The structure of childhood obsessions and compulsions: Dimensions in an outpatient sample☆
Introduction
While the Diagnostic and Statistical Manual (DSM-IV-TR; American Psychiatric Association, 2000) lists obsessive-compulsive disorder (OCD) as a unitary condition, the diagnosis itself does not convey a sense of the complexity of the disorder. OCD is widely considered as one of the most complex anxiety disorders to treat and is generally chronic in course (Steketee & Barlow, 2002). Whereas the DSM-IV provides the perspective to the untrained observer that OCD is comprised of clearly specified symptoms, the manner of clinical presentation and symptom manifestation is often heterogeneous. In addition to the most commonly known symptoms (e.g., contamination fears, washing and checking rituals; Ball, Baer, & Otto, 1996), primary obsessive-compulsive symptoms may include concerns with symmetry or exactness, obsessions of harm, injury or sex, hoarding, repeating rituals, and rituals involving lucky or unlucky numbers, to name a few. While the parsimonious nature of the DSM-IV allows for these various symptoms to appear under the umbrella of OCD, wide variations exist in the manifestations and treatment outcomes associated with these symptoms (Foa et al., 1983; Piacentini, Bergman, Jacobs, McCracken, & Kretchman, 2002; McKay et al., 2004).
The complexity of symptom presentation for OCD is not idiosyncratic to adults. The prevalence of OCD in childhood is similar to that for adults (Rapoport et al., 2000; Rasmussen & Eisen, 1990), and a significant proportion of adults with the disorder report a childhood onset for their illness (Rasmussen & Eisen, 1990). The collective body of evidence suggests similarities in clinical presentation, including the relative prevalence of specific obsessive and compulsive symptoms, across the age span (Swedo, Rapoport, Leonard, Lenane, & Cheslow, 1989; Rasmussen & Eisen, 1990). In research examining the response to treatment in adult samples, it appears that different dimensions of OCD show differential response to treatment (Mataix-Cols, Rauch, Manzo, Jenike, & Baer, 1999). In order to extend this line of inquiry to childhood OCD, the identification of discrete symptom dimensions in children and adolescents is necessary.
One widely used assessment instrument of major symptoms in OCD is the symptom checklist portion of the Yale-Brown Obsessive-Compulsive Scale (Y-BOCS, Goodman et al., 1989). This checklist covers a wide range of OCD symptoms clustered in the following manner: aggressive, contamination, sexual, hoarding, religious, symmetry, and somatic obsessions, cleaning, checking, repeating, counting, ordering/arranging, and hoarding/collecting compulsions. The symptom checklist is also used for assessing symptoms of OCD in children as part of the Children's Yale-Brown Obsessive-Compulsive Scale (CY-BOCS, Scahill et al., 1997), a downward extension of the adult measure, and includes items that are associated with OCD based on developmental considerations (i.e., checking toys and schoolwork). Dimensions of childhood obsessive-compulsive symptoms on the basis of the CY-BOCS checklist has not been well studied, although comparisons have been made in the symptom profiles of children with OCD in the presence or absence of comorbid tic disorders (George, Trimble, Ring, Sallee, & Robertson, 1993; Leckman, Walker, Goodman, Pauls, & Cohen, 1994; Leonard et al., 1992; Zohar et al., 1997).
In adults, there have been several efforts to determine OCD dimensions on the basis of symptom lists including the Y-BOCS checklist. The types of analyzes fall generally into two categories: factor analysis and cluster analysis. First, factor analytic studies. Baer (1994) found that the Y-BOCS symptom checklist yielded three distinct factors in a sample of 107 patients diagnosed with OCD. These factors were identified as symmetry/hoarding, contamination/checking, and pure obsessions. Leckman et al. (1997) examined the factor structure of the Y-BOCS checklist in two separate samples ( and 98). Although they noted a variable pattern of correlations between the dimensions defined in the checklist, they found that the checklist was composed of four factors in each sample, which accounted for over 60% of the variance. These factors were obsessions and checking, symmetry and ordering, cleanliness and washing, and hoarding. They also found that patients with chronic tic disorders or Tourette's syndrome had higher scores for the obsessions and checking, symmetry and ordering, and hoarding factors. Mataix-Cols et al. (1999) examined the factor structure of the Y-BOCS checklist using a sample of 354 OCD patients enrolled in a medication trial of serotonin reuptake inhibitors. These authors identified five factors that they named symmetry/ordering, hoarding, contamination/cleaning, aggressive/checking, and sexual/religious obsessions. In this sample, obsessions were further defined into separate factors, unlike prior research where these were subsumed into one factor.
Summerfeldt, Richter, Antony, and Swinson (1999) used confirmatory factor analysis to examine overall symptom groupings on the Y-BOCS symptom checklist. Using a sample of 203 individuals diagnosed with OCD, four different factor structures (from a single factor model, through to a four factor model) were tested. The best fit was obtained using the four-factor model, structured in similar manner to the factor structure identified by Leckman et al. (1997). These factors were obsessions/checking, symmetry/ordering, contamination/cleanliness, and hoarding. In summary, the research conducted thus far appears to support four major dimensions of OCD in adult samples. While the literature has shown consistency in the factor structure and subtyping of OCD in adult samples, it is unknown whether these same dimensions exist in children or if there is a different pattern of dimensions for the childhood manifestation of the condition. There is reason to hypothesize a similar arrangement of symptoms based upon clinical presentation of OCD in children and the apparent similarities to adults (Swedo et al., 1989). However, it is also possible that the manifestation and distribution of symptoms could vary in children given that there are developmental periods where ritualistic behavior is expected (Evans et al., 1997), creating the possibility that in some children obsessive-compulsive symptoms may represent a developmental delay rather than a manifestation of OCD per se.
Cluster analytic studies have been conducted using the Y-BOCS checklist as well. Cluster analysis is advantageous in that specific symptom groups may be determined with a reduction in overlap between dimensions. That is, individuals who primarily endorse symptoms in one category may form a distinct category, whereas factor analysis describes dimensions that are applicable to all members of the sample. One of the first studies to employ this procedure was Calamari, Weigartz, and Janeck (1999) who found five symptom-based cluster in a sample of 106 individuals diagnosed with OCD. These clusters were: harming, hoarding, contamination, certainty, and obsessions. In a replication study, Calamari et al. (2004) used cluster analysis with a sample of 114 individuals diagnosed with OCD. In this case, seven subgroups were identified: contamination, harming, hoarding, obsessions, symmetry, certainty, and contamination/harming. The authors concluded that the seven subgroup model was a better representation of symptoms as it was replicated when combined with the data from the Calamari et al.(1999) dataset.
One other cluster analytic study has been conducted using the symptom checklist of the Y-BOCS. Abramowitz, Franklin, Schwartz, and Furr (2003, study 1), with a sample of 132 individuals diagnosed with OCD, found five clusters, as follows: harming, contamination, hoarding, symmetry, and unacceptable thoughts. This pattern of subgroups conforms to the previous cluster analytic studies in the following ways. First, hoarding, harming, and contamination are all present in the three studies. Second, obsessions appear in both Calamari and colleagues studies, and are present in the Abramowitz et al. study in the form of unacceptable thoughts. Finally, while there were seven subgroups in the Calamari et al. (2004) paper, two of those may be considered more specific derivations of obsessions, namely symmetry and certainty.
In light of the literature reviewed, the aim of this study was to examine the factor structure of the checklist portion of the CY-BOCS in a sample of children and adolescents with OCD. We chose factor analysis over cluster analysis for the following reasons. First, cluster analysis, while providing evidence of subgroups of participants, can have limited generalizability for the groups identified. For example, out of the combined sample () in the Calamari et al. (2004) study, the hoarding subgroup had a sample of twelve. Second, factor analysis allows for the determination of dimensions along which symptoms may fall across all subjects, capturing aspects of the sample where individuals endorse symptoms in multiple domains. Given the unexplored nature of the structure of childhood obsessive-compulsive symptoms, and the aforementioned issues, factor analysis was considered the superior data analytic approach in this study. It was expected that the factor structure would be similar to that observed in the adult version of the Y-BOCS checklist.
Section snippets
Participants
The sample consisted of a consecutive series of children referred to one of two child OCD specialty clinics for evaluation and/or treatment of OCD. The two study sites were the New York Presbyterian Hospital-Weill Cornell Medical Center Outpatient Childhood OCD/Anxiety Disorders Program () and the UCLA Childhood OCD, Anxiety, and Tourette's Disorder Program (). All children completed an informed assent following completion of a parental/guardian informed consent to use data from the
Results
Principal components analysis showed that, when the number of factors was constrained to a single factor, only 18.88% of the variance was accounted for by the obtained factor. When the analyzes were constrained to two factors, the obtained factors accounted for 12.92% (obsessions) and 15.11% (compulsions) of the variance. For the three-factor model, 13.65% of the variance was accounted by a factor for obsessions, 14.44% was accounted for by compulsions, and 13.19% by a factor for
Discussion
While OCD has been conceptualized as a unitary condition in the diagnostic nomenclature, the heterogeneity of the condition is striking. While it appears that in adults there are identifiable distinct dimensions (Leckman et al., 1997; Mataix-Cols, et al., 1999; Summerfeldt, et al., 1999), this has not been investigated in children. The present study sought to determine whether there were symptom dimensions of OCD in children, and whether these conformed to the same pattern as identified in
References (30)
- et al.
Symptom subtypes of obsessive-compulsive disorder in behavioral treatment studies: A quantitative review
Behaviour Research and Therapy
(1996) - et al.
Obsessive-compulsive disorder subgroups: A symptom-based cluster analytic approach
Behaviour Research and Therapy
(1999) - et al.
Obsessive-compulsive disorder subtypes: An attempted replication and extension of a symptom-based taxonomy
Behaviour Research and Therapy
(2004) - et al.
A critical evaluation of obsessive-compulsive disorder subtypes: Symptoms versus mechanisms
Clinical Psychology Review
(2004) - et al.
Open trial of cognitive behavior therapy for childhood obsessive-compulsive disorder
Journal of Anxiety Disorders
(2002) - et al.
Child obsessive-compulsive disorder in the NIMH MECA study: Parent versus child indentification of cases
Journal of Anxiety Disorders
(2000) - et al.
Obsessions and compulsions across time in 79 children and adolescents with obsessive-compulsive disorder
Journal of the American Academy of Child and Adolescent Psychiatry
(1992) - et al.
Children's Yale-Brown Obsessive-Compulsive Scale: Reliability and validity
Journal of the American Academy of Child and Adolescent Psychiatry
(1997) - et al.
Symptom structure in obsessive-compulsive disorder: A confirmatory factor analytic study
Behaviour Research and Therapy
(1999) - et al.
Symptom presentation and outcome of cognitive-behavioral therapy for obsessive-compulsive disorder
Journal of Consulting and Clinical Psychology
(2003)
Diagnostic and statistical manual of mental disorders
Factor analysis of symptom subtypes of obsessive-compulsive disorder and their relation to personality and tic disorders
Journal of Clinical Psychiatry
Ritual, habit, and perfectionism: The prevalence and development of compulsive-like behavior in normal young children
Child Development
Factor analysis in the development and refinement of clinical assessment instruments
Psychological Assessment
Success and failure in the behavioral treatment of obsessive-compulsives
Journal of Consulting and Clinical Psychology
Cited by (62)
Connectome-wide Functional Connectivity Abnormalities in Youth With Obsessive-Compulsive Symptoms
2022, Biological Psychiatry: Cognitive Neuroscience and NeuroimagingPsychological, addictive, and health behavior implications of the COVID-19 pandemic
2020, Behaviour Research and TherapyThe relation between normative rituals/routines and obsessive-compulsive symptoms at a young age: A systematic review
2020, Developmental ReviewCitation Excerpt :Obsessive-Compulsive Disorder (OCD) is characterized by uncontrollable, intrusive and time-consuming thoughts or images (obsessions) and/or repetitive acts or mental rituals (compulsions) that interfere with daily functioning (American Psychiatric Association [APA], 2013). The content of these obsessions and compulsions generally involves themes of contamination, washing, aggression, symmetry, ordering, repeating and checking (Leckman, Zhang, Alsobrook, & Pauls, 2001; McKay et al., 2006). OCD was historically considered a rare condition in youth (Geller, 2006), until studies indicated that pediatric prevalence rates were considerably higher than initially assumed (Leonard, Ale, Freeman, Garcia, & Ng, 2005; Merlo & Storch, 2006).
Obsessive-Compulsive Symptomatology in Community Youth: Typical Development or a Red Flag for Psychopathology?
2019, Journal of the American Academy of Child and Adolescent Psychiatry
- ☆
Portions of these data were presented at the meeting of the World Congress of Behavioral and Cognitive Therapies, Vancouver, BC. We would like to thank Steven Taylor for very helpful comments on an earlier version of this manuscript.