Content uploaded by Jean-Yves Rotge
Author content
All content in this area was uploaded by Jean-Yves Rotge on Jan 29, 2018
Content may be subject to copyright.
A challenging task for assessment of
checking behaviors in obsessive–compulsive
disorder
Rotge JY, Clair AH, Jaafari N, Hantouche EG, Pelissolo A,
Goillandeau M, Pochon JB, Guehl D, Bioulac B, Burbaud P, Tignol J,
Mallet L, Aouizerate B. A challenging task for assessment of checking
behaviors in obsessive–compulsive disorder.
Objective: The present study concerns the objective and quantitative
measurement of checking activity, which represents the most
frequently observed compulsions in obsessive–compulsive disorder
(OCD). To address this issue, we developed an instrumental task
producing repetitive checking in OCD subjects.
Method: Fifty OCD subjects and 50 normal volunteers (NV) were
administered a delayed matching-to-sample task that offered the
unrestricted opportunity to verify the choice made. Response accuracy,
number of verifications, and response time for choice taken to reflect
the degree of uncertainty and doubt were recorded over 50 consecutive
trials.
Results: Despite similar levels of performance, patients with OCD
demonstrated a greater number of verifications and a longer response
time for choice before checking than NV. Such behavioral patterns
were more pronounced in OCD subjects currently experiencing
checking compulsions.
Conclusion: The present task might be of special relevance for the
quantitative assessment of checking behaviors and for determining
relationships with cognitive processes.
J. Y. Rotge
1,2
, A. H. Clair
3
,
N. Jaafari
4
, E. G. Hantouche
5
,
A. Pelissolo
5,6
, M. Goillandeau
1
,
J. B. Pochon
3
, D. Guehl
1,7
,
B. Bioulac
1,7
, P. Burbaud
1,7
,
J. Tignol
2
, L. Mallet
3
,
B. Aouizerate
1,2
1
Movement, Adaptation, Cognition Laboratory, CNRS
UMR 5227, Bordeaux 2 University, Bordeaux, France,
2
Academic Department of Adult Psychiatry, Charles
Perrens Hospital, Bordeaux, France,
3
Clinical
Investigation Center, Behavior, Emotion and Basal
Ganglia Laboratory, INSERM IFR 70, Piti-SalpÞtrire
Hospital, Paris, France,
4
Academic Department of Adult
Psychiatry, Henri Laborit University Hospital, Poitiers,
France,
5
Academic Department of Adult Psychiatry,
Piti-SalpÞtrire University Hospital, Paris, France,
6
Personnality and Adaptive Behaviors Laboratory, CNRS
UMR 7593, Piti-SalpÞtrire University Hospital, Paris,
France and
7
Academic Department of Neurophysiology,
Pellegrin University Hospital, Bordeaux, France
Key words: checking behavior; delayed matching-to-
sample task; doubt; obsessive–compulsive disorder
Dr Bruno Aouizerate, Academic Department of Adult
Psychiatry, CNRS UMR 5227, Bordeaux 2 University,
Charles Perrens Hospital, 121 rue de la Bchade, 33076
Bordeaux, France.
E-mail: bruno.aouizerate@u-bordeaux2.fr
Accepted for publication January 30, 2008
Significant outcomes
•The present task allows one to assess the propensity to develop checking activity in obsessive–
compulsive disorder and to characterize some cognitive and emotional aspects of this repetitive
behavior.
Limitations
•No neuropsychological test was performed to explore visuospatial abilities and to eliminate deficits
related to spatial working memory and recognition in subjects with obsessive–compulsive disorder.
•Most of the study subjects with obsessive–compulsive disorder were administered with psychotropic
medications. Some among these agents were able to alter cognitive and motor functions, and to
interfere with the task performance.
Acta Psychiatr Scand 2008: 117: 465–473
All rights reserved
DOI: 10.1111/j.1600-0447.2008.01173.x
Copyright 2008 The Authors
Journal Compilation 2008 Blackwell Munksgaard
ACTA PSYCHIATRICA
SCANDINAVICA
465
Introduction
Obsessive–compulsive disorder (OCD) is a rela-
tively frequent anxiety disorder that affects 2–3%
of the general population (1). It is characterized by
persistent, intrusive and unpleasant thoughts,
impulses or images. Recurrent, time-consuming
behaviors are performed in a ritualistic fashion
according to rigid rules to reduce the anxiety and
severe distress caused by obsessions.
Phenomenological considerations suggest that
the core of the obsessional process is the sensation
that Ôsomething is wrongÕ(2, 3). Obsessions may be
understood as the consequence of an excessive
attribution of negative consequences upon expo-
sure to certain behavioral situations (4, 5). Path-
ological doubt, a prominent phenomenon in OCD,
may be related to the permanent perception of a
mistake and ⁄or error in the initial representation,
relating to the cognitive overestimation of the
likelihood of aversive events occurring in response
to the subjectÕs action. Pathological doubt may be
considered not only in terms of cognitive con-
structs, but also in terms of the subjectÕs emotional
reaction to feelings of uncertainty generated by
perceived Ôerror detectionÕsignals (6, 7). Compul-
sions may be construed as behavioral responses
intending to relieve the tensions or anxiety caused
by high levels of self-doubt (8–10). If obtained, this
relief is only transient and leads immediately to
reproducing the behavior in an excessive and
unreasonable manner. Given these considerations,
repetitive checking behaviors, which are the most
frequent form of compulsions in patients with
OCD (1, 11), may be thought of as reducing the
risk of error in the estimated probability of
negative events in specific behavioral situations,
thereby relieving the uncertainty and obsessional
doubt (8–10).
Contemporary strategies for assessment of OCD
severity use a wide variety of psychometric rating
scales. There are several self-report and assessor-
rated instruments that examine different types of
obsessions and compulsions (checking, washing,
etc.) or which are designed to measure the symp-
toms present, regardless of their thematic content.
They also provide information on the efficacy of
conventional treatments in OCD. However,
administration of these rating scales requires clini-
cians who are extensively trained in evaluating
psychopathology and who are well versed in
general clinical interview technique. To be able to
probe the pathophysiological bases of checking
compulsions, a tool is required that would allow us
to assess objectively the propensity to repeated
checking.
Aims of the study
We sought to investigate checking behaviors in
OCD through an objective approach. For this
purpose, we developed an instrumental task that
specifically offered the opportunity to verify once
the choice was made. We examined whether the
present task was able to discriminate checking and
choice behavioral patterns in two subgroups of
patients with OCD who actively experienced
checking compulsions (OCD CC+) or not (OCD
CC)) and normal healthy volunteers (NV).
Material and methods
Subjects
Participants were 50 adult patients with a primary
diagnosis of OCD (28 women, 22 men, mean age
40.3 years, SD 11.8, range 23–65) seen at three
specialized university hospitals (Charles Perrens,
Henri Laborit and Pitie
´-Salpeˆ trie
`re) and 50 normal
healthy volunteers (NV) (28 women, 22 men, mean
age 40.5 years, SD 12.4, range 19–63) who were
recruited from the local community by newspaper
advertisements and word of mouth. All subjects
gave their written informed consent. The groups
did not differ significantly in age or gender. They
were evaluated for OCD and associated psychopa-
thology with the Mini-International Neuropsychi-
atric Interview (MINI) (version 5.0.0) (12).
Patients with OCD with comorbid diagnoses were
not excluded from the study, except for those with
current mood episodes, provided that OCD was
the dominant disorder. Healthy controls were
excluded if they had a history of Axis I psychiatric
disorders. All 50 normal volunteers were free of
medication at the time of study, whereas 45
patients with OCD were receiving psychotropic
drugs, especially antidepressants with serotonin
reuptake-inhibiting properties (n= 45) alone or
combined with atypical antipsychotics (n= 12),
benzodiazepines (n= 12), mood stabilizers
(n= 8) and buspirone (n= 5). Twenty of the 50
patients with OCD were receiving cognitive-behav-
ioral therapies.
Clinical assessment
Instruments included: i) the Yale–Brown Obses-
sive–Compulsive Scale (Y-BOCS) symptom check-
list, a list of more than 50 common obsessions and
compulsions, which was used to identify patients
with OCD with checking compulsions (OCD
CC+) (n= 36) (23 women, 13 men, mean age
41.6 years, SD 11.6, range 20–65) or not (OCD
Rotge et al.
466
CC))(n= 14) (five women, nine men, mean age
37.3 years, SD 12.4, range 23–60) respectively; and
ii) the 10-item clinician-rated Y-BOCS for evalu-
ation of OCD symptom severity (13, 14). Y-BOCS
scores are presented in Table 2.
Task design
The behavioral task was based on a delayed visual
discrimination paradigm with checking option, as
shown in Fig. 1. Under the task condition, the
subjects were presented with the sample, a single
image on a video screen (study phase). Then, they
had to compare the image presented with the initial
image studied and to respond by pressing a left or
right button corresponding to the choice ÔRightÕ
(R) or ÔWrongÕ(W) according to whether these
images were considered as identical or different
respectively (choice phase). After making their
choice, the subjects had the opportunity of pressing
the left button (left arrows) during the subsequent
phase (checking phase) to return to the study phase
to access the initial image presented, before moving
again to the choice phase. Therefore, this paradigm
placed the subject in a potential situation of
unrestricted repetitive checking. Once the subject
had confirmed his ⁄her choice of ÔRightÕ(R) or
ÔWrongÕ(W) by pressing the right button (right
arrows) without recourse to the checking option
during the checking phase, a signal indicating a
correct (Yes) or incorrect (No) response appeared
on the screen (evaluation phase) at the end of each
trial.
To ensure that subjects were familiar with the
procedure, the task was explained and 10 trials
were performed before being completed with the 50
test trials based on the random presentation of
images that were either identical (Id: n= 25) to or
different from the sample (Df: n= 25). The
imagesÕfeature sets were selected from an open
clipart library without attempting to produce OCD
symptoms. Participants were instructed to respond
R
<<
W
>>
YES
or
NO
checking
2s 3.5s 3.5s 2sunlimitedunlimited
Study Choice Checking EvaluationAttention Delay
Phase Phase Phase Phase Phase
Fig. 1. Delayed matching-to-sample checking task. All trials began by a 2-s attention signal on a video screen. This phase was
followed by a presentation of the sample, a single image for 3.5 s (study phase). After a 3.5-s interval, the subject had access to the
subsequent screen displaying a second image, which was compared with the initial image by pressing the left or right button
corresponding to the choice ÔRightÕ(R) or ÔWrongÕ(W) when the two images appeared to be identical or different respectively (choice
phase). Thereafter, the subject had two courses of action available as to either: i) verify the choice made by pressing the left button
corresponding to the two left arrows offering the opportunity to return to the study phase for reviewing the sample image before
participating again in the choice phase; or ii) validate the choice made by pressing the right button corresponding to the two right
arrows, leading to the appearance of a signal indicating a correct (yes) or incorrect response (evaluation phase) at the end of the trial.
Checking behaviors in OCD
467
Ôas efficiently and correctly as possibleÕ. Response
accuracy (expressed as the number of correct
trials), number of verifications and response time
for choice (defined as the time elapsing from the
appearance of the image matching the sample to
the moment the choice was made) were monitored
during the 50 consecutive trials.
Data analyses
Level of performance was compared among the
three groups by using a one-way analysis of variance
(anova). The number of verifications and response
time for choice making were examined by using a
two-way anova with repeated measures on the last
factor followed by a one-way anova to determine
differences between groups according to trial type,
i.e. presenting comparison images identical (Id) to
and different (Df) from the sample or leading to the
production of checking behaviors (CB+) or not
(CB)). The differences for each subject in either the
number of verifications between trials Id and Df or
the response time for choice between trials CB+ and
CB)were analyzed by a one-way anova. Newman–
Keuls tests were used for post hoc analyses where
appropriate. To describe the changes in both
number of verifications and response time for
choice across all 50 trials performed during the
task, linear regression lines were computed for each
group. To determine whether there was a progres-
sive rise in these behavioral markers, as expected in
patients with OCD, correlation coefficients for each
group were converted by using FisherÕsr-to-z
transformation. To test this directional hypothesis,
a one-tailed a= 0.05 was used. PearsonÕs correla-
tions were performed to establish whether there were
any relationships between the behavioral parame-
ters studied and the clinical severity of OCD. An
a= 0.05 (two-tailed) was used as the threshold for
statistical significance for both anova and correla-
tion analyses.
Results
Both subgroups of patients with OCD had similar
illness duration (OCD CC+: mean duration
23.1 years, SD 12.5, range 4–48; OCD CC):
mean duration 22.1 years, SD 14.1, range 4–50)
and psychiatric comorbidity (Table 1). They did
not significantly differ in their Y-BOCS scores
(Table 2). Eight patients with OCD (OCD CC+:
n= 7; OCD CC):n= 1) were in remission
defined as a Y-BOCS score less than 16 (15).
However, none of the patients in this study reached
clinical recovery, as indicated by a Y-BOCS score
less than 8 (15), except one OCD CC).
We first explored the level of performance in the
three groups of subjects OCD CC+, OCD CC)
and NV, as measured by the number of correct
trials. No significant difference was found between
the groups [group effect, F(2,97) = 0.04,
P> 0.96] (Fig. 2a). Thus, the three groups of
subjects had a similar level of performance on the
task. With regard to the number of verifications
during the behavioral task, there was a significant
difference between groups [group effect,
Table 1. Per cent of lifetime and current DSM-IV Axis I comorbidities in patients
with OCD
OCD CC+
(n= 36)
OCD CC)
(n= 14)
n%n%
Lifetime comorbidities
Mood disorders
Major depression 25 69.4 11 78.6
Bipolar I disorder 0 0 0 0
Bipolar II disorder 3 8.3 2 14.3
Anxiety disorders
Panic disorder 1 2.8 0 0
Panic disorder with agoraphobia 2 5.6 0 0
Agoraphobia without panic disorder 0 0 0 0
Social phobia 2 5.6 1 7.1
Post-traumatic stress disorder 1 2.8 0 0
Generalized anxiety disorder 0 0 0 0
Substance use disorders
Alcohol abuse ⁄dependence 0 0 0 0
Drug abuse ⁄dependence 0 0 0 0
Any eating disorder 0 0 0 0
Any psychotic disorder 0 0 0 0
Current comorbidities
Mood disorders
Major depression 0 0 0 0
Bipolar I disorder 0 0 0 0
Bipolar II disorder 0 0 0 0
Anxiety disorders
Panic disorder 1 2.8 0 0
Panic disorder with agoraphobia 2 5.6 0 0
Agoraphobia without panic disorder 0 0 0 0
Social phobia 2 5.6 1 7.1
Post-traumatic stress disorder 1 2.8 0 0
Generalized anxiety disorder 0 0 0 0
Substance use disorders
Alcohol abuse ⁄dependence 0 0 0 0
Drug abuse ⁄dependence 0 0 0 0
Any eating disorder 0 0 0 0
Any psychotic disorder 0 0 0 0
Table 2. Symptom severity scores in patients with OCD
OCD CC+ (n= 36) OCD CC)(n= 14)
Mean SD Range Mean SD Range
Y-BOCS total score 21.8 6.8 9–36 21.9 6.7 7–31
Y-BOCS obsessive subscale score 10.5 4.0 4–18 11.4 3.3 4–15
Y-BOCS compulsive subscale score 11.3 3.7 3–18 10.6 4.0 3–16
OCD CC+, patients with OCD with checking compulsions; OCD CC), patients with
OCD without checking compulsions; Y-BOCS, Yale–Brown Obsessive–compulsive
score.
Rotge et al.
468
F(2,97) = 18.88, P< 0.001] (Fig. 2b). Newman–
Keuls post hoc tests revealed that the number of
verifications in OCD CC+ was greater than in
OCD CC)(P< 0.02), and that of OCD CC)was
higher than that of NV (P< 0.02). When trials
based on the presentation of comparison
images similar to (Id) or different from (Df) the
sample were considered, the pattern of the
groupsÕnumber of verifications differed according
to the type of trial performed [group ·trial inter-
action, F(2,97) = 7.07, P< 0.001]. For either
trials Id or Df, a significant difference was found
between groups [group effect, F(2,97) = 17.07–
18.89, P< 0.001]. Newman–Keuls post hoc tests
indicated that OCD CC+ exhibited a significantly
greater number of verifications than OCD CC)
(P< 0.01–0.03) showing a significantly higher
level of verification than NV (P< 0.01–0.03).
When the magnitude of increase in the number of
verifications from trials Df to Id was examined, a
significant difference was observed among all three
groups of subjects [F(2,97) = 7.07, P< 0.001]
(Fig. 2c). Newman–Keuls post hoc tests showed
that a significant difference was found between
OCD CC+ and NV (P< 0.01), whereas compar-
ison between OCD CC+ and OCD CC)
approached statistical significance (P= 0.06). By
contrast, there was no difference between OCD
CC)and NV (P> 0.31). Thus, OCD CC+
experienced a greater exacerbation of checking
behaviors, when comparison images identical (Id)
to the sample were displayed.
To examine the link between uncertainty and
checking behavior, we then analyzed the response
time for choice making during trials with and
without checking behaviors (CB+ and CB)
respectively). There was a significant difference
between all three groups [group effect,
F(2,87) = 10.80, P< 0.001] in a trial type-depen-
dent manner [group ·trial interaction,
F(2,87) = 6.71, P< 0.001] (Fig. 3a). When trials
CB+ were considered, a significant difference was
found between groups [group effect,
F(2,87) = 9.57, P< 0.001]. Newman–Keuls post
hoc tests indicated that OCD CC+ showed a
significantly higher response time for choice than
NV (P< 0.002), whereas OCD CC)did not
(P> 0.11). For trials CB), there was a significant
difference between groups [group effect,
F(2,87) = 7.18, P< 0.001]. Newman–Keuls post
hoc tests demonstrated that the response time for
choice in both OCD CC+ and OCD CC)was
longer than that of NV (OCD CC+ vs. NV:
P< 0.01; OCD CC)vs. NV: P< 0.02). When
the magnitude of increase in the response time
from trials CB)to CB+ was examined, a signif-
icant difference was observed between all three
groups of subjects [F(2,87) = 6.71, P< 0.001]
(Fig. 3b). Newman–Keuls post hoc tests showed a
significant difference between OCD CC+ and NV
(P< 0.01), whereas there was a tendency towards
a difference between OCD CC+ and OCD CC)
(P= 0.06). By contrast, OCD CC)did not differ
from NV (P> 0.33). Therefore, OCD CC+
showed a greater increase in the response time for
making their choice during trials CB+.
We then studied the course of both number of
verifications and response time for choice across all
Mean level of performance
(number of correct trials)
0
10
20
30
40
50
(a) (b) (c)
Groups
OCD CC+ OCD CC– NV
Mean increase from trials Df to Id
(number of verification)
0
2
4
6
8
OCD CC+ OCD CC– NV
Groups
Mean number of verification
0
5
10
15
20
25
Trials Id
Trials Df
OCD CC+ OCD CC– NV
Groups
Fig. 2. Mean (±SEM) response accuracy (expressed as the number of correct trials) (a), mean (±SEM) number of verifications in
trials where the second image presented was either identical (Id) to or different from the sample (Df) (b), mean (±SEM) increase in
the number of verifications from trials Df to Id (c) among the three groups of subjects, OCD CC+, OCD CC)and NV. No
significant difference was found in terms of performance on the checking task between the groups (a). Both OCD CC+ and OCD
CC)showed a significantly higher number of verifications than NV for both trials Id and Df (b). However, the magnitude of increase
in the level of verification from trials Df to Id was significantly greater only in OCD CC+ (c).
Checking behaviors in OCD
469
50 trials. OCD CC+ showed the expected ascend-
ing pattern of verification (Fig. 4a), which was
more pronounced than that of OCD CC)(Fig. 4b)
[OCD CC+: r= 0.53, P< 0.001; OCD CC):
r= 0.36, P< 0.01]. By contrast, checking in NV
did not increase in a particularly linear manner
from the first to the last trial compared to that
observed in both OCD CC+ and OCD CC).In
particular, the slope of the linear regression curve
was virtually flat [r= 0.21, P> 0.13], indicating a
relatively constant verification level during the
execution of the task (Fig. 4c). Statistical compar-
isons showed that OCD CC+ significantly differed
from NV in their checking profile over trials
(z= 1.66, P< 0.05), although no difference was
detected between OCD CC)and NV (z= 0.49,
P> 0.32). This was not paralleled by the time
response, which did not gradually increase
throughout the task across the three groups of
subjects [OCD CC+: r= 0.12, P> 0.41; OCD
CC):r=)0.09, P> 0.56; NV: r=)0.04,
P> 0.80].
Finally, we examined the relationships between
OCD symptom intensity and the aforementioned
behavioral parameters. Although there was no
significant correlation between individual Y-BOCS
scores and the number of verifications [Y-BOCS
total score: r
2
= 0.001, P> 0.86; obsessive sub-
scale score: r
2
= 0.004, P> 0.70; compulsive
subscale score: r
2
= 0.0001, P> 0.95], the
Y-BOCS scores were positively, but weakly corre-
lated with time response for choice during the task
[Y-BOCS total score: r
2
= 0.11, P< 0.01; obses-
sive subscale score: r
2
= 0.08, P< 0.05; compul-
sive subscale score: r
2
= 0.10, P< 0.03].
Therefore, uncertainty at the time of choice
appeared to increase as a function of OCD
severity.
0
2000
4000
6000
8000
(a) (b)
Groups
Trials CB+
Trials CB–
Mean response time for choice (ms)
NV
OCD CC–
OCD CC+
Mean increase from trials CB– to CB+
(response time for choice) (ms)
0
1000
2000
3000
4000
NV
OCD CC–OCD CC+
Groups
Fig. 3. Mean (±SEM) response time
for choice (determined as the time
between presentation of comparison
image and button pressing for making
choice) during trials leading to the
production of checking behaviors
(CB+) or not (CB)) (a), mean
(±SEM) increase in response time for
choice from trials CB)to CB+ (b) in
the three groups of subjects, OCD
CC+, OCD CC)and NV. OCD CC+
exhibited significantly longer response
time for choice, particularly when trials
CB+ were performed (a). Moreover,
the magnitude of increase in response
time from trials CB)to CB+ was
significantly greater in OCD CC+ (b).
50
1
Trials 1–50
Mean number of verification / trial
1.4
(a) (b) (c)
1.2
1
0.8
0.2
0.6
0.4
0
r = 0.53
501
Trials 1–50
1.4
1.2
1
0.8
0.2
0.6
0.4
0
r = 0.36
501
Trials 1–50
1.4
1.2
1
0.8
0.2
0.6
0.4
0
r = 0.21
Fig. 4. Course of the mean number of verifications over the 50 consecutive trials in the three groups of subjects, OCD CC+ (a),
OCD CC)(b) and NV (c). OCD CC+ demonstrated a gradual increase in their number of verifications throughout the task. This
ascending profile significantly differed from that of NV. OCD CC)showed an intermediate pattern of checking response that was not
significantly different from that of NV.
Rotge et al.
470
Discussion
In the present study, we highlighted distinct
behavioral patterns between patients with OCD
with and without checking compulsions (OCD
CC+ and OCD CC)respectively) and normal
healthy volunteers (NV) by using a delayed
matching-to-sample visuospatial task which
placed subjects in a potential situation of verifica-
tion. We found similar level of performance in both
OCD CC+ and OCD CC)and in NV. Checking
behaviors occurred more frequently in OCD CC+
in trials presenting comparison images identical to
the sample. There was even a progressive rise in
checking, as OCD CC+ successively performed
the 50 trials comprising the task. Similarly, OCD
CC+ took longer to make their choice before
checking. However, time to choose failed to
demonstrate the ascending profile observed for
checking throughout the task.
Phenomenologically, the central point in obses-
sional symptomatology is the sensation that Ôsome-
thing is wrongÕ(2, 3), with the internal perception
of error in response to specific environmental
stimuli. Compulsions are behavioral responses
performed to alleviate worry and anxiety produced
by highly distressing doubt (8–10). Our findings
support and extend these considerations, as dem-
onstrated by the link between uncertainty and
compulsive checking. The presentation of compar-
ison images identical to the sample elicited a more
pronounced increase in verification in OCD CC+.
It is likely that this experimental condition gener-
ated high uncertainty and doubt that persisted
once the choice was made, and then caused the
production of checking behaviors. Response time
for choice making, which is presumably indicative
of the degree of uncertainty and doubt at the
moment of choice, was longer in OCD CC+,
especially during trials leading to repetitive check-
ing behaviors. Interestingly, this behavioral pattern
mainly affected OCD CC+ subjects who displayed
a higher level of verification during the task.
However, the course of the response time for
choice was not superimposed on that of checking
which progressively increased throughout the task
in OCD CC+. This leads to postulate that
verification may be an adaptive behavioral strategy
intending to refrain the possible increase in doubt
throughout the task. In this context, repeated
verification may be expected as adopting an
ascending profile to sufficiently limit discomfort
caused by doubt. Such clinical considerations may
explain the apparent discrepancy concerning the
course of these two behavioral parameters
throughout the task.
Recent studies have provided a classification of
patients with OCD according to their clinical
phenotype (16–20). The symptom subgroups iden-
tified have been found to be linked to distinct
patterns of brain activation on the basis of func-
tional neuroimaging data (21, 22). Findings from
these studies indicate excessive activity in dorsal
cortical areas (dorsolateral prefrontal cortex and
dorsal anterior cingulate cortex) upon presentation
of provocative checking-related pictures. These
brain regions are particularly involved in a wide
range of cognitive functions, such as error recogni-
tion, conflict detection and response selection
(23–26). Consistent with our work suggesting a
link between uncertainty and checking activity,
these findings demonstrated the importance of
excessive monitoring for the occurrence of error-
generating pathological doubt. In this respect, a
critical consideration was to examine the possibility
of an altered error detection process in OCD.
Overactivation in the dorsal anterior cingulate
cortex of patients with OCD was found when
errors were made not only during a high-conflict
non-routine task (exaggerated error signals), but
also when trials were correctly completed (false error
signals) (27, 28). This internal perception of false
error producing uncertainty and doubt may explain
our results, showing that the performance of both
OCD CC+ and OCD CC)was not altered.
Finally, there was relatively unconvincing corre-
lation between the clinical severity of OCD and the
response time for choice. The weakness of this
relationship along with the absence of clear
evidence for a link between OCD severity and the
number of verifications may be because of the use
of Y-BOCS in our study which rated the severity of
a large variety of symptom types instead of
focusing on the examination of pathological
doubt or compulsive checking. The development
of specific tools for assessing the degree of uncer-
tainty at the time of choice and intensity of
ongoing checking behaviors would be of particular
importance in future clinical research, especially
for validation of our interpretation concerning the
response time for choice as a measurement of the
doubt level.
The present study has some limitations. First,
the geometrical complexity of the images (ranging
from abstract to figurative) was not appreciated.
Subsequent research should classify each image
according to its capacity to induce checking.
Second, we did not estimate the test–retest reli-
ability of behavioral responses which would have
been useful for assessing their stability over time.
Third, our findings raise the question of whether
checking activity is increased in patients with OCD
Checking behaviors in OCD
471
intending to compensate a possible impairment in
visuospatial ability, thereby maintaining a normal
level of performance at the behavioral task. How-
ever, deficits related to spatial working memory
and spatial recognition have not been consistently
found in patients with OCD (29–31). Their perfor-
mance accuracy was even reported as higher than
that of controls on a delayed matching-to-sample
task (32). Despite having comparable levels of
response accuracy, both of our subgroups of
patients with OCD differed in their checking
pattern. Therefore, the checking process might be
something other than just a compensatory
mechanism in the phase of spatial working
memory in the present task. Fourth, only OCD
subjects with remitted and uncomplicated mood
disorders were included in our study, thereby
minimizing the potential influence of this factor
on cognitive functions. Finally, it may be prema-
ture to rule out the possibly deleterious effects of
current medications on neuropsychological perfor-
mance. Nonetheless, the power of the study was
inadequate to examine each drug separately.
In conclusion, this study using a delayed match-
ing-to-sample task and proposing the option of
choice checking provides an objective quantifica-
tion of checking in patients with OCD. For future
psychophysiological research, such an experimen-
tal approach would be helpful for understanding
the cognitive and emotional determinants of
checking and for identifying its anatomical sub-
strates.
Acknowledgements
The authors thank Dr Ray Cooke for providing advice on the
English language. This work was supported in part by a grant
from the ÔAgence Nationale de la RechercheÕ. JY Rotge was
supported by a PhD grant from the ÔFondation pour la
Recherche Me
´dicaleÕ.
Declaration of interest
All authors reported no biomedical financial interests or
potential conflicts of interest.
References
1. Rasmussen SA,Eisen JL. The epidemiology and differential
diagnosis of obsessive compulsive disorder. J Clin Psychi-
atry 1992;53:4–10.
2. Schwartz JM. Neuroanatomical aspects of cognitive-
behavioural therapy response in obsessive–compulsive
disorder. An evolving perspective on brain and behaviour.
Br J Psychiatry Suppl 1998;35:38–44.
3. Schwartz JM. A role of volition and attention in the gen-
eration of new brain circuitry. Toward a neurobiology of
mental force. J Consciousness Stud 1999;6:115–142.
4. Salkovskis PM. Obsessional–compulsive problems: a cog-
nitive-behavioural analysis. Behav Res Ther 1985;23:571–
583.
5. Salkovskis PM. Understanding and treating obsessive–
compulsive disorder. Behav Res Ther 1999;37:29–52.
6. Foa EB,Franklin ME,Kozac MJ. Psychosocial treatments
for obsessive–compulsive disorder. Literature review. In:
Swinson RP,Antony MM,Rachman S,Richter MA, eds.
Obsessive–compulsive disorder. New York: Theory, Re-
search, and Treatment, Guilford Press, 1998:258–276.
7. Tolin DF,Abramowitz JS,Brigidi BD,Foa EB. Intolerance
of uncertainty in obsessive–compulsive disorder. J Anxiety
Disord 2003;17:233–242.
8. Aouizerate B,Guehl D,Cuny E et al. Pathophysiology of
obsessive–compulsive disorder: a necessary link between
phenomenology, neuropsychology, imagery and physiol-
ogy. Prog Neurobiol 2004;72:195–221.
9. Aouizerate B,Guehl D,Cuny E et al. Updated overview of
the putative role of the serotoninergic system in obsessive–
compulsive disorder. Neuropsychiatr Dis Treat
2005;1:231–243.
10. Aouizerate B,Rotge JY,Martin-guehl C et al. Neurosur-
gical treatments of obsessive–compulsive disorder. Clin
Neuropsychiatry 2006;3:391–403.
11. Rasmussen SA,Eisen JL. Epidemiology of obsessive com-
pulsive disorder. J Clin Psychiatry 1990;51:10–13.
12. Sheehan DV,Lecrubier Y,Sheehan KH et al. The Mini-
International Neuropsychiatric Interview (M.I.N.I.): the
development and validation of a structured diagnostic
psychiatric interview for DSM-IV and ICD-10. J Clin
Psychiatry 1998;59:22–33.
13. Goodman WK,Price LH,Rasmussen SA et al. The
Yale–Brown Obsessive Compulsive Scale. I. Develop-
ment, use, and reliability. Arch Gen Psychiatry 1989;
46:1006–1011.
14. Goodman WD,Price LH,Rasmussen SA et al. The Yale–
Brown Obsessive Compulsive Scale. II. Validity. Arch Gen
Psychiatry 1989;46:1012–1016.
15. Pallanti S,Hollander E,Bienstock C et al. Treatment non-
response in OCD: methodological issues and operational
definitions. Int J Neuropsychopharmacol 2002;5:181–191.
16. Baer L. Factor analysis of symptom subtypes of obsessive
compulsive disorder and their relation to personality and
tic disorders. J Clin Psychiatry 1994;55:18–23.
17. Leckman JF,Grice DE,Boardman J et al. Symptoms of
obsessive–compulsive disorder. Am J Psychiatry 1997;
154:911–917.
18. Summerfeldt LJ,Richter MA,Antony MM,Swinson RP.
Symptom structure in obsessive–compulsive disorder: a
confirmatory factor-analytic study. Behav Res Ther
1999;37:297–311.
19. Mataix-cols D,Rauch SL,Manzo PA,Jenike MA,Baer L.
Use of factor-analyzed symptom dimensions to predict
outcome with serotonin reuptake inhibitors and placebo in
the treatment of obsessive–compulsive disorder. Am J
Psychiatry 1999;156:1409–1416.
20. Mataix-cols D,Rauch SL,Baer L et al. Symptom stability
in adult obsessive–compulsive disorder: data from a nat-
uralistic two-year follow-up study. Am J Psychiatry
2002;159:263–268.
21. Mataix-cols D,Cullen S,Lange K et al. Neural correlates
of anxiety associated with obsessive–compulsive symptom
dimensions in normal volunteers. Biol Psychiatry
2003;53:482–493.
22. Mataix-cols D,Wooderson S,Lawrence N,Brammer MJ,
Speckens A,Phillips ML. Distinct neural correlates of
Rotge et al.
472
washing, checking, and hoarding symptom dimensions in
obsessive–compulsive disorder. Arch Gen Psychiatry 2004;
61:564–576.
23. Devinsky O,Morrell MJ,Vogt BA. Contributions of
anterior cingulate cortex to behaviour. Brain 1995;
118:279–306.
24. Bush G,Luu P,Posner MI. Cognitive and emotional influ-
ences in anterior cingulate cortex. Trends Cogn Sci 2000;
4:215–222.
25. Ito S,Stuphorn V,Brown JW,Schall JD. Performance
monitoring by the anterior cingulate cortex during saccade
countermanding. Science 2003;302:120–122.
26. Kerns JG,Cohen JD,Macdonald AW 3rd,Cho RY,Stenger
VA,Carter CS. Anterior cingulate conflict monitoring and
adjustments in control. Science 2004;303:1023–1026.
27. Fitzgerald KD,Welsh RC,Gehring WJ et al. Error-related
hyperactivity of the anterior cingulate cortex in obsessive–
compulsive disorder. Biol Psychiatry 2005;57:287–294.
28. Maltby N,Tolin DF,Worhunsky P,OÕkeefe TM,Kiehl KA.
Dysfunctional action monitoring hyperactivates frontal-
striatal circuits in obsessive–compulsive disorder: an event-
related fMRI study. Neuroimage 2005;24:495–503.
29. Head D,Bolton D,Hymas N. Deficit in cognitive shifting
ability in patients with obsessive–compulsive disorder. Biol
Psychiatry 1989;25:929–937.
30. Purcell R,Maruff P,Kyrios M,Pantelis C. Cognitive def-
icits in obsessive–compulsive disorder on tests of frontal-
striatal function. Biol Psychiatry 1998;43:348–357.
31. Van der wee NJ,Ramsey NF,Jansma JM et al. Spatial
working memory deficits in obsessive compulsive disorder
are associated with excessive engagement of the medial
frontal cortex. Neuroimage 2003;20:2271–2280.
32. Ciesielski KT,Hamalainen MS,Lesnik PG,Geller DA,
Ahlfors SP. Increased MEG activation in OCD reflects a
compensatory mechanism specific to the phase of a visual
working memory task. Neuroimage 2005;24:1180–1191.
Checking behaviors in OCD
473
