Executive Function in Compulsive Hoarding with and without a Diagnosis of Obsessive Compulsive Disorder

Transcription

1 Diploma Thesis: Executive Function in Compulsive Hoarding with and without a Diagnosis of Obsessive Compulsive Disorder First examiner: Prof. Dr. Trevor W. Robbins Behavioural and Clinical Neurosciences Institute, University of Cambridge, United Kingdom Second examiner: Prof. Dr. Canan Basar-Eroglu Institute of Psychology and Cognition Research, University of Bremen, Germany Submitted by: Degree programme: Martina Butt Diploma in Psychology, University of Bremen Matriculation number: address: M.Butt@sms.ed.ac.uk Date: 24 th of March 2011

2 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r II "I feel at home here in this chaos because chaos suggests images to me" Irish painter Francis Bacon ( ) about his studio (Sylvester, 1995, p. 190)

3 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r III Acknowledgment I would like to express my sincere gratitude to all the research participants who took part in this project. Above all, I would like to thank the individuals suffering from compulsive hoarding for offering me their time, being so open-minded and never tiring of explaining to me their reasons for hoarding. I learned much during this study not only of my neuropsychological investigations, but also from the conversations I had with my participants. This experience significantly changed my view on hoarding in several ways. Furthermore, I would like to thank my supervisors Professor Trevor W. Robbins and Professor Canan Basar-Eroglu for their kind support, advice, and guidance during this research project and the preparation of this Diploma Thesis. Importantly, I would like to thank Professor Robbins for enabling me to come back to Cambridge several times in order to finish this project and for the organization of financial support. Also, I would like to thank several people who contributed to this research project in several ways: First of all, I would like to thank Dr. Sharon Morein-Zamir for her supervision and support. I am very grateful for her professional and personal guidance during all stages of this research project and the preparation of this thesis. Secondly, I would like to thank all people who shared their scientific ideas with me and/or helped recruiting participants and/or conducted psychiatric interviews. These are Professor Barbara J. Sahakian, Professor Naomi Fineberg, Dr. David Mataix-Cols, Alberto Pertusa, Dr. Danielle Landau, Dr. Alessandra Iervolino, Satwant Singh, and the charity organizations OCD action and OCD UK. Also, I am very grateful for the kind financial support of this project by the Medical Research Counsil (MRC) and the Wellcome Trust. Furthermore, I would like to thank Professor Helmut Hildebrandt for introducing me into the world of clinical neuropsychology in such a vivid way and for offering me the opportunity to present some preliminary results of this research project during the 8th Conference on Cognitive Neuropsychiatry in Bremen in His passion and enthusiasm for science is contagious. Finally, I thank my family and friends for their love and support. Especially, I need to thank Dr. Helen Miranda Knight and my fiancé Marcus Papmeyer for their critical remarks when proofreading this thesis, and for their valuable suggestions and advice.

4 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r IV CONTENTS Index of Tables... VIII Index of Figures... IXX Index of Abbreviations... X Abstract... XII 1 Introduction Contrasting Compulsive Hoarding with Obsessive Compulsive Disorder Clinical Characteristics Clinical Characteristics of Compulsive Hoarding Clinical Characteristics of Obsessive Compulsive Disorder Comparison of Clinical Characteristics Epidemiology Prevalence Prevalence of Compulsive Hoarding Prevalence of Obsessive Compulsive Disorder Comparison of Prevalence Comorbidity Comorbidity in Compulsive Hoarding Comorbidity in Obsessive Compulsive Disorder Comparison of Comorbidity Gender Distribution Gender Distribution in Compulsive Hoarding Gender Distribution in Obsessive Compulsive Disorder Comparison of Gender Distribution Age of Onset Age of Onset of Compulsive Hoarding Age of Onset of Obsessive Compulsive Disorder Comparison of Age of Onset Course of Illness Course of Illness in Compulsive Hoarding Course of Illness in Obsessive Compulsive Disorder Comparison of Course of Illness Treatment Response Pharmacological Treatment Response... 21

5 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r V Pharmacological Treatment Response in Compulsive Hoarding Pharmacological Treatment Response in Obsessive Compulsive Disorder Comparison of Pharmacological Treatment Response Psychotherapeutic Treatment Response Psychotherapeutic Treatment Response in Compulsive Hoarding Psychotherapeutic Treatment Response in Obsessive Compulsive Disorder Comparison of Psychotherapeutic Treatment Response Genetics Familial Aggregation Familial Aggregation in Compulsive Hoarding Familial Aggregation in Obsessive Compulsive Disorder Comparison of Familial Aggregation Twin Studies Twin Studies in Compulsive Hoarding Twin Studies in Obsessive Compulsive Disorder Comparison of Twin Studies Molecular Genetic Analyses Molecular Genetic Analyses in Compulsive Hoarding Molecular Genetic Analyses in Obsessive Compulsive Disorder Comparison of Molecular Genetic Analyses Summary Executive Function in Compulsive Hoarding and Obsessive Compulsive... Disorder Neural Correlates of Executive Function The Prefrontal Cortex Prefrontal Connectivity Dissociating Executive Functions Attentional Set-Shifting Planning Response Inhibition Neuropsychology of Compulsive Hoarding and Obsessive Compulsive... Disorder Neuropsychology of Compulsive Hoarding Attention Attentional Set-Shifting Memory... 45

6 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r VI Categorization Decision making Planning Response Inhibition Neuropsychology of Obsessive Compulsive Disorder Attention Attentional Set-Shifting Memory Categorization Decision making Planning Response Inhibition Neuropsychological Comparison of Compulsive Hoarding and Obsessive Compulsive Disorder Neuropsychological Research Hypotheses Methods Participants Recruitment Inclusion and Exclusion Criteria Procedure Measures Sociodemographic Measures Diagnostic Assessment Clinical Measures Personality Measures Neuropsychological Measures Attentional Set-Shifting Spatial Planning Response Inhibition Statistical Analysis Results Sociodemographic Characteristics Diagnostic Characteristics Clinical Characteristics... 76

7 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r VII 5.4 Personality Characteristics Neuropsychological Performance Attentional Set-Shifting Spatial Planning Response Inhibition Discussion Sociodemographic, Diagnostic, Clinical, and Personality Characteristics Neuropsychological Performance Limitations Future Research Directions References Appendix Declaration of Originality

8 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r VIII Index of Tables Table 1: Proposed Diagnostic Criteria for Hoarding Disorder Table 2: Diagnostic Criteria for Obsessive Compulsive Disorder Table 3: Sociodemographic Characteristics Table 4: Medication Status Table 5: Severity of Hoarding Symptoms Table 6: Current Comorbid Axis I Disorders Table 7: Severity and Types of Obsessive Compulsive Disorder Symptoms Table 8: Severity of Clinical Symptoms.. 79 Table 9: Personality Measures Table 10: Attentional Set-Shifting Performance Table 11: Spatial Planning Performance Table 12: Response Inhibition Performance.. 86 Table A1: Medication Status and Dose of Hoarding Participants Table A2: Transformed Variables Table A3: Levene s Test of Homogeneity of Variance of the Variables 123 Table A4: Shapiro-Wilk Test of Normality of Sample Distributions. 125

9 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r IX Index of Figures Figure 1: Prefronto - Subcortical Connectivity.. 37 Figure 2: The Intradimensional/Extradimensional Set-Shifting Task Figure 3: The Stockings of Cambridge Task. 69 Figure 4: The Stop Signal Task. 71 Figure 5: Descriptive Comparison of Respone Inhibition Performance... 87

10 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r X Index of Abbreviations ADHD ANOVA BA BDI CANTAB CBT CERQ COMT CPAS CVLT DSM DSM-IV-TR DSM-V EF GWAS fmri IED MADRS MCQ-30 MINI NART Attention-deficit/hyperactivity disorder Analysis of variance Brodmann area Beck Depression Inventory Cambridge Neuropsychological Test Automated Battery Cognitive Behavioural Therapy Cognitive Emotion Regulation Questionnaire Catechol-O-methyltransferase Compulsive Personality Assessment Scale California Verbal Learning Test Diagnostic and Statistical Manual of Mental Disorders Diagnostic and Statistical Manual of Mental Disorders, fourth text-revised edition Diagnostic and Statistical Manual of Mental Disorders, fifth edition Executive function Genome-wide association study Functional magnet resonance imaging Intradimensional/Extradimensional Set-Shifting Task Montgomery Asberg Depression Rating Scale Metacognitions Questionnaire, 30-item version Mini-International Neuropsychiatric Interview National Adult Reading Test

11 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r XI NTRK3 Neurotrophic tyrosine kinase receptor type 3 OCD OCI-R OCPD PET RCFT SI-R Obsessive compulsive disorder Obsessive Compulsive Inventory Revised Obsessive - compulsive personality disorder Positron emission tomography Rey-Osterrieth Complex Figure Test Saving Inventory Revised SLC1A1 Solute carrier family 1, member 1 SNP SPSS SOC SPECT SRI SSRI SST STAI ToL Tukey HSD WCST Y-BOCS YGTSS Single nucleotide polymorphism Statistical Package for the Social Sciences Stockings of Cambridge Task Single-photon emission computerized tomography Serotonin reuptake inhibitor Selective serotonin reuptake inhibitor Stop Signal Task State-Trait Anxiety Inventory Tower of London Task Tukey honestly significant difference test Wisconsin Card-Sorting Test Yale-Brown Obsessive Compulsive Scale Yale Global Tic Severity Scale

12 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r XII Abstract Introduction: Compulsive hoarding is a debilitating condition that is characterized by excessive acquiring and saving behaviour. Although compulsive hoarding has been commonly associated with obsessive compulsive disorder, several findings suggest that hoarding may form a clinically distinct disorder from OCD. This Diploma Thesis aims to contribute to this scientific debate by examining the differences and similarities of hoarding and OCD from a neuropsychological perspective. Methods: The neuropsychological performance on attentional set-shifting, spatial planning, and response inhibition tasks was compared between 20 OCD patients suffering from compulsive hoarding symptoms (OCD hoarding group), 20 hoarding participants who did not meet criteria for OCD (non-ocd hoarding group), and 20 control subjects (control group). Moreover, several diagnostic, clinical, and personality measures were assessed. Results: The OCD hoarding group, but not the non-ocd hoarding group, was significantly impaired on tasks of attentional set-shifting and response inhibition as compared to the control group. However, there were no differences in performance between both hoarding groups, indicating that hoarding may not be distinguishable from OCD on the basis of these neuropsychological tasks. Interestingly, the OCD hoarding group suffered from spatial planning deficits in comparison to the control group, with no differences in planning performance between both hoarding groups. Accordingly, spatial planning may form a key neuropsychological deficit in compulsive hoarding that has not been linked to obsessive compulsive disorder. Discussion: The finding that compulsive hoarding may not be distinguished from OCD with respect to attentional set-shifting and response inhibition performance, but planning deficits, is discussed with respect to other neuroscientific studies. The neuropsychology of planning is integrated into a neuroanatomical concept of compulsive hoarding that could potentially stimulate new research.

13 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r 1 1 Introduction The collecting and saving of items is a common human behaviour that is particularly prevalent among young children (Evans, et al., 1997; Landau, et al., 2011). From an evolutionary perspective, saving behaviour appears to be highly adaptive as it likely enhances the chances of survival and reproduction if resources become scarce (Grisham & Barlow, 2005; Polimeni, Reiss, & Sareen, 2005). However, in some individuals the urge to save and inability to discard items becomes excessive and appears to be maladaptive. This condition has been termed compulsive hoarding (e.g., Frost & Gross, 1993; Greenberg, 1987), excessive hoarding (e.g., Rachman, Elliott, Shafran, & Radomsky, 2009), or most recently, hoarding disorder (Landau, et al., 2011; Mataix-Cols, et al., 2010). Compulsive hoarding is a debilitating and chronic condition (Grisham, Frost, Steketee, Kim, & Hood, 2006). Major characteristics include the acquisition of and inability to discard excessive quantities of items, even if they appear to have no apparent value (Frost & Gross, 1993). Commonly saved items include old clothes, magazines, letters, pens, notes, bills, and newspapers (Pertusa, et al., 2008). As a consequence, severe hoarding results in cluttered living spaces that causes distress or impairment in daily functioning (Frost & Hartl, 1996). Moreover, the piles of accumulated items are associated with substantial health and safety risks as they can cause fire, falling or illness due to poor sanitation (Kim, Steketee, & Frost, 2001). Individuals who excessively hoard often suffer from marked occupational impairment, poor physical health, social impairment, and a low quality of life (Saxena, et al., in press; Tolin, Frost, Steketee, & Fitch, 2008). Furthermore, hoarding represents a profound economic and social burden for the society (Tolin, Frost, Steketee, Gray, & Fitch, 2008). Although reports about compulsive hoarding date back to 1947 when the Collyer Brothers were found dead in their cluttered home in New York (Weiss, 2010), scientific research into this condition has only been substantially undertaken during the last decade. As Frost and Hartl (1996) pointed out, hoarding was a little studied phenomenon (p. 341) about 15 years ago. Traditionally, hoarding has been associated with obsessive - compulsive disorder (OCD) and many scientists have proposed hoarding to be a symptom, manifestation or variant of OCD (e.g. Frost, Steketee, Williams, & Warren, 2000; Saxena, et al., 2002; Shafran & Tallis, 1996; Stein, Seedat, & Potocnik, 1999). Currently, OCD is listed as an

14 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r 2 anxiety disorder in the text-revised fourth edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-IV-TR; American Psychiatric Association, 2000). It is characterized by recurrent intrusive thoughts, images or impulses (obsessions) and/or repetitive, purposeful behaviour or mental acts (compulsions) that the person feels compelled to carry out in response to an obsession or according to rigid rules. Both obsessions and compulsions cause marked distress and significantly interfere with the person s life (American Psychiatric Association, 2000). The association of excessive hoarding with OCD presumably emerged on the grounds of case reports that found hoarding to appear in 18% to 42% of OCD patients (Mataix-Cols, Nakatani, Micali, & Heyman, 2008; Rasmussen & Eisen, 1992; Samuels, Bienvenu, et al., 2007). As a result, hoarding has been included as an OCD symptom dimension in the widely used OCD assessment instruments such as the Yale-Brown Obsessive Compulsive Scale (Y-BOCS; Goodman, et al., 1989) and the Obsessive Compulsive Inventory - Revised (OCI-R; Foa, et al., 2002). However, compulsive hoarding has neither been included in the diagnostic criteria nor description of OCD in the DSM-IV-TR (American Psychiatric Association, 2000). In contrast, it is listed as one of the eight diagnostic criteria for obsessive - compulsive personality disorder (OCPD) in the DSM-IV-TR and it is also briefly mentioned in the description of this disorder. Remarkably, one commentary in the criteria for OCPD in the DSM-IV-TR links hoarding to OCD: A diagnosis of obsessive - compulsive disorder should be considered especially when hoarding is extreme (American Psychiatric Association, 2000, p. 728). This remark is particularly confusing as hoarding is not part of the diagnostic criteria for OCD. Thus, the current diagnostic status of compulsive hoarding remains ambiguous. Recently, a controversy among scientists has emerged about the accuracy of a conceptualization and classification of hoarding within the context of OCD (e.g. Grisham, Brown, Liverant, & Campbell-Sills, 2005; Mataix-Cols, et al., 2010; Pertusa, et al., 2008; Steketee & Frost, 2003) and some scientists propose to include hoarding as a distinct disorder from OCD in the upcoming fifth edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-V) that could potentially be called hoarding disorder (Mataix- Cols, et al., 2010). On the basis of their sociodemographic and clinical research, Pertusa and colleagues (2008) suggest that hoarding is a clinically distinct syndrome from OCD in most of the cases. However, in a minority of individuals it may be a symptom of OCD

15 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r 3 when accompanied by distinct clinical features. This view has also been supported by a recent review of Mataix-Cols and colleagues (2010). Further insight into the nature of compulsive hoarding and its relation to OCD is of importance. First, it will affect the diagnostic classification scheme by establishing meaningful criteria for hoarding and thereby improve the identification of this condition. Second, it will potentially stimulate research and increase the public awareness for this condition. Finally, a better understanding of the nosology of hoarding might enhance the medical and psychological treatment of compulsive hoarders and thereby improve not only their quality of life, but also limit the economic and social burden hoarding causes to the society. This Diploma Thesis focuses on the ongoing debate about the relationship between hoarding and OCD. It aims to contribute to our understanding of both conditions, using a neuropsychological approach. First, compulsive hoarding and OCD will be contrasted with respect to clinical characteristics, epidemiological findings, treatment response, and underlying genetics. Second, the neuropsychology of hoarding and OCD will be presented with a main focus on executive function. Finally, methods and results from a conducted research project on executive function in compulsive hoarders with and without a formal diagnosis of OCD will be presented. It will shown if compulsive hoarding shares a similar neurocognitive profile with OCD, and thus, whether or not compulsive hoarding should be best conceptualized as a symptom (dimension) of OCD.

16 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r 4 2 Contrasting Compulsive Hoarding with Obsessive Compulsive Disorder The scientific controversy regarding the nature of compulsive hoarding and its association with OCD has stimulated research on this topic across a variety of academic disciplines, each employing different approaches to examine the similarities and differences between the two conditions. Although this thesis attempts to focus primarily on neuropsychological, and in particular executive function (EF), it is essential to investigate this debate from various perspectives. Therefore, compulsive hoarding and OCD will be contrasted in this chapter in terms of their clinical and epidemiological characteristics, as well as their response to treatment interventions and underlying genetic profile. 2.1 Clinical Characteristics Diagnoses of mental disorders are made on the basis of clinical symptoms that can either be observed in the behaviour of the patient or indirectly assessed through clinical interviews (Miller, 2010). Therefore, it is important to compare compulsive hoarding and OCD with respect to their clinical presentation, as the phenomenological similarities and differences between both conditions are most likely to influence the upcoming DSM-V criteria Clinical Characteristics of Compulsive Hoarding Hoarding has been initially described as the acquisition of and failure to discard possessions that appear to be useless or of limited value (Frost & Gross, 1993, p. 367). While this definition describes the core features of compulsive hoarding, i.e. excessive acquiring and saving behaviour, it lacks the specificity to distinguish pathological from non-pathological hoarding. Accordingly, Frost and Hartl (1996) made a first attempt to define more stringent clinical characteristics of compulsive hoarding. The authors proposed the following criteria: (a) the acquisition of, and failure to discard a large number of possessions that appear to be useless or of limited value; (b) living spaces sufficiently cluttered so as to preclude activities for which those spaces were designed; and (c) significant distress or impairment in functioning caused by the hoarding. These criteria have been widely used in research over the last decade (e.g. Grisham, Norberg, Williams, Certoma, & Kadib, 2010; Hartl, Duffany, Allen, Steketee, & Frost, 2005; Tolin, Kiehl,

17 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r 5 Worhunsky, Book, & Maltby, 2009) and have also been included in common hoarding assessment instruments such as the Saving Inventory Revised (SI-R; Frost, Steketee, & Grisham, 2004), the Hoarding Rating Scale (Tolin, Frost, & Steketee, 2010), and the UCLA Hoarding Severity Scale (Saxena, Brody, Maidment, & Baxter, 2007). Recently, Mataix-Cols and colleagues (2010) adapted Frost and Hartl s (1996) original definition and proposed the following diagnostic criteria of compulsive hoarding for inclusion in DSM-V: (a) persistent difficulty discarding personal belongings even if they are useless or of limited value due to strong saving urges, distress and/or indecisiveness regarding discarding; (b) accumulation of possessions that clutter the active living areas and prevent normal use of the space, if the living space is uncluttered then only because of the efforts of others, e.g. family members; (c) distress or impairment in functioning due to the symptoms; and (d) the hoarding symptoms are not caused by a general medical condition. Owing to the fact that the authors developed these criteria to allow for a diagnosis of hoarding disorder, i.e. a disorder distinct from OCD, they further included the diagnostic criterion that the hoarding behaviour may not be restricted to the symptoms of another mental disorder (e.g. due to obsessions in OCD; see Table 1). Individuals with compulsive hoarding mainly save common items such as letters, magazines, old clothes or newspapers (Seedat & Stein, 2002). However, research has indicated that a minority of individuals hoard bizarre items such as faeces, urine, nails, hair, used diapers, or rotten food (Pertusa, et al., 2008). When asked about their reasons for saving, most individuals indicate that the hoarded items are of intrinsic or emotional value to them. Thus, they either regard the saved items as valuable or think that they might be needed in the future, or they feel sentimentally attached to their possessions (Steketee, Frost, & Kyrios, 2003). On the contrary, a minority of hoarders report keeping items for other reasons than that: they fear that something bad will happen if they discard things, they need to check that no important items are discarded, or they need to throw things away in certain lucky numbers (Pertusa, et al., 2008). Remarkably, the acquiring and storing behaviour in compulsive hoarding is usually accompanied by positive or neutral emotions; only the result of this behaviour, the unmanageable clutter or the intervention of third parties such as family members or local authorities causes emotional distress (Mataix-Cols, et al., 2010; Rachman, et al., 2009). Impairment and suffering only often increase late in the course of the disorder (Tolin, Meunier, Frost, & Steketee, 2010a), presumably because the clutter has accumulated

18 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r 6 excessively by then. Compulsive hoarders usually regard their saving behaviour as reasonable and have a limited insight into their condition (Greenberg, 1987; Rachman, et al., 2009). Table 1 Proposed Diagnostic Criteria for Hoarding Disorder (A) Persistent difficulty discarding or parting with personal possessions, even those of apparently useless or limited value, due to strong urges to save items, distress, and/or indecision associated with discarding. (B) The symptoms result in the accumulation of a large number of possessions that fill up and clutter the active living areas of the home, workplace, or other personal surroundings (e.g., office, vehicle, yard) and prevent normal use of the space. If all living areas are uncluttered, it is only because of others efforts (e.g., family members, authorities) to keep these areas free of possessions. (C) The symptoms cause clinically significant distress or impairment in social, occupational, or other important areas of functioning (including maintaining a safe environment for self and others). (D) The hoarding symptoms are not due to a general medical condition (e.g., brain injury, cerebrovascular disease). (E) The hoarding symptoms are not restricted to the symptoms of another mental disorder (e.g., hoarding due to obsessions in Obsessive Compulsive Disorder (OCD), lack of motivation in Major Depressive Disorder, delusions in Schizophrenia or another Psychotic Disorder, cognitive deficits in Dementia, restricted interests in Autistic Disorder, food storing in Prader Willi Syndrome). These criteria have been proposed by Mataix-Cols et al. (2010) for inclusion in the fifth edition of the Diagnostic and Statistical Manual of Mental Disorders Clinical Characteristics of Obsessive Compulsive Disorder In the DSM IV-TR (American Psychiatric Association, 2000), OCD is classified as an anxiety disorder that is characterized by obsessions and/or compulsions (see Table 2 for DSM-IV-TR criteria of OCD). Obsessions are defined as recurrent intrusive thoughts, impulses or images that are experienced as anxiety-provoking or distressing and are not simply excessive worries about ordinary problems. Individuals with OCD try to ignore or suppress these thoughts or engage in other thoughts or compulsive actions to neutralize them (American Psychiatric Association, 2000). Typical examples of obsessions in OCD include worries about contamination (e.g., excessive thoughts that the hands are dirty) or

19 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r 7 doubts about the individual s safety (e.g., that the stove has not been turned off; Pine, 2009). Compulsions are defined as repetitive mental or behavioural rituals that an individual feels compelled to perform in response to an obsession or according to rigid rules. These rituals are meant to reduce distress and anxiety or to prevent imagined feared events or situations from occurring (American Psychiatric Association, 2000). Examples of common compulsions in OCD include excessive hand washing to reduce obsessions about contamination, or repetitive checking behaviour in response to excessive doubts about personal safety. Notably, compulsions can also be manifested as thoughts, for instance by repeating a silent prayer to relieve the obsessional anxiety of having committed a sin (Pine, 2009). Individuals who suffer from OCD generally experience their symptoms as distressing or time consuming, and/or indicate that they are interfering with their occupational or social functioning. The patients typically have insight into their condition and realize that their obsessions or compulsions are irrational or excessive (American Psychiatric Association, 2000). The clinical presentation of OCD is heterogeneous with great variability in symptoms. Various factor and cluster analytical studies on the symptom structure of OCD have been conducted to identify more homogeneous subgroups of OCD. Most structural analyses made use of the Y-BOCS (Goodman, et al., 1989), a widely used clinical rating scale to assess the severity of OCD symptoms. The Y-BOCS contains 13 major OCD symptom categories in its checklist, and clinicians are asked to indicate the presence of each symptom in a patient and to rate its severity accordingly. Seven categories of obsessions (aggression, contamination, sexual, hoarding, religious, symmetry, somatic) and six categories of compulsions (cleaning, checking, repeating, counting, ordering, hoarding) can be differentially assessed, including hoarding. In a recent meta-analysis of 21 factor analytical studies including 5124 OCD patients, the following four principal OCD symptom dimensions were identified (Bloch, Landeros- Weisenberger, Rosario, Pittenger, & Leckman, 2008): (a) symmetry (symmetry obsessions and repeating, ordering, and counting compulsions); (b) forbidden thoughts (aggression, sexual, religious, and somatic obsessions and checking compulsions); (c) cleaning (cleaning compulsions and contamination obsessions); and (d) hoarding (hoarding obsessions and compulsions). The finding that hoarding constitutes an independent factor

20 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r 8 is in line with the majority of other factor analytical studies (e.g., Cavallini, Di Bella, Siliprandi, Malchiodi, & Bellodi, 2002; Leckman, et al., 1997; Pinto, et al., 2007) and also cluster analytical approaches (e.g., Calamari, Wiegartz, & Janeck, 1999; Hasler, et al., 2005), although it should be noted that some studies have not confirmed this result (e.g., Baer, 1994; Katerberg, et al., 2010). The common identification of a separate hoarding factor among OCD symptoms cannot be taken as evidence to suggest that hoarding is a distinct disorder from OCD. It only proves that hoarding symptoms are distinguishable from typical OCD symptoms assessed but does not address to what extend hoarding and other identified OCD symptom dimensions are related to each other. To examine this latter issue, Abramowitz, Wheaton and Storch (2008) conducted a principal component factor analysis on the extent of OCD symptoms in a combined sample of OCD patients, individuals with other anxiety disorders other than OCD, and students. The authors specified a single factor solution and found strong factor loadings of typical OCD symptoms such as checking or washing, but the weakest factor loading for symptoms of hoarding on the unitary factor. Moreover, the authors reported that the non-hoarding OCD symptoms correlated more strongly with each other than they correlated with hoarding. However, these results were statistically not significant.

21 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r 9 Table 2 Diagnostic Criteria for Obsessive Compulsive Disorder (A) Either obsessions or compulsions: Obsessions as defined by (1), (2), (3), and (4): (1) recurrent and persistent thoughts, impulses, or images that are experienced, at some time during the disturbance, as intrusive and inappropriate and that cause marked anxiety or distress (2) the thoughts, impulses, or images are not simply excessive worries about real-life problems (3) the person attempts to ignore or suppress such thoughts, impulses or images, or to neutralize them with some other thought or action (4) the person recognizes that the obsessional thoughts, impulses, or images are a product of his or her own mind (not imposed from without as in thought insertion) Compulsions are defined by (1) and (2): (1) repetitive behaviours (e.g., hand washing, ordering, checking) or mental acts (e.g., praying, counting, repeating words silently) that the person feels driven to perform in response to an obsession, or according to rules that must be applied rigidly (2) the behaviours or mental acts are aimed at preventing or reducing distress or preventing some dreaded event or situation; however, these behaviours or mental acts either are not connected in a realistic way with what they are designed to neutralize or prevent or are clearly excessive (B) At some point during the course of the disorder, the person has recognized that the obsessions or compulsions are excessive or unreasonable. (C) The obsessions or compulsions cause marked distress, are time consuming (take more than 1 hour a day), or significantly interfere with the person s normal routine, occupational (or academic) functioning, or usual social activities or relationships. (D) If another Axis I disorder is present, the content of the obsessions or compulsions is not restricted to it (e.g., preoccupation with food in the presence of an Eating Disorder; hair pulling in the presence of Trichotillomania; concern with appearance in the presence of Body Dysmorphic Disorder; preoccupation with drugs in the presence of a Substance Use Disorder; preoccupation with having a serious illness in the presence of Hypochondrias; preoccupation with sexual urges or fantasies in the presence of Paraphilia; or guilty ruminations in the presence of Major Depressive Disorder). (E) The disturbance is not due to the direct physiological effects of a substance (e.g., a drug of abuse, a medication) or a general medical condition. Adapted from the Diagnostic and Statistical Manual of Mental Disorders, fourth text-revised edition (American Psychiatric Association, 2000).

22 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r Comparison of Clinical Characteristics The clinical characteristics of compulsive hoarding and OCD clearly share some similarities. For example, the intensive fear of losing important items resembles obsessions, and the urges to save possessions share some overlap with compulsions. However, several phenomenological differences between hoarding and OCD symptoms exist. Therefore, it remains questionable if hoarding related thoughts and actions resemble obsessions and compulsions as observed in OCD. First of all, hoarding related thoughts typically involve an intense and pervasive interest in saving belongings and/or adding new items to one s collection (Rachman, et al., 2009). When confronted with discarding possessions, individuals who suffer from compulsive hoarding often experience the fear of losing something important that might be needed in the future, or they feel sentimentally attached to their possessions (Steketee, et al., 2003). These hoarding-related thoughts resemble obsessions because they can be recurrent and intense. However, they are not repetitive in the same way as typical obsessions (Steketee, et al., 2003) and are not experienced as intrusive, anxiety-provoking or distressing by the individual (Frost & Gross, 1993; Frost, Hartl, Christian, & Williams, 1995; Rachman, et al., 2009). Moreover, individuals who suffer from compulsive hoarding often do not make attempts to ignore these thoughts (Rachman, 1973), and importantly, the distress that can be observed in compulsive hoarding individuals does not come from the hoarding related thoughts themselves, but is the result of the unmanageable clutter that has been accumulated after years of hoarding (Mataix-Cols, et al., 2010). Thus, typical hoarding related thoughts do not meet the criteria for obsessions as defined in the DSM- IV-TR (American Psychiatric Association, 2000). Second, hoarding behaviour is characterized by the avoidance of discarding possessions and/or excessive acquisition of items accompanied by strong repetitive urges to save them that are difficult to resist, and therefore resemble compulsions. However, hoarding individuals do not perform stereotyped rituals as do OCD patients (Pertusa, Frost, & Mataix-Cols, 2010). Furthermore, the avoidance of discarding things is a relatively passive behaviour. This contrasts with typical OCD compulsions such as excessive hand washing or cleaning that are clearly performed in an active way. Remarkably, compulsive hoarders experience the accumulation or saving of items as emotionally positive or neutral, while individuals with OCD experience their symptoms as distressing and emotionally negative (Rachman, et al., 2009). The distress that individuals with compulsive hoarding

23 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r 11 experience is caused by the unmanageable clutter, but not the storing or acquiring behaviour itself (Rachman, et al., 2009). Finally, compulsive hoarders often do not have insight into their condition and regard their saving behaviour as reasonable (Greenberg, 1987). This further contrasts with clinical OCD characteristics as individuals with OCD are supposed to realize at least at one point during their illness that their obsessions or compulsions are excessive or unrealistic. Thus, hoarding also does not meet the criteria for compulsions as defined in the DSM-IV-TR (American Psychiatric Association, 2000). As stated earlier, a minority of individuals who suffer from compulsive hoarding save bizarre items and their hoarding is accompanied by fears that something bad will happen if they discard things, the need to check that no important items are discarded, or the urge to throw things away in certain lucky numbers (Pertusa, Frost, & Mataix-Cols, 2010). These hoarding-related thoughts and actions appear to resemble obsessions and compulsions because they show a strong link to typical OCD symptoms, i.e. the fear of catastrophic consequences, the need for symmetry or order, and the need to perform checking rituals. However, no research has been conducted yet to examine if the hoardingrelated thoughts and actions of this subgroup do meet the criteria for OCD. In conclusion, compulsive hoarding and OCD share some clinical characteristics, and research indicates that hoarding might form a symptom of OCD in a minority of the cases when hoarding is accompanied by typical OCD-related symptoms. However, in the majority of individuals, hoarding-related thoughts and actions clearly differ from typical OCD obsessions and compulsions. Thus, compulsive hoarding symptoms do not generally meet the diagnostic criteria for OCD according to the DSM-IV-TR classification scheme. This finding is also in accordance with the presented structural analyses on the symptom dimensions of OCD which have shown that hoarding commonly forms a separate factor or cluster and is correlated more weakly with typical OCD symptoms than these are intercorrelated. 2.2 Epidemiology In this section, epidemiological findings for compulsive hoarding and OCD will be presented and contrasted. It will be shown if there is evidence from an epidemiological perspective to suggest that compulsive hoarding is a distinct disorder from OCD.

24 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r Prevalence Several epidemiological studies have been conducted to estimate the point and/or lifetime prevalence rates for compulsive hoarding and OCD in the population. Therefore, these prevalence estimates for both conditions will be contrasted. If hoarding was to be classified as a subtype of OCD, it should be expected to occur less commonly in the population over a certain period of time than all OCD subtypes combined Prevalence of Compulsive Hoarding Epidemiological studies indicate that compulsive hoarding is highly prevalent in the population with point prevalence estimates ranging from 2.3% (Iervolino, et al., 2009) and 3.7% (Samuels, et al., 2008) to 4.6% (Mueller, Mitchell, Crosby, Glaesmer, & de Zwaan, 2009). Thus far, lifetime prevalence estimates for excessive hoarding have only been published within the context of individuals diagnosed with OCD. Interestingly however, one study found that as many as 14.4% of the population experience hoarding symptoms at one point in their life (Ruscio, Stein, Chiu, & Kessler, 2010). These epidemiological estimates need to be interpreted with caution. First, the research conducted by Samuels et al. (2008) and Mueller et al. (2009) did not apply the stringent criteria of compulsive hoarding that have been developed by Frost and Hartl (1996). Therefore, it remains unknown if these studies identified individuals with clinically significant compulsive hoarding only. Second, the identification of pathological hoarding in the research projects of Iervolino et al. (2009) and Mueller et al. (2009) was based on self-report measures. Self-report measures are potential confounders when estimating prevalence rates as the diagnosis is purely based on a cut-off score in a questionnaire and is not made by a clinical psychologist or psychiatrist. Therefore, it is likely that some individuals who took part in these studies were either incorrectly classified as suffering from pathological hoarding or incorrectly classified as not suffering from this condition. Third, Iervolino et al. (2009) included 90% women and only 10% men in their sample. Accordingly, the prevalence estimate derived from this study is limited to the female population. Finally, the research design employed by Mueller et al. (2009) included a home visit to hand over questionnaires to participants. Due to the clutter in their houses and the associated fear of inviting people to their homes, it is likely that many hoarding individuals refused to take part in this research which in turn may have resulted in underestimation of prevalence rates.

25 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r Prevalence of Obsessive Compulsive Disorder In comparison to excessive hoarding, lower prevalence rates for OCD have been reported. Point prevalence estimates range from approximately 1.0% to 1.2% (Kessler, Chiu, Demler, Merikangas, & Walters, 2005; Ruscio, et al., 2010; Torres, et al., 2006b), and lifetime prevalence rates vary between 1% and 3% (Rasmussen & Eisen, 1992; Ruscio, et al., 2010). Each of these studies used good methodological designs with large sample sizes and face-to-face clinical interviews. The response rate in these surveys, i.e. the percentage of individuals who took part after being invited, was approximately 70%. This might have resulted in either an over- or underestimation of the actual prevalence rate as 30% of the invited participants did not take part Comparison of Prevalence Although the epidemiological studies on compulsive hoarding have several methodological limitations, there is some evidence to suggest that compulsive hoarding is more prevalent in the population than all of the different OCD subtypes combined. This finding contradicts with the assumption of excessive hoarding being a symptom (dimension) of OCD. However, the prevalence of compulsive hoarding needs to be examined in studies with good methodological designs to verify this point Comorbidity Mental disorders in general are highly comborbid with each other (Kessler, et al., 2005). Research indicates that this is also true for both compulsive hoarding and OCD. Therefore, the specific patterns of comorbidity for both conditions will be examined in this section. It can be argued that if excessive hoarding were a symptom (dimension) of OCD, it should occur frequently within the context of typical OCD symptom dimensions such as washing or checking as most OCD patients suffer from multiple OCD symptoms (Mataix- Cols, Rauch, et al., 2002). For example, in one epidemiological study that examined the life-time prevalence rate of OCD, only 19% of the individuals diagnosed with OCD were monosymptomatic, i.e. their obsessions and/or compulsions were limited to one OCD symptom dimension. Accordingly, 81% of individuals who met with a life-time diagnosis of OCD experienced multiple types of obsessions and compulsions (Ruscio, et al., 2010).

26 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r 14 Likewise, one can expect hoarding to be more frequently comorbid with other mental disorders than OCD if hoarding were to be classified as a distinct disorder from OCD Comorbidity in Compulsive Hoarding Compulsive hoarding has been observed frequently in a variety of clinical conditions. As noted earlier, hoarding behaviour has been closely linked to OCD because it has been found to occur in 18% to 42% of OCD patients (Lochner, et al., 2005; Steketee & Frost, 2003; Wheaton, Timpano, Lasalle-Ricci, & Murphy, 2008), but only forms the primary symptom in approximately 4-11% of these individuals (Mataix-Cols, Rauch, Manzo, Jenike, & Baer, 1999; Saxena, et al., 2002). This finding indicates that compulsive hoarding often occurs within the context of OCD, thus being either comorbid with OCD, or if compulsive hoarding was to be classified as a symptom of OCD, that the symptom dimension hoarding can be observed frequently within OCD patients. In contrast, in a recent study containing a sample-size of 271 individuals with clinically significant hoarding, only 18% were diagnosed with OCD (Frost, Steketee, Tolin, & Glossner, 2010), and in one epidemiological study that included a community-based sample of 742 participants, none of the identified hoarders met with the criteria for OCD (Samuels, et al., 2008). This implies that the majority of individuals suffering from compulsive hoarding do not meet criteria for a diagnosis of OCD, thus demonstrating that either compulsive hoarding does exist in the absence of OCD or that the current diagnostic criteria for OCD according to the DSM-IV-TR (American Psychiatric Association, 2000) should be modified to include this condition. Beyond its association with OCD, excessive hoarding has been documented in various forms of anxiety disorders (Tolin, Meunier, Frost, & Steketee, 2010b), mood disorders (Ayers, Saxena, Golshan, & Wetherell, 2010), eating disorders (Frankenburg, 1984), various types of dementia (Hwang, Tsai, Yang, Liu, & Lirng, 1998), schizophrenia (Stein, Laszlo, Marais, Seedat, & Potocnik, 1997), and Parkinson s disease (O'Sullivan, et al., 2010). Moreover, hoarding has been observed within the context of personality disorders (Samuels, et al., 2008), and following brain injury (Hahm, Kang, Cheong, & Na, 2001; Volle, Beato, Levy, & Dubois, 2002). Research on comorbidity estimates is rare and relies on small sample sizes. However, there is some evidence to conclude that compulsive hoarding is particularly high comorbid with mood disorders and anxiety disorders. For instance, in a research project on 18 elderly compulsive hoarders, Ayers, Saxena, Golshan

27 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r 15 and Wetherell (2010) reported a comorbidity rate of 28% for major depressive disorder, followed by 22% for dysthymia. In contrast, Grisham, Brown, Savage, Steketee, and Barlow (2007) reported a somewhat smaller comorbidity rate of nearly 17% for major depressive disorder in their sample of 30 compulsive hoarders. Moreover, the authors reported social phobia to appear in approximately 33% of the hoarding individuals, followed by panic disorder with agoraphobia in 10%, and generalized anxiety disorder in nearly 7% of the individuals. In addition, a recent study by Tolin et al. (2010b) examined the prevalence of compulsive hoarding in a sample of 130 individuals suffering from anxiety disorders. The authors found clinically significant compulsive hoarding in 29% of individuals diagnosed with generalized anxiety disorder, and in 15% of individuals with social phobia. Although research on comorbidity rates of compulsive hoarding is sparse and relies on small sample sizes, these results suggest that compulsive hoarding is not particularly associated with OCD as it appears as often within the context of other anxiety disorders and/or mood disorders Comorbidity in Obsessive Compulsive Disorder Like compulsive hoarding, OCD is associated with substantial comorbidity rates for various mental disorders. For example, Ruscio et al. (2010) reported that 90% of their participants diagnosed with lifetime OCD met criteria for another lifetime DSM-IV-TR mental disorder. Other studies have found comorbidity to occur in 46% to 62% of individuals diagnosed with OCD (Denys, Tenney, van Megen, de Geus, & Westenberg, 2004; Torres, et al., 2006b). Most notably, OCD has been found to be highly comorbid with mood disorders, other anxiety disorders, impulse-control disorders, and personality disorders. Major depressive disorder is one of the most commonly observed mood disorder within the context of OCD as it occurs in approximately 22% to 37% of the cases (Brown, Campbell, Lehman, Grisham, & Mancill, 2001; Torres, et al., 2006b). Amongst anxiety disorders, social phobia has been found to occur in 17% to 25% of OCD patients, specific phobia in 15% to 17%, generalized anxiety disorder in about 12% to 31%, and panic disorder has been documented to co-occur with OCD in approximately 9% to 22% of the cases (Brown, et al., 2001; Torres, et al., 2006b). Moreover, a higher rate of personality

28 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r 16 disorders varying between 16% to 75% has been observed in OCD patients (Bejerot, Ekselius, & von Knorring, 1998; Denys, et al., 2004; Torres, et al., 2006a) Comparison of Comorbidity To summarise, compulsive hoarding has been frequently observed within the context of OCD. However, the vast majority of individuals suffering from excessive hoarding do not meet a diagnosis of OCD. Both conditions are highly comorbid with other clinical conditions, especially (other) anxiety and mood disorders. Research has shown that compulsive hoarding appears not to be particularly associated with OCD as hoarding occurs as often within the context of other anxiety disorders and/or mood disorders Gender Distribution Some epidemiological research has examined if compulsive hoarding and OCD are equally prevalent among men and women Gender Distribution in Compulsive Hoarding Although most clinical research studies on compulsive hoarding demonstrated a gender bias in recruitment, i.e. more female than male participants (e.g., Frost, Tolin, Steketee, Fitch, & Selbo-Bruns, 2009; Grisham, et al., 2010; Saxena, et al., in press), two epidemiological studies found significantly higher prevalence rates for compulsive hoarding amongst men than women (Iervolino, et al., 2009; Samuels, et al., 2008). Similarly, one epidemiological study by Mueller et al. (2009) reported a higher prevalence of hoarding symptoms amongst male than female individuals. However, this result was not significant. Owing to the difference in recruitment and prevalence rates in the clinical and epidemiological studies, it remains unknown if compulsive hoarding is equally common among men and women. More research into the gender distribution of excessive hoarding is clearly needed Gender Distribution in Obsessive Compulsive Disorder Epidemiological studies of OCD have reported women to be more often affected by this condition than men (Bebbington, 1998; Denys, et al., 2004; Torres, et al., 2006b). However, some authors argue that this finding might only be a reflection of gender

29 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r 17 differences in help-seeking behaviour, in the degree to which the disorder is tolerated by family members, in occupational impairment, and in career expectations (Castle & Groves, 2000; Lochner & Stein, 2001). Moreover, an equal gender distribution of OCD is assumed in DSM-IV-TR (American Psychiatric Association, 2000). Accordingly, it remains controversial if OCD is more common among female then male individuals, or if both gender are equally often affected. Further research on this topic would be of interest Comparison of Gender Distribution In sum, epidemiological research on compulsive hoarding indicates that this condition is more prevalent amongst men than women. This finding is at odds with clinical studies that commonly recruit more female than male hoarding participants. Therefore, the gender distribution of compulsive hoarding remains ambiguous. The same can be concluded for OCD. Some studies have suggested that more women might be affected by OCD than men. However, it has been hypothesized that this may only reflect a general higher tendency in women to seek help, rather than a real difference in gender distribution. Accordingly, more research on this topic is clearly needed Age of Onset Epidemiological research has also examined the mean age of onset of hoarding and OCD, respectively. The age of onset of a mental disorder is thought to reflect the influence of genetic and/or environmental factors in its pathogenesis (O'Connor & Plomin, 2000). If hoarding was a symptom (dimension) of OCD, both conditions are expected to have a similar mean age of onset Age of Onset of Compulsive Hoarding Epidemiological studies investigating the age of onset of compulsive hoarding are rare and rely on small sample sizes and retrospective clinical interviews. Studies consistently report that mild hoarding symptoms mainly start during childhood or early adolescence (Ayers, et al., 2010; Frost & Gross, 1993; Grisham, Frost, Steketee, Kim, & Hood, 2003; Tolin, Meunier, et al., 2010a). However, symptoms only interfere with the individual s everyday functioning around the age of 35 (Frost, et al., 2009; Grisham, et al., 2006). Therefore, the

30 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r 18 mean age of onset of clinically significant compulsive hoarding appears to be around the age of 35, with mild symptoms starting much earlier Age of Onset of Obsessive Compulsive Disorder Studies have consistently shown that the average onset of OCD lies between the ages of 20 and 22 (Brown, et al., 2001; Denys, et al., 2004; Ruscio, et al., 2010). However, there appears to be a gender difference as men have on average an earlier onset of the disorder than women. Research has shown that males develop OCD more often before the age of 10, while females develop this condition more often during or after puberty (Lochner & Stein, 2001; Ruscio, et al., 2010). In contrast, little is known about the age of onset of subclinical OCD. This is presumably owing to the fact that it remains controversial if individuals who experience mild OCD symptoms later in their life develop clinically significant OCD (Fogel, 2003) Comparison of Age of Onset In conclusion, the mean age of onset of clinically significant compulsive hoarding occurs around the age of 35, and thus approximately 10 years later than the average age of onset of OCD. A potential explanation for the relatively late onset of hoarding might be the fact that it takes several years to accumulate a sufficient number of items that they occupy a substantial amount of space, and hence start to interfere with the individual s daily life. In contrast, most hoarding individuals experience subclinical symptoms as early as during childhood or early adolescence, and these symptoms gradually worsen over time. Owing to the fact that it remains controversial if OCD commonly emerges on the grounds of subclinical symptoms, a comparison between conditions on the age of onset of subclinical symptoms cannot be made Course of Illness Several studies have examined the course of illness of compulsive hoarding and OCD over a period of two to three years, or even 40 years. Accordingly, these findings will be summarised and compared.

31 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r Course of Illness in Compulsive Hoarding Although no prospective longitudinal studies with reasonable sample sizes examining the course of compulsive hoarding have been carried out so far, research based on retrospective assessment has consistently reported that compulsive hoarding runs a chronic and progressive course of illness as the severity of symptoms generally increases over time (Ayers, et al., 2010; Seedat & Stein, 2002; Tolin, Meunier, et al., 2010a). Only very a few sufferers report a decreasing course of illness (Grisham & Barlow, 2005; Grisham, et al., 2003). Tolin et al. (2010a) surveyed the course of illness among 751 compulsive hoarders who took part in an online evaluation. The authors found a decreasing course of illness in only 0.7% of their sample. Similarly, another study that followed the course of hoarding in OCD patients for two years (Eisen, et al., 2010) found hoarding to be associated with a three- to fourfold decreased remission rate in comparison to typical OCD subtypes: only one out of 19 (5.3%) hoarding participants experienced a partial remission within the twoyear period. Interestingly, the first symptoms individuals with compulsive hoarding experience tend to be a difficulty in discarding items, followed by the accumulation of clutter. Excessive acquisition appears to occur only later during the course of the disorder (Grisham, et al., 2006; Pertusa, et al., 2008). Mataix-Cols et al. (2010) argue that the later onset of acquisition symptoms during the course of illness might reflect greater financial and physical independence that individuals experience as they get older, thus enabling them to buy significantly more items and to store them accordingly in their own homes Course of Illness in Obsessive Compulsive Disorder Prospective longitudinal studies examining the course of OCD have been mainly carried out within the context of pharmacological and/or cognitive behavioural therapeutic interventions. Studies that followed the course of illness over a period of two or three years report partial remission of OCD symptoms to occur in 24% to 47%, and full recovery in 6% to 48% of the individuals (Catapano, et al., 2006; Eisen, et al., 1999; Eisen, et al., 2010; Reddy, Alur, Manjunath, Kandavel, & Math, 2010). Some of these studies have also indicated that about half of the patients with a significant reduction in the symptom severity experience a subsequent relapse (Catapano, et al., 2006; Eisen, et al., 1999), while others report the probability of a relapse to be very low (Eisen, et al., 2010).

32 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r 20 To study the long-term course of OCD, Skoog and Skoog (1999) followed 144 OCD patients over a time period of 40 years. They observed full or partial remission in 83% of their sample, with 48% recovering completely or experiencing subclinical symptoms only. Although there is some inconsistency in the literature about the probability rate of OCD patients to experience full or at least partial recovery, reported studies indicate that the severity of OCD symptoms decreases in a significant number of patients over a period of two to three years, and that about half of them recover over a longer period of time. There also appears, however, to be a substantial number of OCD patients who do not improve or suffer from a relapse after remission, despite the fact that they received a pharmacological and/or behavioural therapeutic treatment. Importantly, approximately 17% of OCD patients still suffer from this condition after more than 40 years of illness (Skoog & Skoog, 1999). It can be concluded therefore that OCD runs a chronic and persistent course of illness in a substantial amount of individuals. However, in approximately half of the cases full or partial remission occurs within some years, and the majority improves within decades. There is also some evidence to suggest that the course of OCD fluctuates in some individuals, with relapses following remissions and vice versa Comparison of Course of Illness In summary, compulsive hoarding clearly runs a chronic and progressive course of illness with very few individuals experiencing an improvement in hoarding symptoms over time. In contrast, OCD is associated with much higher partial and full remission rates over time, although a substantial amount of patients do not recover within decades. Even though there is some evidence to suggest that compulsive hoarding runs more often a progressive and chronic course, these results should be interpreted carefully as most individuals with compulsive hoarding did not receive any form of treatment, while the opposite is true for OCD patients. 2.3 Treatment Response During the last decades, much research has focused on the efficacy of pharmacotherapy and psychotherapy in treating OCD (Marazziti & Consoli, 2010). By contrast, only a few studies have been conducted to examine the treatment response of individuals suffering from compulsive hoarding. If hoarding were to be classified as a symptom of OCD, it

33 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r 21 would be expected that hoarding patients would benefit similarly to non-hoarding OCD patients from pharmacological and psychotherapeutic interventions Pharmacological Treatment Response Patients suffering from OCD are commonly treated with serotonin reuptake inhibitors (SRIs) which have been shown to ameliorate OCD symptoms in a substantial number of individuals (Marazziti & Consoli, 2010). This pharmacological approach has also been employed to treat compulsive hoarders with or without a diagnosis of OCD. Accordingly, these findings will be summarised and compared in this section Pharmacological Treatment Response in Compulsive Hoarding Research on the efficacy of pharmacological interventions for treating compulsive hoarding has been mainly carried out in individuals diagnosed with OCD. Furthermore, pharmacological studies have only focussed on the tricyclic SRI clomipramine, or the selective SRIs (SSRIs) citalopram, escitalopram, fluvoxamine, fluoxetine, sertraline, and paroxetine. Saxena et al. (2007) conducted a prospective pharmacological study of a mixed group of compulsive hoarders of which some were diagnosed with OCD. The authors found hoarding symptoms to improve as much from paroxetine monotherapy than (other) OCD symptom dimensions. Further research on the efficacy of pharmacotherapy in hoarding OCD patients has yielded mixed results. While some of these studies reported excessive hoarding to be associated with a poorer treatment response to SRIs than classical OCD symptoms (e.g., Black, et al., 1998; Mataix-Cols, et al., 1999; Stein, et al., 2008), others did not confirm this finding (Alarcon, Libb, & Spitler, 1993; Alonso, et al., 2001; Jenike, Baer, Minichiello, Rauch, & Buttolph, 1997). Accordingly, Saxena (in press) argues that compulsive hoarding cannot be associated with a poor response to pharmacotherapy. Owing to the inconsistent findings across studies, it remains controversial if compulsive hoarding and OCD respond equally well to SRIs. The contrasting results are likely to reflect at least partially methodological differences between the studies. For example, the studies were based on different types of SRIs, different sample sizes, as well as distinct treatment response measurement methods. Moreover, the time span of the psychopharmacological interventions differed greatly between the studies, ranging from

34 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r 22 ten weeks (Jenike, et al., 1997) up to one year (Alonso, et al., 2001). Accordingly, more research is needed to draw any conclusions Pharmacological Treatment Response in Obsessive Compulsive Disorder Several pharmacological studies on the efficacy of SSRIs in treating OCD have been conducted, and meta-analytic studies have consistently shown that symptoms improve in a substantial amount of OCD patients when treated with this class of antidepressants (e.g., Ackerman & Greenland, 2002; Piccinelli, Pini, Bellantuono, & Wilkinson, 1995). For example, a good response to citalopram has been observed in approximately 58% of OCD patients (Montgomery, Kasper, Stein, Bang Hedegaard, & Lemming, 2001), fluoxetine is effective in roughly 34% (Tollefson, et al., 1994), fluvoxamine in 45% (Hollander, et al., 2003), paroxetine in 55% (Zohar & Judge, 1996), and sertraline in about 39% (Greist, et al., 1995) of the cases. As outlined earlier, some studies have associated hoarding symptoms with a poor response to SRI treatment. Importantly, several (other) OCD symptom dimensions have been associated with a worse outcome as well. These include cleaning (Alarcon, et al., 1993; Stein, Andersen, & Overo, 2007), contamination (Stein, et al., 2007), and symmetry symptoms (Stein, et al., 2008), as well as sexual or religious obsessions (Alonso, et al., 2001) Comparison of Pharmacological Treatment Response While some research has linked hoarding symptoms to a worse treatment response following pharmacological interventions with SRIs than more typical OCD symptoms, other studies did not confirm this result. Inconsistent findings might be the result of methodological differences between the studies. Moreover, the fact that most studies included hoarding individuals who were diagnosed with OCD may have influenced the results. Notably, poor response to SRIs has also been linked to several (other) OCD symptom dimensions. Accordingly, it remains controversial if compulsive hoarders respond less well to SRIs than OCD patients. A research project examining treatment response to SRIs in a large sample of non-hoarding OCD patients and compulsive hoarders who do not meet a diagnosis of OCD would clearly contribute to our understanding.

35 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r Psychotherapeutic Treatment Response The efficacy of Cognitive Behavioural Therapy (CBT) in reducing OCD symptoms has been assessed for various OCD symptom dimensions, including hoarding. This section will summarise the main findings from compulsive hoarding and OCD CBT studies Psychotherapeutic Treatment Response in Compulsive Hoarding The treatment response of compulsive hoarders to standard CBT has only been assessed in individuals who met diagnostic criteria for OCD. The majority of these studies found hoarding to be associated with a significant poorer outcome than (other) OCD symptoms (Abramowitz, Franklin, Schwartz, & Furr, 2003; Black, et al., 1998; Mataix-Cols, Marks, Greist, Kobak, & Baer, 2002; Matsunaga, Hayashida, Kiriike, Nagata, & Stein, 2010; Rufer, Fricke, Moritz, Kloss, & Hand, 2006; Saxena, et al., 2002). Notably, the research by Matsunaga et al. (2010) and Saxena et al. (2002) assessed treatment response in individuals who received both treatment with SSRIs and CBT. Furthermore, OCD patients in the research project by Black and colleagues (1998) were either treated with SSRIs or CBT, but the statistical analysis was not conducted separately for the different treatment approaches. Therefore, a better treatment outcome of non-hoarding OCD patients in these studies might reflect a response to pharmacological medication rather than CBT. In contrast, one study did not find a link between hoarding symptoms and poor treatment outcome following CBT (Seaman, Oldfield, Gordon, Forrester, & Salkovskis, 2010). However, this research was based on a rather small sample size of 18 OCD patients with, and 20 OCD patients without hoarding symptoms. Thus, the sample size might have been too small to detect significant differences between distinct OCD symptom dimensions with respect to treatment response. In summary, there is some evidence to suggest that compulsive hoarders respond worse to standard CBT treatment than individuals with more classical OCD symptoms. Accordingly, some scientists have developed new psychotherapeutic approaches that are particularly adjusted for treating excessive hoarding (Steketee, Frost, Tolin, Rasmussen, & Brown, 2010; Tolin, Frost, & Steketee, 2007). These treatment strategies have proven effective in treating hoarding symptoms as they result in clinically significant improvement in approximately 41% to 60% of the patients (Steketee, et al., 2010).

36 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r Psychotherapeutic Treatment Response in Obsessive Compulsive Disorder Research over the past four decades has shown CBT to be effective in treating OCD, with approximately 60% to 80% of the patients improving following this psychotherapeutic intervention (Foa, 2010). As outlined above, some research found hoarding symptoms to be associated with a poor response to CBT. Notably, two studies also linked sexual and religious symptoms to a poor outcome (Mataix-Cols, Marks, et al., 2002; Rufer, et al., 2006). However, other research did not confirm this finding (Abramowitz, et al., 2003; Black, et al., 1998) Comparison of Psychotherapeutic Treatment Response The vast majority of studies investigating the treatment response of OCD patients to CBT have demonstrated that hoarding symptoms are linked to a poor outcome. In contrast, most research studies have not found an association between (other) OCD symptoms and a worse outcome following CBT, apart for sexual and religious symptoms. Although all research has been conducted in hoarding individuals with a formal diagnosis of OCD only, this provides strong evidence to suggest that hoarding symptoms do not improve as much as more typical OCD symptoms from this therapeutic approach. 2.4 Genetics Several lines of evidence indicate that genetic factors are likely to play a role in the aetiology of both compulsive hoarding and OCD. So far, several familial aggregation and twin studies have been conducted to examine the influence of genetic factors on the pathogenesis of these disorders. In addition, molecular genetic analyses have been carried out in an attempt to identify susceptibility genes for both conditions. In the ensuing sections, findings derived from these different scientific approaches will be summarised to provide an overview as to whether or not compulsive hoarding and OCD appear to share a similar genetic profile Familial Aggregation The assumption that genetic factors might play a role in the aetiology of compulsive hoarding initially derived from the fact that hoarding appears to run in families. Familial aggregation has also been commonly reported for OCD. Importantly, the familial nature of hoarding and OCD respectively does not in itself establish that genes contribute to these

37 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r 25 conditions since both shared genes and environmental factors cause familial aggregation. Thus, prevalence rates for compulsive hoarding and OCD within affected families do not allow the influence of shared environmental and genetic factors to be dissociated Familial Aggregation in Compulsive Hoarding The first research on familial aggregation of compulsive hoarding was conducted by Frost and Gross (1993). The authors found that as many as 78% in their sample of 32 excessive hoarders reported having a first-degree relative with hoarding problems. By contrast, Pertusa et al. (2008) reported a somewhat smaller percentage: nearly 52% in their sample of 27 compulsive hoarders had at least one hoarding relative. Unfortunately, Pertusa et al. (2008) did not state whether they assessed the presence of hoarding behaviour in firstdegree relatives only, or if they counted in more distant relatives, too. Both studies employed a family history approach so that the diagnostic data about family members were obtained from the hoarding participants themselves. In general, it has been documented that family history data yields underestimates of the true rates of illness within families (Gershon & Guroff, 1984). In addition to this methodological shortcoming, both studies used small sample sizes, self-report measures, and did not include a control group. Therefore, these results need to be interpreted with caution. Further research on the familial aggregation of compulsive hoarding has been carried out within the context of OCD only. Two studies reported compulsive hoarding in family members of hoarding OCD patients to occur in 54% (Pertusa, et al., 2008) and 84% (Winsberg, Cassic, & Koran, 1999) of the cases, respectively. While Winsberg, Cassic, and Koran (1999) assessed the prevalence of compulsive hoarding in first-degree relatives of hoarding participants only, Pertusa et al. (2008) did not indicate if they counted distant relatives in or not. Of note, Seedat and Stein (2002) reported a somewhat smaller percentage. In their sample of excessive hoarders, 6 participants (40%) had at least one hoarding first-degree relative. Owing to the fact that this study included hoarding participants with and without a diagnosis of OCD, this result is difficult to interpret. Importantly, all studies carried out on familial aggregation of compulsive hoarding within the context of OCD so far used a family history approach, self-report measures, small sample sizes ranging from 15 to 25 hoarding OCD participants, and no control group was included.

38 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r 26 Despite the methodological limitations of the research conducted, it should be highlighted that these studies consistently reported hoarding to run in families. The majority of studies indicated that more than half of the individuals who suffer from compulsive hoarding report as having a relative who hoards as well, regardless of an additional diagnosis of OCD or whether the study included first-degree relatives only. Taken together, these findings provide some tentative evidence to suggest that shared environmental and/or genetic factors are likely to play a role in the aetiology of this condition. Once again, further research is clearly needed Familial Aggregation in Obsessive Compulsive Disorder In comparison to compulsive hoarding, considerably more research on the familiality of OCD has been conducted. Using a family history approach, Pertusa et al. (2008) reported that 21% in their sample of 71 non-hoarding OCD patients reported having at least one relative affected by OCD. As stated above, Pertusa et al. (2008) failed to report whether first-degree relatives were included only. Several studies conducted with fair sample sizes and employing semi-structured instruments to assess the lifetime prevalence of OCD in first-degree relatives of OCD patients have been reported. For example, Nestadt et al. (2000) included 343 first-degree relatives of OCD patients, as well as 300 first-degree relatives of healthy comparison subjects. The authors found OCD to occur in 12% of the first-degree relatives of OCD patients, but only in 3% of the first-degree relatives of controls. Other research report OCD to occur between 2.5% to 10.3% among first-degree relatives of OCD patients, while the prevalence of OCD among first-degree relatives of healthy control subjects appears to be approximately 0.0% to 2.3% (Bellodi, Sciuto, Diaferia, Ronchi, & Smeraldi, 1992; Black, Noyes, Goldstein, & Blum, 1992; Fyer, Lipsitz, Mannuzza, Aronowitz, & Chapman, 2005; Grabe, et al., 2006; Pauls, Alsobrook, Goodman, Rasmussen, & Leckman, 1995) Comparison of Familial Aggregation In contrast to OCD, research into the familial aggregation of compulsive hoarding is rare and all studies conducted so far have several methodological shortcomings and thus need to be interpreted with caution. Besides these limitations, the scientifc findings indicate that both compulsive hoarding and OCD appear to run frequently within families.

39 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r 27 Importantly, hoarding appears to aggregate more frequently in families than OCD: at least half of all hoarding participants report having a relative who hoards, while OCD appears to occur only in up to 12% of first-degree relatives of OCD patients. It should be argued that a comparison between studies that used different approaches, i.e. a family history method versus a clinical interview, might be precarious. However, even when applying a family history approach, OCD was found to occur only in around 21% of firstdegree OCD relatives (Pertusa, et al., 2008). In summary, hoarding appears to run more frequently in families than OCD. Although it is not possible to conclude whether it is shared genetic and/or environmental factors which contribute to the familial aggregation in compulsive hoarding, this difference provides some tentative evidence to suggest that the aetiology of compulsive hoarding may differ from OCD Twin Studies Twin studies are a powerful method used to estimate the relative contributions of genetic and environmental factors in the aetiology of a given phenotype or illness. In a twin study design, the concordance rates for a given phenotype are compared between monozygotic twins (who share nearly 100% of their genes) and dizygotic twins (who only share about 50% of their genes). Since pair members of both monozygotic and dizygotic twins are typically exposed to similar family, social, and cultural environmental influences, a comparison of the concordance rates for a phenotype between both groups of twins enables one to estimate the extent to which genetic factors play a role in the pathogenesis of a condition Twin Studies in Compulsive Hoarding A twin study involving 4355 female participants of whom approximately 2% were classified as compulsive hoarders has recently been conducted (Iervolino, et al., 2009). The results indicated that excessive hoarding among women is highly heritable, with genetic factors accounting for approximately 50% of the variance in compulsive hoarding. Notably, nonshared environmental factors and measurement error accounted for the other 50% of the variance. Although the precise number of measurement error is not known, the results imply that not only genetic factors, but also environmental factors are likely to play an important role in the pathogenesis of compulsive hoarding, at least in women.

40 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r Twin Studies in Obsessive Compulsive Disorder Since Lange (1929) described the first cases of twins suffering from OCD more than 80 years ago, several twin studies focussing on OCD have been conducted. However, many studies published are based on very small twin population sizes, provide case descriptions only, and/or do not provide sufficient data to verify a diagnosis of OCD (e.g., Cryan, Butcher, & Webb, 1992; Inouye, 1965; Parker, 1964; Rüdin, 1953; Skre, Onstad, Torgersen, Lygren, & Kringlen, 1993; Tienari, 1963). Three studies have been conducted so far that included somewhat larger sample sizes and employed psychiatric interviews to verify a diagnosis of OCD according to DSM criteria (Andrews, Stewart, Allen, & Henderson, 1990; Carey & Gottesman, 1981; Torgersen, 1983). The research by Carey and Gottesman (1981) will not be outlined here, as their analysis included twins suffering from other psychiatric disorders, with OCD only being secondary. Torgersen (1983) recruited 85 twin pairs from Norway. Of these 170 individuals, three monozygotic twins and nine dizygotic twins were diagnosed with OCD, but none of their co-twins suffered from this condition. Thus, no pairs of twins was concordant for OCD. Identical results were obtained in a study by Andrews et al. (1990). The authors recruited 186 monozygotic and 260 dizygotic twin pairs from Australia. In total, 48 twins were diagnosed with OCD, but none of their co-twins also suffered from it. The results of these studies indicate that the impact of shared genetic and environmental factors on the pathogenesis of OCD appears to be absent or low. In contrast to classical twin studies that make use of categorical data, i.e. the presence or absence of a clinical diagnosis of OCD, several twin studies have been conducted that applied a dimensional approach of OCD. Thus, these studies assessed continuously distributed obsessional traits or symptom scores of twins. The rationale behind this approach is the assumption that OCD can be viewed as an extreme score on symptom or trait measures (van Grootheest, Cath, Beekman, & Boomsma, 2005). As this approach has not been applied using hoarding as a dimensional trait yet, it is out of the scope of this thesis to review the findings derived from this method. However, it should be noted that the heritability estimates for obsessional traits in a twin study with a large sample size of 419 pairs of twins were approximately 44% (Clifford, Hopper, Fulker, & Murray, 1984). Furthermore, van Grootheest et al. (2005) included dimensional approaches into their review on twin studies in OCD and concluded that the genetic influence of obsessive compulsive symptoms appears to be in the range of 27% to 47%.

41 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r Comparison of Twin Studies In summary, only one classical twin study examining the prevalence rates for compulsive hoarding between adult female monozygotic and dizygotic twins has been published so far. The results indicate that the genetic influence on compulsive hoarding appears to be approximately 50% in women. In contrast, more classical twin studies of OCD have been conducted. However, most of them have methodological limitations, and are all based on relatively small sample sizes. Two classical adult twin studies of OCD that used good methodological designs but had small sample sizes did not find evidence for genetic influence on the aetiology of OCD. Owing to the small sample sizes used and the fact that dimensional approaches of OCD found a considerable genetic influence on obsessive compulsive symptoms, the influence of genes in the aetiology of OCD remains unclear. Therefore, the genetic influence on the aetiology of compulsive hoarding and OCD can not be compared as more research is needed to draw any meaningful conclusions Molecular Genetic Analyses Recently, several molecular genetic studies aimed at identifying potentially influencing genes or chromosome loci which house candidate genes, have been conducted either on OCD in general or hoarding behaviour in OCD or other disorders. The two common methods applied in molecular genetic research are linkage and association analyses. Linkage studies are performed on families with at least two family members sharing a common phenotype. This method examines whether genetic markers co-segregate within families. In contrast, association studies compare the frequency of genetic markers in nonrelated individuals with and without a certain phenotype. Thus, if a genetic marker is found to be significantly more common in individuals with a certain phenotype than in individuals without this phenotype, the genetic marker is said to be associated with the phenotype Molecular Genetic Analyses in Compulsive Hoarding One genetic linkage study on compulsive hoarding has been conducted in individuals diagnosed with OCD. The OCD Collaborative Genetics Study (Samuels, Shugart, et al., 2007) found compulsive hoarding in families with early-onset OCD significantly linked to a marker on chromosome 14. This result should be interpreted with caution as the authors included individuals with hoarding symptoms rather than clinically significant hoarding

42 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r 30 into their sample. Importantly, this linkage study potentially identified a specific genomic region that was associated with hoarding behaviour in OCD, but not with (other) OCD symptom dimensions. Moreover, three association studies have been published. Research by Alonso et al. (2008) indicated that two single nucleotide polymorphisms (SNPs) of the neurotrophic tyrosine kinase receptor type 3 (NTRK3) gene on chromosome 15 may contribute to the genetic susceptibility to excessive hoarding in OCD patients. The authors found no evidence for an association of NTRK3 polymorphisms with any (other) OCD symptom dimension. Similarly, Muinos-Gimeno et al. (2009) found a significant association between a genetic variant of NTRK3 and the hoarding phenotype of OCD, but no association between NTRK3 and (other) OCD subtypes or panic disorder. The NTRK3 is a high-affinity receptor of neurotrophin-3. Neurotrophins have been shown to be involved in the proliferation and differentiation of neurons during embryonic development, as well as the regulation of neuronal survival, axonal growth, synaptogenesis and neurotransmission in the adult human brain (e.g., Alcantara, et al., 1997; Lentz, Knudson, Korsmeyer, & Snider, 1999; Lykissas, Batistatou, Charalabopoulos, & Beris, 2007; Martinez, et al., 1998). It has been speculated that impaired signalling of neurotrophin-3 may harm the integrity of dopaminergic neurons (Alonso, et al., 2008). As animal research has suggested a potential involvement of the dopamine system in food hoarding (Lacroix, Broersen, Weiner, & Feldon, 1998), the finding that the NTRK3 gene might be involved into the pathogenesis of excessive hoarding in humans is interesting and needs to be investigated further. In addition, a recent study that included 325 OCD patients found a significant association between a SNP correlating with the solute carrier family 1, member 1 (SLC1A1) gene expression and the hoarding dimension in OCD (Wendland, et al., 2009). This gene encodes high-affinity glutamate transporters that are widely expressed in the human brain. Glutamate transporters are essential for the transport of the neurotransmitter glutamate across plasma membranes and maintaining extracellular glutamate concentrations. Of note, glutamate has been suggested to be involved in the aetiology of OCD (e.g., Chakrabarty, Bhattacharyya, Christopher, & Khanna, 2005). Moreover, one genomewide-scan of hoarding within the context of Gilles de la Tourette Syndrome has been carried out. Zhang et al. (2002) found a significant linkage of excessive hoarding with markers on chromosomes 4q, 5q and 17q.

43 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r 31 Finally, one case-control study by Lochner et al. (2005) reported that the methionine/methionine (met/met) genotype of the catechol-o-methyltransferase (COMT) Val158Met polymorphism on chromosome 22 was significantly more common in OCD patients with excessive hoarding than non-hoarding OCD patients or healthy controls. However, these results were only significant when comparing subjects of Afrikaner descent and did not hold when Caucasian participants were included in the sample. It has been hypothesized that COMT modulates dopaminergic function and possibly executive function in humans (Lipsky, et al., 2005) Molecular Genetic Analyses in Obsessive Compulsive Disorder Many genetic studies on OCD have been published over the last decade, focussing mainly on genes that are involved in the serotonergic, dopaminergic, glutaminergic, and opiod systems (Pauls, 2010). While some studies found a significant association between candidate genes and OCD (e.g., Bengel, et al., 1999; Meira-Lima, et al., 2004; Millet, et al., 2003), others failed to replicate these findings (e.g., Dickel, et al., 2007; Frisch, et al., 2000; Hemmings, et al., 2003). Noteworthy, some studies reported that the COMT gene may contribute to the genetic susceptibility of OCD, while others did not confirm this finding (Erdal, et al., 2003; Karayiorgou, et al., 1999). However, the glutamate transporter gene SLCL1A1 has been consistently found to be associated with OCD (Arnold, Sicard, Burroughs, Richter, & Kennedy, 2006; Dickel, et al., 2006; Stewart, et al., 2007; Wendland, et al., 2009). In addition, two genome-wide linkage studies of OCD have been carried out. They reported suggestive linkage on chromosomes 9p (Hanna, et al., 2002) and 3q, 7p, 1q, 15q, and 6q (Shugart, et al., 2006), respectively. The results did not reach genomewide significance. However, Nicolini et al. (2009) points out that these inconsistent findings may demonstrate that OCD is a heterogeneous condition. Similarly, Pauls (2008) argues that it is unlikely that one gene has a major effect on the manifestation of OCD as the phenotype of this disorder is very complex Comparison of Molecular Genetic Analyses In summary, there is a lack of research on the genetics of compulsive hoarding in the absence of OCD or Gilles de la Tourette Syndrome. Accordingly, our knowledge on the genetics of excessive hoarding is very limited. In contrast, more research on the genetics of

44 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r 32 OCD has been conducted, but with the exception of the glutamate transporter gene SLC1A1, the genetic findings have not been consistently replicated. Importantly, research into the molecular genetics of OCD might be biased as hoarding is often included as a subtype of OCD in these analyses. Therefore, a comparison of the genetic profile of compulsive hoarding and OCD cannot, as yet, be performed. Symptoms of compulsive hoarding have however been associated with a variant in the COMT gene and one SNP within the SLC1A1 gene. Moreover, these associations were not found for (other) OCD dimensions in these studies. Interestingly, the NTRK3 gene has been found to be associated with hoarding, but not with typical OCD symptom dimensions. As outlined earlier, this gene has also been linked to hoarding in animal research and is likely to play a role in the dopamine pathway. Furthermore, several chromosomal regions have been identified that appear to be linked to excessive hoarding, while none of them were linked to OCD in the two recently published genome-wide scans. In conclusion, there is some evidence to suggest that compulsive hoarding might have a different genetic profile than classical OCD subtypes. In particular, the NTRK3 gene is a potential susceptibility gene that might influence risk to developing compulsive hoarding. Furthermore, several chromosomal regions have been identified that might house causative candidate genes which play a role in the aetiology of compulsive hoarding, but there is no evidence to suggest that they are involved in the pathogenesis of classical OCD subtypes. 2.5 Summary To summarise, several lines of evidence suggest that compulsive hoarding might be a distinct disorder from OCD. First, the clinical characteristics of compulsive hoarding generally do not meet the diagnostic criteria to allow for a diagnosis of OCD. Second, epidemiological findings indicate that excessive hoarding is more prevalent in the population than all OCD symptom dimensions combined. Third, hoarding appears to occur predominantly in the absence of typical OCD symptoms and can be observed as least as often in the context of non-ocd anxiety disorders and/or mood disorders. Fourth, the mean age of onset of this condition is about 10 years later than OCD. Fifth, compulsive hoarders respond significantly less well to standard CBT interventions than non-hoarding OCD patients. Sixth, research has indicated that hoarding runs more often within families than is the case for OCD. Finally, genetic markers have been identified that might be involved in the aetiology of hoarding, but do not appear to be linked to OCD.

45 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r 33 3 Executive Function in Compulsive Hoarding and Obsessive Compulsive Disorder The term executive function (EF) refers to the set of processes that help to optimize performance in complex conditions requiring several components of cognitive function (Robbins & Arnsten, 2009, p. 268). It encompasses higher-level cognitive abilities that permit individuals to control and direct more automatic lower-level cognitive processes such as language or perception (Alvarez & Emory, 2006; Stuss, 2007). Broadly accepted types of EF include the anticipation, planning, and goal-directed execution of actions as well as their monitoring and inhibition (Müller & Münte, 2009). Of note, some researchers include in their definition of EF the regulation of social behaviour (Beer, Shimamura, & Knight, 2004). The theoretical conception of EF varies largely across researchers and no consensus has yet been found. Historically, EF has been attributed to the frontal lobes because many patients with focal lesions are strikingly impaired in tasks that rely on this cognitive ability (Burgess, 2010). However, several individuals with frontal lesions do not appear to suffer from executive dysfunctions, while some patients with non-frontal brain lesions have been found to be impaired in EF (Stuss, 1992). Therefore, a strict localizationist approach is misleading and equivalence between the prefrontal cortex and executive functioning cannot of course be assumed (Robbins, 1996, p. 1463). Since the prefrontal cortex is anatomically connected to other cortical as well as subcortical regions, it is now widely assumed that a wider network of brain areas facilitates EF (Burgess, 2010; Robbins, 1996). Moreover, research implicates that there are dissociable forms of EF which are thought to rely at least to some extent on distinct anatomical connections between the prefrontal cortex and other cortical and subcortical brain regions (Fuster, 2009). Neuropsychological research has shown that both compulsive hoarding and OCD are characterized by impairments in EF (e.g., Chamberlain, Fineberg, Blackwell, Robbins, & Sahakian, 2006; Grisham, et al., 2007). Furthermore, brain imaging studies found various structural brain abnormalities and/or abnormal brain activation patterns for OCD (Chamberlain, Blackwell, Fineberg, Robbins, & Sahakian, 2005), and there is some evidence for abnormal brain activation during symptom provocation paradigms in compulsive hoarding (An, et al., 2009; Tolin, et al., 2009). Therefore, it appears plausible

46 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r 34 to suggest that abnormal brain functioning in both conditions reflects disease-specific pathogenetic processes that cause neuropsychological deficits. However, neuroscientific research on compulsive hoarding is still relatively rare and little is known about EF and its potential neural correlates in this condition. Accordingly, it remains unclear if compulsive hoarding and OCD share a similar neurocognitive profile with identical EF deficits. From a neuropsychological perspective, it can be expected that compulsive hoarding and OCD share similar neuroanatomic and neuropsychological characteristics, if hoarding were a symptom (dimension) of OCD. In contrast, differing neuropsychological profiles of these conditions would provide some evidence to suggest that hoarding and OCD do not rely on a common pathology, thus hoarding being a clinically distinct disorder. In order to compare EFs in compulsive hoarding and OCD in detail, this chapter will first summarise the neuroanatomical mechanisms that are thought to underlie these higher order cognitive abilities. Subsequently, the neuropsychological conception of three presumably somewhat dissociable EFs that are potentially impaired in OCD and/or hoarding will be highlighted as well as their neuroanatomical correlates. These are attentional set-shifting, spatial planning, and response inhibition. Finally, the neuropsychology of compulsive hoarding and OCD will be compared to generate research hypotheses that will guide the neuropsychological investigation of compulsive hoarding in the following chapters. 3.1 Neural Correlates of Executive Function The prefrontal cortex forms the most anterior portion of the frontal lobes. From an evolutionary perspective, the prefrontal cortex has undergone more expansion than the rest of the cerebral cortex and its cortical area reaches the greatest relative size in the human brain (Fuster, 2009). Accordingly, it has been proposed that this cortical region forms the neural basis of higher cognitive abilities such as EF (Miller & Wallis, 2008) The Prefrontal Cortex Anatomical definitions of the prefrontal cortex vary in accordance to the methodological approach employed. In general, the prefrontal cortex is thought to be situated anterior to the motor and premotor cortex and superior to the limbic cortex (Fuster, 2009). It encompasses greater parts of the superior, middle, and inferior frontal gyri, the orbital gyri,

47 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r 35 most parts of the medial frontal gyrus, and the anterior half of the cingulate gyrus (Snell, 2001). The prefrontal cortex has also been defined as the region of the cerebral cortex that receives projection fibres from the dorsomedial nucleus of the thalamus (Kolb & Whishaw, 2003). During the last century, several cytoarchitectonic maps of the cerebral cortex have been constructed to define distinct cortical regions with respect to the laminar distribution and packing density of neurons. Amongst those, the cytoarchitectonic map of Brodmann (1909) has received great attention and is still widely used, presumably due to the fact that it has been included into the Talairach and Tournoux stereotaxic atlas of the human brain (Talairach & Tournoux, 1988) which has become the standard reference in functional neuroimaging research (Petrides, 2000). Petrides and Pandya (1994, 1999) reanalysed and compared cytoarchitectonic maps of the human brain with Walker s architectonic map of the prefrontal cortex in the macaque monkey (Walker, 1940), and made some adjustments to the terminology and the boarders of certain structures. The investigation of the microscopic architecture of the cerebral cortex has led to the identification of several distinct prefrontal cortical regions. For example, Brodmann (1909) identified seven distinct regions that are roughly located in the prefrontal area, namely Brodmann areas (BA) 9 to 14, 46 and 47 (Müller & Münte, 2009). Therefore, it is generally accepted that the prefrontal cortex is a heterogeneous region that encompasses several distinct architectonic areas (Petrides, 2000). The prefrontal cortex is presumably the most richly connected of all cortical regions (Fuster, 2009). Besides its heterogeneous cytoarchitectonic organization, the prefrontal cortex varies in terms of its cortical and subcortical connections. Accordingly, it is possible to subdivide the prefrontal cortex into smaller regions with respect to their distinct topological pattern of connectivity that potentially reflects differing functional roles in EF. However, structural and functional investigations of the prefrontal cortex have yielded no consensus in identifying separate prefrontal regions. Several researchers divide the prefrontal cortex into two major components: the dorsolateral prefrontal cortex that corresponds to BA 9 and 46, and the ventral prefrontal cortex that encompasses BA 11 to 14 (Müller & Münte, 2009). However, the anatomical definition of the dorsolateral prefrontal cortex varies. For example, some authors include BA 8 (Miller & Wallis, 2008), and others exclude BA 46 but include BA 10 (Tekin & Cummings, 2002) into their conception of this cortical region. While there appears to be at

48 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r 36 least a consensus to name regions around BA 9 dorsolateral prefrontal cortex, this is not true for the ventral prefrontal cortex. While some scientists use the term orbitofrontal cortex with respect to BA 11 to 14 (Kolb & Whishaw, 2003), others refer to this region as inferior (ventral) prefrontal cortex (Kolb & Whishaw, 2003). Moreover, some researchers have grouped the prefrontal cortex into three regions, namely the lateral prefrontal cortex, the orbital prefrontal cortex, and the medial prefrontal cortex or anterior cingulate cortex (Fuster, 2009). The lateral prefrontal cortex comprises roughly of BA 8 to 10 and BA 46 which correspondents mainly to the anatomical definition of the dorsolateral prefrontal cortex that has been outlined earlier. The orbital prefrontal cortex forms the ventral aspect of the frontal lobe and compromises BA 11 and 13. This definition corresponds to the more ventral parts of the anatomical boundaries of the so-called ventral prefrontal cortex or orbitofrontal cortex as outlined above. The medial prefrontal cortex comprises parts of BA 8 to10, and BA 12, 24, and 32. The regions comprising BA 24 and 32 can be further subdivided as they constitute the anterior cingulate cortex (Fuster, 2009) Prefrontal Connectivity The prefrontal cortex is richly connected to other cortical as well as subcortical regions. In particular, different prefrontal areas are strongly interconnected which might be essential for the proposed role of the prefrontal cortex in integrating disparate information (Miller & Asaad, 2002). Moreover, prefrontal areas form reciprocal connections with almost all cortical sensory systems and motor systems (Miller & Cohen, 2001). An exception is for the basal ganglia which receive some unreciprocal direct projections from the prefrontal cortex (Fuster, 2009). Various areas of the prefrontal cortex receive projection fibres from the dorsomedial thalamic nucleus, but also from other diencephalic structures such as other regions of the thalamus and hypothalamus (Fuster, 2009; Kolb & Whishaw, 2003). Furthermore, the prefrontal cortex is connected to mesencephalic brain regions and limbic structures such as the amygdala and the limbic cortex that are thought to play an important role in memory, affect, and reward (Miller & Wallis, 2008). The role of the prefrontal cortex in EF undoubtedly depends on the functional integration of different brain areas (Fuster, 2009). Influential research by Alexander, Delong, and Strick (1986) has supported the idea of functionally segregated brain circuits that connect relatively distinct areas of the prefrontal

49 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r 37 cortex with subcortical regions. The authors identified five fronto-subcortical circuits of which three are likely to play an important role in EF. All three circuits connect one area of the prefrontal cortex with discrete, essentially non-overlapping regions of the basal ganglia and the thalamus in parallel loops, that project from the thalamic regions back to the prefrontal region of origin (see Figure 1). The dorsolateral prefrontal loop connects the dorsolateral prefrontal cortex (BA 9 and 10) with the dorsolateral head of the caudate nucleus. The latter projects to the lateral dorsomedial internal segment of globus pallidus, and to the rostrolateral portion of substantia nigra pars reticulata. From there, a projection terminates in the parvocellular portion of the ventral anterior as well as the mediodorsal thalamic nucleus, which in turn project back to the dorsolateral prefrontal cortex. The dorsolateral prefrontal loop is thought to be involved in cognitive functions such as problem solving, planning, and setshifting (Tekin & Cummings, 2002). Figure 1. Prefronto - Subcortical Connectivity. Adapted from Alexander et al. (1986). The lateral orbitofrontal loop connects the lateral orbitofrontal cortex (BA 10) with the ventromedial head of the caudate nucleus. This region projects in turn to the medial dorsomedial internal segment of globus pallidus, and to the rostromedial portion of

50 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r 38 substantia nigra pars reticulata. From there, a projection terminates in the medial magnocellular portions of the ventral anterior as well as mediodorsal thalamic nucleus, which finally project back to the lateral orbitofrontal cortex. The lateral orbitofrontal loop has been proposed to play an important role in behavioural inhibition, emotional liability, and adequate social behaviour (Tekin & Cummings, 2002). The anterior cingulate loop connects the anterior cingulate area (BA 24) with the ventral striatum, which in turn projects to the rostrolateral internal segment of globus pallidus, to the ventral pallidum, as well as to the rostrodorsal substantia nigra pars reticulata. From there, projections terminate in posterior and medial portions of the mediodorsal thalamic nucleus, which in turn project back to the anterior cingulate area. The anterior cingulated loop appears to be particularly involved in motivational processes (Tekin & Cummings, 2002). 3.2 Dissociating Executive Functions There has been a major controversy in neuroscience as to whether the prefrontal cortex should be considered unitary or heterogeneous in function (Roberts, 1998). Similarly, it is still unknown if and how EF should be classified into smaller distinct cognitive components. Based on the idea that EFs are dissociable, several neuropsychological tasks have been developed that aim to assess distinct cognitive functions. Moreover, brain imaging studies have investigated the neural substrates of different cognitive abilities that appear to rely at least partially on distinct brain areas. Of particular importance for this thesis are the proposed distinct neuropsychological functions attentional set-shifting, planning, and response inhibition as these are likely to be impaired in compulsive hoarding and/or OCD. Therefore, this section will introduce the neuropsychological concept of each of these cognitive abilities and highlight their proposed underlying neural substrates Attentional Set-Shifting The term attentional set-shifting refers to the ability to switch attention from one aspect of a stimulus to another in an ongoing task, in accordance with changing reinforcement contingencies (Chamberlain, et al., 2005). This cognitive ability is crucial to flexibly alter behaviour according to environmental changes and therefore carries survival value (Kehagia, Murray, & Robbins, 2010).

51 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r 39 Two types of attentional set-shifting have been proposed (Downes, et al., 1989), intradimensional and extradimensional set-shifting. An intradimensional shift requires switching attention from exemplars of one stimulus dimension (e.g., circles) to newly introduced stimuli of the same perceptual dimension or sensory modality (e.g., squares) on the basis of feedback. This cognitive ability reflects rule generalisation when novel stimuli are presented. In contrast, an extradimensional shift occurs when attention needs to be switched between different perceptual dimensions or sensory modalities (e.g., from lines to shapes) on the basis of feedback, and is thought to reflect rule transfer from one stimulus dimension to another (Pantelis, et al., 1999; Robbins & Arnsten, 2009). A computerized version of the Wisconsin Card Sorting Test (WCST; Berg, 1948) has been developed to assess both intra- and extradimensional set-shifting abilities. During this so-called Intradimensional/Extradimensional Set-Shifting Task (IED; Downes, et al., 1989), subjects are required to learn a series of visual discriminations in which only one of two stimuli is relevant according to computer feedback the subjects receive. After six consecutive correct trials, the rule is changed and another stimulus becomes relevant. There are two critical parts of the task in which novel exemplars of a previously relevant stimulus dimension are presented: At one stage, the subject is required to generalize the established rule and keep responding to the stimuli of the same stimulus dimension (intradimensional shift). At another stage, the subject has to shift attention to the previously non-reinforced stimulus dimension, ignoring the other dimension (extradimensional shift). Early research in marmosets has shown that lesions of the lateral prefrontal cortex (BA 9) impair extradimensional, but not intradimensional set-shifting (Dias, Robbins, & Roberts, 1996). Of interest, the same study reported that orbitofrontal lesions selectively impaired reversal learning, the ability to switch responding to a previously non-reinforced stimulus (Kehagia, et al., 2010). Further studies in humans have been conducted using functional magnet resonance imaging (fmri) or positron emission tomography (PET), linking extradimensional set-shifting to anterior and/or dorsolateral prefrontal regions (Nagahama, et al., 2001; Rogers, Andrews, Grasby, Brooks, & Robbins, 2000). Hampshire and Owen (2006) applied an event related fmri paradigm and showed that a broad network of brain areas is activated when subjects are searching for a solution during attentional set-shifting tasks, including the orbitofrontal, dorsolateral prefrontal, ventrolateral prefrontal, and posterior parietal cortices. The authors contrasted activation

52 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r 40 during extradimensional and intradimensional set-shifting and found an increased activation in the ventrolateral prefrontal cortex for the extradimensional shift. When comparing activity during intradimensional shifts with non-shifts, changes in the dorsolateral prefrontal cortex and posterior parietal cortex were observed. Furthermore, they found evidence to suggest that the posterior parietal cortex appears to mediate changes in stimulus-response mapping, and that the lateral orbitofrontal cortex is likely to be involved in response-inhibition. Moreover, the authors showed that the dorsolateral prefrontal cortex appears to be generally involved in the search for a solution, but no other specific components of set-shifting. Accordingly, the authors hypothesised that lateral prefrontal, orbital, and parietal brain areas form a supervisory network that controls the focus of attention, with a potential functional specialisation of each region. In summary, there is growing evidence to suggest that attentional set-shifting is mediated through a network of brain regions, encompassing particularly lateral and orbital, but also parietal and dorsolateral prefrontal areas, and presumably their subcortical connections. Furthermore, noradrenaline is likely to play an important role in the modulation of attentional set-shifting (Kehagia, et al., 2010). It appears plausible that each brain region is functionally specialized in this cognitive function and that the lateral prefrontal cortex is particularly involved in successful extradimensional shifting. However, further research is clearly needed to dissociate the various brain areas involved according to their specific roles Planning Planning has been defined as the ability to organize cognitive behaviour in time and space (Owen, 1997, p. 431) and involves the identification, generation and timing of a series of steps to achieve a certain goal (Lezak, Howieson, & Loring, 2004). According to Shallice and Burgess (1996), the ability to plan is particularly needed to cope with novel situations, but is not important to carry out routine tasks that are thought to rely on established strategies. It is generally accepted that several cognitive components interact to execute an efficient plan (Lezak, et al., 2004; Owen, 1997). Using a neuronal network approach, these subprocesses were differentiated into four major components: plan generation, working memory, internal evaluation, and reward (Dehaene & Changeux, 1997). Moreover, Lezak (2004) highlights the importance of decision making processes, impulse control, and sustained attention during planning.

53 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r 41 Several studies have been conducted to investigate the neural correlates of spatial planning in healthy participants using a computerized version of Tower of London Task (ToL; Shallice, 1982), the Stockings of Cambridge Task (SOC; Owen, Downes, Sahakian, Polkey, & Robbins, 1990). This neuropsychological task requires subjects to move coloured balls on a computer screen to match a specified target pattern in the minimum amount of moves possible. During the test, the minimum amount of moves needed to solve each problem gradually increases, and thus also the task s difficulty and complexity. Using single-photon emission computerized tomography (SPECT), increased cerebral blood flow in the left frontal cortex has been observed in normal controls while performing the SOC (Morris, Ahmed, Syed, & Toone, 1993; Rezai, et al., 1993). Due to the limited spatial resolution of SPECT, further research has investigated the functional specialisation of brain areas in planning by means of PET or fmri paradigms. An increased activation during performance of the task has been identified in several cortical areas, including the prefrontal, cingulate, premotor, parietal, and occipital cortices (Baker, et al., 1996; Dagher, Owen, Boecker, & Brooks, 1999; Owen, Doyon, Petrides, & Evans, 1996; Rowe, Owen, Johnsrude, & Passingham, 2001). Furthermore, some studies reported an increased activation in the basal ganglia, cerebellum, and/or insula (Lazeron, et al., 2000; Schall, et al., 2003; van den Heuvel, et al., 2003). It has been proposed that this broad pattern of brain activation reflects the involvement of several cognitive abilities in planning, including the selection of strategies, visual imagery, working memory, or the execution of movements (Dagher, et al., 1999). To isolate the pattern of activation during planning, some PET studies subtracted the cerebral blood flow during simple SOC problems (that require only a few moves) from that during more difficult ones (that require many moves). While Baker and colleagues (1996) found a significant increase in regional cerebral blood flow in the dorsolateral and rostrolateral prefrontal cortex, parietal and occipital cortices, Owen et al. (1996) reported an increased activation in the head of the left caudate nucleus and the right thalamus. In contrast, one fmri study did not find any differences in brain activation associated with task difficulty (Lazeron, et al., 2000). Applying a parametric approach, Dagher and colleagues (1999) observed a significant correlation between regional cerebral blood flow and task difficulty in prefrontal areas (including the dorsolateral prefrontal cortex), the cingulate cortex, and the right caudate nucleus. Similar results were obtained with fmri and PET studies by Schall and colleagues (2003).

54 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r 42 In summary, imaging research has consistently shown that the dorsolateral prefrontal cortex is activated during planning tasks. However, this region has also been associated with working memory performance (e.g., Owen, et al., 1996) and has been hypothesized to be involved in the active manipulation and monitoring of information in tasks that require for example the generation of plans (Petrides, 1994). Moreover, several other cortical and subcortical areas have also been linked to cognitive planning. Accordingly, it has been suggested that the dorsolateral prefrontal cortex plays a critical role in planning which is mediated through its close functional interactions with multiple cortical and subcortical regions (Owen, 1997). In particular, it seems likely to associate planning with the dorsolateral prefrontal loop that has been proposed by Alexander et al. (1986) which connects the dorsolateral prefrontal cortex with basal ganglia structures and the thalamic regions (Alvarez & Emory, 2006) Response Inhibition The term response inhibition refers to a hypothetical process that is recruited when ongoing responding must be suppressed (Robbins & Arnsten, 2009, p. 273). It has been suggested that this cognitive component is likely to reflect a set of dissociable processes, including attending to and interpreting signals, making decisions, selecting an appropriate inhibitory action, and executing motor actions that counteract the pre-planned motor response (Eagle, Bari, & Robbins, 2008). Typical neuropsychological tasks that aim to assess response inhibition are Go/No-Go and stop-signal paradigms that require subjects to respond quickly on go trials and to inhibit their response on no-go or stop trials. Among these, the Stop Signal Task (SST; Aron, Dowson, Sahakian, & Robbins, 2003) has been widely used in research. During each trial of this test, an arrow (stimulus) pointing randomly to the left or right appears in the centre of the computer screen and subjects are asked to indicate the direction of each arrow as quickly as possible by pressing the left or right key of a button-box. Importantly, on approximately 25% of the trials, an auditory signal (stop-signal) sounds and subjects are instructed to inhibit their response when they hear it. The onset of the stop-signal following the initial stimulus varies during the task, and it is therefore possible to estimate the time it takes an individual to inhibit a prepotent motor response (stop-signal reaction time), using a tracking algorithm.

55 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r 43 In a first attempt to investigate the neural substrates of response inhibition, lesion studies in primates have been conducted that highlighted the role of the right inferior frontal gyrus during Go-/No-Go performance (Robbins, 2007). Several neuroimaging and lesion studies in humans have confirmed an association between this cortical area and response inhibition (e.g., Aron, Monsell, Sahakian, & Robbins, 2004; Goghari & MacDonald, 2009; Rubia, Smith, Brammer, & Taylor, 2003). Based on a functional neuroimaging approach, Aron and Poldrack (2006) hypothesized that go and stop processes during the SST rely on distinct neural pathways, with going being predominantly associated with motor areas and stopping being mainly associated with the right inferior frontal cortex and subthalamic nucleus. The authors suggested that the subthalamic nucleus acts to suppress thalamocortical output and thereby inhibts the fronto-striatal pathway important for response inhibition. Further research has shown that noradrenaline appears to modulate response inhibition relative selectively, while dopaminergic and serotonergic influences appear to be only minor (Robbins, 2007). In sum, research has highlighted the role of the right inferior frontal cortex and its subcortical connections in response inhibition. However, recent evidence indicates that activation in this cortical region is not particularly linked to inhibition processes, but rather reflects a combination of attentional and response control demands (Dodds, Morein-Zamir, & Robbins, 2010). Accordingly, Dodds and colleagues (2010) proposed that the inferior frontal cortex plays a critical role in the integration of bottom-up attentional processing of sensory stimuli (through its connections with posterior, sensory-related brain areas) with top-down processing of response-related information (through its connections to anterior, motor-related brain areas), thereby facilitating flexible, goal-directed behaviour. Although this specialized function of the right inferior frontal cortex is clearly involved in the successful inhibition of motor responses, the neural correlates of response inhibition per se remain unknown. 3.3 Neuropsychology of Compulsive Hoarding and Obsessive Compulsive Disorder This section summarises the neuropsychological findings of research on compulsive hoarding and OCD. First, the neuropsychological profiles of both conditions will be described with a major focus on EFs, and subsequently compared. It will be examined if

56 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r 44 hoarding and OCD appear to share similar neuropsychological deficits and how a comparison of EFs between these conditions can contribute to the scientific debate on the relationship between hoarding and OCD Neuropsychology of Compulsive Hoarding Little is known about the neuropsychology of compulsive hoarding. However, there is some evidence to suggest that certain cognitive abilities and particular EFs are impaired in individuals who hoard Attention During individual and group therapy sessions, Steketee, Frost, Wincze, Greene, and Douglass (2000) observed that individuals suffering from compulsive hoarding had difficulties staying focused on sorting tasks. Since then, some studies have examined whether compulsive hoarding might be linked to impairments in attention and/or symptoms of attention-deficit/hyperactivity disorder (ADHD). First of all, Hartl et al. (2005) compared a group of 26 compulsive hoarders (32% were diagnosed with OCD) with a sample of 36 matched non-clinical controls on the ADHD symptom checklist (Barkley & Murphey, 1998), a self-report questionnaire on ADHD symptoms. The results showed that the group of compulsive hoarders scored significantly higher on both inattention as well as hyperactivity symptoms than the control group. However, 32% of the recruited compulsive hoarders were actually diagnosed with ADHD. The authors stated that, in their experience, ADHD is commonly found in hoarding individuals. It would still be important to know whether symptoms of inattention were also more common in the subgroup of compulsive hoarders who did not meet a diagnosis of ADHD. Secondly, Grisham et al. (2007; 2010) published two studies that assessed ADHD symptoms in compulsive hoarding, also using the ADHD symptom checklist (Barkley & Murphey, 1998). In both studies, the authors compared individuals suffering from compulsive hoarding with a non-clinical control group and a mixed clinical control group of individuals suffering from mood or anxiety disorders, but not hoarding. In the earlier published article, Grisham et al. (2007) included 30 individuals in each group. Of note, approximately 50% of the individuals in the hoarding group were diagnosed with OCD. The results yielded higher scores of both inattention and hyperactivity symptoms in the

57 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r 45 hoarding group, when compared to clinical or non-clinical controls. In their later published article, the authors recruited individuals per group, with only one individual in the hoarding group meeting criteria for OCD (Grisham, et al., 2010). The authors reported that the hoarding group had significantly higher scores on ADHD symptoms than the nonclinical control group. However, the hoarding group did not differ from a clinical comparison group of individuals suffering from mood or anxiety disorders. In summary, the research on attentional deficits in compulsive hoarding has yielded mixed results and do not provide evidence to suggest that compulsive hoarding is particularly linked to inattention. First of all, one study did not find a difference for selfreported attentional deficits between hoarders and a mixed-clinical control group. Secondly, the significant result of one study is likely due to a participant selection bias as one third of the hoarding individuals were diagnosed with ADHD. Finally, all findings derived from self-report measures but not standardized neuropsychological tasks Attentional Set-Shifting Research by Grisham et al. (2010) compared attentional set-shifting performance between a group of compulsive hoarders, a clinical control group suffering from mood or anxiety disorders, and a non-clinical control group. They administered the IED and assessed reversal learning, as well as intradimensional and extradimensional set-shifting abilities. The authors found to differences between the groups. Therefore, hoarding does not appear to be associated with attentional set-shifting impairments Memory Frost and Hartl (1996) hypothesised that compulsive hoarding might be linked to memory deficits. Based on their clinical experience, they reported that individuals who hoard commonly lack confidence in their memory, i.e. keep items like newspapers because they worry not to be able to remember the information contained, and overestimate the importance of remembering. Since then, three neuropsychological studies investigating memory performance in compulsive hoarding have been conducted. Hartl et al. (2004) globally supported Frost and Hartl s assumption of impaired memory function: The authors compared 22 hoarding individuals (of which four were diagnosed with OCD), with 24 matched controls. In comparison to controls, individuals suffering from compulsive hoarding had significantly less confidence in their memory,

58 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r 46 perceived the consequences of forgetting as more catastrophic, and had a stronger desire to keep possessions in sight. Moreover, compulsive hoarding was associated with deficits in verbal and visuo-spatial memory: Hoarding individuals recalled less information on the delayed recall of the California Verbal Learning Test (CVLT; Delis, Kramer, Kaplan, & Ober, 1987) and the Rey-Osterrieth Complex Figure Test (RCFT; Osterrieth, 1944). In contrast, research by Jang and colleagues (2010) did not confirm a deficit on the RCFT. The authors analysed several performance measures of this task among 144 OCD patients and 144 controls. Although several OCD symptom dimensions were negatively correlated with task performance, i.e. OCD patients with these symptom presentations performed worse on the task, this was not the case for the hoarding symptom dimension. In addition, Grisham et al. (2007) compared memory performance of 30 individuals suffering from compulsive hoarding (approximately half were diagnosed with OCD), with 30 non-clinical controls, and 30 individuals of a mixed clinical group. The authors reported no significant differences between the groups on short-term auditory memory as measured by the Digit Span of the Wechsler Adult Intelligence Scale (Wechsler, 1997). However, hoarding individuals performed worse during the forward, but not the backward Visual Memory Span of the Wechsler Memory Scale - Revised (Wechsler, 1987), a test of visuospatial short-term memory. It should be noted however, that the group of compulsive hoarders differed significantly from the control group, and marginally significant from the mixed clinical group, on a test for non-verbal intelligence. Although no differences in verbal intelligence were found between the groups, it is arguable whether differences in general intellectual ability, and thus a potential selection bias in recruiting hoarding participants, might have caused the poor performance on visuo-spatial memory in the hoarding group. Finally, a study that compared memory performance of 7 individuals suffering from compulsive hoarding and late life depression with 45 participants suffering from late life depression only, found no significant differences in performance between the groups when comparing measures of auditory and visual short-term memory, as well as auditory and visual working memory (Mackin, Arean, Delucchi, & Mathews, 2010). However, this result needs to be interpreted with caution as the sample size of compulsive hoarders participating in this research project was probably too small to detect significant differences in performance. Furthermore, the fact that the participants had a mean age of

59 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r years, and that all individuals suffered from late life depression, is likely to have influenced the results. In conclusion, there is some evidence to suggest that individuals suffering from compulsive hoarding commonly lack confidence in their memory and overestimate the importance of remembering information. However, it is currently unclear if the subjectively perceived memory deficits correspond to objectively measurable memory impairments Categorization The assumption that hoarding behaviour might be linked to problems grouping items together into categories has received great attention. Frost and Hartl (1996) hypothesised that the accumulating clutter that occurs within the course of compulsive hoarding might reflect a deficit in the structuring and categorization of information. The authors reported that individuals suffering from excessive hoarding regard each of their possessions as unique which in turn increases the value of each possession and leads them to create greater number of categories when asked to sort them. This has also been termed underinclusive cognitive style (Steketee & Frost, 2003, p. 913). Furthermore, Frost and Hartl (1996) stated that the deficit in categorizing objects leads to problems in organizing possessions which in turn is regarded as one source of the accumulation of clutter. Two studies examined categorization abilities in compulsive hoarding using a nearly identical task. First of all, participants were asked to sort 20 commonly hoarded objects according to their meaning into different piles. In both studies, there were no differences in the amount of piles created by individuals suffering from compulsive hoarding, a nonclinical control group, a matched clinical control group, and a group of individuals suffering from non-hoarding OCD (Grisham, et al., 2010; Wincze, Steketee, & Frost, 2007). Furthermore, there were also no differences between the groups when comparing the amount of time needed to sort items into different categories. In the research project of Wincze and colleagues (2007), all participants completed an additional task using identical objects and were asked to sort these items into the smallest number of meaningful groups possible. Again, there were no significant differences between the groups when comparing the amount of piles created and the time needed to sort the items. In contrast, Grisham and colleagues (2010) asked their participants to group 20 index cards on which the names of the previously grouped items were written into different meaningful piles. There were no

60 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r 48 differences between the groups with regard to the amount of piles created, however hoarding individuals were on average significantly slower than non-clinical controls when sorting the index cards, but not significantly slower than the clinical control group. To assess categorization deficits when sorting personal belongings, both studies first assessed items that were typically saved by the individual, and wrote 20 of them on index cards that matched five idiographic themes. The participants were then asked to group the items into different meaningful piles. The study by Wincze et al. (2007) observed that the compulsive hoarding group created significantly more categories than the non-clinical control group, but that there was no difference between the hoarding group and the nonhoarding OCD group. Furthermore, hoarders took significantly more time to sort personal items than both non-clinical controls as well as non-hoarding OCD patients. Similarly, Grisham and colleagues (2010) found that hoarding individuals created significantly more categories when sorting personal belongings than a clinical-control group, but there were no differences between the hoarding group and a non-clinical control group. There were also no significant differences in time needed to sort the possessions between the groups. Additionally, the participants of this research project were asked to bring 20 items with them that represent the types of items they typically have at their home and were asked to sort them into categories. Owing to the fact that it appears impossible to make sure that these personal possessions of the participants could be grouped in the same amount of categories across individuals, the results will not be presented here. Notably, both studies reported that individuals suffering from compulsive hoarding experienced significantly more anxiety during nearly all stages of the tasks in comparison to the clinical and nonclinical control groups. Furthermore, one study examined categorization abilities in a group of students who identified themselves as packrats (Luchian, McNally, & Hooley, 2007). Although these individuals did not suffer from clinically significant compulsive hoarding, it should be noted that the hoarding individuals took more time and created more categories on a task that required them to sort objects into groups then non-hoarding students. Furthermore, the self-identified packrats experienced the sorting task as more stressful and difficult than the control group. Finally, research by Hartl and colleagues (2004) found significant categorization deficits in a sample of compulsive hoarders in comparison to a control group during performance on the RCFT (Osterrieth, 1944), a measure of visuo-spatial memory.

61 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r 49 However, they found no evidence for categorization deficits when comparing performance on the CVLT (Delis, et al., 1987), a measure of verbal and auditory memory between the groups. In summary, these findings do not provide conclusive evidence to suggest that hoarding is particularly associated with deficits in categorization skills. More research is needed to investigate if this cognitive ability is impaired in hoarding individuals Decision making Based on case reports, Frost and Hartl (1996) hypothesized that decision making deficits may be a hallmark of compulsive hoarding. The authors argued that indecisiveness and the fear of making mistakes appears to be a common reason for individuals suffering from excessive hoarding to avoid throwing things away. Since then, some research on this topic has been conducted. First of all, two studies assessed the extent of decision making difficulties using the Frost Indecisiveness Scale (Frost & Shows, 1993), a self-report questionnaire. When comparing a group of individuals suffering from compulsive hoarding with a non-clinical control group, significant differences between the groups were found, indicating that the hoarding individuals report higher levels of indecisiveness (Grisham, et al., 2010; Wincze, et al., 2007). However, self-rated indecisiveness did not differ between compulsive hoarders and a mixed clinical control group (Grisham, et al., 2010), and was also not significantly different when comparing hoarding individuals to non-hoarding OCD patients (Wincze, et al., 2007). Moreover, two studies examined decision making performance on the Iowa Gambling task (Bechara, Damasio, Damasio, & Anderson, 1994), a computerized card playing game that requires participants to avoid selecting cards that are associated with immediate high rewards, in favour of smaller rewards to achieve a higher net gain in the long term. Using this task, Lawrence et al. (2006) found a significantly poorer performance on decision making in hoarding individuals who were diagnosed with OCD in comparison to nonhoarding OCD patients and non-clinical controls. In contrast, another study did not find a difference in performance on this task between individuals suffering from compulsive hoarding, a mixed clinical control group, and a non-clinical control group (Grisham, et al., 2007).

62 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r 50 In addition, decision making abilities have been studied using the Cambridge Gambling Task (Rogers, et al., 1999) a computerized task that assesses decision making and risktaking behaviour. No significant differences with respect to the quality of decision making and risk-taking adjustment were found when comparing a group of compulsive hoarders with a mixed clinical control group and a non-clinical control group (Grisham, et al., 2010). Owing to the inconsistent results when testing decision making abilities in individuals suffering from excessive hoarding, it has been hypothesized that decision making deficits might only be present when making personal decisions, for example when individuals have to decide whether or not to discard possessions they own (Pertusa, Frost, Fullana, et al., 2010). Accordingly, Tolin and colleagues (2009) designed a task in which participants had to decide whether to keep or discard personal belongings as well as items that did not belong to them but the experimenter. Depending on the decision of the participants, the items were either saved or destroyed. The authors compared decision making abilities in terms of the average time needed to make a decision and the percentage of items that were discarded between a group of 12 individuals suffering from compulsive hoarding and a group of 12 matched controls. The results indicated that compulsive hoarding individuals were significantly slower when deciding whether or not to discard personal possessions in contrast to possessions that did not belong to them. There was no such difference for the control group. Furthermore, the hoarding group took significantly longer than the control group to make decisions about personal items, but there was no such difference when deciding whether or not to discard the possessions of the experimenter. Also, individuals suffering from compulsive hoarding discarded on average fewer personal items than items that did not belonging to them. Such a difference was not found for the control group. Finally, the hoarding group discarded significantly less items than the control group, regardless of whether or not the items belonged to them. Of note, the hoarding group rated this task as more anxiety provoking than the control group. In conclusion, individuals suffering from compulsive hoarding frequently report high levels of indecisiveness. However, this self-rated indecisiveness appears to occur with a similar intensity amongst other individuals suffering from mood disorders or anxiety disorders in the absence of hoarding. Typical neuropsychological tasks examining decision making abilities have yielded inconsistent results. This might be due to the fact that hoarding individuals are only impaired in decision making when emotionally and

63 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r 51 personally affected by a decision, i.e. when they have to decide whether or not to throw things away they own and potentially feel attached to. Such personal consequences of a decision are unlikely to be simulated when using standard neuropsychological measures of decision making such as the Iowa Gambling Task or the Cambridge Gambling Task. However, this hypothesis clearly needs to be investigated further Planning Only one study examining planning ability in compulsive hoarding has been published so far (Grisham, et al., 2010). The SOC (Owen, et al., 1990) was administered to 19 individuals suffering from compulsive hoarding, 17 clinical controls suffering from mood or anxiety disorder, and 20 non-clinical controls. During the tasks, the participants are presented with colourful balls they have to move in a certain way in order to meet a given target pattern. The task requires participants to plan ahead and execute a series of steps to meet the target in the minimum amount of moves possible. Individuals suffering from compulsive hoarding needed significantly more moves to solve the different problems presented during the tasks than both control groups. This finding indicates that compulsive hoarding might be linked to spatial planning deficits Response Inhibition Recently, one study examined response inhibition in compulsive hoarding. Using an affective Go/No-go task, Grisham and colleagues (2010) compared performance measures between a group of hoarding individuals, a matched clinical control group and a nonclinical control group. The authors found no differences between the groups. Therefore, there is no evidence for impaired response inhibition in compulsive hoarding. However, other studies should be conducted, using different response inhibition tasks like the SST, to verify this result Neuropsychology of Obsessive Compulsive Disorder Extensive research on the neuropsychology of OCD has been carried out during the last decade. Accordingly, the underlying neurocognitive profile of this condition will be outlined here (for reviews on neuropsychological function in OCD, see Chamberlain et al., 2005, or Kuelz et al., 2004 ).

64 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r Attention Basic attentional processes generally appear not to be impaired in OCD (Kuelz, Hohagen, & Voderholzer, 2004). For example, OCD patients perform within the normal range on tasks that assess alertness or sustained attention (e.g., Hollander, et al., 1993; Martin, et al., 1993; Milliery, Bouvard, Aupetit, & Cottraux, 2000). Although some studies observed that OCD patients have a slower speed of information processing (Moritz, et al., 2002), others did not confirm this finding (Cohen, et al., 1996; Jurado, Junque, Vallejo, & Salgado, 2001). Owing to the fact that the OCD patients in the studies that found slower speed of information processing were predominantly medicated with SSRIs, it has been suggested that attentional deficits are generally not impaired in OCD, but might reflect the impact of medication. Finally, selective attention in OCD has been examined. Several studies did find deficits in this cognitive ability, however it has been argued that this might only be caused from medication effects on information processing speed (Kuelz, et al., 2004). In summary, there is no evidence to suggest that OCD is particularly associated with attentional deficits. However, further research should investigate if impaired speed of information processing and selective attention are a consequence of medication, or if they form a typical neuropsychological deficit in unmedicated OCD patients as well Attentional Set-Shifting Potential set-shifting deficits in OCD have been investigated by many researchers, using mainly the WCST or IED. When administrating the WCST, some studies identified attentional set-shifting deficits in OCD patients (Hymas, Lees, Bolton, Epps, & Head, 1991; Okasha, et al., 2000; Rao, Reddy, Kumar, Kandavel, & Chandrashekar, 2008), while others did not confirm this finding (Abbruzzese, Ferri, & Scarone, 1995, 1997; Moritz, et al., 2002). However, most research observed impaired performance (primarily selective extradimensional set-shifting deficits) of OCD patients on the IED, (Chamberlain, et al., 2007; Nedeljkovic, et al., 2009; Purcell, Maruff, Kyrios, & Pantelis, 1998; Veale, Sahakian, Owen, & Marks, 1996; Watkins, et al., 2005), although one study did not replicate this (Nielen & Den Boer, 2003). Importantly, Chamberlain and colleagues (2007) observed poor extradimensional set-shifting performance in unaffected first-degree relatives of OCD patients. Accordingly, the authors suggest that extra-dimensional shifting represents a potential endophenotype for OCD.

65 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r 53 Although the performance of OCD patients on the WCST has yielded inconsistent results, the findings derived from the IED indicate that OCD is commonly associated with extradimensional set-shifting deficits. Due to the fact that this deficit is also evident in first-degree relatives of OCD patients, it is likely that extradimensional set-shifting impairments form a key neuropsychological impairment in OCD Memory Several studies have examined memory performance of OCD patients on a wide range of neuropsychological tasks. For example, non-verbal memory has been often assessed with the RCFT (Osterrieth, 1944) or Benton Visual Retention Test (Benton, 1974). Many studies have shown that OCD patients perform poor on these tasks when asked to recall visuo-spatial information (e.g., Kim, Park, Shin, & Kwon, 2002; Penades, Catalan, Andres, Salamero, & Gasto, 2005; Shin, et al., 2010; Simpson, et al., 2006). However, since these non-verbal memory tasks presuppose organisational strategies, it has been suggested that the observed deficits reflect poor strategy formation rather than non-verbal memory impairment (Chamberlain, et al., 2005). This hypothesis is in line with the fact that verbal memory, when assessed with the Rey Auditory Verbal Learning Test (Rey, 1964) is generally unimpaired in OCD (Jurado, et al., 2001; Kitis, et al., 2007; Rao, et al., 2008; Schmidtke, Schorb, Winkelmann, & Hohagen, 1998), presumably because this task has been shown not to rely on organizational skills (Ryan, Rosenberg, & Mittenberg (1984) in Callahan & Johnstone, 1994). Moreover, when verbal memory tasks that require more strategy formation such as the CVLT were applied, OCD patients have been often found to be impaired (Deckersbach, et al., 2004; Savage, et al., 2000). Therefore, verbal and visuo-spatial memory appears to be not impaired in OCD patients. Findings of deficits in these cognitive abilities appear to reflect impaired organizational strategy formation Categorization Categorization skills have not been extensively studied in OCD, yet. However, OCD patients often show deficits in visuo-spatial and auditory memory tasks such as the CVLT and RCFT that require organizational skills (Deckersbach, et al., 2004; Savage, et al., 2000). Therefore, it can be concluded that OCD is presumably linked to some form of categorization deficits. However, OCD patients do not appear to have deficits when asked

66 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r 54 to sort non-personal or personal items into categories (Grisham, et al., 2010; Wincze, et al., 2007) Decision making Decision making abilities in OCD have not been studied extensively yet (Chamberlain, et al., 2005) and the published work on this topic has yielded mixed results. For example, performance on the Iowa Gambling Task (Bechara, et al., 1994) or its computerized version, the Cambridge Gambling Task, has been both shown to be impaired (Cavedini, et al., 2002; Starcke, Tuschen-Caffier, Markowitsch, & Brand, 2010) or intact (Chamberlain, et al., 2007; Lawrence, et al., 2006; Nielen & Den Boer, 2003) in OCD patients when compared to healthy controls. Two studies that administered other decision making tasks than the Iowa Gambling Task found no evidence for impaired decision making abilities in OCD (Starcke, et al., 2010; Watkins, et al., 2005). Accordingly, decision making deficits do not appear to form a key feature of the neuropsychological profile of OCD Planning Spatial planning performance in OCD has been commonly investigated with the SOC. Only one (Nielen & Den Boer, 2003) out of four studies (Purcell, et al., 1998; Veale, et al., 1996; Watkins, et al., 2005) found evidence for impaired planning in OCD. Accordingly, planning abilities as assessed with the SOC do not appear to be commonly impaired in OCD Response Inhibition There are several lines of evidence to suggest that OCD is linked to deficits in response inhibition that might play a key role in the aetiology of the disorder. First, several studies employing Go/No-go paradigms found OCD patients to make significantly more inappropriate motor-responses to non-target stimuli than healthy controls or individuals suffering from panic disorder (Aycicegi, Dinn, Harris, & Erkmen, 2003; Bannon, Gonsalvez, Croft, & Boyce, 2002). Secondly, research has commonly found that OCD patients have a significantly slower stop-signal reaction time, a measure for motor inhibition, during the SST than healthy controls (Chamberlain, et al., 2006; Menzies, et al., 2007; Penades, et al., 2007). Moreover, Chamberlain et al. (2007) and Menzies et al. (2007) also observed impaired response inhibition in unaffected first-degree relatives of

67 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r 55 OCD patients. The authors concluded that response inhibition might represent a potential endophenotype for OCD. Finally, evidence for altered executive control processes in OCD comes from electroencephalographic and (f)mri studies that reported abnormal neural correlates of inhibition during performance of Go-/No-go tasks or the SST (Menzies, et al., 2007; Roth, et al., 2007; Ruchsow, et al., 2007). In conclusion, response inhibition appears to form a key feature of the neuropsychological profile of OCD that is not only present in the patients themselves, but also in their unaffected first-degree relatives Neuropsychological Comparison of Compulsive Hoarding and Obsessive Compulsive Disorder The neuropsychological profiles of compulsive hoarding and OCD clearly share some overlap, but also appear to differ in some important ways. While attentional deficits might be evident to some degree in both conditions, extradimensional set-shifting impairments have only been observed in OCD patients. Moreover, visuo-spatial and verbal memory performance has not been found to be generally impaired in both disorders. Owing to the fact that both conditions might be linked to categorization or strategy formation deficits, it is plausible to suggest that this impairment is likely to cause worse performance during memory tasks that rely more on the generation of strategies. While OCD patients generally appear not to suffer from decision making difficulties, hoarding individuals might experience some decision making problems, but only when they have to make more personal decisions. Furthermore, planning deficits have been observed in hoarding individuals, but there is some evidence to suggest that planning is generally unimpaired in OCD. Finally, OCD has been strongly linked to response inhibition impairments, but no evidence for such deficits has been observed in hoarding individuals as yet. To summarise, extradimensional set-shifting impairments and response inhibition deficits form a hallmark of the neuropsychology of OCD, but do not appear to be associated with compulsive hoarding. In contrast, spatial planning abilities appear to be impaired in excessive hoarding, but are generally intact in OCD. 3.4 Neuropsychological Research Hypotheses On the basis of the hypothesis that compulsive hoarding forms a clinically distinct disorder from OCD in the majority of the cases (Mataix-Cols, et al., 2010; Pertusa, et al., 2008),

68 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r 56 differing neuropsychological profiles of excessive hoarding and OCD would be expected as this would point to an unshared underlying pathogenesis of both conditions. In particular, evidence from neuropsychological research indicates that OCD patients have deficits in extradimensional set-shifting performance and response inhibition, but are generally unimpaired in spatial planning. In contrast, compulsive hoarding appears to be linked to spatial planning deficits, but not to poor performance on extradimensional setshifting and response inhibition tasks. To explore the assumption that compulsive hoarding mainly forms a clinically distinct disorder instead of an OCD symptom, it is important to compare EF in OCD patients who also experience severe hoarding symptoms with a group of individuals who suffer from compulsive hoarding in the absence of any classical OCD symptoms such as washing or checking. If compulsive hoarding in the presence of (other) OCD symptoms were to form a comorbid disorder with OCD, it seems plausible to suggest that OCD patients with prominent hoarding symptoms would show neuropsychological deficits that have been commonly observed in OCD as well as deficits that have been mainly associated with compulsive hoarding. In contrast, it can be assumed that compulsive hoarding in the absence of any (other) OCD symptoms would not be linked to typical executive dysfunctions that have been observed in OCD but hoarding specific impairments only, remaining that compulsive hoarding were to form a distinct disorder from OCD. Accordingly, the following research hypotheses can be generated to test the hypothesis that compulsive hoarding forms a distinct disorder from OCD in the majority of the cases from a neuropsychological perspective: (1) Attentional Set-Shifting - Individuals suffering from compulsive hoarding but no (other) OCD symptoms are expected to perform similar on a task of attentional set-shifting to a healthy control group. - Individuals suffering from compulsive hoarding who do meet a diagnosis of OCD are expected to have attentional set-shifting impairments, and particularly extradimensional set-shifting deficits when compared to a healthy control group. - Hoarding individuals who do meet a diagnosis of OCD are expected to perform significantly worse on attentional set-shifting tasks, and particularly extradimensional set-shifting in comparison to non-ocd hoarders.

69 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r 57 (2) Planning - Individuals suffering from compulsive hoarding but no (other) OCD symptoms are expected to perform worse during a planning task in comparison to a control group. - Individuals suffering from compulsive hoarding who do meet a diagnosis of OCD are expected to suffer from planning deficits in comparison to a control group. - Hoarding individuals who do not meet a diagnosis of OCD are expected to perform similar on a planning task when compared to compulsive hoarders who are diagnosed with OCD. (3) Response Inhibition - Individuals suffering from compulsive hoarding but no (other) OCD symptoms are expected to be unimpaired in response inhibition when compared to a healthy control group. - Individuals suffering from compulsive hoarding who do meet a diagnosis of OCD are expected to suffer from impaired response inhibition when compared to a healthy control group. - Hoarding individuals who meet a diagnosis of OCD are expected to perform significantly worse on a task measuring response inhibition in comparison to non-ocd hoarders.

70 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r 58 4 Methods 4.1 Participants Three groups of participants were recruited to test for differences in neuropsychological performance on three cognitive tasks. The three groups consisted of: 20 patients suffering from compulsive hoarding as well as (other) OCD symptoms (OCD hoarding group); 20 patients suffering from compulsive hoarding but no (other) OCD symptoms (non-ocd hoarding group); and 20 healthy control subjects (control group). In total, 70 participants took part in this research project. In order to match the groups according to their socio-demographical characteristics however, ten participants were excluded from the data analysis, leading to a final set of 20 participants per group. 4.2 Recruitment All participants were recruited between August 2008 and March 2010 in the United Kingdom from various sources. First, both hoarding groups were recruited via Dr. David Mataix-Cols from the Institute of Psychiatry at King s College London and the Central London Hoarding Support Group which is led by Satwant Singh, a nurse consultant in cognitive behavioural therapy. Second, adverts were placed on the websites of the two biggest OCD charities in the United Kingdom, OCD Action and OCD-UK. Furthermore, hoarding participants were recruited via the Bournemouth conference of OCD-UK on 3 rd of October 2009 and the London conference of OCD Action on 6 th of February Moreover, some members of the OCD hoarding group were found via the specialized outpatient clinic for OCD at Queen Elizabeth II Hospital in Welwyn Garden City that is led by Professor Dr. Naomi Fineberg. Finally, healthy control subjects were recruited through the data base of volunteers of the Behavioural and Clinical Neuroscience Institute at Cambridge University and through adverts that were placed in local shops and the central library in Cambridge.

71 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r Inclusion and Exclusion Criteria All participants of this research project were required to be between 18 and 70 years of age and to speak English fluently. To be included in one of the hoarding groups, participants had to score equal or higher than 35 on the Saving Inventory - Revised (SI-R; Frost, et al., 2004), a questionnaires that assesses the severity of hoarding symptoms. In order to further verify a diagnosis of compulsive hoarding, these subjects additionally had to fulfil the following slightly adjusted hoarding criteria that were originally proposed by Frost and Hartl (1996) and have been successfully used in previous research: the acquisition of, and failure to discard a large number of possessions that appear to be useless or of limited value; living spaces sufficiently cluttered so as to preclude activities for which those spaces were designed; significant distress or impairment in functioning caused by hoarding; clutter persisting for at least six months; hoarding behaviours not better accounted for by another mental disorder (e.g., dementia). Hoarding participants who met a diagnosis of OCD according to DSM-IV-TR criteria (see Table 2; American Psychiatric Association, 2000) were included in the OCD hoarding group. In contrast, hoarding subjects who did not meet a diagnosis of OCD were included in the non-ocd hoarding group. Exclusion criteria for all subjects were impaired vision or hearing that could not be corrected during the neuropsychological testing (e.g., by wearing glasses) or movement disabilities that could potentially impair their performance on the tasks. Furthermore, participants were not included if they had a history of excessive drug abuse or head injury, or if they were suffering from neurological deficits, organic diseases, or psychotic disorders. Moreover, subjects of the control group were excluded if they suffered from any current or previous psychiatric disorder. In contrast, hoarding individuals were also included if they fulfilled diagnostic criteria for another axis-i disorder if hoarding was their major complaint.

72 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r Procedure Before being invited to take part in the research project, both patient groups were screened during a face-to-face or telephone interview for compulsive hoarding according to the criteria outlined above, as well as axis-i disorders according to the DSM-IV-TR criteria. The interview was conducted by the experienced psychiatrists Professor Naomi Fineberg, Dr. David Mataix-Cols or Alberto Pertusa. Healthy control subjects were screened for these disorders during a telephone interview that was conducted by the author. Subsequently, all interviewed participants who fulfilled the inclusion criteria stated above received an information sheet about this study and another telephone call with further details about the research project and the possibility to ask questions. During this telephone call, all inclusion and exclusion criteria were checked again. Participants were then invited for a three hour testing session which included a clinical interview, several clinical and personality questionnaires, and various cognitive tasks. During the testing day, participants first read the information sheet again, had the possibility to ask further questions and gave written consent. Then, socio-demographic data were obtained and a diagnostic screening was conducted. Subsequently, neuropsychological tasks and clinical and personality questionnaires were administered which were counterbalanced across subjects. Participants were given breaks whenever and how long they wanted and were free to withdraw at any given point during the research project. This research project got a favourable opinion of the Essex 1 Research Ethics Committee in August Measures Sociodemographic Measures The following sociodemographic measures of each participant were obtained: age (in years), gender, handedness, and level of education (measured in years of school and vocational education). Moreover, it was assessed if subjects were native English speakers and their verbal intelligence was estimated using the National Adult Reading Test (NART; Nelson, 1982). Furthermore, their medication status was recorded Diagnostic Assessment Before finally inviting potential participants to complete the whole test battery, a clinical interview was conducted by an experienced psychiatrist who screened and diagnosed both

73 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r 61 patient groups for mental disorders according to DSM-IV-TR criteria (American Psychiatric Association, 2000). During the testing day, the Mini-International Neuropsychiatric Interview (MINI; Sheehan, et al., 1998), a short structured diagnostic interview for DSM-IV psychiatric disorders was administered to screen all participants for OCD and other axis-i disorders again. This was done to verify the previously made diagnosis by a psychiatrist by taking into account any potential delay between the previous clinical interview and the testing day. Furthermore, the described diagnostic criteria for compulsive hoarding were checked again during the testing day. Also, the SI-R (Frost, et al., 2004), a self-administered questionnaire was employed to measure the severity of hoarding symptoms. It consists of 23 items requesting a response on a Likert-type scale from zero to four. The total score on the SI-R ranges from zero to 92, with higher scores indicating more severe self-rated hoarding symptoms. The items of the SI-R cover the three domains of compulsive hoarding that factor analytical studies have identified: clutter, difficulty discarding, and excessive acquisition (Frost, et al., 2004; Tortella-Feliu, et al., 2006). Therefore, the SI-R can be divided into the clutter subscale which covers nine items, as well as the difficulty discarding/saving subscale and the acquisition subscale which both consist of seven items. The SI-R has proven to be a reliable and valid instrument for the assessment of hoarding symptoms (Frost, et al., 2004). In the present study, the SI-R was used to confirm a diagnosis of compulsive hoarding and to verify that the symptom severity was similar between both hoarding groups Clinical Measures To measure the severity of OCD symptoms, the Y-BOCS (Goodman, et al., 1989) was assessed. It consists of an initial checklist and two separate scales measuring the severity of obsessions on the one hand, and the severity of compulsions on the other hand. At first, a checklist which includes different types of obsessions and compulsions is read out to the participants and they are asked to indicate the three most prominent or distressing OCD symptoms they are experiencing. Subsequently, both symptom scales are used to rate the severity of the identified obsessions and compulsions. Each of the two scales contains five items that are rated by a clinician from zero (no symptoms) to four (extreme symptoms). The items refer to the amount of time that is occupied by obsessions/compulsions, the interference they cause, the distress they cause, the patient s resistance against the

74 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r 62 obsessions/compulsions, and the degree of control that the patient has over the OCD symptoms. The severity of OCD can be interpreted based on the total score of both scales combined. A total score lower than eight is regarded as sub-clinical OCD, a score between eight and 15 indicates mild OCD, a score of 16 to 23 is regarded as moderate OCD, a score of 24 to 31 means severe OCD, and total score of 32 to 40 is interpreted as extreme OCD (Goodman, et al., 1989). Moreover, a computerized version of the OCI-R (Foa, et al., 2002) was administered to assess the levels of OCD and subclinical OCD symptoms. The OCI-R is an 18-item questionnaire and requires participants to rate on a scale from zero (not at all) to four (extremely) the degree to which they were distressed or bothered by OCD symptoms in the past month. These symptoms can be divided into six subscales, each containing three items of the OCD dimensions: washing, checking, obsessions, mental neutralizing, ordering, and hoarding. To assess the severity of depression, the Montgomery-Asberg Depression Rating Scale (MADRS; Montgomery & Asberg, 1979) was used. The MADRS is a clinical questionnaire that consists of ten items. Each item refers to a common symptom of depression and is rated according to its severity during the past week on a scale from zero to six. The total score ranges from zero to 60, with higher scores referring to more severe forms of depression. In addition, the Beck Depression Inventory (BDI; Beck, Ward, Mendelson, Mock, & Erbaugh, 1961), a self-report questionnaire was used to determine the severity of a current depression. It consists of 21 questions that cover various depressive symptoms. Participants rate each item on a scale from zero (symptom is absent) to three (symptom is very severe). Accordingly, the total score ranges from zero to 63, with a total score higher than ten indicating a mild mood disturbance, and a score higher than 16 can be interpreted as borderline clinical depression. A total score of more than 20 indicates moderate depression, a score more than 30 can be regarded as severe depression, and if subjects have a total score higher than 40 an extreme depression is apparent (Beck, et al., 1961). Furthermore, the Yale Global Tic Severity Scale (YGTSS; Leckman, et al., 1989), a semi-structured interview that helps to assess motor and phonic tics over the previous week was administered. The rationale for including the YGTSS was to control for the frequent comorbidity of OCD with tic disorders (Brakoulias, et al., in press). The YGTSS consists of two elements: At first, the clinician notes the presence of motor and/or phonic

75 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r 63 tics on the basis of the subject s self-report and behavioural observations. Then, motor and/or phonic tics are rated according to their number, frequency, intensity, complexity, and interference. These ratings can be summed up and form the total motor tic score or total phonic tic score, respectively. In addition, the experienced distress and impairment that is caused by the tics can be assessed on a separate scale. Finally, all scores can be summed up together to form the overall YGTSS score. Higher scores indicate more severe forms of tics. Finally, the State-Trait Anxiety Inventory (STAI; Spielberger, 1983) was administered to examine state and trait anxiety in the participants. This self-rating questionnaire consists of two subscales, each containing 20 identical statements about emotions and feelings. Subjects are asked to indicate how strongly each item refers to their feelings right now (state anxiety) and generally (trait anxiety) on a scale from one to four, with one meaning not at all and four meaning very much so. The final sum scores range from 20 to 80 on each scale, with higher scores indicating more trait or state anxiety, respectively. The Compulsive Personality Assessment Scale (CPAS; Fineberg, Sharma, Sivakumaran, Sahakian, & Chamberlain, 2007) was used to measure the severity of obsessive - compulsive personality traits. The CPAS was assessed during a semi-structured interview. It includes the eight diagnostic DSM-IV-TR criteria of obsessive compulsive personality disorder which can be rated according to its severity from zero (absent) to four (very severe). The total score ranges from zero to 32, with higher scores indicating the presence of more OCPD symptoms Personality Measures The computerized version of the Cognitive Emotion Regulation Questionnaire (CERQ; Garnefski, Kraaij, & Spinhoven, 2001) was used to indentify cognitive coping strategies of participants after experiencing a negative event. The CERQ contains 36 items and participants are asked to rate on a scale from one (almost never) to five (almost always) how often they cope with a negative event in a given way. In total, nine different cognitive coping strategies can be identified: (a) blaming yourself, (b) accepting, (c) ruminating, (d) concentrating on other positive aspects, (e) putting into perspective, (f) concentrating on planning or thinking what steps must be taken to cope with the event, (g) positive reinterpretation, (h) catastrophising, (i) blaming others. Each of these nine coping strategies is represented by four items of the CERQ. Accordingly, scores on the different

76 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r 64 subscales vary from four to twenty. The higher the score, the more evident is the particular coping style for the participant. The 30-item version of the Metacognitions Questionnaire (MCQ-30; Wells & Cartwright-Hatton, 2004) was used to assess metacognitive traits. Metacognition refers to psychological structures, knowledge, events and processes that are involved in the control, modification and interpretation of thinking itself (Wells & Cartwright-Hatton, 2004, p. 386). Metacognitive traits might form important factor in the development and maintenance of mental disorders like OCD (Janeck, Calamari, Riemann, & Heffelfinger, 2003). The questionnaire contains 30 statements that can be rated from one (do not agree) to four (agree very much). Each item belongs to one of the five following factors: (a) cognitive confidence; (b) positive beliefs; (c) cognitive self-consciousness; (d) uncontrollability and danger; and (e) need to control thoughts. All of these metacognitive factors are covered by six items of the MCQ-30. The average score on each subscale varies between six and 24, with higher scores indicating the more presence of a certain cognitive coping style Neuropsychological Measures The Cambridge Neuropsychological Test Automated Battery (CANTAB) has been developed by Professor Trevor W. Robbins and Professor Barbara J. Sahakian more than 20 years ago at the University of Cambridge. The computer-based test battery enables to assess various cognitive functions with the aid of a touch-sensitive computer monitor. In 1994, CANTAB commercialized and has since then been further developed and marketed by Cambridge Cognition Limited. It has become the world leading computerized cognitive testing battery in its field and has been quoted in more than 600 scientific papers (Cambridge Cognition Limited, 2010) Attentional Set-Shifting The IED (Roberts, Robbins, & Everitt, 1988), a subtest of the CANTAB, was applied to examines different components of cognitive flexibility. The task is a computerized adaptation of the WCST (Berg, 1948), but is less complex and accordingly enables to study the formation, maintenance and shifting of cognitive set in a purer form (Lawrence, Sahakian, & Robbins, 1998, p. 384).

77 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r 65 The IED is made up of nine consecutive stages, with each stage consisting of at least six trials. On each trial, two visual stimuli are presented simultaneously on the computer screen and subjects are asked to touch (i.e. select) the one that subsequently leads to positive visual and auditory feedback. In the beginning of the task, the participants can only figure out which of the two stimuli is the relevant one by trial and error. However, once they learned which stimulus to select, the subjects are instructed to keep on touching the correct stimulus until they realise that the rule has changed, in which case they need to start choosing the newly relevant stimulus. During the task, the two visual stimuli randomly vary between four possible locations on the computer screen. In total, the IED consists of nine blocks, with each block consisting of six consecutive correct responses. After a block has been successfully completed, the next stage with different stimuli and/or a different rule starts. If subjects fail to learn a rule at any stage of the task within 50 trials, the whole task terminates. During the first stage of the IED, subjects are presented two visual stimuli of the same dimension (pink shapes), with one of them being reinforced by positive computer feedback. This block requires simple discrimination learning. After successfully completing this stage, the same two stimuli are presented during the second block of the task. However, the rule is now reversed so that the previously irrelevant stimulus becomes relevant. Therefore, stage two is a good indicator of reversal learning, the ability to switch responding to a previously non-reinforced stimulus (Kehagia, et al., 2010). During the third and fourth stage of the task, the rule remains unchanged. However, another stimulus dimension is introduced: in stage three, white stripes appear adjacent to the pink colourfilled shapes; and in stage four, the same white stripes are randomly superimposed on the relevant dimension (shapes). After successfully completing these stages, the fifth stage begins that contains identical stimuli as in the fourth block, however the rule is reversed so that the previously irrelevant stimulus becomes relevant. Thus, this stage is another indicator of reversal learning. In the sixth stage of the task, novel exemplars of stimuli for both dimensions are introduced and subjects are required to shift attention to the novel exemplar of the previously relevant perceptual dimension of shapes. This block of the IED is a good measure of intradimensional set-shifting, the ability to switch attention from exemplars of one stimulus dimension to novel exemplar of the same perceptual dimension. During stage seven, the rule is reversed so that the previously not reinforced exemplar of the same dimension becomes relevant. Thus, this stage requires reversal learning. The

78 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r 66 eighth stage introduces novel exemplars of stimuli for both dimensions and subjects are required to shift attention to the novel exemplar of the previously unrewarded perceptual dimension of lines. Thus, this stage assesses extradimensional set-shifting, the ability to shift attention between different perceptual dimensions. Finally, the rule is reversed in stage nine so that the previously irrelevant stimulus of the same perceptual dimension is now reinforced. Accordingly, this block is another indicator of reversal learning. In summary, performance on stage one is a good indicator of simple discrimination learning; stages two, five, seven, and nine are a good measure of reversal learning; stage six is an indicator of intradimensional set-shifting; and stage eight is a measure of extradimensional set-shifting (see Figure 2). The following measures were derived from IED: Errors during block one: an indicator of simple discrimination learning; Errors during blocks two, five, and seven: a measure of reversal learning; Errors during block six: an indicator of intradimensional set-shifting ability; Errors during block eight: an indicator of extradimensional set-shifting ability; Total trials adjusted: the total number of trials on all stages during the task, with an adjustment for blocks that were not reached due to failure at an earlier stage. The adjustment adds 50 trials for each non-completed stage. This measure is a good indicator of overall performance.

79 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r 67 Figure 2. The Intradimensional/Extradimensional Set-Shifting Task. Blocks one and two require simple discrimination learning and reversal learning, respectively. In block three and four, the second dimension is introduced but not relevant yet. During stage five, the rule is reversed. Novel exemplars of stimuli are introduced in block six and require an intradimensional shift, which is followed by a reversal of the rule in block seven. In block eight, novel stimulus exemplars are introduced and an extradimensional shift is needed to successfully complete the task, followed by a reversal of the rule in block nine. The green arrows indicate the correct stimulus during each stage of the task Spatial Planning To assess spatial planning and spatial working memory, the SOC (Owen, et al., 1990), a computerized touchscreen adaptation of the ToL (Shallice, 1982) was administered. The

80 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r 68 task can be divided into four blocks: block one and three measure planning abilities, while blocks two and four assess motor control. During the planning stages, three coloured balls are presented on the top, and three identically balls on the bottom of the computer screen. The balls are arranged in stacks and can easily be perceived as being held in stockings (see Figure 3). The subjects are instructed to think ahead and then move the balls in the lower half of the screen in the minimum amount of moves possible in order to match the target pattern presented in the upper half of the monitor. Before the task starts, the participants are taught how to move balls using the touch sensitive computer screen, and receive instructions regarding the rules of moving balls. The number of moves required to complete each trial increases as the task progresses, thereby increasing the task s difficulty. The difficulty of the task spans from one-move problems (only one ball needs to be moved to match the target pattern) gradually to five-move problems (five balls needs to be moved to match the target pattern). Each trial terminates automatically if a subject makes more than double the amount of moves necessary to complete the task. Moreover, the entire task ends if three trials are terminated automatically in a row. The SOC has no time limit. The subject s planning ability is measured by the time taken to complete the patterns and the number of moves required to do so. After each planning block, a block controlling for motor performance is inserted. During these trials, one ball at a time moves on the upper display and the subject has to repeat the moves in the corresponding lower display, thereby repeating the same n-move problems previously solved during the planning block. The motor control condition enables to measure the simple motor speed, i.e. the time a subject needs to move the balls in the given way. By subtracting this simple motor speed from the time it takes a participant to complete a trial during the planning stage, it is possible to estimate the planning and thinking time independently form the actual movement time.

81 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r 69 Figure 3. The Stocking of Cambridge Task. During the planning stage, subjects are required to plan ahead and move balls on the bottom half of the screen to match the target pattern of balls presented on the upper half of the screen. The number on the right hand side of the monitor indicates the minimum amount of moves needed to solve a problem. During the motor control stage of the task, subjects have to imitate movements on the bottom half of the screen that are presented on the upper half of the monitor. The following measures were obtained from the SOC: Problems solved in minimum moves: the number of problems that were solved in the minimum amount of moves possible during the task. This measure indicates overall planning performance; Mean moves for 2-move and 3-move problems: the mean number of moves needed to solve 2-move and 3-move problems. This measure indicates planning performance on easy tasks; Mean moves for 4-move and 5-move problems: the mean number of moves needed to solve 4-move and 5-move problems. This measure indicates planning performance on difficult tasks; Mean initial thinking time for 2-move and 3-move problems: the mean difference in time to select the first ball for the same 2-move or 3-move problem between the planning and the motor control condition. This measure indicates the time taken to plan the moves during easy problems; Mean initial thinking time for 4-move and 5-move problems: the mean difference in time to select the first ball for the same 4-move and 5-move

82 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r 70 problem between the planning and motor control condition. This measure indicates the time take to plan the moves during difficult problems; Mean subsequent thinking time for 2-move and 3-move problems: the mean difference in time between selecting the first ball and completing the 2-move or 3-move problem for the same problem under the planning and motor response condition, divided by the number of moves made. This measure provides an estimate of the subject s speed of movement after the initial move has been made for 2-move and 3-move problems; Mean subsequent thinking time for 4-move and 5-move problems: the mean difference in time between selecting the first ball and completing the 4-move or 5-move problem for the same problem under the planning and motor response condition, divided by the number of moves made. This measure provides an estimate of the subject s speed of movement after the initial move has been made for 4-move and 5-move problems Response Inhibition The SST of the CANTAB was applied to measure response inhibition, i.e. the ability to internally suppress prepotent motor responses. On each trial of the task, subjects are presented with a white circle appearing in the centre of the computer screen. Some milliseconds later, an arrow (stimulus) appears inside of the circle, pointing randomly to the left or right (see Figure 4). The subjects are instructed to respond as quickly as possible to the stimulus by pressing the left or right key of a button-box to indicate the direction in which the arrow is pointing. Thus, subjects press the left key as quickly as possible if a left-pointing arrow is presented and vice versa. The task is made up of two parts: In the first part, participants are introduced to the task by completing 16 trials as described above. The second part of the task consists of five consecutive blocks of 64 trials each. In 25% of the trials, an auditory signal (stop signal) appears shortly after the onset of the arrow and the subjects are instructed to withhold their response when they hear it. The delay between the onset of the stimulus and the stop signal (stop signal delay) varies during the task. It is therefore possible to estimate the average time it takes an individual to inhibit a prepotent motor response (stop signal reaction time) with a tracking algorithm.

83 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r 71 Figure 4. The Stop Signal Task. The following measures were derived from the SST: Direction errors on go trials: the sum of direction errors during all go trials of the task (i.e. when the subject pressed the incorrect button). This is an indicator of impulsivity; Reaction time on go trials: the median of the reaction time in milliseconds on go trials. This is a measure of motor speed; Stop signal delay: the arithmetic mean of the stop signal delay in milliseconds at which the subject was able to successfully stop 50% of the time. The measure is calculated for the last half of the assed blocks in the task. It is necessary to estimate the stop signal reaction time. Stop signal reaction time: an estimate of the length of time between the go stimulus and the stop stimulus at which the subject is successfully inhibiting a response on 50% of the trials. This measure is calculated by the reaction time on go trials and the stop signal delay. It is an indicator of response inhibition performance. 4.6 Statistical Analysis The neuropsychological performance as well as sociodemographic, diagnostic, clinical, and personality measures were statistically compared between the control group, the OCD hoarding group, and the non-ocd hoarding group, using the Statistical Package for the Social Sciences (SPSS) version 17. Several variables were log10 or square-root transformed to reduce the skewness of the data (see Table A2 in Appendix). For dependent measures with interval scale, the normality of the sample distributions and the homogeneity of their variances were tested with Shapiro-Wilk test and Levene s test, respectively (see Tables A3 and A4 in Appendix). If the statistical assumptions of normality and homoscedasticity were not violated (Levene Statistic and Shapiro Wilk

84 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r 72 Statistic p 0.05), one-way analyses of variance (ANOVA) were conducted with group as a between-subject factor and the specified measure as a within-subject factor. Subsequent post hoc Tukey honestly significant difference tests (HSD) were conducted in case of significant (p 0.05) differences between the three groups were observed. When the statistical assumptions of normality and homoscedasticity were not met, the non-parametric Kruskal-Wallis H test was used, with subsequent Mann-Whitney U comparisons in case of significant results (p 0.05) that were not corrected for multiple testing. To analyse the distribution of categorical variables such as gender, handedness, and number of native English speakers across the three groups, chi square (χ 2 ) tests were conducted.

85 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r 73 5 Results 5.1 Sociodemographic Characteristics In total, 60 subjects were included in the final sample. The OCD hoarding group, the non- OCD hoarding group, and the control group consisted of 20 participants each. All subjects were between 29 and 69 years old, with each group having a mean age of 51 to 53 years (see Table 3). The three groups did not differ with respect to age, gender, handedness, education, number of native English speakers, and estimated premorbid verbal intelligence quotient using the NART (Nelson, 1982). Table 3 Sociodemographic Characteristics Controls (n=20) H+OCD (n=20) H only (n=20) Variable Mean/n (SD) Mean/n (SD) Mean/n (SD) Statistic p Age (years) Gender (male:female) Handedness (left:right) Native speaker (n) Education (years) (10.85) (10.79) (8.38) H= :14 7:13 4:16 χ 2 = :18 4:16 2:18 χ 2 = χ 2 = (2.41) (2.25) (2.16) H= Verbal IQ (3.75) (3.58) (3.69) F= Education was assessed in years of school and vocational education. The verbal intelligence quotient (IQ) was estimated with the National Adult Reading Test (NART). There were no significant differences for all the variables listed between the groups. Controls = control group; H+OCD = OCD hoarding group; H only = non-ocd hoarding group. Nearly half of the subjects in the OCD hoarding group received some form of pharmacological treatment for OCD or a comorbid mental disorder (see Table 4). The most commonly subscribed drugs were SSRIs, including fluoxetine, paroxetine, citalopram, escitalopram, and sertraline. In total, 25% of the OCD hoarding group took

86 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r 74 SSRIs as monotherapy, while two patients (10%) received it in combination with an atypical antipsychotic, and one person took it in combination to a benzodiazepine. Moreover, another participant was pharmacologically treated with a combination therapy that included an atypical antipsychotic, a trizyclic antidepressant, as well as a selective noradrenaline reuptake inhibitor. In contrast, one fifth of the non-ocd hoarding group received some form of pharmacotherapy. Of these, all but one participant (15%) were on SSRI monotherapy, while one subject took a norepinephrine dopamine disinhibitor. No statistical comparisons between the two patient groups with respect to pharmacological treatment were conducted as most participants received different types of (SRI) medication and/or different doses (see Table A1 in Appendix). Table 4 Medication Status H+OCD H only Current medication n (%) n (%) SSRI monotherapy 5 (25%) 3 (15%) SSRI and atypical antipsychotic 2 (10%) - SSRI and benzodiazepine 1 (5%) - Atypical antipsychotic, trizyclycic antidepressant, and SNRI 1 (5%) - NDDI - 1 (5%) Total 9 (45%) 4 (20%) In total, 45% of the subjects in the OCD hoarding group (H+OCD) received some form of pharmacological treatment, while only 20% of the participants in the non-ocd hoarding group (H only) were medicated. The majority of the individuals took one type of a selective serotonin reuptake inhibitor (SSRI), either as monotherapy, or as part of a combination therapy. SNRI = selective noradrenaline reuptake inhibitor; NDDI = norepinephrine dopamine disinhibitor. 5.2 Diagnostic Characteristics All hoarding participants met the previously described criteria for compulsive hoarding and scored higher than 35 on the SI-R (Frost, et al., 2004). The control group had a total mean score of 11, while both hoarding groups had on average a score of 60 to 61.

87 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r 75 As expected, there was a significant main effect of groups for the SI-R total score (F 2,57, p<.001). Post hoc Tukey HSD comparisons revealed that both hoarding groups differed significantly from the control group (p<.001), while there were no differences in the severity of compulsive hoarding between the OCD hoarding group and the non-ocd hoarding group (p.90). Moreover, significant overall group differences were observed when comparing the severity of the hoarding symptoms clutter (F 2,57 =151.42, p<.001), difficulty discarding/saving (Kruskal-Wallis H=39.06, df=2, p<.001), and acquisition (F 2,57 =47.80, p<.001) separately. Subsequent tests demonstrated that the control group had significantly lower scores on all three symptom subscales than both hoarding groups (p<.001). As expected, there were no significant differences between the OCD hoarding group and the non-ocd hoarding group. Thus, the severity of hoarding symptoms did not differ between both hoarding groups. Results of the SI-R are summarised in Table 5. Table 5 Severity of Hoarding Symptoms Controls H+OCD H only Measure Mean (SD) Mean (SD) Mean (SD) Statistic p Total score (6.82) (11.81) (14.44) F= <.001 Clutter 3.15 (2.48) (5.94) (7.10) F= <.001 Difficulty discarding/saving 4.30 (3.94) (3.59) (4.62) H=39.06 <.001 Acquisition 3.85 (2.43) (5.46) (5.14) F=47.80 <.001 The severity of hoarding symptoms was assessed with the Saving Inventory Revised (SI-R). Controls = control group; H+OCD = OCD hoarding group; H only = non-ocd hoarding group. The majority of hoarding participants met DSM-IV-TR criteria for a current Axis I disorder other than OCD as assessed with the MINI (Sheehan, et al., 1998). Half of the individuals of the OCD hoarding group, and 30% of the non-ocd hoarding group did not suffer from a comorbid disorder (see Table 6). The most common comorbidity for both hoarding groups was major depressive disorder which occurred in nearly half of the participants. Moreover, anxiety disorders were common in both groups. In the OCD hoarding group, three individuals (15%) were diagnosed with generalized anxiety disorder, and one participant suffered from social

88 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r 76 phobia (5%). The prevalence of anxiety disorders was greater in the non-ocd hoarding group: four subjects (20%) were diagnosed with generalized anxiety disorder, while social phobia, agoraphobia, and panic disorder occurred each in 5% of the sample. Table 6 Current Comorbid Axis I Disorders H+OCD H only Current Axis I Diagnosis n (%) n (%) Major depressive disorder 9 (45%) 10 (50%) Dysthymia - 1 (5%) Social phobia 1 (5%) 1 (5%) Agoraphobia - 1 (5%) Generalized anxiety disorder 3 (15%) 4 (20%) Panic disorder - 1 (5%) Posttraumatic stress disorder - 1 (5%) Anorexia nervosa - 1 (5%) No comorbid disorders 10 (50%) 6 (30%) Current DSM-IV-TR Axis I diagnoses were assessed with the MINI International Neuropsychiatric Interview (MINI). Only 50% of the OCD hoarding group (H+OCD), and 30% of the non-ocd hoarding group (H only) did not suffer from a current Axis I disorder other than OCD. 5.3 Clinical Characteristics To assess the severity and types of OCD symptoms, the Y-BOCS (Goodman, et al., 1989) and the OCI-R (Foa, et al., 2002) were administered. In the OCD hoarding group, the mean total Y-BOCS score, reflecting the severity of obsessions and compulsions was 19.5, indicating moderate OCD (see Table 7). In contrast, only one subject of the non-ocd hoarding group reported mild OCD symptoms that did not meet criteria for a diagnosis of OCD (total Y-BOCS score = 4), and none of the control subjects reported any OCD symptom. Accordingly, no statistical analysis to examine differences between the groups was conducted. Moreover, the severity of OCD symptoms was assessed with the OCI-R (Foa, et al., 2002). Significant group differences were found on the overall symptom scale (Kruskal- Wallis H=44.09, df=2, p<.001). Mann-Whitney U tests demonstrated that the control

89 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r 77 group had significantly lower levels of symptoms than both hoarding groups (p<.001), and that the OCD-hoarding group had significantly higher levels of OCD symptoms than the non-ocd hoarding group (p<.001). Owing to the fact that hoarding symptoms are included in the sum score of the OCI-R, a separate analysis was conducted to examine between-group differences for the overall scale when excluding hoarding symptoms. The Kruskal-Wallis test revealed significant differences between the groups (Kruskal-Wallis H=34.31, df=2, p<.001), with Mann-Whitney U tests indicating that the OCD hoarding group had significantly higher scores than the control group and the non-ocd hoarding group (p<.001). However, the non-ocd hoarding group had significantly higher level of non-hoarding OCD symptoms than the control group (p<.001). Thus, the OCD hoarding group had the most severe OCD symptoms, followed by the non-ocd hoarding group, and the control group showing the mildest symptoms. Additionally, overall significant group differences were observed for the separate OCD symptom subscales of washing, obsessing, ordering, checking, and neutralizing (p<.01). Mann-Whitney U comparisons showed that the OCD hoarding group reported more severe symptoms of all these dimensions than both other groups (p<.02). There were no significant differences between the control group and the non-ocd hoarding group on the washing (p.20) and neutralizing (p.80) symptom dimensions, but the differences were significant on the obsessing, ordering, and checking (p<.01) scale. These findings indicate that the non-ocd hoarding group experienced more severe OCD symptoms when compared to the control group, but generally had significantly lower levels than the OCD hoarding group. Finally, the hoarding subscale of the OCI-R was compared between the groups. As expected, the control group reported significantly lower hoarding symptoms than both hoarding groups (p<.001). The OCD hoarding group and the non-ocd hoarding group did not differ with respect to the severity of compulsive hoarding (p.20). The MADRS (Montgomery & Asberg, 1979) and BDI (Beck, et al., 1961) were included in this study to further assess the severity of depressive symptoms. The total scores on both measures differed significantly between the groups (p<.001; see Table 8). Mann-Whitney U comparisons revealed that the control group scored significantly lower on the two depression scales than both hoarding groups (p<0.001). There were no differences between the hoarding groups with respect to the reported severity of depressive symptoms on the MADRS (p.20) or the BDI (p.60).

90 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r 78 Table 7 Severity and Types of Obsessive Compulsive Symptoms Controls H+OCD H only Measures Mean (SD) Mean (SD) Mean (SD) Y-BOCS Kruskal- Wallis H p Total score (5.15) 0.20 (0.89) - - Obsessions (3.17) 0.15 (0.671) - - Compulsions (2.78) 0.05 (0.22) - - OCI-R Total score 4.75 (3.93) (13.97) (7.43) <.001 Total score excl. hoarding 3.25 (3.52) (13.31) 9.55 (6.58) <.001 Washing 0.35 (0.93) 3.90 (3.50) 0.55 (0.76) <.001 Obsessing 0.70 (1.08) 5.55 (3.10) 2.80 (2.46) <.001 Ordering 1.25 (1.45) 6.30 (3.57) 3.55 (2.44) <.001 Checking 0.55 (0.89) 6.25 (3.34) 1.85 (1.81) <.001 Neutralizing 0.40 (0.75) 3.55 (4.03) 0.80 (1.70) <.001 Hoarding 1.50 (1.28) (2.32) 9.30 (2.68) <.001 The severity of obsessive - compulsive disorder symptoms was assessed with the Yale-Brown Obsessive Compulsive Scale (Y-BOCS) and the Obsessive Compulsive Inventory Revised (OCI- R). A statistical comparison between the control group (Controls), the OCD hoarding group (H+OCD), and the non-ocd hoarding group (H only) was only carried out for the OCI-R scales as only one participant in the non-ocd hoarding group and none of the control subjects reported OCD symptoms that were assessable with the Y-BOCS. The YGTSS (Leckman, et al., 1989) was administered to identify phonic and motor tics, respectively. However, only three individuals in the OCD hoarding group had mild motor tics, with only one of them also having mild phonic tics. Therefore, a comparison of severity of tics between the groups was not carried out (see Table 8). The levels of state and trait anxiety were assessed with the STAI (Spielberger, 1983). Owing to the fact that several individuals of both hoarding groups were diagnosed with an anxiety disorder, higher state and trait anxiety ratings in comparison to the control group were expected. There was a significant main effect of groups for both anxiety measures

91 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r 79 (p<0.001; see Table 8). Subsequent analyses showed that the control group had significantly lower anxiety levels on the state as well as the trait measures when compared to both hoarding groups (p<.001). There was no difference between the OCD hoarding group and the non-ocd hoarding group with respect to the state anxiety levels (p.90) and trait anxiety levels (p.10). Finally, OCPD characteristics were assessed with the CPAS (Fineberg, et al., 2007). A significant main effect of group was observed (Kruskal Wallis H=37.06, df=2, p<.001; see Table 8). Mann Whitney U tests demonstrated that both hoarding groups had significantly more severe OCPD symptoms than the control group (p<.001), with no difference between both hoarding groups (p.40). Table 8 Severity of Clinical Symptoms Controls H+OCD H only Measures Mean (SD) Mean (SD) Mean (SD) MADRS (Total score) BDI (Total score) YGTSS (Total score) STAI State (Total score) STAI Trait (Total score) CPAS (Total score) Kruskal- Wallis H p 1.25 (1.74) (8.93) 9.05 (6.72) < (2.50) (10.19) (10.11) < (9.75) (4.86) (11.41) (11.49) < (5.41) (8.75) (12.82) < (2.08) (5.16) (6.16) <.001 MADRS = Montgomery-Asberg Depression Rating Scale; BDI = Beck Depression Inventory; YGTSS = Yale Global Tic Severity Scale; STAI = State-Trait Anxiety Inventory (STAI); CPAS = Compulsive Personality Assessment Scale; Controls = control group; H+OCD = OCD hoarding group; H only = non-ocd hoarding group. 5.4 Personality Characteristics To examine metacognitive traits, the MCQ-30 (Wells & Cartwright-Hatton, 2004) was administered. There were significant main effects of group for the metacognitive traits

92 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r 80 positive beliefs (Kruskal-Wallis H=8.03, df=2, p<.05), cognitive self-consciousness (Kruskal-Wallis H=19.77, df=2, p<.001), uncontrollability (Kruskal-Wallis H=27.19, df=2, p<.001), and need to control thoughts (Kruskal Wallis H=16.30, df=2, p<.001), but not for cognitive confidence (Kruskal-Wallis H=4.08, df=2, p.10). Following analyses for the first four of these traits showed that the control group reported significantly lower levels of all these traits in comparison to the OCD hoarding group (p<.05), and lower levels of uncontrollability (p<.001) and cognitive self-consciousness (p<.05) then the non-ocd hoarding group. Both hoarding groups only differed with respect to the metacognitive trait need to control thoughts (p<.05) as the OCD hoarding individuals had on average higher levels than the non-ocd hoarding participants on this subscale. Moreover, the CERQ (Garnefski, et al., 2001) was used to identify differences between the groups in their coping strategies with negative events. Overall significant group differences were observed for seven coping strategies (p<.05), except for accepting and blaming others. Post hoc Tukey HSD comparisons revealed that individuals from the control group as well as the non-ocd hoarding group blamed themselves to a lesser extent for the occurrence of negative events than the OCD hoarding group (p<.01). There was no difference between the non-ocd hoarding group and the control group (p.30). Moreover, the control group significantly reinterpreted events more positively and catastrophized them to a lesser extent than both hoarding groups (p<.05). There were no differences for these two cognitive strategies between both hoarding groups. Finally, the control group scored significantly higher on the strategies putting in perspective, planning what to do next, and concentrating on positive aspects than the non-ocd hoarding group (p<.01). There were neither significant differences of these traits between the control group and the OCD hoarding group, nor both hoarding groups. Thus, both hoarding groups appear to cope with negative events differently than the control group, focussing to a lesser extent on positive aspects of such events, reinterpretating to a lesser extent the events positively, and plan less intensively to find a solution.

93 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r 81 Table 9 Personality Measures Control H+OCD H only Measures Mean (SD) Mean (SD) Mean (SD) Statistic p MCQ Cognitive confidence 9.90 (4.14) (4.76) (4.84) H= Positive beliefs 8.15 (2.41) (5.53) 9.95 (4.37) H= Cognitive selfconsciousness 9.70 (3.08) (3.87) (5.74) H=19.77 <.001 Uncontrollability 7.55 (2.63) (4.20) (5.54) H=27.19 <.001 Need to control thoughts CERQ 8.00 (1.81) (4.58) (3.65) H=16.30 <.001 Blaming oneself 7.60 (1.50) (3.57) 8.80 (2.55) F=14.05 <.001 Accepting (2.86) (2.70) (2.71) F= Ruminating 7.35 (2.70) (3.98) (4.44) F=19.16 <.001 Concentrating on pos. aspects Putting into perspective Planning what to do next Positive reinterpretation (4.30) 8.15 (2.08) 7.95 (3.46) F= (3.58) (3.45) (3.08) F= (3.94) (3.06) (4.20) F= (3.20) (3.62) (3.87) F=8.76 <.001 Catastrophising 5.65 (1.60) (3.50) 9.55 (3.87) F=17.63 <.001 Blaming others 7.75 (1.16) 8.50 (2.16) 8.75 (3.48) F= Metacognitive traits were examined with the 30-item version of the Meta-Cognitive Questionnaire (MCQ-30), and coping strategies for negative events were assessed with the Cognitive Emotion Regulation Questionnaire (CERQ). Controls = control group; H+OCD = OCD hoarding group; H only = non-ocd hoarding group.

94 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r Neuropsychological Performance Attentional Set-Shifting Attentional set-shifting performance was assessed with the IED (Downes, et al., 1989). There were no significant differences between the three groups with respect to discrimination learning, reversal learning, or intradimensional set-shifting (see Table 10). Contrarily to the hypothesis that the OCD hoarding group would be impaired on extradimensional set-shifting performance in comparison to the control and non-ocd hoarding group, there were no significant differences between the groups on this IED outcome measure. As expected however, there was a significant group difference for the overall task performance, measured as the mean of the total (adjusted) trials subjects needed to complete all stages of the task (Kruskal Wallis H=7.61, df=2, p<.05). Nonparameter Mann-Whitney U tests revealed that the OCD hoarding group performed significantly worse than the control group (p<0.01). In line with the hypothesis that the non-ocd hoarding group would be unimpaired in attentional set-shifting performance, there were not significant differences between the control group and the non-ocd hoarding group. However, contrarily to this hypothesis, no significant differences between the two patient groups emerged. Accordingly, it can be concluded that the OCD hoarding group was significantly impaired on the overall IED task performance, but not specifically on extradimensional set-shifting. Although the non-ocd hoarding group did not differ significantly on all performance measures from the control group, indicating unimpaired performance on the task, this group also did not differ significantly from the OCD hoarding group. These results can be explained when examining the performance measures descriptively as the non-ocd hoarding group generally performed on average (non-significantly) better on the attentional set-shifting measures than the OCD hoarding group, but worse than the control group.

95 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r 83 Table 10 Attentional Set-Shifting Performance Controls H+OCD H only Measure Mean (SD) Mean (SD) Mean (SD) Discrimination learning (errors) Reversal learning (errors) Intradimensional set-shifting (errors) Extradimensional set-shifting (errors) Overall perform. (total trials adjusted) Shifting Task (IED). Controls = control group; H+OCD = OCD hoarding group; H only = non-ocd hoarding group; Overall perform. = overall performance. Attentional set-shifting performance was assessed with the Intradimensional/Extradimensional Set- Kruskal- Wallis H p 0.75 (1.41) 0.65 (0.49) 0.60 (0.60) (0.67) 3.75 (3.40) 3.50 (0.95) (0.49) 0.40 (0.50) 0.30 (0.92) (10.31) (12.01) (9.91) (34.39) (68.47) (36.24) Spatial Planning To examine spatial planning abilities between the groups, the SOC (Owen, et al., 1990) was administered. For the overall planning performance, i.e. the problems that were solved in the minimum amount of moves possible, a one-way ANOVA was conducted (see Table 11). The results indicated significant overall differences between the groups (F 2,57 =3.488, p<.05). Post hoc Tukey HSD comparisons revealed that the control group solved significantly more problems within the minimum amount of moves than the OCD hoarding group (p<.05). Moreover, both hoarding groups did not differ in their planning abilities (p.30). These findings are in line with the hypothesis that hoarding is characterised by planning deficits that distinguish this condition from the neuropsychological deficits commonly observed in non-hoarding OCD patients. However, there were no significant differences between the control group and the non-ocd hoarding group (p.40). To examine the performance of the groups on easy problems that require two or three moves, and hard problems that require four or five moves to meet the target pattern successfully, the number of problems solved in the minimum amount of moves possible for easy and hard problems were compared between the groups. There were no

96 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r 84 significant differences between the groups for the number of easy problems that were solved in the minimum amount of moves possible (Kruskal Wallis H=0.34, df=2, p.80). However, there was a significant group difference for the number of hard problems solved in the minimum amount of moves (F 2,57 =3.19, p<.05). Post hoc Tukey HSD comparisons revealed that the OCD hoarding group solved significantly less hard problems in the minimum amount of moves than the control group (p<0.05) and that the two hoarding groups did not differ from each other in performance (p.40) which is in line with assumption of hoarding being generally associated with planning deficits. Contrarily to this hypothesis however, there were no significant differences between the control group and the non-ocd hoarding group (p.40). Moreover, the mean initial thinking time and subsequent thinking time for easy and hard problems were compared between the groups. The analyses revealed no differences for the initial and subsequent thinking time on hard problems, and no difference for the subsequent thinking time on easy problems. However, there was a significant overall effect of group for the initial thinking time on easy problems (F 2,57 =5.21, p<0.01). Post hoc Tukey HSD tests showed that the OCD hoarding group needed significantly more time to initially plan the movements on easy trials than the control group (p<0.01). Both hoarding groups did not differ significantly from each other on the initial thinking time on easy problems. These results are in line with the proposed hypothesis. However, no significant differences were found on this measure between the non-ocd hoarding group and the control group. In conclusion, the OCD hoarding group was significantly impaired on several measures of spatial planning ability as compared to the control group. In particular, this patient group performed poor on the overall task (problems solved in minimum moves), especially on trials that required to think several steps ahead, and needed significantly more time to initially plan relatively few sequences of actions to reach the target pattern. Moreover, there were no significant differences in planning performance between both hoarding groups. These findings are in line with the hypothesis of hoarding being generally associated with impaired planning abilities. Contrarily to this hypothesis however, the non- OCD hoarding group did not differ from the control group on these SOC outcome measures. These results are due to the fact that the non-ocd hoarding group performed on average worse on most planning measures than the control group, but better than the OCD hoarding group. These differences were not statistically significant however.

97 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r 85 Table 11 Spatial Planning Performance Controls H+OCD H only Variable Mean (SD) Mean (SD) Mean (SD) Statistic p Problems solved in minimum moves Easy problems solved in minimum moves Hard problems solved in minimum moves Mean initial thinking time easy problems (s) Mean initial thinking time hard problems (s) Mean subsequent thinking time easy problems (s) Mean subsequent thinking time hard problems (s) 9.40 (1.67) 7.65 (2.35) 8.55 (2.21) F= (0.41) 3.74 (0.45) 3.75 (0.55) H= (1.60) 4.05 (2.07) 4.80 (2.04) F= (4.80) (7.66) 8.00 (6.94) F= (9.28) (9.62) (20.83) H= (1.33) 0.72 (1.22) 1.57 (4.92) F= (2.29) 1.48 (1.47) 3.93 (5.08) H= Spatial planning was assessed with various performance measures of the Stocking of Cambridge Task (SOC). Controls = control group; H+OCD = OCD hoarding group; H only = non-ocd hoarding group, s = seconds Response Inhibition To investigate response inhibition performance, the SST (Aron, et al., 2003) was used. Between-group analyses yielded no significant differences for the median reaction time on go trials, direction errors on go trials, or the stop signal delay (see Table 12). However, there was a significant main effect of group for the stop signal reaction time (F 2,57 =3.32, p<.05). As expected, post hoc Tukey HSD comparisons showed that the OCD hoarding group, but not the non-ocd hoarding group had a significant longer stop signal reaction time than the control group (p<.05). However, there were no significant differences for stop signal reaction time between both hoarding groups (p.80).

98 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r 86 Table 12 Response Inhibition Performance Controls H+OCD H only Variable Mean (SD) Mean (SD) Mean (SD) Statistic p Reaction time on go trials (Median, ms) Direction errors on go trials (n) (55.77) (46.66) (82.77) F= (3.77) 3.65 (3.28) 3.05 (2.67) H= Stop signal delay (ms) (58.28) (89.73) (133.35) H= Stop signal reaction time (ms) (40.39) (78.02) (78.23) F= Response inhibition ability was assessed with the Stop Signal Task (SST). Results yielded a significant difference between the control group (Controls), the OCD hoarding group (H+OCD), and the non-ocd hoarding group (H only) for the stop signal reaction time. n = number; ms = milliseconds. In summary, the OCD hoarding group was significantly impaired in response inhibition when compared to the control group, while the non-ocd hoarding group did not differ from the control group. These findings are in line with the hypothesis that only the hoarding OCD patients would suffer from typical neuropsychological deficits that have been previously linked to OCD. Contrasting to this assumption however, performance of the non-ocd hoarding group did not significantly differ from the OCD hoarding group. These results can be understood when examining the mean scores of the groups on the task measures descriptively (see Figure 5), as it took the non-ocd hoarding group on average a longer time to stop a prepotent motor response than the control group, but less time as compared to the OCD hoarding group. Again, these differences were not statistically significant.

99 o f O b s e s s i v e C o m p u l s i v e D i s o r d e r 87 Figure 5. Descriptive Comparison of Response Inhibition Performance. Controls = control group; H+OCD = OCD hoarding group; H only = non-ocd hoarding group; ms = milliseconds.