Molecular Psychiatry (2006) 11, 495504

& 2006 Nature Publishing Group All rights reserved 1359-4184/06 $30.00
www.nature.com/mp

ORIGINAL ARTICLE

The Dimensional YaleBrown ObsessiveCompulsive

Scale (DY-BOCS): an instrument for assessing
obsessivecompulsive symptom dimensions
MC Rosario-Campos1,2, EC Miguel1, S Quatrano2, P Chacon1, Y Ferrao1, D Findley2, L Katsovich2,
L Scahill2, RA King2, SR Woody3, D Tolin4, E Hollander5, Y Kano6 and JF Leckman2
1

Department of Psychiatry, University of Sao Paulo Medical School, Sao Paulo, Brazil; 2Child Study Center, Yale University
School of Medicine, New Haven, CT, USA; 3Department of Psychology, University of British Columbia, Vancouver, BC, Canada;
4
Institute of Living, Hartford, CT, USA; 5Department of Psychiatry, Mt Sinai School of Medicine, New York, NY, USA and
6
Graduate School of Medical Sciences, Kitasato University, Kanagawa, Japan
Obsessivecompulsive disorder (OCD) encompasses a broad range of symptoms representing
multiple domains. This complex phenotype can be summarized using a few consistent and
temporally stable symptom dimensions. The objective of this study was to assess the
psychometric properties of the Dimensional YaleBrown ObsessiveCompulsive Scale
(DY-BOCS). This scale measures the presence and severity of obsessivecompulsive (OC)
symptoms within six distinct dimensions that combine thematically related obsessions and
compulsions. The DY-BOCS includes portions to be used as a self-report instrument and
portions to be used by expert raters, including global ratings of OC symptom severity and
overall impairment. We assessed 137 patients with a Diagnostic and Statistical Manual-IV
diagnosis of OCD, aged 669 years, from sites in the USA, Canada and Brazil. Estimates of the
reliability and validity of both the expert and self-report versions of the DY-BOCS were
calculated and stratified according to age (pediatric vs. adult subjects). The internal
consistency of each of the six symptom dimensions and the global severity score were
excellent. The inter-rater agreement was also excellent for all component scores. Self-report
and expert ratings were highly intercorrelated. The global DY-BOCS score was highly
correlated with the total YaleBrown ObsessiveCompulsive Scale score (Pearson r = 0.82,
P < 0.0001). Severity scores for individual symptom dimensions were largely independent of
one another, only modestly correlated with the global ratings, and were also differentially
related to ratings of depression, anxiety and tic severity. No major differences were observed
when the results were stratified by age. These results indicate that the DY-BOCS is a reliable
and valid instrument for assessing multiple aspects of OCD symptom severity in natural
history, neuroimaging, treatment response and genetic studies when administered by expert
clinicians or their highly trained staff.
Molecular Psychiatry (2006) 11, 495504. doi:10.1038/sj.mp.4001798; published online 24 January 2006
Keywords: obsessivecompulsive disorder; obsessivecompulsive symptom dimensions; scale;
phenotypic heterogeneity

Introduction
Obsessivecompulsive disorder (OCD) has a lifetime
prevalence of 23%,1 affecting all age groups, across
different cultures.2 Obsessivecompulsive disorder is
considered to be the fourth most common mental
disorder and is frequently accompanied by family,
social, school and work dysfunctions.3 The World
Health Organization (WHO) estimated OCD to be
among the top 10 causes of years lived with illnessCorrespondence: Dr JF Leckman, Child Study Center, Yale
University School of Medicine, I-265 SHM, 230 South Frontage
Road, New Haven, CT 06520-7920, USA.
E-mail: james.leckman@yale.edu
Received 19 May 2005; revised 18 November 2005; accepted 15
December 2005; published online 24 January 2006

related disability by 2020.4 Obsessivecompulsive

disorder is characterized by intrusive unwanted
thoughts, fears or images (obsessions) and/or ritualized behaviors or mental acts (compulsions), generally performed to relieve the anxiety and/or distress
caused by the obsessions.
Current classification manuals, such as the DSM-IV
and the ICD-10 (WHO), regard OCD as a unitary
nosological entity. While this approach has added
specificity to research studies, it may be misleading.
In fact, obsessivecompulsive (OC) symptoms are
remarkably heterogeneous.
In addition to the clinical diversity seen in OCD,
genetic and treatment studies also support the view
that OCD is a heterogeneous disorder. This heterogeneity obscures the findings from natural history

Dimensional YaleBrown ObsessiveCompulsive Scale

MC Rosario-Campos et al

496

studies, neuroimaging and clinical trials, and complicates the search for vulnerability genes.5 Therefore,
many investigators have emphasized the need for
identifying more homogenous subgroups in order to
better understand the OCD complexity. Examples
include dividing patients according to the age of
onset of OC symptoms68 and the presence of specific
clinical features, including subjective experiences
preceding OC symptoms,911 comorbidity patterns,12
or clinical course.13 Although these subtyping strategies have been useful, they have had limited success
in identifying predictors of treatment response,
endophenotypic markers, or vulnerability genes.
More recently, dimensional strategies have been
introduced to more accurately characterize individual
differences among OCD patients. Similar efforts have
been undertaken with other disorders, including
schizophrenia,14,15 bipolar disorder,16 Tourettes syndrome,17 eating disorders,18 and learning disabilities.19
In OCD, there have been at least 12 factor-analytic
studies published, involving more than 2000 patients.20 These studies have consistently identified
35 symptom factors or dimensions, accounting for
nearly 70% of the variance.20 A similar factor
structure has been observed in both adults2224 and
children (Evelyn Stewart, personal communication,
2005) and in mixed samples.21 These factors are
temporally stable,25,26 and correlate meaningfully
with various genetic2730 and neuroimaging variables31,32 as well as treatment response.22,3336
Obsessivecompulsive disorder genetic studies
have reported that the use of a dimensional approach
may provide a powerful approach to detect the
genetic susceptibility loci that contribute to the
heterogeneous OCD presentations.5 For instance,
Alsobrook et al.26 found that the relatives of OCD
probands who had high scores on the obsessions/
checking and symmetry/ordering factors were at
greater risk for OCD than were relatives of probands
who had low scores on those factors. Using similar
methods, Leckman et al.28 found that the obsessions/
checking and symmetry/ordering factors were significantly correlated in sib-pairs concordant for Tourette
Syndrome (TS). Using the same data set, Zhang et al.29
observed significant allele sharing for the hoarding
factor for loci at 4q34, 5q35.2 and 17q25. Similarly,
Cavallini et al.30 performed a candidate gene study
with a functional polymorphism in the promoter
region of the serotonin transporter locus at 17q11.
They found a significant association of the long/long
haplotype in patients with tics and high scores on the
repeating/counting factor.
Despite the potential usefulness of using dimensional ratings and the availability of psychometrically
sound self-report measures,37,38 there are currently no
severity scales designed for use by expert raters
capable of properly assessing the dimension-specific
OC symptom severity. In an effort to address this
need, we have developed a new instrument the
Dimensional YaleBrown ObsessiveCompulsive

Molecular Psychiatry

Scale (DY-BOCS). Practically, by dividing symptoms by dimension, it is possible to inquire about

symptom types that are inherently ambiguous. For
example, checking compulsions are now asked
about in several of the domains checking related
to sexual and religious obsessions vs checking related
to contamination worries. Another important innovation is the decision to focus on frequency, distress
and interference as the best estimates of symptom
severity within each dimension, as well as for
all OC symptoms considered in aggregate.39,40 After
an extensive period of refinement and preliminary
testing in both pediatric and adult populations, we
conducted a multinational, multi-site cross-sectional
study in order to determine the psychometric properties of the DY-BOCS. The results of these efforts are
reported below. A secondary aim was to explore
potential differences between pediatric and adult
subjects.

Materials and methods

Subjects
One hundred and thirty-seven outpatient subjects,
aged 669 years were assessed. Fifty-nine subjects
were under 18 years of age and 78 were 18 years of age
or older. Patients were recruited from four different
sites: the Child Study Center at Yale University, USA
(N = 85); the OCD Spectrum Disorders Project at the
OCD Spectrum Disorders Project at Getulio Vargas
Medical Hospital, Porto Alegre, Brazil (N = 39), the
Institute of Living at Hartford Hospital, USA (N = 10)
and the University of British Columbia, Canada
(N = 3). Subjects were either being followed for OCD
treatment or had been followed in the past at one of
the sites. Inclusion criteria required a DSM-IV
Diagnostic and Statistical Manual-IV (American Psychiatric Association) diagnosis of OCD. Exclusion
criteria included an intelligence quotient (IQ) < 75,
head trauma resulting in loss of consciousness and
current psychiatric disorder that could interfere with
an individuals ability to report accurately on their OC
symptoms, such as psychosis or pervasive developmental disorders.
The project was approved by institutional review
boards at each of the sites. After a thorough description of the study and the assurance that their decision
to participate would not interfere with their clinical
treatment, all patients were asked to sign an informed
consent document. A separate assent form was used
to ensure the informed participation of the child and
adolescent subjects.
Instruments
The DY-BOCS was initially developed by three
individuals then at Yale University (JFL, SW and
MCR-C). The instrument was then refined by a larger
group of investigators from the United States (RAK,
DF and LS), Brazil (EM) and Japan (YK). The DYBOCS is based on the YaleBrown ObsessiveCompulsive Scale (Y-BOCS) created by Goodman et al.,41,42

Dimensional YaleBrown ObsessiveCompulsive Scale

MC Rosario-Campos et al

the Schedule for Tourettes and Other Behavior

Syndrome (STOBS) developed by Pauls and Hurst,43
as well as the results of earlier factor analyses2124 and
the DSM-IV field trial for OCD.44
The DY-BOCS consists of semi-structured scales for
assessing the presence and severity of OC symptom
dimensions. The DY-BOCS includes a self-report
instrument and an instrument to be used by expert
raters. The DY-BOCS self-report is composed of an
88-item self-report checklist, designed to provide a
detailed description of obsessions and compulsions
that are divided into six different OC symptom
dimensions: (1) obsessions about harm due to aggression/injury/violence/natural disasters and related
compulsions; (2) obsessions concerning sexual/
moral/religious obsessions and related compulsions;
(3) obsessions about symmetry/just-right perceptions, and compulsions to count or order/arrange;
(4) contamination obsessions and cleaning compulsions; (5) obsessions and compulsions related to
hoarding and (6) miscellaneous obsessions and
compulsions that relate to somatic concerns and
superstitions, among other symptoms.
By dividing OC symptoms according to these
dimensions, the DY-BOCS is capable of inquiring
about symptoms that are inherently ambiguous (such
as checking, mental rituals, repetition and avoidance
behaviors) and that may be present in more than one
symptom domain. For example, checking compulsions could be related to obsessional worries about
harm coming to a close family member or to ensure
that an item is clean or just right. Similarly, mental
rituals, avoidance and repetitive behaviors could be
related to one or more symptom domains depending
on their content. Patients are asked to endorse both
lifetime and current symptoms, which are then
reviewed by the clinician in order to increase the
reliability of the information, and to ensure that the
endorsed items are truly OC symptoms.
In addition to the symptom checklist, the DY-BOCS
self-report also includes items that ask the patient to
assess the overall symptom severity in each of the
dimensions for the previous week, on scales ranging
from 0 (no symptoms) to 10 (symptoms are extremely
troublesome). Patients are also asked explicitly about
avoidance behaviors accompanying their OC symptoms within each dimension.
Expert ratings of symptom severity involve reviewing the symptoms endorsed in each dimension.
Avoidance was included as an integral part of each
of these severity ratings based on the findings of
Woody et al.40 Severity for each dimension is measured on three ordinal scales with six anchor points
that focus on symptom frequency (05), the amount of
distress they cause (05) and the degree to which they
interfered with functioning (05) during the previous
week.
In addition to the symptom severity ratings for
each dimension, the expert raters are asked to
estimate the global OC symptom severity using the
same three ordinal scales. Finally, the expert raters

are asked to assess an individuals overall level of

current impairment due to OC symptoms on a scale
that ranges from none (0 points) to severe
(15 points). The total global score is obtained by
combining the sum of the global severity scores for
frequency, distress and interference and the impairment score (015), yielding a maximum total global
severity score of 30.
The DY-BOCS was piloted over a 6-month period
in the OCD/Tic Disorders Clinic of the Child
Study Center at Yale with adult and child cases.
During this period refinements were made in the
wording of the checklist items, in the descriptive anchor points for the impairment scale and the
instructions manual describing the rationale for
administering the DY-BOCS. Copies of the Self-report
and Clinician Rated versions of the DYBOCS are
available online as supplementary materials at the
journal website.
The Y-BOCS41,42 and the Childrens YaleBrown
ObsessiveCompulsive Scale (CY-BOCS)45 were used
for measurement of convergent validity. These are
clinician-rated, semi-structured interview-based scales
that are widely used to assess OC symptom severity.
Considerable data support the reliability and validity
of both the adult and child versions of this scale. They
yield an obsession score (maximum = 20), a compulsion score (maximum = 20) and a combined total score
(maximum = 40).
The Yale Global Tic Severity Scale (YGTSS),46 the
Childrens Depression Rating Scale (CDRS),47 the
Hamilton Scale for Depression (Ham-D),48 and the
Hamilton Scale for Anxiety (Ham-A)49 were used to
assess the divergent validity of the component scales
of the DY-BOCS.
The YGTSS is a clinician-rated, semi-structured
scale with excellent psychometric properties developed to evaluate the presence and severity of
tics, and the impairment caused by them. The motor
(maximum = 25) is added to the phonic (maximum = 25), and the impairment scores (maximum = 50), thus yielding a maximum total score of
100. The YGTSS has been widely used to assess tic
severity in both pediatric and adult populations.
The CDRS is a clinician-rated, semi-structured
interview for assessing current depressive symptoms
in children. The CDRS yields a total maximum score
of 113, and has been shown to possess good reliability
and validity. A score of 40 is considered to be a
reliable indicator of clinically significant depression.
The Ham-D is a clinician-rated, semi-structured
interview for assessing current depressive symptoms
in adults. The Ham-D yields a total score of 63, and
possesses good psychometric properties. A score of
> 14 is considered to be a reliable indicator of
moderate to severe depression.
The HAM-A is a clinician-rated, semi-structured,
valid and reliable interview for assessing current
anxiety symptoms. It yields a total maximum score of
56. A score of > 14 has been suggested to indicate
clinically significant anxiety.

497

Molecular Psychiatry

Dimensional YaleBrown ObsessiveCompulsive Scale

MC Rosario-Campos et al

498

Procedures
Interviews
Informants included a parent and the child in the case
of younger subjects. Adolescents had the choice to be
interviewed individually and additional information was sought from parents as needed. Adults were
interviewed individually. Each subject was evaluated
separately by two different interviewers. Before
starting the interview the order of the interviews
was decided, according to a coin toss. The first rater
reviewed the consent forms, and asked the subjects
to sign them. If the coin toss was heads, the first
interviewer reviewed the DY-BOCS self-report and
then completed the component ratings of the
DY-BOCS. If the coin toss was tails, the first
interviewer reviewed the Y-BOCS/CY-BOCS self-report
and then completed the Y-BOCS/CY-BOCS scores. The
first interviewer then completed the other clinician
ratings, including the HAM-D or the CDRS, the HAMA and the YGTSS. After completing the first interview,
the subjects were introduced to the second rater and
invited to have a 15-min break. After the break, the
second interviewer completed either the DY-BOCS or
Y-BOCS/CY-BOCS ratings with them, depending upon
which scale still needed to be completed.
Training of the interviewers
In order to improve reliability across sites, all
interviewers were trained on the DY-BOCS by
MCR-C, an experienced clinician. All interviewers
had at least a bachelors degree and were trained to
reliability in the other instruments.
Translation of the instruments
The translation of the DY-BOCS into Portuguese was
made by MCR-C, a fluent speaker of both languages.
Afterwards, a certified translator was hired to do the
back-translation into English. The two versions were
then compared and resolved. All other instruments
had previously been translated into Portuguese.
Statistical analyses
Before pooling the data, possible differences between
the North-American and Brazilian samples were
investigated. Demographic and clinical characteristics in these two groups were compared using
chi-square or Fishers exact tests for categorical data
and t-test for continuous data. All tests were twotailed, with a = 0.05.
The internal consistency of the DY-BOCS was
determined by using Cronbachs alpha to assess the
three severity items (time, distress and interference)
in each of the dimensions. Inter-rater reliability was
determined for a subset of 29 (21.2%) patients drawn
mostly from the Yale and Brazilian sites. These
ratings were done with two raters sitting in the same
room, with no communication between the two
during the interviews. Pearson correlation coefficients were used to evaluate the association between
the patient self-report and the clinician measures of
severity for each dimension on 98 patients. Thirty-

Molecular Psychiatry

nine patients were excluded form this analysis

because they did not fill out the questions assessing
symptoms severity in the self-report.
Pearson correlation coefficients were used to assess
the convergent and divergent validities of the component scores of the DY-BOCS, with the other clinical
instruments described above. Fishers Z transformation was used to compare the correlations within the
pediatric age group matrix with those correlations
obtained from the adult subject matrix. A test statistic
and P-value were computed for each pair of correlations testing the hypothesis that the sample correlations estimate the same population correlation value.

Results
The 137 subjects (75 males and 62 females) had a
mean age of 25.1 years (s.d. = 14.3, range 669), with a
median age of 21. Children and adolescents were
interviewed at two sites in Connecticut (N = 69), with
a mean age of 12.7 (s.d. = 2.9, range 617) and a
median age of 12. When considering only adult
subjects (N = 78), the mean age was 34.5 (s.d. = 12.1,
range 1869).
There were no statistically significant differences
between the Brazilian and the North-American sites
with regard to gender (North-American sites: 57 male
and 41 female subjects; Brazilian site: 18 male and 21
female subjects; w2(1) = 1.6, NS) or ethnicity. Nearly
96% (N = 131) of the subjects were Caucasian. Three
patients from the North-American sites were Hispanic
(2.2%) and one patient was from India (0.7%). All
Brazilian subjects were Caucasian. The mean scores
on each of the DY-BOCS dimensions, on the Y-BOCS/
CY-BOCS and on the other scales are presented in
Table 1. The pediatric and adult samples were largely
indistinguishable, with the adult sample being
slightly more severe on the DY-BOCS Global Score
(t135 = 2.2, P = 0.04, two-tailed), DY-BOCS Global
Impairment rating (t135 = 2.7, P = 0.007, two-tailed),
and Sexual and Religious Dimension (t135 = 2.04,
P = 0.04, two-tailed).
Table 2 presents the number of patients with
symptoms in each of the dimensions for both the
pediatric and adult groups. As a group, the pediatric
sample had a greater frequency of symptoms in the
Symmetry dimension and less within the Aggression
dimension, but these differences in frequency were
not statistically significant. However, adult patients
had significantly higher frequency of symptoms in the
Sexual and religious dimension (w2(1) = 4.1, P = 0.04).
The age of onset of OC symptoms ranged from 2 to 46
years, with a mean of 10.5 years (s.d. = 6.7) and a
mode of 10.
When dividing the sample into adults and children,
the two groups exhibit very similar rates of avoidance
behaviors (78 vs 80%, respectively). This was also
true when each dimension was considered separately:
Aggression dimension (49% of adults and 32% of
children); Sexual/religious dimension (26% of adults
and 25% of children); Contamination dimension

Dimensional YaleBrown ObsessiveCompulsive Scale

MC Rosario-Campos et al

Table 1

499

Mean scores on the DY-BOCS, the Y-BOCS and measures of tic, depression and anxiety severity

Pediatric age group

Adults

Subjects (N)

Mean (s.d.)

Range

Subjects (N)

Mean (s.d.)

Range

DY-BOCS global score

59

17.8 (4.8)*

728

78

19.6 (4.8)*

729

DY-BOCS dimensions
Aggression
Sexual/religious
Symmetry
Contamination
Hoarding
Miscellaneous

59
59
59
59
59
59

4.2
2.3
6.7
5.5
2.3
6.2

(4.3)
(3.6)*
(3.9)
(5.0)
(3.5)
(3.8)

013
011
014
015
011
014

78
78
78
78
78
78

DY-BOCS impairment
Y-BOCS total score
Y-BOCS obsessions
Y-BOCS compulsions
Yale Global Tic Severity Scale
Children Depression Rating Scale Revised
Hamilton Depression Scale
Hamilton Anxiety Scale

59
59
59
59
26
50

7.5
23.0
11.2
11.8
15.7
27.7

(2.9)w
(6.3)
(3.7)
(3.5)
(9.3)
(11.1)

213
835
018
318
234
1773

78
78
78
78
17

73
72

5.3
3.7
6.7
6.1
2.5
6.6

(4.2)
(4.2)*
(4.6)
(4.8)
(3.7)
(4.5)

014
012
015
015
015
015

8.9 (2.8)w
24.2 (6.9)
12.2 (3.8)
12.1 (3.8)
20.9 (11.7)

10.5 (6.5)
11.6 (8.2)

314
538
019
119
338

028
040

DY-BOCS = Dimensional YaleBrown ObsessiveCompulsive Scale; Y-BOCS = YaleBrown ObsessiveCompulsive Scale;

*P < 0.05.
w
P < 0.01.

Table 2 Number of subjects presenting lifetime obsessive

compulsive symptoms in each of the dimensions
Dimension
Aggression
Sexual/religious
Symmetry
Contamination
Hoarding
Miscellaneous

Pediatric group
N (%)
33
20
51
36
22
50

(56)
(34)*
(86)
(61)
(37)
(85)

Adult group
N (%)
55
40
63
54
32
64

(71)
(51)*
(81)
(69)
(41)
(82)

*P < 0.05, w2 test.

(56% of adults and 54% of children); Symmetry

dimension (59% of adults and 52% of children);
Hoarding dimension (29% of adults and 15% of children) and in the Miscellaneous dimension (67% of
adults and 54% of children).
Reliability
Interrater reliability. The interrater reliability between the expert raters on the DY-BOCS was excellent. Intraclass correlation coefficients (ICCs) were
> 0.98 for each component score of the DY-BOCS.
Level of agreement between self-report and expert
ratings. The correlations between the self-report
and clinician measures of severity were highly significant. Pearson correlation coefficients were 0.87 for
the Aggressive, 0.86 for the Sexual/religious, 0.78 for

the Symmetry, 0.83 for the Contamination/washing,

0.86 for the Hoarding and 0.75 for the Miscellaneous
dimensions (P < 0.0001 for all dimensions).
When considering adults and children separately,
the levels of agreement between the self-report and
expert ratings are very similar to those of the entire
sample. Pearson correlation coefficients were 0.84
in adults and 0.90 in children for the Aggressive
dimension; 0.83 and 0.89 (for adults and children,
respectively) for the Sexual/religious dimension; 0.81
and 0.75 (for adults and children, respectively) for the
Symmetry dimension; 0.91 and 0.76 (for adults and
children, respectively) for the Contamination dimension; 0.87 and 0.85 (for adults and children, respectively) for the Hoarding dimension; and 0.82 and 0.67
(for adults and children, respectively) for the Miscellaneous dimension. All of the Pearson correlations
for both adults and children were highly significant
(P < 0.0001). There was a significant difference in the
correlations for the Contamination dimension between the adult and pediatric groups (Fishers Z = 2.48,
P < 0.05).
Internal consistency. The internal consistency across
the domains of time, distress and interference for each
dimension was excellent. Cronbachs alphas were
0.94 for Aggressive, 0.95 for Sexual/religious, 0.95 for
Symmetry, 0.96 for Contamination, 0.95 for Hoarding
and 0.94 for Miscellaneous dimensions.
In adult and child groups, Chronbachs alphas were
0.94 for the Aggressive dimension for both adults and
children; 0.95 for the Sexual/religious dimension for
both adults and children; 0.96 for adults and 0.93 for
Molecular Psychiatry

Dimensional YaleBrown ObsessiveCompulsive Scale

MC Rosario-Campos et al

500

Table 3 Correlations between the dimensional DY-BOCS

global score and the symptom severity ratings for each
dimension, the components of the global severity ratings
and the impairment rating
Pediatric group
Pearson r
Dimensions
Aggression
Sexual/
religious
Symmetry
Contamination/
cleaning
Hoarding
Miscellaneous
Global severity
Time
Distress
Interference
Impairment
rating

Adult group
Pearson r

0.24
0.03

NS
NS

0.28
0.22

0.02
NS

0.33
0.54

0.01
< 0.0001

0.46
0.46

< 0.0001
< 0.0001

0.43
0.39

0.001
0.002

0.20
0.40

NS
< 0.0001

0.76
0.85
0.83

< 0.0001
< 0.0001
< 0.0001

0.76
0.77
0.88

< 0.0001
< 0.0001
< 0.0001

0.94

< 0.0001

0.95

< 0.0001

DY-BOCS = Dimensional YaleBrown ObsessiveCompulsive Scale. There were no significant differences between
the groups in correlations for any dimensions or the
components of the global severity ratings or the impairment
rating (data not shown).

children for Symmetry; 0.97 for Contamination in

both adult and child groups; 096 for adults and 0.94
for children in the Hoarding dimension and 0.96 and
0.93 for the Miscellaneous dimension in adults and
children, respectively.
Validity
Construct validity. There were no statistically
significant differences between the Brazilian and the
North-American sites with regard to any of the scores
on these clinical ratings (data not shown). When
including only adult subjects (N = 78), there were also
no significant differences between the Brazilian and
the North-American sites with regard to the DY-BOCS
(t75 = 0.23, P = NS, two-tailed) or Y-BOCS (t75 = 0.18,
P = NS, two-tailed) total scores.
Correlations between each of the DY-BOCS dimensions and the total DY-BOCS scores are presented in
Table 3. With the exception of the severity ratings for
sexual and religious symptoms (both for children and
adults), aggressive obsessions and compulsions (children only) and hoarding symptoms (adults only),
each of the other severity ratings was positively
correlated with the DY-BOCS Global severity score,
with Pearson correlations ranging from 0.28 to 0.54.
Table 4 presents the intercorrelations between each
of the different OC symptom dimensions. With a few
exceptions, each dimension-specific severity rating
was largely independent of the others.

Table 4 Intercorrelations between estimates of current symptom severity within each of the dimensional components of the
DY-BOCSa
Dimensions

Sexual/religious

Aggressive

0.23
0.27 (P = 0.02)
0.27 (P = 0.001)

Sexual/religious

Symmetry

Contamination

Hoarding

Symmetry

Contamination

Hoarding

Miscellaneous

0.06
0.02
0.03

0.01
0.04
0.02

0.30 (P = 0.02)
0.02
0.12

0.12
0.25 (P = 0.03)
0.20 (P = 0.02)

0.24
0.04
0.11

0.25
0.13
0.18 (P = 0.03)

0.002
0.002
0.0004

0.02
0.27 (P = 0.02)
0.17 (P = 0.04)

0.07
0.15
0.12

0.28 (P = 0.03)
0.35 (P = 0.002)
0.32 (P = 0.0001)

0.27 (P = 0.04)
0.51 (P = 0.0001)
0.43 (P = 0.0001)

0.32 (P = 0.01)
0.06
0.11

0.15
0.003
0.06
0.34 (P = 0.008)
0.25 (P = 0.03)
0.28 (P = 0.001)

Correlations for the pediatric age group are listed first, followed by the adult and then the total group.
Numbers in bold are significant Pearson r correlations, two-tailed.
There were no significant differences in correlations for any of the dimensions between the children and adult groups (data
not shown). DY-BOCS = Dimensional YaleBrown ObsessiveCompulsive Scale
Molecular Psychiatry

Dimensional YaleBrown ObsessiveCompulsive Scale

MC Rosario-Campos et al

Convergent validity. The correlation between the

DY-BOCS total global score and the Y-BOCS total
score was excellent (Pearson r = 0.82, P < 0.0001),
indicating the convergent validity of the DY-BOCS
total global severity score. For the adult and child
groups separately, these correlations were also highly
significant (Pearson radult = 0.84, P < 0.0001, Pearson
rchildren = 0.79, P < 0.0001). Similarly, the correlation
between DY-BOCS impairment rating and the Y-BOCS
total score (Pearson r = 0.71, P < 0.0001) was also
robust across all patients as well as for the pediatric
and adult groups, when considered separately (radult =
0.76, P < 0.0001, Pearson rchildren = 0.65, P < 0.0001).
The correlations between impairment and the obsession (Pearson r = 0.58, P < 0.0001) and the compulsion
(Pearson r = 0.68, P < 0.0001) subscales of the Y-BOCS
were also significant. Among the adults, the correlations between impairment and the obsession and the
compulsion subscales of the Y-BOCS were also
significant (Pearson r = 0.62, P < 0.0001 and Pearson
r = 0.75, P < 0.0001). In the child group, these correlations were slightly lower than in the adult group,
but still significant (Pearson r = 0.51, P < 0.0001 for
obsessions and Pearson r = 0.62, P < 0.0001 for
compulsions).
Divergent validity. Divergent validity uses measurement approaches that differentiate psychometric
constructs from one another. The constructs that are
at the heart of the DY-BOCS are the severity ratings for
the dimensional constructs, not the DY-BOCS Global
Score. Evidence in support of the divergent validity of
the dimensional constructs include the orthogonal
character of the dimensional severity ratings (Table 4)
as well as the differential relationships between the
dimensional severity scores and measures of tic
severity, as well as clinician-rated estimates of mood
and anxiety severity (Table 5).
Of note is the fact that the correlation coefficients
varied according to the specific OC symptom dimensions, with the most robust relationships being seen
between the severity of anxiety and mood symptoms
and the severity of the OC symptoms in the
Aggressive dimension on one hand and the severity
of tics and the severity of Symmetry dimension on the
other. Comparable results were obtained when the
sample was restratified with regard to the presence of
clinical significant levels of anxiety and depression
(data not shown). However, when the sample was
restratified according to the presence of a chronic tic
disorder (N = 43) vs the absence of a tic disorder
(N = 89), the severity of the OC symptoms in the
Contamination dimension was found to be reduced in
the OCD patients with a history of tics (data not
shown).
Feasibility
On average, the self-report of the DY-BOCS usually
takes about 40 min to complete (mean = 41 min;
s.d. = 13.3 min; range 10120 min). On average, expert
raters take slightly longer to complete the DY-BOCS

Table 5 Correlations between the each of the DY-BOCS

dimensions and scores of the HAM-D, the HAM-A and the
YGTSSa
Dimensions

HAM-A

HAM-D

0.46
(P = 0.0001)

0.45
(P = 0.0001)

0.11

0.01

0.09

0.17

0.12

0.13

0.17

0.14

0.64
(P = 0.005)

0.21

0.14

0.31

501

YGTSS

Aggression
Sexual/
religious
Symmetry

0.14
0.18
0.19

Contamination
0.21

Hoarding
Miscellaneous

a
Correlations presented for just the adult age group (N = 78).
Numbers in bold are significant Pearson r correlations, twotailed. DY-BOCS = Dimensional YaleBrown Obsessive
Compulsive Scale; HAM-D = Hamilton Scale for Depression;
HAM-A = Hamilton Scale for Anxiety; YGTSS = Yale Global
Tic Severity Scale.

(mean = 49 min, s.d., 23.3 min, range 2065 min)

depending on the diversity of symptoms endorsed
by the patient. Subsequent ratings of the same patient
take substantially less time, as symptom validation is
less of an issue. There were no significant differences
in DY-BOCS or self-report completion time between
adult and pediatric groups (data not shown).

Discussion
A growing body of data indicates that a dimensional
approach to OC symptoms may have heuristic value
in genetic, neurobiological, comorbidity and treatment response studies.5,20 Some very compelling data
have also come from brain imaging studies, where it
appears that each OC symptom dimension reflects the
dysregulation of complex and partially overlapping
neural systems that serve to detect, appraise
and respond to potential threats that are contentspecific.5,20,31,32
A major limitation in adopting a dimensional
approach to OCD research had been the lack of
assessment tools capable of encompassing the dimensionality of OC symptoms.5,20 This work presents the
development and initial psychometric properties of a
new instrument designed to focus on OC symptom
dimensions. As hypothesized, our results suggest that
the DY-BOCS is a valid and reliable tool for assessing
OC symptom dimensions using either expert clinicians or relying solely on self-reports. This appears to
be true for both adult patients as well as children with
OCD. Indeed, a careful examination of potential
differences between children and adults with OCD
revealed relatively few. This suggests that regardless
Molecular Psychiatry

Dimensional YaleBrown ObsessiveCompulsive Scale

MC Rosario-Campos et al

502

of the age of the patient the psychometric structure of

OC symptoms is similar. This result is consistent with
the recent report (Evelyn Stewart, personal communication, 2005) in which the factor structure was
examined in 327 children and adolescents using the
Y-BOCS symptom checklist and found to be nearly
identical with that regularly seen in adult and mixed
samples.20 We hasten to add that this similarity does
not imply that there are no important differences
with regard to family history, comorbidity and sex
distribution between subjects with an early versus a
late age of onset.5,20
As expected, the OC symptom dimensions showed
only modest correlations with each other, suggesting
that they are largely independent of one another,
and from the estimates of global severity. The one
exception concerns the Miscellaneous dimension,
which was found to have significant correlations
with four of the five other dimensions in the adult
group and with two other dimensions in the pediatric
age group. Additional work is needed to sort through
these miscellaneous symptoms and determine which
symptoms are associated with which of the other
dimensions and which symptoms, if any, uniquely
define a true Miscellaneous category. Similar efforts
are currently underway by other groups of investigators.50,51
The DY-BOCS has excellent content validity. The
self-report checklist assesses the presence of the most
common OC symptoms, and includes items that are
usually underrepresented in other scales, such as
mental rituals, checking, repetition and avoidance
behaviors. Similarly, the severity scales evaluate the
time, distress and interference parameters described
in the DSM-IV diagnostic criteria for OCD.
The DY-BOCS is based in part on the Y-BOCS41,42
and the very similarly constructed CY-BOCS.45
Although the Y-BOCS and CY-BOCS remain the gold
standard for assessing OC symptom severity, the DYBOCS has some advantages over previously developed clinician- and self-rated instruments such as the
Leyton Obsessional Inventory,52 the Maudsley ObsessionalCompulsive Inventory,53 the National Institute
of Mental Health Global ObsessiveCompulsive
Scale,54 ObsessiveCompulsive Inventory37 and the
Vancouver Obsessional Compulsive Inventory.38 First,
the DY-BOCS self-report checklist has greater clarity
and offers examples after each symptom description.
This has made the instrument more patient-friendly,
and likely contributed to the high correlations between the self-report and clinician measures of
severity, ranging from 0.75 to 0.87, and suggests that
either version could be used alone. Second, the fact
that each dimension assesses the severity of both
thematically related obsessions and compulsions has
eliminated the bias in collecting information from
patients with either obsessions or compulsions only.
Third, the DY-BOCS does not include items about
resistance or control as part of the assessment of OC
symptom severity. This decision was based on
previous work suggesting that the resistance/control

Molecular Psychiatry

items in the Y-BOCS do not meaningfully contribute

to the measurement of OCD symptom severity.39
Recent data have reinforced this decision.40 Fourth,
the ordinal scales used for assessing the three severity
domains expanded from 5 to 6 anchor points in the
DY-BOCS. This expansion should allow investigators
to refine the assessment of subjects with subclinical
presentations of OCD.55 In the case of genetic studies,
family members presenting with OC symptoms below
the threshold for a DSM-IV diagnosis could be more
accurately assessed.5 In the general population, there
are periods in life, such as childbirth, when otherwise
normal individuals experience marked OC behaviors
and mental states, and when a diagnosis of OCD
would be inappropriate.56 These are all the unique
and distinguishing features of the DY-BOCS from
earlier instruments.
Another DY-BOCS innovation was the inclusion
of avoidance behaviors in each dimension of the
symptom checklist, and also as an integral part of the
severity ratings. This decision was reinforced by the
findings that more than 75% of our sample reported
some kind of avoidance behavior, and that the
frequencies of these avoidance symptoms varied
according to the specific dimensions. These findings
are in accordance with studies suggesting that
avoidance symptoms are commonly reported by
OCD patients and should be included in assessment
measures.39
The results concerning the reliability and validity
of this scale are promising. For example, the DYBOCS showed an excellent concordant validity with
both the Y-BOCS and the CY-BOCS, currently used as
the gold standards for assessing OCD severity. In
addition, the individual DY-BOCS dimensional
severity ratings showed excellent divergent validity
with anxiety, depression and tic severity scales. The
one exception is that the severity of the Aggression
dimension was related to the severity of mood and
anxiety symptoms. This result largely replicates in a
larger sample of OCD patients the finding by Hasler
et al.24 Similarly, the Symmetry dimension was
differentially associated with the YGTSS scores,
reinforcing its close relationship with the tic-related
phenotype.
Finally, the time burden of the DY-BOCS instruments needs to be emphasized, since the time of
administration may be a major limitation. Its initial
use in clinical settings may provide the therapist with
a valuable perspective on the dimensional landscape
of their patients OC symptoms. In research studies,
the advantages of being able to chart the course of
particular symptom dimensions may or may not
outweigh the time cost associated with their use.

Conclusions and future directions

Despite the time burden of these instruments, they
may have much to offer in clinical practice and
research studies. Specifically, the data suggest that
the DY-BOCS is a useful tool in assessing OC

Dimensional YaleBrown ObsessiveCompulsive Scale

MC Rosario-Campos et al

symptom dimensions, while continuing to provide

valid overall estimates of OC symptom severity.
Future investigations with the DY-BOCS should
include clinical and epidemiological samples, twin
populations, genetic family studies, as well as individuals participating in neuroimaging and treatment
response studies. It is also clear that other instruments may be more time-efficient tools for the initial
screening of large numbers of individuals for possible
OCD.37,38,52
Success in refining the measurement of OC phenotypes is likely to lead to greater clarity concerning
course and outcome, and should facilitate genetic,
neurobiological and treatment response studies. Likewise, dissecting the phenotype into less complex
components may be an important tool for the identification of susceptibility genes for OCD.2730
A dimensional approach may also advance our
nosology, as these dimensions may well cut across
diagnostic boundaries (e.g. eating disorders, body
dysmorphic disorder, as well as the elusive boundary
that divides normal from abnormal psychological
functioning). However, the combined use of both
categorical and dimensional strategies is likely to
offer the greatest promise for a better understanding of
the complex picture of OCD throughout development
and across the spectrum of related diseases.

Acknowledgments
This research was funded by NIH grants MH493515,
MH61940, RR00044, and RR00125; Fundacao de
Amparo a` Pesquisa do Estado de Sao Paulo grant
03/07451-6 (MCR-C) and 99/12205-7 (ECM); grants
from the Tourette Syndrome Association (MCR-C),
the ObsessiveCompulsive Foundation (MCR-C), the
Echlin Foundation and the Kaiser Family. The
authors thank David Mataix-Cols, PhD, Steven A
Rasmussen, MD, David L. Pauls, PhD, and Scott L
Rauch, MD for their help in the development of this
instrument.

References
1 Kessler RC, Chiu WT, Demler O, Walters EE. Prevalence, severity,
and comorbidity of 12-month DSM-IV disorders in the national
comorbidity survey replication. Arch Gen Psychiatry 2005; 62:
617627.
2 Crino R, Slade T, Andrews G. The changing prevalence and
severity of obsessivecompulsive disorder criteria from DSM-III to
DSM-IV. Am J Psychiatry 2005; 162: 876882.
3 Karno M, Golding JM. Obsessivecompulsive disorder. In: Robins
LN, Regier DA (eds). Psychiatric Disorders in America: The
Epidemiologic Catchment Area Study. The Free Press: New York,
1991; pp 204219.
4 Murray CJ, Lopez AD. The global burden of disease: a comprehensive assessment of mortality and disability from diseases,
injuries, and risk factors in 1990 and projected to 2020. In: Murray
CJ, Lopez AD (eds). Harvard University Press: Cambridge, MA,
1996.
5 Miguel EC, Leckman JF, Rauch S, do Rosario-Campos MC, Hounie
AG, Mercadante MT et al. Obsessivecompulsive disorder phenotypes: implications for genetic studies. Mol Psychiatry 2005; 10:
258275.

6 Rasmussen SA, Tsuang MT. Clinical characteristics and family

history in DSM-III obsessivecompulsive disorder. Am J Psychiatry 1986; 143: 317322.
7 Rosario-Campos MC, Leckman JF, Mercadante MT, Shavitt RG,
Prado HS, Sada P et al. Adults with early-onset obsessive
compulsive disorder. Am J Psychiatry 2001; 158: 18991903.
8 Diniz JB, Rosario-Campos MC, Shavitt RG, Curi M, Hounie AG,
Brotto SA et al. Impact of age at onset and duration of illness on
the expression of comorbidities in obsessivecompulsive disorder.
J Clin Psychiatry 2004; 65: 2227.
9 Miguel EC, Baer L, Coffey BJ, Rauch SL, Savage CR, OSullivan RL
et al. Phenomenological differences appearing with repetitive
behaviours in obsessivecompulsive disorder. Br J Psychiatry
1997; 170: 140145.
10 Miguel EC, do Rosario-Campos MC, Prado HS, do Valle R, Rauch
SL, Coffey BJ et al. Sensory phenomena in obsessivecompulsive
disorder and Tourettes disorder. J Clin Psychiatry 2000; 61:
150156.
11 Leckman JF, McDougle CJ, Pauls DL, Peterson BS, Grice DE,
King RA et al. Tic-related versus non-tic related obsessive
compulsive disorder. In: Goodman WK, Rudorfer MV, Mazur JD
(eds). ObsessiveCompulsive Disorder: Contemporary Issues in
Treatment. Lawrence Erlbaum: New York, 2000, pp 4368.
12 Nestadt G, Addington A, Samuels J, Liang KY, Bienvenu OJ,
Riddle M et al. The identification of OCD-related sub-groups based
on comorbidity. Biol Psychiatry 2003; 53: 914920.
13 Ravizza L, Maina G, Bogetto F. Episodic and chronic obsessive
compulsive disorder. Depress Anxiety 1997; 6: 154158.
14 Andreasen NC, Arndt S, Alliger R, Miller D, Flaum M. Symptoms
of schizophrenia. Methods, meanings, and mechanisms. Arch Gen
Psychiatry 1995; 52: 341351.
15 Liddle PF, Friston KJ, Frith CD, Hirsch SR, Jones T, Frackowiak
RS. Patterns of cerebral blood flow in schizophrenia. Br J
Psychiatry 1992; 160: 179186.
16 Cassidy F, Forest K, Murry E, Carroll BJ. A factor analysis of
the signs and symptoms of mania. Arch Gen Psychiatry 1998; 55:
2732.
17 Alsobrook JP, Pauls DL. A factor analysis of tic symptoms in Gilles
de la Tourettes Syndrome. Am J Psychiatry 2002; 159: 291296.
18 Williamson DA, Womble LG, Smeets MAM, Netemeyer RG, Thaw
JM, Kutlesic V et al. Latent structure of eating disorder symptoms:
a factor-analytic and taxonomic investigation. Am J Psychiatry
2002; 159: 412418.
19 Grigorenko EL, Wood FB, Meyer MS, Hart LA, Speed WC, Shuster
A et al. Susceptibility loci for distinct components of developmental dyslexia on chromosomes 6 and 15. Am J Hum Genet 1997;
60: 2739.
20 Mataix-Cols D, do Rosario-Campos MC, Leckman JF. A multidimensional model of obsessivecompulsive disorder. Am J
Psychiatry 2005; 162: 228238.
21 Leckman JF, Grice DE, Boardman J, Zhang H, Vitale A, Bondi C et
al. Symptoms of obsessivecompulsive disorder. Am J Psychiatry
1997; 154: 911917.
22 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 obsessivecompulsive disorder. Am J Psychiatry 1999; 156:
14091416.
23 Summerfeldt LJ, Richter MA, Antony MM, Swinson RP. Symptom
structure in obsessivecompulsive disorder: a confirmatory factoranalytic study. Behav Res Ther 1999; 37: 297311.
24 Hasler G, Lasalle-Ricci VH, Ronquillo JG, Crawley SA, Cochran
LW, Kazuba D et al. Obsessivecompulsive disorder symptom
dimensions show specific relationships to psychiatric comorbidity. Psychiatry Res 2005; 135: 121132.
25 Mataix-Cols D, Rauch SL, Baer L, Shera D, Eisen J, Goodman WK
et al. Symptom stability in adult obsessivecompulsive disorder:
data from a two year naturalistic study. Am J Psychiatry 2002; 159:
263268.
26 Rufer M, Grothusen A, Mass R, Peter H, Hand I. Temporal stability
of symptom dimensions in adult patients with obsessivecompulsive disorder. J Affect Disord 2005; 88: 99102.
27 Alsobrook II JP, Leckman JF, Goodman WK, Rasmussen SA,
Pauls DL. Segregation analysis of obsessivecompulsive disorder

503

Molecular Psychiatry

Dimensional YaleBrown ObsessiveCompulsive Scale

MC Rosario-Campos et al

504
28

29

30

31

32

33

34

35

36

37

38

39

40

using symptom-based factor scores. Am J Med Genet 1999; 88:

669675.
Leckman JF, Pauls DL, Zhang H, Rosario-Campos MC, Katsovich L,
Kidd K et al. Obsessivecompulsive symptom dimensions in
affected sibling pairs diagnosed with Gilles de la Tourette
syndrome. Am J Med Genet 2003; 116: 6068.
Zhang H, Leckman JF, Tsai C-P, Kidd KK, Rosario Campos MC.
The Tourette syndrome association international consortium for
genetics. Genome wide scan of hoarding in sibling pairs both
diagnosed with Gilles de la Tourette syndrome. Am J Hum Genet
2002; 70: 896904.
Cavallini MC, Di Bella D, Siliprandi F, Malchiodi F, Bellodi L.
Exploratory factor-analysis of obsessivecompulsive patients and
association with 5-HTTLPR polymorphism. Am J Med Genet 2002;
114: 347353.
Mataix-Cols D, Cullen S, Lange K, Zelaya F, Andrew C, Amaro E
et al. Neural correlates of anxiety associated with obsessive
compulsive symptom dimensions in normal volunteers. Biol
Psychiatry 2003; 53: 482493.
Mataix-Cols D, Wooderson S, Lawrence N, Brammer MJ, Speckens
A, Phillips ML. Distinct neural correlates of washing, checking
and hoarding symptom dimensions in obsessivecompulsive
disorder. Arch Gen Psychiatry 2004; 61: 564576.
Black DW, Monahan P, Gable J, Blum N, Clancy G, Baker P.
Hoarding and treatment response in 38 nondepressed subjects with
obsessivecompulsive disorder. J Clin Psychiatry 1998; 59: 420425.
Alonso P, Menchon JM, Pifarre J, Mataix-Cols D, Torres L, Salgado
P et al. Long-term follow-up and predictors of clinical outcome in
obsessivecompulsive patients treated with serotonin reuptake
inhibitors and behavioral therapy. J Clin Psychiatry 2001; 62:
535540.
Saxena S, Maidment KM, Vapnik T, Golden G, Rishwain T, Rosen
RM et al. Obsessivecompulsive hoarding: symptom severity and
response to multimodal treatment. J Clin Psychiatry 2002; 63: 2127.
Mataix-Cols D, Marks IM, Greist JH, Kobak KA, Baer L. Obsessive
compulsive symptom dimensions as predictors of compliance
with and response to behaviour therapy: results from a controlled
trial. Psychother Psychosom 2002; 71: 255262.
Foa EB, Huppert JD, Leiberg S, Langner R, Kichic R, Hajcak G et al.
The obsessivecompulsive inventory: development and validation
of a short version. Psychol Assess 2002; 14: 485496.
Thordarson DS, Radomsky AS, Rachman S, Shafran R, Sawchuk
CN, Ralph Hakstian A. The Vancouver obsessional compulsive
inventory (VOCI). Behav Res Ther 2004; 42: 12891314.
Woody SR, Steketee G, Chambless DL. Reliability and validity of
the YaleBrown obsessivecompulsive scale. Behav Res Ther
1995; 33: 597605.
Deacon BJ, Abramowitz JS. The YaleBrown obsessive compulsive
scale: factor analysis, construct validity, and suggestions for
refinement. J Anxiety Disord 2005; 19: 573585.

41 Goodman WK, Price LH, Rasmussen SA, Mazure C, Fleischmann

RL, Hill CL et al. The YaleBrown obsessivecompulsive scale-Idevelopment, use, and reliability. Arch Gen Psychiatry 1989; 46:
10061011.
42 Goodman WK, Price LH, Rasmussen SA, Mazure C, Delgado P,
Heninger GR et al. The YaleBrown obsessivecompulsive scaleII-validity. Arch Gen Psychiatry 1989; 46: 10121016.
43 Pauls D, Hurst CR. Schedule for Tourettes Syndrome and
Other Behavioral Syndrome. Yale University: New Haven,
CT, 1994.
44 Foa EB, Kozak MJ. DSM-IV field trial: obsessivecompulsive
disorder. Am J Psychiatry 1995; 152: 9096.
45 Scahill L, Riddle MA, McSwiggin-Hardin M, Ort SI, King RA,
Goodman WK et al. Childrens YaleBrown obsessivecompulsive
scale: reliability and validity. J Am Acad Child Adolesc Psychiatry
1997; 36: 844852.
46 Leckman JF, Riddle MA, Hardin MT, Ort SI, Swartz KL, Stevenson
J et al. The Yale Global Tic Severity Scale: initial testing of a
clinician-rated scale of tic severity. J Am Acad Child Adolesc
Psychiatry 1989; 28: 566573.
47 Poznanski EO, Cook SC, Carroll BJ. A depression rating scale for
children. Pediatrics 1979; 64: 442450.
48 Hamilton M. A rating scale for depression. J Neurol Neurosurg
Psychiatry 1960; 23: 5662.
49 Maier W, Buller R, Philipp M, Heuser I. The Hamilton anxiety
scale: reliability, validity and sensitivity to change in anxiety and
depressive disorders. J Affect Disord 1988; 14: 6168.
50 Summerfeldt LJ, Kloosterman PH, Antony MM, Richter MA,
Swinson RP. The relationship between miscellaneous symptoms
and major symptom factors in obsessivecompulsive disorder.
Behav Res Ther 2004; 42: 14531467.
51 McKay D, Abramowitz JS, Calamari JE, Kyrios M, Radomsky A,
Sookman D et al. A critical evaluation of obsessivecompulsive
disorder subtypes: symptoms versus mechanisms. Clin Psychol
Rev 2004; 24: 283313.
52 Cooper J. The Leyton obsessional inventory. Psychol Med 1970; 1:
4864.
53 Hodgson RJ, Rachman S. Obsessionalcompulsive complaints.
Behav Res Ther 1977; 15: 389395.
54 Insel TR, Murphy DL, Cohen RM, Alterman I, Kilts C, Linnoila M.
Obsessivecompulsive disorder: a double-blind trial of clomipramine and clorgyline. Arch Gen Psychiatry 1983; 40: 605612.
55 Frost RO, Steketee GS, Krause MS, Trepanier KL. The relationship
of the YaleBrown obsessivecompulsive scale (Y-BOCS) to other
measures of obsessivecompulsive symptoms in a nonclinical
population. J Personal Assess 1995; 65: 158168.
56 Leckman JF, Mayes LC, Feldman R, Evans D, King RA, Cohen DJ.
Early parental preoccupations and behaviors and their possible
relationship to the symptoms of obsessivecompulsive disorder.
Acta Psychiatr Scand 1999; 100: 126.

Supplementary information accompanies the paper on Molecular Psychiatry website (http://www.nature.com/mp)

Molecular Psychiatry