Journal of Child Psychology and Psychiatry **:* (2011), pp ****

doi:10.1111/j.1469-7610.2011.02434.x

Children with very early onset obsessivecompulsive disorder: clinical features and
treatment outcome
Eriko Nakatani,1 Georgina Krebs,2 Nadia Micali,1 Cynthia Turner,1,2
Isobel Heyman,1,2 and David Mataix-Cols1,2
1

Kings College London, Institute of Psychiatry; 2OCD and Related Disorders Clinic for Young People, South London
and Maudsley NHS Foundation Trust, London, UK

Background: There is emerging evidence that early onset obsessive-compulsive disorder (OCD) may be
a phenomenologically distinct subtype of the disorder. Previous research has shown that individuals
who report an early onset display greater severity and persistence of symptoms, and they may be less
responsive to treatment. To date, this question has been investigated solely in adult samples. The
present study represents the first investigation into the effect of age at onset of OCD on clinical characteristics and response to treatment in a paediatric sample. Methods: A total of 365 young people
referred to a specialist OCD clinic were included in the study. Clinical records were used to examine
potential differences in key clinical characteristics between those who had a very early onset of the
disorder (before 10 years) and those who had a late onset (10 years or later). Group differences in
treatment responsiveness were also examined within a subgroup that received cognitive behaviour
therapy (CBT) alone or CBT plus medication (n = 109). Results: The very early onset group were
characterised by a longer duration of illness, higher rates of comorbid tics, more frequent ordering and
repeating compulsions and greater parent-reported psychosocial difficulties. There were no differences
in treatment response between the groups, and when age at onset was examined as a continuous
variable, it did not correlate with treatment response. Conclusions: Very early onset OCD may be
associated with different symptoms and comorbidities compared with late onset OCD. However, these
differences do not appear to impact on responsiveness to developmentally tailored CBT alone or in
combination with medication. These findings further indicate the value in early detection and treatment
of OCD in childhood. Keywords: Obsessive-compulsive disorder, paediatric, age at onset, early onset,
cognitive behaviour therapy.

Introduction
Obsessive-compulsive disorder (OCD) is a chronic
and debilitating condition with a lifetime prevalence
of approximately 2% in general population (Ruscio,
Stein, Chiu, & Kessler, 2010). The disorder often
emerges in childhood or adolescence, with approximately a third to a half of adult patients reporting a
childhood onset (Rasmussen & Eisen, 1990). Paediatric OCD is increasingly recognised as a putative
developmental subtype of the disorder, which is
characterised by a higher preponderance of boys
(Tukel et al., 2005), an increasing familial load for
OCD (Nestadt et al., 2000; Rosario-Campos et al.,
2005) and higher comorbidity with tic disorders
(Diniz et al., 2004; Millet et al., 2004; Rosario-Campos et al., 2001).
Investigations into early onset OCD to date have
been complicated by methodological inconsistencies
in the literature. First, definitions of age at onset
have varied, with some studies reporting the age at
which the patient and/or family members first

Eriko Nakatani and Georgina Krebs are joint first authors.

Conflict of interest statement: No conflicts declared.

noticed the presence of obsessive-compulsive

symptoms (Diniz et al., 2004; de Mathis et al., 2008;
Rosario-Campos et al., 2001), and others reporting
the age at which the patient first fulfilled strict
diagnostic criteria for OCD (Sobin, Blundell, &
Karayiorgou, 2000).
Second, it remains unclear how early onset should
be defined and whether particular age cut-offs
should be employed. Previous studies using adult
and/or paediatric samples have used various
thresholds to select their early onset samples. For
example, early onset has been defined as before 15
(Millet et al., 2004), 17 (Fontenelle, Mendlowicz,
Marques, & Versiani, 2003; Tukel et al., 2005), and
even 18 years (Sobin et al., 2000). In another study,
thresholds of 10 and 17 years were used to categorise early and late onset, respectively (RosarioCampos et al., 2001). Recently, a large-scale study
(n = 330) attempted to establish the most appropriate cut-off points to differentiate early and late onset
OCD (de Mathis et al., 2008). The authors suggested
that ages of 10 and 17 years might be reasonable
thresholds, although concluded that age at onset
may be best measured as a continuous variable.
The lack of consensus in this area was further

 2011 The Authors. Journal of Child Psychology and Psychiatry  2011 Association for Child and Adolescent Mental Health.
Published by Blackwell Publishing, 9600 Garsington Road, Oxford OX4 2DQ, UK and 350 Main St, Malden, MA 02148, USA

Eriko Nakatani et al.

highlighted in a recent worldwide survey among OCD

experts, who failed to agree on the appropriate cutoff for early onset OCD (Mataix-Cols, Pertusa, &
Leckman, 2007).
The clinical utility of differentiating early and late
onset OCD is uncertain. Some studies have reported
that early onset is associated with poorer prognosis
and more severe symptoms after treatment
(Ackerman, Greenland, Bystritsky, Morgenstern, &
Katz, 1994; Fontenelle et al., 2003; Ravizza, Barzega, Bellino, Bogetto, & Maina, 1995; Rosario-Campos et al., 2001), while others have found no
significant relationship between age at onset and
treatment response (Ackerman, Greenland, & Bystritsky, 1998; Alonso et al., 2001; Shavitt et al.,
2006; Uguz, Askin, Cilli, & Besiroglu, 2006). However, most of these studies examined this question in
relation to medication response, rather than cognitive behaviour therapy (CBT), which is the first-line
treatment for the disorder in young people (e.g. National Institute for Health and Clinical Excellence,
2005).
To date, only three studies have examined the
relationship between age at onset of OCD and CBT
response. In a survey of 617 adults with OCD, selfreport data were collected on age at onset and
treatment history (Millet et al., 2004). No differences
were found in response to selective serotonin reuptake inhibitor (SSRI) medication or behaviour therapy between individuals who reported early onset
(before 15 years) and late onset (after 15 years). An
important limitation of this study was the reliance on
retrospective self-report of treatment response. In
another study, case records of 254 adult inpatients
who had received CBT for OCD were reviewed
(Langner et al., 2009). No significant differences
were found between the early onset (before 12 years)
and late onset (15 years or later) groups in terms of
treatment outcome, although there were group differences in terms of the variables that predicted
treatment responsiveness. Lomax, Oldfield, and
Salkovskis (2009) compared CBT response rates
between early onset (12 years or younger; n = 22)
and a late onset (16 years or older; n = 23) OCD.
They found that the two groups were equally
responsive to CBT, although the early onset group
had more severe symptoms both before and after
treatment; the authors suggested that such individuals may require an extended course of CBT.
Arguably, the main limitation of research in this
area is the fact that most studies have ascertained
age at onset retrospectively in adult samples of OCD
patients. This limitation can be partially overcome by
studying the correlates of age at onset in paediatric
samples because children seek help much earlier
than adults with OCD and recall bias is therefore
less likely. For example in a recent UK study, young
people were first seen at a specialist OCD clinic on
average 3 years after the onset of the disorder
(Nakatani, Mataix-Cols, Micali, Turner, & Heyman,

2009). This compares with a reported average delay

of 8 years until diagnosis in adult populations (Stobie, Taylor, Quigley, Ewing, & Salkovskis, 2007). To
date, no study has examined the relationship between age at onset and CBT response in a paediatric
sample.
In this study, we report on a large sample of patients referred to a national specialist clinic for children and adolescents with OCD. The first aim was to
compare the demographic characteristics, clinical
features, and severity of symptoms of patients with
very early (i.e. before 10 years) onset OCD and those
with a late onset (i.e. between 10 and 18 years) OCD.
The second aim was to determine whether there was
a differential effect of age at onset on response to
CBT (delivered as a monotherapy or in combination
with SSRI medication). It was predicted that patients
with very early age at onset would be more likely to
be male, have comorbid tic disorders and present
with more severe symptoms, but that they would be
just as likely to benefit from specialist CBT tailored
to the young persons developmental level.

Methods
Participants
All children and adolescents consecutively referred for
assessment and/or treatment to a national specialist
paediatric OCD clinic at the Maudsley Hospital, London, between the years 1996 and 2007 were included in
the study. Young people are referred to the clinic from
across the United Kingdom and tend to be a relatively
severe or treatment-refractory group, or have complexities regarding diagnosis. All participants met International Classification of Diseases (ICD)-10 (World Health
Organisation, 1996) diagnostic criteria for OCD, as
confirmed by the specialist multidisciplinary team.
Detailed sociodemographic and clinical information,
including age at onset, was gathered from the patients
and their parents at the initial assessment, which lasted approximately 3 hr. In addition to assessing OCD
symptoms, clinical assessment included a careful clinical screen for current/ever tics based on ICD-10 criteria, as well as direct observation and probe for
family history of tic disorders. All data were collected as
part of routine clinical practice at the clinic, and advice
from local ethics committee recommended this project
be classed as audit.
After exclusion of four patients whose age at onset was
missing, a total of 365 patients were included in the
study. Of these, a subgroup of 109 individuals received
CBT. The decision to offer CBT was largely determined by
clinical appropriateness and geographical location (i.e.
feasibility of travelling to the clinic for weekly sessions).
Age at onset was defined as the age at which patients
first displayed significant distress or impairment associated with obsessive-compulsive symptoms. This
information was obtained on the day of the initial
assessment and based primarily on parental report and
supplemented with the young persons report. In the
total sample, the mean age at onset of OCD was 10.2

 2011 The Authors. Journal of Child Psychology and Psychiatry  2011 Association for Child and Adolescent Mental Health.

Children with very early onset OCD

60

Number of patients

50

40

30

20

10

0
3

10

11

12

13

14

15

16

17

Age at onset (years)

Figure 1 Distribution of age at onset in the sample

(SD = 2.9; range = 317) and the median was 10 years

(Figure 1). A median split was used to define the two age
at onset groups: very early onset included children who
reported onset at 9 years or younger and the late onset
group constituted children who reported an onset at
10 years or older. While relatively arbitrary, this
approach ensured similar sized groups and sufficient
statistical power for analysis. In any case, analyses were
repeated using various ages as the defined cut-off point
with no difference on the results obtained (data available
upon request). Furthermore, the role of age at onset was
also examined as a continuous variable.

Measures
The Childrens Yale-Brown Obsessive-Compulsive Scale
(CY-BOCS; Scahill et al., 1997) is a widely used clinician-administered measure of OCD symptom severity.
It includes a symptom checklist and severity scale.
Severity scores range from 0 to 40. It has established
psychometric properties (Scahill et al., 1997) and is
sensitive to treatment effects (Storch, Lewin, De Nadai,
& Murphy, 2010).
The Childrens Obsessive-Compulsive Inventory
(ChOCI; Shafran et al., 2003) is a self-report instrument developed to assess obsessive-compulsive symptoms in children and adolescents. It has a patient and a
parent version, both of which consist of the following
four subscales: obsessions, impairment associated with
obsessions, compulsions and impairment associated
with compulsions. The total score, ranging from 0 to 48,
is constructed by summing the impairment items. It
has been shown to have good internal consistency and
criterion validity and to be significantly correlated with
the CY-BOCS (Uher, Heyman, Turner, & Shafran,
2008).
The Beck Depression Inventory for Youth (BDI-Y;
Beck, Beck, & Jolly, 2001) is a 20-item, self-report
measure of depressive symptoms, which includes
questions about negative thoughts, feelings of sadness
and physiological indications of depression. Total raw
scores range from 0 to 60, and can be translated into a T

score based on the age and gender of the young person.

The BDI-Y displays good internal consistency and testcriterion validity, and correlates highly with other
established measures of depression (Beck et al., 2001).
The BDI-Y was administered to a subset of the current
sample (n = 95).
The Global Assessment of Psychosocial Disability
(GAPD; World Health Organisation, 1996) constitutes
Axis VI of the ICD-10, and is a measure of psychological, social and educational/occupational disability that
has arisen as a consequence of psychiatric or developmental disorders coded on Axes I, II and III. Clinicians
assign a score, ranging from 0 (superior/good social
functioning) to 8 (profound and pervasive social disability). The scale has been shown to have good interrater reliability and comparable properties with the
Childrens Global Assessment Scale, a widely used
measure of global functioning (Dyborg et al., 2000).
The Strengths and Difficulties Questionnaire (SDQ;
Goodman, 1997) is a self-report measure that assesses
psychological adjustment in children and adolescents.
It has self, parent and teacher versions and includes 25
items divided among five subscales that relate to different areas of difficulty. The total difficulties score was
used in the present study, and is constructed by summing the symptom subscales. The measure has good
internal consistency, cross-informant correlation and
retest stability after 46 months, and an elevated score
is predictive of psychiatric diagnosis (Goodman, 2001).

Treatment
CBT was protocol-driven, and broadly based on a
published treatment manual (March & Mulle, 1998). It
involved the following key components: psychoeducation about OCD and anxiety, and development of a
hierarchy of compulsions (Sessions 1 and 2); graded
exposure with response prevention (ERP; Session 3
onwards); and relapse prevention (final session). The
treatment was carefully tailored to the developmental
level of the young person, for example, by modifying the
language and worksheets used. The extent of parental

 2011 The Authors. Journal of Child Psychology and Psychiatry  2011 Association for Child and Adolescent Mental Health.

Eriko Nakatani et al.

involvement varied depending both on the developmental level of the young person and the extent to
which parents were involved in, or accommodating,
compulsive behaviours and avoidance. Sessions usually lasted 1 hr, and were conducted on a weekly basis
whenever possible. In most cases, 812 sessions were
offered. For most patients, CBT was delivered in an
individual, face-to-face format, although some received
group CBT (n = 4) or telephone-based CBT (n = 2). A
proportion of young people (n = 77) also received SSRI
medication in combination with CBT; in most cases,
medication was started and had reached a stable dose
before CBT commenced.

Statistical analyses
Data was analysed using SPSS version 18 (IBM, Chicago,
Illinois). software. Chi-squared tests were used for
between-group comparisons of categorical variables and
independent sample t-tests for continuous variables. A
mixed model ANOVA was used to test for a differential
effect of age at onset on responsiveness to CBT. The
associations between age at onset (as a continuous
variable) and other clinical variables of interest were
examined with Pearson correlations. Finally, a multiple
regression analysis was conducted to investigate the extent to which age at onset predicts OCD symptom severity
following CBT, controlling for other variables of interest.
The significance level was set at p < .05 (two-tailed).

range = 618 years) at assessment, and duration of

illness of 3.6 years (SD = 2.8, range = 013). The
mean total CY-BOCS score was 22.3 (SD = 7.9),
indicating moderately severe OCD. The most frequently assigned grade on the GAPD was 3, corresponding to moderate psychosocial disability
(M = 3.2, SD = 1.5).

Comparison of very early and later onset groups

Demographic and clinical characteristics of the two
groups are shown in Table 1. The very early onset
group was younger at assessment and had a longer
history of OCD than the late onset group. The two
groups were comparable in terms of gender
distribution, family history of OCD, symptom severity (CY-BOCS and ChOCI), depression (BDI) and
psychosocial disability (GAPD). Comorbid chronic tic
disorders were more frequent in the very early onset
group, although the proportion of patients with
comorbid Tourette syndrome was comparable. On
the parent-rated SDQ, patients with very early onset
OCD had significantly greater scores on the total
difficulties. There were no between-group differences
on the self or teacher versions of the SDQ. Repeating
and ordering compulsions were significantly more
frequent in the very early onset group (Table 2).

Effectiveness of CBT

Results
Sample characteristics
The sample consisted predominantly of boys
(58.6%), with a mean age of 13.8 years (SD = 2.5,

Of the total sample, 109 (40 very early onset, 69 later

onset) were treated with CBT at the clinic and had
available CY-BOCS scores before and after the treat-

Table 1: Comparison of demographic and clinical characteristics of patients in very early and later onset group
Very early
onset (n=151)
Age (years) at assessment, mean (SD)
Boys, n (%)
Duration of OCD (years), mean (SD)
Age at onset (years), mean (SD)
Tic disorders
Chronic Tics, n (%)
Tourette syndrome, n (%)
Any Tic disorder, n (%)
Family history of OCD, n (%)
CY-BOCS scores (n=329)
Total, mean (SD)
Obsessions score, mean (SD)
Compulsions score, mean (SD)
ChOCI scores
Self total (n=251), mean (SD)
Parent total (n=228), mean (SD)
SDQ scores
Self total (n=218), mean (SD)
Parent total (n=328), mean (SD)
Teacher total (n=209), mean (SD)
GAPD score (n=365), mean (SD)
BDI-Y score (n=95), mean (SD)

Later onset
(n=214)

Chi square/
t-test

p
<0.0001**
0.91
<0.0001**
<0.0001**

12.5
89
5.1
7.4

(2.9)
(58.9)
(3.4)
(1.7)

14.7
125
2.6
12.5

(1.7)
(58.4)
(1.7)
(2.9)

)8.45
0.01
8.48
)9.18

25
24
49
17

(16.6)
(15.9)
(32.5)
(11.5)

18
23
41
16

(8.4)
(10.7)
(19.2)
(7.7)

5.65
2.09
8.42
1.52

0.02*
0.14
0.004**
0.22

22.2 (7.7)
10.1 (4.6)
12.0 (3.8)

22.4 (8.0)
10.7 (4.7)
11.7 (4.3)

)0.26
)1.07
0.54

0.78
0.28
0.58

29.5 (8.7)
32.4 (8.5)

28.1 (9.8)
32.6 (8.6)

1.16
)0.18

0.24
0.85

17.7
19.2
14.4
3.0
23.3

17.5
17.4
14.3
3.3
21.7

0.31
2.34
0.11
)1.81
0.68

0.75
0.03*
0.90
0.07
0.49

(6.9)
(7.2)
(7.6)
(1.4)
(11.3)

(6.4)
(6.8)
(7.4)
(1.5)
(12.0)

OCD, obsessive-compulsive disorder; CY-BOCS, Childrens Yale-Brown Obsessive-Compulsive Scale; ChOCI, Childrens ObsessiveCompulsive Inventory; SDQ, Strengths and Difficulties Questionnaire; GAPD, Global Assessment of Psychosocial Disability;
BDI-Y,Beck Depression Inventory for Youth. * = significant at .05 level;** = significant at .01 level.
 2011 The Authors. Journal of Child Psychology and Psychiatry  2011 Association for Child and Adolescent Mental Health.

Children with very early onset OCD

Table 2: Frequency of obsessions and compulsions in young
people with very early and late onset OCD according to the
CY-BOCS symptom checklist.
Very early
onset
Later onset
Chi
(n=124)
(n=177)
square
Obsessions (%)
Contamination
Aggressive
Sexual
Hoarding
Magical
Somatic
Religious
Compulsions (%)
Cleaning
Checking
Repeating
Counting
Ordering
Hoarding
Superstitious games
Rituals involving
others

85
89
23
39
42
42
37

(68.5)
(71.8)
(18.5)
(31.5)
(33.9)
(33.9)
(29.8)

123
124
41
43
59
66
67

(69.5)
(70.1)
(23.2)
(24.3)
(33.3)
(33.7)
(37.9)

0.03
0.04
0.92
1.88
0.009
0.37
0.20

0.86
0.75
0.34
0.17
0.92
0.54
0.15

91
91
91
59
74
41
49
72

(71.7)
(71.7)
(71.7)
(46.5)
(58.3)
(32.3)
(38.6)
(56.7)

122
117
102
73
77
60
68
108

(68.9)
(66.1)
(57.6)
(41.2)
(43.5)
(33.9)
(38.4)
(61.0)

0.26
1.05
6.22
1.05
6.44
0.08
0.001
0.57

0.61
0.30
0.01*
0.37
0.01*
0.77
0.98
0.45

Table 3 Mean (standard deviation) Childrens Yale-Brown

Obsessive-Compulsive Scale (CY-BOCS) scores pre- and posttreatment for the very early and late onset groups
Very early
onset
(n = 40)

CY-BOCS total
Pre
23.6 (5.6)
Post
9.8 (5.8)
CY-BOCS obsessions
Pre
10.9 (3.5)
Post
4.3 (3.1)
CY-BOCS compulsions
Pre
12.7 (2.9)
Post
5.7 (3.4)

the within-subjects factor of time (pre- vs. post-treatment) and the between-subjects factor of onset group
(very early vs. later onset) was conducted and revealed
a main effect of time [F(1, 108) = 332.46, p < .001], as
indicated by a significant reduction in CY-BOCS score
over the course of the treatment [M = 23.40,
SD = 5.60 vs. M = 10.76, SD = 6.92; t(108) = 18.51,
p < .001]. There was no main effect of onset group
[F(1, 108) = 0.319, p = .57], and no significant
Time Onset Group interaction [F(1, 108) = 1.64,
p = .208], indicating that the two onset groups responded equally well to treatment.
Participants were classified into the following
severity groups based on their total CY-BOCS score:
subclinical (010); mild (1119); moderate (2029);
and severe (>30). Of the 109 treated patients, 24
(60.0%) in very early onset group and 38 (55.1%) in
the later onset group were classified as having subclinical OCD symptoms following treatment
(v2 = .251, df = 1, p = .69; Figure 2).

Age at onset as a continuous variable

* = significant at .05 level.

Treatment
time point

Late
onset
(n = 69)

23.3 (5.6)
11.3 (7.5)

0.277
)1.076

.782
.284

11.1 (4.0)
5.4 (4.3)

)0.251
)1.318

.802
.190

12.2 (3.2)
6.0 (4.2)

0.807
)0.369

.421
.297

ment. Analysis of this subgroup with respect to

demographic and clinical characteristics yielded
similar findings to the total sample. Patients in the
very early onset were significantly younger (M = 12.7
vs. M = 14.8 years, p < .001) and had a longer duration of illness (M = 5.3 vs. M = 2.4 years, p < .001).
There were no significant differences in terms of
gender distribution in the very early and later onset
groups (boys: 60.5% and 55.1%; v2 = .251, df = 1,
p = .69). Of the 109 treated patients, 75 received CBT
concomitant with SSRI medication. The proportion
of patients receiving combined treatment was equivalent in the very early and later onset groups (77.5%
vs. 63.8%; v2 = 2.22, df = 1, p = .20).
Mean CY-BOCS scores pre- and post-treatment are
shown in Table 3. The mean percentage reduction in
total CY-BOCS score from the baseline to post-treatment was 58.4% for very early onset group and 51.5%
for the later onset group. A mixed-model ANOVA with

Planned correlational analyses were conducted to

further investigate the relationship between age at
onset and other clinical characteristics. On the
parent-rated SDQ, there was a negative correlation
between age at onset and the total difficulties score
(r = )0.14, p < .05). There were no other statistically
significant associations between age at onset and
clinical measurements (CY-BOCS, ChOCI, BDI, SDQ
and GAPD). Among the 109 treated patients, there
was no statistically significant correlation between
age at onset and the outcome variables including
post-treatment CY-BOCS score and percentage
reduction on the CY-BOCS (all p > .05).
An additional exploratory multiple regression
analysis was conducted with the post-treatment
CY-BOCS score as the dependent variable and the
following variables as regressors: pretreatment
CY-BOCS score; age at onset; presence of tic disorder;
chronicity of OCD; medication status; and gender.
This analysis revealed that pretreatment severity
predicted post-treatment severity (b = .367, t = 4.09,
p < .001). Concomitant use of SSRIs (b = .230,
t = 2.60, p = .011) was also a significant predictor;
combined treatment was associated with more severe
symptoms after CBT, controlling for pretreatment
severity. No other variables were significant predictors of post-treatment symptom severity.

Discussion
This study examined the influence of age at onset of
OCD on clinical characteristics and responsiveness
to CBT in a large paediatric sample. As predicted, we
found a number of differences in the demographic
and clinical characteristics of the very early onset
group compared with the late onset group. Individuals who developed OCD before 10 years were found

 2011 The Authors. Journal of Child Psychology and Psychiatry  2011 Association for Child and Adolescent Mental Health.

Eriko Nakatani et al.

0

100

13

15

10

13

80

Percentage of sample

28

60

31
64

60

Severe >30
Moderate 2029
Mild 1119
Sublcinical 010

40
60

55

20
25

25
0

0
Pre

Post

Very early onset

Pre

Post
Later onset

Figure 2 Obsessive-compulsive disorder symptom severity (total Childrens Yale-Brown Obsessive-Compulsive Scale scores) reported by
the very early and later onset groups before and after cognitive behaviour therapy

to have had a longer duration of illness, despite

being younger at assessment, compared with the late
onset group. This may partly be because of obsessive-compulsive symptoms being mistaken as a
normal developmental phase in very young children,
and partly because of poorer insight in this population which in turn means they are less likely to seek
help. The very early onset group also had higher
rates of comorbid tic disorders, and their OCD
symptoms more frequently involved repeating and
ordering/arranging. These findings are consistent
with previous studies which have demonstrated a
relationship between early onset OCD and tic disorders (Diniz et al., 2004; Millet et al., 2004; RosarioCampos et al., 2001), and also elevated levels of
repeating and ordering compulsions among OCD
patients with tic disorders (Scahill et al., 2003).
No relationship was found between age at onset
and OCD symptom severity, an association that has
been demonstrated in previous studies conducted
among adults with OCD (e.g. Jansch et al., 2007;
Lomax et al., 2009). However, parents of the very
early onset group reported a higher level of psychosocial difficulties, possibly indicating greater
impairment associated with OCD symptoms, which
could well be a consequence of greater duration of
illness. Both the very early and later onset groups
were characterised by a male preponderance, which
is a well-replicated finding among paediatric OCD
studies (e.g. Geller et al., 2001; Last & Strauss,
1989; Swedo, Rapoport, Leonard, Lenane, & Cheslow, 1989). Somewhat surprisingly, no group differences were observed in the gender distribution,
unlike previous studies conducted among adults
which have demonstrated a greater ratio of boys
among early onset patients (Tukel et al., 2005). This
inconsistency may be explained by the varying age

thresholds used across studies, and the fact that the

sex ratio in OCD may switch from predominantly
men to predominantly women during teenage years
(Castle, Deale, & Marks, 1995; Geller et al., 2001).
Crucially, both the very early onset and late onset
groups demonstrated significant reductions in OCD
symptoms over the course of CBT, and the extent of
symptom reduction and the proportion of patients
achieving remission was found to be equivalent for
the two groups. When age at onset was examined as
a continuous variable, again no relationship was
found between this factor and CBT response. This
finding is in line with previous studies in adult
populations (Langner et al., 2009; Lomax et al.,
2009; Millet et al., 2004), and suggests that developmentally appropriate CBT for OCD, delivered as a
monotherapy or in combination with SSRI medication, is robust to age at onset and duration of illness
in young people. This was further supported by a
multiple regression analysis, which found that age at
onset did not predict OCD symptom severity after
treatment. More severe pretreatment symptoms and
concomitant medication were both independent
predictors of more severe symptoms at post-treatment. With respect to medication, this finding may
reflect a tendency to prescribe medication for cases
with greater complexities in their clinical presentation (e.g. significant comorbidities), which might
create barriers in CBT.
This study has a number of limitations. First, age at
onset was determined by retrospective recall by the
young person and parents. However, compared with
previous studies that have been conducted in adult
populations and relied solely on the patients report,
this study has the advantage of establishing onset
timings closer to the actual onset date, as well as
utilising multiple informant accounts. Second, this

 2011 The Authors. Journal of Child Psychology and Psychiatry  2011 Association for Child and Adolescent Mental Health.

Children with very early onset OCD

study was conducted in a specialist clinic that tends

to receive referrals for complicated or severe OCD,
and hence the patients may not have been a representative sample. However, as discussed before, a
number of similarities were noted between the sample characteristics reported here and previous studies. Third, no structured diagnostic interviews were
used to assign diagnoses and therefore it was not
possible to examine the influence of age at onset on
comorbidity, other than tic disorders which were
routinely assessed. Fourth, some of the young people
who received CBT were also on SSRI medication,
although in most cases medication was started and
had reached a stable dose before CBT commenced.
Furthermore, the proportion of patients on medication in the very early and late onset groups was
comparable.
In summary, this study represents the first
investigation into the potential influence of age at
onset of OCD on responsiveness to treatment in
young people. The current findings suggest that
although very early onset OCD could be phenomenologically distinct from late onset OCD, the age at
which the disorder emerges is not a prognostic factor
for treatment. Individuals who have a very early
onset respond equally well to CBT that is tailored to
their developmental level, compared with young

people with a late onset. This finding is encouraging

and indicates the value of early detection and treatment of the disorder. At present, OCD in youth often
goes undetected for many years, thus delaying
access to evidence-based treatment. A longer duration of illness has been shown to predict persistence
of OCD symptoms (e.g. Micali et al., 2010), which
can in turn lead to substantial disability that
expands into adulthood. Further investigation into
the phenotype of OCD in childhood has the potential
to assist clinicians in the detection and diagnosis of
the disorder, thereby facilitating early intervention
and improving clinical outcomes.

Acknowledgements
This study was partially funded by a grant from the
South London and Maudsley NHS Foundation
Trust.

Correspondence to
David Mataix-Cols, Kings College London, Institute of
Psychiatry, PO Box 69, De Crespigny Park, London SE5
8AF, UK; Tel: +44 2078480543; Email: david.
mataix-cols@kcl.ac.uk

Key points
The current study represents the first investigation into the potential association between age at onset of OCD
and responsiveness to treatment in young people.
Young people who developed OCD before 10 years of age were more likely to present with comorbid tics,
ordering and repeating compulsions, and their parents reported greater psychosocial difficulties, compared
with those who developed OCD at 10 years or later.
Importantly, the very early onset and late onset groups were equally responsive to CBT, delivered alone or in
combination with medication.
CBT tailored to the developmental age of the child is a powerful treatment either as monotherapy or in
combination with SSRI medication.

References
Ackerman, D.L., Greenland, S., & Bystritsky, A. (1998).
Clinical characteristics of response to fluoxetine treatment
of obsessive-compulsive disorder. Journal of Clinical Psychopharmacology, 18, 185192.
Ackerman, D.L., Greenland, S., Bystritsky, A., Morgenstern,
H., & Katz, R.J. (1994). Predictors of treatment response in
obsessive-compulsive disorder: Multivariate analyses from a
multicenter trial of clomipramine. Journal of Clinical Psychopharmacology, 14, 247254.
Alonso, P., Menchon, J.M., Pifarre, J., Mataix-Cols, D., Torres,
L., Salgado, P., & Vallejo, J. (2001). Long-term follow-up and
predictors of clinical outcome in obsessive-compulsive patients treated with serotonin reuptake inhibitors and behavioural therapy. Journal of Clinical Psychiatry, 62, 535540.
Beck, J.S., Beck, A.T., & Jolly, J.B. (2001). Beck youth
inventories for children and adolescents. San Antonio, TX:
The Psychological Corporation.
Castle, D.J., Deale, A., & Marks, I.M. (1995). Gender differences in obsessive compulsive disorder. Australian and New
Zealand Journal of Psychiatry, 29, 114117.

Diniz, J.B., Rosario-Campos, M.C., Shavitt, R.G., Curi, M.,

Hounie, A.G., Brotto, S., & Miguel, E.C. (2004). Impact of age
at onset and duration of illness on the expression of
comorbidities in obsessive-compulsive disorder. Journal of
Clinical Psychiatry, 65, 2227.
Dyborg, J., Warborg Larsen, F., Nielson, S., Byman, J., Buhl
Neilson, B., & Gautre`-Delay, F. (2000). The Childrens
Global Assessment Scale (CGAS) and Global Assessment of
Psychosocial Disability (GAPD) in clinical practice Substance and reliability as judged by intraclass correlations.
European Child and Adolescent Psychiatry, 9, 195201.
Fontenelle, L., Mendlowicz, M.V., Marques, C., & Versiani,
M. (2003). Early- and late onset obsessive-compulsive
disorder in adult patients: An exploratory clinical and
therapeutic study. Journal of Psychiatric Research, 37,
127133.
Geller, D.A., Biederman, J., Faraone, S., Agranat, A., Cradock,
K., Hagermoser, L., , Coffey, B.J. (2001). Developmental
aspects of obsessive compulsive disorder: Findings in children, adolescents, and adults. Journal of Nervous and
Mental Disease, 189, 471477.

 2011 The Authors. Journal of Child Psychology and Psychiatry  2011 Association for Child and Adolescent Mental Health.

Eriko Nakatani et al.

Goodman, R. (1997). The Strengths and Difficulties Questionnaire: A research note. Journal of Child Psychology and
Psychiatry, 38, 581586.
Goodman, R. (2001). Psychometric properties of the strengths
and difficulties questionnaire. Journal of the American
Academy of Child and Adolescent Psychiatry, 40, 1337
1345.
nsch, P., Zaudig, M., Roper, G., Hauke, W., Piesbergen, C., &
Ja
Butollo, W. (2007). The early onset of obsessive-compulsive
disorder and its impact on symptomatology and severity.
Verhaltenstherapie, 17, 3743.
Langner, J., Laws, M., Roper, G., Zaudig, M., Hauke, W., &
Piesbergen, C. (2009). Predicting therapy outcome in
patients with early and late obsessive-compulsive disorder
(EOCD and LOCD). Behavioural and Cognitive Psychotherapy, 37, 485496.
Last, C.G., & Strauss, C.C. (1989). Obsessive-compulsive
disorder in childhood. Journal of Anxiety Disorders, 3,
295302.
Lomax, C.L., Oldfield, V.B., & Salkovskis, P.M. (2009). Clinical
and treatment comparisons between adults with earlyand late-onset obsessive-compulsive disorder. Behaviour
Research and Therapy, 47, 99104.
March, J., & Mulle, K. (1998). OCD in children and adolescents:
A cognitive-behavioral treatment manual. New York: The
Guilford Press.
Mataix-Cols, D., Pertusa, A., & Leckman, J.F. (2007). Issues
for DSM-V: How should obsessive-compulsive and related
disorders be classified? American Journal of Psychiatry, 164,
13131314.
de Mathis, M.A., do Rosario, M.C., Diniz, J.B., Torres, A.R.,
Shavitt, R.G., Ferra
o, Y.A., , Miguel, E.C. (2008).
Obsessive compulsive disorder: Influence of age at onset
on comorbidity patterns. European Psychiatry, 23, 187194.
Micali, N., Heyman, I., Perez, M., Hilton, K., Nakatani, E.,
Turner, C., & Mataix-Cols, D. (2010). Long-term outcomes of
obsessive-compulsive disorder: Follow-up of 142 children
and adolescents. British Journal of Psychiatry, 197, 128
134.
Millet, B., Kochman, F., Gallarda, T., Krebs, M.O., Demonfaucon, F., Barrot, I., , Hantouche, E.G. (2004). Phenomenological and comorbid features associated in obsessive
compulsive disorder: Influence of age of onset. Journal of
Affective Disorders, 79, 241246.
Nakatani, E., Mataix-Cols, D., Micali, N., Turner, C., &
Heyman, I. (2009). Outcomes of cognitive behaviour therapy
for obsessive compulsive disorder in a clinical setting: A 10year experience from a specialist OCD service for children
and adolescents. Child and Adolescent Mental Health, 14,
133139.
National Institute for Health and Clinical Excellence (2005).
Obsessive-compulsive disorder: Core interventions in the
treatment of obsessive-compulsive disorder and body dysmorphic disorder. London: NICE.
Nestadt, G., Samuels, J., Riddle, M.A., Bienvenu, O.J., Liang,
K.Y., La Buda, M., , Hochn-Saric, R. (2000). A family study
of obsessive-compulsive disorder. Archives of General Psychiatry, 57, 358363.
Rasmussen, S.A., & Eisen, J.L. (1990). Epidemiology of
obsessive compulsive disorder. Journal of Clinical Psychiatry, 53, 1014.
Ravizza, L., Barzega, G., Bellino, S., Bogetto, F., & Maina, G.
(1995). Predictors of drug treatment response in obsessivecompulsive disorder. Journal of Clinical Psychiatry, 56, 368
373.
Rosario-Campos, M.C., Leckman, J.F., Curi, M., Quatrano, S.,
Katsovitch, L., Miguel, E.C., & Pauls, D.L. (2005). A family

study of early-onset obsessive-compulsive disorder. American Journal of Medical Genetics Part B: Neuropsychiatric
Genetics, 136, 9297.
Rosario-Campos, M., Leckman, J., Mercadante, M.T., Shavitt,
R., da Silva Prado, H., Sada, P., , Miguel, E.C. (2001).
Adults with early-onset obsessive-compulsive disorder.
American Journal of Psychiatry, 158, 18991903.
Ruscio, A.M., Stein, D.J., Chiu, W.T., & Kessler, R.C. (2010).
The epidemiology of obsessive-compulsive disorder in the
National Comorbidity Survey Replication. Molecular Psychiatry, 15, 5363.
Scahill, L., Kano, Y., King, R.A., Carlson, A., Peller, A., LeBrun,
U., , Leckman, J.F. (2003). Influence of age and tic
disorders on obsessive-compulsive disorder in a pediatric
sample. Journal of Child and Adolescent Psychopharmacology, 13, 717.
Scahill, L., Riddle, M.A., McSwiggin-Hardin, M., Ort, S.I., King,
R.A., Goodman, W.K., , Leckman, J.F. (1997). Childrens
Yale-Brown Obsessive Compulsive Scale: Reliability and
validity. Journal of the American Academy of Child and
Adolescent Psychiatry, 36, 844852.
Shafran, R., Frampton, I., Heyman, I., Reynolds, M., Teachman, B., & Rachman, S. (2003). Preliminary development of
a new self report measure for OCD in young people. Journal
of Adolescence, 26, 137142.
Shavitt, R.G., Belotto, C., Curi, M., Hounie, A.G., RosarioCampos, M.C., Diniz, J.B., , Miguel, E.C. (2006). Clinical
features associated with treatment response in obsessive
compulsive disorder. Comprehensive Psychiatry, 47, 276
281.
Sobin, C., Blundell, M.L., & Karayiorgou, M. (2000). Phenotypic differences in early and late-onset obsessive-compulsive disorder. Comprehensive Psychiatry, 41, 373379.
Stobie, B., Taylor, T., Quigley, A., Ewing, S., & Salkovskis, P.M.
(2007). Contents May Vary: A pilot study of treatment
histories of OCD patients. Behavioural and Cognitive Psychotherapy, 35, 273282.
Storch, E.A., Lewin, A.B., De Nadai, A.S., & Murphy, T.K.
(2010). Defining treatment response and treatment remission in obsessive-compulsive disorder: A signal detection
analysis of the Childrens Yale-Brown Obsessive Compulsive
Scale. Journal of the American Academy of Child and
Adolescent Psychiatry, 49, 708717.
Swedo, S.E., Rapoport, J.L., Leonard, H., Lenane, M., &
Cheslow, D. (1989). Obsessive-compulsive disorder in children and adolescents. Clinical phenomenology of 70 consecutive cases. Archives of General Psychiatry, 49, 335341.
Tukel, R., Ertekin, E., Batmaz, S., Alyanak, F., Sozen, A.,
Aslantas, B., , Ozyildirim, I. (2005). Influence of age of
onset on clinical features in obsessive-compulsive disorder.
Depression and Anxiety, 21, 112117.
Uguz, F., Askin, R., Cilli, A.S., & Besiroglu, L. (2006).
Comparison of treatment responses and clinical characteristics of early-onset and late-onset obsessive compulsive
disorder. International Journal of Psychiatry in Clinical
Practice, 10, 291296.
Uher, R., Heyman, I., Turner, C.M., & Shafran, R. (2008). Self-,
parent-report and interview measures of obsessivecompulsive disorder in children and adolescents. Journal of Anxiety
Disorders, 22, 979990.
World Health Organisation (1996). The ICD-10 classification of
mental and behavioural disorders. Multiaxial classification of
child and adolescent psychiatric disorders. Cambridge:
Cambridge University Press.

Manuscript accepted 16 May 2011

 2011 The Authors. Journal of Child Psychology and Psychiatry  2011 Association for Child and Adolescent Mental Health.