbreviations: BSAT, Brixton Spatial Anticipation Test; CTMT F5‐F1, Comprehensive Trait Making Test time of completion of form 5 minus form 1; EDE‐Q, Eating Disorder Examination ‐ Questionnaire; GEFT,
Group Embedded Figures Test; HADS, Hospital Anxiety and Depression Scale; MOCI, Maudsley Obsessive Compulsive Inventory; RCF CCI, Rey Complex Figure Central Coherence Index.
6
-
BERTHOZ
ET AL.

acquired difculties. A strength of our study design was
the possibility to control for the potential inuence of the
most frequently implicated confounding factors: (a) the
subtype of AN (premenarcheal or not; restrictive vs.
binge‐purging), (b) the severity of malnutrition, (c) the
heterogeneity of comorbid mood or anxiety and perceived
levels of dysphoric affects (as measured by the HADS),
and (d) psychotropic medications.
Unlike age, we found a connection between a longer
illness duration and greater cognitive rigidity (as
measured by WCST percentage of perseverative errors)
on the one hand, and poorer central coherence (during
the recall condition of the RCF) on the other. If adjust-
ments for multiple comparisons were made, these asso-
ciations would not remain signicant. However, this
nding is consistent with the conclusion of a recent
systematic review and meta‐analysis showing that prob-
lems in cognitive exibility, central coherence and
emotion recognition are observable among individuals
with prolonged AN (7 years or over) but to a lesser degree
among those with a shorter duration of illness (Saure
et al., 2020). Thus, although age and duration of illness
are highly correlated, future studies would need to more
systematically disentangle their potential effects. More-
over, our age groups were dichotomised based on the
national legal age of adulthood, which corresponds
closely to the benchmarks of the traditional denition of
older adolescents. However, with the advances of neu-
rosciences relative to biological growth and/or brain
development, the need to adopt new denitions of
adolescence that extend this period to the age of 24 is
increasingly discussed (Sawyer et al., 2018). If longitudi-
nal designs will be necessary to draw rm conclusions,
the current ndings are coherent with the assumption
posited in the Cognitive‐Interpersonal model of AN that
neurocognitive difculties would contribute to a more
chronic form of this disorder (Treasure & Schmidt, 2013).
We found only negligible and non‐statistically signif-
icant associations between neurocognitive performances
and levels of malnutrition or ED symptomatology, sug-
gesting at rst glance that neurocognitive weaknesses
and the clinical severity of AN are not related. However,
our sample included only severe forms of this disorder,
with little amount of variability among the observations ‐
a characteristic that is known to impact the rvalues
(Goodwin & Leech, 2006). For this reason, our results
may not be generalisable to younger adolescents or less
severe cases of AN. Nevertheless, this nding as well as
the absence of association between the different test
scores and dysphoric affects are in line with the conclu-
sion of Fuglset's systematic review (Fuglset, 2021) that
BMI, depression and anxiety do not inuence set‐shifting
or central coherence in AN.
In summary, additional studies on the relationships
between the duration of AN and neuropsychological
difculties are needed. Concerning treatment implica-
tions, cognitive remediation therapy (CRT) for AN is an
intervention that targets skills in set‐shifting and central
coherence. It aims at improving the awareness of the
potential negative impact of these difculties on the
ability to recover, notably by favouring intolerance for
changes in ED routines, stereotyped behaviours, erro-
neous beliefs and ‘jumping to conclusions’. This pro-
gramme has been initially developed for adults, and
progressively adapted for younger individuals (Giombini
et al., 2022; Tchanturia et al., 2014,2017). Since set‐
shifting and central coherence difculties do not seem
related to age but to the progression of the disorder,
delivering CRT from the onset of the illness may help
change and improve the recovery trajectory.
ACKNOWLEDGEMENTS
We thank Laura Bignami, Maurice Corcos, Annaïg
Courty, Jeanne Duclos, Delphine de Hauteclocque,
Marion Jahan, Fanny Kahalé, Marie‐Rose Moro, Leslie
Radon for their invaluable help in setting up the study
and/or collecting the data. This research received
funding from the Protocole Hospitalier de Recherche
Clinique (PHRC‐2010‐AOR10099) and a Research
Award from the Institut Benjamin Delessert (IBD AO‐
2012). AS Maria received a PhD fellowship award from
the Regional Council of the Parisian Area Ile de France
(PICRI‐2011).
CONFLICT OF INTEREST
Sylvie Berthoz has received fees for editorial activities
with Elsevier. The other authors declare no conict of
interest.
DATA AVAILABILITY STATEMENT
The data that support the ndings of this study are
available on request from the corresponding author. The
data are not publicly available due to privacy or ethical
restrictions.
ORCID
Sylvie Berthoz
https://orcid.org/0000-0002-6862-2362
Corinne Blanchet https://orcid.org/0000-0003-3135-
5210
REFERENCES
Booth, R., & Denise, L. (2006). Local‐global processing and cognitive
style in autism spectrum disorders and typical development.
King’s College London, University of London.
Burgess, P. W., & Shallice, T. (1997). The Hayling and Brixton tests.
Thames Valley Test Company.
BERTHOZ ET AL.
-
7

Fairburn, C. G., & Beglin, S. J. (1994). Assessment of eating disor-
ders: Interview or self‐report questionnaire? International
Journal of Eating Disorders,16(4), 363–370.
Fuglset, T. S. (2021). Is set‐shifting and central coherence in
anorexia nervosa inuenced by body mass index, anxiety or
depression? A systematic review. BMC Psychiatry,21(1), 137.
https://doi.org/10.1186/s12888‐021‐03120‐6
Giombini, L., Nesbitt, S., Kusosa, R., Hinallas, K., Easter, A., &
Tchanturia, K. (2022). Neuropsychological and clinical
ndings of cognitive remediation therapy feasibility rando-
mised controlled trial in young people with anorexia nervosa.
European Eating Disorders Review,30(1), 50–60. https://doi.
org/10.1002/erv.2874
Goodwin, L. D., & Leech, N. L. (2006). Understanding correlation:
Factors that affect the size of r. The Journal of Experimental
Education,74(3), 249–266. https://doi.org/10.3200/JEXE.74.3.
249‐266
Grégoire, J., & Wierzbicki, C. (2009). Comparaison de quatre formes
abrégées de l’échelle d’intelligence de Wechsler pour adultes –
troisième édition (WAIS‐III). European Review of Applied
Psychology,59(1), 17–24. https://doi.org/10.1016/j.erap.2007.
08.003
Hantouche, E., & Guel, J. D. (1993). Self‐evaluation of obsessive‐
compulsive disorder. Adaptation and validation of two psycho-
metric scales to the French version. L’Encephale,19(3), 241–248.
Heaton, R. K., & Psychological Assessment Resources. (2003).
WCST:CV4
TM
. Wisconsin Card Sorting Test®: Computer Version
4—Research Edition. Psychological Assessment Resources.
Hodgson, R. J., & Rachman, S. (1977). Obsessional‐compulsive
complaints. Behaviour Research and Therapy,15(5), 389–395.
https://doi.org/10.1016/0005‐7967(77)90042‐0
Keegan, E., Tchanturia, K., & Wade, T. D. (2021). Central coherence
and set‐shifting between nonunderweight eating disorders and
anorexia nervosa: A systematic review and meta‐analysis. In-
ternational Journal of Eating Disorders,54(3), 229–243. https://
doi.org/10.1002/eat.23430
Lang, K., Lopez, C., Stahl, D., Tchanturia, K., & Treasure, J. (2014).
Central coherence in eating disorders: An updated systematic
review and meta‐analysis. World Journal of Biological Psychi-
atry,15(8), 586–598. https://doi.org/10.3109/15622975.2014.
909606
Lang, K., Lloyd, S., Khondoker, M., Simic, M., Treasure, J., &
Tchanturia, K. (2015). Do children and adolescents with
Anorexia Nervosa display an inefcient cognitive processing
style? PLoS ONE,10(7), e0131724. https://doi.org/10.1371/
journal.pone.0131724
Lang, K., Roberts, M., Lopez, C., Goddard, E., Khondoker, M.,
Treasure, J., & Tchanturia, K. (2016). An investigation of central
coherence in eating disorders: A synthesis of studies using the
Rey Osterrieth Complex Figure Test. PLoS ONE,11(11),
e0165467. https://doi.org/10.1371/journal.pone.0165467
Lang, K., Stahl, D., Espie, J., Treasure, J., & Tchanturia, K. (2014).
Set shifting in children and adolescents with anorexia nervosa:
An exploratory systematic review and meta‐analysis: Set
shifting in adolescents with An. International Journal of
Eating Disorders,47(4), 394–399. https://doi.org/10.1002/eat.
22235
Lang, K., & Tchanturia, K. (2014). A systematic review of central
coherence in young people with anorexia nervosa. Journal of
Child and Adolescent Behaviour,02(03). https://doi.org/10.
4172/2375‐4494.1000140
Lépine, J. P., Godchau, M., Brun, P., & Lempérière, T. (1985).
Evaluation de l’anxiété et de la dépression chez des patients
hospitalisés dans un service de médecine interne. Annales
Médico‐Psychologiques,143(2), 175–189.
Osterrieth, P. A. (1944). Le test de copie d’une gure complexe:
Contribution à l’étude de la perception et de la mémoire. Ar-
chives de Psychologie,30, 206–356.
R Core Team. (2019). R: A language and environment for statistical
computing. R Foundation for Statistical (R Foundation for
Statistical Computing). https://www.R‐project.org/
Reitan, R. M. (1956). Trail making test. Manual for administration,
scoring, and interpretation. Indiana University Press.
Rey, A. (1941). L’examen psychologique dans les cas d’encephalo-
pathie traumatique. Archives de Psychologie,28, 215–285.
Reynolds, C. R. (2002). Comprehensive trail‐making test: Examiner’s
manual. PRO‐ED Inc.
Saure, E., Laasonen, M., Lepistö‐Paisley, T., Mikkola, K., Ålgars,
M., & Raevuori, A. (2020). Characteristics of autism spectrum
disorders are associated with longer duration of anorexia
nervosa: A systematic review and meta‐analysis. International
Journal of Eating Disorders,53(7), 1056–1079. https://doi.org/
10.1002/eat.23259
Savage, C. R., Baer, L., Keuthen, N. J., Brown, H. D., Rauch, S. L., &
Jenike, M. A. (1999). Organizational strategies mediate
nonverbal memory impairment in obsessive‐compulsive dis-
order. Biological Psychiatry,45(7), 905–916.
Sawyer, S. M., Azzopardi, P. S., Wickremarathne, D., & Patton,
G. C. (2018). The age of adolescence. The Lancet Child &
Adolescent Health,2(3), 223–228. https://doi.org/10.1016/
S2352‐4642(18)30022‐1
Smith, K. E., Mason, T. B., Johnson, J. S., Lavender, J. M., &
Wonderlich, S. A. (2018). A systematic review of reviews of
neurocognitive functioning in eating disorders: The state‐of‐
the‐literature and future directions. International Journal of
Eating Disorders,51(8), 798–821. https://doi.org/10.1002/eat.
22929
Tchanturia, K., Giombini, L., & Leppanen, J. (2017). Evidence for
cognitive remediation therapy in young people with anorexia
nervosa: Systematic review and meta-analysis of the literature.
European Eating Disorder Review,25(4), 227–236. https://doi.
org/10.1002/erv.2522
Tchanturia, K., Davies, H., Roberts, M., Harrison, A., Nakazato, M.,
Schmidt, U., Treasure, J., & Morris, R. (2012). Poor cognitive
exibility in eating disorders: Examining the evidence using
the Wisconsin Card Sorting Task. PLoS One,7(1), e28331.
https://doi.org/10.1371/journal.pone.0028331
Tchanturia, K., Lounes, N., & Holttum, S. (2014). Cognitive reme-
diation in anorexia nervosa and related conditions: A sys-
tematic review: CRT in anorexia. European Eating Disorders
Review,22(6), 454–462. https://doi.org/10.1002/erv.2326
Treasure, J., & Schmidt, U. (2013). The cognitive‐interpersonal
maintenance model of anorexia nervosa revisited: A sum-
mary of the evidence for cognitive, socio‐emotional and
interpersonal predisposing and perpetuating factors. Journal of
Eating Disorders,1, 13. https://doi.org/10.1186/2050‐2974‐1‐13
Wechsler, D. (1997). WAIS‐III. Administration and scoring manual.
The Psychological Association.
8
-
BERTHOZ
ET AL.

Weinbach, N., Bohon, C., & Lock, J. (2019). Set‐shifting in adoles-
cents with weight‐restored anorexia nervosa and their unaf-
fected family members. Journal of Psychiatric Research,112,
71–76. https://doi.org/10.1016/j.jpsychires.2019.02.022
Westwood, H., Stahl, D., Mandy, W., & Tchanturia, K. (2016). The
set‐shifting proles of anorexia nervosa and autism spectrum
disorder using the Wisconsin Card Sorting Test: A systematic
review and meta‐analysis. Psychological Medicine,46(9),
1809–1827. https://doi.org/10.1017/S0033291716000581
Witkin, H., Oltman, P., Raskin, E., & Karp, S. A. (1971). A manual
for the group embedded gures test. Mind Garden Inc.
Wu, M., Brockmeyer, T., Hartmann, M., Skunde, M., Herzog, W., &
Friederich, H.‐C. (2014). Set‐shifting ability across the spectrum
of eating disorders and in overweight and obesity: A systematic
review and meta‐analysis. Psychological Medicine,44(16),
3365–3385. https://doi.org/10.1017/S0033291714000294
Zhou, Z. C., McAdam, D. B., & Donnelly, D. R. (2018). Endophe-
notypes: A conceptual link between anorexia nervosa and
autism spectrum disorder. Research in Developmental Disabil-
ities,82, 153–165. https://doi.org/10.1016/j.ridd.2017.11.008
Zigmond, A. S., & Snaith, R. P. (1983). The hospital anxiety and
depression scale. Acta Psychiatrica Scandinavica,67(6),
361–370. https://doi.org/10.1111/j.1600‐0447.1983.tb09716.x
SUPPORTING INFORMATION
Additional supporting information may be found in the
online version of the article at the publisher's website.
How to cite this article: Berthoz, S., Maria, A.‐S.,
Ringuenet, D., Bourdier, L., Nicolas, I., Blanchet,
C., Foulon, C., Lavoisy, G., Godart, N., & Barry, C.
(2022). Cognitive exibility and attention to detail
in adolescents and adults with severe forms of
anorexia nervosa. European Eating Disorders
Review, 1–9. https://doi.org/10.1002/erv.2883
BERTHOZ ET AL.
-
9