Journal of Child Psychology and Psychiatry 52:12 (2011), pp 12611268

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 rst 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 difculties. 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 ndings 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, denitions of age at onset have varied, with some studies reporting the age at which the patient and/or family members rst

Eriko Nakatani and Georgina Krebs are joint rst authors. Conict of interest statement: No conicts 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 rst fullled strict diagnostic criteria for OCD (Sobin, Blundell, & Karayiorgou, 2000). Second, it remains unclear how early onset should be dened 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 dened 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

1262

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 signicant 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 rst-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 signicant 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, Oldeld, 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 rst 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 rst 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 benet 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 Classication of Diseases (ICD)-10 (World Health Organisation, 1996) diagnostic criteria for OCD, as conrmed 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 dened as the age at which patients rst displayed signicant 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

1263

60

50

Number of patients

40

30

20

10

0 3 4 5 6 7 8 9 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 dene 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 sufcient statistical power for analysis. In any case, analyses were repeated using various ages as the dened 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 signicantly 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 Difculties 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 ve subscales that relate to different areas of difculty. The total difculties 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 (nal 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.

1264

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.

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 signicantly greater scores on the total difculties. There were no between-group differences on the self or teacher versions of the SDQ. Repeating and ordering compulsions were signicantly more frequent in the very early onset group (Table 2).

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 signicance level was set at p < .05 (two-tailed).

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

Effectiveness of CBT
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) 12.5 89 5.1 7.4 25 24 49 17 (2.9) (58.9) (3.4) (1.7) (16.6) (15.9) (32.5) (11.5) Later onset (n=214) 14.7 125 2.6 12.5 18 23 41 16 (1.7) (58.4) (1.7) (2.9) (8.4) (10.7) (19.2) (7.7) Chi square/ t-test )8.45 0.01 8.48 )9.18 5.65 2.09 8.42 1.52 )0.26 )1.07 0.54 1.16 )0.18 0.31 2.34 0.11 )1.81 0.68 p <0.0001** 0.91 <0.0001** <0.0001** 0.02* 0.14 0.004** 0.22 0.78 0.28 0.58 0.24 0.85 0.75 0.03* 0.90 0.07 0.49

22.2 (7.7) 10.1 (4.6) 12.0 (3.8) 29.5 (8.7) 32.4 (8.5) 17.7 19.2 14.4 3.0 23.3 (6.9) (7.2) (7.6) (1.4) (11.3)

22.4 (8.0) 10.7 (4.7) 11.7 (4.3) 28.1 (9.8) 32.6 (8.6) 17.5 17.4 14.3 3.3 21.7 (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 Difculties Questionnaire; GAPD, Global Assessment of Psychosocial Disability; BDI-Y,Beck Depression Inventory for Youth. * = signicant at .05 level;** = signicant 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 91 91 91 59 74 41 49 72 (68.5) (71.8) (18.5) (31.5) (33.9) (33.9) (29.8) (71.7) (71.7) (71.7) (46.5) (58.3) (32.3) (38.6) (56.7) 123 124 41 43 59 66 67 122 117 102 73 77 60 68 108 (69.5) (70.1) (23.2) (24.3) (33.3) (33.7) (37.9) (68.9) (66.1) (57.6) (41.2) (43.5) (33.9) (38.4) (61.0) 0.03 0.04 0.92 1.88 0.009 0.37 0.20 0.26 1.05 6.22 1.05 6.44 0.08 0.001 0.57

1265

p 0.86 0.75 0.34 0.17 0.92 0.54 0.15 0.61 0.30 0.01* 0.37 0.01* 0.77 0.98 0.45

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 signicant 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 signicant Time Onset Group interaction [F(1, 108) = 1.64, p = .208], indicating that the two onset groups responded equally well to treatment. Participants were classied 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 classied as having subclinical OCD symptoms following treatment (v2 = .251, df = 1, p = .69; Figure 2).

* = signicant at .05 level.

Age at onset as a continuous variable

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 difculties score (r = )0.14, p < .05). There were no other statistically signicant associations between age at onset and clinical measurements (CY-BOCS, ChOCI, BDI, SDQ and GAPD). Among the 109 treated patients, there was no statistically signicant 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 signicant predictor; combined treatment was associated with more severe symptoms after CBT, controlling for pretreatment severity. No other variables were signicant predictors of post-treatment symptom severity.

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) Late onset (n = 69) 23.3 (5.6) 11.3 (7.5) 11.1 (4.0) 5.4 (4.3) 12.2 (3.2) 6.0 (4.2)

Treatment time point

t 0.277 )1.076 )0.251 )1.318 0.807 )0.369

p .782 .284 .802 .190 .421 .297

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)

ment. Analysis of this subgroup with respect to demographic and clinical characteristics yielded similar ndings to the total sample. Patients in the very early onset were signicantly 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 signicant 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

Discussion
This study examined the inuence 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.

1266

Eriko Nakatani et al.

0 15 80 28 31 64 Severe >30 Moderate 2029 Mild 1119 40 60 20 25 0 0 Pre Post Very early onset 25 1 Pre Later onset Post 55 Sublcinical 010 13 1 13

100

10

Percentage of sample

60 60

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 ndings 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 nsch et al., 2007; among adults with OCD (e.g. Ja Lomax et al., 2009). However, parents of the very early onset group reported a higher level of psychosocial difculties, 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 nding 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 signicant 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 nding 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 nding may reect a tendency to prescribe medication for cases with greater complexities in their clinical presentation (e.g. signicant 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

1267

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 inuence 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 rst investigation into the potential inuence of age at onset of OCD on responsiveness to treatment in young people. The current ndings 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 nding 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 rst 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 difculties, 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 uoxetine 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.

1268

Eriko Nakatani et al. can 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). Inuence 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., Rosa rioCampos, 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). Dening 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). Inuence 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 classication of mental and behavioural disorders. Multiaxial classication of child and adolescent psychiatric disorders. Cambridge: Cambridge University Press.

Goodman, R. (1997). The Strengths and Difculties Questionnaire: A research note. Journal of Child Psychology and Psychiatry, 38, 581586. Goodman, R. (2001). Psychometric properties of the strengths and difculties questionnaire. Journal of the American Academy of Child and Adolescent Psychiatry, 40, 1337 1345. nsch, P., Zaudig, M., Ro per, 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. per, G., Zaudig, M., Hauke, W., & Langner, J., Laws, M., Ro 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., Oldeld, 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 classied? 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: Inuence 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: Inuence 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. Ameri-

Accepted for publication: 16 May 2011 Published online: 5 July 2011

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

Copyright of Journal of Child Psychology & Psychiatry is the property of Wiley-Blackwell and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.