Categorization in compulsive hoarding

Abstract

Based on hypothesizing about the role of information processing, and in particular, underinclusive categorization in compulsive hoarding, this study examined categorization processes in people with clinically significant compulsive hoarding problems. Twenty-one participants with primary compulsive hoarding, 21 with OCD without hoarding, and 21 non-psychiatric controls completed three categorization tasks. Hoarding and OCD participants reported significantly more distress prior to each of the three tasks than did controls. On tasks sorting common household items, the groups did not differ on the number of piles created nor on the amount of time taken to sort. However, on a task sorting personally relevant items, hoarding participants took more time, created more piles, and reported more anxiety than non-psychiatric controls. Hoarders also took more time than the OCD group, and tended to create more piles. Hoarding severity was correlated with the number of piles created, but only when the objects were personally relevant. Results support under-inclusive categorizing for people with compulsive hoarding, but the effect was largely confined to objects of personal relevance.

Introduction

Compulsive hoarding, defined as the acquisition of and failure to discard possessions that appear to be useless or of limited value, is a complex and little studied disorder. Although hoarding occurs in a number of clinical disorders, including obsessive–compulsive personality disorder (American Psychological Association, 1994), anorexia nervosa (Frankenburg, 1984), and schizophrenia (Chong, Tan, & Lee, 1996), it is thought to be a syndrome associated with obsessive–compulsive disorder (OCD) (Frost & Gross, 1993; Frost, Krause, & Steketee, 1996). In fact, approximately 25% of cases presenting for treatment of OCD exhibit hoarding symptoms (Samuels et al., 2002; Sobin et al., 2000). Although the prevalence of this condition is unknown, a recent survey of hoarding complaints to Massachusetts health departments found 26.3 per 100,000 population over a period of 5 years (Frost, Steketee, & Williams, 2000). Over half of health departments sampled had received complaints about hoarding behavior within the past 2.5 years. Thus, hoarding appears to be a significant problem for many people.

Mataix-Cols, Rosario-Campos, and Leckman (2005) suggest that hoarding is a distinct dimension of OCD about which we know very little. The phenomenological model developed by Frost and Hartl (1996) and elaborated in Steketee and Frost (2003) suggests that hoarding is a multi-faceted disorder that stems from four primary deficits or difficulties: excessive emotional attachment to possessions, significant behavioral avoidance, erroneous beliefs about the nature and importance of possessions, and information processing deficits. The latter aspect includes problems with categorization, memory, and the use of information to draw conclusions and make decisions. Although some recent research has examined some aspects of information processing (see Steketee & Frost, 2003 for summary), as well as neuroimaging (Mataix-Cols et al., 2003), categorization and decision-making have been little studied. In studying memory problems, Hartl et al. (2004) observed that people with problematic hoarding used less effective organizational strategies on the Rey-Osterrieth Complex Figure Test, and, perhaps because of this, recalled less information on a delayed recall task. Decision-making problems have also been observed in hoarding samples (Frost & Gross, 1993; Frost & Shows, 1993; Steketee, Frost, & Kyrios, 2003).

Individuals who hoard appear to exhibit deficits in the ability to arrange and categorize information as evident from the large amount of disorganized clutter and difficulty finding desired possessions when needed, as well as decision-making problems. Such deficits have also been noted among non-hoarding OCD patients (i.e., Frost, Lahart, Dugas, & Sher, 1988; Persons & Foa, 1984; Reed (1969a), Reed (1969b), 1977, 1985). These studies found that individuals with OCD exhibit more “complex concepts”, considering many more details and requiring more information to make decisions, compared to non-OCD individuals. Reed (1969a), Reed (1969b) defined this as an “under-inclusive” style of thinking in which individuals with OCD created smaller sizes and larger numbers of categories than non-OCD individuals when sorting items. Reed (1985) suggested that one explanation for this is that people with OCD perceive a larger number of essential features of objects, leading to more categories. What may occur in those who hoard is that the larger number of details obscures their importance, making decisions about organizing possessions more difficult.

In their phenomenological model, Frost and Hartl (1996) hypothesized that information-processing deficits in decision-making, organization, and categorization contribute to compulsive hoarding behavior. The purpose of the present study was to directly test this hypothesis by examining the relationship between compulsive hoarding and under-inclusion in 3 groups of participants: those with principal symptoms of compulsive hoarding, patients with OCD but no hoarding symptoms, and non-psychiatric controls. Participants completed 3 categorization tasks using common household items while sorting time, number of categories and discomfort were measured. We hypothesized that, compared to OCD and non-psychiatric control groups, individuals with hoarding would (1) over-specify categories, allotting fewer items to each class, (2) take more time, reflecting difficulty determining category membership, and (3) experience more discomfort.